diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,40831 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5827,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017161489617298782,
+      "grad_norm": 30.668089447930665,
+      "learning_rate": 1.142857142857143e-07,
+      "loss": 4.5603,
+      "step": 1
+    },
+    {
+      "epoch": 0.00034322979234597563,
+      "grad_norm": 25.93890164506215,
+      "learning_rate": 2.285714285714286e-07,
+      "loss": 3.8759,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005148446885189634,
+      "grad_norm": 36.96142326314757,
+      "learning_rate": 3.428571428571429e-07,
+      "loss": 4.7173,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006864595846919513,
+      "grad_norm": 34.785801715634655,
+      "learning_rate": 4.571428571428572e-07,
+      "loss": 4.8091,
+      "step": 4
+    },
+    {
+      "epoch": 0.000858074480864939,
+      "grad_norm": 28.830290581224215,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 4.1964,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010296893770379268,
+      "grad_norm": 22.707707298395775,
+      "learning_rate": 6.857142857142858e-07,
+      "loss": 4.1035,
+      "step": 6
+    },
+    {
+      "epoch": 0.0012013042732109147,
+      "grad_norm": 16.854320496736147,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 3.5957,
+      "step": 7
+    },
+    {
+      "epoch": 0.0013729191693839025,
+      "grad_norm": 26.88002151772228,
+      "learning_rate": 9.142857142857144e-07,
+      "loss": 4.4156,
+      "step": 8
+    },
+    {
+      "epoch": 0.0015445340655568904,
+      "grad_norm": 30.606793141390057,
+      "learning_rate": 1.0285714285714286e-06,
+      "loss": 4.5317,
+      "step": 9
+    },
+    {
+      "epoch": 0.001716148961729878,
+      "grad_norm": 29.070794542104583,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 4.4011,
+      "step": 10
+    },
+    {
+      "epoch": 0.001887763857902866,
+      "grad_norm": 29.849593762767654,
+      "learning_rate": 1.2571428571428573e-06,
+      "loss": 4.5002,
+      "step": 11
+    },
+    {
+      "epoch": 0.0020593787540758536,
+      "grad_norm": 24.231850340704394,
+      "learning_rate": 1.3714285714285717e-06,
+      "loss": 4.1261,
+      "step": 12
+    },
+    {
+      "epoch": 0.0022309936502488414,
+      "grad_norm": 21.178239012309003,
+      "learning_rate": 1.4857142857142858e-06,
+      "loss": 4.0331,
+      "step": 13
+    },
+    {
+      "epoch": 0.0024026085464218293,
+      "grad_norm": 17.404912868999176,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 4.0249,
+      "step": 14
+    },
+    {
+      "epoch": 0.002574223442594817,
+      "grad_norm": 18.87587660804914,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 3.9221,
+      "step": 15
+    },
+    {
+      "epoch": 0.002745838338767805,
+      "grad_norm": 13.678354367731275,
+      "learning_rate": 1.8285714285714288e-06,
+      "loss": 3.5854,
+      "step": 16
+    },
+    {
+      "epoch": 0.002917453234940793,
+      "grad_norm": 17.11526727983393,
+      "learning_rate": 1.942857142857143e-06,
+      "loss": 3.6071,
+      "step": 17
+    },
+    {
+      "epoch": 0.003089068131113781,
+      "grad_norm": 23.592418185912088,
+      "learning_rate": 2.0571428571428573e-06,
+      "loss": 3.444,
+      "step": 18
+    },
+    {
+      "epoch": 0.0032606830272867687,
+      "grad_norm": 14.012203304419085,
+      "learning_rate": 2.1714285714285716e-06,
+      "loss": 3.4874,
+      "step": 19
+    },
+    {
+      "epoch": 0.003432297923459756,
+      "grad_norm": 15.458808624656582,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 3.5889,
+      "step": 20
+    },
+    {
+      "epoch": 0.003603912819632744,
+      "grad_norm": 11.584259212284152,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 3.3634,
+      "step": 21
+    },
+    {
+      "epoch": 0.003775527715805732,
+      "grad_norm": 14.249117085121204,
+      "learning_rate": 2.5142857142857147e-06,
+      "loss": 3.6045,
+      "step": 22
+    },
+    {
+      "epoch": 0.00394714261197872,
+      "grad_norm": 14.355055747286821,
+      "learning_rate": 2.6285714285714286e-06,
+      "loss": 3.4743,
+      "step": 23
+    },
+    {
+      "epoch": 0.004118757508151707,
+      "grad_norm": 15.986607989434727,
+      "learning_rate": 2.7428571428571433e-06,
+      "loss": 3.5071,
+      "step": 24
+    },
+    {
+      "epoch": 0.004290372404324695,
+      "grad_norm": 15.009222302006576,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 3.4385,
+      "step": 25
+    },
+    {
+      "epoch": 0.004461987300497683,
+      "grad_norm": 15.00054408070156,
+      "learning_rate": 2.9714285714285716e-06,
+      "loss": 3.6082,
+      "step": 26
+    },
+    {
+      "epoch": 0.004633602196670671,
+      "grad_norm": 26.73604290038376,
+      "learning_rate": 3.085714285714286e-06,
+      "loss": 3.5343,
+      "step": 27
+    },
+    {
+      "epoch": 0.004805217092843659,
+      "grad_norm": 14.263684436385656,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 3.1143,
+      "step": 28
+    },
+    {
+      "epoch": 0.0049768319890166465,
+      "grad_norm": 16.56950614762184,
+      "learning_rate": 3.314285714285714e-06,
+      "loss": 3.035,
+      "step": 29
+    },
+    {
+      "epoch": 0.005148446885189634,
+      "grad_norm": 14.087675339145806,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 3.1071,
+      "step": 30
+    },
+    {
+      "epoch": 0.005320061781362622,
+      "grad_norm": 12.840712423623462,
+      "learning_rate": 3.542857142857143e-06,
+      "loss": 3.0384,
+      "step": 31
+    },
+    {
+      "epoch": 0.00549167667753561,
+      "grad_norm": 17.39947494853309,
+      "learning_rate": 3.6571428571428576e-06,
+      "loss": 3.0405,
+      "step": 32
+    },
+    {
+      "epoch": 0.005663291573708598,
+      "grad_norm": 11.505629537032268,
+      "learning_rate": 3.771428571428572e-06,
+      "loss": 3.0336,
+      "step": 33
+    },
+    {
+      "epoch": 0.005834906469881586,
+      "grad_norm": 21.545273331248218,
+      "learning_rate": 3.885714285714286e-06,
+      "loss": 2.8799,
+      "step": 34
+    },
+    {
+      "epoch": 0.006006521366054574,
+      "grad_norm": 9.353883606209362,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.9076,
+      "step": 35
+    },
+    {
+      "epoch": 0.006178136262227562,
+      "grad_norm": 10.617631469620065,
+      "learning_rate": 4.114285714285715e-06,
+      "loss": 2.8172,
+      "step": 36
+    },
+    {
+      "epoch": 0.0063497511584005495,
+      "grad_norm": 11.770045423370412,
+      "learning_rate": 4.228571428571429e-06,
+      "loss": 2.8444,
+      "step": 37
+    },
+    {
+      "epoch": 0.006521366054573537,
+      "grad_norm": 11.645109020898436,
+      "learning_rate": 4.342857142857143e-06,
+      "loss": 2.8348,
+      "step": 38
+    },
+    {
+      "epoch": 0.006692980950746525,
+      "grad_norm": 10.90549455003319,
+      "learning_rate": 4.457142857142858e-06,
+      "loss": 2.6774,
+      "step": 39
+    },
+    {
+      "epoch": 0.006864595846919512,
+      "grad_norm": 14.637239016287888,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 2.7251,
+      "step": 40
+    },
+    {
+      "epoch": 0.0070362107430925,
+      "grad_norm": 13.431102067923531,
+      "learning_rate": 4.685714285714286e-06,
+      "loss": 2.6214,
+      "step": 41
+    },
+    {
+      "epoch": 0.007207825639265488,
+      "grad_norm": 8.653917935373755,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.6324,
+      "step": 42
+    },
+    {
+      "epoch": 0.007379440535438476,
+      "grad_norm": 16.187652389258293,
+      "learning_rate": 4.9142857142857145e-06,
+      "loss": 2.5056,
+      "step": 43
+    },
+    {
+      "epoch": 0.007551055431611464,
+      "grad_norm": 13.482647329846788,
+      "learning_rate": 5.028571428571429e-06,
+      "loss": 2.5724,
+      "step": 44
+    },
+    {
+      "epoch": 0.0077226703277844515,
+      "grad_norm": 12.801695593411296,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 2.2592,
+      "step": 45
+    },
+    {
+      "epoch": 0.00789428522395744,
+      "grad_norm": 9.137131346103367,
+      "learning_rate": 5.257142857142857e-06,
+      "loss": 2.5445,
+      "step": 46
+    },
+    {
+      "epoch": 0.008065900120130427,
+      "grad_norm": 33.6158256121922,
+      "learning_rate": 5.371428571428572e-06,
+      "loss": 2.3522,
+      "step": 47
+    },
+    {
+      "epoch": 0.008237515016303414,
+      "grad_norm": 25.862161525403106,
+      "learning_rate": 5.485714285714287e-06,
+      "loss": 2.4477,
+      "step": 48
+    },
+    {
+      "epoch": 0.008409129912476403,
+      "grad_norm": 12.286103253639418,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 2.407,
+      "step": 49
+    },
+    {
+      "epoch": 0.00858074480864939,
+      "grad_norm": 17.697613851103835,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 2.4449,
+      "step": 50
+    },
+    {
+      "epoch": 0.008752359704822379,
+      "grad_norm": 15.87446154785113,
+      "learning_rate": 5.828571428571429e-06,
+      "loss": 2.5336,
+      "step": 51
+    },
+    {
+      "epoch": 0.008923974600995366,
+      "grad_norm": 8.762927637790245,
+      "learning_rate": 5.942857142857143e-06,
+      "loss": 2.3772,
+      "step": 52
+    },
+    {
+      "epoch": 0.009095589497168355,
+      "grad_norm": 13.553911318253906,
+      "learning_rate": 6.057142857142858e-06,
+      "loss": 2.512,
+      "step": 53
+    },
+    {
+      "epoch": 0.009267204393341342,
+      "grad_norm": 13.707503059814044,
+      "learning_rate": 6.171428571428572e-06,
+      "loss": 2.7997,
+      "step": 54
+    },
+    {
+      "epoch": 0.00943881928951433,
+      "grad_norm": 6.5694772116241476,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 2.552,
+      "step": 55
+    },
+    {
+      "epoch": 0.009610434185687317,
+      "grad_norm": 7.882295102416172,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 2.6071,
+      "step": 56
+    },
+    {
+      "epoch": 0.009782049081860306,
+      "grad_norm": 6.763526523613429,
+      "learning_rate": 6.514285714285715e-06,
+      "loss": 2.3997,
+      "step": 57
+    },
+    {
+      "epoch": 0.009953663978033293,
+      "grad_norm": 7.0746264269142385,
+      "learning_rate": 6.628571428571428e-06,
+      "loss": 2.2753,
+      "step": 58
+    },
+    {
+      "epoch": 0.010125278874206282,
+      "grad_norm": 4.734415443054102,
+      "learning_rate": 6.742857142857143e-06,
+      "loss": 2.4382,
+      "step": 59
+    },
+    {
+      "epoch": 0.010296893770379269,
+      "grad_norm": 32.23294843067819,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 2.4524,
+      "step": 60
+    },
+    {
+      "epoch": 0.010468508666552257,
+      "grad_norm": 35.15987114461738,
+      "learning_rate": 6.971428571428573e-06,
+      "loss": 2.26,
+      "step": 61
+    },
+    {
+      "epoch": 0.010640123562725244,
+      "grad_norm": 11.266343217208767,
+      "learning_rate": 7.085714285714286e-06,
+      "loss": 2.5029,
+      "step": 62
+    },
+    {
+      "epoch": 0.010811738458898233,
+      "grad_norm": 6.319958231301672,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 2.4196,
+      "step": 63
+    },
+    {
+      "epoch": 0.01098335335507122,
+      "grad_norm": 8.529824841325285,
+      "learning_rate": 7.314285714285715e-06,
+      "loss": 2.637,
+      "step": 64
+    },
+    {
+      "epoch": 0.011154968251244207,
+      "grad_norm": 8.812275589036922,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 2.4787,
+      "step": 65
+    },
+    {
+      "epoch": 0.011326583147417196,
+      "grad_norm": 9.688015408033584,
+      "learning_rate": 7.542857142857144e-06,
+      "loss": 2.2467,
+      "step": 66
+    },
+    {
+      "epoch": 0.011498198043590183,
+      "grad_norm": 9.16959427521578,
+      "learning_rate": 7.657142857142858e-06,
+      "loss": 2.4296,
+      "step": 67
+    },
+    {
+      "epoch": 0.011669812939763172,
+      "grad_norm": 7.62012774598611,
+      "learning_rate": 7.771428571428572e-06,
+      "loss": 2.3844,
+      "step": 68
+    },
+    {
+      "epoch": 0.011841427835936159,
+      "grad_norm": 7.769904677161206,
+      "learning_rate": 7.885714285714286e-06,
+      "loss": 2.372,
+      "step": 69
+    },
+    {
+      "epoch": 0.012013042732109147,
+      "grad_norm": 12.849978359761737,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.3955,
+      "step": 70
+    },
+    {
+      "epoch": 0.012184657628282134,
+      "grad_norm": 12.718571975350164,
+      "learning_rate": 8.114285714285715e-06,
+      "loss": 2.4096,
+      "step": 71
+    },
+    {
+      "epoch": 0.012356272524455123,
+      "grad_norm": 7.981698156226107,
+      "learning_rate": 8.22857142857143e-06,
+      "loss": 2.517,
+      "step": 72
+    },
+    {
+      "epoch": 0.01252788742062811,
+      "grad_norm": 5.642451140300902,
+      "learning_rate": 8.342857142857143e-06,
+      "loss": 2.5422,
+      "step": 73
+    },
+    {
+      "epoch": 0.012699502316801099,
+      "grad_norm": 14.312535444678016,
+      "learning_rate": 8.457142857142859e-06,
+      "loss": 2.4205,
+      "step": 74
+    },
+    {
+      "epoch": 0.012871117212974086,
+      "grad_norm": 8.508921817132261,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 2.4116,
+      "step": 75
+    },
+    {
+      "epoch": 0.013042732109147075,
+      "grad_norm": 6.281352630353262,
+      "learning_rate": 8.685714285714287e-06,
+      "loss": 2.315,
+      "step": 76
+    },
+    {
+      "epoch": 0.013214347005320062,
+      "grad_norm": 5.748376635296371,
+      "learning_rate": 8.8e-06,
+      "loss": 2.4858,
+      "step": 77
+    },
+    {
+      "epoch": 0.01338596190149305,
+      "grad_norm": 5.939096917118166,
+      "learning_rate": 8.914285714285716e-06,
+      "loss": 2.6016,
+      "step": 78
+    },
+    {
+      "epoch": 0.013557576797666037,
+      "grad_norm": 5.907623734872398,
+      "learning_rate": 9.028571428571428e-06,
+      "loss": 2.3556,
+      "step": 79
+    },
+    {
+      "epoch": 0.013729191693839024,
+      "grad_norm": 5.028447859126793,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 2.2946,
+      "step": 80
+    },
+    {
+      "epoch": 0.013900806590012013,
+      "grad_norm": 8.371317073248003,
+      "learning_rate": 9.257142857142858e-06,
+      "loss": 2.1278,
+      "step": 81
+    },
+    {
+      "epoch": 0.014072421486185,
+      "grad_norm": 4.831054716031191,
+      "learning_rate": 9.371428571428572e-06,
+      "loss": 2.4365,
+      "step": 82
+    },
+    {
+      "epoch": 0.014244036382357989,
+      "grad_norm": 5.63478432227459,
+      "learning_rate": 9.485714285714287e-06,
+      "loss": 2.4448,
+      "step": 83
+    },
+    {
+      "epoch": 0.014415651278530976,
+      "grad_norm": 4.044331484487414,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 2.152,
+      "step": 84
+    },
+    {
+      "epoch": 0.014587266174703965,
+      "grad_norm": 6.395798824205685,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 2.0984,
+      "step": 85
+    },
+    {
+      "epoch": 0.014758881070876952,
+      "grad_norm": 9.457886889171613,
+      "learning_rate": 9.828571428571429e-06,
+      "loss": 2.1834,
+      "step": 86
+    },
+    {
+      "epoch": 0.01493049596704994,
+      "grad_norm": 6.977054550828756,
+      "learning_rate": 9.942857142857145e-06,
+      "loss": 2.0698,
+      "step": 87
+    },
+    {
+      "epoch": 0.015102110863222927,
+      "grad_norm": 4.6302943156432965,
+      "learning_rate": 1.0057142857142859e-05,
+      "loss": 2.2912,
+      "step": 88
+    },
+    {
+      "epoch": 0.015273725759395916,
+      "grad_norm": 4.637121728781503,
+      "learning_rate": 1.0171428571428573e-05,
+      "loss": 2.4046,
+      "step": 89
+    },
+    {
+      "epoch": 0.015445340655568903,
+      "grad_norm": 4.969664179177465,
+      "learning_rate": 1.0285714285714285e-05,
+      "loss": 2.479,
+      "step": 90
+    },
+    {
+      "epoch": 0.015616955551741892,
+      "grad_norm": 4.676793232734069,
+      "learning_rate": 1.04e-05,
+      "loss": 2.4297,
+      "step": 91
+    },
+    {
+      "epoch": 0.01578857044791488,
+      "grad_norm": 9.070247616235994,
+      "learning_rate": 1.0514285714285714e-05,
+      "loss": 2.2559,
+      "step": 92
+    },
+    {
+      "epoch": 0.015960185344087868,
+      "grad_norm": 4.5134907813658955,
+      "learning_rate": 1.062857142857143e-05,
+      "loss": 2.2815,
+      "step": 93
+    },
+    {
+      "epoch": 0.016131800240260855,
+      "grad_norm": 4.632550177557325,
+      "learning_rate": 1.0742857142857144e-05,
+      "loss": 2.2171,
+      "step": 94
+    },
+    {
+      "epoch": 0.01630341513643384,
+      "grad_norm": 4.483767211266106,
+      "learning_rate": 1.0857142857142858e-05,
+      "loss": 2.0609,
+      "step": 95
+    },
+    {
+      "epoch": 0.01647503003260683,
+      "grad_norm": 5.813321926149965,
+      "learning_rate": 1.0971428571428573e-05,
+      "loss": 2.0531,
+      "step": 96
+    },
+    {
+      "epoch": 0.01664664492877982,
+      "grad_norm": 4.691478793006599,
+      "learning_rate": 1.1085714285714286e-05,
+      "loss": 2.4327,
+      "step": 97
+    },
+    {
+      "epoch": 0.016818259824952806,
+      "grad_norm": 16.13534090304662,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 2.2145,
+      "step": 98
+    },
+    {
+      "epoch": 0.016989874721125793,
+      "grad_norm": 4.619099358239232,
+      "learning_rate": 1.1314285714285715e-05,
+      "loss": 2.2148,
+      "step": 99
+    },
+    {
+      "epoch": 0.01716148961729878,
+      "grad_norm": 6.83011343785692,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 2.2016,
+      "step": 100
+    },
+    {
+      "epoch": 0.01733310451347177,
+      "grad_norm": 4.285660770520661,
+      "learning_rate": 1.1542857142857145e-05,
+      "loss": 2.2128,
+      "step": 101
+    },
+    {
+      "epoch": 0.017504719409644758,
+      "grad_norm": 6.34310824249887,
+      "learning_rate": 1.1657142857142859e-05,
+      "loss": 2.2913,
+      "step": 102
+    },
+    {
+      "epoch": 0.017676334305817745,
+      "grad_norm": 5.574044081010341,
+      "learning_rate": 1.177142857142857e-05,
+      "loss": 2.5,
+      "step": 103
+    },
+    {
+      "epoch": 0.01784794920199073,
+      "grad_norm": 7.788022677364951,
+      "learning_rate": 1.1885714285714286e-05,
+      "loss": 1.9828,
+      "step": 104
+    },
+    {
+      "epoch": 0.018019564098163722,
+      "grad_norm": 7.23517441529495,
+      "learning_rate": 1.2e-05,
+      "loss": 2.3067,
+      "step": 105
+    },
+    {
+      "epoch": 0.01819117899433671,
+      "grad_norm": 7.056350065099166,
+      "learning_rate": 1.2114285714285716e-05,
+      "loss": 2.1428,
+      "step": 106
+    },
+    {
+      "epoch": 0.018362793890509696,
+      "grad_norm": 5.066019952689693,
+      "learning_rate": 1.222857142857143e-05,
+      "loss": 2.3834,
+      "step": 107
+    },
+    {
+      "epoch": 0.018534408786682683,
+      "grad_norm": 5.033652563830539,
+      "learning_rate": 1.2342857142857144e-05,
+      "loss": 2.4106,
+      "step": 108
+    },
+    {
+      "epoch": 0.018706023682855673,
+      "grad_norm": 8.82881241227772,
+      "learning_rate": 1.245714285714286e-05,
+      "loss": 2.1689,
+      "step": 109
+    },
+    {
+      "epoch": 0.01887763857902866,
+      "grad_norm": 4.020974319459271,
+      "learning_rate": 1.2571428571428572e-05,
+      "loss": 2.0145,
+      "step": 110
+    },
+    {
+      "epoch": 0.019049253475201647,
+      "grad_norm": 5.151370367980264,
+      "learning_rate": 1.2685714285714286e-05,
+      "loss": 2.5154,
+      "step": 111
+    },
+    {
+      "epoch": 0.019220868371374634,
+      "grad_norm": 11.56478613792912,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 2.1864,
+      "step": 112
+    },
+    {
+      "epoch": 0.01939248326754762,
+      "grad_norm": 5.167901030570373,
+      "learning_rate": 1.2914285714285715e-05,
+      "loss": 2.3358,
+      "step": 113
+    },
+    {
+      "epoch": 0.019564098163720612,
+      "grad_norm": 9.787831322651368,
+      "learning_rate": 1.302857142857143e-05,
+      "loss": 2.1912,
+      "step": 114
+    },
+    {
+      "epoch": 0.0197357130598936,
+      "grad_norm": 5.234233774052135,
+      "learning_rate": 1.3142857142857145e-05,
+      "loss": 2.1578,
+      "step": 115
+    },
+    {
+      "epoch": 0.019907327956066586,
+      "grad_norm": 5.664095721805089,
+      "learning_rate": 1.3257142857142857e-05,
+      "loss": 2.381,
+      "step": 116
+    },
+    {
+      "epoch": 0.020078942852239573,
+      "grad_norm": 5.977125695283194,
+      "learning_rate": 1.3371428571428572e-05,
+      "loss": 2.1459,
+      "step": 117
+    },
+    {
+      "epoch": 0.020250557748412563,
+      "grad_norm": 5.007933101525984,
+      "learning_rate": 1.3485714285714286e-05,
+      "loss": 2.283,
+      "step": 118
+    },
+    {
+      "epoch": 0.02042217264458555,
+      "grad_norm": 7.801138402645364,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 2.169,
+      "step": 119
+    },
+    {
+      "epoch": 0.020593787540758537,
+      "grad_norm": 9.627003751448669,
+      "learning_rate": 1.3714285714285716e-05,
+      "loss": 2.1238,
+      "step": 120
+    },
+    {
+      "epoch": 0.020765402436931524,
+      "grad_norm": 6.201321750691995,
+      "learning_rate": 1.382857142857143e-05,
+      "loss": 2.2845,
+      "step": 121
+    },
+    {
+      "epoch": 0.020937017333104515,
+      "grad_norm": 6.739802120345924,
+      "learning_rate": 1.3942857142857145e-05,
+      "loss": 2.0094,
+      "step": 122
+    },
+    {
+      "epoch": 0.021108632229277502,
+      "grad_norm": 6.140016667915861,
+      "learning_rate": 1.4057142857142858e-05,
+      "loss": 1.9595,
+      "step": 123
+    },
+    {
+      "epoch": 0.02128024712545049,
+      "grad_norm": 9.847414600678896,
+      "learning_rate": 1.4171428571428572e-05,
+      "loss": 2.1787,
+      "step": 124
+    },
+    {
+      "epoch": 0.021451862021623476,
+      "grad_norm": 4.69911054437477,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 2.2823,
+      "step": 125
+    },
+    {
+      "epoch": 0.021623476917796466,
+      "grad_norm": 7.500875628785198,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 2.42,
+      "step": 126
+    },
+    {
+      "epoch": 0.021795091813969453,
+      "grad_norm": 5.860447483933667,
+      "learning_rate": 1.4514285714285717e-05,
+      "loss": 2.3998,
+      "step": 127
+    },
+    {
+      "epoch": 0.02196670671014244,
+      "grad_norm": 5.0042507863008625,
+      "learning_rate": 1.462857142857143e-05,
+      "loss": 2.1521,
+      "step": 128
+    },
+    {
+      "epoch": 0.022138321606315427,
+      "grad_norm": 10.48288589489657,
+      "learning_rate": 1.4742857142857143e-05,
+      "loss": 2.2169,
+      "step": 129
+    },
+    {
+      "epoch": 0.022309936502488414,
+      "grad_norm": 4.170799239900871,
+      "learning_rate": 1.4857142857142858e-05,
+      "loss": 2.0158,
+      "step": 130
+    },
+    {
+      "epoch": 0.022481551398661405,
+      "grad_norm": 5.252861584712857,
+      "learning_rate": 1.4971428571428572e-05,
+      "loss": 2.5714,
+      "step": 131
+    },
+    {
+      "epoch": 0.022653166294834392,
+      "grad_norm": 4.7504203282586595,
+      "learning_rate": 1.5085714285714288e-05,
+      "loss": 2.1937,
+      "step": 132
+    },
+    {
+      "epoch": 0.02282478119100738,
+      "grad_norm": 3.8862421876989717,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 2.4138,
+      "step": 133
+    },
+    {
+      "epoch": 0.022996396087180366,
+      "grad_norm": 5.52634558713055,
+      "learning_rate": 1.5314285714285716e-05,
+      "loss": 2.2333,
+      "step": 134
+    },
+    {
+      "epoch": 0.023168010983353356,
+      "grad_norm": 6.031380436251491,
+      "learning_rate": 1.542857142857143e-05,
+      "loss": 2.2482,
+      "step": 135
+    },
+    {
+      "epoch": 0.023339625879526343,
+      "grad_norm": 5.846899422230628,
+      "learning_rate": 1.5542857142857144e-05,
+      "loss": 2.2351,
+      "step": 136
+    },
+    {
+      "epoch": 0.02351124077569933,
+      "grad_norm": 4.499833447260017,
+      "learning_rate": 1.5657142857142856e-05,
+      "loss": 2.0721,
+      "step": 137
+    },
+    {
+      "epoch": 0.023682855671872317,
+      "grad_norm": 5.5578739267495925,
+      "learning_rate": 1.577142857142857e-05,
+      "loss": 2.3898,
+      "step": 138
+    },
+    {
+      "epoch": 0.023854470568045308,
+      "grad_norm": 4.507883690316529,
+      "learning_rate": 1.5885714285714287e-05,
+      "loss": 2.1942,
+      "step": 139
+    },
+    {
+      "epoch": 0.024026085464218295,
+      "grad_norm": 5.465098089669164,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.2732,
+      "step": 140
+    },
+    {
+      "epoch": 0.024197700360391282,
+      "grad_norm": 3.606924457210559,
+      "learning_rate": 1.6114285714285715e-05,
+      "loss": 2.4358,
+      "step": 141
+    },
+    {
+      "epoch": 0.02436931525656427,
+      "grad_norm": 7.609974907749964,
+      "learning_rate": 1.622857142857143e-05,
+      "loss": 2.3202,
+      "step": 142
+    },
+    {
+      "epoch": 0.024540930152737256,
+      "grad_norm": 8.3992014193753,
+      "learning_rate": 1.6342857142857146e-05,
+      "loss": 2.3107,
+      "step": 143
+    },
+    {
+      "epoch": 0.024712545048910246,
+      "grad_norm": 7.943361068023809,
+      "learning_rate": 1.645714285714286e-05,
+      "loss": 2.5093,
+      "step": 144
+    },
+    {
+      "epoch": 0.024884159945083233,
+      "grad_norm": 9.683401712750092,
+      "learning_rate": 1.6571428571428574e-05,
+      "loss": 2.0652,
+      "step": 145
+    },
+    {
+      "epoch": 0.02505577484125622,
+      "grad_norm": 4.851349498462054,
+      "learning_rate": 1.6685714285714286e-05,
+      "loss": 2.1429,
+      "step": 146
+    },
+    {
+      "epoch": 0.025227389737429207,
+      "grad_norm": 6.116749168180479,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 2.2171,
+      "step": 147
+    },
+    {
+      "epoch": 0.025399004633602198,
+      "grad_norm": 21.619649174097454,
+      "learning_rate": 1.6914285714285717e-05,
+      "loss": 2.2228,
+      "step": 148
+    },
+    {
+      "epoch": 0.025570619529775185,
+      "grad_norm": 6.191470229031899,
+      "learning_rate": 1.702857142857143e-05,
+      "loss": 1.9655,
+      "step": 149
+    },
+    {
+      "epoch": 0.025742234425948172,
+      "grad_norm": 6.523677989815507,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 2.1148,
+      "step": 150
+    },
+    {
+      "epoch": 0.02591384932212116,
+      "grad_norm": 3.698357238067846,
+      "learning_rate": 1.7257142857142857e-05,
+      "loss": 2.0033,
+      "step": 151
+    },
+    {
+      "epoch": 0.02608546421829415,
+      "grad_norm": 5.04695114971071,
+      "learning_rate": 1.7371428571428573e-05,
+      "loss": 1.9735,
+      "step": 152
+    },
+    {
+      "epoch": 0.026257079114467136,
+      "grad_norm": 5.282682948320752,
+      "learning_rate": 1.748571428571429e-05,
+      "loss": 2.0038,
+      "step": 153
+    },
+    {
+      "epoch": 0.026428694010640123,
+      "grad_norm": 4.800801803810252,
+      "learning_rate": 1.76e-05,
+      "loss": 2.1585,
+      "step": 154
+    },
+    {
+      "epoch": 0.02660030890681311,
+      "grad_norm": 9.028618098248373,
+      "learning_rate": 1.7714285714285717e-05,
+      "loss": 2.1533,
+      "step": 155
+    },
+    {
+      "epoch": 0.0267719238029861,
+      "grad_norm": 12.347764899908109,
+      "learning_rate": 1.7828571428571432e-05,
+      "loss": 2.323,
+      "step": 156
+    },
+    {
+      "epoch": 0.026943538699159088,
+      "grad_norm": 12.727288499723587,
+      "learning_rate": 1.7942857142857144e-05,
+      "loss": 2.1587,
+      "step": 157
+    },
+    {
+      "epoch": 0.027115153595332075,
+      "grad_norm": 8.45403493722998,
+      "learning_rate": 1.8057142857142857e-05,
+      "loss": 2.2363,
+      "step": 158
+    },
+    {
+      "epoch": 0.027286768491505062,
+      "grad_norm": 3.6371063335499394,
+      "learning_rate": 1.8171428571428572e-05,
+      "loss": 2.2642,
+      "step": 159
+    },
+    {
+      "epoch": 0.02745838338767805,
+      "grad_norm": 8.503641705960693,
+      "learning_rate": 1.8285714285714288e-05,
+      "loss": 2.0775,
+      "step": 160
+    },
+    {
+      "epoch": 0.02762999828385104,
+      "grad_norm": 3.9593201776203903,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 2.0017,
+      "step": 161
+    },
+    {
+      "epoch": 0.027801613180024026,
+      "grad_norm": 4.047122537768316,
+      "learning_rate": 1.8514285714285716e-05,
+      "loss": 2.0842,
+      "step": 162
+    },
+    {
+      "epoch": 0.027973228076197013,
+      "grad_norm": 4.133796081789339,
+      "learning_rate": 1.8628571428571428e-05,
+      "loss": 2.2323,
+      "step": 163
+    },
+    {
+      "epoch": 0.02814484297237,
+      "grad_norm": 3.5505724791516142,
+      "learning_rate": 1.8742857142857143e-05,
+      "loss": 2.4431,
+      "step": 164
+    },
+    {
+      "epoch": 0.02831645786854299,
+      "grad_norm": 6.368793705040636,
+      "learning_rate": 1.885714285714286e-05,
+      "loss": 2.0516,
+      "step": 165
+    },
+    {
+      "epoch": 0.028488072764715978,
+      "grad_norm": 5.05590285307787,
+      "learning_rate": 1.8971428571428575e-05,
+      "loss": 2.0304,
+      "step": 166
+    },
+    {
+      "epoch": 0.028659687660888965,
+      "grad_norm": 4.610753004740335,
+      "learning_rate": 1.9085714285714287e-05,
+      "loss": 2.1065,
+      "step": 167
+    },
+    {
+      "epoch": 0.028831302557061952,
+      "grad_norm": 4.7447497256392,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 2.4162,
+      "step": 168
+    },
+    {
+      "epoch": 0.029002917453234942,
+      "grad_norm": 3.856569889357456,
+      "learning_rate": 1.9314285714285718e-05,
+      "loss": 2.2449,
+      "step": 169
+    },
+    {
+      "epoch": 0.02917453234940793,
+      "grad_norm": 4.010537453686156,
+      "learning_rate": 1.942857142857143e-05,
+      "loss": 2.267,
+      "step": 170
+    },
+    {
+      "epoch": 0.029346147245580916,
+      "grad_norm": 3.895556359814577,
+      "learning_rate": 1.9542857142857143e-05,
+      "loss": 2.1048,
+      "step": 171
+    },
+    {
+      "epoch": 0.029517762141753903,
+      "grad_norm": 3.945479769866957,
+      "learning_rate": 1.9657142857142858e-05,
+      "loss": 1.9937,
+      "step": 172
+    },
+    {
+      "epoch": 0.029689377037926894,
+      "grad_norm": 3.2695958663509788,
+      "learning_rate": 1.9771428571428574e-05,
+      "loss": 1.9953,
+      "step": 173
+    },
+    {
+      "epoch": 0.02986099193409988,
+      "grad_norm": 3.9714084675208574,
+      "learning_rate": 1.988571428571429e-05,
+      "loss": 2.1219,
+      "step": 174
+    },
+    {
+      "epoch": 0.030032606830272868,
+      "grad_norm": 3.662956223801034,
+      "learning_rate": 2e-05,
+      "loss": 2.0746,
+      "step": 175
+    },
+    {
+      "epoch": 0.030204221726445855,
+      "grad_norm": 4.011090108652777,
+      "learning_rate": 1.999999845522429e-05,
+      "loss": 2.2858,
+      "step": 176
+    },
+    {
+      "epoch": 0.03037583662261884,
+      "grad_norm": 4.066209701080919,
+      "learning_rate": 1.9999993820897636e-05,
+      "loss": 1.7863,
+      "step": 177
+    },
+    {
+      "epoch": 0.030547451518791832,
+      "grad_norm": 5.472757350285933,
+      "learning_rate": 1.9999986097021468e-05,
+      "loss": 2.2069,
+      "step": 178
+    },
+    {
+      "epoch": 0.03071906641496482,
+      "grad_norm": 5.834262069284485,
+      "learning_rate": 1.9999975283598174e-05,
+      "loss": 2.2528,
+      "step": 179
+    },
+    {
+      "epoch": 0.030890681311137806,
+      "grad_norm": 3.3155284708163912,
+      "learning_rate": 1.99999613806311e-05,
+      "loss": 1.9184,
+      "step": 180
+    },
+    {
+      "epoch": 0.031062296207310793,
+      "grad_norm": 4.14717746097735,
+      "learning_rate": 1.999994438812453e-05,
+      "loss": 2.1259,
+      "step": 181
+    },
+    {
+      "epoch": 0.031233911103483784,
+      "grad_norm": 3.263252511795784,
+      "learning_rate": 1.999992430608372e-05,
+      "loss": 1.9938,
+      "step": 182
+    },
+    {
+      "epoch": 0.03140552599965677,
+      "grad_norm": 3.4793563126214995,
+      "learning_rate": 1.9999901134514877e-05,
+      "loss": 2.0116,
+      "step": 183
+    },
+    {
+      "epoch": 0.03157714089582976,
+      "grad_norm": 4.728527397407396,
+      "learning_rate": 1.999987487342516e-05,
+      "loss": 1.9359,
+      "step": 184
+    },
+    {
+      "epoch": 0.031748755792002745,
+      "grad_norm": 6.228075711112214,
+      "learning_rate": 1.9999845522822675e-05,
+      "loss": 2.32,
+      "step": 185
+    },
+    {
+      "epoch": 0.031920370688175735,
+      "grad_norm": 8.62147957543855,
+      "learning_rate": 1.9999813082716498e-05,
+      "loss": 2.1879,
+      "step": 186
+    },
+    {
+      "epoch": 0.03209198558434872,
+      "grad_norm": 5.94206466512737,
+      "learning_rate": 1.999977755311665e-05,
+      "loss": 2.2037,
+      "step": 187
+    },
+    {
+      "epoch": 0.03226360048052171,
+      "grad_norm": 4.4989390981166135,
+      "learning_rate": 1.99997389340341e-05,
+      "loss": 2.2205,
+      "step": 188
+    },
+    {
+      "epoch": 0.0324352153766947,
+      "grad_norm": 3.182881699543917,
+      "learning_rate": 1.9999697225480793e-05,
+      "loss": 2.1409,
+      "step": 189
+    },
+    {
+      "epoch": 0.03260683027286768,
+      "grad_norm": 5.806182784255027,
+      "learning_rate": 1.999965242746961e-05,
+      "loss": 2.0681,
+      "step": 190
+    },
+    {
+      "epoch": 0.032778445169040674,
+      "grad_norm": 3.9949497913052134,
+      "learning_rate": 1.9999604540014385e-05,
+      "loss": 1.9099,
+      "step": 191
+    },
+    {
+      "epoch": 0.03295006006521366,
+      "grad_norm": 4.264121839402924,
+      "learning_rate": 1.999955356312992e-05,
+      "loss": 2.3339,
+      "step": 192
+    },
+    {
+      "epoch": 0.03312167496138665,
+      "grad_norm": 3.6493480360975665,
+      "learning_rate": 1.9999499496831964e-05,
+      "loss": 2.1846,
+      "step": 193
+    },
+    {
+      "epoch": 0.03329328985755964,
+      "grad_norm": 3.9791360047438102,
+      "learning_rate": 1.9999442341137216e-05,
+      "loss": 1.7587,
+      "step": 194
+    },
+    {
+      "epoch": 0.03346490475373262,
+      "grad_norm": 4.739405910717689,
+      "learning_rate": 1.999938209606334e-05,
+      "loss": 2.1572,
+      "step": 195
+    },
+    {
+      "epoch": 0.03363651964990561,
+      "grad_norm": 4.698002599709917,
+      "learning_rate": 1.9999318761628946e-05,
+      "loss": 2.0548,
+      "step": 196
+    },
+    {
+      "epoch": 0.0338081345460786,
+      "grad_norm": 4.8135736509450755,
+      "learning_rate": 1.9999252337853605e-05,
+      "loss": 2.2919,
+      "step": 197
+    },
+    {
+      "epoch": 0.033979749442251586,
+      "grad_norm": 4.717374624286988,
+      "learning_rate": 1.9999182824757835e-05,
+      "loss": 2.1388,
+      "step": 198
+    },
+    {
+      "epoch": 0.03415136433842458,
+      "grad_norm": 5.443390808863331,
+      "learning_rate": 1.9999110222363117e-05,
+      "loss": 1.9442,
+      "step": 199
+    },
+    {
+      "epoch": 0.03432297923459756,
+      "grad_norm": 3.5416871939501346,
+      "learning_rate": 1.9999034530691873e-05,
+      "loss": 2.3343,
+      "step": 200
+    },
+    {
+      "epoch": 0.03449459413077055,
+      "grad_norm": 3.928628846334232,
+      "learning_rate": 1.99989557497675e-05,
+      "loss": 2.0561,
+      "step": 201
+    },
+    {
+      "epoch": 0.03466620902694354,
+      "grad_norm": 3.8231391715035294,
+      "learning_rate": 1.999887387961433e-05,
+      "loss": 1.8915,
+      "step": 202
+    },
+    {
+      "epoch": 0.034837823923116525,
+      "grad_norm": 5.148709447014063,
+      "learning_rate": 1.9998788920257658e-05,
+      "loss": 1.9244,
+      "step": 203
+    },
+    {
+      "epoch": 0.035009438819289515,
+      "grad_norm": 5.225707198161678,
+      "learning_rate": 1.9998700871723736e-05,
+      "loss": 2.3055,
+      "step": 204
+    },
+    {
+      "epoch": 0.035181053715462506,
+      "grad_norm": 3.964128867071377,
+      "learning_rate": 1.999860973403976e-05,
+      "loss": 2.2706,
+      "step": 205
+    },
+    {
+      "epoch": 0.03535266861163549,
+      "grad_norm": 4.439156598745672,
+      "learning_rate": 1.99985155072339e-05,
+      "loss": 1.9415,
+      "step": 206
+    },
+    {
+      "epoch": 0.03552428350780848,
+      "grad_norm": 3.747634015520336,
+      "learning_rate": 1.9998418191335257e-05,
+      "loss": 1.9203,
+      "step": 207
+    },
+    {
+      "epoch": 0.03569589840398146,
+      "grad_norm": 9.971101569080222,
+      "learning_rate": 1.99983177863739e-05,
+      "loss": 2.036,
+      "step": 208
+    },
+    {
+      "epoch": 0.035867513300154454,
+      "grad_norm": 3.3921949604506754,
+      "learning_rate": 1.999821429238085e-05,
+      "loss": 2.1034,
+      "step": 209
+    },
+    {
+      "epoch": 0.036039128196327444,
+      "grad_norm": 3.6407447317392494,
+      "learning_rate": 1.9998107709388084e-05,
+      "loss": 2.1111,
+      "step": 210
+    },
+    {
+      "epoch": 0.03621074309250043,
+      "grad_norm": 4.485329884784654,
+      "learning_rate": 1.9997998037428528e-05,
+      "loss": 2.275,
+      "step": 211
+    },
+    {
+      "epoch": 0.03638235798867342,
+      "grad_norm": 4.190281083470862,
+      "learning_rate": 1.9997885276536067e-05,
+      "loss": 2.0305,
+      "step": 212
+    },
+    {
+      "epoch": 0.0365539728848464,
+      "grad_norm": 3.8190741377340967,
+      "learning_rate": 1.9997769426745538e-05,
+      "loss": 2.0532,
+      "step": 213
+    },
+    {
+      "epoch": 0.03672558778101939,
+      "grad_norm": 5.295757263330625,
+      "learning_rate": 1.9997650488092737e-05,
+      "loss": 2.3212,
+      "step": 214
+    },
+    {
+      "epoch": 0.03689720267719238,
+      "grad_norm": 4.7603003206845695,
+      "learning_rate": 1.999752846061441e-05,
+      "loss": 2.0086,
+      "step": 215
+    },
+    {
+      "epoch": 0.037068817573365366,
+      "grad_norm": 5.2877008090048605,
+      "learning_rate": 1.9997403344348255e-05,
+      "loss": 2.0967,
+      "step": 216
+    },
+    {
+      "epoch": 0.037240432469538357,
+      "grad_norm": 5.746606057083038,
+      "learning_rate": 1.9997275139332926e-05,
+      "loss": 2.3167,
+      "step": 217
+    },
+    {
+      "epoch": 0.03741204736571135,
+      "grad_norm": 4.914828526948254,
+      "learning_rate": 1.9997143845608038e-05,
+      "loss": 1.9359,
+      "step": 218
+    },
+    {
+      "epoch": 0.03758366226188433,
+      "grad_norm": 4.192150027643096,
+      "learning_rate": 1.999700946321415e-05,
+      "loss": 2.0423,
+      "step": 219
+    },
+    {
+      "epoch": 0.03775527715805732,
+      "grad_norm": 7.885046348932198,
+      "learning_rate": 1.9996871992192784e-05,
+      "loss": 1.9538,
+      "step": 220
+    },
+    {
+      "epoch": 0.037926892054230305,
+      "grad_norm": 5.930802630560254,
+      "learning_rate": 1.999673143258641e-05,
+      "loss": 2.2622,
+      "step": 221
+    },
+    {
+      "epoch": 0.038098506950403295,
+      "grad_norm": 6.7277426797522875,
+      "learning_rate": 1.9996587784438458e-05,
+      "loss": 1.9707,
+      "step": 222
+    },
+    {
+      "epoch": 0.038270121846576285,
+      "grad_norm": 5.08161908508106,
+      "learning_rate": 1.99964410477933e-05,
+      "loss": 2.1961,
+      "step": 223
+    },
+    {
+      "epoch": 0.03844173674274927,
+      "grad_norm": 7.548243152560011,
+      "learning_rate": 1.9996291222696284e-05,
+      "loss": 1.9572,
+      "step": 224
+    },
+    {
+      "epoch": 0.03861335163892226,
+      "grad_norm": 4.8097191364480185,
+      "learning_rate": 1.9996138309193687e-05,
+      "loss": 1.8955,
+      "step": 225
+    },
+    {
+      "epoch": 0.03878496653509524,
+      "grad_norm": 4.373817752431658,
+      "learning_rate": 1.9995982307332763e-05,
+      "loss": 2.168,
+      "step": 226
+    },
+    {
+      "epoch": 0.038956581431268233,
+      "grad_norm": 3.5792006126973126,
+      "learning_rate": 1.99958232171617e-05,
+      "loss": 2.0136,
+      "step": 227
+    },
+    {
+      "epoch": 0.039128196327441224,
+      "grad_norm": 3.3180076759778605,
+      "learning_rate": 1.9995661038729656e-05,
+      "loss": 2.0724,
+      "step": 228
+    },
+    {
+      "epoch": 0.03929981122361421,
+      "grad_norm": 4.413172277841362,
+      "learning_rate": 1.9995495772086735e-05,
+      "loss": 2.0184,
+      "step": 229
+    },
+    {
+      "epoch": 0.0394714261197872,
+      "grad_norm": 3.50017394385066,
+      "learning_rate": 1.9995327417283997e-05,
+      "loss": 1.9525,
+      "step": 230
+    },
+    {
+      "epoch": 0.03964304101596019,
+      "grad_norm": 4.104073070939531,
+      "learning_rate": 1.9995155974373456e-05,
+      "loss": 2.0843,
+      "step": 231
+    },
+    {
+      "epoch": 0.03981465591213317,
+      "grad_norm": 3.5664780974854615,
+      "learning_rate": 1.999498144340808e-05,
+      "loss": 1.8218,
+      "step": 232
+    },
+    {
+      "epoch": 0.03998627080830616,
+      "grad_norm": 5.408943361359781,
+      "learning_rate": 1.9994803824441794e-05,
+      "loss": 2.2513,
+      "step": 233
+    },
+    {
+      "epoch": 0.040157885704479146,
+      "grad_norm": 3.6523538502640216,
+      "learning_rate": 1.9994623117529467e-05,
+      "loss": 2.0021,
+      "step": 234
+    },
+    {
+      "epoch": 0.040329500600652136,
+      "grad_norm": 8.254834921731893,
+      "learning_rate": 1.999443932272694e-05,
+      "loss": 2.2235,
+      "step": 235
+    },
+    {
+      "epoch": 0.04050111549682513,
+      "grad_norm": 3.7270234847996426,
+      "learning_rate": 1.9994252440090985e-05,
+      "loss": 1.9841,
+      "step": 236
+    },
+    {
+      "epoch": 0.04067273039299811,
+      "grad_norm": 5.974287742095654,
+      "learning_rate": 1.999406246967935e-05,
+      "loss": 2.3204,
+      "step": 237
+    },
+    {
+      "epoch": 0.0408443452891711,
+      "grad_norm": 3.394315816956387,
+      "learning_rate": 1.9993869411550724e-05,
+      "loss": 2.1603,
+      "step": 238
+    },
+    {
+      "epoch": 0.041015960185344084,
+      "grad_norm": 2.592530619481467,
+      "learning_rate": 1.9993673265764753e-05,
+      "loss": 1.8293,
+      "step": 239
+    },
+    {
+      "epoch": 0.041187575081517075,
+      "grad_norm": 3.0827127181917815,
+      "learning_rate": 1.9993474032382035e-05,
+      "loss": 2.0716,
+      "step": 240
+    },
+    {
+      "epoch": 0.041359189977690065,
+      "grad_norm": 3.6407775794491193,
+      "learning_rate": 1.9993271711464133e-05,
+      "loss": 2.0632,
+      "step": 241
+    },
+    {
+      "epoch": 0.04153080487386305,
+      "grad_norm": 3.539435874588333,
+      "learning_rate": 1.9993066303073546e-05,
+      "loss": 2.1099,
+      "step": 242
+    },
+    {
+      "epoch": 0.04170241977003604,
+      "grad_norm": 4.533093018347158,
+      "learning_rate": 1.9992857807273738e-05,
+      "loss": 2.0409,
+      "step": 243
+    },
+    {
+      "epoch": 0.04187403466620903,
+      "grad_norm": 3.417663569586204,
+      "learning_rate": 1.9992646224129127e-05,
+      "loss": 2.1697,
+      "step": 244
+    },
+    {
+      "epoch": 0.04204564956238201,
+      "grad_norm": 4.287451168460292,
+      "learning_rate": 1.9992431553705082e-05,
+      "loss": 2.0607,
+      "step": 245
+    },
+    {
+      "epoch": 0.042217264458555004,
+      "grad_norm": 3.847626705709783,
+      "learning_rate": 1.9992213796067924e-05,
+      "loss": 2.0346,
+      "step": 246
+    },
+    {
+      "epoch": 0.04238887935472799,
+      "grad_norm": 6.781181795874853,
+      "learning_rate": 1.999199295128493e-05,
+      "loss": 2.1162,
+      "step": 247
+    },
+    {
+      "epoch": 0.04256049425090098,
+      "grad_norm": 3.8423922077002626,
+      "learning_rate": 1.999176901942434e-05,
+      "loss": 2.1391,
+      "step": 248
+    },
+    {
+      "epoch": 0.04273210914707397,
+      "grad_norm": 2.944651891036526,
+      "learning_rate": 1.999154200055533e-05,
+      "loss": 2.1784,
+      "step": 249
+    },
+    {
+      "epoch": 0.04290372404324695,
+      "grad_norm": 3.6257027439944634,
+      "learning_rate": 1.9991311894748037e-05,
+      "loss": 2.1781,
+      "step": 250
+    },
+    {
+      "epoch": 0.04307533893941994,
+      "grad_norm": 3.153539075259829,
+      "learning_rate": 1.999107870207356e-05,
+      "loss": 1.8551,
+      "step": 251
+    },
+    {
+      "epoch": 0.04324695383559293,
+      "grad_norm": 4.125627135993159,
+      "learning_rate": 1.9990842422603945e-05,
+      "loss": 1.9995,
+      "step": 252
+    },
+    {
+      "epoch": 0.043418568731765916,
+      "grad_norm": 6.0088583399602555,
+      "learning_rate": 1.9990603056412187e-05,
+      "loss": 2.0258,
+      "step": 253
+    },
+    {
+      "epoch": 0.04359018362793891,
+      "grad_norm": 3.5960539809680356,
+      "learning_rate": 1.9990360603572244e-05,
+      "loss": 2.1001,
+      "step": 254
+    },
+    {
+      "epoch": 0.04376179852411189,
+      "grad_norm": 3.0334651823685213,
+      "learning_rate": 1.9990115064159017e-05,
+      "loss": 2.2487,
+      "step": 255
+    },
+    {
+      "epoch": 0.04393341342028488,
+      "grad_norm": 3.5943226010838436,
+      "learning_rate": 1.9989866438248372e-05,
+      "loss": 2.1614,
+      "step": 256
+    },
+    {
+      "epoch": 0.04410502831645787,
+      "grad_norm": 3.051716911652185,
+      "learning_rate": 1.998961472591712e-05,
+      "loss": 1.936,
+      "step": 257
+    },
+    {
+      "epoch": 0.044276643212630855,
+      "grad_norm": 3.3852311554081944,
+      "learning_rate": 1.9989359927243034e-05,
+      "loss": 2.0865,
+      "step": 258
+    },
+    {
+      "epoch": 0.044448258108803845,
+      "grad_norm": 5.102975445206884,
+      "learning_rate": 1.998910204230483e-05,
+      "loss": 2.0388,
+      "step": 259
+    },
+    {
+      "epoch": 0.04461987300497683,
+      "grad_norm": 4.013708890038843,
+      "learning_rate": 1.9988841071182182e-05,
+      "loss": 2.1261,
+      "step": 260
+    },
+    {
+      "epoch": 0.04479148790114982,
+      "grad_norm": 3.391615244801847,
+      "learning_rate": 1.9988577013955726e-05,
+      "loss": 2.0906,
+      "step": 261
+    },
+    {
+      "epoch": 0.04496310279732281,
+      "grad_norm": 3.8082378978033837,
+      "learning_rate": 1.9988309870707035e-05,
+      "loss": 1.8132,
+      "step": 262
+    },
+    {
+      "epoch": 0.04513471769349579,
+      "grad_norm": 3.9267193517957324,
+      "learning_rate": 1.998803964151865e-05,
+      "loss": 2.0355,
+      "step": 263
+    },
+    {
+      "epoch": 0.045306332589668784,
+      "grad_norm": 3.20937664441744,
+      "learning_rate": 1.998776632647406e-05,
+      "loss": 2.1182,
+      "step": 264
+    },
+    {
+      "epoch": 0.045477947485841774,
+      "grad_norm": 4.126309440434599,
+      "learning_rate": 1.99874899256577e-05,
+      "loss": 2.184,
+      "step": 265
+    },
+    {
+      "epoch": 0.04564956238201476,
+      "grad_norm": 3.137038546242247,
+      "learning_rate": 1.9987210439154973e-05,
+      "loss": 2.0854,
+      "step": 266
+    },
+    {
+      "epoch": 0.04582117727818775,
+      "grad_norm": 3.7249125051993626,
+      "learning_rate": 1.9986927867052226e-05,
+      "loss": 2.2648,
+      "step": 267
+    },
+    {
+      "epoch": 0.04599279217436073,
+      "grad_norm": 5.210752851368842,
+      "learning_rate": 1.9986642209436758e-05,
+      "loss": 1.932,
+      "step": 268
+    },
+    {
+      "epoch": 0.04616440707053372,
+      "grad_norm": 3.5378426429474272,
+      "learning_rate": 1.998635346639683e-05,
+      "loss": 2.1835,
+      "step": 269
+    },
+    {
+      "epoch": 0.04633602196670671,
+      "grad_norm": 2.8272341094270814,
+      "learning_rate": 1.9986061638021643e-05,
+      "loss": 1.932,
+      "step": 270
+    },
+    {
+      "epoch": 0.046507636862879696,
+      "grad_norm": 15.551161972920074,
+      "learning_rate": 1.9985766724401365e-05,
+      "loss": 1.7904,
+      "step": 271
+    },
+    {
+      "epoch": 0.04667925175905269,
+      "grad_norm": 3.4232639279297934,
+      "learning_rate": 1.998546872562711e-05,
+      "loss": 2.2103,
+      "step": 272
+    },
+    {
+      "epoch": 0.04685086665522567,
+      "grad_norm": 3.7729142339088333,
+      "learning_rate": 1.9985167641790948e-05,
+      "loss": 1.7944,
+      "step": 273
+    },
+    {
+      "epoch": 0.04702248155139866,
+      "grad_norm": 3.3560768641924814,
+      "learning_rate": 1.9984863472985892e-05,
+      "loss": 1.744,
+      "step": 274
+    },
+    {
+      "epoch": 0.04719409644757165,
+      "grad_norm": 4.2064486421624245,
+      "learning_rate": 1.998455621930593e-05,
+      "loss": 2.0007,
+      "step": 275
+    },
+    {
+      "epoch": 0.047365711343744635,
+      "grad_norm": 3.193568853483708,
+      "learning_rate": 1.9984245880845975e-05,
+      "loss": 1.9463,
+      "step": 276
+    },
+    {
+      "epoch": 0.047537326239917625,
+      "grad_norm": 3.410498022019692,
+      "learning_rate": 1.9983932457701918e-05,
+      "loss": 1.7809,
+      "step": 277
+    },
+    {
+      "epoch": 0.047708941136090616,
+      "grad_norm": 3.480581671341832,
+      "learning_rate": 1.998361594997059e-05,
+      "loss": 2.1692,
+      "step": 278
+    },
+    {
+      "epoch": 0.0478805560322636,
+      "grad_norm": 3.5109612424418626,
+      "learning_rate": 1.9983296357749774e-05,
+      "loss": 1.9474,
+      "step": 279
+    },
+    {
+      "epoch": 0.04805217092843659,
+      "grad_norm": 3.6369447796663796,
+      "learning_rate": 1.9982973681138215e-05,
+      "loss": 2.198,
+      "step": 280
+    },
+    {
+      "epoch": 0.04822378582460957,
+      "grad_norm": 4.494218086331917,
+      "learning_rate": 1.9982647920235604e-05,
+      "loss": 2.093,
+      "step": 281
+    },
+    {
+      "epoch": 0.048395400720782564,
+      "grad_norm": 4.5341565606847976,
+      "learning_rate": 1.9982319075142585e-05,
+      "loss": 2.0615,
+      "step": 282
+    },
+    {
+      "epoch": 0.048567015616955554,
+      "grad_norm": 3.775401733957357,
+      "learning_rate": 1.998198714596076e-05,
+      "loss": 2.0879,
+      "step": 283
+    },
+    {
+      "epoch": 0.04873863051312854,
+      "grad_norm": 3.224281406562129,
+      "learning_rate": 1.998165213279267e-05,
+      "loss": 2.3028,
+      "step": 284
+    },
+    {
+      "epoch": 0.04891024540930153,
+      "grad_norm": 3.1701862227332374,
+      "learning_rate": 1.9981314035741835e-05,
+      "loss": 2.2169,
+      "step": 285
+    },
+    {
+      "epoch": 0.04908186030547451,
+      "grad_norm": 3.8078015062216886,
+      "learning_rate": 1.99809728549127e-05,
+      "loss": 1.9385,
+      "step": 286
+    },
+    {
+      "epoch": 0.0492534752016475,
+      "grad_norm": 3.4956049530811146,
+      "learning_rate": 1.9980628590410675e-05,
+      "loss": 1.7499,
+      "step": 287
+    },
+    {
+      "epoch": 0.04942509009782049,
+      "grad_norm": 3.1104021548104503,
+      "learning_rate": 1.9980281242342127e-05,
+      "loss": 2.1662,
+      "step": 288
+    },
+    {
+      "epoch": 0.049596704993993476,
+      "grad_norm": 3.5602611180813244,
+      "learning_rate": 1.997993081081437e-05,
+      "loss": 1.7943,
+      "step": 289
+    },
+    {
+      "epoch": 0.04976831989016647,
+      "grad_norm": 2.9550996231707765,
+      "learning_rate": 1.997957729593567e-05,
+      "loss": 2.0344,
+      "step": 290
+    },
+    {
+      "epoch": 0.04993993478633946,
+      "grad_norm": 12.10802116117018,
+      "learning_rate": 1.997922069781525e-05,
+      "loss": 2.2949,
+      "step": 291
+    },
+    {
+      "epoch": 0.05011154968251244,
+      "grad_norm": 3.991348866827493,
+      "learning_rate": 1.997886101656328e-05,
+      "loss": 2.0757,
+      "step": 292
+    },
+    {
+      "epoch": 0.05028316457868543,
+      "grad_norm": 2.9214814563011853,
+      "learning_rate": 1.9978498252290887e-05,
+      "loss": 2.1655,
+      "step": 293
+    },
+    {
+      "epoch": 0.050454779474858415,
+      "grad_norm": 10.328640309606229,
+      "learning_rate": 1.9978132405110148e-05,
+      "loss": 2.0256,
+      "step": 294
+    },
+    {
+      "epoch": 0.050626394371031405,
+      "grad_norm": 3.0822074962729933,
+      "learning_rate": 1.997776347513409e-05,
+      "loss": 1.981,
+      "step": 295
+    },
+    {
+      "epoch": 0.050798009267204396,
+      "grad_norm": 3.7977003070847384,
+      "learning_rate": 1.9977391462476704e-05,
+      "loss": 2.0104,
+      "step": 296
+    },
+    {
+      "epoch": 0.05096962416337738,
+      "grad_norm": 3.129260857004221,
+      "learning_rate": 1.9977016367252916e-05,
+      "loss": 2.0055,
+      "step": 297
+    },
+    {
+      "epoch": 0.05114123905955037,
+      "grad_norm": 2.49936465622327,
+      "learning_rate": 1.9976638189578623e-05,
+      "loss": 2.1314,
+      "step": 298
+    },
+    {
+      "epoch": 0.05131285395572336,
+      "grad_norm": 5.627534615826663,
+      "learning_rate": 1.9976256929570657e-05,
+      "loss": 1.9054,
+      "step": 299
+    },
+    {
+      "epoch": 0.051484468851896344,
+      "grad_norm": 6.644104933532107,
+      "learning_rate": 1.997587258734681e-05,
+      "loss": 2.0553,
+      "step": 300
+    },
+    {
+      "epoch": 0.051656083748069334,
+      "grad_norm": 6.20781166333256,
+      "learning_rate": 1.9975485163025837e-05,
+      "loss": 1.7756,
+      "step": 301
+    },
+    {
+      "epoch": 0.05182769864424232,
+      "grad_norm": 4.777678957881222,
+      "learning_rate": 1.9975094656727424e-05,
+      "loss": 1.9228,
+      "step": 302
+    },
+    {
+      "epoch": 0.05199931354041531,
+      "grad_norm": 3.2413464766698676,
+      "learning_rate": 1.9974701068572223e-05,
+      "loss": 1.9268,
+      "step": 303
+    },
+    {
+      "epoch": 0.0521709284365883,
+      "grad_norm": 3.4243989209979513,
+      "learning_rate": 1.9974304398681837e-05,
+      "loss": 2.0327,
+      "step": 304
+    },
+    {
+      "epoch": 0.05234254333276128,
+      "grad_norm": 3.6407848281373316,
+      "learning_rate": 1.9973904647178816e-05,
+      "loss": 1.8295,
+      "step": 305
+    },
+    {
+      "epoch": 0.05251415822893427,
+      "grad_norm": 3.277014675061404,
+      "learning_rate": 1.997350181418667e-05,
+      "loss": 1.7963,
+      "step": 306
+    },
+    {
+      "epoch": 0.052685773125107256,
+      "grad_norm": 4.215647349610302,
+      "learning_rate": 1.997309589982985e-05,
+      "loss": 1.8516,
+      "step": 307
+    },
+    {
+      "epoch": 0.05285738802128025,
+      "grad_norm": 3.546260309218347,
+      "learning_rate": 1.997268690423377e-05,
+      "loss": 2.0199,
+      "step": 308
+    },
+    {
+      "epoch": 0.05302900291745324,
+      "grad_norm": 5.221183858099176,
+      "learning_rate": 1.997227482752479e-05,
+      "loss": 1.976,
+      "step": 309
+    },
+    {
+      "epoch": 0.05320061781362622,
+      "grad_norm": 3.7676627879817124,
+      "learning_rate": 1.997185966983022e-05,
+      "loss": 1.9,
+      "step": 310
+    },
+    {
+      "epoch": 0.05337223270979921,
+      "grad_norm": 3.52804983582057,
+      "learning_rate": 1.9971441431278337e-05,
+      "loss": 1.9173,
+      "step": 311
+    },
+    {
+      "epoch": 0.0535438476059722,
+      "grad_norm": 6.530673481853173,
+      "learning_rate": 1.9971020111998342e-05,
+      "loss": 2.0977,
+      "step": 312
+    },
+    {
+      "epoch": 0.053715462502145185,
+      "grad_norm": 5.0170849453421855,
+      "learning_rate": 1.9970595712120414e-05,
+      "loss": 2.0448,
+      "step": 313
+    },
+    {
+      "epoch": 0.053887077398318176,
+      "grad_norm": 4.130987581333207,
+      "learning_rate": 1.997016823177567e-05,
+      "loss": 1.8148,
+      "step": 314
+    },
+    {
+      "epoch": 0.05405869229449116,
+      "grad_norm": 6.215053078586563,
+      "learning_rate": 1.9969737671096182e-05,
+      "loss": 1.944,
+      "step": 315
+    },
+    {
+      "epoch": 0.05423030719066415,
+      "grad_norm": 3.8873415625275602,
+      "learning_rate": 1.9969304030214976e-05,
+      "loss": 1.9012,
+      "step": 316
+    },
+    {
+      "epoch": 0.05440192208683714,
+      "grad_norm": 3.9648633375059834,
+      "learning_rate": 1.9968867309266025e-05,
+      "loss": 1.834,
+      "step": 317
+    },
+    {
+      "epoch": 0.054573536983010124,
+      "grad_norm": 3.260714173660746,
+      "learning_rate": 1.9968427508384256e-05,
+      "loss": 2.0917,
+      "step": 318
+    },
+    {
+      "epoch": 0.054745151879183114,
+      "grad_norm": 3.3986200126057033,
+      "learning_rate": 1.996798462770555e-05,
+      "loss": 2.0721,
+      "step": 319
+    },
+    {
+      "epoch": 0.0549167667753561,
+      "grad_norm": 2.8063111215512357,
+      "learning_rate": 1.9967538667366738e-05,
+      "loss": 1.7492,
+      "step": 320
+    },
+    {
+      "epoch": 0.05508838167152909,
+      "grad_norm": 3.148377573614974,
+      "learning_rate": 1.99670896275056e-05,
+      "loss": 1.7018,
+      "step": 321
+    },
+    {
+      "epoch": 0.05525999656770208,
+      "grad_norm": 3.036364302057122,
+      "learning_rate": 1.996663750826087e-05,
+      "loss": 1.9847,
+      "step": 322
+    },
+    {
+      "epoch": 0.05543161146387506,
+      "grad_norm": 3.4340609311578767,
+      "learning_rate": 1.996618230977223e-05,
+      "loss": 1.9873,
+      "step": 323
+    },
+    {
+      "epoch": 0.05560322636004805,
+      "grad_norm": 3.483425040286511,
+      "learning_rate": 1.996572403218032e-05,
+      "loss": 2.2108,
+      "step": 324
+    },
+    {
+      "epoch": 0.05577484125622104,
+      "grad_norm": 2.6320552294886586,
+      "learning_rate": 1.9965262675626726e-05,
+      "loss": 1.977,
+      "step": 325
+    },
+    {
+      "epoch": 0.055946456152394027,
+      "grad_norm": 3.7381712119779547,
+      "learning_rate": 1.996479824025398e-05,
+      "loss": 1.9211,
+      "step": 326
+    },
+    {
+      "epoch": 0.05611807104856702,
+      "grad_norm": 3.014971799857585,
+      "learning_rate": 1.9964330726205585e-05,
+      "loss": 2.0162,
+      "step": 327
+    },
+    {
+      "epoch": 0.05628968594474,
+      "grad_norm": 3.7417585676111815,
+      "learning_rate": 1.9963860133625967e-05,
+      "loss": 2.0456,
+      "step": 328
+    },
+    {
+      "epoch": 0.05646130084091299,
+      "grad_norm": 3.517090546394605,
+      "learning_rate": 1.996338646266053e-05,
+      "loss": 2.041,
+      "step": 329
+    },
+    {
+      "epoch": 0.05663291573708598,
+      "grad_norm": 3.51522607791768,
+      "learning_rate": 1.996290971345561e-05,
+      "loss": 1.9018,
+      "step": 330
+    },
+    {
+      "epoch": 0.056804530633258965,
+      "grad_norm": 4.245594218887946,
+      "learning_rate": 1.9962429886158503e-05,
+      "loss": 2.1062,
+      "step": 331
+    },
+    {
+      "epoch": 0.056976145529431955,
+      "grad_norm": 3.9658374350384107,
+      "learning_rate": 1.9961946980917457e-05,
+      "loss": 2.0405,
+      "step": 332
+    },
+    {
+      "epoch": 0.057147760425604946,
+      "grad_norm": 2.798536101245012,
+      "learning_rate": 1.9961460997881664e-05,
+      "loss": 1.7389,
+      "step": 333
+    },
+    {
+      "epoch": 0.05731937532177793,
+      "grad_norm": 2.622222621230687,
+      "learning_rate": 1.9960971937201273e-05,
+      "loss": 1.8965,
+      "step": 334
+    },
+    {
+      "epoch": 0.05749099021795092,
+      "grad_norm": 3.5051816941479963,
+      "learning_rate": 1.996047979902738e-05,
+      "loss": 1.9965,
+      "step": 335
+    },
+    {
+      "epoch": 0.057662605114123903,
+      "grad_norm": 4.023153275223512,
+      "learning_rate": 1.9959984583512034e-05,
+      "loss": 2.0692,
+      "step": 336
+    },
+    {
+      "epoch": 0.057834220010296894,
+      "grad_norm": 3.3095562234726668,
+      "learning_rate": 1.995948629080824e-05,
+      "loss": 2.1669,
+      "step": 337
+    },
+    {
+      "epoch": 0.058005834906469884,
+      "grad_norm": 3.250292635314978,
+      "learning_rate": 1.995898492106994e-05,
+      "loss": 1.963,
+      "step": 338
+    },
+    {
+      "epoch": 0.05817744980264287,
+      "grad_norm": 3.7160346187443776,
+      "learning_rate": 1.9958480474452038e-05,
+      "loss": 2.0932,
+      "step": 339
+    },
+    {
+      "epoch": 0.05834906469881586,
+      "grad_norm": 5.363449367366593,
+      "learning_rate": 1.9957972951110384e-05,
+      "loss": 1.8988,
+      "step": 340
+    },
+    {
+      "epoch": 0.05852067959498884,
+      "grad_norm": 2.9097716754228724,
+      "learning_rate": 1.9957462351201787e-05,
+      "loss": 2.0416,
+      "step": 341
+    },
+    {
+      "epoch": 0.05869229449116183,
+      "grad_norm": 3.998268814523747,
+      "learning_rate": 1.995694867488399e-05,
+      "loss": 1.9339,
+      "step": 342
+    },
+    {
+      "epoch": 0.05886390938733482,
+      "grad_norm": 3.263953790011657,
+      "learning_rate": 1.9956431922315696e-05,
+      "loss": 1.8995,
+      "step": 343
+    },
+    {
+      "epoch": 0.059035524283507806,
+      "grad_norm": 3.567779664486138,
+      "learning_rate": 1.995591209365657e-05,
+      "loss": 1.9859,
+      "step": 344
+    },
+    {
+      "epoch": 0.0592071391796808,
+      "grad_norm": 3.2861151704697247,
+      "learning_rate": 1.9955389189067205e-05,
+      "loss": 1.9084,
+      "step": 345
+    },
+    {
+      "epoch": 0.05937875407585379,
+      "grad_norm": 3.5321753686438653,
+      "learning_rate": 1.9954863208709155e-05,
+      "loss": 2.0222,
+      "step": 346
+    },
+    {
+      "epoch": 0.05955036897202677,
+      "grad_norm": 3.4816258808995637,
+      "learning_rate": 1.995433415274493e-05,
+      "loss": 1.8883,
+      "step": 347
+    },
+    {
+      "epoch": 0.05972198386819976,
+      "grad_norm": 2.9457399292289317,
+      "learning_rate": 1.9953802021337984e-05,
+      "loss": 1.8481,
+      "step": 348
+    },
+    {
+      "epoch": 0.059893598764372745,
+      "grad_norm": 3.0190464376361485,
+      "learning_rate": 1.9953266814652715e-05,
+      "loss": 1.9166,
+      "step": 349
+    },
+    {
+      "epoch": 0.060065213660545735,
+      "grad_norm": 2.74366104722361,
+      "learning_rate": 1.9952728532854488e-05,
+      "loss": 2.1952,
+      "step": 350
+    },
+    {
+      "epoch": 0.060236828556718726,
+      "grad_norm": 4.406679756642423,
+      "learning_rate": 1.9952187176109598e-05,
+      "loss": 2.3529,
+      "step": 351
+    },
+    {
+      "epoch": 0.06040844345289171,
+      "grad_norm": 3.489249349264093,
+      "learning_rate": 1.9951642744585307e-05,
+      "loss": 2.0213,
+      "step": 352
+    },
+    {
+      "epoch": 0.0605800583490647,
+      "grad_norm": 4.451462316787464,
+      "learning_rate": 1.9951095238449816e-05,
+      "loss": 1.9524,
+      "step": 353
+    },
+    {
+      "epoch": 0.06075167324523768,
+      "grad_norm": 3.346050256969279,
+      "learning_rate": 1.995054465787228e-05,
+      "loss": 2.1326,
+      "step": 354
+    },
+    {
+      "epoch": 0.060923288141410674,
+      "grad_norm": 3.274753224765989,
+      "learning_rate": 1.994999100302281e-05,
+      "loss": 1.9726,
+      "step": 355
+    },
+    {
+      "epoch": 0.061094903037583664,
+      "grad_norm": 3.452040592920918,
+      "learning_rate": 1.994943427407245e-05,
+      "loss": 2.0665,
+      "step": 356
+    },
+    {
+      "epoch": 0.06126651793375665,
+      "grad_norm": 3.325702290390705,
+      "learning_rate": 1.994887447119321e-05,
+      "loss": 1.8927,
+      "step": 357
+    },
+    {
+      "epoch": 0.06143813282992964,
+      "grad_norm": 2.8228953686847116,
+      "learning_rate": 1.9948311594558044e-05,
+      "loss": 1.9875,
+      "step": 358
+    },
+    {
+      "epoch": 0.06160974772610263,
+      "grad_norm": 3.1045240616279597,
+      "learning_rate": 1.994774564434086e-05,
+      "loss": 2.274,
+      "step": 359
+    },
+    {
+      "epoch": 0.06178136262227561,
+      "grad_norm": 3.616583977726727,
+      "learning_rate": 1.99471766207165e-05,
+      "loss": 1.8978,
+      "step": 360
+    },
+    {
+      "epoch": 0.0619529775184486,
+      "grad_norm": 5.010761526550373,
+      "learning_rate": 1.9946604523860775e-05,
+      "loss": 2.0328,
+      "step": 361
+    },
+    {
+      "epoch": 0.062124592414621586,
+      "grad_norm": 3.1813612302118255,
+      "learning_rate": 1.9946029353950438e-05,
+      "loss": 2.1191,
+      "step": 362
+    },
+    {
+      "epoch": 0.06229620731079458,
+      "grad_norm": 8.461598649395224,
+      "learning_rate": 1.9945451111163184e-05,
+      "loss": 2.1219,
+      "step": 363
+    },
+    {
+      "epoch": 0.06246782220696757,
+      "grad_norm": 3.6026486209249753,
+      "learning_rate": 1.9944869795677673e-05,
+      "loss": 1.9317,
+      "step": 364
+    },
+    {
+      "epoch": 0.06263943710314056,
+      "grad_norm": 3.1743447509788703,
+      "learning_rate": 1.99442854076735e-05,
+      "loss": 1.8406,
+      "step": 365
+    },
+    {
+      "epoch": 0.06281105199931354,
+      "grad_norm": 2.9018765647258724,
+      "learning_rate": 1.994369794733121e-05,
+      "loss": 1.9747,
+      "step": 366
+    },
+    {
+      "epoch": 0.06298266689548652,
+      "grad_norm": 4.509269773211525,
+      "learning_rate": 1.9943107414832314e-05,
+      "loss": 1.7696,
+      "step": 367
+    },
+    {
+      "epoch": 0.06315428179165952,
+      "grad_norm": 3.418384148627606,
+      "learning_rate": 1.994251381035925e-05,
+      "loss": 1.9637,
+      "step": 368
+    },
+    {
+      "epoch": 0.0633258966878325,
+      "grad_norm": 2.693871077608062,
+      "learning_rate": 1.994191713409542e-05,
+      "loss": 2.0829,
+      "step": 369
+    },
+    {
+      "epoch": 0.06349751158400549,
+      "grad_norm": 3.6544102213339675,
+      "learning_rate": 1.9941317386225165e-05,
+      "loss": 2.1403,
+      "step": 370
+    },
+    {
+      "epoch": 0.06366912648017847,
+      "grad_norm": 3.4147609805280825,
+      "learning_rate": 1.9940714566933787e-05,
+      "loss": 2.0245,
+      "step": 371
+    },
+    {
+      "epoch": 0.06384074137635147,
+      "grad_norm": 3.51893148223146,
+      "learning_rate": 1.994010867640753e-05,
+      "loss": 1.9533,
+      "step": 372
+    },
+    {
+      "epoch": 0.06401235627252445,
+      "grad_norm": 6.446889997539744,
+      "learning_rate": 1.993949971483358e-05,
+      "loss": 2.2087,
+      "step": 373
+    },
+    {
+      "epoch": 0.06418397116869744,
+      "grad_norm": 3.5840239062714696,
+      "learning_rate": 1.9938887682400084e-05,
+      "loss": 2.0374,
+      "step": 374
+    },
+    {
+      "epoch": 0.06435558606487043,
+      "grad_norm": 3.564481556922706,
+      "learning_rate": 1.993827257929613e-05,
+      "loss": 2.0596,
+      "step": 375
+    },
+    {
+      "epoch": 0.06452720096104342,
+      "grad_norm": 6.000711626236975,
+      "learning_rate": 1.9937654405711758e-05,
+      "loss": 1.9471,
+      "step": 376
+    },
+    {
+      "epoch": 0.0646988158572164,
+      "grad_norm": 3.593852534204912,
+      "learning_rate": 1.993703316183796e-05,
+      "loss": 1.9825,
+      "step": 377
+    },
+    {
+      "epoch": 0.0648704307533894,
+      "grad_norm": 3.3036832169777335,
+      "learning_rate": 1.9936408847866668e-05,
+      "loss": 2.0102,
+      "step": 378
+    },
+    {
+      "epoch": 0.06504204564956238,
+      "grad_norm": 3.052793501014722,
+      "learning_rate": 1.9935781463990765e-05,
+      "loss": 2.0017,
+      "step": 379
+    },
+    {
+      "epoch": 0.06521366054573537,
+      "grad_norm": 3.6612148008631733,
+      "learning_rate": 1.993515101040409e-05,
+      "loss": 1.8878,
+      "step": 380
+    },
+    {
+      "epoch": 0.06538527544190836,
+      "grad_norm": 3.475842753389438,
+      "learning_rate": 1.9934517487301423e-05,
+      "loss": 1.9538,
+      "step": 381
+    },
+    {
+      "epoch": 0.06555689033808135,
+      "grad_norm": 4.998090766931655,
+      "learning_rate": 1.9933880894878492e-05,
+      "loss": 1.8712,
+      "step": 382
+    },
+    {
+      "epoch": 0.06572850523425433,
+      "grad_norm": 2.589300930701225,
+      "learning_rate": 1.993324123333198e-05,
+      "loss": 2.044,
+      "step": 383
+    },
+    {
+      "epoch": 0.06590012013042731,
+      "grad_norm": 3.780280718282726,
+      "learning_rate": 1.9932598502859507e-05,
+      "loss": 2.0189,
+      "step": 384
+    },
+    {
+      "epoch": 0.06607173502660031,
+      "grad_norm": 3.730079966278093,
+      "learning_rate": 1.9931952703659655e-05,
+      "loss": 1.7114,
+      "step": 385
+    },
+    {
+      "epoch": 0.0662433499227733,
+      "grad_norm": 3.5938600807071572,
+      "learning_rate": 1.9931303835931942e-05,
+      "loss": 1.9016,
+      "step": 386
+    },
+    {
+      "epoch": 0.06641496481894628,
+      "grad_norm": 2.898200117914242,
+      "learning_rate": 1.993065189987684e-05,
+      "loss": 1.9806,
+      "step": 387
+    },
+    {
+      "epoch": 0.06658657971511928,
+      "grad_norm": 3.945719611649247,
+      "learning_rate": 1.992999689569577e-05,
+      "loss": 1.9177,
+      "step": 388
+    },
+    {
+      "epoch": 0.06675819461129226,
+      "grad_norm": 2.747466871189157,
+      "learning_rate": 1.9929338823591097e-05,
+      "loss": 2.1244,
+      "step": 389
+    },
+    {
+      "epoch": 0.06692980950746524,
+      "grad_norm": 3.18948722466054,
+      "learning_rate": 1.9928677683766135e-05,
+      "loss": 2.2223,
+      "step": 390
+    },
+    {
+      "epoch": 0.06710142440363824,
+      "grad_norm": 3.181927887612582,
+      "learning_rate": 1.992801347642515e-05,
+      "loss": 1.8363,
+      "step": 391
+    },
+    {
+      "epoch": 0.06727303929981122,
+      "grad_norm": 3.495407532764056,
+      "learning_rate": 1.992734620177335e-05,
+      "loss": 1.9259,
+      "step": 392
+    },
+    {
+      "epoch": 0.06744465419598421,
+      "grad_norm": 3.155084394755952,
+      "learning_rate": 1.9926675860016897e-05,
+      "loss": 1.8842,
+      "step": 393
+    },
+    {
+      "epoch": 0.0676162690921572,
+      "grad_norm": 4.434655326993586,
+      "learning_rate": 1.9926002451362886e-05,
+      "loss": 2.1352,
+      "step": 394
+    },
+    {
+      "epoch": 0.06778788398833019,
+      "grad_norm": 3.173446973582143,
+      "learning_rate": 1.992532597601938e-05,
+      "loss": 2.0347,
+      "step": 395
+    },
+    {
+      "epoch": 0.06795949888450317,
+      "grad_norm": 4.424686902944709,
+      "learning_rate": 1.9924646434195375e-05,
+      "loss": 1.834,
+      "step": 396
+    },
+    {
+      "epoch": 0.06813111378067616,
+      "grad_norm": 3.019727201660959,
+      "learning_rate": 1.992396382610082e-05,
+      "loss": 1.9919,
+      "step": 397
+    },
+    {
+      "epoch": 0.06830272867684915,
+      "grad_norm": 3.7108070834984654,
+      "learning_rate": 1.9923278151946614e-05,
+      "loss": 1.8066,
+      "step": 398
+    },
+    {
+      "epoch": 0.06847434357302214,
+      "grad_norm": 5.942983928233282,
+      "learning_rate": 1.9922589411944592e-05,
+      "loss": 1.9729,
+      "step": 399
+    },
+    {
+      "epoch": 0.06864595846919512,
+      "grad_norm": 3.633377819528754,
+      "learning_rate": 1.992189760630755e-05,
+      "loss": 1.8059,
+      "step": 400
+    },
+    {
+      "epoch": 0.06881757336536812,
+      "grad_norm": 4.220312113768063,
+      "learning_rate": 1.9921202735249224e-05,
+      "loss": 1.7966,
+      "step": 401
+    },
+    {
+      "epoch": 0.0689891882615411,
+      "grad_norm": 3.5661471053114893,
+      "learning_rate": 1.9920504798984293e-05,
+      "loss": 1.9462,
+      "step": 402
+    },
+    {
+      "epoch": 0.06916080315771408,
+      "grad_norm": 3.5535047546791594,
+      "learning_rate": 1.9919803797728393e-05,
+      "loss": 1.8868,
+      "step": 403
+    },
+    {
+      "epoch": 0.06933241805388708,
+      "grad_norm": 3.3674287157828013,
+      "learning_rate": 1.9919099731698103e-05,
+      "loss": 1.9673,
+      "step": 404
+    },
+    {
+      "epoch": 0.06950403295006007,
+      "grad_norm": 2.8131842269351166,
+      "learning_rate": 1.991839260111094e-05,
+      "loss": 2.0662,
+      "step": 405
+    },
+    {
+      "epoch": 0.06967564784623305,
+      "grad_norm": 3.782756029035063,
+      "learning_rate": 1.9917682406185388e-05,
+      "loss": 2.1042,
+      "step": 406
+    },
+    {
+      "epoch": 0.06984726274240605,
+      "grad_norm": 3.2660228549491497,
+      "learning_rate": 1.9916969147140856e-05,
+      "loss": 1.9055,
+      "step": 407
+    },
+    {
+      "epoch": 0.07001887763857903,
+      "grad_norm": 3.769049985211678,
+      "learning_rate": 1.991625282419771e-05,
+      "loss": 1.7759,
+      "step": 408
+    },
+    {
+      "epoch": 0.07019049253475201,
+      "grad_norm": 4.273349197912016,
+      "learning_rate": 1.9915533437577264e-05,
+      "loss": 2.1668,
+      "step": 409
+    },
+    {
+      "epoch": 0.07036210743092501,
+      "grad_norm": 4.4434702887682835,
+      "learning_rate": 1.9914810987501776e-05,
+      "loss": 1.9974,
+      "step": 410
+    },
+    {
+      "epoch": 0.070533722327098,
+      "grad_norm": 3.417576560816798,
+      "learning_rate": 1.991408547419445e-05,
+      "loss": 1.962,
+      "step": 411
+    },
+    {
+      "epoch": 0.07070533722327098,
+      "grad_norm": 3.700950392107592,
+      "learning_rate": 1.991335689787944e-05,
+      "loss": 1.9546,
+      "step": 412
+    },
+    {
+      "epoch": 0.07087695211944396,
+      "grad_norm": 3.26947208946921,
+      "learning_rate": 1.9912625258781835e-05,
+      "loss": 1.7231,
+      "step": 413
+    },
+    {
+      "epoch": 0.07104856701561696,
+      "grad_norm": 4.046723169623149,
+      "learning_rate": 1.9911890557127685e-05,
+      "loss": 1.9454,
+      "step": 414
+    },
+    {
+      "epoch": 0.07122018191178994,
+      "grad_norm": 4.412385919431148,
+      "learning_rate": 1.991115279314398e-05,
+      "loss": 1.6827,
+      "step": 415
+    },
+    {
+      "epoch": 0.07139179680796293,
+      "grad_norm": 3.05634464104461,
+      "learning_rate": 1.9910411967058656e-05,
+      "loss": 2.0428,
+      "step": 416
+    },
+    {
+      "epoch": 0.07156341170413592,
+      "grad_norm": 3.249064810850331,
+      "learning_rate": 1.9909668079100595e-05,
+      "loss": 1.8849,
+      "step": 417
+    },
+    {
+      "epoch": 0.07173502660030891,
+      "grad_norm": 3.187057708839728,
+      "learning_rate": 1.9908921129499624e-05,
+      "loss": 2.0543,
+      "step": 418
+    },
+    {
+      "epoch": 0.07190664149648189,
+      "grad_norm": 12.91007819072769,
+      "learning_rate": 1.9908171118486514e-05,
+      "loss": 1.8321,
+      "step": 419
+    },
+    {
+      "epoch": 0.07207825639265489,
+      "grad_norm": 2.7855151460674326,
+      "learning_rate": 1.990741804629299e-05,
+      "loss": 2.0461,
+      "step": 420
+    },
+    {
+      "epoch": 0.07224987128882787,
+      "grad_norm": 2.9537147805502277,
+      "learning_rate": 1.9906661913151716e-05,
+      "loss": 1.8317,
+      "step": 421
+    },
+    {
+      "epoch": 0.07242148618500086,
+      "grad_norm": 3.927997817390927,
+      "learning_rate": 1.99059027192963e-05,
+      "loss": 1.9384,
+      "step": 422
+    },
+    {
+      "epoch": 0.07259310108117385,
+      "grad_norm": 4.094598512833103,
+      "learning_rate": 1.9905140464961304e-05,
+      "loss": 1.9198,
+      "step": 423
+    },
+    {
+      "epoch": 0.07276471597734684,
+      "grad_norm": 4.239176127453716,
+      "learning_rate": 1.9904375150382223e-05,
+      "loss": 1.8482,
+      "step": 424
+    },
+    {
+      "epoch": 0.07293633087351982,
+      "grad_norm": 4.024275611068595,
+      "learning_rate": 1.9903606775795518e-05,
+      "loss": 2.0661,
+      "step": 425
+    },
+    {
+      "epoch": 0.0731079457696928,
+      "grad_norm": 3.1193533477311663,
+      "learning_rate": 1.9902835341438565e-05,
+      "loss": 1.8587,
+      "step": 426
+    },
+    {
+      "epoch": 0.0732795606658658,
+      "grad_norm": 2.512491975849034,
+      "learning_rate": 1.9902060847549716e-05,
+      "loss": 1.8338,
+      "step": 427
+    },
+    {
+      "epoch": 0.07345117556203878,
+      "grad_norm": 2.658662449947897,
+      "learning_rate": 1.990128329436825e-05,
+      "loss": 1.8456,
+      "step": 428
+    },
+    {
+      "epoch": 0.07362279045821177,
+      "grad_norm": 3.6995282944953716,
+      "learning_rate": 1.99005026821344e-05,
+      "loss": 2.0521,
+      "step": 429
+    },
+    {
+      "epoch": 0.07379440535438477,
+      "grad_norm": 2.6358572188183444,
+      "learning_rate": 1.989971901108933e-05,
+      "loss": 1.7836,
+      "step": 430
+    },
+    {
+      "epoch": 0.07396602025055775,
+      "grad_norm": 4.127850827221215,
+      "learning_rate": 1.989893228147517e-05,
+      "loss": 1.8566,
+      "step": 431
+    },
+    {
+      "epoch": 0.07413763514673073,
+      "grad_norm": 3.0195471907690847,
+      "learning_rate": 1.9898142493534975e-05,
+      "loss": 1.9363,
+      "step": 432
+    },
+    {
+      "epoch": 0.07430925004290373,
+      "grad_norm": 3.3071702203798874,
+      "learning_rate": 1.9897349647512762e-05,
+      "loss": 2.1171,
+      "step": 433
+    },
+    {
+      "epoch": 0.07448086493907671,
+      "grad_norm": 3.6398663295664493,
+      "learning_rate": 1.9896553743653484e-05,
+      "loss": 1.9054,
+      "step": 434
+    },
+    {
+      "epoch": 0.0746524798352497,
+      "grad_norm": 2.830364598806842,
+      "learning_rate": 1.9895754782203035e-05,
+      "loss": 1.8125,
+      "step": 435
+    },
+    {
+      "epoch": 0.0748240947314227,
+      "grad_norm": 3.0352347000519364,
+      "learning_rate": 1.989495276340826e-05,
+      "loss": 1.9944,
+      "step": 436
+    },
+    {
+      "epoch": 0.07499570962759568,
+      "grad_norm": 3.0099781708301045,
+      "learning_rate": 1.989414768751695e-05,
+      "loss": 1.9338,
+      "step": 437
+    },
+    {
+      "epoch": 0.07516732452376866,
+      "grad_norm": 3.315329810451368,
+      "learning_rate": 1.9893339554777834e-05,
+      "loss": 1.9966,
+      "step": 438
+    },
+    {
+      "epoch": 0.07533893941994164,
+      "grad_norm": 7.665918309754256,
+      "learning_rate": 1.989252836544059e-05,
+      "loss": 1.9866,
+      "step": 439
+    },
+    {
+      "epoch": 0.07551055431611464,
+      "grad_norm": 3.408241793814834,
+      "learning_rate": 1.989171411975584e-05,
+      "loss": 2.0964,
+      "step": 440
+    },
+    {
+      "epoch": 0.07568216921228763,
+      "grad_norm": 4.008997099121279,
+      "learning_rate": 1.9890896817975145e-05,
+      "loss": 1.9554,
+      "step": 441
+    },
+    {
+      "epoch": 0.07585378410846061,
+      "grad_norm": 2.965087523211052,
+      "learning_rate": 1.989007646035102e-05,
+      "loss": 1.6524,
+      "step": 442
+    },
+    {
+      "epoch": 0.0760253990046336,
+      "grad_norm": 2.945400897183803,
+      "learning_rate": 1.9889253047136913e-05,
+      "loss": 2.0752,
+      "step": 443
+    },
+    {
+      "epoch": 0.07619701390080659,
+      "grad_norm": 3.5975580699086382,
+      "learning_rate": 1.988842657858723e-05,
+      "loss": 2.0144,
+      "step": 444
+    },
+    {
+      "epoch": 0.07636862879697957,
+      "grad_norm": 3.741142530595054,
+      "learning_rate": 1.9887597054957304e-05,
+      "loss": 1.9509,
+      "step": 445
+    },
+    {
+      "epoch": 0.07654024369315257,
+      "grad_norm": 2.6574967818936566,
+      "learning_rate": 1.9886764476503425e-05,
+      "loss": 1.7284,
+      "step": 446
+    },
+    {
+      "epoch": 0.07671185858932555,
+      "grad_norm": 2.819419196217178,
+      "learning_rate": 1.9885928843482824e-05,
+      "loss": 1.9558,
+      "step": 447
+    },
+    {
+      "epoch": 0.07688347348549854,
+      "grad_norm": 5.046569462324691,
+      "learning_rate": 1.988509015615367e-05,
+      "loss": 1.8855,
+      "step": 448
+    },
+    {
+      "epoch": 0.07705508838167154,
+      "grad_norm": 2.408377661475034,
+      "learning_rate": 1.9884248414775083e-05,
+      "loss": 1.8529,
+      "step": 449
+    },
+    {
+      "epoch": 0.07722670327784452,
+      "grad_norm": 4.936975052755598,
+      "learning_rate": 1.9883403619607123e-05,
+      "loss": 2.1046,
+      "step": 450
+    },
+    {
+      "epoch": 0.0773983181740175,
+      "grad_norm": 3.1958831225716393,
+      "learning_rate": 1.9882555770910793e-05,
+      "loss": 1.7649,
+      "step": 451
+    },
+    {
+      "epoch": 0.07756993307019049,
+      "grad_norm": 2.6807885905387816,
+      "learning_rate": 1.9881704868948035e-05,
+      "loss": 1.8609,
+      "step": 452
+    },
+    {
+      "epoch": 0.07774154796636348,
+      "grad_norm": 10.419998758920823,
+      "learning_rate": 1.9880850913981748e-05,
+      "loss": 1.958,
+      "step": 453
+    },
+    {
+      "epoch": 0.07791316286253647,
+      "grad_norm": 3.8775785410147225,
+      "learning_rate": 1.9879993906275767e-05,
+      "loss": 2.065,
+      "step": 454
+    },
+    {
+      "epoch": 0.07808477775870945,
+      "grad_norm": 2.849534708153432,
+      "learning_rate": 1.9879133846094856e-05,
+      "loss": 1.8052,
+      "step": 455
+    },
+    {
+      "epoch": 0.07825639265488245,
+      "grad_norm": 3.5578531827828814,
+      "learning_rate": 1.987827073370475e-05,
+      "loss": 1.8499,
+      "step": 456
+    },
+    {
+      "epoch": 0.07842800755105543,
+      "grad_norm": 2.5982057469746147,
+      "learning_rate": 1.98774045693721e-05,
+      "loss": 1.7817,
+      "step": 457
+    },
+    {
+      "epoch": 0.07859962244722841,
+      "grad_norm": 3.5287356306423914,
+      "learning_rate": 1.987653535336452e-05,
+      "loss": 2.0148,
+      "step": 458
+    },
+    {
+      "epoch": 0.07877123734340141,
+      "grad_norm": 2.7055000404778546,
+      "learning_rate": 1.9875663085950556e-05,
+      "loss": 1.7614,
+      "step": 459
+    },
+    {
+      "epoch": 0.0789428522395744,
+      "grad_norm": 2.805128544342488,
+      "learning_rate": 1.9874787767399703e-05,
+      "loss": 2.061,
+      "step": 460
+    },
+    {
+      "epoch": 0.07911446713574738,
+      "grad_norm": 2.7474189241133495,
+      "learning_rate": 1.9873909397982385e-05,
+      "loss": 1.7367,
+      "step": 461
+    },
+    {
+      "epoch": 0.07928608203192038,
+      "grad_norm": 3.7147745663511347,
+      "learning_rate": 1.987302797796999e-05,
+      "loss": 1.9826,
+      "step": 462
+    },
+    {
+      "epoch": 0.07945769692809336,
+      "grad_norm": 4.23555155133059,
+      "learning_rate": 1.987214350763483e-05,
+      "loss": 2.182,
+      "step": 463
+    },
+    {
+      "epoch": 0.07962931182426634,
+      "grad_norm": 2.9681099902927888,
+      "learning_rate": 1.9871255987250172e-05,
+      "loss": 1.8332,
+      "step": 464
+    },
+    {
+      "epoch": 0.07980092672043933,
+      "grad_norm": 2.644610422448092,
+      "learning_rate": 1.9870365417090217e-05,
+      "loss": 1.9012,
+      "step": 465
+    },
+    {
+      "epoch": 0.07997254161661232,
+      "grad_norm": 3.085899852043181,
+      "learning_rate": 1.9869471797430112e-05,
+      "loss": 1.8466,
+      "step": 466
+    },
+    {
+      "epoch": 0.08014415651278531,
+      "grad_norm": 2.718756855467533,
+      "learning_rate": 1.986857512854594e-05,
+      "loss": 2.1873,
+      "step": 467
+    },
+    {
+      "epoch": 0.08031577140895829,
+      "grad_norm": 2.822960592290468,
+      "learning_rate": 1.9867675410714744e-05,
+      "loss": 2.2561,
+      "step": 468
+    },
+    {
+      "epoch": 0.08048738630513129,
+      "grad_norm": 2.7080760757881848,
+      "learning_rate": 1.9866772644214484e-05,
+      "loss": 1.8981,
+      "step": 469
+    },
+    {
+      "epoch": 0.08065900120130427,
+      "grad_norm": 2.8606334579442803,
+      "learning_rate": 1.986586682932408e-05,
+      "loss": 1.8657,
+      "step": 470
+    },
+    {
+      "epoch": 0.08083061609747726,
+      "grad_norm": 2.9320038272501243,
+      "learning_rate": 1.986495796632339e-05,
+      "loss": 2.1208,
+      "step": 471
+    },
+    {
+      "epoch": 0.08100223099365025,
+      "grad_norm": 3.3310414941403983,
+      "learning_rate": 1.9864046055493205e-05,
+      "loss": 1.8079,
+      "step": 472
+    },
+    {
+      "epoch": 0.08117384588982324,
+      "grad_norm": 4.173030541213135,
+      "learning_rate": 1.9863131097115273e-05,
+      "loss": 2.1978,
+      "step": 473
+    },
+    {
+      "epoch": 0.08134546078599622,
+      "grad_norm": 3.2631957690217166,
+      "learning_rate": 1.9862213091472267e-05,
+      "loss": 1.6694,
+      "step": 474
+    },
+    {
+      "epoch": 0.08151707568216922,
+      "grad_norm": 3.6549279485111072,
+      "learning_rate": 1.9861292038847818e-05,
+      "loss": 2.1436,
+      "step": 475
+    },
+    {
+      "epoch": 0.0816886905783422,
+      "grad_norm": 3.4579723786448406,
+      "learning_rate": 1.9860367939526482e-05,
+      "loss": 2.0133,
+      "step": 476
+    },
+    {
+      "epoch": 0.08186030547451519,
+      "grad_norm": 3.502710560745024,
+      "learning_rate": 1.9859440793793766e-05,
+      "loss": 1.9742,
+      "step": 477
+    },
+    {
+      "epoch": 0.08203192037068817,
+      "grad_norm": 3.544396288978015,
+      "learning_rate": 1.9858510601936124e-05,
+      "loss": 2.1805,
+      "step": 478
+    },
+    {
+      "epoch": 0.08220353526686117,
+      "grad_norm": 3.443255565471956,
+      "learning_rate": 1.9857577364240933e-05,
+      "loss": 2.0108,
+      "step": 479
+    },
+    {
+      "epoch": 0.08237515016303415,
+      "grad_norm": 2.810740942597221,
+      "learning_rate": 1.9856641080996527e-05,
+      "loss": 2.1126,
+      "step": 480
+    },
+    {
+      "epoch": 0.08254676505920713,
+      "grad_norm": 2.6304720077998645,
+      "learning_rate": 1.9855701752492174e-05,
+      "loss": 1.9508,
+      "step": 481
+    },
+    {
+      "epoch": 0.08271837995538013,
+      "grad_norm": 3.4067239352352443,
+      "learning_rate": 1.985475937901809e-05,
+      "loss": 2.0391,
+      "step": 482
+    },
+    {
+      "epoch": 0.08288999485155311,
+      "grad_norm": 2.7479962805697555,
+      "learning_rate": 1.985381396086542e-05,
+      "loss": 1.7972,
+      "step": 483
+    },
+    {
+      "epoch": 0.0830616097477261,
+      "grad_norm": 3.4900530274096995,
+      "learning_rate": 1.9852865498326253e-05,
+      "loss": 1.961,
+      "step": 484
+    },
+    {
+      "epoch": 0.0832332246438991,
+      "grad_norm": 2.5491101161030776,
+      "learning_rate": 1.9851913991693626e-05,
+      "loss": 2.1167,
+      "step": 485
+    },
+    {
+      "epoch": 0.08340483954007208,
+      "grad_norm": 2.924649694838515,
+      "learning_rate": 1.9850959441261514e-05,
+      "loss": 1.8073,
+      "step": 486
+    },
+    {
+      "epoch": 0.08357645443624506,
+      "grad_norm": 3.536472957231003,
+      "learning_rate": 1.9850001847324825e-05,
+      "loss": 1.9382,
+      "step": 487
+    },
+    {
+      "epoch": 0.08374806933241806,
+      "grad_norm": 3.054182657170999,
+      "learning_rate": 1.984904121017942e-05,
+      "loss": 2.0425,
+      "step": 488
+    },
+    {
+      "epoch": 0.08391968422859104,
+      "grad_norm": 2.7927409849462945,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 2.0228,
+      "step": 489
+    },
+    {
+      "epoch": 0.08409129912476403,
+      "grad_norm": 2.57527649206855,
+      "learning_rate": 1.9847110807450555e-05,
+      "loss": 1.8985,
+      "step": 490
+    },
+    {
+      "epoch": 0.08426291402093702,
+      "grad_norm": 3.4586774197398036,
+      "learning_rate": 1.9846141042463505e-05,
+      "loss": 1.7976,
+      "step": 491
+    },
+    {
+      "epoch": 0.08443452891711001,
+      "grad_norm": 2.7668380943368076,
+      "learning_rate": 1.9845168235460553e-05,
+      "loss": 1.9634,
+      "step": 492
+    },
+    {
+      "epoch": 0.08460614381328299,
+      "grad_norm": 3.4242371710848163,
+      "learning_rate": 1.984419238674225e-05,
+      "loss": 2.0589,
+      "step": 493
+    },
+    {
+      "epoch": 0.08477775870945597,
+      "grad_norm": 2.7113501894419514,
+      "learning_rate": 1.9843213496610086e-05,
+      "loss": 1.964,
+      "step": 494
+    },
+    {
+      "epoch": 0.08494937360562897,
+      "grad_norm": 3.058832849825258,
+      "learning_rate": 1.98422315653665e-05,
+      "loss": 1.7989,
+      "step": 495
+    },
+    {
+      "epoch": 0.08512098850180196,
+      "grad_norm": 4.080728644514222,
+      "learning_rate": 1.984124659331486e-05,
+      "loss": 1.9278,
+      "step": 496
+    },
+    {
+      "epoch": 0.08529260339797494,
+      "grad_norm": 3.0226243545744182,
+      "learning_rate": 1.984025858075948e-05,
+      "loss": 1.8404,
+      "step": 497
+    },
+    {
+      "epoch": 0.08546421829414794,
+      "grad_norm": 3.370632563740033,
+      "learning_rate": 1.983926752800561e-05,
+      "loss": 1.9167,
+      "step": 498
+    },
+    {
+      "epoch": 0.08563583319032092,
+      "grad_norm": 3.978424419074891,
+      "learning_rate": 1.9838273435359447e-05,
+      "loss": 1.9387,
+      "step": 499
+    },
+    {
+      "epoch": 0.0858074480864939,
+      "grad_norm": 3.9857709854862575,
+      "learning_rate": 1.983727630312811e-05,
+      "loss": 1.9806,
+      "step": 500
+    },
+    {
+      "epoch": 0.0859790629826669,
+      "grad_norm": 3.078979700594615,
+      "learning_rate": 1.983627613161968e-05,
+      "loss": 2.0171,
+      "step": 501
+    },
+    {
+      "epoch": 0.08615067787883988,
+      "grad_norm": 2.5471736174880104,
+      "learning_rate": 1.9835272921143156e-05,
+      "loss": 2.1804,
+      "step": 502
+    },
+    {
+      "epoch": 0.08632229277501287,
+      "grad_norm": 3.4829829985310585,
+      "learning_rate": 1.983426667200849e-05,
+      "loss": 2.0033,
+      "step": 503
+    },
+    {
+      "epoch": 0.08649390767118587,
+      "grad_norm": 2.95167871963179,
+      "learning_rate": 1.983325738452657e-05,
+      "loss": 2.0968,
+      "step": 504
+    },
+    {
+      "epoch": 0.08666552256735885,
+      "grad_norm": 3.360596300617668,
+      "learning_rate": 1.983224505900921e-05,
+      "loss": 2.0263,
+      "step": 505
+    },
+    {
+      "epoch": 0.08683713746353183,
+      "grad_norm": 8.967978708069104,
+      "learning_rate": 1.9831229695769184e-05,
+      "loss": 2.1001,
+      "step": 506
+    },
+    {
+      "epoch": 0.08700875235970482,
+      "grad_norm": 2.5752979411092767,
+      "learning_rate": 1.983021129512019e-05,
+      "loss": 1.8587,
+      "step": 507
+    },
+    {
+      "epoch": 0.08718036725587781,
+      "grad_norm": 2.657936696316239,
+      "learning_rate": 1.982918985737687e-05,
+      "loss": 1.8626,
+      "step": 508
+    },
+    {
+      "epoch": 0.0873519821520508,
+      "grad_norm": 4.0533900902450855,
+      "learning_rate": 1.9828165382854798e-05,
+      "loss": 1.9509,
+      "step": 509
+    },
+    {
+      "epoch": 0.08752359704822378,
+      "grad_norm": 4.45646664322224,
+      "learning_rate": 1.9827137871870494e-05,
+      "loss": 1.9707,
+      "step": 510
+    },
+    {
+      "epoch": 0.08769521194439678,
+      "grad_norm": 2.8055586834923765,
+      "learning_rate": 1.9826107324741413e-05,
+      "loss": 1.6581,
+      "step": 511
+    },
+    {
+      "epoch": 0.08786682684056976,
+      "grad_norm": 3.013466058812041,
+      "learning_rate": 1.9825073741785946e-05,
+      "loss": 2.0867,
+      "step": 512
+    },
+    {
+      "epoch": 0.08803844173674275,
+      "grad_norm": 2.760088907057879,
+      "learning_rate": 1.9824037123323423e-05,
+      "loss": 2.1439,
+      "step": 513
+    },
+    {
+      "epoch": 0.08821005663291574,
+      "grad_norm": 3.117211945576528,
+      "learning_rate": 1.9822997469674116e-05,
+      "loss": 1.9068,
+      "step": 514
+    },
+    {
+      "epoch": 0.08838167152908873,
+      "grad_norm": 3.3500210513213684,
+      "learning_rate": 1.9821954781159224e-05,
+      "loss": 1.7608,
+      "step": 515
+    },
+    {
+      "epoch": 0.08855328642526171,
+      "grad_norm": 3.0002791868067815,
+      "learning_rate": 1.9820909058100904e-05,
+      "loss": 1.8047,
+      "step": 516
+    },
+    {
+      "epoch": 0.08872490132143471,
+      "grad_norm": 2.51459532589054,
+      "learning_rate": 1.9819860300822227e-05,
+      "loss": 1.9093,
+      "step": 517
+    },
+    {
+      "epoch": 0.08889651621760769,
+      "grad_norm": 8.390605370624675,
+      "learning_rate": 1.9818808509647215e-05,
+      "loss": 2.1025,
+      "step": 518
+    },
+    {
+      "epoch": 0.08906813111378067,
+      "grad_norm": 3.9247418992069014,
+      "learning_rate": 1.9817753684900826e-05,
+      "loss": 1.8867,
+      "step": 519
+    },
+    {
+      "epoch": 0.08923974600995366,
+      "grad_norm": 3.067543628105918,
+      "learning_rate": 1.9816695826908948e-05,
+      "loss": 1.7059,
+      "step": 520
+    },
+    {
+      "epoch": 0.08941136090612666,
+      "grad_norm": 3.0720250388049624,
+      "learning_rate": 1.9815634935998417e-05,
+      "loss": 1.9277,
+      "step": 521
+    },
+    {
+      "epoch": 0.08958297580229964,
+      "grad_norm": 4.512210623445909,
+      "learning_rate": 1.9814571012497e-05,
+      "loss": 2.1153,
+      "step": 522
+    },
+    {
+      "epoch": 0.08975459069847262,
+      "grad_norm": 3.3849956847818654,
+      "learning_rate": 1.9813504056733403e-05,
+      "loss": 1.9046,
+      "step": 523
+    },
+    {
+      "epoch": 0.08992620559464562,
+      "grad_norm": 2.975847294834541,
+      "learning_rate": 1.981243406903726e-05,
+      "loss": 2.0535,
+      "step": 524
+    },
+    {
+      "epoch": 0.0900978204908186,
+      "grad_norm": 5.271145305845035,
+      "learning_rate": 1.981136104973916e-05,
+      "loss": 2.0219,
+      "step": 525
+    },
+    {
+      "epoch": 0.09026943538699159,
+      "grad_norm": 3.6965403948842974,
+      "learning_rate": 1.981028499917061e-05,
+      "loss": 1.8159,
+      "step": 526
+    },
+    {
+      "epoch": 0.09044105028316458,
+      "grad_norm": 3.370738627265268,
+      "learning_rate": 1.9809205917664064e-05,
+      "loss": 1.7585,
+      "step": 527
+    },
+    {
+      "epoch": 0.09061266517933757,
+      "grad_norm": 3.3626169915707336,
+      "learning_rate": 1.9808123805552908e-05,
+      "loss": 1.7938,
+      "step": 528
+    },
+    {
+      "epoch": 0.09078428007551055,
+      "grad_norm": 4.035214519495458,
+      "learning_rate": 1.980703866317147e-05,
+      "loss": 2.1773,
+      "step": 529
+    },
+    {
+      "epoch": 0.09095589497168355,
+      "grad_norm": 7.440776600318674,
+      "learning_rate": 1.9805950490855003e-05,
+      "loss": 1.8127,
+      "step": 530
+    },
+    {
+      "epoch": 0.09112750986785653,
+      "grad_norm": 3.4549605133639236,
+      "learning_rate": 1.980485928893971e-05,
+      "loss": 2.0026,
+      "step": 531
+    },
+    {
+      "epoch": 0.09129912476402952,
+      "grad_norm": 5.6791317707563325,
+      "learning_rate": 1.9803765057762723e-05,
+      "loss": 1.8348,
+      "step": 532
+    },
+    {
+      "epoch": 0.0914707396602025,
+      "grad_norm": 3.2830422913475052,
+      "learning_rate": 1.9802667797662113e-05,
+      "loss": 1.53,
+      "step": 533
+    },
+    {
+      "epoch": 0.0916423545563755,
+      "grad_norm": 3.356830892760497,
+      "learning_rate": 1.9801567508976874e-05,
+      "loss": 1.8351,
+      "step": 534
+    },
+    {
+      "epoch": 0.09181396945254848,
+      "grad_norm": 4.870122528034997,
+      "learning_rate": 1.9800464192046956e-05,
+      "loss": 2.1619,
+      "step": 535
+    },
+    {
+      "epoch": 0.09198558434872146,
+      "grad_norm": 2.9268404406739297,
+      "learning_rate": 1.979935784721323e-05,
+      "loss": 1.8741,
+      "step": 536
+    },
+    {
+      "epoch": 0.09215719924489446,
+      "grad_norm": 3.8006197984773853,
+      "learning_rate": 1.9798248474817506e-05,
+      "loss": 1.7688,
+      "step": 537
+    },
+    {
+      "epoch": 0.09232881414106744,
+      "grad_norm": 3.1193126068672266,
+      "learning_rate": 1.9797136075202532e-05,
+      "loss": 2.1704,
+      "step": 538
+    },
+    {
+      "epoch": 0.09250042903724043,
+      "grad_norm": 3.724141320543389,
+      "learning_rate": 1.979602064871199e-05,
+      "loss": 2.056,
+      "step": 539
+    },
+    {
+      "epoch": 0.09267204393341343,
+      "grad_norm": 4.097581038475727,
+      "learning_rate": 1.9794902195690494e-05,
+      "loss": 2.0488,
+      "step": 540
+    },
+    {
+      "epoch": 0.09284365882958641,
+      "grad_norm": 2.577689561051121,
+      "learning_rate": 1.97937807164836e-05,
+      "loss": 1.9698,
+      "step": 541
+    },
+    {
+      "epoch": 0.09301527372575939,
+      "grad_norm": 3.311870050584975,
+      "learning_rate": 1.979265621143779e-05,
+      "loss": 1.7667,
+      "step": 542
+    },
+    {
+      "epoch": 0.09318688862193239,
+      "grad_norm": 3.7420977288530444,
+      "learning_rate": 1.9791528680900494e-05,
+      "loss": 1.8669,
+      "step": 543
+    },
+    {
+      "epoch": 0.09335850351810537,
+      "grad_norm": 3.128448264247861,
+      "learning_rate": 1.9790398125220056e-05,
+      "loss": 2.1067,
+      "step": 544
+    },
+    {
+      "epoch": 0.09353011841427836,
+      "grad_norm": 2.8846794252262726,
+      "learning_rate": 1.9789264544745776e-05,
+      "loss": 1.749,
+      "step": 545
+    },
+    {
+      "epoch": 0.09370173331045134,
+      "grad_norm": 3.151280219152059,
+      "learning_rate": 1.978812793982788e-05,
+      "loss": 2.0316,
+      "step": 546
+    },
+    {
+      "epoch": 0.09387334820662434,
+      "grad_norm": 3.0588625058884094,
+      "learning_rate": 1.9786988310817523e-05,
+      "loss": 2.0002,
+      "step": 547
+    },
+    {
+      "epoch": 0.09404496310279732,
+      "grad_norm": 3.19267554194025,
+      "learning_rate": 1.97858456580668e-05,
+      "loss": 1.7847,
+      "step": 548
+    },
+    {
+      "epoch": 0.0942165779989703,
+      "grad_norm": 4.359375800283079,
+      "learning_rate": 1.9784699981928742e-05,
+      "loss": 2.0769,
+      "step": 549
+    },
+    {
+      "epoch": 0.0943881928951433,
+      "grad_norm": 3.7311402607739437,
+      "learning_rate": 1.9783551282757312e-05,
+      "loss": 1.9001,
+      "step": 550
+    },
+    {
+      "epoch": 0.09455980779131629,
+      "grad_norm": 4.065594334215484,
+      "learning_rate": 1.9782399560907402e-05,
+      "loss": 1.747,
+      "step": 551
+    },
+    {
+      "epoch": 0.09473142268748927,
+      "grad_norm": 3.8955969320944206,
+      "learning_rate": 1.9781244816734846e-05,
+      "loss": 1.974,
+      "step": 552
+    },
+    {
+      "epoch": 0.09490303758366227,
+      "grad_norm": 5.863358963140666,
+      "learning_rate": 1.9780087050596407e-05,
+      "loss": 1.9237,
+      "step": 553
+    },
+    {
+      "epoch": 0.09507465247983525,
+      "grad_norm": 5.016966928614574,
+      "learning_rate": 1.9778926262849786e-05,
+      "loss": 1.8072,
+      "step": 554
+    },
+    {
+      "epoch": 0.09524626737600823,
+      "grad_norm": 4.709542800475869,
+      "learning_rate": 1.9777762453853607e-05,
+      "loss": 1.7808,
+      "step": 555
+    },
+    {
+      "epoch": 0.09541788227218123,
+      "grad_norm": 3.6423757917494117,
+      "learning_rate": 1.9776595623967443e-05,
+      "loss": 1.7195,
+      "step": 556
+    },
+    {
+      "epoch": 0.09558949716835421,
+      "grad_norm": 3.6524386217677813,
+      "learning_rate": 1.9775425773551788e-05,
+      "loss": 1.6905,
+      "step": 557
+    },
+    {
+      "epoch": 0.0957611120645272,
+      "grad_norm": 4.827821550933328,
+      "learning_rate": 1.977425290296807e-05,
+      "loss": 1.9539,
+      "step": 558
+    },
+    {
+      "epoch": 0.09593272696070018,
+      "grad_norm": 5.935373963702225,
+      "learning_rate": 1.977307701257866e-05,
+      "loss": 1.8562,
+      "step": 559
+    },
+    {
+      "epoch": 0.09610434185687318,
+      "grad_norm": 3.373470791522794,
+      "learning_rate": 1.9771898102746852e-05,
+      "loss": 1.9662,
+      "step": 560
+    },
+    {
+      "epoch": 0.09627595675304616,
+      "grad_norm": 3.2249559209608067,
+      "learning_rate": 1.9770716173836877e-05,
+      "loss": 2.1233,
+      "step": 561
+    },
+    {
+      "epoch": 0.09644757164921915,
+      "grad_norm": 2.8806275169368702,
+      "learning_rate": 1.9769531226213897e-05,
+      "loss": 1.9166,
+      "step": 562
+    },
+    {
+      "epoch": 0.09661918654539214,
+      "grad_norm": 2.846031982418901,
+      "learning_rate": 1.9768343260244008e-05,
+      "loss": 2.1811,
+      "step": 563
+    },
+    {
+      "epoch": 0.09679080144156513,
+      "grad_norm": 2.6801343098220665,
+      "learning_rate": 1.9767152276294238e-05,
+      "loss": 1.9713,
+      "step": 564
+    },
+    {
+      "epoch": 0.09696241633773811,
+      "grad_norm": 3.3555235026297776,
+      "learning_rate": 1.976595827473255e-05,
+      "loss": 1.8181,
+      "step": 565
+    },
+    {
+      "epoch": 0.09713403123391111,
+      "grad_norm": 2.7266374781964697,
+      "learning_rate": 1.9764761255927838e-05,
+      "loss": 1.9583,
+      "step": 566
+    },
+    {
+      "epoch": 0.09730564613008409,
+      "grad_norm": 2.921284455513616,
+      "learning_rate": 1.976356122024992e-05,
+      "loss": 1.9281,
+      "step": 567
+    },
+    {
+      "epoch": 0.09747726102625708,
+      "grad_norm": 6.026462412464931,
+      "learning_rate": 1.976235816806956e-05,
+      "loss": 1.9916,
+      "step": 568
+    },
+    {
+      "epoch": 0.09764887592243007,
+      "grad_norm": 3.908176239370151,
+      "learning_rate": 1.976115209975844e-05,
+      "loss": 1.9847,
+      "step": 569
+    },
+    {
+      "epoch": 0.09782049081860306,
+      "grad_norm": 3.741167402922778,
+      "learning_rate": 1.975994301568919e-05,
+      "loss": 2.0098,
+      "step": 570
+    },
+    {
+      "epoch": 0.09799210571477604,
+      "grad_norm": 3.248760159620241,
+      "learning_rate": 1.9758730916235356e-05,
+      "loss": 1.9698,
+      "step": 571
+    },
+    {
+      "epoch": 0.09816372061094902,
+      "grad_norm": 2.840623470202775,
+      "learning_rate": 1.9757515801771426e-05,
+      "loss": 1.9504,
+      "step": 572
+    },
+    {
+      "epoch": 0.09833533550712202,
+      "grad_norm": 4.6024284522510825,
+      "learning_rate": 1.9756297672672812e-05,
+      "loss": 1.9659,
+      "step": 573
+    },
+    {
+      "epoch": 0.098506950403295,
+      "grad_norm": 2.8757555299447914,
+      "learning_rate": 1.9755076529315867e-05,
+      "loss": 1.9931,
+      "step": 574
+    },
+    {
+      "epoch": 0.09867856529946799,
+      "grad_norm": 2.8663417701819576,
+      "learning_rate": 1.9753852372077863e-05,
+      "loss": 2.0503,
+      "step": 575
+    },
+    {
+      "epoch": 0.09885018019564099,
+      "grad_norm": 3.2065741772887955,
+      "learning_rate": 1.9752625201337017e-05,
+      "loss": 2.1571,
+      "step": 576
+    },
+    {
+      "epoch": 0.09902179509181397,
+      "grad_norm": 3.8679304110141564,
+      "learning_rate": 1.9751395017472462e-05,
+      "loss": 1.7899,
+      "step": 577
+    },
+    {
+      "epoch": 0.09919340998798695,
+      "grad_norm": 3.746094886403712,
+      "learning_rate": 1.9750161820864274e-05,
+      "loss": 2.0576,
+      "step": 578
+    },
+    {
+      "epoch": 0.09936502488415995,
+      "grad_norm": 3.0513299621498415,
+      "learning_rate": 1.974892561189346e-05,
+      "loss": 1.8414,
+      "step": 579
+    },
+    {
+      "epoch": 0.09953663978033293,
+      "grad_norm": 2.5881152892209793,
+      "learning_rate": 1.9747686390941944e-05,
+      "loss": 1.5498,
+      "step": 580
+    },
+    {
+      "epoch": 0.09970825467650592,
+      "grad_norm": 3.594256534330549,
+      "learning_rate": 1.974644415839259e-05,
+      "loss": 2.0439,
+      "step": 581
+    },
+    {
+      "epoch": 0.09987986957267891,
+      "grad_norm": 2.690148745647811,
+      "learning_rate": 1.97451989146292e-05,
+      "loss": 1.976,
+      "step": 582
+    },
+    {
+      "epoch": 0.1000514844688519,
+      "grad_norm": 3.0016291317241035,
+      "learning_rate": 1.974395066003649e-05,
+      "loss": 1.8832,
+      "step": 583
+    },
+    {
+      "epoch": 0.10022309936502488,
+      "grad_norm": 3.548342018027038,
+      "learning_rate": 1.974269939500012e-05,
+      "loss": 2.0379,
+      "step": 584
+    },
+    {
+      "epoch": 0.10039471426119788,
+      "grad_norm": 3.6550894878737146,
+      "learning_rate": 1.974144511990668e-05,
+      "loss": 2.0104,
+      "step": 585
+    },
+    {
+      "epoch": 0.10056632915737086,
+      "grad_norm": 3.245162718313138,
+      "learning_rate": 1.9740187835143673e-05,
+      "loss": 1.9786,
+      "step": 586
+    },
+    {
+      "epoch": 0.10073794405354385,
+      "grad_norm": 4.968097520781465,
+      "learning_rate": 1.973892754109955e-05,
+      "loss": 2.0406,
+      "step": 587
+    },
+    {
+      "epoch": 0.10090955894971683,
+      "grad_norm": 3.551799572281438,
+      "learning_rate": 1.9737664238163683e-05,
+      "loss": 1.9049,
+      "step": 588
+    },
+    {
+      "epoch": 0.10108117384588983,
+      "grad_norm": 7.569469865852626,
+      "learning_rate": 1.973639792672638e-05,
+      "loss": 2.0975,
+      "step": 589
+    },
+    {
+      "epoch": 0.10125278874206281,
+      "grad_norm": 5.167328577189202,
+      "learning_rate": 1.9735128607178868e-05,
+      "loss": 1.8706,
+      "step": 590
+    },
+    {
+      "epoch": 0.1014244036382358,
+      "grad_norm": 2.6708530788301115,
+      "learning_rate": 1.9733856279913314e-05,
+      "loss": 1.9269,
+      "step": 591
+    },
+    {
+      "epoch": 0.10159601853440879,
+      "grad_norm": 3.50025775398842,
+      "learning_rate": 1.9732580945322813e-05,
+      "loss": 1.8695,
+      "step": 592
+    },
+    {
+      "epoch": 0.10176763343058177,
+      "grad_norm": 3.17379544886861,
+      "learning_rate": 1.973130260380138e-05,
+      "loss": 1.7495,
+      "step": 593
+    },
+    {
+      "epoch": 0.10193924832675476,
+      "grad_norm": 4.612389180089776,
+      "learning_rate": 1.9730021255743967e-05,
+      "loss": 2.1471,
+      "step": 594
+    },
+    {
+      "epoch": 0.10211086322292776,
+      "grad_norm": 4.730339771370154,
+      "learning_rate": 1.9728736901546454e-05,
+      "loss": 2.026,
+      "step": 595
+    },
+    {
+      "epoch": 0.10228247811910074,
+      "grad_norm": 3.186485286061052,
+      "learning_rate": 1.972744954160565e-05,
+      "loss": 1.8984,
+      "step": 596
+    },
+    {
+      "epoch": 0.10245409301527372,
+      "grad_norm": 3.096466402518034,
+      "learning_rate": 1.9726159176319292e-05,
+      "loss": 1.7394,
+      "step": 597
+    },
+    {
+      "epoch": 0.10262570791144672,
+      "grad_norm": 3.2725057351977216,
+      "learning_rate": 1.972486580608604e-05,
+      "loss": 2.057,
+      "step": 598
+    },
+    {
+      "epoch": 0.1027973228076197,
+      "grad_norm": 3.3517565050028657,
+      "learning_rate": 1.972356943130549e-05,
+      "loss": 1.9072,
+      "step": 599
+    },
+    {
+      "epoch": 0.10296893770379269,
+      "grad_norm": 2.740538428486397,
+      "learning_rate": 1.9722270052378167e-05,
+      "loss": 1.5391,
+      "step": 600
+    },
+    {
+      "epoch": 0.10314055259996567,
+      "grad_norm": 2.947851109288251,
+      "learning_rate": 1.972096766970552e-05,
+      "loss": 2.2493,
+      "step": 601
+    },
+    {
+      "epoch": 0.10331216749613867,
+      "grad_norm": 3.4665930207183964,
+      "learning_rate": 1.971966228368992e-05,
+      "loss": 1.7292,
+      "step": 602
+    },
+    {
+      "epoch": 0.10348378239231165,
+      "grad_norm": 2.7646787832771547,
+      "learning_rate": 1.9718353894734677e-05,
+      "loss": 1.9337,
+      "step": 603
+    },
+    {
+      "epoch": 0.10365539728848464,
+      "grad_norm": 2.816464405924286,
+      "learning_rate": 1.9717042503244032e-05,
+      "loss": 1.8134,
+      "step": 604
+    },
+    {
+      "epoch": 0.10382701218465763,
+      "grad_norm": 2.793797413732475,
+      "learning_rate": 1.9715728109623134e-05,
+      "loss": 2.0687,
+      "step": 605
+    },
+    {
+      "epoch": 0.10399862708083062,
+      "grad_norm": 2.813962776958442,
+      "learning_rate": 1.971441071427808e-05,
+      "loss": 1.6167,
+      "step": 606
+    },
+    {
+      "epoch": 0.1041702419770036,
+      "grad_norm": 4.155914786093373,
+      "learning_rate": 1.9713090317615877e-05,
+      "loss": 2.1836,
+      "step": 607
+    },
+    {
+      "epoch": 0.1043418568731766,
+      "grad_norm": 3.6848162922317202,
+      "learning_rate": 1.971176692004448e-05,
+      "loss": 2.1002,
+      "step": 608
+    },
+    {
+      "epoch": 0.10451347176934958,
+      "grad_norm": 4.041843806508891,
+      "learning_rate": 1.971044052197275e-05,
+      "loss": 1.8046,
+      "step": 609
+    },
+    {
+      "epoch": 0.10468508666552256,
+      "grad_norm": 3.641295749603698,
+      "learning_rate": 1.970911112381049e-05,
+      "loss": 1.7766,
+      "step": 610
+    },
+    {
+      "epoch": 0.10485670156169556,
+      "grad_norm": 3.172220566704014,
+      "learning_rate": 1.9707778725968427e-05,
+      "loss": 1.7329,
+      "step": 611
+    },
+    {
+      "epoch": 0.10502831645786855,
+      "grad_norm": 4.395258415141472,
+      "learning_rate": 1.97064433288582e-05,
+      "loss": 1.9403,
+      "step": 612
+    },
+    {
+      "epoch": 0.10519993135404153,
+      "grad_norm": 3.439688326160458,
+      "learning_rate": 1.9705104932892398e-05,
+      "loss": 1.8274,
+      "step": 613
+    },
+    {
+      "epoch": 0.10537154625021451,
+      "grad_norm": 3.5392487781649966,
+      "learning_rate": 1.9703763538484525e-05,
+      "loss": 1.9361,
+      "step": 614
+    },
+    {
+      "epoch": 0.10554316114638751,
+      "grad_norm": 5.462091347217853,
+      "learning_rate": 1.9702419146049004e-05,
+      "loss": 2.1089,
+      "step": 615
+    },
+    {
+      "epoch": 0.1057147760425605,
+      "grad_norm": 3.408580501453271,
+      "learning_rate": 1.97010717560012e-05,
+      "loss": 1.9725,
+      "step": 616
+    },
+    {
+      "epoch": 0.10588639093873348,
+      "grad_norm": 3.013125385132649,
+      "learning_rate": 1.969972136875739e-05,
+      "loss": 1.9127,
+      "step": 617
+    },
+    {
+      "epoch": 0.10605800583490647,
+      "grad_norm": 4.047263351078423,
+      "learning_rate": 1.9698367984734786e-05,
+      "loss": 1.7079,
+      "step": 618
+    },
+    {
+      "epoch": 0.10622962073107946,
+      "grad_norm": 3.579643650818717,
+      "learning_rate": 1.9697011604351524e-05,
+      "loss": 1.9259,
+      "step": 619
+    },
+    {
+      "epoch": 0.10640123562725244,
+      "grad_norm": 3.216916581264715,
+      "learning_rate": 1.9695652228026664e-05,
+      "loss": 1.7663,
+      "step": 620
+    },
+    {
+      "epoch": 0.10657285052342544,
+      "grad_norm": 2.4013460522156334,
+      "learning_rate": 1.969428985618019e-05,
+      "loss": 1.5992,
+      "step": 621
+    },
+    {
+      "epoch": 0.10674446541959842,
+      "grad_norm": 3.536308684264547,
+      "learning_rate": 1.9692924489233015e-05,
+      "loss": 1.8546,
+      "step": 622
+    },
+    {
+      "epoch": 0.1069160803157714,
+      "grad_norm": 4.486968367615817,
+      "learning_rate": 1.969155612760698e-05,
+      "loss": 2.0518,
+      "step": 623
+    },
+    {
+      "epoch": 0.1070876952119444,
+      "grad_norm": 3.1146699884436604,
+      "learning_rate": 1.9690184771724843e-05,
+      "loss": 1.9719,
+      "step": 624
+    },
+    {
+      "epoch": 0.10725931010811739,
+      "grad_norm": 2.723565022988268,
+      "learning_rate": 1.968881042201029e-05,
+      "loss": 1.6757,
+      "step": 625
+    },
+    {
+      "epoch": 0.10743092500429037,
+      "grad_norm": 2.5503330278692276,
+      "learning_rate": 1.968743307888794e-05,
+      "loss": 1.8799,
+      "step": 626
+    },
+    {
+      "epoch": 0.10760253990046335,
+      "grad_norm": 2.5208950324474064,
+      "learning_rate": 1.9686052742783324e-05,
+      "loss": 1.9736,
+      "step": 627
+    },
+    {
+      "epoch": 0.10777415479663635,
+      "grad_norm": 2.7016050673369527,
+      "learning_rate": 1.9684669414122905e-05,
+      "loss": 1.7507,
+      "step": 628
+    },
+    {
+      "epoch": 0.10794576969280933,
+      "grad_norm": 4.3722822684553435,
+      "learning_rate": 1.968328309333407e-05,
+      "loss": 1.8855,
+      "step": 629
+    },
+    {
+      "epoch": 0.10811738458898232,
+      "grad_norm": 4.463669412847087,
+      "learning_rate": 1.9681893780845135e-05,
+      "loss": 1.6386,
+      "step": 630
+    },
+    {
+      "epoch": 0.10828899948515532,
+      "grad_norm": 5.27048101943173,
+      "learning_rate": 1.9680501477085326e-05,
+      "loss": 1.9182,
+      "step": 631
+    },
+    {
+      "epoch": 0.1084606143813283,
+      "grad_norm": 2.9816302262616774,
+      "learning_rate": 1.967910618248481e-05,
+      "loss": 1.8867,
+      "step": 632
+    },
+    {
+      "epoch": 0.10863222927750128,
+      "grad_norm": 4.1278007050382595,
+      "learning_rate": 1.9677707897474666e-05,
+      "loss": 2.2158,
+      "step": 633
+    },
+    {
+      "epoch": 0.10880384417367428,
+      "grad_norm": 2.7763390061635174,
+      "learning_rate": 1.9676306622486905e-05,
+      "loss": 1.8851,
+      "step": 634
+    },
+    {
+      "epoch": 0.10897545906984726,
+      "grad_norm": 2.6991119336027625,
+      "learning_rate": 1.9674902357954452e-05,
+      "loss": 1.873,
+      "step": 635
+    },
+    {
+      "epoch": 0.10914707396602025,
+      "grad_norm": 3.247815678741392,
+      "learning_rate": 1.9673495104311173e-05,
+      "loss": 2.0542,
+      "step": 636
+    },
+    {
+      "epoch": 0.10931868886219324,
+      "grad_norm": 4.928723626057146,
+      "learning_rate": 1.967208486199183e-05,
+      "loss": 2.044,
+      "step": 637
+    },
+    {
+      "epoch": 0.10949030375836623,
+      "grad_norm": 8.847684867762078,
+      "learning_rate": 1.967067163143214e-05,
+      "loss": 1.8528,
+      "step": 638
+    },
+    {
+      "epoch": 0.10966191865453921,
+      "grad_norm": 4.138800403458856,
+      "learning_rate": 1.966925541306872e-05,
+      "loss": 1.9944,
+      "step": 639
+    },
+    {
+      "epoch": 0.1098335335507122,
+      "grad_norm": 3.2290060979119546,
+      "learning_rate": 1.966783620733912e-05,
+      "loss": 1.7702,
+      "step": 640
+    },
+    {
+      "epoch": 0.11000514844688519,
+      "grad_norm": 3.5287648021841154,
+      "learning_rate": 1.966641401468181e-05,
+      "loss": 1.7717,
+      "step": 641
+    },
+    {
+      "epoch": 0.11017676334305818,
+      "grad_norm": 4.85279740952659,
+      "learning_rate": 1.9664988835536182e-05,
+      "loss": 1.9606,
+      "step": 642
+    },
+    {
+      "epoch": 0.11034837823923116,
+      "grad_norm": 3.252648610700815,
+      "learning_rate": 1.966356067034256e-05,
+      "loss": 2.1478,
+      "step": 643
+    },
+    {
+      "epoch": 0.11051999313540416,
+      "grad_norm": 2.4114083412323613,
+      "learning_rate": 1.966212951954217e-05,
+      "loss": 1.6045,
+      "step": 644
+    },
+    {
+      "epoch": 0.11069160803157714,
+      "grad_norm": 3.81809633012379,
+      "learning_rate": 1.9660695383577186e-05,
+      "loss": 1.8116,
+      "step": 645
+    },
+    {
+      "epoch": 0.11086322292775012,
+      "grad_norm": 4.1287708371918495,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 1.9988,
+      "step": 646
+    },
+    {
+      "epoch": 0.11103483782392312,
+      "grad_norm": 6.137133294763594,
+      "learning_rate": 1.9657818157926677e-05,
+      "loss": 2.1113,
+      "step": 647
+    },
+    {
+      "epoch": 0.1112064527200961,
+      "grad_norm": 2.6768078380956473,
+      "learning_rate": 1.965637506913008e-05,
+      "loss": 1.833,
+      "step": 648
+    },
+    {
+      "epoch": 0.11137806761626909,
+      "grad_norm": 2.9178277443453053,
+      "learning_rate": 1.9654928996946758e-05,
+      "loss": 1.8341,
+      "step": 649
+    },
+    {
+      "epoch": 0.11154968251244209,
+      "grad_norm": 4.2718234230700745,
+      "learning_rate": 1.9653479941823473e-05,
+      "loss": 2.0536,
+      "step": 650
+    },
+    {
+      "epoch": 0.11172129740861507,
+      "grad_norm": 3.0820747168150713,
+      "learning_rate": 1.9652027904207923e-05,
+      "loss": 1.9344,
+      "step": 651
+    },
+    {
+      "epoch": 0.11189291230478805,
+      "grad_norm": 2.892465972319917,
+      "learning_rate": 1.9650572884548715e-05,
+      "loss": 1.9195,
+      "step": 652
+    },
+    {
+      "epoch": 0.11206452720096104,
+      "grad_norm": 2.179473271607477,
+      "learning_rate": 1.9649114883295393e-05,
+      "loss": 1.6974,
+      "step": 653
+    },
+    {
+      "epoch": 0.11223614209713403,
+      "grad_norm": 2.8238338911288143,
+      "learning_rate": 1.9647653900898408e-05,
+      "loss": 1.8772,
+      "step": 654
+    },
+    {
+      "epoch": 0.11240775699330702,
+      "grad_norm": 2.9846740287773157,
+      "learning_rate": 1.9646189937809145e-05,
+      "loss": 1.8713,
+      "step": 655
+    },
+    {
+      "epoch": 0.11257937188948,
+      "grad_norm": 3.1163205886163605,
+      "learning_rate": 1.9644722994479898e-05,
+      "loss": 1.9222,
+      "step": 656
+    },
+    {
+      "epoch": 0.112750986785653,
+      "grad_norm": 3.6345679253974135,
+      "learning_rate": 1.9643253071363885e-05,
+      "loss": 2.0683,
+      "step": 657
+    },
+    {
+      "epoch": 0.11292260168182598,
+      "grad_norm": 3.6614704530939512,
+      "learning_rate": 1.9641780168915247e-05,
+      "loss": 2.0771,
+      "step": 658
+    },
+    {
+      "epoch": 0.11309421657799897,
+      "grad_norm": 3.057366967433169,
+      "learning_rate": 1.964030428758905e-05,
+      "loss": 1.9282,
+      "step": 659
+    },
+    {
+      "epoch": 0.11326583147417196,
+      "grad_norm": 5.2361901605509145,
+      "learning_rate": 1.963882542784127e-05,
+      "loss": 2.0172,
+      "step": 660
+    },
+    {
+      "epoch": 0.11343744637034495,
+      "grad_norm": 2.615700006965504,
+      "learning_rate": 1.963734359012881e-05,
+      "loss": 1.9016,
+      "step": 661
+    },
+    {
+      "epoch": 0.11360906126651793,
+      "grad_norm": 2.909416564121831,
+      "learning_rate": 1.9635858774909492e-05,
+      "loss": 1.6456,
+      "step": 662
+    },
+    {
+      "epoch": 0.11378067616269093,
+      "grad_norm": 3.9147334343482876,
+      "learning_rate": 1.9634370982642056e-05,
+      "loss": 2.0586,
+      "step": 663
+    },
+    {
+      "epoch": 0.11395229105886391,
+      "grad_norm": 3.2735393451851533,
+      "learning_rate": 1.9632880213786164e-05,
+      "loss": 2.086,
+      "step": 664
+    },
+    {
+      "epoch": 0.1141239059550369,
+      "grad_norm": 3.9843737550808327,
+      "learning_rate": 1.9631386468802397e-05,
+      "loss": 1.7187,
+      "step": 665
+    },
+    {
+      "epoch": 0.11429552085120989,
+      "grad_norm": 2.8896273668287393,
+      "learning_rate": 1.9629889748152252e-05,
+      "loss": 1.8384,
+      "step": 666
+    },
+    {
+      "epoch": 0.11446713574738288,
+      "grad_norm": 3.5645648642231187,
+      "learning_rate": 1.9628390052298155e-05,
+      "loss": 1.8703,
+      "step": 667
+    },
+    {
+      "epoch": 0.11463875064355586,
+      "grad_norm": 2.5079097656523603,
+      "learning_rate": 1.962688738170344e-05,
+      "loss": 1.9289,
+      "step": 668
+    },
+    {
+      "epoch": 0.11481036553972884,
+      "grad_norm": 2.8016352796236306,
+      "learning_rate": 1.9625381736832363e-05,
+      "loss": 1.986,
+      "step": 669
+    },
+    {
+      "epoch": 0.11498198043590184,
+      "grad_norm": 2.8181450705992424,
+      "learning_rate": 1.9623873118150105e-05,
+      "loss": 1.8151,
+      "step": 670
+    },
+    {
+      "epoch": 0.11515359533207482,
+      "grad_norm": 2.557212183770673,
+      "learning_rate": 1.962236152612276e-05,
+      "loss": 1.8604,
+      "step": 671
+    },
+    {
+      "epoch": 0.11532521022824781,
+      "grad_norm": 3.112788750888735,
+      "learning_rate": 1.9620846961217342e-05,
+      "loss": 1.6763,
+      "step": 672
+    },
+    {
+      "epoch": 0.1154968251244208,
+      "grad_norm": 3.583075138051105,
+      "learning_rate": 1.9619329423901782e-05,
+      "loss": 1.842,
+      "step": 673
+    },
+    {
+      "epoch": 0.11566844002059379,
+      "grad_norm": 3.0603181131811112,
+      "learning_rate": 1.9617808914644932e-05,
+      "loss": 2.0202,
+      "step": 674
+    },
+    {
+      "epoch": 0.11584005491676677,
+      "grad_norm": 2.7594121002432157,
+      "learning_rate": 1.9616285433916564e-05,
+      "loss": 1.9104,
+      "step": 675
+    },
+    {
+      "epoch": 0.11601166981293977,
+      "grad_norm": 2.0308915383609016,
+      "learning_rate": 1.961475898218736e-05,
+      "loss": 1.5745,
+      "step": 676
+    },
+    {
+      "epoch": 0.11618328470911275,
+      "grad_norm": 2.6653189468467957,
+      "learning_rate": 1.961322955992893e-05,
+      "loss": 1.904,
+      "step": 677
+    },
+    {
+      "epoch": 0.11635489960528574,
+      "grad_norm": 3.722748499584908,
+      "learning_rate": 1.9611697167613796e-05,
+      "loss": 1.7403,
+      "step": 678
+    },
+    {
+      "epoch": 0.11652651450145873,
+      "grad_norm": 3.4872868367061134,
+      "learning_rate": 1.96101618057154e-05,
+      "loss": 1.7692,
+      "step": 679
+    },
+    {
+      "epoch": 0.11669812939763172,
+      "grad_norm": 3.041586950031621,
+      "learning_rate": 1.960862347470809e-05,
+      "loss": 1.7355,
+      "step": 680
+    },
+    {
+      "epoch": 0.1168697442938047,
+      "grad_norm": 3.13573484165979,
+      "learning_rate": 1.9607082175067153e-05,
+      "loss": 1.9715,
+      "step": 681
+    },
+    {
+      "epoch": 0.11704135918997768,
+      "grad_norm": 2.702380065137708,
+      "learning_rate": 1.9605537907268776e-05,
+      "loss": 1.7196,
+      "step": 682
+    },
+    {
+      "epoch": 0.11721297408615068,
+      "grad_norm": 3.3994377792098733,
+      "learning_rate": 1.960399067179007e-05,
+      "loss": 1.8419,
+      "step": 683
+    },
+    {
+      "epoch": 0.11738458898232366,
+      "grad_norm": 3.5898581174795403,
+      "learning_rate": 1.9602440469109062e-05,
+      "loss": 1.8036,
+      "step": 684
+    },
+    {
+      "epoch": 0.11755620387849665,
+      "grad_norm": 2.549191167890007,
+      "learning_rate": 1.9600887299704694e-05,
+      "loss": 1.9944,
+      "step": 685
+    },
+    {
+      "epoch": 0.11772781877466965,
+      "grad_norm": 2.8067638997963087,
+      "learning_rate": 1.9599331164056823e-05,
+      "loss": 2.1468,
+      "step": 686
+    },
+    {
+      "epoch": 0.11789943367084263,
+      "grad_norm": 3.2503872066757644,
+      "learning_rate": 1.959777206264623e-05,
+      "loss": 1.7988,
+      "step": 687
+    },
+    {
+      "epoch": 0.11807104856701561,
+      "grad_norm": 2.5655492270826046,
+      "learning_rate": 1.9596209995954606e-05,
+      "loss": 1.9896,
+      "step": 688
+    },
+    {
+      "epoch": 0.11824266346318861,
+      "grad_norm": 3.4038373020167003,
+      "learning_rate": 1.9594644964464552e-05,
+      "loss": 1.8029,
+      "step": 689
+    },
+    {
+      "epoch": 0.1184142783593616,
+      "grad_norm": 2.2470274647188173,
+      "learning_rate": 1.9593076968659606e-05,
+      "loss": 1.8847,
+      "step": 690
+    },
+    {
+      "epoch": 0.11858589325553458,
+      "grad_norm": 3.3166196832228665,
+      "learning_rate": 1.95915060090242e-05,
+      "loss": 1.8632,
+      "step": 691
+    },
+    {
+      "epoch": 0.11875750815170757,
+      "grad_norm": 3.1911420202071628,
+      "learning_rate": 1.958993208604369e-05,
+      "loss": 2.1536,
+      "step": 692
+    },
+    {
+      "epoch": 0.11892912304788056,
+      "grad_norm": 3.610909895768703,
+      "learning_rate": 1.9588355200204356e-05,
+      "loss": 1.9012,
+      "step": 693
+    },
+    {
+      "epoch": 0.11910073794405354,
+      "grad_norm": 3.8160243655922845,
+      "learning_rate": 1.958677535199337e-05,
+      "loss": 1.8011,
+      "step": 694
+    },
+    {
+      "epoch": 0.11927235284022653,
+      "grad_norm": 2.980075594564947,
+      "learning_rate": 1.9585192541898847e-05,
+      "loss": 1.7993,
+      "step": 695
+    },
+    {
+      "epoch": 0.11944396773639952,
+      "grad_norm": 3.67289866574711,
+      "learning_rate": 1.9583606770409798e-05,
+      "loss": 1.8665,
+      "step": 696
+    },
+    {
+      "epoch": 0.1196155826325725,
+      "grad_norm": 3.260134076805343,
+      "learning_rate": 1.9582018038016156e-05,
+      "loss": 2.0675,
+      "step": 697
+    },
+    {
+      "epoch": 0.11978719752874549,
+      "grad_norm": 2.6784053949660467,
+      "learning_rate": 1.9580426345208773e-05,
+      "loss": 1.5103,
+      "step": 698
+    },
+    {
+      "epoch": 0.11995881242491849,
+      "grad_norm": 3.074552583094222,
+      "learning_rate": 1.95788316924794e-05,
+      "loss": 1.7144,
+      "step": 699
+    },
+    {
+      "epoch": 0.12013042732109147,
+      "grad_norm": 5.172743445453561,
+      "learning_rate": 1.9577234080320725e-05,
+      "loss": 1.8502,
+      "step": 700
+    },
+    {
+      "epoch": 0.12030204221726445,
+      "grad_norm": 3.056563009474087,
+      "learning_rate": 1.957563350922633e-05,
+      "loss": 1.8261,
+      "step": 701
+    },
+    {
+      "epoch": 0.12047365711343745,
+      "grad_norm": 3.1842263189341007,
+      "learning_rate": 1.957402997969073e-05,
+      "loss": 1.9883,
+      "step": 702
+    },
+    {
+      "epoch": 0.12064527200961044,
+      "grad_norm": 5.0630167880209065,
+      "learning_rate": 1.957242349220933e-05,
+      "loss": 1.6892,
+      "step": 703
+    },
+    {
+      "epoch": 0.12081688690578342,
+      "grad_norm": 3.0565241361852205,
+      "learning_rate": 1.9570814047278473e-05,
+      "loss": 1.7964,
+      "step": 704
+    },
+    {
+      "epoch": 0.12098850180195642,
+      "grad_norm": 2.785422063330768,
+      "learning_rate": 1.95692016453954e-05,
+      "loss": 1.7392,
+      "step": 705
+    },
+    {
+      "epoch": 0.1211601166981294,
+      "grad_norm": 2.849393011403411,
+      "learning_rate": 1.9567586287058274e-05,
+      "loss": 1.8318,
+      "step": 706
+    },
+    {
+      "epoch": 0.12133173159430238,
+      "grad_norm": 3.6018994122316577,
+      "learning_rate": 1.9565967972766164e-05,
+      "loss": 1.9907,
+      "step": 707
+    },
+    {
+      "epoch": 0.12150334649047537,
+      "grad_norm": 2.891249821928477,
+      "learning_rate": 1.956434670301906e-05,
+      "loss": 1.9956,
+      "step": 708
+    },
+    {
+      "epoch": 0.12167496138664836,
+      "grad_norm": 3.58828544806441,
+      "learning_rate": 1.956272247831786e-05,
+      "loss": 2.0249,
+      "step": 709
+    },
+    {
+      "epoch": 0.12184657628282135,
+      "grad_norm": 2.8488348227855935,
+      "learning_rate": 1.956109529916438e-05,
+      "loss": 1.9498,
+      "step": 710
+    },
+    {
+      "epoch": 0.12201819117899433,
+      "grad_norm": 2.8837550100220803,
+      "learning_rate": 1.9559465166061344e-05,
+      "loss": 1.6279,
+      "step": 711
+    },
+    {
+      "epoch": 0.12218980607516733,
+      "grad_norm": 3.320573283921507,
+      "learning_rate": 1.9557832079512387e-05,
+      "loss": 1.8533,
+      "step": 712
+    },
+    {
+      "epoch": 0.12236142097134031,
+      "grad_norm": 2.4239358301693277,
+      "learning_rate": 1.9556196040022063e-05,
+      "loss": 1.8177,
+      "step": 713
+    },
+    {
+      "epoch": 0.1225330358675133,
+      "grad_norm": 3.5073093120306496,
+      "learning_rate": 1.955455704809583e-05,
+      "loss": 1.8524,
+      "step": 714
+    },
+    {
+      "epoch": 0.1227046507636863,
+      "grad_norm": 3.072178814401796,
+      "learning_rate": 1.9552915104240067e-05,
+      "loss": 1.9913,
+      "step": 715
+    },
+    {
+      "epoch": 0.12287626565985928,
+      "grad_norm": 2.623549589919317,
+      "learning_rate": 1.955127020896206e-05,
+      "loss": 2.1435,
+      "step": 716
+    },
+    {
+      "epoch": 0.12304788055603226,
+      "grad_norm": 3.132919700835491,
+      "learning_rate": 1.954962236277001e-05,
+      "loss": 1.695,
+      "step": 717
+    },
+    {
+      "epoch": 0.12321949545220526,
+      "grad_norm": 2.913405218348578,
+      "learning_rate": 1.9547971566173025e-05,
+      "loss": 1.8305,
+      "step": 718
+    },
+    {
+      "epoch": 0.12339111034837824,
+      "grad_norm": 3.6656251904198123,
+      "learning_rate": 1.9546317819681128e-05,
+      "loss": 2.0613,
+      "step": 719
+    },
+    {
+      "epoch": 0.12356272524455122,
+      "grad_norm": 3.6722745723633623,
+      "learning_rate": 1.9544661123805245e-05,
+      "loss": 1.748,
+      "step": 720
+    },
+    {
+      "epoch": 0.12373434014072421,
+      "grad_norm": 2.828121884091384,
+      "learning_rate": 1.954300147905724e-05,
+      "loss": 1.9915,
+      "step": 721
+    },
+    {
+      "epoch": 0.1239059550368972,
+      "grad_norm": 2.8714512600803612,
+      "learning_rate": 1.9541338885949845e-05,
+      "loss": 1.848,
+      "step": 722
+    },
+    {
+      "epoch": 0.12407756993307019,
+      "grad_norm": 3.8996471448338035,
+      "learning_rate": 1.9539673344996744e-05,
+      "loss": 1.9279,
+      "step": 723
+    },
+    {
+      "epoch": 0.12424918482924317,
+      "grad_norm": 2.9536823996782346,
+      "learning_rate": 1.953800485671251e-05,
+      "loss": 1.7804,
+      "step": 724
+    },
+    {
+      "epoch": 0.12442079972541617,
+      "grad_norm": 4.7489942812156976,
+      "learning_rate": 1.9536333421612626e-05,
+      "loss": 1.7883,
+      "step": 725
+    },
+    {
+      "epoch": 0.12459241462158915,
+      "grad_norm": 3.2145611731803405,
+      "learning_rate": 1.9534659040213498e-05,
+      "loss": 1.8362,
+      "step": 726
+    },
+    {
+      "epoch": 0.12476402951776214,
+      "grad_norm": 4.454136603463675,
+      "learning_rate": 1.9532981713032432e-05,
+      "loss": 1.9277,
+      "step": 727
+    },
+    {
+      "epoch": 0.12493564441393513,
+      "grad_norm": 3.61009760856778,
+      "learning_rate": 1.953130144058764e-05,
+      "loss": 2.0101,
+      "step": 728
+    },
+    {
+      "epoch": 0.1251072593101081,
+      "grad_norm": 3.7323789053868346,
+      "learning_rate": 1.9529618223398258e-05,
+      "loss": 2.0167,
+      "step": 729
+    },
+    {
+      "epoch": 0.12527887420628112,
+      "grad_norm": 3.461211435335458,
+      "learning_rate": 1.9527932061984327e-05,
+      "loss": 1.9127,
+      "step": 730
+    },
+    {
+      "epoch": 0.1254504891024541,
+      "grad_norm": 2.976985533695183,
+      "learning_rate": 1.952624295686679e-05,
+      "loss": 1.694,
+      "step": 731
+    },
+    {
+      "epoch": 0.12562210399862708,
+      "grad_norm": 3.3575090676614496,
+      "learning_rate": 1.9524550908567505e-05,
+      "loss": 1.9081,
+      "step": 732
+    },
+    {
+      "epoch": 0.12579371889480007,
+      "grad_norm": 3.6069574862657485,
+      "learning_rate": 1.9522855917609243e-05,
+      "loss": 1.8363,
+      "step": 733
+    },
+    {
+      "epoch": 0.12596533379097305,
+      "grad_norm": 3.1836154152273406,
+      "learning_rate": 1.9521157984515672e-05,
+      "loss": 2.0397,
+      "step": 734
+    },
+    {
+      "epoch": 0.12613694868714603,
+      "grad_norm": 3.5354867259088816,
+      "learning_rate": 1.951945710981139e-05,
+      "loss": 1.9146,
+      "step": 735
+    },
+    {
+      "epoch": 0.12630856358331904,
+      "grad_norm": 2.812869580394354,
+      "learning_rate": 1.9517753294021876e-05,
+      "loss": 1.7738,
+      "step": 736
+    },
+    {
+      "epoch": 0.12648017847949203,
+      "grad_norm": 2.7574905789818307,
+      "learning_rate": 1.9516046537673543e-05,
+      "loss": 1.9723,
+      "step": 737
+    },
+    {
+      "epoch": 0.126651793375665,
+      "grad_norm": 2.56161869953953,
+      "learning_rate": 1.95143368412937e-05,
+      "loss": 1.8765,
+      "step": 738
+    },
+    {
+      "epoch": 0.126823408271838,
+      "grad_norm": 3.3025814410973964,
+      "learning_rate": 1.9512624205410568e-05,
+      "loss": 1.9226,
+      "step": 739
+    },
+    {
+      "epoch": 0.12699502316801098,
+      "grad_norm": 2.65364915839108,
+      "learning_rate": 1.951090863055327e-05,
+      "loss": 1.8904,
+      "step": 740
+    },
+    {
+      "epoch": 0.12716663806418396,
+      "grad_norm": 4.403950648327619,
+      "learning_rate": 1.9509190117251842e-05,
+      "loss": 1.8477,
+      "step": 741
+    },
+    {
+      "epoch": 0.12733825296035695,
+      "grad_norm": 2.7297214077424994,
+      "learning_rate": 1.950746866603723e-05,
+      "loss": 1.6244,
+      "step": 742
+    },
+    {
+      "epoch": 0.12750986785652996,
+      "grad_norm": 2.8623623789273953,
+      "learning_rate": 1.950574427744129e-05,
+      "loss": 1.8283,
+      "step": 743
+    },
+    {
+      "epoch": 0.12768148275270294,
+      "grad_norm": 2.776423003806021,
+      "learning_rate": 1.9504016951996774e-05,
+      "loss": 2.0191,
+      "step": 744
+    },
+    {
+      "epoch": 0.12785309764887592,
+      "grad_norm": 2.6959180765798445,
+      "learning_rate": 1.950228669023735e-05,
+      "loss": 1.6939,
+      "step": 745
+    },
+    {
+      "epoch": 0.1280247125450489,
+      "grad_norm": 2.8895795018010526,
+      "learning_rate": 1.9500553492697587e-05,
+      "loss": 1.7744,
+      "step": 746
+    },
+    {
+      "epoch": 0.1281963274412219,
+      "grad_norm": 2.959800013563544,
+      "learning_rate": 1.9498817359912973e-05,
+      "loss": 1.9661,
+      "step": 747
+    },
+    {
+      "epoch": 0.12836794233739487,
+      "grad_norm": 2.5285430397908923,
+      "learning_rate": 1.949707829241989e-05,
+      "loss": 1.7182,
+      "step": 748
+    },
+    {
+      "epoch": 0.12853955723356789,
+      "grad_norm": 4.091924561463437,
+      "learning_rate": 1.949533629075563e-05,
+      "loss": 2.043,
+      "step": 749
+    },
+    {
+      "epoch": 0.12871117212974087,
+      "grad_norm": 5.1680584893973585,
+      "learning_rate": 1.9493591355458404e-05,
+      "loss": 1.8322,
+      "step": 750
+    },
+    {
+      "epoch": 0.12888278702591385,
+      "grad_norm": 3.7773885086932535,
+      "learning_rate": 1.9491843487067305e-05,
+      "loss": 1.7617,
+      "step": 751
+    },
+    {
+      "epoch": 0.12905440192208684,
+      "grad_norm": 4.43437457571488,
+      "learning_rate": 1.9490092686122356e-05,
+      "loss": 1.8913,
+      "step": 752
+    },
+    {
+      "epoch": 0.12922601681825982,
+      "grad_norm": 4.634918982978317,
+      "learning_rate": 1.948833895316447e-05,
+      "loss": 2.0737,
+      "step": 753
+    },
+    {
+      "epoch": 0.1293976317144328,
+      "grad_norm": 3.5909650599406215,
+      "learning_rate": 1.9486582288735475e-05,
+      "loss": 2.0325,
+      "step": 754
+    },
+    {
+      "epoch": 0.1295692466106058,
+      "grad_norm": 4.173974377007036,
+      "learning_rate": 1.94848226933781e-05,
+      "loss": 1.6162,
+      "step": 755
+    },
+    {
+      "epoch": 0.1297408615067788,
+      "grad_norm": 3.5284936584114917,
+      "learning_rate": 1.9483060167635983e-05,
+      "loss": 2.0698,
+      "step": 756
+    },
+    {
+      "epoch": 0.12991247640295178,
+      "grad_norm": 3.4981107122899817,
+      "learning_rate": 1.9481294712053663e-05,
+      "loss": 1.7509,
+      "step": 757
+    },
+    {
+      "epoch": 0.13008409129912477,
+      "grad_norm": 2.5605155857801964,
+      "learning_rate": 1.947952632717659e-05,
+      "loss": 1.6244,
+      "step": 758
+    },
+    {
+      "epoch": 0.13025570619529775,
+      "grad_norm": 3.9763571740288612,
+      "learning_rate": 1.9477755013551106e-05,
+      "loss": 2.1433,
+      "step": 759
+    },
+    {
+      "epoch": 0.13042732109147073,
+      "grad_norm": 2.759892536779822,
+      "learning_rate": 1.947598077172448e-05,
+      "loss": 1.7429,
+      "step": 760
+    },
+    {
+      "epoch": 0.13059893598764372,
+      "grad_norm": 5.046535409841607,
+      "learning_rate": 1.9474203602244866e-05,
+      "loss": 2.0881,
+      "step": 761
+    },
+    {
+      "epoch": 0.13077055088381673,
+      "grad_norm": 4.706604367700716,
+      "learning_rate": 1.9472423505661332e-05,
+      "loss": 1.8053,
+      "step": 762
+    },
+    {
+      "epoch": 0.1309421657799897,
+      "grad_norm": 2.6280885019018974,
+      "learning_rate": 1.947064048252385e-05,
+      "loss": 2.0394,
+      "step": 763
+    },
+    {
+      "epoch": 0.1311137806761627,
+      "grad_norm": 3.014891797747624,
+      "learning_rate": 1.946885453338329e-05,
+      "loss": 1.7792,
+      "step": 764
+    },
+    {
+      "epoch": 0.13128539557233568,
+      "grad_norm": 2.3571690425386596,
+      "learning_rate": 1.946706565879143e-05,
+      "loss": 1.7023,
+      "step": 765
+    },
+    {
+      "epoch": 0.13145701046850866,
+      "grad_norm": 3.361034360801788,
+      "learning_rate": 1.9465273859300957e-05,
+      "loss": 1.7349,
+      "step": 766
+    },
+    {
+      "epoch": 0.13162862536468165,
+      "grad_norm": 2.5269476555856345,
+      "learning_rate": 1.946347913546545e-05,
+      "loss": 1.8084,
+      "step": 767
+    },
+    {
+      "epoch": 0.13180024026085463,
+      "grad_norm": 3.926594466583261,
+      "learning_rate": 1.9461681487839402e-05,
+      "loss": 2.0435,
+      "step": 768
+    },
+    {
+      "epoch": 0.13197185515702764,
+      "grad_norm": 2.8687039485104453,
+      "learning_rate": 1.945988091697821e-05,
+      "loss": 1.7041,
+      "step": 769
+    },
+    {
+      "epoch": 0.13214347005320062,
+      "grad_norm": 2.536793958815223,
+      "learning_rate": 1.9458077423438157e-05,
+      "loss": 1.7624,
+      "step": 770
+    },
+    {
+      "epoch": 0.1323150849493736,
+      "grad_norm": 2.3145146220422963,
+      "learning_rate": 1.9456271007776456e-05,
+      "loss": 1.96,
+      "step": 771
+    },
+    {
+      "epoch": 0.1324866998455466,
+      "grad_norm": 3.4557279192210286,
+      "learning_rate": 1.9454461670551198e-05,
+      "loss": 1.8301,
+      "step": 772
+    },
+    {
+      "epoch": 0.13265831474171957,
+      "grad_norm": 3.052242910129911,
+      "learning_rate": 1.945264941232139e-05,
+      "loss": 1.5877,
+      "step": 773
+    },
+    {
+      "epoch": 0.13282992963789256,
+      "grad_norm": 3.2937598733249565,
+      "learning_rate": 1.945083423364694e-05,
+      "loss": 1.8908,
+      "step": 774
+    },
+    {
+      "epoch": 0.13300154453406557,
+      "grad_norm": 8.45529425515086,
+      "learning_rate": 1.9449016135088657e-05,
+      "loss": 1.9223,
+      "step": 775
+    },
+    {
+      "epoch": 0.13317315943023855,
+      "grad_norm": 3.867949124556896,
+      "learning_rate": 1.944719511720825e-05,
+      "loss": 1.8183,
+      "step": 776
+    },
+    {
+      "epoch": 0.13334477432641154,
+      "grad_norm": 3.502460333702417,
+      "learning_rate": 1.9445371180568335e-05,
+      "loss": 2.1246,
+      "step": 777
+    },
+    {
+      "epoch": 0.13351638922258452,
+      "grad_norm": 2.8627687844724217,
+      "learning_rate": 1.944354432573242e-05,
+      "loss": 2.085,
+      "step": 778
+    },
+    {
+      "epoch": 0.1336880041187575,
+      "grad_norm": 2.916270177542365,
+      "learning_rate": 1.944171455326493e-05,
+      "loss": 1.9664,
+      "step": 779
+    },
+    {
+      "epoch": 0.1338596190149305,
+      "grad_norm": 5.241825710417449,
+      "learning_rate": 1.9439881863731177e-05,
+      "loss": 1.7361,
+      "step": 780
+    },
+    {
+      "epoch": 0.13403123391110347,
+      "grad_norm": 4.2172481694906745,
+      "learning_rate": 1.943804625769738e-05,
+      "loss": 1.8512,
+      "step": 781
+    },
+    {
+      "epoch": 0.13420284880727648,
+      "grad_norm": 2.602852498480944,
+      "learning_rate": 1.9436207735730658e-05,
+      "loss": 1.9726,
+      "step": 782
+    },
+    {
+      "epoch": 0.13437446370344946,
+      "grad_norm": 3.1822616104029664,
+      "learning_rate": 1.9434366298399037e-05,
+      "loss": 1.9234,
+      "step": 783
+    },
+    {
+      "epoch": 0.13454607859962245,
+      "grad_norm": 2.915442004940933,
+      "learning_rate": 1.9432521946271434e-05,
+      "loss": 1.894,
+      "step": 784
+    },
+    {
+      "epoch": 0.13471769349579543,
+      "grad_norm": 3.0864641358886624,
+      "learning_rate": 1.9430674679917672e-05,
+      "loss": 1.9447,
+      "step": 785
+    },
+    {
+      "epoch": 0.13488930839196842,
+      "grad_norm": 4.769603918101526,
+      "learning_rate": 1.9428824499908473e-05,
+      "loss": 1.8205,
+      "step": 786
+    },
+    {
+      "epoch": 0.1350609232881414,
+      "grad_norm": 3.1381352998719145,
+      "learning_rate": 1.9426971406815464e-05,
+      "loss": 1.8191,
+      "step": 787
+    },
+    {
+      "epoch": 0.1352325381843144,
+      "grad_norm": 3.5527016762460235,
+      "learning_rate": 1.942511540121116e-05,
+      "loss": 1.8117,
+      "step": 788
+    },
+    {
+      "epoch": 0.1354041530804874,
+      "grad_norm": 3.335609160798985,
+      "learning_rate": 1.942325648366899e-05,
+      "loss": 1.7949,
+      "step": 789
+    },
+    {
+      "epoch": 0.13557576797666038,
+      "grad_norm": 3.5589128617081522,
+      "learning_rate": 1.9421394654763275e-05,
+      "loss": 1.7756,
+      "step": 790
+    },
+    {
+      "epoch": 0.13574738287283336,
+      "grad_norm": 3.3931035007393,
+      "learning_rate": 1.9419529915069234e-05,
+      "loss": 1.9588,
+      "step": 791
+    },
+    {
+      "epoch": 0.13591899776900634,
+      "grad_norm": 2.732770910885144,
+      "learning_rate": 1.9417662265162988e-05,
+      "loss": 1.8161,
+      "step": 792
+    },
+    {
+      "epoch": 0.13609061266517933,
+      "grad_norm": 2.8714399790740592,
+      "learning_rate": 1.941579170562156e-05,
+      "loss": 2.0133,
+      "step": 793
+    },
+    {
+      "epoch": 0.1362622275613523,
+      "grad_norm": 2.8201340231038565,
+      "learning_rate": 1.9413918237022865e-05,
+      "loss": 1.9192,
+      "step": 794
+    },
+    {
+      "epoch": 0.13643384245752532,
+      "grad_norm": 2.8829937524306026,
+      "learning_rate": 1.9412041859945723e-05,
+      "loss": 1.8748,
+      "step": 795
+    },
+    {
+      "epoch": 0.1366054573536983,
+      "grad_norm": 2.6121298207375196,
+      "learning_rate": 1.941016257496985e-05,
+      "loss": 1.7875,
+      "step": 796
+    },
+    {
+      "epoch": 0.1367770722498713,
+      "grad_norm": 3.7901313395030893,
+      "learning_rate": 1.9408280382675862e-05,
+      "loss": 2.0532,
+      "step": 797
+    },
+    {
+      "epoch": 0.13694868714604427,
+      "grad_norm": 2.845924585151732,
+      "learning_rate": 1.9406395283645273e-05,
+      "loss": 1.7259,
+      "step": 798
+    },
+    {
+      "epoch": 0.13712030204221726,
+      "grad_norm": 2.5725454432234596,
+      "learning_rate": 1.9404507278460492e-05,
+      "loss": 1.9039,
+      "step": 799
+    },
+    {
+      "epoch": 0.13729191693839024,
+      "grad_norm": 2.8917593842067366,
+      "learning_rate": 1.940261636770483e-05,
+      "loss": 2.061,
+      "step": 800
+    },
+    {
+      "epoch": 0.13746353183456325,
+      "grad_norm": 2.1796790945538778,
+      "learning_rate": 1.9400722551962487e-05,
+      "loss": 1.697,
+      "step": 801
+    },
+    {
+      "epoch": 0.13763514673073624,
+      "grad_norm": 2.763119759861247,
+      "learning_rate": 1.939882583181857e-05,
+      "loss": 1.5279,
+      "step": 802
+    },
+    {
+      "epoch": 0.13780676162690922,
+      "grad_norm": 3.682365440487759,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 1.814,
+      "step": 803
+    },
+    {
+      "epoch": 0.1379783765230822,
+      "grad_norm": 3.162536467592949,
+      "learning_rate": 1.939502368067093e-05,
+      "loss": 1.9678,
+      "step": 804
+    },
+    {
+      "epoch": 0.13814999141925519,
+      "grad_norm": 3.143446298871448,
+      "learning_rate": 1.9393118250841897e-05,
+      "loss": 2.0622,
+      "step": 805
+    },
+    {
+      "epoch": 0.13832160631542817,
+      "grad_norm": 2.3456851919733004,
+      "learning_rate": 1.9391209918960677e-05,
+      "loss": 1.7169,
+      "step": 806
+    },
+    {
+      "epoch": 0.13849322121160118,
+      "grad_norm": 2.2080044222294903,
+      "learning_rate": 1.9389298685616866e-05,
+      "loss": 1.7335,
+      "step": 807
+    },
+    {
+      "epoch": 0.13866483610777416,
+      "grad_norm": 2.3748221133733436,
+      "learning_rate": 1.9387384551400942e-05,
+      "loss": 1.8508,
+      "step": 808
+    },
+    {
+      "epoch": 0.13883645100394715,
+      "grad_norm": 3.4599546197719966,
+      "learning_rate": 1.9385467516904288e-05,
+      "loss": 1.7573,
+      "step": 809
+    },
+    {
+      "epoch": 0.13900806590012013,
+      "grad_norm": 2.823313781110258,
+      "learning_rate": 1.9383547582719186e-05,
+      "loss": 1.9186,
+      "step": 810
+    },
+    {
+      "epoch": 0.13917968079629311,
+      "grad_norm": 2.7898428340211727,
+      "learning_rate": 1.9381624749438807e-05,
+      "loss": 1.6144,
+      "step": 811
+    },
+    {
+      "epoch": 0.1393512956924661,
+      "grad_norm": 3.014886138789283,
+      "learning_rate": 1.937969901765722e-05,
+      "loss": 1.947,
+      "step": 812
+    },
+    {
+      "epoch": 0.13952291058863908,
+      "grad_norm": 2.365393743197819,
+      "learning_rate": 1.9377770387969388e-05,
+      "loss": 2.0048,
+      "step": 813
+    },
+    {
+      "epoch": 0.1396945254848121,
+      "grad_norm": 2.862184098927013,
+      "learning_rate": 1.937583886097117e-05,
+      "loss": 1.6929,
+      "step": 814
+    },
+    {
+      "epoch": 0.13986614038098508,
+      "grad_norm": 5.177811782665515,
+      "learning_rate": 1.937390443725933e-05,
+      "loss": 2.0268,
+      "step": 815
+    },
+    {
+      "epoch": 0.14003775527715806,
+      "grad_norm": 2.6597735151341095,
+      "learning_rate": 1.9371967117431513e-05,
+      "loss": 1.9714,
+      "step": 816
+    },
+    {
+      "epoch": 0.14020937017333104,
+      "grad_norm": 2.537487351793122,
+      "learning_rate": 1.937002690208626e-05,
+      "loss": 1.8663,
+      "step": 817
+    },
+    {
+      "epoch": 0.14038098506950403,
+      "grad_norm": 2.5155501706236736,
+      "learning_rate": 1.9368083791823013e-05,
+      "loss": 1.9868,
+      "step": 818
+    },
+    {
+      "epoch": 0.140552599965677,
+      "grad_norm": 2.70501004307245,
+      "learning_rate": 1.9366137787242107e-05,
+      "loss": 1.8901,
+      "step": 819
+    },
+    {
+      "epoch": 0.14072421486185002,
+      "grad_norm": 2.7554807612409076,
+      "learning_rate": 1.936418888894477e-05,
+      "loss": 1.7269,
+      "step": 820
+    },
+    {
+      "epoch": 0.140895829758023,
+      "grad_norm": 2.8951157418624254,
+      "learning_rate": 1.9362237097533123e-05,
+      "loss": 1.8491,
+      "step": 821
+    },
+    {
+      "epoch": 0.141067444654196,
+      "grad_norm": 2.685476353499669,
+      "learning_rate": 1.9360282413610185e-05,
+      "loss": 1.8163,
+      "step": 822
+    },
+    {
+      "epoch": 0.14123905955036897,
+      "grad_norm": 2.914179530331329,
+      "learning_rate": 1.9358324837779864e-05,
+      "loss": 1.7363,
+      "step": 823
+    },
+    {
+      "epoch": 0.14141067444654196,
+      "grad_norm": 2.605303386454893,
+      "learning_rate": 1.935636437064696e-05,
+      "loss": 1.7258,
+      "step": 824
+    },
+    {
+      "epoch": 0.14158228934271494,
+      "grad_norm": 2.8321319406795085,
+      "learning_rate": 1.9354401012817172e-05,
+      "loss": 2.0022,
+      "step": 825
+    },
+    {
+      "epoch": 0.14175390423888792,
+      "grad_norm": 3.0643757031727996,
+      "learning_rate": 1.935243476489709e-05,
+      "loss": 2.0587,
+      "step": 826
+    },
+    {
+      "epoch": 0.14192551913506093,
+      "grad_norm": 3.2031184109178086,
+      "learning_rate": 1.9350465627494196e-05,
+      "loss": 1.651,
+      "step": 827
+    },
+    {
+      "epoch": 0.14209713403123392,
+      "grad_norm": 4.370719302377024,
+      "learning_rate": 1.9348493601216864e-05,
+      "loss": 1.8206,
+      "step": 828
+    },
+    {
+      "epoch": 0.1422687489274069,
+      "grad_norm": 4.842158211369917,
+      "learning_rate": 1.9346518686674365e-05,
+      "loss": 2.2634,
+      "step": 829
+    },
+    {
+      "epoch": 0.14244036382357989,
+      "grad_norm": 3.76928527886839,
+      "learning_rate": 1.9344540884476856e-05,
+      "loss": 2.0082,
+      "step": 830
+    },
+    {
+      "epoch": 0.14261197871975287,
+      "grad_norm": 2.762609253767019,
+      "learning_rate": 1.9342560195235386e-05,
+      "loss": 1.8976,
+      "step": 831
+    },
+    {
+      "epoch": 0.14278359361592585,
+      "grad_norm": 2.410302542310939,
+      "learning_rate": 1.9340576619561907e-05,
+      "loss": 1.9664,
+      "step": 832
+    },
+    {
+      "epoch": 0.14295520851209886,
+      "grad_norm": 3.7311121338365396,
+      "learning_rate": 1.933859015806925e-05,
+      "loss": 1.6968,
+      "step": 833
+    },
+    {
+      "epoch": 0.14312682340827185,
+      "grad_norm": 5.519498021808017,
+      "learning_rate": 1.9336600811371143e-05,
+      "loss": 1.8068,
+      "step": 834
+    },
+    {
+      "epoch": 0.14329843830444483,
+      "grad_norm": 3.9452926697799198,
+      "learning_rate": 1.9334608580082204e-05,
+      "loss": 1.9151,
+      "step": 835
+    },
+    {
+      "epoch": 0.14347005320061781,
+      "grad_norm": 2.47590519165114,
+      "learning_rate": 1.9332613464817947e-05,
+      "loss": 2.1557,
+      "step": 836
+    },
+    {
+      "epoch": 0.1436416680967908,
+      "grad_norm": 2.2773478151251596,
+      "learning_rate": 1.933061546619477e-05,
+      "loss": 1.8475,
+      "step": 837
+    },
+    {
+      "epoch": 0.14381328299296378,
+      "grad_norm": 3.0465913365439588,
+      "learning_rate": 1.9328614584829963e-05,
+      "loss": 1.5076,
+      "step": 838
+    },
+    {
+      "epoch": 0.14398489788913676,
+      "grad_norm": 3.1285160686984628,
+      "learning_rate": 1.9326610821341713e-05,
+      "loss": 1.9832,
+      "step": 839
+    },
+    {
+      "epoch": 0.14415651278530978,
+      "grad_norm": 2.7302090464369035,
+      "learning_rate": 1.932460417634909e-05,
+      "loss": 1.893,
+      "step": 840
+    },
+    {
+      "epoch": 0.14432812768148276,
+      "grad_norm": 2.9357521601780316,
+      "learning_rate": 1.932259465047206e-05,
+      "loss": 1.9015,
+      "step": 841
+    },
+    {
+      "epoch": 0.14449974257765574,
+      "grad_norm": 2.932565203055729,
+      "learning_rate": 1.9320582244331475e-05,
+      "loss": 1.9993,
+      "step": 842
+    },
+    {
+      "epoch": 0.14467135747382873,
+      "grad_norm": 2.5392560938244033,
+      "learning_rate": 1.9318566958549075e-05,
+      "loss": 1.8044,
+      "step": 843
+    },
+    {
+      "epoch": 0.1448429723700017,
+      "grad_norm": 2.931105428412681,
+      "learning_rate": 1.9316548793747495e-05,
+      "loss": 1.6959,
+      "step": 844
+    },
+    {
+      "epoch": 0.1450145872661747,
+      "grad_norm": 3.076745313072229,
+      "learning_rate": 1.9314527750550263e-05,
+      "loss": 2.0188,
+      "step": 845
+    },
+    {
+      "epoch": 0.1451862021623477,
+      "grad_norm": 3.1249281361631684,
+      "learning_rate": 1.931250382958178e-05,
+      "loss": 1.863,
+      "step": 846
+    },
+    {
+      "epoch": 0.1453578170585207,
+      "grad_norm": 3.425673091348935,
+      "learning_rate": 1.9310477031467357e-05,
+      "loss": 1.9239,
+      "step": 847
+    },
+    {
+      "epoch": 0.14552943195469367,
+      "grad_norm": 3.1376131163028655,
+      "learning_rate": 1.930844735683318e-05,
+      "loss": 2.0033,
+      "step": 848
+    },
+    {
+      "epoch": 0.14570104685086666,
+      "grad_norm": 2.865218109427285,
+      "learning_rate": 1.9306414806306323e-05,
+      "loss": 1.6858,
+      "step": 849
+    },
+    {
+      "epoch": 0.14587266174703964,
+      "grad_norm": 2.836261917489325,
+      "learning_rate": 1.930437938051476e-05,
+      "loss": 2.0244,
+      "step": 850
+    },
+    {
+      "epoch": 0.14604427664321262,
+      "grad_norm": 2.4022066479421054,
+      "learning_rate": 1.930234108008734e-05,
+      "loss": 1.7304,
+      "step": 851
+    },
+    {
+      "epoch": 0.1462158915393856,
+      "grad_norm": 2.9197060396009182,
+      "learning_rate": 1.9300299905653813e-05,
+      "loss": 1.7783,
+      "step": 852
+    },
+    {
+      "epoch": 0.14638750643555862,
+      "grad_norm": 3.5324989987112168,
+      "learning_rate": 1.9298255857844803e-05,
+      "loss": 2.1322,
+      "step": 853
+    },
+    {
+      "epoch": 0.1465591213317316,
+      "grad_norm": 2.6958341222475983,
+      "learning_rate": 1.9296208937291835e-05,
+      "loss": 1.8372,
+      "step": 854
+    },
+    {
+      "epoch": 0.14673073622790458,
+      "grad_norm": 3.279463392162738,
+      "learning_rate": 1.929415914462731e-05,
+      "loss": 1.9398,
+      "step": 855
+    },
+    {
+      "epoch": 0.14690235112407757,
+      "grad_norm": 2.9746504195005925,
+      "learning_rate": 1.9292106480484533e-05,
+      "loss": 1.9512,
+      "step": 856
+    },
+    {
+      "epoch": 0.14707396602025055,
+      "grad_norm": 2.599058511769964,
+      "learning_rate": 1.9290050945497668e-05,
+      "loss": 1.7489,
+      "step": 857
+    },
+    {
+      "epoch": 0.14724558091642354,
+      "grad_norm": 2.4030634214006206,
+      "learning_rate": 1.9287992540301802e-05,
+      "loss": 1.7582,
+      "step": 858
+    },
+    {
+      "epoch": 0.14741719581259655,
+      "grad_norm": 3.827757116809412,
+      "learning_rate": 1.9285931265532875e-05,
+      "loss": 1.7235,
+      "step": 859
+    },
+    {
+      "epoch": 0.14758881070876953,
+      "grad_norm": 2.7678919341185644,
+      "learning_rate": 1.9283867121827736e-05,
+      "loss": 1.7356,
+      "step": 860
+    },
+    {
+      "epoch": 0.1477604256049425,
+      "grad_norm": 3.2867412465832855,
+      "learning_rate": 1.928180010982411e-05,
+      "loss": 1.9071,
+      "step": 861
+    },
+    {
+      "epoch": 0.1479320405011155,
+      "grad_norm": 2.656270649072383,
+      "learning_rate": 1.927973023016061e-05,
+      "loss": 1.911,
+      "step": 862
+    },
+    {
+      "epoch": 0.14810365539728848,
+      "grad_norm": 2.434197098758412,
+      "learning_rate": 1.9277657483476736e-05,
+      "loss": 1.6403,
+      "step": 863
+    },
+    {
+      "epoch": 0.14827527029346146,
+      "grad_norm": 2.8058482059718233,
+      "learning_rate": 1.9275581870412883e-05,
+      "loss": 1.863,
+      "step": 864
+    },
+    {
+      "epoch": 0.14844688518963445,
+      "grad_norm": 3.230606706076792,
+      "learning_rate": 1.9273503391610307e-05,
+      "loss": 1.8928,
+      "step": 865
+    },
+    {
+      "epoch": 0.14861850008580746,
+      "grad_norm": 3.0741959612497345,
+      "learning_rate": 1.927142204771118e-05,
+      "loss": 1.9563,
+      "step": 866
+    },
+    {
+      "epoch": 0.14879011498198044,
+      "grad_norm": 2.4980609904025406,
+      "learning_rate": 1.926933783935853e-05,
+      "loss": 1.9163,
+      "step": 867
+    },
+    {
+      "epoch": 0.14896172987815343,
+      "grad_norm": 3.357732561087617,
+      "learning_rate": 1.926725076719629e-05,
+      "loss": 1.7384,
+      "step": 868
+    },
+    {
+      "epoch": 0.1491333447743264,
+      "grad_norm": 2.911414826424691,
+      "learning_rate": 1.9265160831869276e-05,
+      "loss": 1.9063,
+      "step": 869
+    },
+    {
+      "epoch": 0.1493049596704994,
+      "grad_norm": 2.2814824068799466,
+      "learning_rate": 1.926306803402318e-05,
+      "loss": 2.0599,
+      "step": 870
+    },
+    {
+      "epoch": 0.14947657456667238,
+      "grad_norm": 2.0531841217194304,
+      "learning_rate": 1.9260972374304582e-05,
+      "loss": 1.6243,
+      "step": 871
+    },
+    {
+      "epoch": 0.1496481894628454,
+      "grad_norm": 2.4565323616356896,
+      "learning_rate": 1.9258873853360946e-05,
+      "loss": 1.6371,
+      "step": 872
+    },
+    {
+      "epoch": 0.14981980435901837,
+      "grad_norm": 2.5500894867398007,
+      "learning_rate": 1.9256772471840627e-05,
+      "loss": 1.983,
+      "step": 873
+    },
+    {
+      "epoch": 0.14999141925519135,
+      "grad_norm": 3.018479297539399,
+      "learning_rate": 1.925466823039285e-05,
+      "loss": 1.8799,
+      "step": 874
+    },
+    {
+      "epoch": 0.15016303415136434,
+      "grad_norm": 3.3599963072443138,
+      "learning_rate": 1.9252561129667736e-05,
+      "loss": 1.9047,
+      "step": 875
+    },
+    {
+      "epoch": 0.15033464904753732,
+      "grad_norm": 2.876939590060346,
+      "learning_rate": 1.925045117031628e-05,
+      "loss": 1.8091,
+      "step": 876
+    },
+    {
+      "epoch": 0.1505062639437103,
+      "grad_norm": 4.587035200278965,
+      "learning_rate": 1.924833835299037e-05,
+      "loss": 1.8766,
+      "step": 877
+    },
+    {
+      "epoch": 0.1506778788398833,
+      "grad_norm": 3.2603079675805917,
+      "learning_rate": 1.924622267834277e-05,
+      "loss": 1.818,
+      "step": 878
+    },
+    {
+      "epoch": 0.1508494937360563,
+      "grad_norm": 2.914222052871642,
+      "learning_rate": 1.9244104147027127e-05,
+      "loss": 1.9931,
+      "step": 879
+    },
+    {
+      "epoch": 0.15102110863222928,
+      "grad_norm": 2.869009191272744,
+      "learning_rate": 1.9241982759697976e-05,
+      "loss": 1.8678,
+      "step": 880
+    },
+    {
+      "epoch": 0.15119272352840227,
+      "grad_norm": 3.760143196914323,
+      "learning_rate": 1.9239858517010724e-05,
+      "loss": 1.8291,
+      "step": 881
+    },
+    {
+      "epoch": 0.15136433842457525,
+      "grad_norm": 2.870117971628503,
+      "learning_rate": 1.9237731419621672e-05,
+      "loss": 1.8694,
+      "step": 882
+    },
+    {
+      "epoch": 0.15153595332074823,
+      "grad_norm": 5.117646025360141,
+      "learning_rate": 1.9235601468187997e-05,
+      "loss": 1.7764,
+      "step": 883
+    },
+    {
+      "epoch": 0.15170756821692122,
+      "grad_norm": 2.034057626859885,
+      "learning_rate": 1.9233468663367755e-05,
+      "loss": 1.8281,
+      "step": 884
+    },
+    {
+      "epoch": 0.15187918311309423,
+      "grad_norm": 1.9387544199806614,
+      "learning_rate": 1.9231333005819893e-05,
+      "loss": 1.5317,
+      "step": 885
+    },
+    {
+      "epoch": 0.1520507980092672,
+      "grad_norm": 2.3356128278812727,
+      "learning_rate": 1.9229194496204226e-05,
+      "loss": 1.9282,
+      "step": 886
+    },
+    {
+      "epoch": 0.1522224129054402,
+      "grad_norm": 2.7405137688083383,
+      "learning_rate": 1.9227053135181467e-05,
+      "loss": 1.8164,
+      "step": 887
+    },
+    {
+      "epoch": 0.15239402780161318,
+      "grad_norm": 3.214610331040803,
+      "learning_rate": 1.922490892341319e-05,
+      "loss": 1.7349,
+      "step": 888
+    },
+    {
+      "epoch": 0.15256564269778616,
+      "grad_norm": 3.465707142240606,
+      "learning_rate": 1.9222761861561868e-05,
+      "loss": 1.7254,
+      "step": 889
+    },
+    {
+      "epoch": 0.15273725759395915,
+      "grad_norm": 3.4556275043365927,
+      "learning_rate": 1.9220611950290845e-05,
+      "loss": 1.8544,
+      "step": 890
+    },
+    {
+      "epoch": 0.15290887249013213,
+      "grad_norm": 3.4041490208093257,
+      "learning_rate": 1.9218459190264346e-05,
+      "loss": 1.829,
+      "step": 891
+    },
+    {
+      "epoch": 0.15308048738630514,
+      "grad_norm": 2.5276474455165263,
+      "learning_rate": 1.9216303582147475e-05,
+      "loss": 1.7018,
+      "step": 892
+    },
+    {
+      "epoch": 0.15325210228247813,
+      "grad_norm": 2.4042077778658797,
+      "learning_rate": 1.9214145126606224e-05,
+      "loss": 1.7764,
+      "step": 893
+    },
+    {
+      "epoch": 0.1534237171786511,
+      "grad_norm": 2.7129688878164537,
+      "learning_rate": 1.9211983824307455e-05,
+      "loss": 1.5837,
+      "step": 894
+    },
+    {
+      "epoch": 0.1535953320748241,
+      "grad_norm": 3.2627235950578886,
+      "learning_rate": 1.920981967591891e-05,
+      "loss": 2.0954,
+      "step": 895
+    },
+    {
+      "epoch": 0.15376694697099708,
+      "grad_norm": 2.7252865505248867,
+      "learning_rate": 1.9207652682109223e-05,
+      "loss": 1.7785,
+      "step": 896
+    },
+    {
+      "epoch": 0.15393856186717006,
+      "grad_norm": 2.8096998910455153,
+      "learning_rate": 1.920548284354789e-05,
+      "loss": 1.7101,
+      "step": 897
+    },
+    {
+      "epoch": 0.15411017676334307,
+      "grad_norm": 2.537356504744846,
+      "learning_rate": 1.9203310160905296e-05,
+      "loss": 2.1799,
+      "step": 898
+    },
+    {
+      "epoch": 0.15428179165951605,
+      "grad_norm": 2.9497464860444316,
+      "learning_rate": 1.9201134634852706e-05,
+      "loss": 1.7241,
+      "step": 899
+    },
+    {
+      "epoch": 0.15445340655568904,
+      "grad_norm": 5.570811784311937,
+      "learning_rate": 1.9198956266062258e-05,
+      "loss": 1.8488,
+      "step": 900
+    },
+    {
+      "epoch": 0.15462502145186202,
+      "grad_norm": 2.9716800160992327,
+      "learning_rate": 1.9196775055206965e-05,
+      "loss": 2.1307,
+      "step": 901
+    },
+    {
+      "epoch": 0.154796636348035,
+      "grad_norm": 2.620842956655345,
+      "learning_rate": 1.9194591002960732e-05,
+      "loss": 1.9544,
+      "step": 902
+    },
+    {
+      "epoch": 0.154968251244208,
+      "grad_norm": 3.3286638067892254,
+      "learning_rate": 1.9192404109998325e-05,
+      "loss": 1.8323,
+      "step": 903
+    },
+    {
+      "epoch": 0.15513986614038097,
+      "grad_norm": 2.5659862954745294,
+      "learning_rate": 1.91902143769954e-05,
+      "loss": 1.9745,
+      "step": 904
+    },
+    {
+      "epoch": 0.15531148103655398,
+      "grad_norm": 2.661306312749646,
+      "learning_rate": 1.9188021804628486e-05,
+      "loss": 1.6669,
+      "step": 905
+    },
+    {
+      "epoch": 0.15548309593272697,
+      "grad_norm": 3.308562108860248,
+      "learning_rate": 1.9185826393574994e-05,
+      "loss": 1.9903,
+      "step": 906
+    },
+    {
+      "epoch": 0.15565471082889995,
+      "grad_norm": 2.9098122788690106,
+      "learning_rate": 1.9183628144513196e-05,
+      "loss": 1.6804,
+      "step": 907
+    },
+    {
+      "epoch": 0.15582632572507293,
+      "grad_norm": 3.728510596350371,
+      "learning_rate": 1.9181427058122265e-05,
+      "loss": 1.7867,
+      "step": 908
+    },
+    {
+      "epoch": 0.15599794062124592,
+      "grad_norm": 3.7122472196052425,
+      "learning_rate": 1.9179223135082226e-05,
+      "loss": 1.9655,
+      "step": 909
+    },
+    {
+      "epoch": 0.1561695555174189,
+      "grad_norm": 2.958838137311516,
+      "learning_rate": 1.9177016376074006e-05,
+      "loss": 1.7651,
+      "step": 910
+    },
+    {
+      "epoch": 0.1563411704135919,
+      "grad_norm": 3.026374820962408,
+      "learning_rate": 1.9174806781779384e-05,
+      "loss": 1.8354,
+      "step": 911
+    },
+    {
+      "epoch": 0.1565127853097649,
+      "grad_norm": 4.492953610295223,
+      "learning_rate": 1.9172594352881028e-05,
+      "loss": 1.9504,
+      "step": 912
+    },
+    {
+      "epoch": 0.15668440020593788,
+      "grad_norm": 2.900144753968202,
+      "learning_rate": 1.917037909006248e-05,
+      "loss": 2.048,
+      "step": 913
+    },
+    {
+      "epoch": 0.15685601510211086,
+      "grad_norm": 2.6953935878760378,
+      "learning_rate": 1.9168160994008157e-05,
+      "loss": 1.8892,
+      "step": 914
+    },
+    {
+      "epoch": 0.15702762999828385,
+      "grad_norm": 4.945101883397236,
+      "learning_rate": 1.916594006540335e-05,
+      "loss": 1.8452,
+      "step": 915
+    },
+    {
+      "epoch": 0.15719924489445683,
+      "grad_norm": 2.8266649868888125,
+      "learning_rate": 1.9163716304934227e-05,
+      "loss": 1.759,
+      "step": 916
+    },
+    {
+      "epoch": 0.1573708597906298,
+      "grad_norm": 2.2645644715819957,
+      "learning_rate": 1.916148971328783e-05,
+      "loss": 1.8898,
+      "step": 917
+    },
+    {
+      "epoch": 0.15754247468680282,
+      "grad_norm": 2.55348016259605,
+      "learning_rate": 1.9159260291152077e-05,
+      "loss": 1.8977,
+      "step": 918
+    },
+    {
+      "epoch": 0.1577140895829758,
+      "grad_norm": 2.4832091991606204,
+      "learning_rate": 1.915702803921576e-05,
+      "loss": 1.8246,
+      "step": 919
+    },
+    {
+      "epoch": 0.1578857044791488,
+      "grad_norm": 2.4629536181266203,
+      "learning_rate": 1.9154792958168542e-05,
+      "loss": 1.9943,
+      "step": 920
+    },
+    {
+      "epoch": 0.15805731937532178,
+      "grad_norm": 3.0165092747286177,
+      "learning_rate": 1.9152555048700966e-05,
+      "loss": 1.9152,
+      "step": 921
+    },
+    {
+      "epoch": 0.15822893427149476,
+      "grad_norm": 2.391542522749986,
+      "learning_rate": 1.9150314311504444e-05,
+      "loss": 1.9288,
+      "step": 922
+    },
+    {
+      "epoch": 0.15840054916766774,
+      "grad_norm": 4.100846860532991,
+      "learning_rate": 1.9148070747271262e-05,
+      "loss": 1.7039,
+      "step": 923
+    },
+    {
+      "epoch": 0.15857216406384075,
+      "grad_norm": 3.2340100229029347,
+      "learning_rate": 1.914582435669458e-05,
+      "loss": 2.0645,
+      "step": 924
+    },
+    {
+      "epoch": 0.15874377896001374,
+      "grad_norm": 3.4940397651555535,
+      "learning_rate": 1.914357514046844e-05,
+      "loss": 1.8363,
+      "step": 925
+    },
+    {
+      "epoch": 0.15891539385618672,
+      "grad_norm": 2.4999340289881244,
+      "learning_rate": 1.9141323099287737e-05,
+      "loss": 1.8756,
+      "step": 926
+    },
+    {
+      "epoch": 0.1590870087523597,
+      "grad_norm": 2.9114026610321053,
+      "learning_rate": 1.913906823384826e-05,
+      "loss": 1.8852,
+      "step": 927
+    },
+    {
+      "epoch": 0.1592586236485327,
+      "grad_norm": 3.114045024461276,
+      "learning_rate": 1.913681054484666e-05,
+      "loss": 1.8964,
+      "step": 928
+    },
+    {
+      "epoch": 0.15943023854470567,
+      "grad_norm": 3.0381278250794588,
+      "learning_rate": 1.9134550032980456e-05,
+      "loss": 1.8356,
+      "step": 929
+    },
+    {
+      "epoch": 0.15960185344087865,
+      "grad_norm": 3.5818913304011235,
+      "learning_rate": 1.913228669894805e-05,
+      "loss": 1.7601,
+      "step": 930
+    },
+    {
+      "epoch": 0.15977346833705167,
+      "grad_norm": 3.1344752832343414,
+      "learning_rate": 1.9130020543448705e-05,
+      "loss": 1.9302,
+      "step": 931
+    },
+    {
+      "epoch": 0.15994508323322465,
+      "grad_norm": 3.0816071140514105,
+      "learning_rate": 1.9127751567182568e-05,
+      "loss": 1.8898,
+      "step": 932
+    },
+    {
+      "epoch": 0.16011669812939763,
+      "grad_norm": 3.017305791318649,
+      "learning_rate": 1.9125479770850646e-05,
+      "loss": 1.7578,
+      "step": 933
+    },
+    {
+      "epoch": 0.16028831302557062,
+      "grad_norm": 2.906786195900609,
+      "learning_rate": 1.9123205155154827e-05,
+      "loss": 2.0959,
+      "step": 934
+    },
+    {
+      "epoch": 0.1604599279217436,
+      "grad_norm": 4.058030755809474,
+      "learning_rate": 1.912092772079786e-05,
+      "loss": 1.9833,
+      "step": 935
+    },
+    {
+      "epoch": 0.16063154281791658,
+      "grad_norm": 2.722762915083046,
+      "learning_rate": 1.9118647468483375e-05,
+      "loss": 1.7963,
+      "step": 936
+    },
+    {
+      "epoch": 0.1608031577140896,
+      "grad_norm": 3.4478066449412546,
+      "learning_rate": 1.9116364398915866e-05,
+      "loss": 1.7697,
+      "step": 937
+    },
+    {
+      "epoch": 0.16097477261026258,
+      "grad_norm": 3.8843306634289583,
+      "learning_rate": 1.9114078512800692e-05,
+      "loss": 1.751,
+      "step": 938
+    },
+    {
+      "epoch": 0.16114638750643556,
+      "grad_norm": 2.619121463614514,
+      "learning_rate": 1.9111789810844104e-05,
+      "loss": 1.7204,
+      "step": 939
+    },
+    {
+      "epoch": 0.16131800240260855,
+      "grad_norm": 3.1020226140291887,
+      "learning_rate": 1.9109498293753196e-05,
+      "loss": 1.7465,
+      "step": 940
+    },
+    {
+      "epoch": 0.16148961729878153,
+      "grad_norm": 2.816955507887721,
+      "learning_rate": 1.9107203962235947e-05,
+      "loss": 1.6794,
+      "step": 941
+    },
+    {
+      "epoch": 0.1616612321949545,
+      "grad_norm": 3.2824724213697447,
+      "learning_rate": 1.91049068170012e-05,
+      "loss": 1.7686,
+      "step": 942
+    },
+    {
+      "epoch": 0.1618328470911275,
+      "grad_norm": 2.297484792157275,
+      "learning_rate": 1.910260685875868e-05,
+      "loss": 1.6778,
+      "step": 943
+    },
+    {
+      "epoch": 0.1620044619873005,
+      "grad_norm": 2.663356348667494,
+      "learning_rate": 1.9100304088218964e-05,
+      "loss": 1.6639,
+      "step": 944
+    },
+    {
+      "epoch": 0.1621760768834735,
+      "grad_norm": 3.2511637087509366,
+      "learning_rate": 1.9097998506093503e-05,
+      "loss": 1.7046,
+      "step": 945
+    },
+    {
+      "epoch": 0.16234769177964647,
+      "grad_norm": 2.7278840599373817,
+      "learning_rate": 1.909569011309462e-05,
+      "loss": 1.8123,
+      "step": 946
+    },
+    {
+      "epoch": 0.16251930667581946,
+      "grad_norm": 3.414306504376067,
+      "learning_rate": 1.9093378909935503e-05,
+      "loss": 1.8098,
+      "step": 947
+    },
+    {
+      "epoch": 0.16269092157199244,
+      "grad_norm": 3.2035688150224404,
+      "learning_rate": 1.9091064897330213e-05,
+      "loss": 1.8737,
+      "step": 948
+    },
+    {
+      "epoch": 0.16286253646816543,
+      "grad_norm": 2.836132541693522,
+      "learning_rate": 1.908874807599368e-05,
+      "loss": 2.0398,
+      "step": 949
+    },
+    {
+      "epoch": 0.16303415136433844,
+      "grad_norm": 3.1036563643922706,
+      "learning_rate": 1.908642844664169e-05,
+      "loss": 1.8861,
+      "step": 950
+    },
+    {
+      "epoch": 0.16320576626051142,
+      "grad_norm": 2.983245282448459,
+      "learning_rate": 1.9084106009990905e-05,
+      "loss": 1.7373,
+      "step": 951
+    },
+    {
+      "epoch": 0.1633773811566844,
+      "grad_norm": 6.702190132656186,
+      "learning_rate": 1.9081780766758862e-05,
+      "loss": 1.7819,
+      "step": 952
+    },
+    {
+      "epoch": 0.1635489960528574,
+      "grad_norm": 5.130541009613373,
+      "learning_rate": 1.9079452717663946e-05,
+      "loss": 2.0151,
+      "step": 953
+    },
+    {
+      "epoch": 0.16372061094903037,
+      "grad_norm": 3.410438731271111,
+      "learning_rate": 1.9077121863425428e-05,
+      "loss": 1.7795,
+      "step": 954
+    },
+    {
+      "epoch": 0.16389222584520335,
+      "grad_norm": 2.976219591231255,
+      "learning_rate": 1.9074788204763438e-05,
+      "loss": 1.917,
+      "step": 955
+    },
+    {
+      "epoch": 0.16406384074137634,
+      "grad_norm": 3.2954198967420467,
+      "learning_rate": 1.907245174239896e-05,
+      "loss": 1.9899,
+      "step": 956
+    },
+    {
+      "epoch": 0.16423545563754935,
+      "grad_norm": 3.4173110130146602,
+      "learning_rate": 1.9070112477053874e-05,
+      "loss": 1.9594,
+      "step": 957
+    },
+    {
+      "epoch": 0.16440707053372233,
+      "grad_norm": 2.714965944147371,
+      "learning_rate": 1.906777040945089e-05,
+      "loss": 1.6845,
+      "step": 958
+    },
+    {
+      "epoch": 0.16457868542989532,
+      "grad_norm": 3.653926983431517,
+      "learning_rate": 1.9065425540313613e-05,
+      "loss": 1.9838,
+      "step": 959
+    },
+    {
+      "epoch": 0.1647503003260683,
+      "grad_norm": 3.5435716850442054,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 2.0632,
+      "step": 960
+    },
+    {
+      "epoch": 0.16492191522224128,
+      "grad_norm": 2.7207032706526957,
+      "learning_rate": 1.9060727400334875e-05,
+      "loss": 1.8097,
+      "step": 961
+    },
+    {
+      "epoch": 0.16509353011841427,
+      "grad_norm": 2.892369982518829,
+      "learning_rate": 1.905837413094493e-05,
+      "loss": 1.9241,
+      "step": 962
+    },
+    {
+      "epoch": 0.16526514501458728,
+      "grad_norm": 3.6079034976895206,
+      "learning_rate": 1.9056018062923717e-05,
+      "loss": 1.7801,
+      "step": 963
+    },
+    {
+      "epoch": 0.16543675991076026,
+      "grad_norm": 3.1529193450761888,
+      "learning_rate": 1.9053659196999157e-05,
+      "loss": 2.0072,
+      "step": 964
+    },
+    {
+      "epoch": 0.16560837480693325,
+      "grad_norm": 2.6680618292760654,
+      "learning_rate": 1.9051297533900032e-05,
+      "loss": 1.58,
+      "step": 965
+    },
+    {
+      "epoch": 0.16577998970310623,
+      "grad_norm": 2.7497719371609786,
+      "learning_rate": 1.9048933074355988e-05,
+      "loss": 1.8281,
+      "step": 966
+    },
+    {
+      "epoch": 0.1659516045992792,
+      "grad_norm": 2.6238643035791043,
+      "learning_rate": 1.9046565819097546e-05,
+      "loss": 1.5967,
+      "step": 967
+    },
+    {
+      "epoch": 0.1661232194954522,
+      "grad_norm": 3.805257049881217,
+      "learning_rate": 1.9044195768856068e-05,
+      "loss": 2.084,
+      "step": 968
+    },
+    {
+      "epoch": 0.16629483439162518,
+      "grad_norm": 3.6411979141620403,
+      "learning_rate": 1.9041822924363808e-05,
+      "loss": 2.1312,
+      "step": 969
+    },
+    {
+      "epoch": 0.1664664492877982,
+      "grad_norm": 3.6989042208560092,
+      "learning_rate": 1.9039447286353855e-05,
+      "loss": 1.9246,
+      "step": 970
+    },
+    {
+      "epoch": 0.16663806418397117,
+      "grad_norm": 2.952442496832039,
+      "learning_rate": 1.9037068855560184e-05,
+      "loss": 1.9249,
+      "step": 971
+    },
+    {
+      "epoch": 0.16680967908014416,
+      "grad_norm": 2.947003125909465,
+      "learning_rate": 1.903468763271762e-05,
+      "loss": 1.8068,
+      "step": 972
+    },
+    {
+      "epoch": 0.16698129397631714,
+      "grad_norm": 2.9669996228262083,
+      "learning_rate": 1.9032303618561855e-05,
+      "loss": 1.6313,
+      "step": 973
+    },
+    {
+      "epoch": 0.16715290887249012,
+      "grad_norm": 3.851865592926603,
+      "learning_rate": 1.9029916813829438e-05,
+      "loss": 1.9488,
+      "step": 974
+    },
+    {
+      "epoch": 0.1673245237686631,
+      "grad_norm": 4.1352847565813065,
+      "learning_rate": 1.902752721925779e-05,
+      "loss": 1.6762,
+      "step": 975
+    },
+    {
+      "epoch": 0.16749613866483612,
+      "grad_norm": 4.25371793127814,
+      "learning_rate": 1.9025134835585187e-05,
+      "loss": 2.0314,
+      "step": 976
+    },
+    {
+      "epoch": 0.1676677535610091,
+      "grad_norm": 2.9739894442464103,
+      "learning_rate": 1.902273966355077e-05,
+      "loss": 1.9101,
+      "step": 977
+    },
+    {
+      "epoch": 0.1678393684571821,
+      "grad_norm": 2.63201072002799,
+      "learning_rate": 1.9020341703894537e-05,
+      "loss": 1.4468,
+      "step": 978
+    },
+    {
+      "epoch": 0.16801098335335507,
+      "grad_norm": 2.739177288565023,
+      "learning_rate": 1.901794095735735e-05,
+      "loss": 1.7929,
+      "step": 979
+    },
+    {
+      "epoch": 0.16818259824952805,
+      "grad_norm": 3.045465062594323,
+      "learning_rate": 1.9015537424680933e-05,
+      "loss": 1.7605,
+      "step": 980
+    },
+    {
+      "epoch": 0.16835421314570104,
+      "grad_norm": 2.4326117471284188,
+      "learning_rate": 1.901313110660787e-05,
+      "loss": 1.7179,
+      "step": 981
+    },
+    {
+      "epoch": 0.16852582804187405,
+      "grad_norm": 2.4453337083070497,
+      "learning_rate": 1.9010722003881604e-05,
+      "loss": 1.6435,
+      "step": 982
+    },
+    {
+      "epoch": 0.16869744293804703,
+      "grad_norm": 4.61559448035971,
+      "learning_rate": 1.900831011724644e-05,
+      "loss": 2.024,
+      "step": 983
+    },
+    {
+      "epoch": 0.16886905783422002,
+      "grad_norm": 2.8038211548060734,
+      "learning_rate": 1.9005895447447543e-05,
+      "loss": 1.7816,
+      "step": 984
+    },
+    {
+      "epoch": 0.169040672730393,
+      "grad_norm": 2.3615213486971633,
+      "learning_rate": 1.9003477995230942e-05,
+      "loss": 1.8099,
+      "step": 985
+    },
+    {
+      "epoch": 0.16921228762656598,
+      "grad_norm": 2.440052416415365,
+      "learning_rate": 1.9001057761343512e-05,
+      "loss": 1.9574,
+      "step": 986
+    },
+    {
+      "epoch": 0.16938390252273897,
+      "grad_norm": 2.460941416495098,
+      "learning_rate": 1.8998634746533004e-05,
+      "loss": 1.8803,
+      "step": 987
+    },
+    {
+      "epoch": 0.16955551741891195,
+      "grad_norm": 4.451366701398478,
+      "learning_rate": 1.8996208951548022e-05,
+      "loss": 1.7882,
+      "step": 988
+    },
+    {
+      "epoch": 0.16972713231508496,
+      "grad_norm": 2.6603009507375437,
+      "learning_rate": 1.899378037713802e-05,
+      "loss": 1.7881,
+      "step": 989
+    },
+    {
+      "epoch": 0.16989874721125794,
+      "grad_norm": 2.7031088505383796,
+      "learning_rate": 1.8991349024053326e-05,
+      "loss": 1.693,
+      "step": 990
+    },
+    {
+      "epoch": 0.17007036210743093,
+      "grad_norm": 2.503672741774846,
+      "learning_rate": 1.898891489304511e-05,
+      "loss": 1.6861,
+      "step": 991
+    },
+    {
+      "epoch": 0.1702419770036039,
+      "grad_norm": 2.3955969538278223,
+      "learning_rate": 1.8986477984865424e-05,
+      "loss": 1.7797,
+      "step": 992
+    },
+    {
+      "epoch": 0.1704135918997769,
+      "grad_norm": 2.5312842171580057,
+      "learning_rate": 1.8984038300267153e-05,
+      "loss": 1.8429,
+      "step": 993
+    },
+    {
+      "epoch": 0.17058520679594988,
+      "grad_norm": 2.7008350326751676,
+      "learning_rate": 1.898159584000405e-05,
+      "loss": 1.7478,
+      "step": 994
+    },
+    {
+      "epoch": 0.1707568216921229,
+      "grad_norm": 3.2241673640984114,
+      "learning_rate": 1.8979150604830726e-05,
+      "loss": 1.9911,
+      "step": 995
+    },
+    {
+      "epoch": 0.17092843658829587,
+      "grad_norm": 3.08809992159637,
+      "learning_rate": 1.897670259550265e-05,
+      "loss": 1.9767,
+      "step": 996
+    },
+    {
+      "epoch": 0.17110005148446886,
+      "grad_norm": 2.2915437722533296,
+      "learning_rate": 1.8974251812776155e-05,
+      "loss": 1.8417,
+      "step": 997
+    },
+    {
+      "epoch": 0.17127166638064184,
+      "grad_norm": 3.2335440033475193,
+      "learning_rate": 1.8971798257408408e-05,
+      "loss": 1.9661,
+      "step": 998
+    },
+    {
+      "epoch": 0.17144328127681482,
+      "grad_norm": 2.5960898381278934,
+      "learning_rate": 1.8969341930157458e-05,
+      "loss": 1.8926,
+      "step": 999
+    },
+    {
+      "epoch": 0.1716148961729878,
+      "grad_norm": 2.595363848437771,
+      "learning_rate": 1.8966882831782197e-05,
+      "loss": 1.9053,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1717865110691608,
+      "grad_norm": 3.1288593573128125,
+      "learning_rate": 1.8964420963042375e-05,
+      "loss": 1.8298,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1719581259653338,
+      "grad_norm": 3.066352066925028,
+      "learning_rate": 1.8961956324698603e-05,
+      "loss": 1.7616,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17212974086150679,
+      "grad_norm": 2.205000671917753,
+      "learning_rate": 1.895948891751234e-05,
+      "loss": 1.6981,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17230135575767977,
+      "grad_norm": 2.3401231677521013,
+      "learning_rate": 1.8957018742245906e-05,
+      "loss": 1.8939,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17247297065385275,
+      "grad_norm": 2.6038708153000956,
+      "learning_rate": 1.895454579966247e-05,
+      "loss": 1.9144,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17264458555002574,
+      "grad_norm": 5.204708163513282,
+      "learning_rate": 1.8952070090526066e-05,
+      "loss": 1.7359,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17281620044619872,
+      "grad_norm": 2.7438381612121736,
+      "learning_rate": 1.894959161560157e-05,
+      "loss": 1.9809,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17298781534237173,
+      "grad_norm": 2.8151981464137448,
+      "learning_rate": 1.8947110375654732e-05,
+      "loss": 1.6311,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17315943023854471,
+      "grad_norm": 2.653602968512543,
+      "learning_rate": 1.894462637145213e-05,
+      "loss": 1.9819,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1733310451347177,
+      "grad_norm": 2.7489311926086932,
+      "learning_rate": 1.8942139603761222e-05,
+      "loss": 2.0842,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17350266003089068,
+      "grad_norm": 2.4668126904731174,
+      "learning_rate": 1.8939650073350296e-05,
+      "loss": 1.5547,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17367427492706367,
+      "grad_norm": 3.123375410077176,
+      "learning_rate": 1.8937157780988512e-05,
+      "loss": 1.9815,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17384588982323665,
+      "grad_norm": 2.8327118437264525,
+      "learning_rate": 1.8934662727445877e-05,
+      "loss": 1.7706,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17401750471940963,
+      "grad_norm": 2.3726618036391045,
+      "learning_rate": 1.893216491349325e-05,
+      "loss": 2.0324,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17418911961558264,
+      "grad_norm": 2.650378326824613,
+      "learning_rate": 1.8929664339902342e-05,
+      "loss": 1.8853,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17436073451175563,
+      "grad_norm": 2.497401438566898,
+      "learning_rate": 1.8927161007445715e-05,
+      "loss": 1.8044,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1745323494079286,
+      "grad_norm": 3.1593763637345225,
+      "learning_rate": 1.8924654916896793e-05,
+      "loss": 2.0135,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1747039643041016,
+      "grad_norm": 3.1305083044842155,
+      "learning_rate": 1.8922146069029843e-05,
+      "loss": 1.918,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17487557920027458,
+      "grad_norm": 2.739394354944574,
+      "learning_rate": 1.8919634464619985e-05,
+      "loss": 1.7404,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17504719409644756,
+      "grad_norm": 3.3322367314107,
+      "learning_rate": 1.8917120104443195e-05,
+      "loss": 1.7972,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17521880899262057,
+      "grad_norm": 2.9088292950504195,
+      "learning_rate": 1.8914602989276294e-05,
+      "loss": 2.0117,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17539042388879356,
+      "grad_norm": 3.219667799774691,
+      "learning_rate": 1.8912083119896956e-05,
+      "loss": 1.9661,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17556203878496654,
+      "grad_norm": 3.495996324018158,
+      "learning_rate": 1.8909560497083718e-05,
+      "loss": 1.8407,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17573365368113952,
+      "grad_norm": 9.414813858534298,
+      "learning_rate": 1.8907035121615946e-05,
+      "loss": 2.1415,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1759052685773125,
+      "grad_norm": 3.4640684834247084,
+      "learning_rate": 1.8904506994273872e-05,
+      "loss": 1.7853,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1760768834734855,
+      "grad_norm": 3.4016429899923306,
+      "learning_rate": 1.8901976115838576e-05,
+      "loss": 1.9636,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17624849836965847,
+      "grad_norm": 3.079043867385513,
+      "learning_rate": 1.889944248709198e-05,
+      "loss": 1.6901,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17642011326583149,
+      "grad_norm": 3.796691967016937,
+      "learning_rate": 1.889690610881687e-05,
+      "loss": 1.9453,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17659172816200447,
+      "grad_norm": 2.952102254205979,
+      "learning_rate": 1.8894366981796865e-05,
+      "loss": 1.8344,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17676334305817745,
+      "grad_norm": 3.029431693632875,
+      "learning_rate": 1.8891825106816444e-05,
+      "loss": 1.743,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17693495795435044,
+      "grad_norm": 3.4461112507548926,
+      "learning_rate": 1.888928048466094e-05,
+      "loss": 1.6788,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17710657285052342,
+      "grad_norm": 3.2071682047553898,
+      "learning_rate": 1.8886733116116515e-05,
+      "loss": 1.619,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1772781877466964,
+      "grad_norm": 3.206436516046294,
+      "learning_rate": 1.8884183001970197e-05,
+      "loss": 2.2403,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17744980264286941,
+      "grad_norm": 3.6350091791623265,
+      "learning_rate": 1.8881630143009856e-05,
+      "loss": 1.892,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1776214175390424,
+      "grad_norm": 2.8812949312018206,
+      "learning_rate": 1.8879074540024215e-05,
+      "loss": 1.8083,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17779303243521538,
+      "grad_norm": 2.589130333556869,
+      "learning_rate": 1.8876516193802838e-05,
+      "loss": 1.6781,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17796464733138836,
+      "grad_norm": 3.804334864048245,
+      "learning_rate": 1.8873955105136134e-05,
+      "loss": 1.7368,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17813626222756135,
+      "grad_norm": 3.2877881347476134,
+      "learning_rate": 1.8871391274815373e-05,
+      "loss": 1.8528,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17830787712373433,
+      "grad_norm": 3.7272616273923775,
+      "learning_rate": 1.8868824703632658e-05,
+      "loss": 1.7402,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17847949201990732,
+      "grad_norm": 2.585434900267579,
+      "learning_rate": 1.8866255392380945e-05,
+      "loss": 1.894,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17865110691608033,
+      "grad_norm": 2.3910608720654256,
+      "learning_rate": 1.886368334185404e-05,
+      "loss": 1.8734,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1788227218122533,
+      "grad_norm": 4.566912535500701,
+      "learning_rate": 1.8861108552846587e-05,
+      "loss": 2.2143,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1789943367084263,
+      "grad_norm": 2.7544536727928857,
+      "learning_rate": 1.8858531026154077e-05,
+      "loss": 2.0066,
+      "step": 1043
+    },
+    {
+      "epoch": 0.17916595160459928,
+      "grad_norm": 4.827569095465524,
+      "learning_rate": 1.885595076257286e-05,
+      "loss": 1.7604,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17933756650077226,
+      "grad_norm": 3.556294587132145,
+      "learning_rate": 1.8853367762900117e-05,
+      "loss": 1.7561,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17950918139694524,
+      "grad_norm": 13.180361400090435,
+      "learning_rate": 1.8850782027933876e-05,
+      "loss": 1.986,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17968079629311826,
+      "grad_norm": 2.679281925560701,
+      "learning_rate": 1.8848193558473014e-05,
+      "loss": 1.828,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17985241118929124,
+      "grad_norm": 2.4857864808750425,
+      "learning_rate": 1.8845602355317257e-05,
+      "loss": 1.7873,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18002402608546422,
+      "grad_norm": 2.573732073133298,
+      "learning_rate": 1.8843008419267164e-05,
+      "loss": 1.8803,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1801956409816372,
+      "grad_norm": 2.6001052561724993,
+      "learning_rate": 1.8840411751124152e-05,
+      "loss": 2.0575,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1803672558778102,
+      "grad_norm": 3.947639652408305,
+      "learning_rate": 1.883781235169047e-05,
+      "loss": 1.8958,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18053887077398317,
+      "grad_norm": 2.348621344786185,
+      "learning_rate": 1.8835210221769214e-05,
+      "loss": 1.5452,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18071048567015616,
+      "grad_norm": 3.324274276550934,
+      "learning_rate": 1.8832605362164326e-05,
+      "loss": 1.6412,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18088210056632917,
+      "grad_norm": 2.6759678165627534,
+      "learning_rate": 1.8829997773680598e-05,
+      "loss": 1.7532,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18105371546250215,
+      "grad_norm": 2.6371836469673315,
+      "learning_rate": 1.882738745712365e-05,
+      "loss": 1.9407,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18122533035867514,
+      "grad_norm": 2.674592344303099,
+      "learning_rate": 1.8824774413299954e-05,
+      "loss": 1.9499,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18139694525484812,
+      "grad_norm": 2.912062735091322,
+      "learning_rate": 1.882215864301683e-05,
+      "loss": 1.7272,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1815685601510211,
+      "grad_norm": 2.8826820866102403,
+      "learning_rate": 1.8819540147082424e-05,
+      "loss": 1.8309,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18174017504719409,
+      "grad_norm": 4.750127936117877,
+      "learning_rate": 1.881691892630574e-05,
+      "loss": 2.1138,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1819117899433671,
+      "grad_norm": 3.036601337031838,
+      "learning_rate": 1.8814294981496616e-05,
+      "loss": 1.8791,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18208340483954008,
+      "grad_norm": 2.965717773074859,
+      "learning_rate": 1.8811668313465736e-05,
+      "loss": 2.0173,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18225501973571306,
+      "grad_norm": 3.2967787137875204,
+      "learning_rate": 1.8809038923024613e-05,
+      "loss": 1.7907,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18242663463188605,
+      "grad_norm": 2.77407663563314,
+      "learning_rate": 1.8806406810985624e-05,
+      "loss": 1.6843,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18259824952805903,
+      "grad_norm": 3.816663346127986,
+      "learning_rate": 1.8803771978161964e-05,
+      "loss": 1.7359,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18276986442423201,
+      "grad_norm": 2.5044013306095363,
+      "learning_rate": 1.8801134425367676e-05,
+      "loss": 1.7789,
+      "step": 1065
+    },
+    {
+      "epoch": 0.182941479320405,
+      "grad_norm": 2.930811755950049,
+      "learning_rate": 1.8798494153417657e-05,
+      "loss": 1.955,
+      "step": 1066
+    },
+    {
+      "epoch": 0.183113094216578,
+      "grad_norm": 3.0432956618503297,
+      "learning_rate": 1.8795851163127626e-05,
+      "loss": 1.7762,
+      "step": 1067
+    },
+    {
+      "epoch": 0.183284709112751,
+      "grad_norm": 3.6274248005457004,
+      "learning_rate": 1.8793205455314147e-05,
+      "loss": 1.8944,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18345632400892398,
+      "grad_norm": 4.374457586779356,
+      "learning_rate": 1.8790557030794627e-05,
+      "loss": 1.8761,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18362793890509696,
+      "grad_norm": 2.829705551125168,
+      "learning_rate": 1.878790589038731e-05,
+      "loss": 1.7654,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18379955380126994,
+      "grad_norm": 2.5614000331429962,
+      "learning_rate": 1.878525203491128e-05,
+      "loss": 1.8806,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18397116869744293,
+      "grad_norm": 2.6451778633835787,
+      "learning_rate": 1.878259546518646e-05,
+      "loss": 1.9004,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18414278359361594,
+      "grad_norm": 3.1207731834877106,
+      "learning_rate": 1.8779936182033606e-05,
+      "loss": 1.5279,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18431439848978892,
+      "grad_norm": 2.869087511211109,
+      "learning_rate": 1.8777274186274325e-05,
+      "loss": 1.7406,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1844860133859619,
+      "grad_norm": 2.5419712105170182,
+      "learning_rate": 1.8774609478731048e-05,
+      "loss": 1.7459,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1846576282821349,
+      "grad_norm": 2.434385372131046,
+      "learning_rate": 1.8771942060227052e-05,
+      "loss": 1.8075,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18482924317830787,
+      "grad_norm": 2.7126029989324043,
+      "learning_rate": 1.876927193158645e-05,
+      "loss": 1.8576,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18500085807448086,
+      "grad_norm": 3.1373612003072893,
+      "learning_rate": 1.876659909363419e-05,
+      "loss": 1.9765,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18517247297065384,
+      "grad_norm": 2.127902015337168,
+      "learning_rate": 1.8763923547196065e-05,
+      "loss": 1.7472,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18534408786682685,
+      "grad_norm": 3.987208635479881,
+      "learning_rate": 1.8761245293098695e-05,
+      "loss": 1.7846,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18551570276299983,
+      "grad_norm": 2.589976903982755,
+      "learning_rate": 1.875856433216954e-05,
+      "loss": 1.7642,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18568731765917282,
+      "grad_norm": 3.288323301571295,
+      "learning_rate": 1.8755880665236895e-05,
+      "loss": 1.9984,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1858589325553458,
+      "grad_norm": 2.488300473967599,
+      "learning_rate": 1.8753194293129892e-05,
+      "loss": 1.706,
+      "step": 1083
+    },
+    {
+      "epoch": 0.18603054745151879,
+      "grad_norm": 4.019839208045935,
+      "learning_rate": 1.8750505216678505e-05,
+      "loss": 1.8421,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18620216234769177,
+      "grad_norm": 3.2029714550456485,
+      "learning_rate": 1.8747813436713534e-05,
+      "loss": 1.7128,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18637377724386478,
+      "grad_norm": 3.2149925112779614,
+      "learning_rate": 1.874511895406662e-05,
+      "loss": 1.9479,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18654539214003776,
+      "grad_norm": 2.39394180672311,
+      "learning_rate": 1.8742421769570235e-05,
+      "loss": 1.4238,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18671700703621075,
+      "grad_norm": 2.8354364046359386,
+      "learning_rate": 1.8739721884057693e-05,
+      "loss": 2.0304,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18688862193238373,
+      "grad_norm": 2.365789454238505,
+      "learning_rate": 1.873701929836313e-05,
+      "loss": 1.8165,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18706023682855671,
+      "grad_norm": 2.4695820556846306,
+      "learning_rate": 1.873431401332153e-05,
+      "loss": 1.6697,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1872318517247297,
+      "grad_norm": 2.3797983670935396,
+      "learning_rate": 1.8731606029768704e-05,
+      "loss": 1.536,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18740346662090268,
+      "grad_norm": 2.2859708842423925,
+      "learning_rate": 1.8728895348541294e-05,
+      "loss": 1.9074,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1875750815170757,
+      "grad_norm": 2.796711393950907,
+      "learning_rate": 1.872618197047678e-05,
+      "loss": 2.0874,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18774669641324868,
+      "grad_norm": 2.692776743704972,
+      "learning_rate": 1.8723465896413476e-05,
+      "loss": 1.9903,
+      "step": 1094
+    },
+    {
+      "epoch": 0.18791831130942166,
+      "grad_norm": 2.9551043800793715,
+      "learning_rate": 1.872074712719053e-05,
+      "loss": 1.9889,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18808992620559464,
+      "grad_norm": 3.269566715140991,
+      "learning_rate": 1.871802566364791e-05,
+      "loss": 1.7872,
+      "step": 1096
+    },
+    {
+      "epoch": 0.18826154110176763,
+      "grad_norm": 2.2442935183112196,
+      "learning_rate": 1.8715301506626435e-05,
+      "loss": 1.7215,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1884331559979406,
+      "grad_norm": 3.826784809565196,
+      "learning_rate": 1.8712574656967743e-05,
+      "loss": 1.8555,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18860477089411362,
+      "grad_norm": 2.4230380154716307,
+      "learning_rate": 1.8709845115514306e-05,
+      "loss": 1.6155,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1887763857902866,
+      "grad_norm": 3.7052796319842707,
+      "learning_rate": 1.8707112883109438e-05,
+      "loss": 1.6623,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1889480006864596,
+      "grad_norm": 2.308489916424168,
+      "learning_rate": 1.870437796059727e-05,
+      "loss": 1.6847,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18911961558263257,
+      "grad_norm": 3.6337660185440246,
+      "learning_rate": 1.870164034882277e-05,
+      "loss": 2.1511,
+      "step": 1102
+    },
+    {
+      "epoch": 0.18929123047880556,
+      "grad_norm": 3.886332399848284,
+      "learning_rate": 1.8698900048631744e-05,
+      "loss": 1.627,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18946284537497854,
+      "grad_norm": 2.758130505540072,
+      "learning_rate": 1.8696157060870815e-05,
+      "loss": 1.6426,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18963446027115152,
+      "grad_norm": 2.978496918541371,
+      "learning_rate": 1.8693411386387445e-05,
+      "loss": 1.7926,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18980607516732453,
+      "grad_norm": 2.6291954597409717,
+      "learning_rate": 1.869066302602992e-05,
+      "loss": 1.7945,
+      "step": 1106
+    },
+    {
+      "epoch": 0.18997769006349752,
+      "grad_norm": 2.6400358893561733,
+      "learning_rate": 1.8687911980647375e-05,
+      "loss": 2.0097,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1901493049596705,
+      "grad_norm": 2.5078371492775817,
+      "learning_rate": 1.868515825108974e-05,
+      "loss": 1.8463,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19032091985584348,
+      "grad_norm": 3.2672353956382456,
+      "learning_rate": 1.8682401838207808e-05,
+      "loss": 1.988,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19049253475201647,
+      "grad_norm": 2.684958790285497,
+      "learning_rate": 1.8679642742853178e-05,
+      "loss": 1.7406,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19066414964818945,
+      "grad_norm": 3.1484450070703427,
+      "learning_rate": 1.8676880965878293e-05,
+      "loss": 1.8318,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19083576454436246,
+      "grad_norm": 2.6696280789216167,
+      "learning_rate": 1.8674116508136415e-05,
+      "loss": 2.0488,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19100737944053545,
+      "grad_norm": 2.631949173273026,
+      "learning_rate": 1.8671349370481636e-05,
+      "loss": 1.9126,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19117899433670843,
+      "grad_norm": 2.887587241068506,
+      "learning_rate": 1.8668579553768882e-05,
+      "loss": 1.9419,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1913506092328814,
+      "grad_norm": 3.1809261401885265,
+      "learning_rate": 1.86658070588539e-05,
+      "loss": 1.9567,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1915222241290544,
+      "grad_norm": 3.348456752055784,
+      "learning_rate": 1.8663031886593268e-05,
+      "loss": 1.8364,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19169383902522738,
+      "grad_norm": 2.5074685817563416,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.7228,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19186545392140036,
+      "grad_norm": 2.4983148708824325,
+      "learning_rate": 1.865747351346549e-05,
+      "loss": 1.7511,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19203706881757338,
+      "grad_norm": 2.4450778760072547,
+      "learning_rate": 1.865469031431563e-05,
+      "loss": 1.7601,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19220868371374636,
+      "grad_norm": 2.802171577959105,
+      "learning_rate": 1.8651904441254696e-05,
+      "loss": 2.0065,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19238029860991934,
+      "grad_norm": 2.5400839595023994,
+      "learning_rate": 1.8649115895143394e-05,
+      "loss": 1.7311,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19255191350609233,
+      "grad_norm": 3.1573210941620284,
+      "learning_rate": 1.8646324676843263e-05,
+      "loss": 1.7388,
+      "step": 1122
+    },
+    {
+      "epoch": 0.1927235284022653,
+      "grad_norm": 2.196387772358243,
+      "learning_rate": 1.864353078721666e-05,
+      "loss": 1.7639,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1928951432984383,
+      "grad_norm": 3.3422846022968375,
+      "learning_rate": 1.8640734227126778e-05,
+      "loss": 1.8086,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1930667581946113,
+      "grad_norm": 2.468942421946988,
+      "learning_rate": 1.8637934997437623e-05,
+      "loss": 1.6612,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1932383730907843,
+      "grad_norm": 2.0826885772486308,
+      "learning_rate": 1.8635133099014032e-05,
+      "loss": 1.8644,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19340998798695727,
+      "grad_norm": 2.4326412367426373,
+      "learning_rate": 1.8632328532721665e-05,
+      "loss": 1.931,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19358160288313025,
+      "grad_norm": 2.6605196466013163,
+      "learning_rate": 1.862952129942701e-05,
+      "loss": 1.6765,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19375321777930324,
+      "grad_norm": 2.979635303821038,
+      "learning_rate": 1.862671139999738e-05,
+      "loss": 1.5364,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19392483267547622,
+      "grad_norm": 5.439138700619377,
+      "learning_rate": 1.86238988353009e-05,
+      "loss": 1.7336,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1940964475716492,
+      "grad_norm": 2.4756700631713473,
+      "learning_rate": 1.862108360620653e-05,
+      "loss": 1.971,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19426806246782222,
+      "grad_norm": 2.7693302346575805,
+      "learning_rate": 1.8618265713584047e-05,
+      "loss": 1.9736,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1944396773639952,
+      "grad_norm": 3.188145891790364,
+      "learning_rate": 1.861544515830406e-05,
+      "loss": 1.8882,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19461129226016818,
+      "grad_norm": 3.2927482288339642,
+      "learning_rate": 1.8612621941237986e-05,
+      "loss": 1.9117,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19478290715634117,
+      "grad_norm": 2.6877567394223614,
+      "learning_rate": 1.8609796063258076e-05,
+      "loss": 1.6008,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19495452205251415,
+      "grad_norm": 2.3054717979221953,
+      "learning_rate": 1.86069675252374e-05,
+      "loss": 1.6678,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19512613694868713,
+      "grad_norm": 3.268207156742372,
+      "learning_rate": 1.860413632804985e-05,
+      "loss": 1.8276,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19529775184486015,
+      "grad_norm": 3.3650128262023067,
+      "learning_rate": 1.8601302472570137e-05,
+      "loss": 1.9564,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19546936674103313,
+      "grad_norm": 3.13641456636503,
+      "learning_rate": 1.8598465959673793e-05,
+      "loss": 1.6997,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1956409816372061,
+      "grad_norm": 2.4888234966190854,
+      "learning_rate": 1.859562679023718e-05,
+      "loss": 1.7335,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1958125965333791,
+      "grad_norm": 2.326502908304316,
+      "learning_rate": 1.859278496513747e-05,
+      "loss": 1.9794,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19598421142955208,
+      "grad_norm": 3.791019241688972,
+      "learning_rate": 1.858994048525266e-05,
+      "loss": 1.9333,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19615582632572506,
+      "grad_norm": 3.445224678397263,
+      "learning_rate": 1.8587093351461564e-05,
+      "loss": 1.8693,
+      "step": 1143
+    },
+    {
+      "epoch": 0.19632744122189805,
+      "grad_norm": 2.188082419713291,
+      "learning_rate": 1.858424356464382e-05,
+      "loss": 1.6474,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19649905611807106,
+      "grad_norm": 2.935323731316312,
+      "learning_rate": 1.858139112567989e-05,
+      "loss": 1.6986,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19667067101424404,
+      "grad_norm": 5.780751942109886,
+      "learning_rate": 1.8578536035451043e-05,
+      "loss": 1.4593,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19684228591041703,
+      "grad_norm": 2.884530891955222,
+      "learning_rate": 1.857567829483937e-05,
+      "loss": 1.9083,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19701390080659,
+      "grad_norm": 3.1345637189603353,
+      "learning_rate": 1.85728179047278e-05,
+      "loss": 1.6807,
+      "step": 1148
+    },
+    {
+      "epoch": 0.197185515702763,
+      "grad_norm": 3.1244632897824394,
+      "learning_rate": 1.8569954866000052e-05,
+      "loss": 1.7671,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19735713059893598,
+      "grad_norm": 2.4182516923417046,
+      "learning_rate": 1.856708917954068e-05,
+      "loss": 1.4882,
+      "step": 1150
+    },
+    {
+      "epoch": 0.197528745495109,
+      "grad_norm": 2.8595853140129854,
+      "learning_rate": 1.856422084623505e-05,
+      "loss": 1.9436,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19770036039128197,
+      "grad_norm": 2.4916865800570274,
+      "learning_rate": 1.8561349866969357e-05,
+      "loss": 1.5929,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19787197528745495,
+      "grad_norm": 2.7711134221112976,
+      "learning_rate": 1.8558476242630595e-05,
+      "loss": 1.6756,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19804359018362794,
+      "grad_norm": 3.822574826162146,
+      "learning_rate": 1.855559997410659e-05,
+      "loss": 1.9019,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19821520507980092,
+      "grad_norm": 2.1176687607265112,
+      "learning_rate": 1.855272106228598e-05,
+      "loss": 1.5265,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1983868199759739,
+      "grad_norm": 2.9471459296488343,
+      "learning_rate": 1.8549839508058215e-05,
+      "loss": 1.6747,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19855843487214692,
+      "grad_norm": 3.3245797370321943,
+      "learning_rate": 1.8546955312313574e-05,
+      "loss": 1.7418,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1987300497683199,
+      "grad_norm": 3.7590341939006287,
+      "learning_rate": 1.8544068475943134e-05,
+      "loss": 2.1677,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19890166466449288,
+      "grad_norm": 2.684332707940541,
+      "learning_rate": 1.854117899983881e-05,
+      "loss": 1.8058,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19907327956066587,
+      "grad_norm": 2.1433615192323,
+      "learning_rate": 1.8538286884893308e-05,
+      "loss": 1.3951,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19924489445683885,
+      "grad_norm": 2.493613543357034,
+      "learning_rate": 1.853539213200017e-05,
+      "loss": 1.5709,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19941650935301183,
+      "grad_norm": 2.6739102443974967,
+      "learning_rate": 1.8532494742053747e-05,
+      "loss": 2.0268,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19958812424918482,
+      "grad_norm": 2.5447858904373395,
+      "learning_rate": 1.8529594715949194e-05,
+      "loss": 2.0656,
+      "step": 1163
+    },
+    {
+      "epoch": 0.19975973914535783,
+      "grad_norm": 2.7842776677002115,
+      "learning_rate": 1.8526692054582493e-05,
+      "loss": 1.7685,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1999313540415308,
+      "grad_norm": 2.5921572243031625,
+      "learning_rate": 1.8523786758850436e-05,
+      "loss": 1.9377,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2001029689377038,
+      "grad_norm": 2.7408247723424535,
+      "learning_rate": 1.852087882965063e-05,
+      "loss": 1.8249,
+      "step": 1166
+    },
+    {
+      "epoch": 0.20027458383387678,
+      "grad_norm": 3.2137183902875375,
+      "learning_rate": 1.8517968267881495e-05,
+      "loss": 1.9897,
+      "step": 1167
+    },
+    {
+      "epoch": 0.20044619873004976,
+      "grad_norm": 5.345600159412697,
+      "learning_rate": 1.851505507444226e-05,
+      "loss": 1.9508,
+      "step": 1168
+    },
+    {
+      "epoch": 0.20061781362622275,
+      "grad_norm": 3.193065256899719,
+      "learning_rate": 1.851213925023298e-05,
+      "loss": 2.0566,
+      "step": 1169
+    },
+    {
+      "epoch": 0.20078942852239576,
+      "grad_norm": 3.506086311061711,
+      "learning_rate": 1.8509220796154504e-05,
+      "loss": 1.9816,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20096104341856874,
+      "grad_norm": 2.8444637855886468,
+      "learning_rate": 1.850629971310851e-05,
+      "loss": 1.9854,
+      "step": 1171
+    },
+    {
+      "epoch": 0.20113265831474172,
+      "grad_norm": 2.443856784353825,
+      "learning_rate": 1.8503376001997474e-05,
+      "loss": 1.7351,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2013042732109147,
+      "grad_norm": 3.746737368687077,
+      "learning_rate": 1.85004496637247e-05,
+      "loss": 1.7701,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2014758881070877,
+      "grad_norm": 2.711946987588241,
+      "learning_rate": 1.8497520699194295e-05,
+      "loss": 1.8346,
+      "step": 1174
+    },
+    {
+      "epoch": 0.20164750300326068,
+      "grad_norm": 2.284625633085561,
+      "learning_rate": 1.8494589109311168e-05,
+      "loss": 1.5929,
+      "step": 1175
+    },
+    {
+      "epoch": 0.20181911789943366,
+      "grad_norm": 2.3956770077097667,
+      "learning_rate": 1.849165489498106e-05,
+      "loss": 1.8978,
+      "step": 1176
+    },
+    {
+      "epoch": 0.20199073279560667,
+      "grad_norm": 2.8082931937038174,
+      "learning_rate": 1.8488718057110503e-05,
+      "loss": 1.7675,
+      "step": 1177
+    },
+    {
+      "epoch": 0.20216234769177965,
+      "grad_norm": 2.8546909357049204,
+      "learning_rate": 1.8485778596606855e-05,
+      "loss": 2.0379,
+      "step": 1178
+    },
+    {
+      "epoch": 0.20233396258795264,
+      "grad_norm": 4.375546502142223,
+      "learning_rate": 1.8482836514378278e-05,
+      "loss": 1.9957,
+      "step": 1179
+    },
+    {
+      "epoch": 0.20250557748412562,
+      "grad_norm": 2.737635853754392,
+      "learning_rate": 1.8479891811333735e-05,
+      "loss": 1.6637,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2026771923802986,
+      "grad_norm": 2.54730197947167,
+      "learning_rate": 1.8476944488383012e-05,
+      "loss": 1.7397,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2028488072764716,
+      "grad_norm": 2.988740542354885,
+      "learning_rate": 1.8473994546436702e-05,
+      "loss": 1.672,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2030204221726446,
+      "grad_norm": 2.931000226684086,
+      "learning_rate": 1.84710419864062e-05,
+      "loss": 1.7761,
+      "step": 1183
+    },
+    {
+      "epoch": 0.20319203706881758,
+      "grad_norm": 2.4484389732536855,
+      "learning_rate": 1.846808680920372e-05,
+      "loss": 1.8635,
+      "step": 1184
+    },
+    {
+      "epoch": 0.20336365196499057,
+      "grad_norm": 2.346915203819909,
+      "learning_rate": 1.8465129015742273e-05,
+      "loss": 1.6783,
+      "step": 1185
+    },
+    {
+      "epoch": 0.20353526686116355,
+      "grad_norm": 2.483798335639931,
+      "learning_rate": 1.846216860693569e-05,
+      "loss": 1.8088,
+      "step": 1186
+    },
+    {
+      "epoch": 0.20370688175733653,
+      "grad_norm": 2.16574537927803,
+      "learning_rate": 1.8459205583698598e-05,
+      "loss": 1.665,
+      "step": 1187
+    },
+    {
+      "epoch": 0.20387849665350952,
+      "grad_norm": 2.2468116921548233,
+      "learning_rate": 1.8456239946946448e-05,
+      "loss": 1.5618,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2040501115496825,
+      "grad_norm": 2.3846735595775996,
+      "learning_rate": 1.845327169759548e-05,
+      "loss": 1.7469,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2042217264458555,
+      "grad_norm": 2.700436306282447,
+      "learning_rate": 1.8450300836562755e-05,
+      "loss": 1.9616,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2043933413420285,
+      "grad_norm": 2.538884538732009,
+      "learning_rate": 1.8447327364766135e-05,
+      "loss": 1.79,
+      "step": 1191
+    },
+    {
+      "epoch": 0.20456495623820148,
+      "grad_norm": 2.2957387150236483,
+      "learning_rate": 1.8444351283124288e-05,
+      "loss": 1.7263,
+      "step": 1192
+    },
+    {
+      "epoch": 0.20473657113437446,
+      "grad_norm": 2.584075714012207,
+      "learning_rate": 1.8441372592556687e-05,
+      "loss": 1.85,
+      "step": 1193
+    },
+    {
+      "epoch": 0.20490818603054745,
+      "grad_norm": 2.5752319438934306,
+      "learning_rate": 1.8438391293983616e-05,
+      "loss": 1.6855,
+      "step": 1194
+    },
+    {
+      "epoch": 0.20507980092672043,
+      "grad_norm": 3.1925834615752176,
+      "learning_rate": 1.8435407388326167e-05,
+      "loss": 2.0197,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20525141582289344,
+      "grad_norm": 2.8515077395439414,
+      "learning_rate": 1.8432420876506226e-05,
+      "loss": 1.8082,
+      "step": 1196
+    },
+    {
+      "epoch": 0.20542303071906642,
+      "grad_norm": 2.7004022660370786,
+      "learning_rate": 1.84294317594465e-05,
+      "loss": 1.7037,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2055946456152394,
+      "grad_norm": 3.4536004497020576,
+      "learning_rate": 1.842644003807048e-05,
+      "loss": 1.5347,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2057662605114124,
+      "grad_norm": 2.887357245320869,
+      "learning_rate": 1.8423445713302485e-05,
+      "loss": 1.7538,
+      "step": 1199
+    },
+    {
+      "epoch": 0.20593787540758537,
+      "grad_norm": 2.6758799097048067,
+      "learning_rate": 1.842044878606762e-05,
+      "loss": 1.8117,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20610949030375836,
+      "grad_norm": 2.962525042169495,
+      "learning_rate": 1.8417449257291802e-05,
+      "loss": 1.9437,
+      "step": 1201
+    },
+    {
+      "epoch": 0.20628110519993134,
+      "grad_norm": 3.9939244474141806,
+      "learning_rate": 1.8414447127901758e-05,
+      "loss": 1.805,
+      "step": 1202
+    },
+    {
+      "epoch": 0.20645272009610435,
+      "grad_norm": 3.063423974601498,
+      "learning_rate": 1.8411442398825002e-05,
+      "loss": 1.8419,
+      "step": 1203
+    },
+    {
+      "epoch": 0.20662433499227734,
+      "grad_norm": 2.8047779634491796,
+      "learning_rate": 1.8408435070989866e-05,
+      "loss": 1.7394,
+      "step": 1204
+    },
+    {
+      "epoch": 0.20679594988845032,
+      "grad_norm": 5.474388927535528,
+      "learning_rate": 1.8405425145325473e-05,
+      "loss": 1.6637,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2069675647846233,
+      "grad_norm": 2.552992855264127,
+      "learning_rate": 1.8402412622761762e-05,
+      "loss": 1.8176,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2071391796807963,
+      "grad_norm": 2.71903146785116,
+      "learning_rate": 1.8399397504229464e-05,
+      "loss": 1.7233,
+      "step": 1207
+    },
+    {
+      "epoch": 0.20731079457696927,
+      "grad_norm": 3.527684060923723,
+      "learning_rate": 1.8396379790660118e-05,
+      "loss": 1.7322,
+      "step": 1208
+    },
+    {
+      "epoch": 0.20748240947314228,
+      "grad_norm": 2.49097674953974,
+      "learning_rate": 1.839335948298606e-05,
+      "loss": 1.8015,
+      "step": 1209
+    },
+    {
+      "epoch": 0.20765402436931527,
+      "grad_norm": 2.3350030981195875,
+      "learning_rate": 1.839033658214043e-05,
+      "loss": 1.7555,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20782563926548825,
+      "grad_norm": 2.8352093798802587,
+      "learning_rate": 1.838731108905717e-05,
+      "loss": 1.9105,
+      "step": 1211
+    },
+    {
+      "epoch": 0.20799725416166123,
+      "grad_norm": 2.9389040097425534,
+      "learning_rate": 1.8384283004671014e-05,
+      "loss": 1.7127,
+      "step": 1212
+    },
+    {
+      "epoch": 0.20816886905783422,
+      "grad_norm": 2.5682400122707616,
+      "learning_rate": 1.8381252329917515e-05,
+      "loss": 1.7001,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2083404839540072,
+      "grad_norm": 2.7530152665991356,
+      "learning_rate": 1.837821906573301e-05,
+      "loss": 1.8195,
+      "step": 1214
+    },
+    {
+      "epoch": 0.20851209885018018,
+      "grad_norm": 2.336997282502789,
+      "learning_rate": 1.8375183213054644e-05,
+      "loss": 1.8306,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2086837137463532,
+      "grad_norm": 2.411554579117792,
+      "learning_rate": 1.837214477282036e-05,
+      "loss": 1.9118,
+      "step": 1216
+    },
+    {
+      "epoch": 0.20885532864252618,
+      "grad_norm": 2.5506688469463734,
+      "learning_rate": 1.8369103745968894e-05,
+      "loss": 1.8777,
+      "step": 1217
+    },
+    {
+      "epoch": 0.20902694353869916,
+      "grad_norm": 2.138704374144811,
+      "learning_rate": 1.8366060133439793e-05,
+      "loss": 1.5769,
+      "step": 1218
+    },
+    {
+      "epoch": 0.20919855843487214,
+      "grad_norm": 2.5674488506950803,
+      "learning_rate": 1.8363013936173393e-05,
+      "loss": 1.8477,
+      "step": 1219
+    },
+    {
+      "epoch": 0.20937017333104513,
+      "grad_norm": 2.295409175087095,
+      "learning_rate": 1.8359965155110836e-05,
+      "loss": 1.9262,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2095417882272181,
+      "grad_norm": 2.763266998801072,
+      "learning_rate": 1.835691379119405e-05,
+      "loss": 1.9518,
+      "step": 1221
+    },
+    {
+      "epoch": 0.20971340312339112,
+      "grad_norm": 3.432923194789523,
+      "learning_rate": 1.835385984536578e-05,
+      "loss": 1.7289,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2098850180195641,
+      "grad_norm": 2.547176472872316,
+      "learning_rate": 1.8350803318569554e-05,
+      "loss": 1.6852,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2100566329157371,
+      "grad_norm": 4.619370165881908,
+      "learning_rate": 1.8347744211749703e-05,
+      "loss": 1.8217,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21022824781191007,
+      "grad_norm": 2.4645850233578805,
+      "learning_rate": 1.834468252585135e-05,
+      "loss": 2.0458,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21039986270808306,
+      "grad_norm": 3.447385594195895,
+      "learning_rate": 1.8341618261820425e-05,
+      "loss": 1.5174,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21057147760425604,
+      "grad_norm": 2.9922073043169095,
+      "learning_rate": 1.833855142060364e-05,
+      "loss": 1.8422,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21074309250042902,
+      "grad_norm": 3.1440995019189537,
+      "learning_rate": 1.8335482003148518e-05,
+      "loss": 1.6751,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21091470739660204,
+      "grad_norm": 2.650246113047022,
+      "learning_rate": 1.8332410010403365e-05,
+      "loss": 1.5431,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21108632229277502,
+      "grad_norm": 2.74448628429144,
+      "learning_rate": 1.83293354433173e-05,
+      "loss": 1.7961,
+      "step": 1230
+    },
+    {
+      "epoch": 0.211257937188948,
+      "grad_norm": 5.173773725496545,
+      "learning_rate": 1.8326258302840214e-05,
+      "loss": 2.0131,
+      "step": 1231
+    },
+    {
+      "epoch": 0.211429552085121,
+      "grad_norm": 2.7630293480019543,
+      "learning_rate": 1.832317858992281e-05,
+      "loss": 1.8199,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21160116698129397,
+      "grad_norm": 2.7474099777145344,
+      "learning_rate": 1.8320096305516585e-05,
+      "loss": 1.7734,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21177278187746695,
+      "grad_norm": 2.907300185016793,
+      "learning_rate": 1.831701145057382e-05,
+      "loss": 1.854,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21194439677363996,
+      "grad_norm": 3.2163083969198385,
+      "learning_rate": 1.83139240260476e-05,
+      "loss": 1.9678,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21211601166981295,
+      "grad_norm": 2.7971666941663,
+      "learning_rate": 1.8310834032891804e-05,
+      "loss": 1.7643,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21228762656598593,
+      "grad_norm": 3.1159513156006846,
+      "learning_rate": 1.8307741472061097e-05,
+      "loss": 1.7535,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21245924146215892,
+      "grad_norm": 2.9782603015278504,
+      "learning_rate": 1.8304646344510943e-05,
+      "loss": 1.7584,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2126308563583319,
+      "grad_norm": 3.1394862566183743,
+      "learning_rate": 1.8301548651197597e-05,
+      "loss": 1.7506,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21280247125450488,
+      "grad_norm": 2.5306688690412233,
+      "learning_rate": 1.8298448393078106e-05,
+      "loss": 1.7415,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21297408615067787,
+      "grad_norm": 2.223856789977762,
+      "learning_rate": 1.8295345571110312e-05,
+      "loss": 1.525,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21314570104685088,
+      "grad_norm": 2.6192094911500874,
+      "learning_rate": 1.829224018625285e-05,
+      "loss": 1.7954,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21331731594302386,
+      "grad_norm": 2.356543372228777,
+      "learning_rate": 1.828913223946514e-05,
+      "loss": 1.5326,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21348893083919684,
+      "grad_norm": 3.0237846537149617,
+      "learning_rate": 1.8286021731707404e-05,
+      "loss": 1.6203,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21366054573536983,
+      "grad_norm": 2.7389379975079104,
+      "learning_rate": 1.8282908663940646e-05,
+      "loss": 1.6309,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2138321606315428,
+      "grad_norm": 2.4741614224392046,
+      "learning_rate": 1.8279793037126662e-05,
+      "loss": 1.5894,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2140037755277158,
+      "grad_norm": 2.8590232662588333,
+      "learning_rate": 1.8276674852228044e-05,
+      "loss": 1.8544,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2141753904238888,
+      "grad_norm": 2.3960262226820928,
+      "learning_rate": 1.827355411020817e-05,
+      "loss": 1.7588,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2143470053200618,
+      "grad_norm": 3.3339359266518667,
+      "learning_rate": 1.827043081203121e-05,
+      "loss": 1.8232,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21451862021623477,
+      "grad_norm": 2.8951852696813556,
+      "learning_rate": 1.8267304958662126e-05,
+      "loss": 1.7857,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21469023511240776,
+      "grad_norm": 3.582896282984653,
+      "learning_rate": 1.826417655106666e-05,
+      "loss": 1.8604,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21486185000858074,
+      "grad_norm": 2.693191603662857,
+      "learning_rate": 1.826104559021135e-05,
+      "loss": 1.7705,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21503346490475372,
+      "grad_norm": 2.5019122964623204,
+      "learning_rate": 1.825791207706353e-05,
+      "loss": 1.6929,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2152050798009267,
+      "grad_norm": 2.3388420194713175,
+      "learning_rate": 1.8254776012591312e-05,
+      "loss": 1.7999,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21537669469709972,
+      "grad_norm": 2.7547199322143108,
+      "learning_rate": 1.8251637397763597e-05,
+      "loss": 1.8926,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2155483095932727,
+      "grad_norm": 2.348766823611328,
+      "learning_rate": 1.8248496233550076e-05,
+      "loss": 1.8291,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21571992448944569,
+      "grad_norm": 1.9923310630157454,
+      "learning_rate": 1.824535252092123e-05,
+      "loss": 1.7714,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21589153938561867,
+      "grad_norm": 2.2958424357813274,
+      "learning_rate": 1.8242206260848322e-05,
+      "loss": 1.7504,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21606315428179165,
+      "grad_norm": 2.9922646790827128,
+      "learning_rate": 1.8239057454303406e-05,
+      "loss": 1.9608,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21623476917796464,
+      "grad_norm": 2.586994668504713,
+      "learning_rate": 1.8235906102259326e-05,
+      "loss": 1.7195,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21640638407413765,
+      "grad_norm": 2.640498253324088,
+      "learning_rate": 1.82327522056897e-05,
+      "loss": 1.8595,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21657799897031063,
+      "grad_norm": 3.4541835233235725,
+      "learning_rate": 1.8229595765568953e-05,
+      "loss": 1.6397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21674961386648361,
+      "grad_norm": 2.5020780651740298,
+      "learning_rate": 1.8226436782872272e-05,
+      "loss": 1.9658,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2169212287626566,
+      "grad_norm": 2.481859236058716,
+      "learning_rate": 1.822327525857565e-05,
+      "loss": 2.0094,
+      "step": 1264
+    },
+    {
+      "epoch": 0.21709284365882958,
+      "grad_norm": 3.332307394423074,
+      "learning_rate": 1.8220111193655847e-05,
+      "loss": 1.7374,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21726445855500257,
+      "grad_norm": 2.971761817217155,
+      "learning_rate": 1.8216944589090424e-05,
+      "loss": 1.8184,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21743607345117555,
+      "grad_norm": 3.3408218773785503,
+      "learning_rate": 1.8213775445857716e-05,
+      "loss": 2.0943,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21760768834734856,
+      "grad_norm": 2.4465288144681248,
+      "learning_rate": 1.821060376493685e-05,
+      "loss": 1.8451,
+      "step": 1268
+    },
+    {
+      "epoch": 0.21777930324352154,
+      "grad_norm": 3.260912485465419,
+      "learning_rate": 1.820742954730773e-05,
+      "loss": 1.8552,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21795091813969453,
+      "grad_norm": 2.4056093886019827,
+      "learning_rate": 1.8204252793951046e-05,
+      "loss": 1.8165,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2181225330358675,
+      "grad_norm": 3.3773877831868293,
+      "learning_rate": 1.8201073505848273e-05,
+      "loss": 1.9007,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2182941479320405,
+      "grad_norm": 3.116863891826737,
+      "learning_rate": 1.8197891683981673e-05,
+      "loss": 1.916,
+      "step": 1272
+    },
+    {
+      "epoch": 0.21846576282821348,
+      "grad_norm": 4.077474826951421,
+      "learning_rate": 1.8194707329334277e-05,
+      "loss": 1.7717,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2186373777243865,
+      "grad_norm": 3.3668874942453626,
+      "learning_rate": 1.819152044288992e-05,
+      "loss": 1.8545,
+      "step": 1274
+    },
+    {
+      "epoch": 0.21880899262055947,
+      "grad_norm": 2.3761737828260365,
+      "learning_rate": 1.8188331025633197e-05,
+      "loss": 1.9399,
+      "step": 1275
+    },
+    {
+      "epoch": 0.21898060751673246,
+      "grad_norm": 3.9393008530823654,
+      "learning_rate": 1.81851390785495e-05,
+      "loss": 1.4977,
+      "step": 1276
+    },
+    {
+      "epoch": 0.21915222241290544,
+      "grad_norm": 3.50209640033252,
+      "learning_rate": 1.818194460262499e-05,
+      "loss": 1.8757,
+      "step": 1277
+    },
+    {
+      "epoch": 0.21932383730907842,
+      "grad_norm": 2.403092229844327,
+      "learning_rate": 1.8178747598846627e-05,
+      "loss": 1.9168,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2194954522052514,
+      "grad_norm": 5.331340745084284,
+      "learning_rate": 1.8175548068202138e-05,
+      "loss": 1.8661,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2196670671014244,
+      "grad_norm": 3.3034480373973945,
+      "learning_rate": 1.817234601168003e-05,
+      "loss": 1.9273,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2198386819975974,
+      "grad_norm": 2.4959495149983595,
+      "learning_rate": 1.81691414302696e-05,
+      "loss": 1.5519,
+      "step": 1281
+    },
+    {
+      "epoch": 0.22001029689377039,
+      "grad_norm": 3.6827836948667434,
+      "learning_rate": 1.816593432496092e-05,
+      "loss": 1.5516,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22018191178994337,
+      "grad_norm": 2.973148329891034,
+      "learning_rate": 1.8162724696744837e-05,
+      "loss": 1.8896,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22035352668611635,
+      "grad_norm": 3.194126737333955,
+      "learning_rate": 1.8159512546612984e-05,
+      "loss": 1.6747,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22052514158228934,
+      "grad_norm": 2.6032348659036004,
+      "learning_rate": 1.8156297875557777e-05,
+      "loss": 1.6555,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22069675647846232,
+      "grad_norm": 2.6929426894436177,
+      "learning_rate": 1.81530806845724e-05,
+      "loss": 1.8846,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22086837137463533,
+      "grad_norm": 2.6280832287436175,
+      "learning_rate": 1.814986097465082e-05,
+      "loss": 1.6971,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22103998627080831,
+      "grad_norm": 2.4539562110003357,
+      "learning_rate": 1.814663874678778e-05,
+      "loss": 1.8418,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2212116011669813,
+      "grad_norm": 2.5155084452298118,
+      "learning_rate": 1.8143414001978813e-05,
+      "loss": 1.6249,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22138321606315428,
+      "grad_norm": 2.119527510131496,
+      "learning_rate": 1.8140186741220214e-05,
+      "loss": 1.821,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22155483095932726,
+      "grad_norm": 2.9593482572765653,
+      "learning_rate": 1.8136956965509064e-05,
+      "loss": 1.7918,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22172644585550025,
+      "grad_norm": 2.6152796447587825,
+      "learning_rate": 1.813372467584322e-05,
+      "loss": 1.7522,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22189806075167323,
+      "grad_norm": 2.895574453223862,
+      "learning_rate": 1.8130489873221307e-05,
+      "loss": 1.7293,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22206967564784624,
+      "grad_norm": 2.4988734751677395,
+      "learning_rate": 1.8127252558642742e-05,
+      "loss": 1.9581,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22224129054401923,
+      "grad_norm": 3.3725290672729584,
+      "learning_rate": 1.812401273310771e-05,
+      "loss": 1.6294,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2224129054401922,
+      "grad_norm": 7.791377195449263,
+      "learning_rate": 1.8120770397617166e-05,
+      "loss": 1.8664,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2225845203363652,
+      "grad_norm": 3.275834011748801,
+      "learning_rate": 1.8117525553172853e-05,
+      "loss": 1.8551,
+      "step": 1297
+    },
+    {
+      "epoch": 0.22275613523253818,
+      "grad_norm": 4.0812507031430645,
+      "learning_rate": 1.8114278200777278e-05,
+      "loss": 1.838,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22292775012871116,
+      "grad_norm": 2.7604762269869307,
+      "learning_rate": 1.8111028341433726e-05,
+      "loss": 1.6942,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22309936502488417,
+      "grad_norm": 2.796400334012174,
+      "learning_rate": 1.810777597614626e-05,
+      "loss": 1.8064,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22327097992105716,
+      "grad_norm": 2.772557371729177,
+      "learning_rate": 1.8104521105919715e-05,
+      "loss": 1.6805,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22344259481723014,
+      "grad_norm": 4.088203707979652,
+      "learning_rate": 1.8101263731759703e-05,
+      "loss": 1.6589,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22361420971340312,
+      "grad_norm": 2.8045707689302946,
+      "learning_rate": 1.8098003854672602e-05,
+      "loss": 1.7349,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2237858246095761,
+      "grad_norm": 2.492347450003987,
+      "learning_rate": 1.809474147566557e-05,
+      "loss": 1.8442,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2239574395057491,
+      "grad_norm": 2.353291139640294,
+      "learning_rate": 1.8091476595746528e-05,
+      "loss": 1.6864,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22412905440192207,
+      "grad_norm": 2.7782602611246605,
+      "learning_rate": 1.8088209215924187e-05,
+      "loss": 1.9521,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22430066929809508,
+      "grad_norm": 3.019808912758091,
+      "learning_rate": 1.808493933720802e-05,
+      "loss": 1.8772,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22447228419426807,
+      "grad_norm": 2.6094264648910146,
+      "learning_rate": 1.8081666960608272e-05,
+      "loss": 1.7142,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22464389909044105,
+      "grad_norm": 2.7692784413758047,
+      "learning_rate": 1.8078392087135957e-05,
+      "loss": 1.7153,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22481551398661404,
+      "grad_norm": 2.929090732988897,
+      "learning_rate": 1.807511471780287e-05,
+      "loss": 1.8716,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22498712888278702,
+      "grad_norm": 2.3056804310234993,
+      "learning_rate": 1.8071834853621568e-05,
+      "loss": 1.5846,
+      "step": 1311
+    },
+    {
+      "epoch": 0.22515874377896,
+      "grad_norm": 2.4403923199911794,
+      "learning_rate": 1.8068552495605375e-05,
+      "loss": 1.9402,
+      "step": 1312
+    },
+    {
+      "epoch": 0.225330358675133,
+      "grad_norm": 2.3832362798720834,
+      "learning_rate": 1.8065267644768406e-05,
+      "loss": 1.5751,
+      "step": 1313
+    },
+    {
+      "epoch": 0.225501973571306,
+      "grad_norm": 2.9258795068855523,
+      "learning_rate": 1.806198030212552e-05,
+      "loss": 1.7419,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22567358846747898,
+      "grad_norm": 3.0735089245704237,
+      "learning_rate": 1.8058690468692366e-05,
+      "loss": 1.5883,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22584520336365196,
+      "grad_norm": 2.639123869972019,
+      "learning_rate": 1.8055398145485354e-05,
+      "loss": 2.0046,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22601681825982495,
+      "grad_norm": 2.512944399199607,
+      "learning_rate": 1.8052103333521664e-05,
+      "loss": 2.0952,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22618843315599793,
+      "grad_norm": 2.6527425669107014,
+      "learning_rate": 1.804880603381924e-05,
+      "loss": 1.8944,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22636004805217091,
+      "grad_norm": 4.923912428369115,
+      "learning_rate": 1.8045506247396804e-05,
+      "loss": 1.6397,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22653166294834393,
+      "grad_norm": 3.0224951366991784,
+      "learning_rate": 1.8042203975273844e-05,
+      "loss": 1.7171,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2267032778445169,
+      "grad_norm": 6.949072684049702,
+      "learning_rate": 1.803889921847061e-05,
+      "loss": 1.8849,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2268748927406899,
+      "grad_norm": 3.072239052808573,
+      "learning_rate": 1.8035591978008127e-05,
+      "loss": 1.7037,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22704650763686288,
+      "grad_norm": 2.817660301036934,
+      "learning_rate": 1.803228225490818e-05,
+      "loss": 1.7688,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22721812253303586,
+      "grad_norm": 2.8730499684429307,
+      "learning_rate": 1.8028970050193327e-05,
+      "loss": 1.7906,
+      "step": 1324
+    },
+    {
+      "epoch": 0.22738973742920884,
+      "grad_norm": 3.5997366282468355,
+      "learning_rate": 1.8025655364886896e-05,
+      "loss": 1.734,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22756135232538185,
+      "grad_norm": 3.5787006512726056,
+      "learning_rate": 1.802233820001297e-05,
+      "loss": 1.9117,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22773296722155484,
+      "grad_norm": 2.6218468563756883,
+      "learning_rate": 1.8019018556596402e-05,
+      "loss": 1.4182,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22790458211772782,
+      "grad_norm": 2.5995495564158153,
+      "learning_rate": 1.8015696435662814e-05,
+      "loss": 1.6654,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2280761970139008,
+      "grad_norm": 2.5899646383440054,
+      "learning_rate": 1.80123718382386e-05,
+      "loss": 1.8941,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2282478119100738,
+      "grad_norm": 2.931079932476063,
+      "learning_rate": 1.8009044765350904e-05,
+      "loss": 2.008,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22841942680624677,
+      "grad_norm": 3.229911226927399,
+      "learning_rate": 1.800571521802764e-05,
+      "loss": 1.9049,
+      "step": 1331
+    },
+    {
+      "epoch": 0.22859104170241978,
+      "grad_norm": 2.256689366904933,
+      "learning_rate": 1.8002383197297496e-05,
+      "loss": 1.6809,
+      "step": 1332
+    },
+    {
+      "epoch": 0.22876265659859277,
+      "grad_norm": 2.2728948666540827,
+      "learning_rate": 1.7999048704189914e-05,
+      "loss": 1.6756,
+      "step": 1333
+    },
+    {
+      "epoch": 0.22893427149476575,
+      "grad_norm": 2.5291378252916368,
+      "learning_rate": 1.79957117397351e-05,
+      "loss": 1.6907,
+      "step": 1334
+    },
+    {
+      "epoch": 0.22910588639093873,
+      "grad_norm": 2.55714269970868,
+      "learning_rate": 1.7992372304964036e-05,
+      "loss": 1.9827,
+      "step": 1335
+    },
+    {
+      "epoch": 0.22927750128711172,
+      "grad_norm": 2.433216575873825,
+      "learning_rate": 1.798903040090844e-05,
+      "loss": 1.7122,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2294491161832847,
+      "grad_norm": 2.474618219899491,
+      "learning_rate": 1.798568602860083e-05,
+      "loss": 1.6089,
+      "step": 1337
+    },
+    {
+      "epoch": 0.22962073107945769,
+      "grad_norm": 2.734379193098395,
+      "learning_rate": 1.7982339189074456e-05,
+      "loss": 1.6649,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2297923459756307,
+      "grad_norm": 3.5470203745874263,
+      "learning_rate": 1.7978989883363344e-05,
+      "loss": 1.5472,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22996396087180368,
+      "grad_norm": 2.7103573954179203,
+      "learning_rate": 1.7975638112502274e-05,
+      "loss": 1.6411,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23013557576797666,
+      "grad_norm": 2.699358838758311,
+      "learning_rate": 1.7972283877526798e-05,
+      "loss": 1.7652,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23030719066414965,
+      "grad_norm": 3.3102573102159347,
+      "learning_rate": 1.7968927179473228e-05,
+      "loss": 1.8146,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23047880556032263,
+      "grad_norm": 2.651795657014917,
+      "learning_rate": 1.7965568019378625e-05,
+      "loss": 1.8207,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23065042045649561,
+      "grad_norm": 2.4361985608756638,
+      "learning_rate": 1.7962206398280823e-05,
+      "loss": 1.9642,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23082203535266863,
+      "grad_norm": 6.106790097286908,
+      "learning_rate": 1.7958842317218413e-05,
+      "loss": 1.9332,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2309936502488416,
+      "grad_norm": 2.309904666725919,
+      "learning_rate": 1.7955475777230744e-05,
+      "loss": 1.6604,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2311652651450146,
+      "grad_norm": 3.0506367973564976,
+      "learning_rate": 1.7952106779357922e-05,
+      "loss": 1.6837,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23133688004118758,
+      "grad_norm": 2.335794128089584,
+      "learning_rate": 1.794873532464082e-05,
+      "loss": 1.6833,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23150849493736056,
+      "grad_norm": 2.542333416599226,
+      "learning_rate": 1.7945361414121068e-05,
+      "loss": 1.9018,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23168010983353354,
+      "grad_norm": 2.4958905772007705,
+      "learning_rate": 1.7941985048841052e-05,
+      "loss": 1.9232,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23185172472970653,
+      "grad_norm": 4.471762978169001,
+      "learning_rate": 1.7938606229843913e-05,
+      "loss": 1.9173,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23202333962587954,
+      "grad_norm": 2.5679115457969943,
+      "learning_rate": 1.793522495817356e-05,
+      "loss": 1.8792,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23219495452205252,
+      "grad_norm": 3.6312333444594587,
+      "learning_rate": 1.793184123487465e-05,
+      "loss": 1.6292,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2323665694182255,
+      "grad_norm": 2.4187153049270553,
+      "learning_rate": 1.7928455060992606e-05,
+      "loss": 1.5971,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2325381843143985,
+      "grad_norm": 3.362483509005112,
+      "learning_rate": 1.79250664375736e-05,
+      "loss": 1.8521,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23270979921057147,
+      "grad_norm": 2.5889713021074403,
+      "learning_rate": 1.7921675365664567e-05,
+      "loss": 1.7494,
+      "step": 1356
+    },
+    {
+      "epoch": 0.23288141410674446,
+      "grad_norm": 2.876865875021111,
+      "learning_rate": 1.79182818463132e-05,
+      "loss": 1.8587,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23305302900291747,
+      "grad_norm": 2.468009015890249,
+      "learning_rate": 1.7914885880567933e-05,
+      "loss": 1.4021,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23322464389909045,
+      "grad_norm": 2.8125242032724955,
+      "learning_rate": 1.7911487469477976e-05,
+      "loss": 1.8282,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23339625879526343,
+      "grad_norm": 3.399451031904387,
+      "learning_rate": 1.7908086614093287e-05,
+      "loss": 1.7086,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23356787369143642,
+      "grad_norm": 2.5880088775536074,
+      "learning_rate": 1.790468331546457e-05,
+      "loss": 1.8835,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2337394885876094,
+      "grad_norm": 2.7956652592360864,
+      "learning_rate": 1.7901277574643294e-05,
+      "loss": 1.6929,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23391110348378238,
+      "grad_norm": 2.3506902463350605,
+      "learning_rate": 1.7897869392681685e-05,
+      "loss": 1.7962,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23408271837995537,
+      "grad_norm": 3.2541471863807496,
+      "learning_rate": 1.7894458770632715e-05,
+      "loss": 1.7539,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23425433327612838,
+      "grad_norm": 3.10385362412242,
+      "learning_rate": 1.7891045709550113e-05,
+      "loss": 1.7094,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23442594817230136,
+      "grad_norm": 2.7168375602859705,
+      "learning_rate": 1.788763021048836e-05,
+      "loss": 1.8563,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23459756306847435,
+      "grad_norm": 4.079176469881752,
+      "learning_rate": 1.7884212274502697e-05,
+      "loss": 1.52,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23476917796464733,
+      "grad_norm": 2.6009107561273628,
+      "learning_rate": 1.788079190264911e-05,
+      "loss": 1.6671,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2349407928608203,
+      "grad_norm": 2.3774728943343106,
+      "learning_rate": 1.7877369095984338e-05,
+      "loss": 1.8783,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2351124077569933,
+      "grad_norm": 3.088598340648571,
+      "learning_rate": 1.7873943855565882e-05,
+      "loss": 1.8504,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2352840226531663,
+      "grad_norm": 3.07208213209689,
+      "learning_rate": 1.787051618245198e-05,
+      "loss": 1.9409,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2354556375493393,
+      "grad_norm": 3.513877457366281,
+      "learning_rate": 1.786708607770163e-05,
+      "loss": 1.6769,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23562725244551228,
+      "grad_norm": 2.517107634291877,
+      "learning_rate": 1.7863653542374588e-05,
+      "loss": 1.4963,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23579886734168526,
+      "grad_norm": 2.9849945508609,
+      "learning_rate": 1.7860218577531344e-05,
+      "loss": 1.6938,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23597048223785824,
+      "grad_norm": 2.423453837335957,
+      "learning_rate": 1.7856781184233152e-05,
+      "loss": 1.6792,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23614209713403123,
+      "grad_norm": 2.4427223482429743,
+      "learning_rate": 1.785334136354201e-05,
+      "loss": 1.6274,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2363137120302042,
+      "grad_norm": 2.6957913548685277,
+      "learning_rate": 1.784989911652068e-05,
+      "loss": 2.2019,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23648532692637722,
+      "grad_norm": 2.5341505572678917,
+      "learning_rate": 1.7846454444232643e-05,
+      "loss": 1.8136,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2366569418225502,
+      "grad_norm": 2.3019151750662807,
+      "learning_rate": 1.7843007347742164e-05,
+      "loss": 1.7144,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2368285567187232,
+      "grad_norm": 2.9669999947902537,
+      "learning_rate": 1.7839557828114233e-05,
+      "loss": 1.8496,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23700017161489617,
+      "grad_norm": 2.27762542111608,
+      "learning_rate": 1.7836105886414596e-05,
+      "loss": 1.729,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23717178651106915,
+      "grad_norm": 2.816074391136964,
+      "learning_rate": 1.7832651523709757e-05,
+      "loss": 1.8827,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23734340140724214,
+      "grad_norm": 2.2802898280291872,
+      "learning_rate": 1.782919474106695e-05,
+      "loss": 1.6265,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23751501630341515,
+      "grad_norm": 2.688405805464489,
+      "learning_rate": 1.7825735539554167e-05,
+      "loss": 1.9403,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23768663119958813,
+      "grad_norm": 2.3731957502721763,
+      "learning_rate": 1.7822273920240148e-05,
+      "loss": 1.6611,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23785824609576112,
+      "grad_norm": 2.651584382583664,
+      "learning_rate": 1.781880988419438e-05,
+      "loss": 1.8196,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2380298609919341,
+      "grad_norm": 3.4841510671078098,
+      "learning_rate": 1.7815343432487094e-05,
+      "loss": 1.8005,
+      "step": 1387
+    },
+    {
+      "epoch": 0.23820147588810708,
+      "grad_norm": 2.368677978744536,
+      "learning_rate": 1.7811874566189267e-05,
+      "loss": 1.5585,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23837309078428007,
+      "grad_norm": 2.6138080533155423,
+      "learning_rate": 1.7808403286372622e-05,
+      "loss": 2.0689,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23854470568045305,
+      "grad_norm": 2.2824559511668276,
+      "learning_rate": 1.780492959410963e-05,
+      "loss": 1.7085,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23871632057662606,
+      "grad_norm": 2.548842199787288,
+      "learning_rate": 1.7801453490473503e-05,
+      "loss": 1.8766,
+      "step": 1391
+    },
+    {
+      "epoch": 0.23888793547279905,
+      "grad_norm": 3.47246445505934,
+      "learning_rate": 1.7797974976538207e-05,
+      "loss": 1.8677,
+      "step": 1392
+    },
+    {
+      "epoch": 0.23905955036897203,
+      "grad_norm": 3.026151651429602,
+      "learning_rate": 1.7794494053378445e-05,
+      "loss": 1.5491,
+      "step": 1393
+    },
+    {
+      "epoch": 0.239231165265145,
+      "grad_norm": 2.7824251939868105,
+      "learning_rate": 1.7791010722069662e-05,
+      "loss": 1.693,
+      "step": 1394
+    },
+    {
+      "epoch": 0.239402780161318,
+      "grad_norm": 2.9721199199088457,
+      "learning_rate": 1.7787524983688057e-05,
+      "loss": 1.6517,
+      "step": 1395
+    },
+    {
+      "epoch": 0.23957439505749098,
+      "grad_norm": 1.9539471315163741,
+      "learning_rate": 1.7784036839310558e-05,
+      "loss": 1.6446,
+      "step": 1396
+    },
+    {
+      "epoch": 0.239746009953664,
+      "grad_norm": 2.4267104977038256,
+      "learning_rate": 1.7780546290014858e-05,
+      "loss": 1.8531,
+      "step": 1397
+    },
+    {
+      "epoch": 0.23991762484983697,
+      "grad_norm": 2.7463526453657328,
+      "learning_rate": 1.777705333687937e-05,
+      "loss": 1.5894,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24008923974600996,
+      "grad_norm": 2.6599455598055735,
+      "learning_rate": 1.7773557980983264e-05,
+      "loss": 1.9886,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24026085464218294,
+      "grad_norm": 2.776970592981246,
+      "learning_rate": 1.7770060223406443e-05,
+      "loss": 1.8507,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24043246953835593,
+      "grad_norm": 2.7128708669109134,
+      "learning_rate": 1.7766560065229564e-05,
+      "loss": 1.8109,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2406040844345289,
+      "grad_norm": 2.644803081104447,
+      "learning_rate": 1.7763057507534016e-05,
+      "loss": 1.8833,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2407756993307019,
+      "grad_norm": 2.987861621290512,
+      "learning_rate": 1.7759552551401932e-05,
+      "loss": 1.676,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2409473142268749,
+      "grad_norm": 2.7283913541229157,
+      "learning_rate": 1.7756045197916188e-05,
+      "loss": 1.8763,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2411189291230479,
+      "grad_norm": 2.4132102439700587,
+      "learning_rate": 1.7752535448160395e-05,
+      "loss": 1.8745,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24129054401922087,
+      "grad_norm": 2.1056526579604204,
+      "learning_rate": 1.774902330321891e-05,
+      "loss": 1.7353,
+      "step": 1406
+    },
+    {
+      "epoch": 0.24146215891539385,
+      "grad_norm": 2.675575563607133,
+      "learning_rate": 1.774550876417683e-05,
+      "loss": 1.9649,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24163377381156684,
+      "grad_norm": 2.3912318188900543,
+      "learning_rate": 1.7741991832119993e-05,
+      "loss": 1.7006,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24180538870773982,
+      "grad_norm": 2.384030479853452,
+      "learning_rate": 1.7738472508134967e-05,
+      "loss": 1.6052,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24197700360391283,
+      "grad_norm": 2.463324580430006,
+      "learning_rate": 1.7734950793309065e-05,
+      "loss": 1.6741,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24214861850008582,
+      "grad_norm": 2.6220867721211496,
+      "learning_rate": 1.773142668873034e-05,
+      "loss": 1.8201,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2423202333962588,
+      "grad_norm": 2.7285078368761875,
+      "learning_rate": 1.7727900195487588e-05,
+      "loss": 1.8388,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24249184829243178,
+      "grad_norm": 2.3986630335661223,
+      "learning_rate": 1.772437131467033e-05,
+      "loss": 1.7534,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24266346318860477,
+      "grad_norm": 3.1325205047159477,
+      "learning_rate": 1.7720840047368834e-05,
+      "loss": 1.9348,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24283507808477775,
+      "grad_norm": 2.9702218771485476,
+      "learning_rate": 1.7717306394674104e-05,
+      "loss": 1.6474,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24300669298095073,
+      "grad_norm": 2.339625482387283,
+      "learning_rate": 1.771377035767788e-05,
+      "loss": 1.8135,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24317830787712374,
+      "grad_norm": 2.802924814798818,
+      "learning_rate": 1.771023193747264e-05,
+      "loss": 1.8663,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24334992277329673,
+      "grad_norm": 2.379822394146494,
+      "learning_rate": 1.7706691135151594e-05,
+      "loss": 1.8176,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2435215376694697,
+      "grad_norm": 2.9006252523047418,
+      "learning_rate": 1.7703147951808695e-05,
+      "loss": 1.831,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2436931525656427,
+      "grad_norm": 3.477911826741412,
+      "learning_rate": 1.7699602388538622e-05,
+      "loss": 1.8788,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24386476746181568,
+      "grad_norm": 3.045349739992863,
+      "learning_rate": 1.7696054446436805e-05,
+      "loss": 1.8765,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24403638235798866,
+      "grad_norm": 3.164984400246872,
+      "learning_rate": 1.769250412659939e-05,
+      "loss": 1.521,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24420799725416167,
+      "grad_norm": 2.591559155834003,
+      "learning_rate": 1.768895143012327e-05,
+      "loss": 1.8061,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24437961215033466,
+      "grad_norm": 2.5013907866691327,
+      "learning_rate": 1.7685396358106063e-05,
+      "loss": 1.6322,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24455122704650764,
+      "grad_norm": 2.5616949021669635,
+      "learning_rate": 1.7681838911646133e-05,
+      "loss": 1.7826,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24472284194268062,
+      "grad_norm": 2.544567571032494,
+      "learning_rate": 1.7678279091842575e-05,
+      "loss": 1.876,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2448944568388536,
+      "grad_norm": 2.437683598841562,
+      "learning_rate": 1.7674716899795205e-05,
+      "loss": 1.635,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2450660717350266,
+      "grad_norm": 2.1796824396879666,
+      "learning_rate": 1.7671152336604584e-05,
+      "loss": 1.7428,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24523768663119958,
+      "grad_norm": 2.6926158715778503,
+      "learning_rate": 1.7667585403372003e-05,
+      "loss": 1.8446,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2454093015273726,
+      "grad_norm": 4.587942031167987,
+      "learning_rate": 1.7664016101199484e-05,
+      "loss": 1.7406,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24558091642354557,
+      "grad_norm": 2.978819599352636,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 1.8303,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24575253131971855,
+      "grad_norm": 2.672755091915462,
+      "learning_rate": 1.7656870394446382e-05,
+      "loss": 1.8897,
+      "step": 1432
+    },
+    {
+      "epoch": 0.24592414621589154,
+      "grad_norm": 2.7354248383324737,
+      "learning_rate": 1.76532939920735e-05,
+      "loss": 1.8856,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24609576111206452,
+      "grad_norm": 2.547167265850652,
+      "learning_rate": 1.7649715225176085e-05,
+      "loss": 1.8571,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2462673760082375,
+      "grad_norm": 2.61646567103413,
+      "learning_rate": 1.7646134094859816e-05,
+      "loss": 1.4694,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24643899090441052,
+      "grad_norm": 2.746827887403261,
+      "learning_rate": 1.7642550602231097e-05,
+      "loss": 1.6877,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2466106058005835,
+      "grad_norm": 4.199810816255603,
+      "learning_rate": 1.7638964748397075e-05,
+      "loss": 1.6023,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24678222069675648,
+      "grad_norm": 3.87698775931774,
+      "learning_rate": 1.7635376534465614e-05,
+      "loss": 1.9767,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24695383559292947,
+      "grad_norm": 2.5567221580273247,
+      "learning_rate": 1.763178596154531e-05,
+      "loss": 1.7989,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24712545048910245,
+      "grad_norm": 3.5848240947039063,
+      "learning_rate": 1.7628193030745486e-05,
+      "loss": 1.6871,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24729706538527543,
+      "grad_norm": 2.4514066562829138,
+      "learning_rate": 1.76245977431762e-05,
+      "loss": 1.8,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24746868028144842,
+      "grad_norm": 2.299893926652702,
+      "learning_rate": 1.7621000099948237e-05,
+      "loss": 1.7775,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24764029517762143,
+      "grad_norm": 2.727368128692727,
+      "learning_rate": 1.7617400102173107e-05,
+      "loss": 1.8309,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2478119100737944,
+      "grad_norm": 3.4162596655754096,
+      "learning_rate": 1.761379775096304e-05,
+      "loss": 1.7726,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2479835249699674,
+      "grad_norm": 2.247357177017449,
+      "learning_rate": 1.7610193047431013e-05,
+      "loss": 1.7182,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24815513986614038,
+      "grad_norm": 2.4713394155198585,
+      "learning_rate": 1.760658599269071e-05,
+      "loss": 1.8464,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24832675476231336,
+      "grad_norm": 2.280633934041197,
+      "learning_rate": 1.7602976587856547e-05,
+      "loss": 1.7349,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24849836965848635,
+      "grad_norm": 2.6766718813593338,
+      "learning_rate": 1.759936483404367e-05,
+      "loss": 1.6545,
+      "step": 1448
+    },
+    {
+      "epoch": 0.24866998455465936,
+      "grad_norm": 2.2588082631025284,
+      "learning_rate": 1.7595750732367954e-05,
+      "loss": 1.7213,
+      "step": 1449
+    },
+    {
+      "epoch": 0.24884159945083234,
+      "grad_norm": 2.3445653041723125,
+      "learning_rate": 1.759213428394599e-05,
+      "loss": 1.7557,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24901321434700532,
+      "grad_norm": 3.116227979413196,
+      "learning_rate": 1.75885154898951e-05,
+      "loss": 1.9927,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2491848292431783,
+      "grad_norm": 2.1168598695968677,
+      "learning_rate": 1.7584894351333327e-05,
+      "loss": 1.6726,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2493564441393513,
+      "grad_norm": 2.2926478860029573,
+      "learning_rate": 1.7581270869379443e-05,
+      "loss": 1.5857,
+      "step": 1453
+    },
+    {
+      "epoch": 0.24952805903552427,
+      "grad_norm": 2.7400574126544464,
+      "learning_rate": 1.757764504515294e-05,
+      "loss": 1.6226,
+      "step": 1454
+    },
+    {
+      "epoch": 0.24969967393169726,
+      "grad_norm": 3.4065919777432727,
+      "learning_rate": 1.7574016879774035e-05,
+      "loss": 1.7629,
+      "step": 1455
+    },
+    {
+      "epoch": 0.24987128882787027,
+      "grad_norm": 3.557445655104279,
+      "learning_rate": 1.7570386374363665e-05,
+      "loss": 1.6168,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25004290372404325,
+      "grad_norm": 2.3393883174838495,
+      "learning_rate": 1.7566753530043502e-05,
+      "loss": 1.9891,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2502145186202162,
+      "grad_norm": 3.2523515815884543,
+      "learning_rate": 1.7563118347935924e-05,
+      "loss": 1.7324,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2503861335163892,
+      "grad_norm": 3.8686911441253997,
+      "learning_rate": 1.755948082916404e-05,
+      "loss": 1.8791,
+      "step": 1459
+    },
+    {
+      "epoch": 0.25055774841256223,
+      "grad_norm": 2.205555296824094,
+      "learning_rate": 1.7555840974851684e-05,
+      "loss": 1.6512,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2507293633087352,
+      "grad_norm": 2.629025346770507,
+      "learning_rate": 1.7552198786123403e-05,
+      "loss": 1.8813,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2509009782049082,
+      "grad_norm": 2.6630022483785907,
+      "learning_rate": 1.7548554264104477e-05,
+      "loss": 1.8821,
+      "step": 1462
+    },
+    {
+      "epoch": 0.25107259310108115,
+      "grad_norm": 2.4017256577402164,
+      "learning_rate": 1.754490740992089e-05,
+      "loss": 1.6335,
+      "step": 1463
+    },
+    {
+      "epoch": 0.25124420799725417,
+      "grad_norm": 2.5078923640316875,
+      "learning_rate": 1.7541258224699368e-05,
+      "loss": 1.7603,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2514158228934272,
+      "grad_norm": 4.425776665753366,
+      "learning_rate": 1.7537606709567336e-05,
+      "loss": 1.7626,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25158743778960013,
+      "grad_norm": 2.5935220142426223,
+      "learning_rate": 1.753395286565295e-05,
+      "loss": 1.8444,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25175905268577314,
+      "grad_norm": 2.4942397685647872,
+      "learning_rate": 1.753029669408509e-05,
+      "loss": 1.7513,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2519306675819461,
+      "grad_norm": 2.2868951218461118,
+      "learning_rate": 1.752663819599334e-05,
+      "loss": 1.6668,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2521022824781191,
+      "grad_norm": 2.9498312765626884,
+      "learning_rate": 1.752297737250802e-05,
+      "loss": 1.8869,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25227389737429207,
+      "grad_norm": 2.654651176624841,
+      "learning_rate": 1.7519314224760154e-05,
+      "loss": 1.6669,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2524455122704651,
+      "grad_norm": 2.9431139101478765,
+      "learning_rate": 1.7515648753881495e-05,
+      "loss": 1.8762,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2526171271666381,
+      "grad_norm": 2.3754066326325436,
+      "learning_rate": 1.7511980961004503e-05,
+      "loss": 1.5924,
+      "step": 1472
+    },
+    {
+      "epoch": 0.25278874206281104,
+      "grad_norm": 3.917715848659881,
+      "learning_rate": 1.7508310847262365e-05,
+      "loss": 1.7453,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25296035695898406,
+      "grad_norm": 2.4358327704685068,
+      "learning_rate": 1.7504638413788987e-05,
+      "loss": 1.7729,
+      "step": 1474
+    },
+    {
+      "epoch": 0.253131971855157,
+      "grad_norm": 3.2779406397859923,
+      "learning_rate": 1.750096366171898e-05,
+      "loss": 1.7636,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25330358675133,
+      "grad_norm": 2.4807346281284284,
+      "learning_rate": 1.749728659218768e-05,
+      "loss": 1.8396,
+      "step": 1476
+    },
+    {
+      "epoch": 0.253475201647503,
+      "grad_norm": 2.45213241820684,
+      "learning_rate": 1.749360720633113e-05,
+      "loss": 1.7953,
+      "step": 1477
+    },
+    {
+      "epoch": 0.253646816543676,
+      "grad_norm": 3.4834473154388794,
+      "learning_rate": 1.7489925505286105e-05,
+      "loss": 1.8063,
+      "step": 1478
+    },
+    {
+      "epoch": 0.253818431439849,
+      "grad_norm": 3.036110684046407,
+      "learning_rate": 1.748624149019008e-05,
+      "loss": 1.6119,
+      "step": 1479
+    },
+    {
+      "epoch": 0.25399004633602196,
+      "grad_norm": 3.069745673025301,
+      "learning_rate": 1.748255516218125e-05,
+      "loss": 1.6641,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25416166123219497,
+      "grad_norm": 2.6950722319517,
+      "learning_rate": 1.7478866522398527e-05,
+      "loss": 1.7586,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2543332761283679,
+      "grad_norm": 3.101920783319619,
+      "learning_rate": 1.7475175571981535e-05,
+      "loss": 2.0566,
+      "step": 1482
+    },
+    {
+      "epoch": 0.25450489102454094,
+      "grad_norm": 2.9183366539911493,
+      "learning_rate": 1.7471482312070612e-05,
+      "loss": 1.9166,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2546765059207139,
+      "grad_norm": 2.5171956846663837,
+      "learning_rate": 1.7467786743806808e-05,
+      "loss": 1.8213,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2548481208168869,
+      "grad_norm": 3.1250940964360447,
+      "learning_rate": 1.746408886833189e-05,
+      "loss": 1.5925,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2550197357130599,
+      "grad_norm": 2.6527166911987687,
+      "learning_rate": 1.7460388686788334e-05,
+      "loss": 1.7109,
+      "step": 1486
+    },
+    {
+      "epoch": 0.25519135060923287,
+      "grad_norm": 2.490174400317957,
+      "learning_rate": 1.745668620031933e-05,
+      "loss": 1.4786,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2553629655054059,
+      "grad_norm": 3.1067721861369773,
+      "learning_rate": 1.7452981410068784e-05,
+      "loss": 1.7134,
+      "step": 1488
+    },
+    {
+      "epoch": 0.25553458040157884,
+      "grad_norm": 3.088538156934454,
+      "learning_rate": 1.7449274317181304e-05,
+      "loss": 1.831,
+      "step": 1489
+    },
+    {
+      "epoch": 0.25570619529775185,
+      "grad_norm": 2.6600478300050976,
+      "learning_rate": 1.7445564922802222e-05,
+      "loss": 1.7196,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25587781019392486,
+      "grad_norm": 2.9236034817135077,
+      "learning_rate": 1.7441853228077564e-05,
+      "loss": 1.7451,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2560494250900978,
+      "grad_norm": 3.6463660055971,
+      "learning_rate": 1.743813923415409e-05,
+      "loss": 1.8362,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2562210399862708,
+      "grad_norm": 4.043323228103443,
+      "learning_rate": 1.743442294217925e-05,
+      "loss": 1.6671,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2563926548824438,
+      "grad_norm": 2.8808692502806204,
+      "learning_rate": 1.7430704353301217e-05,
+      "loss": 1.8547,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2565642697786168,
+      "grad_norm": 2.939420360005542,
+      "learning_rate": 1.742698346866886e-05,
+      "loss": 1.9184,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25673588467478975,
+      "grad_norm": 3.2706136908468704,
+      "learning_rate": 1.742326028943177e-05,
+      "loss": 1.7767,
+      "step": 1496
+    },
+    {
+      "epoch": 0.25690749957096276,
+      "grad_norm": 3.350443254319093,
+      "learning_rate": 1.7419534816740245e-05,
+      "loss": 1.9367,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25707911446713577,
+      "grad_norm": 2.868776685089619,
+      "learning_rate": 1.7415807051745283e-05,
+      "loss": 1.7656,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2572507293633087,
+      "grad_norm": 3.3845682656877805,
+      "learning_rate": 1.7412076995598603e-05,
+      "loss": 1.7656,
+      "step": 1499
+    },
+    {
+      "epoch": 0.25742234425948174,
+      "grad_norm": 2.7619366802629917,
+      "learning_rate": 1.7408344649452617e-05,
+      "loss": 1.967,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2575939591556547,
+      "grad_norm": 2.6326147345309248,
+      "learning_rate": 1.740461001446046e-05,
+      "loss": 1.7066,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2577655740518277,
+      "grad_norm": 2.466792432066721,
+      "learning_rate": 1.740087309177596e-05,
+      "loss": 1.6788,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25793718894800066,
+      "grad_norm": 3.3394562383508486,
+      "learning_rate": 1.7397133882553664e-05,
+      "loss": 1.6711,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2581088038441737,
+      "grad_norm": 3.1221146038995684,
+      "learning_rate": 1.7393392387948818e-05,
+      "loss": 1.6528,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2582804187403467,
+      "grad_norm": 2.4062463872397846,
+      "learning_rate": 1.738964860911738e-05,
+      "loss": 1.6402,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25845203363651964,
+      "grad_norm": 2.5417128723454363,
+      "learning_rate": 1.7385902547215998e-05,
+      "loss": 1.5287,
+      "step": 1506
+    },
+    {
+      "epoch": 0.25862364853269265,
+      "grad_norm": 3.3850179167964725,
+      "learning_rate": 1.738215420340205e-05,
+      "loss": 1.9801,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2587952634288656,
+      "grad_norm": 2.7775339780385013,
+      "learning_rate": 1.7378403578833596e-05,
+      "loss": 1.9936,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2589668783250386,
+      "grad_norm": 2.633031896440434,
+      "learning_rate": 1.7374650674669418e-05,
+      "loss": 1.7873,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2591384932212116,
+      "grad_norm": 3.9769145239712183,
+      "learning_rate": 1.7370895492068994e-05,
+      "loss": 1.7204,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2593101081173846,
+      "grad_norm": 2.6529149414305078,
+      "learning_rate": 1.7367138032192503e-05,
+      "loss": 1.8777,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2594817230135576,
+      "grad_norm": 3.416918983967256,
+      "learning_rate": 1.7363378296200835e-05,
+      "loss": 1.8218,
+      "step": 1512
+    },
+    {
+      "epoch": 0.25965333790973055,
+      "grad_norm": 3.5573971863837963,
+      "learning_rate": 1.735961628525558e-05,
+      "loss": 1.8993,
+      "step": 1513
+    },
+    {
+      "epoch": 0.25982495280590356,
+      "grad_norm": 2.583276001721227,
+      "learning_rate": 1.735585200051903e-05,
+      "loss": 1.9064,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2599965677020765,
+      "grad_norm": 2.4255648144395763,
+      "learning_rate": 1.735208544315418e-05,
+      "loss": 1.4997,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26016818259824953,
+      "grad_norm": 2.6465476156962655,
+      "learning_rate": 1.7348316614324724e-05,
+      "loss": 1.758,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26033979749442254,
+      "grad_norm": 2.8495224000784254,
+      "learning_rate": 1.7344545515195067e-05,
+      "loss": 1.7262,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2605114123905955,
+      "grad_norm": 2.648169152536129,
+      "learning_rate": 1.73407721469303e-05,
+      "loss": 1.7133,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2606830272867685,
+      "grad_norm": 2.321574008172646,
+      "learning_rate": 1.7336996510696238e-05,
+      "loss": 1.7457,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26085464218294147,
+      "grad_norm": 3.7261904644744863,
+      "learning_rate": 1.7333218607659374e-05,
+      "loss": 1.9682,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2610262570791145,
+      "grad_norm": 3.114392867977243,
+      "learning_rate": 1.732943843898691e-05,
+      "loss": 1.9151,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26119787197528743,
+      "grad_norm": 2.902245709693817,
+      "learning_rate": 1.7325656005846755e-05,
+      "loss": 1.96,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26136948687146044,
+      "grad_norm": 3.898436401554328,
+      "learning_rate": 1.732187130940751e-05,
+      "loss": 1.8937,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26154110176763345,
+      "grad_norm": 3.0327781734508923,
+      "learning_rate": 1.7318084350838467e-05,
+      "loss": 1.8831,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2617127166638064,
+      "grad_norm": 2.601655199053984,
+      "learning_rate": 1.731429513130964e-05,
+      "loss": 1.805,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2618843315599794,
+      "grad_norm": 2.6124111405189185,
+      "learning_rate": 1.7310503651991717e-05,
+      "loss": 1.6163,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2620559464561524,
+      "grad_norm": 3.1563553182485857,
+      "learning_rate": 1.7306709914056096e-05,
+      "loss": 1.9093,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2622275613523254,
+      "grad_norm": 3.193772459653954,
+      "learning_rate": 1.730291391867488e-05,
+      "loss": 1.8741,
+      "step": 1528
+    },
+    {
+      "epoch": 0.26239917624849834,
+      "grad_norm": 2.625082431929434,
+      "learning_rate": 1.7299115667020858e-05,
+      "loss": 1.5032,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26257079114467136,
+      "grad_norm": 3.0272481014923476,
+      "learning_rate": 1.7295315160267512e-05,
+      "loss": 1.9747,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26274240604084437,
+      "grad_norm": 2.5129039812489125,
+      "learning_rate": 1.7291512399589038e-05,
+      "loss": 1.6281,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2629140209370173,
+      "grad_norm": 3.2791597949419624,
+      "learning_rate": 1.728770738616031e-05,
+      "loss": 1.6892,
+      "step": 1532
+    },
+    {
+      "epoch": 0.26308563583319033,
+      "grad_norm": 2.957135847451443,
+      "learning_rate": 1.7283900121156915e-05,
+      "loss": 1.7248,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2632572507293633,
+      "grad_norm": 2.998075608106472,
+      "learning_rate": 1.728009060575512e-05,
+      "loss": 1.6913,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2634288656255363,
+      "grad_norm": 3.732057986843457,
+      "learning_rate": 1.7276278841131895e-05,
+      "loss": 1.6593,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26360048052170926,
+      "grad_norm": 2.347499499080433,
+      "learning_rate": 1.727246482846491e-05,
+      "loss": 1.7998,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26377209541788227,
+      "grad_norm": 2.522931851278705,
+      "learning_rate": 1.726864856893252e-05,
+      "loss": 1.7454,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2639437103140553,
+      "grad_norm": 2.4481814095036527,
+      "learning_rate": 1.7264830063713777e-05,
+      "loss": 1.7267,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26411532521022824,
+      "grad_norm": 2.6801758151298487,
+      "learning_rate": 1.726100931398843e-05,
+      "loss": 1.7726,
+      "step": 1539
+    },
+    {
+      "epoch": 0.26428694010640125,
+      "grad_norm": 3.1449830391147713,
+      "learning_rate": 1.7257186320936918e-05,
+      "loss": 1.7759,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2644585550025742,
+      "grad_norm": 2.681847025467261,
+      "learning_rate": 1.7253361085740375e-05,
+      "loss": 1.6681,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2646301698987472,
+      "grad_norm": 2.6528031377273837,
+      "learning_rate": 1.7249533609580625e-05,
+      "loss": 1.5982,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2648017847949202,
+      "grad_norm": 3.280010562924151,
+      "learning_rate": 1.7245703893640188e-05,
+      "loss": 1.955,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2649733996910932,
+      "grad_norm": 3.049427729239868,
+      "learning_rate": 1.7241871939102277e-05,
+      "loss": 1.6879,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2651450145872662,
+      "grad_norm": 2.869023059104075,
+      "learning_rate": 1.723803774715079e-05,
+      "loss": 1.6638,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26531662948343915,
+      "grad_norm": 3.3736156192315585,
+      "learning_rate": 1.7234201318970325e-05,
+      "loss": 1.6949,
+      "step": 1546
+    },
+    {
+      "epoch": 0.26548824437961216,
+      "grad_norm": 3.1229525212485427,
+      "learning_rate": 1.723036265574616e-05,
+      "loss": 1.9317,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2656598592757851,
+      "grad_norm": 3.1484809692342415,
+      "learning_rate": 1.7226521758664273e-05,
+      "loss": 1.9209,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2658314741719581,
+      "grad_norm": 3.462602389684985,
+      "learning_rate": 1.7222678628911335e-05,
+      "loss": 2.0217,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26600308906813114,
+      "grad_norm": 2.5931798602928327,
+      "learning_rate": 1.7218833267674685e-05,
+      "loss": 1.7286,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2661747039643041,
+      "grad_norm": 3.081804671422973,
+      "learning_rate": 1.7214985676142382e-05,
+      "loss": 1.841,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2663463188604771,
+      "grad_norm": 3.1711487358492216,
+      "learning_rate": 1.7211135855503152e-05,
+      "loss": 1.7588,
+      "step": 1552
+    },
+    {
+      "epoch": 0.26651793375665006,
+      "grad_norm": 2.3529762976251503,
+      "learning_rate": 1.7207283806946422e-05,
+      "loss": 1.5071,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26668954865282307,
+      "grad_norm": 3.014107897364318,
+      "learning_rate": 1.7203429531662298e-05,
+      "loss": 1.6855,
+      "step": 1554
+    },
+    {
+      "epoch": 0.266861163548996,
+      "grad_norm": 2.6722626682838655,
+      "learning_rate": 1.7199573030841577e-05,
+      "loss": 1.8494,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26703277844516904,
+      "grad_norm": 3.067349525108301,
+      "learning_rate": 1.719571430567575e-05,
+      "loss": 1.6488,
+      "step": 1556
+    },
+    {
+      "epoch": 0.26720439334134205,
+      "grad_norm": 2.560878066453498,
+      "learning_rate": 1.7191853357356982e-05,
+      "loss": 1.5641,
+      "step": 1557
+    },
+    {
+      "epoch": 0.267376008237515,
+      "grad_norm": 3.5250648369135713,
+      "learning_rate": 1.718799018707814e-05,
+      "loss": 2.0262,
+      "step": 1558
+    },
+    {
+      "epoch": 0.267547623133688,
+      "grad_norm": 2.543328423488226,
+      "learning_rate": 1.718412479603277e-05,
+      "loss": 1.9759,
+      "step": 1559
+    },
+    {
+      "epoch": 0.267719238029861,
+      "grad_norm": 3.1583250595208927,
+      "learning_rate": 1.71802571854151e-05,
+      "loss": 1.8491,
+      "step": 1560
+    },
+    {
+      "epoch": 0.267890852926034,
+      "grad_norm": 3.2639542961971677,
+      "learning_rate": 1.717638735642005e-05,
+      "loss": 1.7902,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26806246782220694,
+      "grad_norm": 2.5168345894504798,
+      "learning_rate": 1.7172515310243228e-05,
+      "loss": 1.8908,
+      "step": 1562
+    },
+    {
+      "epoch": 0.26823408271837995,
+      "grad_norm": 2.7279437081151534,
+      "learning_rate": 1.716864104808092e-05,
+      "loss": 1.916,
+      "step": 1563
+    },
+    {
+      "epoch": 0.26840569761455296,
+      "grad_norm": 2.60769802784938,
+      "learning_rate": 1.716476457113009e-05,
+      "loss": 1.5688,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2685773125107259,
+      "grad_norm": 2.4160706593606527,
+      "learning_rate": 1.716088588058841e-05,
+      "loss": 1.6185,
+      "step": 1565
+    },
+    {
+      "epoch": 0.26874892740689893,
+      "grad_norm": 2.4514751729693276,
+      "learning_rate": 1.7157004977654216e-05,
+      "loss": 1.6715,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2689205423030719,
+      "grad_norm": 2.92845139445762,
+      "learning_rate": 1.7153121863526526e-05,
+      "loss": 1.6777,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2690921571992449,
+      "grad_norm": 2.68958774103703,
+      "learning_rate": 1.714923653940506e-05,
+      "loss": 1.7783,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2692637720954179,
+      "grad_norm": 2.8310014903790264,
+      "learning_rate": 1.71453490064902e-05,
+      "loss": 1.7727,
+      "step": 1569
+    },
+    {
+      "epoch": 0.26943538699159086,
+      "grad_norm": 2.722793387256109,
+      "learning_rate": 1.714145926598302e-05,
+      "loss": 1.347,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2696070018877639,
+      "grad_norm": 3.2556959061412947,
+      "learning_rate": 1.713756731908528e-05,
+      "loss": 1.7703,
+      "step": 1571
+    },
+    {
+      "epoch": 0.26977861678393683,
+      "grad_norm": 3.1541691642039655,
+      "learning_rate": 1.7133673166999414e-05,
+      "loss": 1.66,
+      "step": 1572
+    },
+    {
+      "epoch": 0.26995023168010984,
+      "grad_norm": 2.3661478153638447,
+      "learning_rate": 1.712977681092854e-05,
+      "loss": 1.9131,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2701218465762828,
+      "grad_norm": 3.0677843253879384,
+      "learning_rate": 1.7125878252076458e-05,
+      "loss": 1.7586,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2702934614724558,
+      "grad_norm": 5.631857968375433,
+      "learning_rate": 1.7121977491647647e-05,
+      "loss": 1.7252,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2704650763686288,
+      "grad_norm": 2.8771813830397344,
+      "learning_rate": 1.711807453084727e-05,
+      "loss": 1.6886,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2706366912648018,
+      "grad_norm": 2.7516856693041265,
+      "learning_rate": 1.7114169370881163e-05,
+      "loss": 1.7839,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2708083061609748,
+      "grad_norm": 2.806482208721437,
+      "learning_rate": 1.7110262012955847e-05,
+      "loss": 1.7277,
+      "step": 1578
+    },
+    {
+      "epoch": 0.27097992105714774,
+      "grad_norm": 2.6109556523263624,
+      "learning_rate": 1.7106352458278524e-05,
+      "loss": 1.7107,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27115153595332075,
+      "grad_norm": 3.2363262353487547,
+      "learning_rate": 1.7102440708057064e-05,
+      "loss": 1.905,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2713231508494937,
+      "grad_norm": 2.7751988988842835,
+      "learning_rate": 1.7098526763500022e-05,
+      "loss": 1.7993,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2714947657456667,
+      "grad_norm": 3.0662848666488096,
+      "learning_rate": 1.7094610625816637e-05,
+      "loss": 1.7867,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27166638064183973,
+      "grad_norm": 2.9289235256200197,
+      "learning_rate": 1.709069229621682e-05,
+      "loss": 1.7546,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2718379955380127,
+      "grad_norm": 2.5006326172097872,
+      "learning_rate": 1.7086771775911156e-05,
+      "loss": 1.7583,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2720096104341857,
+      "grad_norm": 4.074310350851328,
+      "learning_rate": 1.708284906611091e-05,
+      "loss": 1.8577,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27218122533035866,
+      "grad_norm": 3.2317818211099216,
+      "learning_rate": 1.7078924168028025e-05,
+      "loss": 1.8612,
+      "step": 1586
+    },
+    {
+      "epoch": 0.27235284022653167,
+      "grad_norm": 2.170859712314514,
+      "learning_rate": 1.7074997082875113e-05,
+      "loss": 1.8406,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2725244551227046,
+      "grad_norm": 2.3826014722338185,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.7281,
+      "step": 1588
+    },
+    {
+      "epoch": 0.27269607001887763,
+      "grad_norm": 2.213821435848086,
+      "learning_rate": 1.706713635621308e-05,
+      "loss": 1.7968,
+      "step": 1589
+    },
+    {
+      "epoch": 0.27286768491505065,
+      "grad_norm": 2.473429836073356,
+      "learning_rate": 1.7063202717132562e-05,
+      "loss": 1.5221,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2730392998112236,
+      "grad_norm": 3.117154404756111,
+      "learning_rate": 1.7059266895839245e-05,
+      "loss": 1.5559,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2732109147073966,
+      "grad_norm": 3.178333493719182,
+      "learning_rate": 1.7055328893549123e-05,
+      "loss": 1.8261,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27338252960356957,
+      "grad_norm": 2.696705231677164,
+      "learning_rate": 1.705138871147886e-05,
+      "loss": 1.8176,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2735541444997426,
+      "grad_norm": 3.258694724397034,
+      "learning_rate": 1.7047446350845795e-05,
+      "loss": 1.8876,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2737257593959156,
+      "grad_norm": 2.4654279160585757,
+      "learning_rate": 1.704350181286794e-05,
+      "loss": 1.8809,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27389737429208855,
+      "grad_norm": 2.9227767149954764,
+      "learning_rate": 1.7039555098763983e-05,
+      "loss": 1.9477,
+      "step": 1596
+    },
+    {
+      "epoch": 0.27406898918826156,
+      "grad_norm": 4.082100811735219,
+      "learning_rate": 1.7035606209753276e-05,
+      "loss": 1.6407,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2742406040844345,
+      "grad_norm": 2.774090413521391,
+      "learning_rate": 1.7031655147055855e-05,
+      "loss": 1.6077,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2744122189806075,
+      "grad_norm": 3.0554846264110433,
+      "learning_rate": 1.702770191189242e-05,
+      "loss": 1.5991,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2745838338767805,
+      "grad_norm": 2.489644653111394,
+      "learning_rate": 1.7023746505484337e-05,
+      "loss": 1.665,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2747554487729535,
+      "grad_norm": 2.7572665668713787,
+      "learning_rate": 1.7019788929053658e-05,
+      "loss": 1.8244,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2749270636691265,
+      "grad_norm": 3.106052522476244,
+      "learning_rate": 1.7015829183823094e-05,
+      "loss": 1.8388,
+      "step": 1602
+    },
+    {
+      "epoch": 0.27509867856529946,
+      "grad_norm": 2.8961483418989658,
+      "learning_rate": 1.7011867271016023e-05,
+      "loss": 1.798,
+      "step": 1603
+    },
+    {
+      "epoch": 0.27527029346147247,
+      "grad_norm": 2.3577958382625455,
+      "learning_rate": 1.7007903191856505e-05,
+      "loss": 1.7003,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2754419083576454,
+      "grad_norm": 3.037560530789631,
+      "learning_rate": 1.7003936947569263e-05,
+      "loss": 1.8744,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27561352325381844,
+      "grad_norm": 2.647717286798901,
+      "learning_rate": 1.699996853937968e-05,
+      "loss": 2.0544,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2757851381499914,
+      "grad_norm": 3.1170986190203176,
+      "learning_rate": 1.6995997968513823e-05,
+      "loss": 1.825,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2759567530461644,
+      "grad_norm": 3.0588809610215266,
+      "learning_rate": 1.6992025236198425e-05,
+      "loss": 1.5916,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2761283679423374,
+      "grad_norm": 2.823637237424643,
+      "learning_rate": 1.698805034366087e-05,
+      "loss": 1.8787,
+      "step": 1609
+    },
+    {
+      "epoch": 0.27629998283851037,
+      "grad_norm": 2.63789669727732,
+      "learning_rate": 1.6984073292129234e-05,
+      "loss": 1.6229,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2764715977346834,
+      "grad_norm": 6.665079470920686,
+      "learning_rate": 1.6980094082832237e-05,
+      "loss": 1.8691,
+      "step": 1611
+    },
+    {
+      "epoch": 0.27664321263085634,
+      "grad_norm": 3.0561724574659372,
+      "learning_rate": 1.697611271699928e-05,
+      "loss": 1.8726,
+      "step": 1612
+    },
+    {
+      "epoch": 0.27681482752702935,
+      "grad_norm": 2.571808848759333,
+      "learning_rate": 1.6972129195860427e-05,
+      "loss": 1.9544,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27698644242320236,
+      "grad_norm": 2.7739072655367214,
+      "learning_rate": 1.696814352064641e-05,
+      "loss": 1.9057,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2771580573193753,
+      "grad_norm": 3.193789246699656,
+      "learning_rate": 1.696415569258862e-05,
+      "loss": 1.8527,
+      "step": 1615
+    },
+    {
+      "epoch": 0.27732967221554833,
+      "grad_norm": 3.704355570715207,
+      "learning_rate": 1.6960165712919115e-05,
+      "loss": 1.67,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2775012871117213,
+      "grad_norm": 2.6885371927955024,
+      "learning_rate": 1.6956173582870627e-05,
+      "loss": 1.824,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2776729020078943,
+      "grad_norm": 2.6720311768588503,
+      "learning_rate": 1.6952179303676537e-05,
+      "loss": 1.8115,
+      "step": 1618
+    },
+    {
+      "epoch": 0.27784451690406725,
+      "grad_norm": 2.6531640238603305,
+      "learning_rate": 1.6948182876570906e-05,
+      "loss": 1.7719,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27801613180024026,
+      "grad_norm": 2.49510053108389,
+      "learning_rate": 1.6944184302788443e-05,
+      "loss": 1.8015,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2781877466964133,
+      "grad_norm": 2.885702356718763,
+      "learning_rate": 1.6940183583564536e-05,
+      "loss": 1.8309,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27835936159258623,
+      "grad_norm": 3.0244940282814157,
+      "learning_rate": 1.693618072013522e-05,
+      "loss": 1.8973,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27853097648875924,
+      "grad_norm": 2.6813830208876706,
+      "learning_rate": 1.6932175713737207e-05,
+      "loss": 1.781,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2787025913849322,
+      "grad_norm": 2.8750531359160023,
+      "learning_rate": 1.6928168565607857e-05,
+      "loss": 1.9627,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2788742062811052,
+      "grad_norm": 3.3877500204967577,
+      "learning_rate": 1.6924159276985206e-05,
+      "loss": 1.7016,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27904582117727816,
+      "grad_norm": 3.508382203916613,
+      "learning_rate": 1.692014784910794e-05,
+      "loss": 1.8295,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2792174360734512,
+      "grad_norm": 2.4402284092690527,
+      "learning_rate": 1.6916134283215412e-05,
+      "loss": 1.6662,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2793890509696242,
+      "grad_norm": 2.5523165571215287,
+      "learning_rate": 1.6912118580547633e-05,
+      "loss": 1.6677,
+      "step": 1628
+    },
+    {
+      "epoch": 0.27956066586579714,
+      "grad_norm": 2.8045635978020296,
+      "learning_rate": 1.690810074234528e-05,
+      "loss": 1.8417,
+      "step": 1629
+    },
+    {
+      "epoch": 0.27973228076197015,
+      "grad_norm": 3.1647165942502697,
+      "learning_rate": 1.6904080769849674e-05,
+      "loss": 1.5957,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2799038956581431,
+      "grad_norm": 2.4324886715140694,
+      "learning_rate": 1.6900058664302814e-05,
+      "loss": 1.6026,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2800755105543161,
+      "grad_norm": 2.489925365723236,
+      "learning_rate": 1.689603442694735e-05,
+      "loss": 1.8996,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2802471254504891,
+      "grad_norm": 2.4716070950887077,
+      "learning_rate": 1.6892008059026587e-05,
+      "loss": 1.667,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2804187403466621,
+      "grad_norm": 2.6020160833977743,
+      "learning_rate": 1.6887979561784496e-05,
+      "loss": 1.7024,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2805903552428351,
+      "grad_norm": 2.5866720867462965,
+      "learning_rate": 1.6883948936465705e-05,
+      "loss": 1.6527,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28076197013900805,
+      "grad_norm": 2.5099339412957296,
+      "learning_rate": 1.6879916184315485e-05,
+      "loss": 1.67,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28093358503518107,
+      "grad_norm": 2.0252511010099923,
+      "learning_rate": 1.6875881306579784e-05,
+      "loss": 1.5212,
+      "step": 1637
+    },
+    {
+      "epoch": 0.281105199931354,
+      "grad_norm": 2.4233829720045352,
+      "learning_rate": 1.6871844304505197e-05,
+      "loss": 1.6831,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28127681482752703,
+      "grad_norm": 2.890156029905627,
+      "learning_rate": 1.6867805179338977e-05,
+      "loss": 1.6371,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28144842972370004,
+      "grad_norm": 3.068250416761238,
+      "learning_rate": 1.686376393232903e-05,
+      "loss": 1.8581,
+      "step": 1640
+    },
+    {
+      "epoch": 0.281620044619873,
+      "grad_norm": 2.493913046417287,
+      "learning_rate": 1.6859720564723925e-05,
+      "loss": 1.9127,
+      "step": 1641
+    },
+    {
+      "epoch": 0.281791659516046,
+      "grad_norm": 2.6231773084668464,
+      "learning_rate": 1.6855675077772875e-05,
+      "loss": 1.6763,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28196327441221897,
+      "grad_norm": 2.337251737496812,
+      "learning_rate": 1.6851627472725758e-05,
+      "loss": 1.7893,
+      "step": 1643
+    },
+    {
+      "epoch": 0.282134889308392,
+      "grad_norm": 2.1506087144821358,
+      "learning_rate": 1.68475777508331e-05,
+      "loss": 1.88,
+      "step": 1644
+    },
+    {
+      "epoch": 0.28230650420456493,
+      "grad_norm": 2.3454309138999454,
+      "learning_rate": 1.6843525913346087e-05,
+      "loss": 1.6818,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28247811910073795,
+      "grad_norm": 4.557589192260528,
+      "learning_rate": 1.683947196151655e-05,
+      "loss": 1.8218,
+      "step": 1646
+    },
+    {
+      "epoch": 0.28264973399691096,
+      "grad_norm": 2.764353171609156,
+      "learning_rate": 1.6835415896596984e-05,
+      "loss": 1.6638,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2828213488930839,
+      "grad_norm": 3.051986174698658,
+      "learning_rate": 1.6831357719840523e-05,
+      "loss": 1.7782,
+      "step": 1648
+    },
+    {
+      "epoch": 0.2829929637892569,
+      "grad_norm": 2.5851504946670842,
+      "learning_rate": 1.682729743250097e-05,
+      "loss": 1.8075,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2831645786854299,
+      "grad_norm": 2.6029778063290716,
+      "learning_rate": 1.682323503583277e-05,
+      "loss": 1.664,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2833361935816029,
+      "grad_norm": 2.5199148605020203,
+      "learning_rate": 1.6819170531091018e-05,
+      "loss": 1.7638,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28350780847777585,
+      "grad_norm": 2.516737709379798,
+      "learning_rate": 1.6815103919531465e-05,
+      "loss": 1.6673,
+      "step": 1652
+    },
+    {
+      "epoch": 0.28367942337394886,
+      "grad_norm": 2.776563180780891,
+      "learning_rate": 1.6811035202410515e-05,
+      "loss": 1.4634,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28385103827012187,
+      "grad_norm": 2.5234906313531704,
+      "learning_rate": 1.680696438098521e-05,
+      "loss": 1.6784,
+      "step": 1654
+    },
+    {
+      "epoch": 0.2840226531662948,
+      "grad_norm": 3.67380008187408,
+      "learning_rate": 1.6802891456513262e-05,
+      "loss": 1.6739,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28419426806246784,
+      "grad_norm": 3.9941325004560295,
+      "learning_rate": 1.6798816430253014e-05,
+      "loss": 1.8279,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2843658829586408,
+      "grad_norm": 2.368148211691596,
+      "learning_rate": 1.679473930346347e-05,
+      "loss": 1.6006,
+      "step": 1657
+    },
+    {
+      "epoch": 0.2845374978548138,
+      "grad_norm": 2.9245378767409167,
+      "learning_rate": 1.6790660077404282e-05,
+      "loss": 1.6498,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28470911275098676,
+      "grad_norm": 3.9455610570534123,
+      "learning_rate": 1.678657875333574e-05,
+      "loss": 1.75,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28488072764715977,
+      "grad_norm": 2.5422718575340446,
+      "learning_rate": 1.67824953325188e-05,
+      "loss": 1.5199,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2850523425433328,
+      "grad_norm": 3.3288356163393438,
+      "learning_rate": 1.6778409816215046e-05,
+      "loss": 1.912,
+      "step": 1661
+    },
+    {
+      "epoch": 0.28522395743950574,
+      "grad_norm": 2.3863330644939516,
+      "learning_rate": 1.6774322205686725e-05,
+      "loss": 1.625,
+      "step": 1662
+    },
+    {
+      "epoch": 0.28539557233567875,
+      "grad_norm": 2.6525840724139904,
+      "learning_rate": 1.6770232502196722e-05,
+      "loss": 1.6626,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2855671872318517,
+      "grad_norm": 3.029489306269049,
+      "learning_rate": 1.6766140707008577e-05,
+      "loss": 1.8314,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2857388021280247,
+      "grad_norm": 3.5909002708597946,
+      "learning_rate": 1.6762046821386466e-05,
+      "loss": 1.8696,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2859104170241977,
+      "grad_norm": 2.621637169884585,
+      "learning_rate": 1.675795084659522e-05,
+      "loss": 1.8434,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2860820319203707,
+      "grad_norm": 2.831122436219414,
+      "learning_rate": 1.6753852783900306e-05,
+      "loss": 1.5602,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2862536468165437,
+      "grad_norm": 5.808637009262689,
+      "learning_rate": 1.674975263456785e-05,
+      "loss": 1.8426,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28642526171271665,
+      "grad_norm": 2.8667219509840685,
+      "learning_rate": 1.6745650399864608e-05,
+      "loss": 1.7443,
+      "step": 1669
+    },
+    {
+      "epoch": 0.28659687660888966,
+      "grad_norm": 3.3201404922473516,
+      "learning_rate": 1.6741546081057985e-05,
+      "loss": 1.8115,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2867684915050626,
+      "grad_norm": 2.654697033579173,
+      "learning_rate": 1.673743967941604e-05,
+      "loss": 1.7685,
+      "step": 1671
+    },
+    {
+      "epoch": 0.28694010640123563,
+      "grad_norm": 2.6299824779881447,
+      "learning_rate": 1.6733331196207453e-05,
+      "loss": 1.4929,
+      "step": 1672
+    },
+    {
+      "epoch": 0.28711172129740864,
+      "grad_norm": 4.237803955494624,
+      "learning_rate": 1.6729220632701574e-05,
+      "loss": 2.1493,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2872833361935816,
+      "grad_norm": 2.8000766703041045,
+      "learning_rate": 1.6725107990168382e-05,
+      "loss": 1.7877,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2874549510897546,
+      "grad_norm": 2.9569304176937643,
+      "learning_rate": 1.6720993269878486e-05,
+      "loss": 1.7197,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28762656598592756,
+      "grad_norm": 2.563568384070091,
+      "learning_rate": 1.6716876473103165e-05,
+      "loss": 1.8082,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2877981808821006,
+      "grad_norm": 2.7210295979197374,
+      "learning_rate": 1.6712757601114318e-05,
+      "loss": 1.8958,
+      "step": 1677
+    },
+    {
+      "epoch": 0.28796979577827353,
+      "grad_norm": 2.4161550886158056,
+      "learning_rate": 1.670863665518449e-05,
+      "loss": 1.7253,
+      "step": 1678
+    },
+    {
+      "epoch": 0.28814141067444654,
+      "grad_norm": 3.054748835078986,
+      "learning_rate": 1.670451363658687e-05,
+      "loss": 1.8488,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28831302557061955,
+      "grad_norm": 2.9790578293152254,
+      "learning_rate": 1.670038854659529e-05,
+      "loss": 1.674,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2884846404667925,
+      "grad_norm": 3.1763609316647625,
+      "learning_rate": 1.6696261386484215e-05,
+      "loss": 1.6418,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2886562553629655,
+      "grad_norm": 2.346245286394302,
+      "learning_rate": 1.6692132157528748e-05,
+      "loss": 1.6809,
+      "step": 1682
+    },
+    {
+      "epoch": 0.2888278702591385,
+      "grad_norm": 2.3873415849252693,
+      "learning_rate": 1.6688000861004644e-05,
+      "loss": 1.7811,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2889994851553115,
+      "grad_norm": 2.4419197472336562,
+      "learning_rate": 1.6683867498188277e-05,
+      "loss": 1.697,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28917110005148444,
+      "grad_norm": 2.335833050990565,
+      "learning_rate": 1.6679732070356677e-05,
+      "loss": 1.5549,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28934271494765745,
+      "grad_norm": 2.462280441941548,
+      "learning_rate": 1.6675594578787513e-05,
+      "loss": 1.5161,
+      "step": 1686
+    },
+    {
+      "epoch": 0.28951432984383046,
+      "grad_norm": 2.71573419847621,
+      "learning_rate": 1.667145502475907e-05,
+      "loss": 1.5512,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2896859447400034,
+      "grad_norm": 3.1488969101767545,
+      "learning_rate": 1.6667313409550294e-05,
+      "loss": 1.5307,
+      "step": 1688
+    },
+    {
+      "epoch": 0.28985755963617643,
+      "grad_norm": 2.624929739552495,
+      "learning_rate": 1.666316973444076e-05,
+      "loss": 1.6547,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2900291745323494,
+      "grad_norm": 3.4616030530486097,
+      "learning_rate": 1.6659024000710667e-05,
+      "loss": 1.6294,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2902007894285224,
+      "grad_norm": 3.306567426869511,
+      "learning_rate": 1.6654876209640868e-05,
+      "loss": 1.7439,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2903724043246954,
+      "grad_norm": 3.4716458157336962,
+      "learning_rate": 1.6650726362512846e-05,
+      "loss": 1.9092,
+      "step": 1692
+    },
+    {
+      "epoch": 0.29054401922086837,
+      "grad_norm": 2.7797580081438875,
+      "learning_rate": 1.6646574460608717e-05,
+      "loss": 1.7246,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2907156341170414,
+      "grad_norm": 3.1061958235586906,
+      "learning_rate": 1.664242050521123e-05,
+      "loss": 1.9607,
+      "step": 1694
+    },
+    {
+      "epoch": 0.29088724901321433,
+      "grad_norm": 2.8194543111195305,
+      "learning_rate": 1.663826449760377e-05,
+      "loss": 1.6998,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29105886390938734,
+      "grad_norm": 2.682792652281103,
+      "learning_rate": 1.6634106439070357e-05,
+      "loss": 1.8813,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2912304788055603,
+      "grad_norm": 4.0851173561301595,
+      "learning_rate": 1.6629946330895647e-05,
+      "loss": 1.7592,
+      "step": 1697
+    },
+    {
+      "epoch": 0.2914020937017333,
+      "grad_norm": 2.7368511886836533,
+      "learning_rate": 1.662578417436493e-05,
+      "loss": 1.7512,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2915737085979063,
+      "grad_norm": 2.6454508121555556,
+      "learning_rate": 1.6621619970764117e-05,
+      "loss": 1.7595,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2917453234940793,
+      "grad_norm": 3.133916992668201,
+      "learning_rate": 1.6617453721379763e-05,
+      "loss": 1.6196,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2919169383902523,
+      "grad_norm": 2.3737226116721923,
+      "learning_rate": 1.661328542749906e-05,
+      "loss": 1.7328,
+      "step": 1701
+    },
+    {
+      "epoch": 0.29208855328642525,
+      "grad_norm": 2.395553125597497,
+      "learning_rate": 1.6609115090409814e-05,
+      "loss": 1.3693,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29226016818259826,
+      "grad_norm": 2.3687314750831177,
+      "learning_rate": 1.6604942711400478e-05,
+      "loss": 1.5641,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2924317830787712,
+      "grad_norm": 2.7890356531540714,
+      "learning_rate": 1.6600768291760124e-05,
+      "loss": 1.7778,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2926033979749442,
+      "grad_norm": 2.8093810531534555,
+      "learning_rate": 1.659659183277847e-05,
+      "loss": 1.8098,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29277501287111724,
+      "grad_norm": 2.8532103382386147,
+      "learning_rate": 1.6592413335745845e-05,
+      "loss": 1.9331,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2929466277672902,
+      "grad_norm": 2.253964944522325,
+      "learning_rate": 1.6588232801953223e-05,
+      "loss": 1.7093,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2931182426634632,
+      "grad_norm": 2.6162674723508275,
+      "learning_rate": 1.6584050232692198e-05,
+      "loss": 1.8555,
+      "step": 1708
+    },
+    {
+      "epoch": 0.29328985755963616,
+      "grad_norm": 3.127102212282499,
+      "learning_rate": 1.6579865629254994e-05,
+      "loss": 1.83,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29346147245580917,
+      "grad_norm": 2.739809753957868,
+      "learning_rate": 1.6575678992934472e-05,
+      "loss": 1.6619,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2936330873519821,
+      "grad_norm": 3.234910572210153,
+      "learning_rate": 1.6571490325024114e-05,
+      "loss": 1.693,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29380470224815514,
+      "grad_norm": 2.591279935651367,
+      "learning_rate": 1.656729962681803e-05,
+      "loss": 1.8893,
+      "step": 1712
+    },
+    {
+      "epoch": 0.29397631714432815,
+      "grad_norm": 2.3652394414385993,
+      "learning_rate": 1.656310689961095e-05,
+      "loss": 1.6505,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2941479320405011,
+      "grad_norm": 2.73000103313852,
+      "learning_rate": 1.6558912144698247e-05,
+      "loss": 1.8394,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2943195469366741,
+      "grad_norm": 3.0177166287416166,
+      "learning_rate": 1.655471536337591e-05,
+      "loss": 1.6607,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29449116183284707,
+      "grad_norm": 4.053047444419392,
+      "learning_rate": 1.655051655694056e-05,
+      "loss": 1.7652,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2946627767290201,
+      "grad_norm": 2.33951076411324,
+      "learning_rate": 1.6546315726689432e-05,
+      "loss": 1.6842,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2948343916251931,
+      "grad_norm": 2.5701408895094944,
+      "learning_rate": 1.65421128739204e-05,
+      "loss": 1.3195,
+      "step": 1718
+    },
+    {
+      "epoch": 0.29500600652136605,
+      "grad_norm": 3.0616093421212356,
+      "learning_rate": 1.6537907999931958e-05,
+      "loss": 2.0091,
+      "step": 1719
+    },
+    {
+      "epoch": 0.29517762141753906,
+      "grad_norm": 2.4744924345100214,
+      "learning_rate": 1.6533701106023216e-05,
+      "loss": 1.5621,
+      "step": 1720
+    },
+    {
+      "epoch": 0.295349236313712,
+      "grad_norm": 2.462587815356701,
+      "learning_rate": 1.6529492193493923e-05,
+      "loss": 1.6705,
+      "step": 1721
+    },
+    {
+      "epoch": 0.295520851209885,
+      "grad_norm": 2.36585828138444,
+      "learning_rate": 1.6525281263644437e-05,
+      "loss": 1.8422,
+      "step": 1722
+    },
+    {
+      "epoch": 0.295692466106058,
+      "grad_norm": 3.0389682522565264,
+      "learning_rate": 1.6521068317775756e-05,
+      "loss": 1.7972,
+      "step": 1723
+    },
+    {
+      "epoch": 0.295864081002231,
+      "grad_norm": 3.001886018804657,
+      "learning_rate": 1.6516853357189483e-05,
+      "loss": 1.7513,
+      "step": 1724
+    },
+    {
+      "epoch": 0.296035695898404,
+      "grad_norm": 2.6329209350646017,
+      "learning_rate": 1.6512636383187853e-05,
+      "loss": 1.8232,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29620731079457696,
+      "grad_norm": 3.324148958321875,
+      "learning_rate": 1.6508417397073727e-05,
+      "loss": 1.8672,
+      "step": 1726
+    },
+    {
+      "epoch": 0.29637892569075,
+      "grad_norm": 2.790546686918112,
+      "learning_rate": 1.6504196400150578e-05,
+      "loss": 1.7081,
+      "step": 1727
+    },
+    {
+      "epoch": 0.29655054058692293,
+      "grad_norm": 2.1758807378295266,
+      "learning_rate": 1.6499973393722504e-05,
+      "loss": 1.8012,
+      "step": 1728
+    },
+    {
+      "epoch": 0.29672215548309594,
+      "grad_norm": 3.22765072068188,
+      "learning_rate": 1.6495748379094227e-05,
+      "loss": 1.725,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2968937703792689,
+      "grad_norm": 2.6520320172964196,
+      "learning_rate": 1.6491521357571086e-05,
+      "loss": 1.729,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2970653852754419,
+      "grad_norm": 2.627842688883535,
+      "learning_rate": 1.6487292330459047e-05,
+      "loss": 1.7606,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2972370001716149,
+      "grad_norm": 2.207401780762021,
+      "learning_rate": 1.648306129906468e-05,
+      "loss": 1.4981,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2974086150677879,
+      "grad_norm": 2.536885985861452,
+      "learning_rate": 1.6478828264695183e-05,
+      "loss": 1.6621,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2975802299639609,
+      "grad_norm": 4.213339771160178,
+      "learning_rate": 1.647459322865838e-05,
+      "loss": 1.4314,
+      "step": 1734
+    },
+    {
+      "epoch": 0.29775184486013384,
+      "grad_norm": 3.15147452213016,
+      "learning_rate": 1.647035619226271e-05,
+      "loss": 1.9904,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29792345975630685,
+      "grad_norm": 3.0136237246467283,
+      "learning_rate": 1.646611715681722e-05,
+      "loss": 1.8061,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2980950746524798,
+      "grad_norm": 2.5644202184853824,
+      "learning_rate": 1.6461876123631587e-05,
+      "loss": 1.7583,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2982666895486528,
+      "grad_norm": 4.084269607841627,
+      "learning_rate": 1.64576330940161e-05,
+      "loss": 1.8055,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29843830444482583,
+      "grad_norm": 3.278608000435779,
+      "learning_rate": 1.645338806928166e-05,
+      "loss": 1.4704,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2986099193409988,
+      "grad_norm": 2.774297752748693,
+      "learning_rate": 1.644914105073979e-05,
+      "loss": 1.724,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2987815342371718,
+      "grad_norm": 2.9731654672955283,
+      "learning_rate": 1.644489203970263e-05,
+      "loss": 2.0269,
+      "step": 1741
+    },
+    {
+      "epoch": 0.29895314913334475,
+      "grad_norm": 2.442395412713044,
+      "learning_rate": 1.6440641037482934e-05,
+      "loss": 1.5501,
+      "step": 1742
+    },
+    {
+      "epoch": 0.29912476402951776,
+      "grad_norm": 3.259578702903151,
+      "learning_rate": 1.6436388045394072e-05,
+      "loss": 1.7035,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2992963789256908,
+      "grad_norm": 3.204422537908801,
+      "learning_rate": 1.6432133064750026e-05,
+      "loss": 1.8385,
+      "step": 1744
+    },
+    {
+      "epoch": 0.29946799382186373,
+      "grad_norm": 6.683879740831648,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 1.6804,
+      "step": 1745
+    },
+    {
+      "epoch": 0.29963960871803674,
+      "grad_norm": 3.127057468542608,
+      "learning_rate": 1.642361714305539e-05,
+      "loss": 1.674,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2998112236142097,
+      "grad_norm": 2.718917309799963,
+      "learning_rate": 1.641935620463584e-05,
+      "loss": 1.7098,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2999828385103827,
+      "grad_norm": 2.550536402175234,
+      "learning_rate": 1.641509328292318e-05,
+      "loss": 1.6888,
+      "step": 1748
+    },
+    {
+      "epoch": 0.30015445340655567,
+      "grad_norm": 3.3331862042007745,
+      "learning_rate": 1.6410828379234464e-05,
+      "loss": 1.9296,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3003260683027287,
+      "grad_norm": 2.9491079315939768,
+      "learning_rate": 1.640656149488735e-05,
+      "loss": 1.8108,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3004976831989017,
+      "grad_norm": 3.1241231978724366,
+      "learning_rate": 1.640229263120012e-05,
+      "loss": 1.7843,
+      "step": 1751
+    },
+    {
+      "epoch": 0.30066929809507464,
+      "grad_norm": 3.121422839882914,
+      "learning_rate": 1.6398021789491663e-05,
+      "loss": 1.6403,
+      "step": 1752
+    },
+    {
+      "epoch": 0.30084091299124766,
+      "grad_norm": 3.6887422859317414,
+      "learning_rate": 1.6393748971081475e-05,
+      "loss": 1.7536,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3010125278874206,
+      "grad_norm": 2.587404898098536,
+      "learning_rate": 1.6389474177289663e-05,
+      "loss": 1.471,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3011841427835936,
+      "grad_norm": 2.2049938633288613,
+      "learning_rate": 1.6385197409436952e-05,
+      "loss": 1.4469,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3013557576797666,
+      "grad_norm": 2.6534542979014883,
+      "learning_rate": 1.6380918668844663e-05,
+      "loss": 1.7421,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3015273725759396,
+      "grad_norm": 2.5086732034298667,
+      "learning_rate": 1.6376637956834743e-05,
+      "loss": 1.4126,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3016989874721126,
+      "grad_norm": 3.037704432094303,
+      "learning_rate": 1.6372355274729735e-05,
+      "loss": 1.6292,
+      "step": 1758
+    },
+    {
+      "epoch": 0.30187060236828556,
+      "grad_norm": 2.3380990696325465,
+      "learning_rate": 1.6368070623852794e-05,
+      "loss": 1.8651,
+      "step": 1759
+    },
+    {
+      "epoch": 0.30204221726445857,
+      "grad_norm": 2.472032221925406,
+      "learning_rate": 1.6363784005527693e-05,
+      "loss": 1.6272,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3022138321606315,
+      "grad_norm": 2.63741323464442,
+      "learning_rate": 1.6359495421078802e-05,
+      "loss": 1.9139,
+      "step": 1761
+    },
+    {
+      "epoch": 0.30238544705680453,
+      "grad_norm": 4.346864817646478,
+      "learning_rate": 1.6355204871831094e-05,
+      "loss": 1.9148,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3025570619529775,
+      "grad_norm": 2.82903428799297,
+      "learning_rate": 1.635091235911016e-05,
+      "loss": 1.8922,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3027286768491505,
+      "grad_norm": 2.7497813392873707,
+      "learning_rate": 1.6346617884242204e-05,
+      "loss": 1.6227,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3029002917453235,
+      "grad_norm": 2.730719694488057,
+      "learning_rate": 1.634232144855401e-05,
+      "loss": 1.788,
+      "step": 1765
+    },
+    {
+      "epoch": 0.30307190664149647,
+      "grad_norm": 2.668694156588302,
+      "learning_rate": 1.6338023053372994e-05,
+      "loss": 1.6601,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3032435215376695,
+      "grad_norm": 2.569734190081979,
+      "learning_rate": 1.6333722700027163e-05,
+      "loss": 1.656,
+      "step": 1767
+    },
+    {
+      "epoch": 0.30341513643384244,
+      "grad_norm": 3.060594153632445,
+      "learning_rate": 1.632942038984514e-05,
+      "loss": 1.6883,
+      "step": 1768
+    },
+    {
+      "epoch": 0.30358675133001545,
+      "grad_norm": 2.382066490108781,
+      "learning_rate": 1.6325116124156136e-05,
+      "loss": 1.7599,
+      "step": 1769
+    },
+    {
+      "epoch": 0.30375836622618846,
+      "grad_norm": 2.4498203789125514,
+      "learning_rate": 1.6320809904289982e-05,
+      "loss": 1.5753,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3039299811223614,
+      "grad_norm": 3.588753215770666,
+      "learning_rate": 1.6316501731577103e-05,
+      "loss": 1.6,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3041015960185344,
+      "grad_norm": 2.764391826647667,
+      "learning_rate": 1.6312191607348537e-05,
+      "loss": 1.9093,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3042732109147074,
+      "grad_norm": 2.8390715985245527,
+      "learning_rate": 1.6307879532935915e-05,
+      "loss": 1.7381,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3044448258108804,
+      "grad_norm": 2.6472469549054054,
+      "learning_rate": 1.6303565509671476e-05,
+      "loss": 1.6926,
+      "step": 1774
+    },
+    {
+      "epoch": 0.30461644070705335,
+      "grad_norm": 2.722237747392849,
+      "learning_rate": 1.629924953888806e-05,
+      "loss": 1.7755,
+      "step": 1775
+    },
+    {
+      "epoch": 0.30478805560322636,
+      "grad_norm": 2.45239831264883,
+      "learning_rate": 1.6294931621919102e-05,
+      "loss": 1.7763,
+      "step": 1776
+    },
+    {
+      "epoch": 0.30495967049939937,
+      "grad_norm": 2.3149445676952984,
+      "learning_rate": 1.6290611760098655e-05,
+      "loss": 1.5728,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3051312853955723,
+      "grad_norm": 2.8559190565308463,
+      "learning_rate": 1.6286289954761356e-05,
+      "loss": 1.9329,
+      "step": 1778
+    },
+    {
+      "epoch": 0.30530290029174534,
+      "grad_norm": 2.3993261673035615,
+      "learning_rate": 1.628196620724245e-05,
+      "loss": 1.4961,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3054745151879183,
+      "grad_norm": 2.4723113811590536,
+      "learning_rate": 1.627764051887778e-05,
+      "loss": 1.9386,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3056461300840913,
+      "grad_norm": 3.044971374821416,
+      "learning_rate": 1.6273312891003795e-05,
+      "loss": 1.7095,
+      "step": 1781
+    },
+    {
+      "epoch": 0.30581774498026426,
+      "grad_norm": 2.6244314868994714,
+      "learning_rate": 1.6268983324957536e-05,
+      "loss": 1.5359,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3059893598764373,
+      "grad_norm": 3.156057128599002,
+      "learning_rate": 1.6264651822076636e-05,
+      "loss": 1.9254,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3061609747726103,
+      "grad_norm": 3.1314213120396066,
+      "learning_rate": 1.6260318383699342e-05,
+      "loss": 1.707,
+      "step": 1784
+    },
+    {
+      "epoch": 0.30633258966878324,
+      "grad_norm": 2.5100764173079617,
+      "learning_rate": 1.62559830111645e-05,
+      "loss": 1.4651,
+      "step": 1785
+    },
+    {
+      "epoch": 0.30650420456495625,
+      "grad_norm": 2.7570370383903144,
+      "learning_rate": 1.625164570581153e-05,
+      "loss": 1.7877,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3066758194611292,
+      "grad_norm": 2.172394923864771,
+      "learning_rate": 1.6247306468980475e-05,
+      "loss": 1.7598,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3068474343573022,
+      "grad_norm": 2.1017876594405087,
+      "learning_rate": 1.624296530201196e-05,
+      "loss": 1.5646,
+      "step": 1788
+    },
+    {
+      "epoch": 0.30701904925347523,
+      "grad_norm": 2.7120160979800514,
+      "learning_rate": 1.6238622206247218e-05,
+      "loss": 1.6566,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3071906641496482,
+      "grad_norm": 2.6644639905803933,
+      "learning_rate": 1.623427718302806e-05,
+      "loss": 1.9006,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3073622790458212,
+      "grad_norm": 3.936265447486098,
+      "learning_rate": 1.6229930233696913e-05,
+      "loss": 1.7847,
+      "step": 1791
+    },
+    {
+      "epoch": 0.30753389394199415,
+      "grad_norm": 2.328242938953495,
+      "learning_rate": 1.622558135959678e-05,
+      "loss": 1.6377,
+      "step": 1792
+    },
+    {
+      "epoch": 0.30770550883816716,
+      "grad_norm": 3.427588101333531,
+      "learning_rate": 1.622123056207128e-05,
+      "loss": 1.7046,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3078771237343401,
+      "grad_norm": 2.6373744651708613,
+      "learning_rate": 1.62168778424646e-05,
+      "loss": 1.7134,
+      "step": 1794
+    },
+    {
+      "epoch": 0.30804873863051313,
+      "grad_norm": 2.7219651990903384,
+      "learning_rate": 1.6212523202121547e-05,
+      "loss": 1.7167,
+      "step": 1795
+    },
+    {
+      "epoch": 0.30822035352668614,
+      "grad_norm": 2.963670856455669,
+      "learning_rate": 1.6208166642387502e-05,
+      "loss": 1.7537,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3083919684228591,
+      "grad_norm": 2.5976787611875096,
+      "learning_rate": 1.6203808164608454e-05,
+      "loss": 1.7955,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3085635833190321,
+      "grad_norm": 2.5656805423137263,
+      "learning_rate": 1.619944777013097e-05,
+      "loss": 1.6677,
+      "step": 1798
+    },
+    {
+      "epoch": 0.30873519821520506,
+      "grad_norm": 3.005673973507895,
+      "learning_rate": 1.619508546030222e-05,
+      "loss": 1.5713,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3089068131113781,
+      "grad_norm": 2.962738495974791,
+      "learning_rate": 1.619072123646996e-05,
+      "loss": 1.9836,
+      "step": 1800
+    },
+    {
+      "epoch": 0.30907842800755103,
+      "grad_norm": 2.9875705315600465,
+      "learning_rate": 1.6186355099982537e-05,
+      "loss": 1.7827,
+      "step": 1801
+    },
+    {
+      "epoch": 0.30925004290372404,
+      "grad_norm": 2.6151243091238814,
+      "learning_rate": 1.61819870521889e-05,
+      "loss": 1.6703,
+      "step": 1802
+    },
+    {
+      "epoch": 0.30942165779989705,
+      "grad_norm": 3.204141538019496,
+      "learning_rate": 1.6177617094438572e-05,
+      "loss": 1.7566,
+      "step": 1803
+    },
+    {
+      "epoch": 0.30959327269607,
+      "grad_norm": 2.740639238720865,
+      "learning_rate": 1.617324522808168e-05,
+      "loss": 1.5124,
+      "step": 1804
+    },
+    {
+      "epoch": 0.309764887592243,
+      "grad_norm": 3.396120257408002,
+      "learning_rate": 1.6168871454468925e-05,
+      "loss": 1.6964,
+      "step": 1805
+    },
+    {
+      "epoch": 0.309936502488416,
+      "grad_norm": 2.3611667032023935,
+      "learning_rate": 1.616449577495162e-05,
+      "loss": 1.7037,
+      "step": 1806
+    },
+    {
+      "epoch": 0.310108117384589,
+      "grad_norm": 2.929923445896195,
+      "learning_rate": 1.6160118190881643e-05,
+      "loss": 1.9567,
+      "step": 1807
+    },
+    {
+      "epoch": 0.31027973228076194,
+      "grad_norm": 2.528745624589099,
+      "learning_rate": 1.615573870361147e-05,
+      "loss": 1.54,
+      "step": 1808
+    },
+    {
+      "epoch": 0.31045134717693496,
+      "grad_norm": 3.603022535568774,
+      "learning_rate": 1.615135731449418e-05,
+      "loss": 1.9665,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31062296207310797,
+      "grad_norm": 2.6122507840237046,
+      "learning_rate": 1.6146974024883414e-05,
+      "loss": 1.6567,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3107945769692809,
+      "grad_norm": 2.581608715522504,
+      "learning_rate": 1.6142588836133413e-05,
+      "loss": 1.6351,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31096619186545393,
+      "grad_norm": 2.547061230280792,
+      "learning_rate": 1.6138201749599e-05,
+      "loss": 1.6298,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3111378067616269,
+      "grad_norm": 2.877789041695864,
+      "learning_rate": 1.6133812766635603e-05,
+      "loss": 1.9615,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3113094216577999,
+      "grad_norm": 2.679265069265271,
+      "learning_rate": 1.61294218885992e-05,
+      "loss": 1.6718,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3114810365539729,
+      "grad_norm": 2.701284834429953,
+      "learning_rate": 1.6125029116846394e-05,
+      "loss": 1.6384,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31165265145014587,
+      "grad_norm": 2.908716215284335,
+      "learning_rate": 1.6120634452734346e-05,
+      "loss": 1.7692,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3118242663463189,
+      "grad_norm": 2.9046631749123666,
+      "learning_rate": 1.611623789762081e-05,
+      "loss": 1.8231,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31199588124249183,
+      "grad_norm": 2.3493255594838502,
+      "learning_rate": 1.6111839452864123e-05,
+      "loss": 1.433,
+      "step": 1818
+    },
+    {
+      "epoch": 0.31216749613866485,
+      "grad_norm": 2.40880166929836,
+      "learning_rate": 1.610743911982321e-05,
+      "loss": 1.6991,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3123391110348378,
+      "grad_norm": 3.0096267354692903,
+      "learning_rate": 1.6103036899857574e-05,
+      "loss": 1.6666,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3125107259310108,
+      "grad_norm": 2.5063850198276074,
+      "learning_rate": 1.6098632794327308e-05,
+      "loss": 1.7437,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3126823408271838,
+      "grad_norm": 2.769509742980168,
+      "learning_rate": 1.609422680459308e-05,
+      "loss": 1.8533,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3128539557233568,
+      "grad_norm": 2.1296962849740355,
+      "learning_rate": 1.608981893201614e-05,
+      "loss": 1.6587,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3130255706195298,
+      "grad_norm": 2.815288268734331,
+      "learning_rate": 1.608540917795833e-05,
+      "loss": 1.8576,
+      "step": 1824
+    },
+    {
+      "epoch": 0.31319718551570275,
+      "grad_norm": 2.755759963811611,
+      "learning_rate": 1.6080997543782063e-05,
+      "loss": 1.5347,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31336880041187576,
+      "grad_norm": 2.6526205228509063,
+      "learning_rate": 1.6076584030850333e-05,
+      "loss": 1.701,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3135404153080487,
+      "grad_norm": 2.814971385706839,
+      "learning_rate": 1.607216864052672e-05,
+      "loss": 1.9335,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3137120302042217,
+      "grad_norm": 2.662273618993851,
+      "learning_rate": 1.6067751374175383e-05,
+      "loss": 1.8275,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31388364510039474,
+      "grad_norm": 2.7467918376897575,
+      "learning_rate": 1.6063332233161055e-05,
+      "loss": 1.7759,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3140552599965677,
+      "grad_norm": 2.737059645175828,
+      "learning_rate": 1.6058911218849058e-05,
+      "loss": 1.7352,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3142268748927407,
+      "grad_norm": 2.9202493578050563,
+      "learning_rate": 1.6054488332605282e-05,
+      "loss": 1.8544,
+      "step": 1831
+    },
+    {
+      "epoch": 0.31439848978891366,
+      "grad_norm": 2.698061435397696,
+      "learning_rate": 1.6050063575796205e-05,
+      "loss": 1.6063,
+      "step": 1832
+    },
+    {
+      "epoch": 0.31457010468508667,
+      "grad_norm": 3.0613954905950136,
+      "learning_rate": 1.6045636949788875e-05,
+      "loss": 2.0303,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3147417195812596,
+      "grad_norm": 3.6848696029277663,
+      "learning_rate": 1.6041208455950924e-05,
+      "loss": 1.7077,
+      "step": 1834
+    },
+    {
+      "epoch": 0.31491333447743264,
+      "grad_norm": 2.532169487777431,
+      "learning_rate": 1.603677809565055e-05,
+      "loss": 1.8292,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31508494937360565,
+      "grad_norm": 2.5141319417602026,
+      "learning_rate": 1.6032345870256546e-05,
+      "loss": 1.591,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3152565642697786,
+      "grad_norm": 2.8644374084060997,
+      "learning_rate": 1.6027911781138263e-05,
+      "loss": 1.5595,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3154281791659516,
+      "grad_norm": 3.4260682868759376,
+      "learning_rate": 1.6023475829665643e-05,
+      "loss": 1.6461,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3155997940621246,
+      "grad_norm": 2.157983539703558,
+      "learning_rate": 1.6019038017209186e-05,
+      "loss": 1.3875,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3157714089582976,
+      "grad_norm": 2.776073279910533,
+      "learning_rate": 1.6014598345139992e-05,
+      "loss": 1.5049,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3159430238544706,
+      "grad_norm": 2.647326741111861,
+      "learning_rate": 1.6010156814829704e-05,
+      "loss": 1.6051,
+      "step": 1841
+    },
+    {
+      "epoch": 0.31611463875064355,
+      "grad_norm": 3.4766660709705284,
+      "learning_rate": 1.6005713427650564e-05,
+      "loss": 1.9456,
+      "step": 1842
+    },
+    {
+      "epoch": 0.31628625364681656,
+      "grad_norm": 2.6280522302863423,
+      "learning_rate": 1.600126818497538e-05,
+      "loss": 1.7731,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3164578685429895,
+      "grad_norm": 2.6315015095568346,
+      "learning_rate": 1.599682108817753e-05,
+      "loss": 1.6893,
+      "step": 1844
+    },
+    {
+      "epoch": 0.31662948343916253,
+      "grad_norm": 2.5175976048693265,
+      "learning_rate": 1.599237213863097e-05,
+      "loss": 1.4549,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3168010983353355,
+      "grad_norm": 2.7836272032881113,
+      "learning_rate": 1.5987921337710224e-05,
+      "loss": 1.764,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3169727132315085,
+      "grad_norm": 2.7474789235034867,
+      "learning_rate": 1.5983468686790386e-05,
+      "loss": 1.902,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3171443281276815,
+      "grad_norm": 2.497226916613902,
+      "learning_rate": 1.5979014187247132e-05,
+      "loss": 1.7118,
+      "step": 1848
+    },
+    {
+      "epoch": 0.31731594302385446,
+      "grad_norm": 2.5860230214370237,
+      "learning_rate": 1.59745578404567e-05,
+      "loss": 1.7818,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3174875579200275,
+      "grad_norm": 2.8044051230590212,
+      "learning_rate": 1.59700996477959e-05,
+      "loss": 1.3873,
+      "step": 1850
+    },
+    {
+      "epoch": 0.31765917281620043,
+      "grad_norm": 3.5705993055093193,
+      "learning_rate": 1.5965639610642117e-05,
+      "loss": 1.878,
+      "step": 1851
+    },
+    {
+      "epoch": 0.31783078771237344,
+      "grad_norm": 2.604514100015129,
+      "learning_rate": 1.5961177730373298e-05,
+      "loss": 1.8874,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3180024026085464,
+      "grad_norm": 2.540428213850227,
+      "learning_rate": 1.5956714008367968e-05,
+      "loss": 1.5936,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3181740175047194,
+      "grad_norm": 2.7223857558162554,
+      "learning_rate": 1.5952248446005213e-05,
+      "loss": 1.7457,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3183456324008924,
+      "grad_norm": 2.6516437941224007,
+      "learning_rate": 1.5947781044664696e-05,
+      "loss": 1.758,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3185172472970654,
+      "grad_norm": 2.3760378743832264,
+      "learning_rate": 1.5943311805726637e-05,
+      "loss": 1.6384,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3186888621932384,
+      "grad_norm": 2.5776847113896117,
+      "learning_rate": 1.5938840730571836e-05,
+      "loss": 1.786,
+      "step": 1857
+    },
+    {
+      "epoch": 0.31886047708941134,
+      "grad_norm": 2.2439622164339497,
+      "learning_rate": 1.5934367820581654e-05,
+      "loss": 1.426,
+      "step": 1858
+    },
+    {
+      "epoch": 0.31903209198558435,
+      "grad_norm": 2.5053452595628007,
+      "learning_rate": 1.592989307713802e-05,
+      "loss": 1.8876,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3192037068817573,
+      "grad_norm": 2.645167967187714,
+      "learning_rate": 1.5925416501623427e-05,
+      "loss": 1.5788,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3193753217779303,
+      "grad_norm": 2.899862135235689,
+      "learning_rate": 1.5920938095420934e-05,
+      "loss": 1.5669,
+      "step": 1861
+    },
+    {
+      "epoch": 0.31954693667410333,
+      "grad_norm": 2.750590749978957,
+      "learning_rate": 1.591645785991417e-05,
+      "loss": 1.7793,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3197185515702763,
+      "grad_norm": 2.8164409715169536,
+      "learning_rate": 1.5911975796487324e-05,
+      "loss": 1.7453,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3198901664664493,
+      "grad_norm": 3.347725327106241,
+      "learning_rate": 1.5907491906525162e-05,
+      "loss": 1.9353,
+      "step": 1864
+    },
+    {
+      "epoch": 0.32006178136262226,
+      "grad_norm": 2.6253304162682403,
+      "learning_rate": 1.5903006191412996e-05,
+      "loss": 1.7538,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32023339625879527,
+      "grad_norm": 3.4345456825590195,
+      "learning_rate": 1.5898518652536713e-05,
+      "loss": 1.9821,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3204050111549683,
+      "grad_norm": 3.147781304132002,
+      "learning_rate": 1.589402929128276e-05,
+      "loss": 1.8258,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32057662605114123,
+      "grad_norm": 3.146856291330822,
+      "learning_rate": 1.588953810903815e-05,
+      "loss": 1.9101,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32074824094731424,
+      "grad_norm": 2.568333247726322,
+      "learning_rate": 1.588504510719046e-05,
+      "loss": 1.4917,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3209198558434872,
+      "grad_norm": 3.3756964269092795,
+      "learning_rate": 1.588055028712782e-05,
+      "loss": 1.8438,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3210914707396602,
+      "grad_norm": 2.872424762146188,
+      "learning_rate": 1.5876053650238927e-05,
+      "loss": 1.802,
+      "step": 1871
+    },
+    {
+      "epoch": 0.32126308563583317,
+      "grad_norm": 2.9384308438565214,
+      "learning_rate": 1.5871555197913045e-05,
+      "loss": 1.5228,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3214347005320062,
+      "grad_norm": 2.49210532271169,
+      "learning_rate": 1.5867054931539995e-05,
+      "loss": 1.4623,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3216063154281792,
+      "grad_norm": 2.895751148600379,
+      "learning_rate": 1.5862552852510152e-05,
+      "loss": 1.9324,
+      "step": 1874
+    },
+    {
+      "epoch": 0.32177793032435215,
+      "grad_norm": 2.9423350219662696,
+      "learning_rate": 1.5858048962214457e-05,
+      "loss": 1.7377,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32194954522052516,
+      "grad_norm": 3.367244935430937,
+      "learning_rate": 1.585354326204442e-05,
+      "loss": 1.8729,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3221211601166981,
+      "grad_norm": 2.4835125115604066,
+      "learning_rate": 1.5849035753392085e-05,
+      "loss": 1.8261,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3222927750128711,
+      "grad_norm": 2.5006489815683213,
+      "learning_rate": 1.5844526437650077e-05,
+      "loss": 1.698,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3224643899090441,
+      "grad_norm": 3.0168484659777333,
+      "learning_rate": 1.584001531621158e-05,
+      "loss": 1.9001,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3226360048052171,
+      "grad_norm": 3.108095293381667,
+      "learning_rate": 1.5835502390470313e-05,
+      "loss": 1.704,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3228076197013901,
+      "grad_norm": 3.7276105585304857,
+      "learning_rate": 1.583098766182058e-05,
+      "loss": 1.6952,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32297923459756306,
+      "grad_norm": 2.3685535660673667,
+      "learning_rate": 1.5826471131657227e-05,
+      "loss": 1.5413,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32315084949373607,
+      "grad_norm": 2.7584718545490277,
+      "learning_rate": 1.5821952801375653e-05,
+      "loss": 1.6635,
+      "step": 1883
+    },
+    {
+      "epoch": 0.323322464389909,
+      "grad_norm": 2.928502150111893,
+      "learning_rate": 1.5817432672371826e-05,
+      "loss": 1.8562,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32349407928608204,
+      "grad_norm": 2.5635450690970223,
+      "learning_rate": 1.581291074604226e-05,
+      "loss": 1.7794,
+      "step": 1885
+    },
+    {
+      "epoch": 0.323665694182255,
+      "grad_norm": 2.5824971920222155,
+      "learning_rate": 1.5808387023784025e-05,
+      "loss": 1.5943,
+      "step": 1886
+    },
+    {
+      "epoch": 0.323837309078428,
+      "grad_norm": 2.8634913336222794,
+      "learning_rate": 1.5803861506994755e-05,
+      "loss": 1.8735,
+      "step": 1887
+    },
+    {
+      "epoch": 0.324008923974601,
+      "grad_norm": 2.8348081198018935,
+      "learning_rate": 1.579933419707263e-05,
+      "loss": 1.8429,
+      "step": 1888
+    },
+    {
+      "epoch": 0.32418053887077397,
+      "grad_norm": 2.764798734979702,
+      "learning_rate": 1.579480509541638e-05,
+      "loss": 1.6969,
+      "step": 1889
+    },
+    {
+      "epoch": 0.324352153766947,
+      "grad_norm": 3.7615717487044424,
+      "learning_rate": 1.5790274203425302e-05,
+      "loss": 1.7232,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32452376866311994,
+      "grad_norm": 3.2134527252429295,
+      "learning_rate": 1.5785741522499232e-05,
+      "loss": 1.8074,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32469538355929295,
+      "grad_norm": 3.321026230072206,
+      "learning_rate": 1.5781207054038567e-05,
+      "loss": 1.8601,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32486699845546596,
+      "grad_norm": 2.9492855961278712,
+      "learning_rate": 1.5776670799444257e-05,
+      "loss": 1.7617,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3250386133516389,
+      "grad_norm": 2.391870319321163,
+      "learning_rate": 1.5772132760117797e-05,
+      "loss": 1.7567,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3252102282478119,
+      "grad_norm": 19.824506052124576,
+      "learning_rate": 1.576759293746124e-05,
+      "loss": 1.6844,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3253818431439849,
+      "grad_norm": 2.6017121668062444,
+      "learning_rate": 1.576305133287719e-05,
+      "loss": 1.5208,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3255534580401579,
+      "grad_norm": 2.078778427972936,
+      "learning_rate": 1.5758507947768794e-05,
+      "loss": 1.6443,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32572507293633085,
+      "grad_norm": 2.643878666518309,
+      "learning_rate": 1.5753962783539755e-05,
+      "loss": 1.6504,
+      "step": 1898
+    },
+    {
+      "epoch": 0.32589668783250386,
+      "grad_norm": 3.038387434675897,
+      "learning_rate": 1.5749415841594328e-05,
+      "loss": 1.5466,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3260683027286769,
+      "grad_norm": 3.5630501290476975,
+      "learning_rate": 1.5744867123337313e-05,
+      "loss": 1.9429,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32623991762484983,
+      "grad_norm": 2.6487908946909977,
+      "learning_rate": 1.5740316630174057e-05,
+      "loss": 1.743,
+      "step": 1901
+    },
+    {
+      "epoch": 0.32641153252102284,
+      "grad_norm": 4.256919439739053,
+      "learning_rate": 1.573576436351046e-05,
+      "loss": 1.4377,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3265831474171958,
+      "grad_norm": 2.7554072056259042,
+      "learning_rate": 1.573121032475297e-05,
+      "loss": 1.7038,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3267547623133688,
+      "grad_norm": 2.503034490131831,
+      "learning_rate": 1.5726654515308584e-05,
+      "loss": 1.7936,
+      "step": 1904
+    },
+    {
+      "epoch": 0.32692637720954176,
+      "grad_norm": 2.487938973126537,
+      "learning_rate": 1.5722096936584834e-05,
+      "loss": 1.6215,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3270979921057148,
+      "grad_norm": 3.861228850303636,
+      "learning_rate": 1.5717537589989813e-05,
+      "loss": 1.7414,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3272696070018878,
+      "grad_norm": 3.6118083360444397,
+      "learning_rate": 1.571297647693215e-05,
+      "loss": 1.8911,
+      "step": 1907
+    },
+    {
+      "epoch": 0.32744122189806074,
+      "grad_norm": 3.741059011806432,
+      "learning_rate": 1.570841359882103e-05,
+      "loss": 1.6635,
+      "step": 1908
+    },
+    {
+      "epoch": 0.32761283679423375,
+      "grad_norm": 3.259956696249125,
+      "learning_rate": 1.5703848957066178e-05,
+      "loss": 1.7691,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3277844516904067,
+      "grad_norm": 2.4287767734968564,
+      "learning_rate": 1.5699282553077855e-05,
+      "loss": 1.7686,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3279560665865797,
+      "grad_norm": 2.4410890155679907,
+      "learning_rate": 1.5694714388266885e-05,
+      "loss": 1.7687,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3281276814827527,
+      "grad_norm": 2.7041966544528284,
+      "learning_rate": 1.569014446404462e-05,
+      "loss": 1.895,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3282992963789257,
+      "grad_norm": 2.6875235258849757,
+      "learning_rate": 1.5685572781822965e-05,
+      "loss": 1.6322,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3284709112750987,
+      "grad_norm": 2.7392496188784095,
+      "learning_rate": 1.568099934301436e-05,
+      "loss": 1.8038,
+      "step": 1914
+    },
+    {
+      "epoch": 0.32864252617127165,
+      "grad_norm": 2.6903997978654797,
+      "learning_rate": 1.5676424149031798e-05,
+      "loss": 1.7666,
+      "step": 1915
+    },
+    {
+      "epoch": 0.32881414106744467,
+      "grad_norm": 3.286426688341425,
+      "learning_rate": 1.5671847201288804e-05,
+      "loss": 1.7416,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3289857559636176,
+      "grad_norm": 3.6508779053828895,
+      "learning_rate": 1.566726850119945e-05,
+      "loss": 1.619,
+      "step": 1917
+    },
+    {
+      "epoch": 0.32915737085979063,
+      "grad_norm": 2.3118380129352367,
+      "learning_rate": 1.566268805017836e-05,
+      "loss": 1.4385,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32932898575596364,
+      "grad_norm": 2.9656824861662607,
+      "learning_rate": 1.565810584964067e-05,
+      "loss": 2.0173,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3295006006521366,
+      "grad_norm": 3.488396055734131,
+      "learning_rate": 1.5653521901002084e-05,
+      "loss": 1.7711,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3296722155483096,
+      "grad_norm": 2.9336110109305236,
+      "learning_rate": 1.564893620567884e-05,
+      "loss": 1.4975,
+      "step": 1921
+    },
+    {
+      "epoch": 0.32984383044448257,
+      "grad_norm": 2.325084121870113,
+      "learning_rate": 1.5644348765087704e-05,
+      "loss": 1.379,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3300154453406556,
+      "grad_norm": 3.4427731001891595,
+      "learning_rate": 1.563975958064599e-05,
+      "loss": 1.7365,
+      "step": 1923
+    },
+    {
+      "epoch": 0.33018706023682853,
+      "grad_norm": 2.9082236290856414,
+      "learning_rate": 1.5635168653771564e-05,
+      "loss": 1.7818,
+      "step": 1924
+    },
+    {
+      "epoch": 0.33035867513300154,
+      "grad_norm": 10.580924998264482,
+      "learning_rate": 1.5630575985882793e-05,
+      "loss": 1.6502,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33053029002917456,
+      "grad_norm": 2.4764236020524204,
+      "learning_rate": 1.5625981578398624e-05,
+      "loss": 1.6864,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3307019049253475,
+      "grad_norm": 3.1169721081121198,
+      "learning_rate": 1.5621385432738516e-05,
+      "loss": 1.6375,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3308735198215205,
+      "grad_norm": 3.5357190805063747,
+      "learning_rate": 1.5616787550322473e-05,
+      "loss": 1.6085,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3310451347176935,
+      "grad_norm": 3.444090357473441,
+      "learning_rate": 1.561218793257103e-05,
+      "loss": 1.8197,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3312167496138665,
+      "grad_norm": 3.535636207498067,
+      "learning_rate": 1.5607586580905273e-05,
+      "loss": 1.6458,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33138836451003945,
+      "grad_norm": 2.5580470797877384,
+      "learning_rate": 1.5602983496746805e-05,
+      "loss": 1.7406,
+      "step": 1931
+    },
+    {
+      "epoch": 0.33155997940621246,
+      "grad_norm": 2.533293873721004,
+      "learning_rate": 1.5598378681517772e-05,
+      "loss": 1.8296,
+      "step": 1932
+    },
+    {
+      "epoch": 0.33173159430238547,
+      "grad_norm": 3.100048834019278,
+      "learning_rate": 1.559377213664086e-05,
+      "loss": 1.5878,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3319032091985584,
+      "grad_norm": 2.41638592842297,
+      "learning_rate": 1.558916386353928e-05,
+      "loss": 1.7117,
+      "step": 1934
+    },
+    {
+      "epoch": 0.33207482409473144,
+      "grad_norm": 3.4660030647597546,
+      "learning_rate": 1.5584553863636783e-05,
+      "loss": 1.5978,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3322464389909044,
+      "grad_norm": 2.628983376924701,
+      "learning_rate": 1.557994213835766e-05,
+      "loss": 1.8324,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3324180538870774,
+      "grad_norm": 2.626539935619845,
+      "learning_rate": 1.5575328689126714e-05,
+      "loss": 1.7423,
+      "step": 1937
+    },
+    {
+      "epoch": 0.33258966878325036,
+      "grad_norm": 2.981021549860697,
+      "learning_rate": 1.5570713517369305e-05,
+      "loss": 1.656,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33276128367942337,
+      "grad_norm": 2.6712964260904624,
+      "learning_rate": 1.5566096624511306e-05,
+      "loss": 1.6129,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3329328985755964,
+      "grad_norm": 3.7767999352759816,
+      "learning_rate": 1.5561478011979133e-05,
+      "loss": 2.0338,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33310451347176934,
+      "grad_norm": 2.752406480350029,
+      "learning_rate": 1.555685768119973e-05,
+      "loss": 1.646,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33327612836794235,
+      "grad_norm": 2.7957883842159035,
+      "learning_rate": 1.5552235633600576e-05,
+      "loss": 1.8039,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3334477432641153,
+      "grad_norm": 2.1821824060511137,
+      "learning_rate": 1.5547611870609672e-05,
+      "loss": 1.5384,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3336193581602883,
+      "grad_norm": 2.5623110844738504,
+      "learning_rate": 1.554298639365555e-05,
+      "loss": 1.834,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3337909730564613,
+      "grad_norm": 2.248116150945698,
+      "learning_rate": 1.5538359204167285e-05,
+      "loss": 1.6514,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3339625879526343,
+      "grad_norm": 2.700377628007367,
+      "learning_rate": 1.553373030357446e-05,
+      "loss": 1.7759,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3341342028488073,
+      "grad_norm": 3.4977059833200115,
+      "learning_rate": 1.5529099693307205e-05,
+      "loss": 1.864,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33430581774498025,
+      "grad_norm": 3.2840104204802123,
+      "learning_rate": 1.5524467374796165e-05,
+      "loss": 1.5823,
+      "step": 1948
+    },
+    {
+      "epoch": 0.33447743264115326,
+      "grad_norm": 3.5584161556558014,
+      "learning_rate": 1.5519833349472523e-05,
+      "loss": 1.69,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3346490475373262,
+      "grad_norm": 2.6078108692928343,
+      "learning_rate": 1.5515197618767985e-05,
+      "loss": 1.8914,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3348206624334992,
+      "grad_norm": 2.7110173387307137,
+      "learning_rate": 1.5510560184114785e-05,
+      "loss": 1.9239,
+      "step": 1951
+    },
+    {
+      "epoch": 0.33499227732967224,
+      "grad_norm": 2.8094977439747564,
+      "learning_rate": 1.5505921046945677e-05,
+      "loss": 1.6266,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3351638922258452,
+      "grad_norm": 2.834909679423824,
+      "learning_rate": 1.5501280208693948e-05,
+      "loss": 1.5762,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3353355071220182,
+      "grad_norm": 2.674892517060095,
+      "learning_rate": 1.5496637670793415e-05,
+      "loss": 2.0049,
+      "step": 1954
+    },
+    {
+      "epoch": 0.33550712201819116,
+      "grad_norm": 3.1138810121668072,
+      "learning_rate": 1.5491993434678404e-05,
+      "loss": 1.7359,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3356787369143642,
+      "grad_norm": 3.103606516363571,
+      "learning_rate": 1.5487347501783783e-05,
+      "loss": 1.6512,
+      "step": 1956
+    },
+    {
+      "epoch": 0.33585035181053713,
+      "grad_norm": 2.43027994642771,
+      "learning_rate": 1.5482699873544937e-05,
+      "loss": 1.8532,
+      "step": 1957
+    },
+    {
+      "epoch": 0.33602196670671014,
+      "grad_norm": 2.6413441206454977,
+      "learning_rate": 1.547805055139777e-05,
+      "loss": 1.7201,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33619358160288315,
+      "grad_norm": 4.0106005749293825,
+      "learning_rate": 1.5473399536778715e-05,
+      "loss": 1.7079,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3363651964990561,
+      "grad_norm": 3.206214786988806,
+      "learning_rate": 1.546874683112473e-05,
+      "loss": 1.7608,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3365368113952291,
+      "grad_norm": 2.6561795997581634,
+      "learning_rate": 1.546409243587329e-05,
+      "loss": 1.8837,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3367084262914021,
+      "grad_norm": 2.58641854118283,
+      "learning_rate": 1.5459436352462395e-05,
+      "loss": 1.6578,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3368800411875751,
+      "grad_norm": 2.768846127982534,
+      "learning_rate": 1.5454778582330567e-05,
+      "loss": 1.6507,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3370516560837481,
+      "grad_norm": 3.2016114245323686,
+      "learning_rate": 1.5450119126916844e-05,
+      "loss": 1.6859,
+      "step": 1964
+    },
+    {
+      "epoch": 0.33722327097992105,
+      "grad_norm": 4.707555669442656,
+      "learning_rate": 1.544545798766079e-05,
+      "loss": 1.8789,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33739488587609406,
+      "grad_norm": 4.461137310127853,
+      "learning_rate": 1.5440795166002492e-05,
+      "loss": 1.9813,
+      "step": 1966
+    },
+    {
+      "epoch": 0.337566500772267,
+      "grad_norm": 2.3129466837544266,
+      "learning_rate": 1.543613066338255e-05,
+      "loss": 1.711,
+      "step": 1967
+    },
+    {
+      "epoch": 0.33773811566844003,
+      "grad_norm": 3.7437345946752036,
+      "learning_rate": 1.5431464481242085e-05,
+      "loss": 1.8767,
+      "step": 1968
+    },
+    {
+      "epoch": 0.337909730564613,
+      "grad_norm": 2.3874183390289034,
+      "learning_rate": 1.5426796621022746e-05,
+      "loss": 1.6342,
+      "step": 1969
+    },
+    {
+      "epoch": 0.338081345460786,
+      "grad_norm": 2.4332292102526085,
+      "learning_rate": 1.542212708416668e-05,
+      "loss": 1.5295,
+      "step": 1970
+    },
+    {
+      "epoch": 0.338252960356959,
+      "grad_norm": 2.79228284272914,
+      "learning_rate": 1.541745587211657e-05,
+      "loss": 1.6769,
+      "step": 1971
+    },
+    {
+      "epoch": 0.33842457525313197,
+      "grad_norm": 2.871427530232779,
+      "learning_rate": 1.541278298631561e-05,
+      "loss": 1.5641,
+      "step": 1972
+    },
+    {
+      "epoch": 0.338596190149305,
+      "grad_norm": 2.5055446947492768,
+      "learning_rate": 1.5408108428207512e-05,
+      "loss": 1.3681,
+      "step": 1973
+    },
+    {
+      "epoch": 0.33876780504547793,
+      "grad_norm": 2.6964311460399637,
+      "learning_rate": 1.540343219923651e-05,
+      "loss": 1.8414,
+      "step": 1974
+    },
+    {
+      "epoch": 0.33893941994165094,
+      "grad_norm": 2.577644863267545,
+      "learning_rate": 1.5398754300847346e-05,
+      "loss": 1.7075,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3391110348378239,
+      "grad_norm": 2.516837313293274,
+      "learning_rate": 1.5394074734485275e-05,
+      "loss": 1.7644,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3392826497339969,
+      "grad_norm": 2.763000775632793,
+      "learning_rate": 1.538939350159608e-05,
+      "loss": 1.8224,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3394542646301699,
+      "grad_norm": 2.697539371400862,
+      "learning_rate": 1.538471060362605e-05,
+      "loss": 1.4923,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3396258795263429,
+      "grad_norm": 3.3502490194731966,
+      "learning_rate": 1.538002604202199e-05,
+      "loss": 1.7466,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3397974944225159,
+      "grad_norm": 2.3468363389541684,
+      "learning_rate": 1.5375339818231213e-05,
+      "loss": 1.6252,
+      "step": 1980
+    },
+    {
+      "epoch": 0.33996910931868884,
+      "grad_norm": 2.7686591333033763,
+      "learning_rate": 1.5370651933701566e-05,
+      "loss": 1.5589,
+      "step": 1981
+    },
+    {
+      "epoch": 0.34014072421486186,
+      "grad_norm": 3.116237920971837,
+      "learning_rate": 1.5365962389881386e-05,
+      "loss": 1.7829,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3403123391110348,
+      "grad_norm": 3.0234370384875153,
+      "learning_rate": 1.536127118821953e-05,
+      "loss": 1.7208,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3404839540072078,
+      "grad_norm": 2.516128941965993,
+      "learning_rate": 1.5356578330165377e-05,
+      "loss": 1.6726,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34065556890338083,
+      "grad_norm": 2.5691896583270966,
+      "learning_rate": 1.53518838171688e-05,
+      "loss": 1.7392,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3408271837995538,
+      "grad_norm": 2.608913854061275,
+      "learning_rate": 1.53471876506802e-05,
+      "loss": 1.8245,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3409987986957268,
+      "grad_norm": 2.564202386189652,
+      "learning_rate": 1.534248983215048e-05,
+      "loss": 1.7224,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34117041359189976,
+      "grad_norm": 3.1016738123791434,
+      "learning_rate": 1.533779036303105e-05,
+      "loss": 1.8091,
+      "step": 1988
+    },
+    {
+      "epoch": 0.34134202848807277,
+      "grad_norm": 3.2323081803279052,
+      "learning_rate": 1.5333089244773844e-05,
+      "loss": 1.7104,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3415136433842458,
+      "grad_norm": 2.9098203649235654,
+      "learning_rate": 1.5328386478831295e-05,
+      "loss": 1.6782,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34168525828041874,
+      "grad_norm": 3.172383459169986,
+      "learning_rate": 1.532368206665634e-05,
+      "loss": 1.7733,
+      "step": 1991
+    },
+    {
+      "epoch": 0.34185687317659175,
+      "grad_norm": 3.656300195554205,
+      "learning_rate": 1.531897600970243e-05,
+      "loss": 1.7086,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3420284880727647,
+      "grad_norm": 2.6344018311027892,
+      "learning_rate": 1.531426830942354e-05,
+      "loss": 1.8706,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3422001029689377,
+      "grad_norm": 4.22851473057233,
+      "learning_rate": 1.5309558967274128e-05,
+      "loss": 1.8577,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34237171786511067,
+      "grad_norm": 2.7353850524020715,
+      "learning_rate": 1.5304847984709166e-05,
+      "loss": 1.7469,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3425433327612837,
+      "grad_norm": 2.680517246832729,
+      "learning_rate": 1.5300135363184145e-05,
+      "loss": 1.5369,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3427149476574567,
+      "grad_norm": 2.8979963835467095,
+      "learning_rate": 1.5295421104155047e-05,
+      "loss": 1.5516,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34288656255362965,
+      "grad_norm": 3.0841503793201888,
+      "learning_rate": 1.529070520907837e-05,
+      "loss": 1.757,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34305817744980266,
+      "grad_norm": 3.082521287986174,
+      "learning_rate": 1.5285987679411114e-05,
+      "loss": 1.8098,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3432297923459756,
+      "grad_norm": 3.029738568939894,
+      "learning_rate": 1.528126851661078e-05,
+      "loss": 1.5466,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3434014072421486,
+      "grad_norm": 2.9328255696848897,
+      "learning_rate": 1.5276547722135385e-05,
+      "loss": 1.4992,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3435730221383216,
+      "grad_norm": 3.014343216455571,
+      "learning_rate": 1.5271825297443436e-05,
+      "loss": 1.7196,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3437446370344946,
+      "grad_norm": 2.92714191232286,
+      "learning_rate": 1.5267101243993954e-05,
+      "loss": 1.7171,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3439162519306676,
+      "grad_norm": 2.8871505875491783,
+      "learning_rate": 1.5262375563246457e-05,
+      "loss": 1.5217,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34408786682684056,
+      "grad_norm": 3.2902156930892623,
+      "learning_rate": 1.525764825666097e-05,
+      "loss": 1.8849,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34425948172301357,
+      "grad_norm": 3.4048275595491377,
+      "learning_rate": 1.5252919325698022e-05,
+      "loss": 1.692,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3444310966191865,
+      "grad_norm": 3.002158590807369,
+      "learning_rate": 1.524818877181863e-05,
+      "loss": 1.7875,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34460271151535954,
+      "grad_norm": 2.854718207276763,
+      "learning_rate": 1.5243456596484339e-05,
+      "loss": 1.7951,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3447743264115325,
+      "grad_norm": 2.4585399399326,
+      "learning_rate": 1.5238722801157164e-05,
+      "loss": 1.8094,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3449459413077055,
+      "grad_norm": 3.269496738087618,
+      "learning_rate": 1.5233987387299643e-05,
+      "loss": 1.7617,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3451175562038785,
+      "grad_norm": 3.1029576986317187,
+      "learning_rate": 1.5229250356374804e-05,
+      "loss": 1.7609,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3452891711000515,
+      "grad_norm": 2.9945120369588176,
+      "learning_rate": 1.5224511709846176e-05,
+      "loss": 1.7917,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3454607859962245,
+      "grad_norm": 3.4834330715867123,
+      "learning_rate": 1.5219771449177794e-05,
+      "loss": 1.72,
+      "step": 2013
+    },
+    {
+      "epoch": 0.34563240089239744,
+      "grad_norm": 2.695216788358231,
+      "learning_rate": 1.5215029575834182e-05,
+      "loss": 1.7781,
+      "step": 2014
+    },
+    {
+      "epoch": 0.34580401578857045,
+      "grad_norm": 2.7967827467215365,
+      "learning_rate": 1.5210286091280364e-05,
+      "loss": 1.841,
+      "step": 2015
+    },
+    {
+      "epoch": 0.34597563068474346,
+      "grad_norm": 2.2228868569973383,
+      "learning_rate": 1.5205540996981865e-05,
+      "loss": 1.4912,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3461472455809164,
+      "grad_norm": 3.5814295288562366,
+      "learning_rate": 1.5200794294404709e-05,
+      "loss": 1.6561,
+      "step": 2017
+    },
+    {
+      "epoch": 0.34631886047708943,
+      "grad_norm": 3.406112532646996,
+      "learning_rate": 1.5196045985015414e-05,
+      "loss": 1.937,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3464904753732624,
+      "grad_norm": 3.1318675983162025,
+      "learning_rate": 1.519129607028099e-05,
+      "loss": 1.6298,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3466620902694354,
+      "grad_norm": 2.6200220724003436,
+      "learning_rate": 1.5186544551668955e-05,
+      "loss": 1.9008,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34683370516560835,
+      "grad_norm": 2.97887899292917,
+      "learning_rate": 1.5181791430647306e-05,
+      "loss": 1.71,
+      "step": 2021
+    },
+    {
+      "epoch": 0.34700532006178136,
+      "grad_norm": 2.6720768431569835,
+      "learning_rate": 1.5177036708684551e-05,
+      "loss": 1.9192,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3471769349579544,
+      "grad_norm": 2.675272761796172,
+      "learning_rate": 1.5172280387249685e-05,
+      "loss": 1.6388,
+      "step": 2023
+    },
+    {
+      "epoch": 0.34734854985412733,
+      "grad_norm": 2.7498697524694884,
+      "learning_rate": 1.5167522467812194e-05,
+      "loss": 1.8369,
+      "step": 2024
+    },
+    {
+      "epoch": 0.34752016475030034,
+      "grad_norm": 3.560804837565884,
+      "learning_rate": 1.5162762951842068e-05,
+      "loss": 1.784,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3476917796464733,
+      "grad_norm": 2.955387252319994,
+      "learning_rate": 1.5158001840809776e-05,
+      "loss": 1.7898,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3478633945426463,
+      "grad_norm": 2.099486300165839,
+      "learning_rate": 1.5153239136186297e-05,
+      "loss": 1.4766,
+      "step": 2027
+    },
+    {
+      "epoch": 0.34803500943881926,
+      "grad_norm": 4.120103882263827,
+      "learning_rate": 1.5148474839443084e-05,
+      "loss": 1.8699,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3482066243349923,
+      "grad_norm": 4.002060875482322,
+      "learning_rate": 1.5143708952052099e-05,
+      "loss": 1.7598,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3483782392311653,
+      "grad_norm": 2.69509799145148,
+      "learning_rate": 1.5138941475485779e-05,
+      "loss": 1.7916,
+      "step": 2030
+    },
+    {
+      "epoch": 0.34854985412733824,
+      "grad_norm": 2.9079292525794873,
+      "learning_rate": 1.5134172411217067e-05,
+      "loss": 1.6887,
+      "step": 2031
+    },
+    {
+      "epoch": 0.34872146902351125,
+      "grad_norm": 2.342411628101549,
+      "learning_rate": 1.5129401760719393e-05,
+      "loss": 1.6308,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3488930839196842,
+      "grad_norm": 2.5179014930477526,
+      "learning_rate": 1.5124629525466664e-05,
+      "loss": 1.5402,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3490646988158572,
+      "grad_norm": 2.5460701327780955,
+      "learning_rate": 1.511985570693329e-05,
+      "loss": 1.6214,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3492363137120302,
+      "grad_norm": 3.2936216283547908,
+      "learning_rate": 1.5115080306594172e-05,
+      "loss": 1.6879,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3494079286082032,
+      "grad_norm": 2.6621086998012973,
+      "learning_rate": 1.5110303325924686e-05,
+      "loss": 1.6966,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3495795435043762,
+      "grad_norm": 3.760328579284915,
+      "learning_rate": 1.5105524766400717e-05,
+      "loss": 1.5663,
+      "step": 2037
+    },
+    {
+      "epoch": 0.34975115840054916,
+      "grad_norm": 3.0909287373416454,
+      "learning_rate": 1.5100744629498611e-05,
+      "loss": 1.7214,
+      "step": 2038
+    },
+    {
+      "epoch": 0.34992277329672217,
+      "grad_norm": 3.2255819420554452,
+      "learning_rate": 1.5095962916695227e-05,
+      "loss": 1.7462,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3500943881928951,
+      "grad_norm": 3.859164936293867,
+      "learning_rate": 1.5091179629467894e-05,
+      "loss": 2.0157,
+      "step": 2040
+    },
+    {
+      "epoch": 0.35026600308906813,
+      "grad_norm": 3.7505339654604204,
+      "learning_rate": 1.5086394769294437e-05,
+      "loss": 1.709,
+      "step": 2041
+    },
+    {
+      "epoch": 0.35043761798524115,
+      "grad_norm": 2.323223409781269,
+      "learning_rate": 1.5081608337653157e-05,
+      "loss": 1.4799,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3506092328814141,
+      "grad_norm": 3.9169386800850434,
+      "learning_rate": 1.5076820336022855e-05,
+      "loss": 1.7923,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3507808477775871,
+      "grad_norm": 2.5083349690438252,
+      "learning_rate": 1.5072030765882804e-05,
+      "loss": 1.5316,
+      "step": 2044
+    },
+    {
+      "epoch": 0.35095246267376007,
+      "grad_norm": 3.0950812152066747,
+      "learning_rate": 1.5067239628712764e-05,
+      "loss": 1.6214,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3511240775699331,
+      "grad_norm": 2.83291084959603,
+      "learning_rate": 1.5062446925992987e-05,
+      "loss": 1.7143,
+      "step": 2046
+    },
+    {
+      "epoch": 0.35129569246610604,
+      "grad_norm": 3.865533320579865,
+      "learning_rate": 1.5057652659204198e-05,
+      "loss": 1.7135,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35146730736227905,
+      "grad_norm": 3.417673960643634,
+      "learning_rate": 1.5052856829827613e-05,
+      "loss": 1.9202,
+      "step": 2048
+    },
+    {
+      "epoch": 0.35163892225845206,
+      "grad_norm": 2.392249161136908,
+      "learning_rate": 1.5048059439344931e-05,
+      "loss": 1.5183,
+      "step": 2049
+    },
+    {
+      "epoch": 0.351810537154625,
+      "grad_norm": 2.4482902650331186,
+      "learning_rate": 1.5043260489238325e-05,
+      "loss": 1.4724,
+      "step": 2050
+    },
+    {
+      "epoch": 0.351982152050798,
+      "grad_norm": 3.002896171553363,
+      "learning_rate": 1.5038459980990459e-05,
+      "loss": 1.7874,
+      "step": 2051
+    },
+    {
+      "epoch": 0.352153766946971,
+      "grad_norm": 2.8087411566253118,
+      "learning_rate": 1.503365791608447e-05,
+      "loss": 1.6975,
+      "step": 2052
+    },
+    {
+      "epoch": 0.352325381843144,
+      "grad_norm": 2.4170815033256496,
+      "learning_rate": 1.5028854296003987e-05,
+      "loss": 1.8364,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35249699673931695,
+      "grad_norm": 3.264864338389202,
+      "learning_rate": 1.5024049122233112e-05,
+      "loss": 1.527,
+      "step": 2054
+    },
+    {
+      "epoch": 0.35266861163548996,
+      "grad_norm": 2.866232775421392,
+      "learning_rate": 1.501924239625642e-05,
+      "loss": 1.5965,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35284022653166297,
+      "grad_norm": 2.314526606451571,
+      "learning_rate": 1.5014434119558988e-05,
+      "loss": 1.2629,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3530118414278359,
+      "grad_norm": 3.429784914737218,
+      "learning_rate": 1.5009624293626343e-05,
+      "loss": 1.6308,
+      "step": 2057
+    },
+    {
+      "epoch": 0.35318345632400894,
+      "grad_norm": 2.971184465812407,
+      "learning_rate": 1.5004812919944515e-05,
+      "loss": 1.8593,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3533550712201819,
+      "grad_norm": 2.3982470302601238,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.462,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3535266861163549,
+      "grad_norm": 2.8484372719852504,
+      "learning_rate": 1.4995185535279775e-05,
+      "loss": 1.8415,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35369830101252786,
+      "grad_norm": 2.599762066061954,
+      "learning_rate": 1.4990369527271294e-05,
+      "loss": 1.625,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35386991590870087,
+      "grad_norm": 3.10493471529143,
+      "learning_rate": 1.4985551977462484e-05,
+      "loss": 1.84,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3540415308048739,
+      "grad_norm": 3.021332212801247,
+      "learning_rate": 1.4980732887341756e-05,
+      "loss": 1.5465,
+      "step": 2063
+    },
+    {
+      "epoch": 0.35421314570104684,
+      "grad_norm": 2.768989936483566,
+      "learning_rate": 1.4975912258397988e-05,
+      "loss": 1.7901,
+      "step": 2064
+    },
+    {
+      "epoch": 0.35438476059721985,
+      "grad_norm": 2.558841054645275,
+      "learning_rate": 1.4971090092120544e-05,
+      "loss": 1.7264,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3545563754933928,
+      "grad_norm": 2.8361183453990755,
+      "learning_rate": 1.496626638999925e-05,
+      "loss": 1.6502,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3547279903895658,
+      "grad_norm": 2.6257259033632967,
+      "learning_rate": 1.4961441153524421e-05,
+      "loss": 1.7741,
+      "step": 2067
+    },
+    {
+      "epoch": 0.35489960528573883,
+      "grad_norm": 3.1519146906299635,
+      "learning_rate": 1.4956614384186837e-05,
+      "loss": 1.9824,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3550712201819118,
+      "grad_norm": 2.701134143660292,
+      "learning_rate": 1.4951786083477746e-05,
+      "loss": 1.8847,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3552428350780848,
+      "grad_norm": 2.4973986732019595,
+      "learning_rate": 1.4946956252888887e-05,
+      "loss": 1.7222,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35541444997425775,
+      "grad_norm": 3.5328549448900732,
+      "learning_rate": 1.4942124893912453e-05,
+      "loss": 1.6673,
+      "step": 2071
+    },
+    {
+      "epoch": 0.35558606487043076,
+      "grad_norm": 2.9775574819790185,
+      "learning_rate": 1.4937292008041117e-05,
+      "loss": 1.7622,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3557576797666037,
+      "grad_norm": 2.452580712385438,
+      "learning_rate": 1.4932457596768032e-05,
+      "loss": 1.4865,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35592929466277673,
+      "grad_norm": 2.751209670642272,
+      "learning_rate": 1.4927621661586806e-05,
+      "loss": 1.7595,
+      "step": 2074
+    },
+    {
+      "epoch": 0.35610090955894974,
+      "grad_norm": 2.594070489724724,
+      "learning_rate": 1.4922784203991531e-05,
+      "loss": 1.5734,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3562725244551227,
+      "grad_norm": 2.919945339360576,
+      "learning_rate": 1.4917945225476762e-05,
+      "loss": 1.7262,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3564441393512957,
+      "grad_norm": 2.674845919320873,
+      "learning_rate": 1.4913104727537526e-05,
+      "loss": 1.6933,
+      "step": 2077
+    },
+    {
+      "epoch": 0.35661575424746866,
+      "grad_norm": 3.4430302297906032,
+      "learning_rate": 1.4908262711669324e-05,
+      "loss": 1.6939,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3567873691436417,
+      "grad_norm": 2.9284917044842507,
+      "learning_rate": 1.4903419179368113e-05,
+      "loss": 1.5715,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35695898403981463,
+      "grad_norm": 2.854764720449646,
+      "learning_rate": 1.4898574132130337e-05,
+      "loss": 1.7242,
+      "step": 2080
+    },
+    {
+      "epoch": 0.35713059893598764,
+      "grad_norm": 2.581716627236574,
+      "learning_rate": 1.489372757145289e-05,
+      "loss": 1.6838,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35730221383216065,
+      "grad_norm": 3.03205355443985,
+      "learning_rate": 1.4888879498833149e-05,
+      "loss": 1.6537,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3574738287283336,
+      "grad_norm": 2.609054169161796,
+      "learning_rate": 1.4884029915768945e-05,
+      "loss": 1.7836,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3576454436245066,
+      "grad_norm": 2.8147834592908714,
+      "learning_rate": 1.4879178823758584e-05,
+      "loss": 1.7895,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3578170585206796,
+      "grad_norm": 4.421250672672322,
+      "learning_rate": 1.4874326224300837e-05,
+      "loss": 1.5548,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3579886734168526,
+      "grad_norm": 3.5349509665651397,
+      "learning_rate": 1.4869472118894939e-05,
+      "loss": 1.7912,
+      "step": 2086
+    },
+    {
+      "epoch": 0.35816028831302554,
+      "grad_norm": 3.2974578297900976,
+      "learning_rate": 1.4864616509040591e-05,
+      "loss": 1.9432,
+      "step": 2087
+    },
+    {
+      "epoch": 0.35833190320919855,
+      "grad_norm": 3.027398437498267,
+      "learning_rate": 1.4859759396237956e-05,
+      "loss": 1.8214,
+      "step": 2088
+    },
+    {
+      "epoch": 0.35850351810537157,
+      "grad_norm": 2.219266843034949,
+      "learning_rate": 1.4854900781987667e-05,
+      "loss": 1.8622,
+      "step": 2089
+    },
+    {
+      "epoch": 0.3586751330015445,
+      "grad_norm": 3.0725642478698556,
+      "learning_rate": 1.4850040667790814e-05,
+      "loss": 1.7087,
+      "step": 2090
+    },
+    {
+      "epoch": 0.35884674789771753,
+      "grad_norm": 2.456343930505925,
+      "learning_rate": 1.4845179055148959e-05,
+      "loss": 1.8706,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3590183627938905,
+      "grad_norm": 3.208164530094859,
+      "learning_rate": 1.4840315945564126e-05,
+      "loss": 1.7426,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3591899776900635,
+      "grad_norm": 2.3661968785355163,
+      "learning_rate": 1.4835451340538785e-05,
+      "loss": 1.6436,
+      "step": 2093
+    },
+    {
+      "epoch": 0.3593615925862365,
+      "grad_norm": 2.7368973716615366,
+      "learning_rate": 1.483058524157589e-05,
+      "loss": 1.5317,
+      "step": 2094
+    },
+    {
+      "epoch": 0.35953320748240947,
+      "grad_norm": 2.504767054026463,
+      "learning_rate": 1.4825717650178846e-05,
+      "loss": 1.7232,
+      "step": 2095
+    },
+    {
+      "epoch": 0.3597048223785825,
+      "grad_norm": 3.4042583327702323,
+      "learning_rate": 1.4820848567851516e-05,
+      "loss": 1.6432,
+      "step": 2096
+    },
+    {
+      "epoch": 0.35987643727475543,
+      "grad_norm": 2.991618591792345,
+      "learning_rate": 1.4815977996098234e-05,
+      "loss": 1.77,
+      "step": 2097
+    },
+    {
+      "epoch": 0.36004805217092845,
+      "grad_norm": 2.325091528558994,
+      "learning_rate": 1.4811105936423787e-05,
+      "loss": 1.4872,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3602196670671014,
+      "grad_norm": 2.3382210155462726,
+      "learning_rate": 1.4806232390333419e-05,
+      "loss": 1.4459,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3603912819632744,
+      "grad_norm": 2.4281709203534114,
+      "learning_rate": 1.480135735933284e-05,
+      "loss": 1.5677,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3605628968594474,
+      "grad_norm": 3.1766231144340877,
+      "learning_rate": 1.4796480844928218e-05,
+      "loss": 1.6072,
+      "step": 2101
+    },
+    {
+      "epoch": 0.3607345117556204,
+      "grad_norm": 3.214952600143939,
+      "learning_rate": 1.4791602848626173e-05,
+      "loss": 1.6929,
+      "step": 2102
+    },
+    {
+      "epoch": 0.3609061266517934,
+      "grad_norm": 3.332856283116005,
+      "learning_rate": 1.4786723371933786e-05,
+      "loss": 1.7261,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36107774154796635,
+      "grad_norm": 2.2312020367976437,
+      "learning_rate": 1.4781842416358604e-05,
+      "loss": 1.5619,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36124935644413936,
+      "grad_norm": 3.345718261205169,
+      "learning_rate": 1.4776959983408618e-05,
+      "loss": 1.7613,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3614209713403123,
+      "grad_norm": 2.137050076720357,
+      "learning_rate": 1.4772076074592279e-05,
+      "loss": 1.3221,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3615925862364853,
+      "grad_norm": 3.8357399228736773,
+      "learning_rate": 1.47671906914185e-05,
+      "loss": 1.6096,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36176420113265834,
+      "grad_norm": 4.116236903039708,
+      "learning_rate": 1.4762303835396643e-05,
+      "loss": 1.6687,
+      "step": 2108
+    },
+    {
+      "epoch": 0.3619358160288313,
+      "grad_norm": 3.202549333640833,
+      "learning_rate": 1.4757415508036528e-05,
+      "loss": 1.7655,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3621074309250043,
+      "grad_norm": 2.88016909904521,
+      "learning_rate": 1.4752525710848426e-05,
+      "loss": 1.7907,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36227904582117726,
+      "grad_norm": 2.759735373177994,
+      "learning_rate": 1.4747634445343072e-05,
+      "loss": 1.6621,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36245066071735027,
+      "grad_norm": 2.5521293862134544,
+      "learning_rate": 1.4742741713031636e-05,
+      "loss": 1.7276,
+      "step": 2112
+    },
+    {
+      "epoch": 0.3626222756135232,
+      "grad_norm": 2.682126368759819,
+      "learning_rate": 1.4737847515425766e-05,
+      "loss": 1.6332,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36279389050969624,
+      "grad_norm": 2.3282003454522866,
+      "learning_rate": 1.473295185403754e-05,
+      "loss": 1.446,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36296550540586925,
+      "grad_norm": 2.5090572008598295,
+      "learning_rate": 1.4728054730379503e-05,
+      "loss": 1.7423,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3631371203020422,
+      "grad_norm": 2.5973100709418326,
+      "learning_rate": 1.4723156145964647e-05,
+      "loss": 1.6289,
+      "step": 2116
+    },
+    {
+      "epoch": 0.3633087351982152,
+      "grad_norm": 3.0858678293813724,
+      "learning_rate": 1.4718256102306407e-05,
+      "loss": 1.7861,
+      "step": 2117
+    },
+    {
+      "epoch": 0.36348035009438817,
+      "grad_norm": 2.1602360735875847,
+      "learning_rate": 1.4713354600918688e-05,
+      "loss": 1.503,
+      "step": 2118
+    },
+    {
+      "epoch": 0.3636519649905612,
+      "grad_norm": 2.906901057828225,
+      "learning_rate": 1.4708451643315827e-05,
+      "loss": 1.7194,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3638235798867342,
+      "grad_norm": 3.1450995314553243,
+      "learning_rate": 1.4703547231012616e-05,
+      "loss": 1.6388,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36399519478290715,
+      "grad_norm": 3.888297884721375,
+      "learning_rate": 1.4698641365524303e-05,
+      "loss": 1.9896,
+      "step": 2121
+    },
+    {
+      "epoch": 0.36416680967908016,
+      "grad_norm": 2.5314979232206514,
+      "learning_rate": 1.4693734048366582e-05,
+      "loss": 1.5323,
+      "step": 2122
+    },
+    {
+      "epoch": 0.3643384245752531,
+      "grad_norm": 2.625170574968521,
+      "learning_rate": 1.4688825281055587e-05,
+      "loss": 1.4178,
+      "step": 2123
+    },
+    {
+      "epoch": 0.36451003947142613,
+      "grad_norm": 3.8181101327414804,
+      "learning_rate": 1.4683915065107912e-05,
+      "loss": 1.806,
+      "step": 2124
+    },
+    {
+      "epoch": 0.3646816543675991,
+      "grad_norm": 2.4735566729714784,
+      "learning_rate": 1.4679003402040593e-05,
+      "loss": 1.7117,
+      "step": 2125
+    },
+    {
+      "epoch": 0.3648532692637721,
+      "grad_norm": 2.4898196807743935,
+      "learning_rate": 1.4674090293371113e-05,
+      "loss": 1.5495,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3650248841599451,
+      "grad_norm": 3.971874749674553,
+      "learning_rate": 1.4669175740617401e-05,
+      "loss": 1.7275,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36519649905611806,
+      "grad_norm": 2.2521467588803996,
+      "learning_rate": 1.4664259745297838e-05,
+      "loss": 1.7097,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3653681139522911,
+      "grad_norm": 3.6612101128507755,
+      "learning_rate": 1.4659342308931237e-05,
+      "loss": 1.8567,
+      "step": 2129
+    },
+    {
+      "epoch": 0.36553972884846403,
+      "grad_norm": 3.187715967776502,
+      "learning_rate": 1.4654423433036873e-05,
+      "loss": 1.5872,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36571134374463704,
+      "grad_norm": 2.1535557633300217,
+      "learning_rate": 1.4649503119134456e-05,
+      "loss": 1.5298,
+      "step": 2131
+    },
+    {
+      "epoch": 0.36588295864081,
+      "grad_norm": 2.917728221700515,
+      "learning_rate": 1.464458136874414e-05,
+      "loss": 1.9201,
+      "step": 2132
+    },
+    {
+      "epoch": 0.366054573536983,
+      "grad_norm": 3.233510869815508,
+      "learning_rate": 1.463965818338653e-05,
+      "loss": 1.6774,
+      "step": 2133
+    },
+    {
+      "epoch": 0.366226188433156,
+      "grad_norm": 2.6731811836923196,
+      "learning_rate": 1.4634733564582666e-05,
+      "loss": 1.8772,
+      "step": 2134
+    },
+    {
+      "epoch": 0.366397803329329,
+      "grad_norm": 2.625132452814094,
+      "learning_rate": 1.4629807513854033e-05,
+      "loss": 1.7352,
+      "step": 2135
+    },
+    {
+      "epoch": 0.366569418225502,
+      "grad_norm": 3.016963574875201,
+      "learning_rate": 1.4624880032722561e-05,
+      "loss": 1.6212,
+      "step": 2136
+    },
+    {
+      "epoch": 0.36674103312167494,
+      "grad_norm": 2.770623140822268,
+      "learning_rate": 1.4619951122710624e-05,
+      "loss": 1.5459,
+      "step": 2137
+    },
+    {
+      "epoch": 0.36691264801784795,
+      "grad_norm": 3.366373600057806,
+      "learning_rate": 1.461502078534103e-05,
+      "loss": 1.842,
+      "step": 2138
+    },
+    {
+      "epoch": 0.36708426291402096,
+      "grad_norm": 3.06896023646229,
+      "learning_rate": 1.461008902213703e-05,
+      "loss": 1.6927,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3672558778101939,
+      "grad_norm": 3.0015100240991637,
+      "learning_rate": 1.460515583462233e-05,
+      "loss": 1.721,
+      "step": 2140
+    },
+    {
+      "epoch": 0.36742749270636693,
+      "grad_norm": 3.0118975738551232,
+      "learning_rate": 1.4600221224321045e-05,
+      "loss": 1.7634,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3675991076025399,
+      "grad_norm": 3.371891528487684,
+      "learning_rate": 1.4595285192757763e-05,
+      "loss": 1.8535,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3677707224987129,
+      "grad_norm": 3.0411025316178204,
+      "learning_rate": 1.4590347741457489e-05,
+      "loss": 1.7882,
+      "step": 2143
+    },
+    {
+      "epoch": 0.36794233739488585,
+      "grad_norm": 3.0786992984465322,
+      "learning_rate": 1.4585408871945675e-05,
+      "loss": 1.6479,
+      "step": 2144
+    },
+    {
+      "epoch": 0.36811395229105887,
+      "grad_norm": 3.033010693532836,
+      "learning_rate": 1.4580468585748212e-05,
+      "loss": 1.7905,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3682855671872319,
+      "grad_norm": 2.8625627407557386,
+      "learning_rate": 1.457552688439143e-05,
+      "loss": 1.8307,
+      "step": 2146
+    },
+    {
+      "epoch": 0.36845718208340483,
+      "grad_norm": 2.324754800708334,
+      "learning_rate": 1.4570583769402085e-05,
+      "loss": 1.6194,
+      "step": 2147
+    },
+    {
+      "epoch": 0.36862879697957784,
+      "grad_norm": 2.5836327999141115,
+      "learning_rate": 1.4565639242307381e-05,
+      "loss": 1.7015,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3688004118757508,
+      "grad_norm": 2.316074404157186,
+      "learning_rate": 1.4560693304634956e-05,
+      "loss": 1.677,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3689720267719238,
+      "grad_norm": 2.563453653505684,
+      "learning_rate": 1.4555745957912886e-05,
+      "loss": 1.7119,
+      "step": 2150
+    },
+    {
+      "epoch": 0.36914364166809677,
+      "grad_norm": 2.865707726633867,
+      "learning_rate": 1.4550797203669673e-05,
+      "loss": 1.7043,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3693152565642698,
+      "grad_norm": 5.011378547660993,
+      "learning_rate": 1.4545847043434266e-05,
+      "loss": 1.5669,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3694868714604428,
+      "grad_norm": 3.202085857937417,
+      "learning_rate": 1.4540895478736037e-05,
+      "loss": 1.5395,
+      "step": 2153
+    },
+    {
+      "epoch": 0.36965848635661575,
+      "grad_norm": 2.6501341424564138,
+      "learning_rate": 1.45359425111048e-05,
+      "loss": 1.6759,
+      "step": 2154
+    },
+    {
+      "epoch": 0.36983010125278876,
+      "grad_norm": 2.621896430240285,
+      "learning_rate": 1.4530988142070802e-05,
+      "loss": 1.7386,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3700017161489617,
+      "grad_norm": 2.395999084791078,
+      "learning_rate": 1.4526032373164717e-05,
+      "loss": 1.5637,
+      "step": 2156
+    },
+    {
+      "epoch": 0.3701733310451347,
+      "grad_norm": 3.6772351195774786,
+      "learning_rate": 1.4521075205917656e-05,
+      "loss": 1.8141,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3703449459413077,
+      "grad_norm": 9.020394448334791,
+      "learning_rate": 1.4516116641861164e-05,
+      "loss": 1.6614,
+      "step": 2158
+    },
+    {
+      "epoch": 0.3705165608374807,
+      "grad_norm": 2.8309602729085053,
+      "learning_rate": 1.4511156682527215e-05,
+      "loss": 1.6501,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3706881757336537,
+      "grad_norm": 3.5653747504992825,
+      "learning_rate": 1.4506195329448205e-05,
+      "loss": 1.7632,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37085979062982666,
+      "grad_norm": 3.3034069919113422,
+      "learning_rate": 1.4501232584156984e-05,
+      "loss": 1.6739,
+      "step": 2161
+    },
+    {
+      "epoch": 0.37103140552599967,
+      "grad_norm": 4.178189283199575,
+      "learning_rate": 1.4496268448186806e-05,
+      "loss": 1.7542,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3712030204221726,
+      "grad_norm": 4.365822456310529,
+      "learning_rate": 1.449130292307137e-05,
+      "loss": 1.8823,
+      "step": 2163
+    },
+    {
+      "epoch": 0.37137463531834564,
+      "grad_norm": 2.899211322850207,
+      "learning_rate": 1.4486336010344804e-05,
+      "loss": 1.6283,
+      "step": 2164
+    },
+    {
+      "epoch": 0.37154625021451865,
+      "grad_norm": 3.9142505904887166,
+      "learning_rate": 1.4481367711541653e-05,
+      "loss": 1.5994,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3717178651106916,
+      "grad_norm": 2.8814960731444406,
+      "learning_rate": 1.4476398028196909e-05,
+      "loss": 1.7413,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3718894800068646,
+      "grad_norm": 2.487125036029448,
+      "learning_rate": 1.4471426961845976e-05,
+      "loss": 1.7801,
+      "step": 2167
+    },
+    {
+      "epoch": 0.37206109490303757,
+      "grad_norm": 2.7121970885323274,
+      "learning_rate": 1.4466454514024685e-05,
+      "loss": 1.6655,
+      "step": 2168
+    },
+    {
+      "epoch": 0.3722327097992106,
+      "grad_norm": 3.1423092690643526,
+      "learning_rate": 1.446148068626931e-05,
+      "loss": 1.7162,
+      "step": 2169
+    },
+    {
+      "epoch": 0.37240432469538354,
+      "grad_norm": 2.862760014745284,
+      "learning_rate": 1.4456505480116533e-05,
+      "loss": 1.8156,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37257593959155655,
+      "grad_norm": 2.6915273388266487,
+      "learning_rate": 1.4451528897103476e-05,
+      "loss": 1.6721,
+      "step": 2171
+    },
+    {
+      "epoch": 0.37274755448772956,
+      "grad_norm": 2.5239709085159605,
+      "learning_rate": 1.4446550938767669e-05,
+      "loss": 1.6769,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3729191693839025,
+      "grad_norm": 2.67868547969852,
+      "learning_rate": 1.4441571606647089e-05,
+      "loss": 1.8929,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3730907842800755,
+      "grad_norm": 3.404084445351775,
+      "learning_rate": 1.4436590902280122e-05,
+      "loss": 1.7555,
+      "step": 2174
+    },
+    {
+      "epoch": 0.3732623991762485,
+      "grad_norm": 2.736853970976809,
+      "learning_rate": 1.4431608827205579e-05,
+      "loss": 1.7508,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3734340140724215,
+      "grad_norm": 3.0522568393109863,
+      "learning_rate": 1.4426625382962706e-05,
+      "loss": 1.838,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37360562896859445,
+      "grad_norm": 3.2251956796296506,
+      "learning_rate": 1.4421640571091153e-05,
+      "loss": 1.7563,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37377724386476746,
+      "grad_norm": 2.326256885234364,
+      "learning_rate": 1.4416654393131013e-05,
+      "loss": 1.6253,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3739488587609405,
+      "grad_norm": 2.929950220726013,
+      "learning_rate": 1.4411666850622788e-05,
+      "loss": 1.8398,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37412047365711343,
+      "grad_norm": 2.882205056595399,
+      "learning_rate": 1.4406677945107398e-05,
+      "loss": 1.7602,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37429208855328644,
+      "grad_norm": 2.154015153842205,
+      "learning_rate": 1.4401687678126205e-05,
+      "loss": 1.5701,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3744637034494594,
+      "grad_norm": 2.370836467763457,
+      "learning_rate": 1.4396696051220965e-05,
+      "loss": 1.6229,
+      "step": 2182
+    },
+    {
+      "epoch": 0.3746353183456324,
+      "grad_norm": 2.760721971689484,
+      "learning_rate": 1.4391703065933873e-05,
+      "loss": 1.7536,
+      "step": 2183
+    },
+    {
+      "epoch": 0.37480693324180536,
+      "grad_norm": 2.8574886068966108,
+      "learning_rate": 1.4386708723807532e-05,
+      "loss": 1.6223,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3749785481379784,
+      "grad_norm": 2.371703482108517,
+      "learning_rate": 1.438171302638498e-05,
+      "loss": 1.7262,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3751501630341514,
+      "grad_norm": 2.4793168684302165,
+      "learning_rate": 1.4376715975209654e-05,
+      "loss": 1.6545,
+      "step": 2186
+    },
+    {
+      "epoch": 0.37532177793032434,
+      "grad_norm": 2.532404862793482,
+      "learning_rate": 1.437171757182542e-05,
+      "loss": 1.5066,
+      "step": 2187
+    },
+    {
+      "epoch": 0.37549339282649735,
+      "grad_norm": 3.7874693942879265,
+      "learning_rate": 1.4366717817776567e-05,
+      "loss": 1.6264,
+      "step": 2188
+    },
+    {
+      "epoch": 0.3756650077226703,
+      "grad_norm": 3.116036982270046,
+      "learning_rate": 1.4361716714607785e-05,
+      "loss": 1.5684,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3758366226188433,
+      "grad_norm": 2.5175876506273043,
+      "learning_rate": 1.4356714263864198e-05,
+      "loss": 1.6901,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37600823751501633,
+      "grad_norm": 2.7037305648012486,
+      "learning_rate": 1.4351710467091337e-05,
+      "loss": 1.8673,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3761798524111893,
+      "grad_norm": 2.511107908132675,
+      "learning_rate": 1.4346705325835148e-05,
+      "loss": 1.5539,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3763514673073623,
+      "grad_norm": 2.489911802380476,
+      "learning_rate": 1.4341698841641997e-05,
+      "loss": 1.8653,
+      "step": 2193
+    },
+    {
+      "epoch": 0.37652308220353525,
+      "grad_norm": 2.4525580717293223,
+      "learning_rate": 1.4336691016058665e-05,
+      "loss": 1.4228,
+      "step": 2194
+    },
+    {
+      "epoch": 0.37669469709970826,
+      "grad_norm": 2.367772194895001,
+      "learning_rate": 1.4331681850632344e-05,
+      "loss": 1.6974,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3768663119958812,
+      "grad_norm": 3.47541828158448,
+      "learning_rate": 1.4326671346910641e-05,
+      "loss": 1.9933,
+      "step": 2196
+    },
+    {
+      "epoch": 0.37703792689205423,
+      "grad_norm": 2.7142388534405297,
+      "learning_rate": 1.4321659506441577e-05,
+      "loss": 1.6537,
+      "step": 2197
+    },
+    {
+      "epoch": 0.37720954178822724,
+      "grad_norm": 2.652141709157279,
+      "learning_rate": 1.4316646330773585e-05,
+      "loss": 1.7776,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3773811566844002,
+      "grad_norm": 3.019675057855687,
+      "learning_rate": 1.431163182145551e-05,
+      "loss": 1.7079,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3775527715805732,
+      "grad_norm": 2.8810525572386636,
+      "learning_rate": 1.4306615980036618e-05,
+      "loss": 1.5222,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37772438647674617,
+      "grad_norm": 2.4813778721533546,
+      "learning_rate": 1.430159880806657e-05,
+      "loss": 1.4897,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3778960013729192,
+      "grad_norm": 2.918744639476238,
+      "learning_rate": 1.429658030709545e-05,
+      "loss": 1.8064,
+      "step": 2202
+    },
+    {
+      "epoch": 0.37806761626909213,
+      "grad_norm": 2.486611141805446,
+      "learning_rate": 1.4291560478673756e-05,
+      "loss": 1.6809,
+      "step": 2203
+    },
+    {
+      "epoch": 0.37823923116526514,
+      "grad_norm": 2.525480350108589,
+      "learning_rate": 1.4286539324352378e-05,
+      "loss": 1.4924,
+      "step": 2204
+    },
+    {
+      "epoch": 0.37841084606143816,
+      "grad_norm": 2.7024109695281577,
+      "learning_rate": 1.4281516845682637e-05,
+      "loss": 1.7283,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3785824609576111,
+      "grad_norm": 3.4908404464474008,
+      "learning_rate": 1.427649304421625e-05,
+      "loss": 1.7113,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3787540758537841,
+      "grad_norm": 2.9712620329475303,
+      "learning_rate": 1.4271467921505348e-05,
+      "loss": 1.8099,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3789256907499571,
+      "grad_norm": 2.777802338360159,
+      "learning_rate": 1.4266441479102464e-05,
+      "loss": 1.8899,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3790973056461301,
+      "grad_norm": 2.5937609206430783,
+      "learning_rate": 1.4261413718560549e-05,
+      "loss": 1.6308,
+      "step": 2209
+    },
+    {
+      "epoch": 0.37926892054230305,
+      "grad_norm": 2.90205576377282,
+      "learning_rate": 1.4256384641432952e-05,
+      "loss": 1.7082,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37944053543847606,
+      "grad_norm": 2.811984879702953,
+      "learning_rate": 1.4251354249273432e-05,
+      "loss": 1.6481,
+      "step": 2211
+    },
+    {
+      "epoch": 0.37961215033464907,
+      "grad_norm": 2.5821967392923155,
+      "learning_rate": 1.424632254363616e-05,
+      "loss": 1.9296,
+      "step": 2212
+    },
+    {
+      "epoch": 0.379783765230822,
+      "grad_norm": 2.434964606347429,
+      "learning_rate": 1.4241289526075697e-05,
+      "loss": 1.7321,
+      "step": 2213
+    },
+    {
+      "epoch": 0.37995538012699503,
+      "grad_norm": 2.288895494712603,
+      "learning_rate": 1.4236255198147029e-05,
+      "loss": 1.5558,
+      "step": 2214
+    },
+    {
+      "epoch": 0.380126995023168,
+      "grad_norm": 2.8732707429446687,
+      "learning_rate": 1.4231219561405533e-05,
+      "loss": 1.9385,
+      "step": 2215
+    },
+    {
+      "epoch": 0.380298609919341,
+      "grad_norm": 3.56963046537789,
+      "learning_rate": 1.4226182617406996e-05,
+      "loss": 1.9864,
+      "step": 2216
+    },
+    {
+      "epoch": 0.380470224815514,
+      "grad_norm": 3.302977319531517,
+      "learning_rate": 1.422114436770761e-05,
+      "loss": 1.5754,
+      "step": 2217
+    },
+    {
+      "epoch": 0.38064183971168697,
+      "grad_norm": 2.5965918821644167,
+      "learning_rate": 1.4216104813863962e-05,
+      "loss": 1.7756,
+      "step": 2218
+    },
+    {
+      "epoch": 0.38081345460786,
+      "grad_norm": 3.002135474152398,
+      "learning_rate": 1.4211063957433056e-05,
+      "loss": 1.8891,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38098506950403294,
+      "grad_norm": 2.461394979703817,
+      "learning_rate": 1.4206021799972283e-05,
+      "loss": 1.7562,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38115668440020595,
+      "grad_norm": 2.320356021719385,
+      "learning_rate": 1.4200978343039448e-05,
+      "loss": 1.6289,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3813282992963789,
+      "grad_norm": 2.6768833977800015,
+      "learning_rate": 1.4195933588192751e-05,
+      "loss": 1.642,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3814999141925519,
+      "grad_norm": 2.780108773902873,
+      "learning_rate": 1.4190887536990797e-05,
+      "loss": 1.7184,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3816715290887249,
+      "grad_norm": 2.4604900837998565,
+      "learning_rate": 1.418584019099259e-05,
+      "loss": 1.5972,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3818431439848979,
+      "grad_norm": 3.0313746678478672,
+      "learning_rate": 1.418079155175753e-05,
+      "loss": 1.6771,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3820147588810709,
+      "grad_norm": 3.11334474632266,
+      "learning_rate": 1.4175741620845422e-05,
+      "loss": 1.7264,
+      "step": 2226
+    },
+    {
+      "epoch": 0.38218637377724385,
+      "grad_norm": 3.3460943489633057,
+      "learning_rate": 1.4170690399816469e-05,
+      "loss": 1.6929,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38235798867341686,
+      "grad_norm": 2.6861366414202537,
+      "learning_rate": 1.4165637890231267e-05,
+      "loss": 1.7905,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3825296035695898,
+      "grad_norm": 3.492722916148379,
+      "learning_rate": 1.4160584093650825e-05,
+      "loss": 1.743,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3827012184657628,
+      "grad_norm": 4.023257395500161,
+      "learning_rate": 1.4155529011636528e-05,
+      "loss": 2.0173,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38287283336193584,
+      "grad_norm": 2.0917941336297705,
+      "learning_rate": 1.4150472645750175e-05,
+      "loss": 1.3025,
+      "step": 2231
+    },
+    {
+      "epoch": 0.3830444482581088,
+      "grad_norm": 3.006190689162569,
+      "learning_rate": 1.4145414997553956e-05,
+      "loss": 1.8324,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3832160631542818,
+      "grad_norm": 3.122814496435496,
+      "learning_rate": 1.4140356068610459e-05,
+      "loss": 1.7966,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38338767805045476,
+      "grad_norm": 3.2009987043855195,
+      "learning_rate": 1.4135295860482663e-05,
+      "loss": 1.7276,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3835592929466278,
+      "grad_norm": 2.554782866016197,
+      "learning_rate": 1.4130234374733944e-05,
+      "loss": 1.5416,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38373090784280073,
+      "grad_norm": 2.5950255478358644,
+      "learning_rate": 1.4125171612928083e-05,
+      "loss": 1.5395,
+      "step": 2236
+    },
+    {
+      "epoch": 0.38390252273897374,
+      "grad_norm": 2.8831155695688375,
+      "learning_rate": 1.4120107576629235e-05,
+      "loss": 1.7688,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38407413763514675,
+      "grad_norm": 2.5397518896498155,
+      "learning_rate": 1.4115042267401968e-05,
+      "loss": 1.6666,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3842457525313197,
+      "grad_norm": 3.6718608295390442,
+      "learning_rate": 1.4109975686811231e-05,
+      "loss": 1.5902,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3844173674274927,
+      "grad_norm": 2.363713596141586,
+      "learning_rate": 1.410490783642237e-05,
+      "loss": 1.5944,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3845889823236657,
+      "grad_norm": 3.165948877022219,
+      "learning_rate": 1.4099838717801126e-05,
+      "loss": 1.6349,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3847605972198387,
+      "grad_norm": 2.640233580352414,
+      "learning_rate": 1.4094768332513629e-05,
+      "loss": 1.704,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3849322121160117,
+      "grad_norm": 2.923584546226495,
+      "learning_rate": 1.4089696682126399e-05,
+      "loss": 1.7403,
+      "step": 2243
+    },
+    {
+      "epoch": 0.38510382701218465,
+      "grad_norm": 2.443877454341655,
+      "learning_rate": 1.4084623768206349e-05,
+      "loss": 1.6041,
+      "step": 2244
+    },
+    {
+      "epoch": 0.38527544190835766,
+      "grad_norm": 3.0326227011507334,
+      "learning_rate": 1.4079549592320782e-05,
+      "loss": 1.5833,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3854470568045306,
+      "grad_norm": 3.1657560521997956,
+      "learning_rate": 1.4074474156037393e-05,
+      "loss": 1.537,
+      "step": 2246
+    },
+    {
+      "epoch": 0.38561867170070363,
+      "grad_norm": 2.5425969518956966,
+      "learning_rate": 1.4069397460924259e-05,
+      "loss": 1.7082,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3857902865968766,
+      "grad_norm": 3.120102526551189,
+      "learning_rate": 1.4064319508549858e-05,
+      "loss": 1.8363,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3859619014930496,
+      "grad_norm": 2.6598909836762177,
+      "learning_rate": 1.4059240300483041e-05,
+      "loss": 1.8236,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3861335163892226,
+      "grad_norm": 2.6328046782494843,
+      "learning_rate": 1.4054159838293063e-05,
+      "loss": 1.6207,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38630513128539556,
+      "grad_norm": 2.7213315919935273,
+      "learning_rate": 1.4049078123549556e-05,
+      "loss": 1.6323,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3864767461815686,
+      "grad_norm": 2.9374698143724487,
+      "learning_rate": 1.4043995157822539e-05,
+      "loss": 1.7496,
+      "step": 2252
+    },
+    {
+      "epoch": 0.38664836107774153,
+      "grad_norm": 2.789279696962537,
+      "learning_rate": 1.4038910942682424e-05,
+      "loss": 1.7075,
+      "step": 2253
+    },
+    {
+      "epoch": 0.38681997597391454,
+      "grad_norm": 2.544224307618862,
+      "learning_rate": 1.4033825479700008e-05,
+      "loss": 1.7613,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3869915908700875,
+      "grad_norm": 2.847108668594769,
+      "learning_rate": 1.4028738770446463e-05,
+      "loss": 1.4704,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3871632057662605,
+      "grad_norm": 2.6030943868529692,
+      "learning_rate": 1.4023650816493362e-05,
+      "loss": 1.6116,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3873348206624335,
+      "grad_norm": 2.34249074933021,
+      "learning_rate": 1.401856161941265e-05,
+      "loss": 1.3736,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3875064355586065,
+      "grad_norm": 2.2890401834443255,
+      "learning_rate": 1.4013471180776661e-05,
+      "loss": 1.5081,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3876780504547795,
+      "grad_norm": 3.156247168187958,
+      "learning_rate": 1.4008379502158111e-05,
+      "loss": 1.7655,
+      "step": 2259
+    },
+    {
+      "epoch": 0.38784966535095244,
+      "grad_norm": 3.1658938242976715,
+      "learning_rate": 1.4003286585130106e-05,
+      "loss": 1.6149,
+      "step": 2260
+    },
+    {
+      "epoch": 0.38802128024712546,
+      "grad_norm": 2.9296203967992644,
+      "learning_rate": 1.3998192431266123e-05,
+      "loss": 1.6677,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3881928951432984,
+      "grad_norm": 4.16357736507998,
+      "learning_rate": 1.399309704214003e-05,
+      "loss": 1.572,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3883645100394714,
+      "grad_norm": 2.515372102550573,
+      "learning_rate": 1.3988000419326073e-05,
+      "loss": 1.7159,
+      "step": 2263
+    },
+    {
+      "epoch": 0.38853612493564443,
+      "grad_norm": 3.0463829188010574,
+      "learning_rate": 1.3982902564398876e-05,
+      "loss": 1.7418,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3887077398318174,
+      "grad_norm": 3.460164535727435,
+      "learning_rate": 1.3977803478933455e-05,
+      "loss": 1.6016,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3888793547279904,
+      "grad_norm": 2.6182099348643346,
+      "learning_rate": 1.3972703164505195e-05,
+      "loss": 1.5533,
+      "step": 2266
+    },
+    {
+      "epoch": 0.38905096962416336,
+      "grad_norm": 2.76380799104909,
+      "learning_rate": 1.3967601622689864e-05,
+      "loss": 1.6983,
+      "step": 2267
+    },
+    {
+      "epoch": 0.38922258452033637,
+      "grad_norm": 2.8298724202907732,
+      "learning_rate": 1.3962498855063606e-05,
+      "loss": 1.7373,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3893941994165094,
+      "grad_norm": 2.7153669429483562,
+      "learning_rate": 1.3957394863202955e-05,
+      "loss": 1.6579,
+      "step": 2269
+    },
+    {
+      "epoch": 0.38956581431268233,
+      "grad_norm": 2.390543625370323,
+      "learning_rate": 1.395228964868481e-05,
+      "loss": 1.5644,
+      "step": 2270
+    },
+    {
+      "epoch": 0.38973742920885535,
+      "grad_norm": 3.0979753570571087,
+      "learning_rate": 1.3947183213086455e-05,
+      "loss": 1.5409,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3899090441050283,
+      "grad_norm": 2.4886974863525704,
+      "learning_rate": 1.394207555798555e-05,
+      "loss": 1.528,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3900806590012013,
+      "grad_norm": 3.028471839823337,
+      "learning_rate": 1.393696668496013e-05,
+      "loss": 1.7404,
+      "step": 2273
+    },
+    {
+      "epoch": 0.39025227389737427,
+      "grad_norm": 2.6314129268290167,
+      "learning_rate": 1.3931856595588609e-05,
+      "loss": 1.5899,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3904238887935473,
+      "grad_norm": 2.7668023188081143,
+      "learning_rate": 1.3926745291449773e-05,
+      "loss": 1.8042,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3905955036897203,
+      "grad_norm": 2.7059509113929123,
+      "learning_rate": 1.3921632774122787e-05,
+      "loss": 1.7132,
+      "step": 2276
+    },
+    {
+      "epoch": 0.39076711858589325,
+      "grad_norm": 3.3045309371531886,
+      "learning_rate": 1.3916519045187192e-05,
+      "loss": 1.9183,
+      "step": 2277
+    },
+    {
+      "epoch": 0.39093873348206626,
+      "grad_norm": 2.5797282027846453,
+      "learning_rate": 1.3911404106222896e-05,
+      "loss": 1.7507,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3911103483782392,
+      "grad_norm": 2.7191289241677663,
+      "learning_rate": 1.3906287958810192e-05,
+      "loss": 1.6681,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3912819632744122,
+      "grad_norm": 2.963087524542881,
+      "learning_rate": 1.3901170604529735e-05,
+      "loss": 1.907,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3914535781705852,
+      "grad_norm": 2.6569891391232145,
+      "learning_rate": 1.3896052044962558e-05,
+      "loss": 1.6908,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3916251930667582,
+      "grad_norm": 2.879023993696917,
+      "learning_rate": 1.3890932281690068e-05,
+      "loss": 1.5593,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3917968079629312,
+      "grad_norm": 3.2212504368048727,
+      "learning_rate": 1.3885811316294042e-05,
+      "loss": 1.9828,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39196842285910416,
+      "grad_norm": 3.254504131122837,
+      "learning_rate": 1.3880689150356633e-05,
+      "loss": 1.5324,
+      "step": 2284
+    },
+    {
+      "epoch": 0.39214003775527717,
+      "grad_norm": 2.849534157439421,
+      "learning_rate": 1.3875565785460348e-05,
+      "loss": 1.6472,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3923116526514501,
+      "grad_norm": 3.093295169429665,
+      "learning_rate": 1.387044122318809e-05,
+      "loss": 1.6151,
+      "step": 2286
+    },
+    {
+      "epoch": 0.39248326754762314,
+      "grad_norm": 2.9988856293266926,
+      "learning_rate": 1.3865315465123112e-05,
+      "loss": 1.5012,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3926548824437961,
+      "grad_norm": 3.3798210062064844,
+      "learning_rate": 1.3860188512849044e-05,
+      "loss": 1.8044,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3928264973399691,
+      "grad_norm": 3.184967592285296,
+      "learning_rate": 1.3855060367949889e-05,
+      "loss": 1.7279,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3929981122361421,
+      "grad_norm": 4.083903163647933,
+      "learning_rate": 1.384993103201001e-05,
+      "loss": 1.7003,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39316972713231507,
+      "grad_norm": 3.239030991580772,
+      "learning_rate": 1.3844800506614138e-05,
+      "loss": 1.8288,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3933413420284881,
+      "grad_norm": 2.5920789014683554,
+      "learning_rate": 1.383966879334738e-05,
+      "loss": 1.5926,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39351295692466104,
+      "grad_norm": 4.2451886486307,
+      "learning_rate": 1.3834535893795204e-05,
+      "loss": 1.9014,
+      "step": 2293
+    },
+    {
+      "epoch": 0.39368457182083405,
+      "grad_norm": 2.7500562719662014,
+      "learning_rate": 1.3829401809543448e-05,
+      "loss": 1.7669,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39385618671700706,
+      "grad_norm": 3.538445334694933,
+      "learning_rate": 1.3824266542178308e-05,
+      "loss": 1.5683,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39402780161318,
+      "grad_norm": 2.352844912970664,
+      "learning_rate": 1.3819130093286359e-05,
+      "loss": 1.3802,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39419941650935303,
+      "grad_norm": 2.739344128865409,
+      "learning_rate": 1.3813992464454526e-05,
+      "loss": 1.5835,
+      "step": 2297
+    },
+    {
+      "epoch": 0.394371031405526,
+      "grad_norm": 2.849176805763815,
+      "learning_rate": 1.3808853657270112e-05,
+      "loss": 1.725,
+      "step": 2298
+    },
+    {
+      "epoch": 0.394542646301699,
+      "grad_norm": 2.888560274894531,
+      "learning_rate": 1.3803713673320773e-05,
+      "loss": 1.6273,
+      "step": 2299
+    },
+    {
+      "epoch": 0.39471426119787195,
+      "grad_norm": 2.9455165438999744,
+      "learning_rate": 1.3798572514194538e-05,
+      "loss": 1.851,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39488587609404496,
+      "grad_norm": 3.119256667747611,
+      "learning_rate": 1.3793430181479788e-05,
+      "loss": 1.6184,
+      "step": 2301
+    },
+    {
+      "epoch": 0.395057490990218,
+      "grad_norm": 3.3863437314608413,
+      "learning_rate": 1.378828667676528e-05,
+      "loss": 1.6019,
+      "step": 2302
+    },
+    {
+      "epoch": 0.39522910588639093,
+      "grad_norm": 3.4413384787839294,
+      "learning_rate": 1.3783142001640122e-05,
+      "loss": 1.8528,
+      "step": 2303
+    },
+    {
+      "epoch": 0.39540072078256394,
+      "grad_norm": 2.6511737602765915,
+      "learning_rate": 1.3777996157693787e-05,
+      "loss": 1.7517,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3955723356787369,
+      "grad_norm": 2.5101174885312374,
+      "learning_rate": 1.3772849146516114e-05,
+      "loss": 1.7305,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3957439505749099,
+      "grad_norm": 3.5403962745653415,
+      "learning_rate": 1.37677009696973e-05,
+      "loss": 1.7702,
+      "step": 2306
+    },
+    {
+      "epoch": 0.39591556547108286,
+      "grad_norm": 3.083670404857814,
+      "learning_rate": 1.3762551628827892e-05,
+      "loss": 1.7577,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3960871803672559,
+      "grad_norm": 2.7012615370850206,
+      "learning_rate": 1.3757401125498815e-05,
+      "loss": 1.6658,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3962587952634289,
+      "grad_norm": 5.123512818791215,
+      "learning_rate": 1.3752249461301337e-05,
+      "loss": 1.8537,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39643041015960184,
+      "grad_norm": 2.7960007122865425,
+      "learning_rate": 1.3747096637827092e-05,
+      "loss": 1.5898,
+      "step": 2310
+    },
+    {
+      "epoch": 0.39660202505577485,
+      "grad_norm": 3.1495010762429465,
+      "learning_rate": 1.3741942656668075e-05,
+      "loss": 1.8736,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3967736399519478,
+      "grad_norm": 3.1866458287117445,
+      "learning_rate": 1.3736787519416632e-05,
+      "loss": 1.6445,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3969452548481208,
+      "grad_norm": 3.141208465191903,
+      "learning_rate": 1.3731631227665472e-05,
+      "loss": 1.8147,
+      "step": 2313
+    },
+    {
+      "epoch": 0.39711686974429383,
+      "grad_norm": 3.4641279223813055,
+      "learning_rate": 1.372647378300765e-05,
+      "loss": 1.7662,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3972884846404668,
+      "grad_norm": 2.769606806978043,
+      "learning_rate": 1.3721315187036594e-05,
+      "loss": 1.5127,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3974600995366398,
+      "grad_norm": 2.7590882144200206,
+      "learning_rate": 1.3716155441346078e-05,
+      "loss": 1.732,
+      "step": 2316
+    },
+    {
+      "epoch": 0.39763171443281276,
+      "grad_norm": 6.126264387812704,
+      "learning_rate": 1.3710994547530225e-05,
+      "loss": 1.7298,
+      "step": 2317
+    },
+    {
+      "epoch": 0.39780332932898577,
+      "grad_norm": 3.6753904077094206,
+      "learning_rate": 1.3705832507183528e-05,
+      "loss": 1.7542,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3979749442251587,
+      "grad_norm": 3.9291222513709148,
+      "learning_rate": 1.3700669321900819e-05,
+      "loss": 1.7058,
+      "step": 2319
+    },
+    {
+      "epoch": 0.39814655912133173,
+      "grad_norm": 3.2643754407351784,
+      "learning_rate": 1.3695504993277296e-05,
+      "loss": 1.6015,
+      "step": 2320
+    },
+    {
+      "epoch": 0.39831817401750474,
+      "grad_norm": 2.6949351056020636,
+      "learning_rate": 1.36903395229085e-05,
+      "loss": 1.6077,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3984897889136777,
+      "grad_norm": 3.225576783310746,
+      "learning_rate": 1.3685172912390332e-05,
+      "loss": 1.7984,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3986614038098507,
+      "grad_norm": 2.567416719134394,
+      "learning_rate": 1.3680005163319045e-05,
+      "loss": 1.6596,
+      "step": 2323
+    },
+    {
+      "epoch": 0.39883301870602367,
+      "grad_norm": 2.7342326308550757,
+      "learning_rate": 1.367483627729124e-05,
+      "loss": 1.7329,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3990046336021967,
+      "grad_norm": 3.606800295580907,
+      "learning_rate": 1.366966625590387e-05,
+      "loss": 1.6412,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39917624849836963,
+      "grad_norm": 2.5606493236508103,
+      "learning_rate": 1.3664495100754238e-05,
+      "loss": 1.4087,
+      "step": 2326
+    },
+    {
+      "epoch": 0.39934786339454265,
+      "grad_norm": 3.0716962891561366,
+      "learning_rate": 1.3659322813440005e-05,
+      "loss": 1.5322,
+      "step": 2327
+    },
+    {
+      "epoch": 0.39951947829071566,
+      "grad_norm": 3.42781339334228,
+      "learning_rate": 1.3654149395559171e-05,
+      "loss": 1.9995,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3996910931868886,
+      "grad_norm": 2.943030860004717,
+      "learning_rate": 1.3648974848710088e-05,
+      "loss": 1.4968,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3998627080830616,
+      "grad_norm": 2.853168126549502,
+      "learning_rate": 1.3643799174491467e-05,
+      "loss": 1.6148,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4000343229792346,
+      "grad_norm": 3.402481678784887,
+      "learning_rate": 1.363862237450235e-05,
+      "loss": 1.6967,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4002059378754076,
+      "grad_norm": 5.980167444239729,
+      "learning_rate": 1.3633444450342142e-05,
+      "loss": 1.6832,
+      "step": 2332
+    },
+    {
+      "epoch": 0.40037755277158055,
+      "grad_norm": 3.202001229439437,
+      "learning_rate": 1.3628265403610588e-05,
+      "loss": 1.6174,
+      "step": 2333
+    },
+    {
+      "epoch": 0.40054916766775356,
+      "grad_norm": 2.619648985772445,
+      "learning_rate": 1.3623085235907779e-05,
+      "loss": 1.723,
+      "step": 2334
+    },
+    {
+      "epoch": 0.40072078256392657,
+      "grad_norm": 3.2512830265800767,
+      "learning_rate": 1.3617903948834155e-05,
+      "loss": 1.7123,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4008923974600995,
+      "grad_norm": 2.7856136346218574,
+      "learning_rate": 1.3612721543990505e-05,
+      "loss": 1.3425,
+      "step": 2336
+    },
+    {
+      "epoch": 0.40106401235627254,
+      "grad_norm": 4.225185232909143,
+      "learning_rate": 1.3607538022977954e-05,
+      "loss": 1.7757,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4012356272524455,
+      "grad_norm": 2.344413381213637,
+      "learning_rate": 1.3602353387397983e-05,
+      "loss": 1.3488,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4014072421486185,
+      "grad_norm": 2.39469252637547,
+      "learning_rate": 1.3597167638852407e-05,
+      "loss": 1.4525,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4015788570447915,
+      "grad_norm": 2.6378953989172995,
+      "learning_rate": 1.3591980778943394e-05,
+      "loss": 1.8389,
+      "step": 2340
+    },
+    {
+      "epoch": 0.40175047194096447,
+      "grad_norm": 3.2946064861122837,
+      "learning_rate": 1.3586792809273447e-05,
+      "loss": 1.6678,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4019220868371375,
+      "grad_norm": 2.977018322829481,
+      "learning_rate": 1.358160373144542e-05,
+      "loss": 1.6902,
+      "step": 2342
+    },
+    {
+      "epoch": 0.40209370173331044,
+      "grad_norm": 3.5461002481268245,
+      "learning_rate": 1.3576413547062502e-05,
+      "loss": 1.5975,
+      "step": 2343
+    },
+    {
+      "epoch": 0.40226531662948345,
+      "grad_norm": 3.0455946538432683,
+      "learning_rate": 1.3571222257728228e-05,
+      "loss": 1.6142,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4024369315256564,
+      "grad_norm": 3.761022202203974,
+      "learning_rate": 1.3566029865046473e-05,
+      "loss": 1.6955,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4026085464218294,
+      "grad_norm": 2.8239873542178158,
+      "learning_rate": 1.3560836370621452e-05,
+      "loss": 1.6632,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4027801613180024,
+      "grad_norm": 2.7781326596790885,
+      "learning_rate": 1.3555641776057729e-05,
+      "loss": 1.7321,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4029517762141754,
+      "grad_norm": 3.5853792791806525,
+      "learning_rate": 1.3550446082960193e-05,
+      "loss": 1.8062,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4031233911103484,
+      "grad_norm": 3.178743085830482,
+      "learning_rate": 1.3545249292934084e-05,
+      "loss": 1.5969,
+      "step": 2349
+    },
+    {
+      "epoch": 0.40329500600652135,
+      "grad_norm": 2.594948001736792,
+      "learning_rate": 1.3540051407584973e-05,
+      "loss": 1.7455,
+      "step": 2350
+    },
+    {
+      "epoch": 0.40346662090269436,
+      "grad_norm": 3.302336978380856,
+      "learning_rate": 1.353485242851878e-05,
+      "loss": 1.8062,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4036382357988673,
+      "grad_norm": 3.371819015339554,
+      "learning_rate": 1.352965235734175e-05,
+      "loss": 1.8473,
+      "step": 2352
+    },
+    {
+      "epoch": 0.40380985069504033,
+      "grad_norm": 2.9330305728459813,
+      "learning_rate": 1.3524451195660472e-05,
+      "loss": 1.6595,
+      "step": 2353
+    },
+    {
+      "epoch": 0.40398146559121334,
+      "grad_norm": 2.777512233639564,
+      "learning_rate": 1.351924894508188e-05,
+      "loss": 1.6931,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4041530804873863,
+      "grad_norm": 2.9965345275741218,
+      "learning_rate": 1.3514045607213227e-05,
+      "loss": 1.6028,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4043246953835593,
+      "grad_norm": 3.1946643934831753,
+      "learning_rate": 1.3508841183662114e-05,
+      "loss": 1.849,
+      "step": 2356
+    },
+    {
+      "epoch": 0.40449631027973226,
+      "grad_norm": 3.6069982189040086,
+      "learning_rate": 1.3503635676036475e-05,
+      "loss": 1.6741,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4046679251759053,
+      "grad_norm": 3.8117258427189014,
+      "learning_rate": 1.3498429085944573e-05,
+      "loss": 1.6118,
+      "step": 2358
+    },
+    {
+      "epoch": 0.40483954007207823,
+      "grad_norm": 3.3113205168932427,
+      "learning_rate": 1.3493221414995021e-05,
+      "loss": 1.8446,
+      "step": 2359
+    },
+    {
+      "epoch": 0.40501115496825124,
+      "grad_norm": 3.2308369849589953,
+      "learning_rate": 1.3488012664796748e-05,
+      "loss": 1.6205,
+      "step": 2360
+    },
+    {
+      "epoch": 0.40518276986442425,
+      "grad_norm": 2.2148697739498564,
+      "learning_rate": 1.3482802836959028e-05,
+      "loss": 1.8271,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4053543847605972,
+      "grad_norm": 3.3635089818766457,
+      "learning_rate": 1.347759193309146e-05,
+      "loss": 1.8421,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4055259996567702,
+      "grad_norm": 2.535274577703987,
+      "learning_rate": 1.3472379954803984e-05,
+      "loss": 1.5083,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4056976145529432,
+      "grad_norm": 2.6862321049664497,
+      "learning_rate": 1.3467166903706865e-05,
+      "loss": 1.6044,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4058692294491162,
+      "grad_norm": 3.899943993554853,
+      "learning_rate": 1.34619527814107e-05,
+      "loss": 1.5995,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4060408443452892,
+      "grad_norm": 2.979986999450808,
+      "learning_rate": 1.3456737589526428e-05,
+      "loss": 1.6707,
+      "step": 2366
+    },
+    {
+      "epoch": 0.40621245924146215,
+      "grad_norm": 2.9725980275188215,
+      "learning_rate": 1.3451521329665297e-05,
+      "loss": 1.7393,
+      "step": 2367
+    },
+    {
+      "epoch": 0.40638407413763517,
+      "grad_norm": 2.6612301630184656,
+      "learning_rate": 1.3446304003438904e-05,
+      "loss": 1.5188,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4065556890338081,
+      "grad_norm": 3.584309254377374,
+      "learning_rate": 1.3441085612459168e-05,
+      "loss": 1.6723,
+      "step": 2369
+    },
+    {
+      "epoch": 0.40672730392998113,
+      "grad_norm": 2.9770255494421045,
+      "learning_rate": 1.3435866158338335e-05,
+      "loss": 1.6215,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4068989188261541,
+      "grad_norm": 3.719197503530315,
+      "learning_rate": 1.343064564268899e-05,
+      "loss": 1.8583,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4070705337223271,
+      "grad_norm": 2.930683693078028,
+      "learning_rate": 1.3425424067124028e-05,
+      "loss": 1.7005,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4072421486185001,
+      "grad_norm": 2.643044662897078,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 1.7285,
+      "step": 2373
+    },
+    {
+      "epoch": 0.40741376351467307,
+      "grad_norm": 3.266690906591515,
+      "learning_rate": 1.3414977742700528e-05,
+      "loss": 1.6536,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4075853784108461,
+      "grad_norm": 3.272319531950153,
+      "learning_rate": 1.3409752997069437e-05,
+      "loss": 1.6254,
+      "step": 2375
+    },
+    {
+      "epoch": 0.40775699330701903,
+      "grad_norm": 3.6428979251322273,
+      "learning_rate": 1.340452719797762e-05,
+      "loss": 1.6311,
+      "step": 2376
+    },
+    {
+      "epoch": 0.40792860820319204,
+      "grad_norm": 3.1756703884844137,
+      "learning_rate": 1.339930034703962e-05,
+      "loss": 1.6189,
+      "step": 2377
+    },
+    {
+      "epoch": 0.408100223099365,
+      "grad_norm": 2.8586468184662275,
+      "learning_rate": 1.3394072445870301e-05,
+      "loss": 1.7896,
+      "step": 2378
+    },
+    {
+      "epoch": 0.408271837995538,
+      "grad_norm": 4.178899943719079,
+      "learning_rate": 1.3388843496084843e-05,
+      "loss": 1.7515,
+      "step": 2379
+    },
+    {
+      "epoch": 0.408443452891711,
+      "grad_norm": 3.882722390523321,
+      "learning_rate": 1.3383613499298762e-05,
+      "loss": 1.708,
+      "step": 2380
+    },
+    {
+      "epoch": 0.408615067787884,
+      "grad_norm": 3.6748393682110043,
+      "learning_rate": 1.3378382457127893e-05,
+      "loss": 1.7525,
+      "step": 2381
+    },
+    {
+      "epoch": 0.408786682684057,
+      "grad_norm": 3.3915391598273312,
+      "learning_rate": 1.3373150371188388e-05,
+      "loss": 1.4559,
+      "step": 2382
+    },
+    {
+      "epoch": 0.40895829758022995,
+      "grad_norm": 3.133401380831929,
+      "learning_rate": 1.3367917243096736e-05,
+      "loss": 1.5121,
+      "step": 2383
+    },
+    {
+      "epoch": 0.40912991247640296,
+      "grad_norm": 3.0944717255055734,
+      "learning_rate": 1.3362683074469729e-05,
+      "loss": 1.6274,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4093015273725759,
+      "grad_norm": 2.8739402739034885,
+      "learning_rate": 1.3357447866924498e-05,
+      "loss": 1.5392,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4094731422687489,
+      "grad_norm": 4.222459155722447,
+      "learning_rate": 1.335221162207848e-05,
+      "loss": 1.7225,
+      "step": 2386
+    },
+    {
+      "epoch": 0.40964475716492194,
+      "grad_norm": 2.813910407334618,
+      "learning_rate": 1.3346974341549448e-05,
+      "loss": 1.6782,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4098163720610949,
+      "grad_norm": 3.0641748631449603,
+      "learning_rate": 1.334173602695548e-05,
+      "loss": 1.6292,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4099879869572679,
+      "grad_norm": 2.6603098511112746,
+      "learning_rate": 1.3336496679914982e-05,
+      "loss": 1.735,
+      "step": 2389
+    },
+    {
+      "epoch": 0.41015960185344086,
+      "grad_norm": 2.8265725469990577,
+      "learning_rate": 1.3331256302046682e-05,
+      "loss": 1.7555,
+      "step": 2390
+    },
+    {
+      "epoch": 0.41033121674961387,
+      "grad_norm": 2.4889522278307656,
+      "learning_rate": 1.3326014894969614e-05,
+      "loss": 1.4641,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4105028316457869,
+      "grad_norm": 2.31789604208341,
+      "learning_rate": 1.3320772460303145e-05,
+      "loss": 1.7797,
+      "step": 2392
+    },
+    {
+      "epoch": 0.41067444654195984,
+      "grad_norm": 3.938043751594953,
+      "learning_rate": 1.3315528999666944e-05,
+      "loss": 1.7933,
+      "step": 2393
+    },
+    {
+      "epoch": 0.41084606143813285,
+      "grad_norm": 3.038438219113065,
+      "learning_rate": 1.3310284514681012e-05,
+      "loss": 1.5516,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4110176763343058,
+      "grad_norm": 2.6736953526838154,
+      "learning_rate": 1.3305039006965657e-05,
+      "loss": 1.7417,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4111892912304788,
+      "grad_norm": 3.063000070423636,
+      "learning_rate": 1.3299792478141507e-05,
+      "loss": 1.8241,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41136090612665177,
+      "grad_norm": 2.9226457280070544,
+      "learning_rate": 1.3294544929829503e-05,
+      "loss": 1.5404,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4115325210228248,
+      "grad_norm": 2.703950525665708,
+      "learning_rate": 1.32892963636509e-05,
+      "loss": 1.7731,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4117041359189978,
+      "grad_norm": 3.382890800600639,
+      "learning_rate": 1.328404678122727e-05,
+      "loss": 1.552,
+      "step": 2399
+    },
+    {
+      "epoch": 0.41187575081517075,
+      "grad_norm": 2.640777331411481,
+      "learning_rate": 1.3278796184180504e-05,
+      "loss": 1.7507,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41204736571134376,
+      "grad_norm": 2.7852046489690165,
+      "learning_rate": 1.3273544574132792e-05,
+      "loss": 1.4589,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4122189806075167,
+      "grad_norm": 3.028048895113507,
+      "learning_rate": 1.3268291952706659e-05,
+      "loss": 1.7759,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4123905955036897,
+      "grad_norm": 2.3876080773025636,
+      "learning_rate": 1.3263038321524912e-05,
+      "loss": 1.6372,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4125622103998627,
+      "grad_norm": 3.017411027837272,
+      "learning_rate": 1.3257783682210704e-05,
+      "loss": 1.7451,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4127338252960357,
+      "grad_norm": 3.094744460017128,
+      "learning_rate": 1.325252803638747e-05,
+      "loss": 1.56,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4129054401922087,
+      "grad_norm": 2.7251727995724853,
+      "learning_rate": 1.3247271385678975e-05,
+      "loss": 1.6569,
+      "step": 2406
+    },
+    {
+      "epoch": 0.41307705508838166,
+      "grad_norm": 2.841943556975302,
+      "learning_rate": 1.324201373170929e-05,
+      "loss": 1.6427,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4132486699845547,
+      "grad_norm": 2.6731575017083515,
+      "learning_rate": 1.323675507610279e-05,
+      "loss": 1.487,
+      "step": 2408
+    },
+    {
+      "epoch": 0.41342028488072763,
+      "grad_norm": 2.607854231517078,
+      "learning_rate": 1.3231495420484167e-05,
+      "loss": 1.7433,
+      "step": 2409
+    },
+    {
+      "epoch": 0.41359189977690064,
+      "grad_norm": 3.010281649227339,
+      "learning_rate": 1.3226234766478413e-05,
+      "loss": 1.7266,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4137635146730736,
+      "grad_norm": 2.4791949736503627,
+      "learning_rate": 1.3220973115710842e-05,
+      "loss": 1.5271,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4139351295692466,
+      "grad_norm": 2.8976582872021557,
+      "learning_rate": 1.3215710469807063e-05,
+      "loss": 1.6556,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4141067444654196,
+      "grad_norm": 2.9294474805422532,
+      "learning_rate": 1.3210446830392997e-05,
+      "loss": 1.6142,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4142783593615926,
+      "grad_norm": 2.6853147595404505,
+      "learning_rate": 1.3205182199094878e-05,
+      "loss": 1.5329,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4144499742577656,
+      "grad_norm": 3.7958933499250382,
+      "learning_rate": 1.3199916577539233e-05,
+      "loss": 1.6015,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41462158915393854,
+      "grad_norm": 3.1940831806276875,
+      "learning_rate": 1.3194649967352913e-05,
+      "loss": 1.7348,
+      "step": 2416
+    },
+    {
+      "epoch": 0.41479320405011155,
+      "grad_norm": 3.1472176380998236,
+      "learning_rate": 1.3189382370163053e-05,
+      "loss": 1.7635,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41496481894628456,
+      "grad_norm": 3.8697757346519266,
+      "learning_rate": 1.318411378759711e-05,
+      "loss": 1.6577,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4151364338424575,
+      "grad_norm": 2.925311450489173,
+      "learning_rate": 1.317884422128284e-05,
+      "loss": 1.4937,
+      "step": 2419
+    },
+    {
+      "epoch": 0.41530804873863053,
+      "grad_norm": 3.953168321266579,
+      "learning_rate": 1.31735736728483e-05,
+      "loss": 1.7816,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4154796636348035,
+      "grad_norm": 2.7129619524129147,
+      "learning_rate": 1.3168302143921858e-05,
+      "loss": 1.6684,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4156512785309765,
+      "grad_norm": 3.218811371511017,
+      "learning_rate": 1.3163029636132176e-05,
+      "loss": 1.7717,
+      "step": 2422
+    },
+    {
+      "epoch": 0.41582289342714945,
+      "grad_norm": 2.904854076316205,
+      "learning_rate": 1.3157756151108222e-05,
+      "loss": 1.7628,
+      "step": 2423
+    },
+    {
+      "epoch": 0.41599450832332246,
+      "grad_norm": 3.3180758454482455,
+      "learning_rate": 1.315248169047927e-05,
+      "loss": 1.9778,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4161661232194955,
+      "grad_norm": 2.504613827973436,
+      "learning_rate": 1.3147206255874886e-05,
+      "loss": 1.4349,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41633773811566843,
+      "grad_norm": 2.861541770015547,
+      "learning_rate": 1.3141929848924947e-05,
+      "loss": 1.7279,
+      "step": 2426
+    },
+    {
+      "epoch": 0.41650935301184144,
+      "grad_norm": 3.3695662367857526,
+      "learning_rate": 1.3136652471259624e-05,
+      "loss": 1.7407,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4166809679080144,
+      "grad_norm": 3.136116894075004,
+      "learning_rate": 1.3131374124509393e-05,
+      "loss": 1.5172,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4168525828041874,
+      "grad_norm": 3.025610091269555,
+      "learning_rate": 1.3126094810305024e-05,
+      "loss": 1.7815,
+      "step": 2429
+    },
+    {
+      "epoch": 0.41702419770036037,
+      "grad_norm": 2.8365900127076413,
+      "learning_rate": 1.3120814530277585e-05,
+      "loss": 1.6956,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4171958125965334,
+      "grad_norm": 2.813125905370307,
+      "learning_rate": 1.3115533286058453e-05,
+      "loss": 1.5698,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4173674274927064,
+      "grad_norm": 3.4584539128938063,
+      "learning_rate": 1.3110251079279292e-05,
+      "loss": 1.4555,
+      "step": 2432
+    },
+    {
+      "epoch": 0.41753904238887934,
+      "grad_norm": 2.5870102182830137,
+      "learning_rate": 1.3104967911572066e-05,
+      "loss": 1.4355,
+      "step": 2433
+    },
+    {
+      "epoch": 0.41771065728505236,
+      "grad_norm": 3.009675308180714,
+      "learning_rate": 1.3099683784569036e-05,
+      "loss": 1.5287,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4178822721812253,
+      "grad_norm": 3.428458138210264,
+      "learning_rate": 1.3094398699902761e-05,
+      "loss": 1.4897,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4180538870773983,
+      "grad_norm": 3.363686747175313,
+      "learning_rate": 1.3089112659206098e-05,
+      "loss": 1.783,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4182255019735713,
+      "grad_norm": 2.99965081566929,
+      "learning_rate": 1.3083825664112193e-05,
+      "loss": 1.7388,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4183971168697443,
+      "grad_norm": 2.7174261587803015,
+      "learning_rate": 1.3078537716254496e-05,
+      "loss": 1.7099,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4185687317659173,
+      "grad_norm": 2.537515826114268,
+      "learning_rate": 1.3073248817266734e-05,
+      "loss": 1.5045,
+      "step": 2439
+    },
+    {
+      "epoch": 0.41874034666209026,
+      "grad_norm": 2.8839849866664653,
+      "learning_rate": 1.3067958968782952e-05,
+      "loss": 1.8548,
+      "step": 2440
+    },
+    {
+      "epoch": 0.41891196155826327,
+      "grad_norm": 2.6449038262998634,
+      "learning_rate": 1.306266817243747e-05,
+      "loss": 1.6012,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4190835764544362,
+      "grad_norm": 2.9347101693935147,
+      "learning_rate": 1.3057376429864905e-05,
+      "loss": 1.6123,
+      "step": 2442
+    },
+    {
+      "epoch": 0.41925519135060924,
+      "grad_norm": 3.1284678526488965,
+      "learning_rate": 1.3052083742700172e-05,
+      "loss": 1.6253,
+      "step": 2443
+    },
+    {
+      "epoch": 0.41942680624678225,
+      "grad_norm": 3.311166284082919,
+      "learning_rate": 1.3046790112578474e-05,
+      "loss": 1.751,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4195984211429552,
+      "grad_norm": 2.8633745386941145,
+      "learning_rate": 1.3041495541135303e-05,
+      "loss": 1.7057,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4197700360391282,
+      "grad_norm": 3.6268574744930064,
+      "learning_rate": 1.3036200030006443e-05,
+      "loss": 1.7908,
+      "step": 2446
+    },
+    {
+      "epoch": 0.41994165093530117,
+      "grad_norm": 3.1110996297433777,
+      "learning_rate": 1.3030903580827974e-05,
+      "loss": 1.8263,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4201132658314742,
+      "grad_norm": 3.2418486119301906,
+      "learning_rate": 1.3025606195236255e-05,
+      "loss": 1.8223,
+      "step": 2448
+    },
+    {
+      "epoch": 0.42028488072764714,
+      "grad_norm": 2.3828221789929533,
+      "learning_rate": 1.3020307874867946e-05,
+      "loss": 1.3161,
+      "step": 2449
+    },
+    {
+      "epoch": 0.42045649562382015,
+      "grad_norm": 2.656760492514176,
+      "learning_rate": 1.301500862135999e-05,
+      "loss": 1.6232,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42062811051999316,
+      "grad_norm": 3.9695577341569015,
+      "learning_rate": 1.3009708436349613e-05,
+      "loss": 1.8714,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4207997254161661,
+      "grad_norm": 3.011636285342456,
+      "learning_rate": 1.300440732147434e-05,
+      "loss": 1.6104,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4209713403123391,
+      "grad_norm": 3.1213909645924116,
+      "learning_rate": 1.2999105278371977e-05,
+      "loss": 1.8676,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4211429552085121,
+      "grad_norm": 2.581359865085188,
+      "learning_rate": 1.2993802308680615e-05,
+      "loss": 1.4625,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4213145701046851,
+      "grad_norm": 3.5425208790720073,
+      "learning_rate": 1.2988498414038635e-05,
+      "loss": 1.6825,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42148618500085805,
+      "grad_norm": 3.146873128662197,
+      "learning_rate": 1.2983193596084704e-05,
+      "loss": 1.6977,
+      "step": 2456
+    },
+    {
+      "epoch": 0.42165779989703106,
+      "grad_norm": 2.768754597582966,
+      "learning_rate": 1.2977887856457772e-05,
+      "loss": 1.4894,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42182941479320407,
+      "grad_norm": 3.582524608584026,
+      "learning_rate": 1.297258119679707e-05,
+      "loss": 1.6369,
+      "step": 2458
+    },
+    {
+      "epoch": 0.422001029689377,
+      "grad_norm": 2.4417208540393163,
+      "learning_rate": 1.2967273618742125e-05,
+      "loss": 1.466,
+      "step": 2459
+    },
+    {
+      "epoch": 0.42217264458555004,
+      "grad_norm": 3.510206071642065,
+      "learning_rate": 1.2961965123932738e-05,
+      "loss": 1.701,
+      "step": 2460
+    },
+    {
+      "epoch": 0.422344259481723,
+      "grad_norm": 2.659599926797607,
+      "learning_rate": 1.2956655714008992e-05,
+      "loss": 1.7297,
+      "step": 2461
+    },
+    {
+      "epoch": 0.422515874377896,
+      "grad_norm": 3.259899806471716,
+      "learning_rate": 1.2951345390611265e-05,
+      "loss": 1.5041,
+      "step": 2462
+    },
+    {
+      "epoch": 0.42268748927406896,
+      "grad_norm": 5.281876660516446,
+      "learning_rate": 1.2946034155380196e-05,
+      "loss": 1.7176,
+      "step": 2463
+    },
+    {
+      "epoch": 0.422859104170242,
+      "grad_norm": 2.889970708269304,
+      "learning_rate": 1.2940722009956731e-05,
+      "loss": 1.6845,
+      "step": 2464
+    },
+    {
+      "epoch": 0.423030719066415,
+      "grad_norm": 3.0943982460867576,
+      "learning_rate": 1.2935408955982079e-05,
+      "loss": 1.6487,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42320233396258794,
+      "grad_norm": 3.192130937293403,
+      "learning_rate": 1.2930094995097732e-05,
+      "loss": 1.7028,
+      "step": 2466
+    },
+    {
+      "epoch": 0.42337394885876095,
+      "grad_norm": 4.018392818013617,
+      "learning_rate": 1.2924780128945473e-05,
+      "loss": 1.9719,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4235455637549339,
+      "grad_norm": 3.023979770776539,
+      "learning_rate": 1.2919464359167356e-05,
+      "loss": 1.5915,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4237171786511069,
+      "grad_norm": 2.9867689907831028,
+      "learning_rate": 1.291414768740571e-05,
+      "loss": 1.623,
+      "step": 2469
+    },
+    {
+      "epoch": 0.42388879354727993,
+      "grad_norm": 3.792612530145307,
+      "learning_rate": 1.2908830115303149e-05,
+      "loss": 1.5804,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4240604084434529,
+      "grad_norm": 2.6266491565189667,
+      "learning_rate": 1.2903511644502568e-05,
+      "loss": 1.7264,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4242320233396259,
+      "grad_norm": 5.1754152604025725,
+      "learning_rate": 1.2898192276647135e-05,
+      "loss": 1.5763,
+      "step": 2472
+    },
+    {
+      "epoch": 0.42440363823579885,
+      "grad_norm": 3.440325934940613,
+      "learning_rate": 1.2892872013380293e-05,
+      "loss": 1.6736,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42457525313197186,
+      "grad_norm": 2.7819449336433157,
+      "learning_rate": 1.288755085634577e-05,
+      "loss": 1.694,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4247468680281448,
+      "grad_norm": 3.376997882586126,
+      "learning_rate": 1.2882228807187559e-05,
+      "loss": 1.6484,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42491848292431783,
+      "grad_norm": 3.114681558827002,
+      "learning_rate": 1.2876905867549938e-05,
+      "loss": 1.7088,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42509009782049084,
+      "grad_norm": 3.334036704464508,
+      "learning_rate": 1.2871582039077456e-05,
+      "loss": 1.6168,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4252617127166638,
+      "grad_norm": 3.369345663161789,
+      "learning_rate": 1.2866257323414936e-05,
+      "loss": 1.7059,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4254333276128368,
+      "grad_norm": 2.650222544411065,
+      "learning_rate": 1.286093172220748e-05,
+      "loss": 1.5365,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42560494250900976,
+      "grad_norm": 3.276973658428143,
+      "learning_rate": 1.2855605237100454e-05,
+      "loss": 1.7649,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4257765574051828,
+      "grad_norm": 2.7366032614387468,
+      "learning_rate": 1.2850277869739509e-05,
+      "loss": 1.7724,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42594817230135573,
+      "grad_norm": 3.4422640524324453,
+      "learning_rate": 1.2844949621770558e-05,
+      "loss": 1.7285,
+      "step": 2482
+    },
+    {
+      "epoch": 0.42611978719752874,
+      "grad_norm": 3.3374359649605543,
+      "learning_rate": 1.283962049483979e-05,
+      "loss": 1.6632,
+      "step": 2483
+    },
+    {
+      "epoch": 0.42629140209370175,
+      "grad_norm": 3.032369116566574,
+      "learning_rate": 1.2834290490593673e-05,
+      "loss": 1.5318,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4264630169898747,
+      "grad_norm": 3.600448353448306,
+      "learning_rate": 1.282895961067893e-05,
+      "loss": 1.5226,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4266346318860477,
+      "grad_norm": 5.2462530037375075,
+      "learning_rate": 1.2823627856742573e-05,
+      "loss": 1.8112,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4268062467822207,
+      "grad_norm": 3.370526626330154,
+      "learning_rate": 1.2818295230431868e-05,
+      "loss": 1.8661,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4269778616783937,
+      "grad_norm": 3.4255163512798243,
+      "learning_rate": 1.2812961733394357e-05,
+      "loss": 1.8682,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4271494765745667,
+      "grad_norm": 3.971550612724134,
+      "learning_rate": 1.2807627367277858e-05,
+      "loss": 1.6008,
+      "step": 2489
+    },
+    {
+      "epoch": 0.42732109147073966,
+      "grad_norm": 3.026515880260373,
+      "learning_rate": 1.2802292133730443e-05,
+      "loss": 1.5783,
+      "step": 2490
+    },
+    {
+      "epoch": 0.42749270636691267,
+      "grad_norm": 3.0694245871201504,
+      "learning_rate": 1.2796956034400467e-05,
+      "loss": 1.5052,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4276643212630856,
+      "grad_norm": 3.6119330081681067,
+      "learning_rate": 1.2791619070936538e-05,
+      "loss": 1.6925,
+      "step": 2492
+    },
+    {
+      "epoch": 0.42783593615925863,
+      "grad_norm": 3.0229506148327543,
+      "learning_rate": 1.2786281244987546e-05,
+      "loss": 1.7646,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4280075510554316,
+      "grad_norm": 3.453197359685171,
+      "learning_rate": 1.2780942558202633e-05,
+      "loss": 1.5679,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4281791659516046,
+      "grad_norm": 2.982202091343988,
+      "learning_rate": 1.2775603012231218e-05,
+      "loss": 1.6655,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4283507808477776,
+      "grad_norm": 3.2921049966667373,
+      "learning_rate": 1.277026260872298e-05,
+      "loss": 1.8113,
+      "step": 2496
+    },
+    {
+      "epoch": 0.42852239574395057,
+      "grad_norm": 3.0090259415730416,
+      "learning_rate": 1.2764921349327864e-05,
+      "loss": 1.6267,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4286940106401236,
+      "grad_norm": 3.0477994828356643,
+      "learning_rate": 1.275957923569608e-05,
+      "loss": 1.7909,
+      "step": 2498
+    },
+    {
+      "epoch": 0.42886562553629654,
+      "grad_norm": 3.275654908227982,
+      "learning_rate": 1.27542362694781e-05,
+      "loss": 1.7368,
+      "step": 2499
+    },
+    {
+      "epoch": 0.42903724043246955,
+      "grad_norm": 3.3348718752774444,
+      "learning_rate": 1.2748892452324661e-05,
+      "loss": 1.7296,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4292088553286425,
+      "grad_norm": 3.391286033750643,
+      "learning_rate": 1.2743547785886765e-05,
+      "loss": 1.4789,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4293804702248155,
+      "grad_norm": 4.007492121640316,
+      "learning_rate": 1.2738202271815671e-05,
+      "loss": 1.6838,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4295520851209885,
+      "grad_norm": 3.098277320150405,
+      "learning_rate": 1.2732855911762907e-05,
+      "loss": 1.8444,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4297237000171615,
+      "grad_norm": 3.1982147931971388,
+      "learning_rate": 1.2727508707380257e-05,
+      "loss": 1.7262,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4298953149133345,
+      "grad_norm": 2.746371600249128,
+      "learning_rate": 1.2722160660319766e-05,
+      "loss": 1.5578,
+      "step": 2505
+    },
+    {
+      "epoch": 0.43006692980950745,
+      "grad_norm": 2.964089110595723,
+      "learning_rate": 1.271681177223374e-05,
+      "loss": 1.4712,
+      "step": 2506
+    },
+    {
+      "epoch": 0.43023854470568046,
+      "grad_norm": 2.9879752219020377,
+      "learning_rate": 1.2711462044774747e-05,
+      "loss": 1.8062,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4304101596018534,
+      "grad_norm": 2.6547171499983238,
+      "learning_rate": 1.2706111479595613e-05,
+      "loss": 1.563,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4305817744980264,
+      "grad_norm": 3.289466648602883,
+      "learning_rate": 1.270076007834942e-05,
+      "loss": 1.8348,
+      "step": 2509
+    },
+    {
+      "epoch": 0.43075338939419944,
+      "grad_norm": 2.70348806902513,
+      "learning_rate": 1.2695407842689518e-05,
+      "loss": 1.5201,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4309250042903724,
+      "grad_norm": 3.371299777829686,
+      "learning_rate": 1.26900547742695e-05,
+      "loss": 2.019,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4310966191865454,
+      "grad_norm": 2.6533607468246148,
+      "learning_rate": 1.2684700874743228e-05,
+      "loss": 1.2952,
+      "step": 2512
+    },
+    {
+      "epoch": 0.43126823408271836,
+      "grad_norm": 2.913953722768904,
+      "learning_rate": 1.2679346145764812e-05,
+      "loss": 1.8159,
+      "step": 2513
+    },
+    {
+      "epoch": 0.43143984897889137,
+      "grad_norm": 3.1425985200207114,
+      "learning_rate": 1.267399058898863e-05,
+      "loss": 1.7227,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4316114638750644,
+      "grad_norm": 3.1812145975071457,
+      "learning_rate": 1.2668634206069305e-05,
+      "loss": 1.8711,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43178307877123734,
+      "grad_norm": 2.778731491270658,
+      "learning_rate": 1.266327699866172e-05,
+      "loss": 1.5984,
+      "step": 2516
+    },
+    {
+      "epoch": 0.43195469366741035,
+      "grad_norm": 2.6114293162120568,
+      "learning_rate": 1.265791896842101e-05,
+      "loss": 1.6444,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4321263085635833,
+      "grad_norm": 3.159336600099177,
+      "learning_rate": 1.2652560117002567e-05,
+      "loss": 1.7862,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4322979234597563,
+      "grad_norm": 2.6823753096696907,
+      "learning_rate": 1.2647200446062037e-05,
+      "loss": 1.4726,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4324695383559293,
+      "grad_norm": 2.6394155243791904,
+      "learning_rate": 1.2641839957255317e-05,
+      "loss": 1.6948,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4326411532521023,
+      "grad_norm": 3.379387051767385,
+      "learning_rate": 1.2636478652238556e-05,
+      "loss": 1.5138,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4328127681482753,
+      "grad_norm": 2.90660532210399,
+      "learning_rate": 1.2631116532668162e-05,
+      "loss": 1.7663,
+      "step": 2522
+    },
+    {
+      "epoch": 0.43298438304444825,
+      "grad_norm": 2.5621277355929717,
+      "learning_rate": 1.2625753600200784e-05,
+      "loss": 1.5962,
+      "step": 2523
+    },
+    {
+      "epoch": 0.43315599794062126,
+      "grad_norm": 2.6803743014738775,
+      "learning_rate": 1.2620389856493326e-05,
+      "loss": 1.505,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4333276128367942,
+      "grad_norm": 2.7100021043004774,
+      "learning_rate": 1.2615025303202952e-05,
+      "loss": 1.5001,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43349922773296723,
+      "grad_norm": 3.2384546908452214,
+      "learning_rate": 1.260965994198706e-05,
+      "loss": 1.5471,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4336708426291402,
+      "grad_norm": 2.504570366147136,
+      "learning_rate": 1.2604293774503311e-05,
+      "loss": 1.3906,
+      "step": 2527
+    },
+    {
+      "epoch": 0.4338424575253132,
+      "grad_norm": 2.6517604418568257,
+      "learning_rate": 1.259892680240961e-05,
+      "loss": 1.754,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4340140724214862,
+      "grad_norm": 3.5500915066469014,
+      "learning_rate": 1.2593559027364108e-05,
+      "loss": 1.8265,
+      "step": 2529
+    },
+    {
+      "epoch": 0.43418568731765916,
+      "grad_norm": 2.6060685890118305,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 1.8128,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4343573022138322,
+      "grad_norm": 3.0960523832519042,
+      "learning_rate": 1.2582821075051561e-05,
+      "loss": 1.5468,
+      "step": 2531
+    },
+    {
+      "epoch": 0.43452891711000513,
+      "grad_norm": 3.2563715708076546,
+      "learning_rate": 1.257745090110206e-05,
+      "loss": 1.5619,
+      "step": 2532
+    },
+    {
+      "epoch": 0.43470053200617814,
+      "grad_norm": 2.618860505704093,
+      "learning_rate": 1.257207993083585e-05,
+      "loss": 1.8292,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4348721469023511,
+      "grad_norm": 3.2065405593672485,
+      "learning_rate": 1.2566708165912322e-05,
+      "loss": 1.711,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4350437617985241,
+      "grad_norm": 3.081878371852855,
+      "learning_rate": 1.2561335607991104e-05,
+      "loss": 1.5996,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4352153766946971,
+      "grad_norm": 3.4555479753699974,
+      "learning_rate": 1.255596225873208e-05,
+      "loss": 1.9782,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4353869915908701,
+      "grad_norm": 2.781853752290773,
+      "learning_rate": 1.2550588119795372e-05,
+      "loss": 1.5864,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4355586064870431,
+      "grad_norm": 3.623781301884262,
+      "learning_rate": 1.254521319284135e-05,
+      "loss": 1.5382,
+      "step": 2538
+    },
+    {
+      "epoch": 0.43573022138321604,
+      "grad_norm": 3.4257640805569958,
+      "learning_rate": 1.2539837479530626e-05,
+      "loss": 1.6121,
+      "step": 2539
+    },
+    {
+      "epoch": 0.43590183627938905,
+      "grad_norm": 3.0553389652452947,
+      "learning_rate": 1.253446098152405e-05,
+      "loss": 1.8595,
+      "step": 2540
+    },
+    {
+      "epoch": 0.43607345117556207,
+      "grad_norm": 2.687459921468701,
+      "learning_rate": 1.2529083700482722e-05,
+      "loss": 1.6432,
+      "step": 2541
+    },
+    {
+      "epoch": 0.436245066071735,
+      "grad_norm": 3.1545775205254984,
+      "learning_rate": 1.252370563806798e-05,
+      "loss": 1.5666,
+      "step": 2542
+    },
+    {
+      "epoch": 0.43641668096790803,
+      "grad_norm": 2.8986822006365127,
+      "learning_rate": 1.2518326795941405e-05,
+      "loss": 1.6598,
+      "step": 2543
+    },
+    {
+      "epoch": 0.436588295864081,
+      "grad_norm": 4.054391636555573,
+      "learning_rate": 1.2512947175764815e-05,
+      "loss": 1.7002,
+      "step": 2544
+    },
+    {
+      "epoch": 0.436759910760254,
+      "grad_norm": 3.578548290273076,
+      "learning_rate": 1.2507566779200273e-05,
+      "loss": 1.5659,
+      "step": 2545
+    },
+    {
+      "epoch": 0.43693152565642696,
+      "grad_norm": 3.2339857848140503,
+      "learning_rate": 1.2502185607910082e-05,
+      "loss": 1.6849,
+      "step": 2546
+    },
+    {
+      "epoch": 0.43710314055259997,
+      "grad_norm": 2.678436382132492,
+      "learning_rate": 1.249680366355678e-05,
+      "loss": 1.8115,
+      "step": 2547
+    },
+    {
+      "epoch": 0.437274755448773,
+      "grad_norm": 3.488463153827224,
+      "learning_rate": 1.2491420947803148e-05,
+      "loss": 1.5979,
+      "step": 2548
+    },
+    {
+      "epoch": 0.43744637034494593,
+      "grad_norm": 3.0841260106537947,
+      "learning_rate": 1.2486037462312202e-05,
+      "loss": 1.8077,
+      "step": 2549
+    },
+    {
+      "epoch": 0.43761798524111895,
+      "grad_norm": 3.329999027823043,
+      "learning_rate": 1.2480653208747198e-05,
+      "loss": 1.6684,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4377896001372919,
+      "grad_norm": 3.355867133233169,
+      "learning_rate": 1.2475268188771628e-05,
+      "loss": 1.8114,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4379612150334649,
+      "grad_norm": 2.9116058326019285,
+      "learning_rate": 1.2469882404049227e-05,
+      "loss": 1.6893,
+      "step": 2552
+    },
+    {
+      "epoch": 0.43813282992963787,
+      "grad_norm": 3.819052716207462,
+      "learning_rate": 1.2464495856243955e-05,
+      "loss": 1.7297,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4383044448258109,
+      "grad_norm": 6.273556976261411,
+      "learning_rate": 1.2459108547020014e-05,
+      "loss": 1.8098,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4384760597219839,
+      "grad_norm": 6.812357201085777,
+      "learning_rate": 1.2453720478041842e-05,
+      "loss": 1.5496,
+      "step": 2555
+    },
+    {
+      "epoch": 0.43864767461815685,
+      "grad_norm": 3.101693527083838,
+      "learning_rate": 1.2448331650974113e-05,
+      "loss": 1.6682,
+      "step": 2556
+    },
+    {
+      "epoch": 0.43881928951432986,
+      "grad_norm": 3.7034135251626763,
+      "learning_rate": 1.244294206748173e-05,
+      "loss": 1.5942,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4389909044105028,
+      "grad_norm": 3.5477499843622877,
+      "learning_rate": 1.2437551729229835e-05,
+      "loss": 1.7179,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4391625193066758,
+      "grad_norm": 3.0806240986978275,
+      "learning_rate": 1.2432160637883794e-05,
+      "loss": 1.7759,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4393341342028488,
+      "grad_norm": 3.552008839972291,
+      "learning_rate": 1.2426768795109223e-05,
+      "loss": 1.7965,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4395057490990218,
+      "grad_norm": 3.1483070739792582,
+      "learning_rate": 1.2421376202571951e-05,
+      "loss": 1.5987,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4396773639951948,
+      "grad_norm": 2.7763357843333165,
+      "learning_rate": 1.2415982861938049e-05,
+      "loss": 1.5757,
+      "step": 2562
+    },
+    {
+      "epoch": 0.43984897889136776,
+      "grad_norm": 2.922576816265704,
+      "learning_rate": 1.2410588774873822e-05,
+      "loss": 1.8297,
+      "step": 2563
+    },
+    {
+      "epoch": 0.44002059378754077,
+      "grad_norm": 2.641247574186035,
+      "learning_rate": 1.2405193943045797e-05,
+      "loss": 1.382,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4401922086837137,
+      "grad_norm": 3.122360740773097,
+      "learning_rate": 1.2399798368120735e-05,
+      "loss": 1.4621,
+      "step": 2565
+    },
+    {
+      "epoch": 0.44036382357988674,
+      "grad_norm": 3.6300781946347467,
+      "learning_rate": 1.2394402051765624e-05,
+      "loss": 1.732,
+      "step": 2566
+    },
+    {
+      "epoch": 0.44053543847605975,
+      "grad_norm": 3.6406195478337113,
+      "learning_rate": 1.2389004995647689e-05,
+      "loss": 2.0176,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4407070533722327,
+      "grad_norm": 3.270127926494256,
+      "learning_rate": 1.2383607201434378e-05,
+      "loss": 1.618,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4408786682684057,
+      "grad_norm": 2.622356348934535,
+      "learning_rate": 1.2378208670793361e-05,
+      "loss": 1.7736,
+      "step": 2569
+    },
+    {
+      "epoch": 0.44105028316457867,
+      "grad_norm": 2.9566685792430785,
+      "learning_rate": 1.2372809405392551e-05,
+      "loss": 1.6427,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4412218980607517,
+      "grad_norm": 2.5076323684799164,
+      "learning_rate": 1.236740940690007e-05,
+      "loss": 1.5429,
+      "step": 2571
+    },
+    {
+      "epoch": 0.44139351295692464,
+      "grad_norm": 2.6831721794783587,
+      "learning_rate": 1.2362008676984281e-05,
+      "loss": 1.5456,
+      "step": 2572
+    },
+    {
+      "epoch": 0.44156512785309765,
+      "grad_norm": 2.823296283918164,
+      "learning_rate": 1.2356607217313765e-05,
+      "loss": 1.7139,
+      "step": 2573
+    },
+    {
+      "epoch": 0.44173674274927066,
+      "grad_norm": 4.120667728216354,
+      "learning_rate": 1.235120502955733e-05,
+      "loss": 1.9691,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4419083576454436,
+      "grad_norm": 2.8717970757022693,
+      "learning_rate": 1.2345802115384014e-05,
+      "loss": 1.674,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44207997254161663,
+      "grad_norm": 2.53361818564715,
+      "learning_rate": 1.2340398476463069e-05,
+      "loss": 1.5349,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4422515874377896,
+      "grad_norm": 2.4607102792534556,
+      "learning_rate": 1.233499411446398e-05,
+      "loss": 1.6835,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4424232023339626,
+      "grad_norm": 8.685491447194668,
+      "learning_rate": 1.2329589031056454e-05,
+      "loss": 1.5918,
+      "step": 2578
+    },
+    {
+      "epoch": 0.44259481723013555,
+      "grad_norm": 3.0363771532608754,
+      "learning_rate": 1.2324183227910418e-05,
+      "loss": 1.6851,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44276643212630856,
+      "grad_norm": 2.5378929876143603,
+      "learning_rate": 1.2318776706696021e-05,
+      "loss": 1.9006,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4429380470224816,
+      "grad_norm": 2.374330167558309,
+      "learning_rate": 1.2313369469083636e-05,
+      "loss": 1.423,
+      "step": 2581
+    },
+    {
+      "epoch": 0.44310966191865453,
+      "grad_norm": 3.053470888666547,
+      "learning_rate": 1.2307961516743862e-05,
+      "loss": 1.3583,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44328127681482754,
+      "grad_norm": 3.2647427085424625,
+      "learning_rate": 1.2302552851347506e-05,
+      "loss": 1.5868,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4434528917110005,
+      "grad_norm": 3.886072547187632,
+      "learning_rate": 1.229714347456561e-05,
+      "loss": 1.7482,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4436245066071735,
+      "grad_norm": 3.0849422437361693,
+      "learning_rate": 1.2291733388069422e-05,
+      "loss": 1.5925,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44379612150334646,
+      "grad_norm": 2.966983680827161,
+      "learning_rate": 1.228632259353042e-05,
+      "loss": 1.6899,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4439677363995195,
+      "grad_norm": 2.546985271873544,
+      "learning_rate": 1.2280911092620298e-05,
+      "loss": 1.7253,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4441393512956925,
+      "grad_norm": 3.1319908880495766,
+      "learning_rate": 1.2275498887010963e-05,
+      "loss": 1.7757,
+      "step": 2588
+    },
+    {
+      "epoch": 0.44431096619186544,
+      "grad_norm": 4.507214578446561,
+      "learning_rate": 1.2270085978374545e-05,
+      "loss": 1.6496,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44448258108803845,
+      "grad_norm": 3.400663752161522,
+      "learning_rate": 1.2264672368383392e-05,
+      "loss": 1.7275,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4446541959842114,
+      "grad_norm": 4.2543266896916005,
+      "learning_rate": 1.225925805871007e-05,
+      "loss": 1.6991,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4448258108803844,
+      "grad_norm": 3.014388705452953,
+      "learning_rate": 1.2253843051027347e-05,
+      "loss": 1.7271,
+      "step": 2592
+    },
+    {
+      "epoch": 0.44499742577655743,
+      "grad_norm": 2.7174643747910965,
+      "learning_rate": 1.2248427347008224e-05,
+      "loss": 1.8008,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4451690406727304,
+      "grad_norm": 3.1056334300432025,
+      "learning_rate": 1.2243010948325918e-05,
+      "loss": 1.8224,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4453406555689034,
+      "grad_norm": 3.0515225248042803,
+      "learning_rate": 1.2237593856653838e-05,
+      "loss": 1.7033,
+      "step": 2595
+    },
+    {
+      "epoch": 0.44551227046507635,
+      "grad_norm": 3.784391122633235,
+      "learning_rate": 1.2232176073665631e-05,
+      "loss": 1.5208,
+      "step": 2596
+    },
+    {
+      "epoch": 0.44568388536124937,
+      "grad_norm": 2.8289385952920276,
+      "learning_rate": 1.222675760103515e-05,
+      "loss": 1.5897,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4458555002574223,
+      "grad_norm": 2.4983918490154715,
+      "learning_rate": 1.2221338440436456e-05,
+      "loss": 1.7346,
+      "step": 2598
+    },
+    {
+      "epoch": 0.44602711515359533,
+      "grad_norm": 3.269913607176651,
+      "learning_rate": 1.221591859354383e-05,
+      "loss": 1.6967,
+      "step": 2599
+    },
+    {
+      "epoch": 0.44619873004976834,
+      "grad_norm": 3.2950159496429117,
+      "learning_rate": 1.2210498062031759e-05,
+      "loss": 1.6056,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4463703449459413,
+      "grad_norm": 2.9395221388436417,
+      "learning_rate": 1.2205076847574945e-05,
+      "loss": 1.5197,
+      "step": 2601
+    },
+    {
+      "epoch": 0.4465419598421143,
+      "grad_norm": 3.1248826064471156,
+      "learning_rate": 1.2199654951848301e-05,
+      "loss": 1.9075,
+      "step": 2602
+    },
+    {
+      "epoch": 0.44671357473828727,
+      "grad_norm": 3.304105023360127,
+      "learning_rate": 1.2194232376526948e-05,
+      "loss": 1.8778,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4468851896344603,
+      "grad_norm": 2.982924041085048,
+      "learning_rate": 1.218880912328622e-05,
+      "loss": 1.7241,
+      "step": 2604
+    },
+    {
+      "epoch": 0.44705680453063323,
+      "grad_norm": 2.437170032811617,
+      "learning_rate": 1.2183385193801655e-05,
+      "loss": 1.5812,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44722841942680625,
+      "grad_norm": 3.027329515915909,
+      "learning_rate": 1.2177960589749014e-05,
+      "loss": 1.9349,
+      "step": 2606
+    },
+    {
+      "epoch": 0.44740003432297926,
+      "grad_norm": 3.851117223695634,
+      "learning_rate": 1.2172535312804243e-05,
+      "loss": 1.8446,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4475716492191522,
+      "grad_norm": 3.193010343206526,
+      "learning_rate": 1.2167109364643518e-05,
+      "loss": 1.6222,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4477432641153252,
+      "grad_norm": 3.537264674635834,
+      "learning_rate": 1.216168274694321e-05,
+      "loss": 1.8511,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4479148790114982,
+      "grad_norm": 2.751241739903351,
+      "learning_rate": 1.2156255461379903e-05,
+      "loss": 1.7101,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4480864939076712,
+      "grad_norm": 2.603927554587191,
+      "learning_rate": 1.2150827509630383e-05,
+      "loss": 1.6773,
+      "step": 2611
+    },
+    {
+      "epoch": 0.44825810880384415,
+      "grad_norm": 3.0629529017505117,
+      "learning_rate": 1.2145398893371645e-05,
+      "loss": 1.7964,
+      "step": 2612
+    },
+    {
+      "epoch": 0.44842972370001716,
+      "grad_norm": 2.9105349444727318,
+      "learning_rate": 1.2139969614280886e-05,
+      "loss": 1.4692,
+      "step": 2613
+    },
+    {
+      "epoch": 0.44860133859619017,
+      "grad_norm": 3.2146378611423043,
+      "learning_rate": 1.2134539674035511e-05,
+      "loss": 1.967,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4487729534923631,
+      "grad_norm": 4.810921855735679,
+      "learning_rate": 1.2129109074313126e-05,
+      "loss": 1.7202,
+      "step": 2615
+    },
+    {
+      "epoch": 0.44894456838853614,
+      "grad_norm": 3.859967797231811,
+      "learning_rate": 1.2123677816791546e-05,
+      "loss": 1.7944,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4491161832847091,
+      "grad_norm": 2.6340297117677736,
+      "learning_rate": 1.2118245903148782e-05,
+      "loss": 1.519,
+      "step": 2617
+    },
+    {
+      "epoch": 0.4492877981808821,
+      "grad_norm": 3.1083475224696118,
+      "learning_rate": 1.2112813335063058e-05,
+      "loss": 1.7049,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4494594130770551,
+      "grad_norm": 2.857870209351344,
+      "learning_rate": 1.2107380114212788e-05,
+      "loss": 1.6021,
+      "step": 2619
+    },
+    {
+      "epoch": 0.44963102797322807,
+      "grad_norm": 2.7377907307525353,
+      "learning_rate": 1.2101946242276594e-05,
+      "loss": 1.5989,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4498026428694011,
+      "grad_norm": 6.086643739019491,
+      "learning_rate": 1.2096511720933301e-05,
+      "loss": 1.8588,
+      "step": 2621
+    },
+    {
+      "epoch": 0.44997425776557404,
+      "grad_norm": 2.8857916443680582,
+      "learning_rate": 1.2091076551861931e-05,
+      "loss": 1.6858,
+      "step": 2622
+    },
+    {
+      "epoch": 0.45014587266174705,
+      "grad_norm": 3.31440482177447,
+      "learning_rate": 1.2085640736741708e-05,
+      "loss": 1.6271,
+      "step": 2623
+    },
+    {
+      "epoch": 0.45031748755792,
+      "grad_norm": 3.7446601859902917,
+      "learning_rate": 1.2080204277252055e-05,
+      "loss": 1.8249,
+      "step": 2624
+    },
+    {
+      "epoch": 0.450489102454093,
+      "grad_norm": 2.708340819102802,
+      "learning_rate": 1.2074767175072592e-05,
+      "loss": 1.6096,
+      "step": 2625
+    },
+    {
+      "epoch": 0.450660717350266,
+      "grad_norm": 2.604841225539455,
+      "learning_rate": 1.2069329431883144e-05,
+      "loss": 1.6638,
+      "step": 2626
+    },
+    {
+      "epoch": 0.450832332246439,
+      "grad_norm": 4.256384406631859,
+      "learning_rate": 1.2063891049363725e-05,
+      "loss": 1.4591,
+      "step": 2627
+    },
+    {
+      "epoch": 0.451003947142612,
+      "grad_norm": 2.8697090254577313,
+      "learning_rate": 1.2058452029194556e-05,
+      "loss": 1.7297,
+      "step": 2628
+    },
+    {
+      "epoch": 0.45117556203878495,
+      "grad_norm": 3.262288916173502,
+      "learning_rate": 1.2053012373056043e-05,
+      "loss": 1.5255,
+      "step": 2629
+    },
+    {
+      "epoch": 0.45134717693495796,
+      "grad_norm": 3.032200308239193,
+      "learning_rate": 1.2047572082628807e-05,
+      "loss": 1.8503,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4515187918311309,
+      "grad_norm": 3.1295044735038404,
+      "learning_rate": 1.2042131159593643e-05,
+      "loss": 1.6148,
+      "step": 2631
+    },
+    {
+      "epoch": 0.45169040672730393,
+      "grad_norm": 2.883092782789536,
+      "learning_rate": 1.2036689605631557e-05,
+      "loss": 1.8782,
+      "step": 2632
+    },
+    {
+      "epoch": 0.45186202162347694,
+      "grad_norm": 3.4864386866907044,
+      "learning_rate": 1.2031247422423746e-05,
+      "loss": 1.6105,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4520336365196499,
+      "grad_norm": 3.420573358637991,
+      "learning_rate": 1.2025804611651595e-05,
+      "loss": 1.6429,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4522052514158229,
+      "grad_norm": 2.9253390573924696,
+      "learning_rate": 1.2020361174996694e-05,
+      "loss": 1.4733,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45237686631199586,
+      "grad_norm": 3.2411935585420366,
+      "learning_rate": 1.2014917114140823e-05,
+      "loss": 1.665,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4525484812081689,
+      "grad_norm": 3.558327855457531,
+      "learning_rate": 1.2009472430765945e-05,
+      "loss": 1.7214,
+      "step": 2637
+    },
+    {
+      "epoch": 0.45272009610434183,
+      "grad_norm": 3.7710362239244937,
+      "learning_rate": 1.2004027126554225e-05,
+      "loss": 1.5841,
+      "step": 2638
+    },
+    {
+      "epoch": 0.45289171100051484,
+      "grad_norm": 3.9690226416401333,
+      "learning_rate": 1.1998581203188022e-05,
+      "loss": 1.6546,
+      "step": 2639
+    },
+    {
+      "epoch": 0.45306332589668785,
+      "grad_norm": 2.7141630145160724,
+      "learning_rate": 1.199313466234988e-05,
+      "loss": 1.4723,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4532349407928608,
+      "grad_norm": 3.1166574525253345,
+      "learning_rate": 1.1987687505722532e-05,
+      "loss": 1.4558,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4534065556890338,
+      "grad_norm": 3.4493588934008863,
+      "learning_rate": 1.1982239734988912e-05,
+      "loss": 1.6947,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4535781705852068,
+      "grad_norm": 2.8326284353603284,
+      "learning_rate": 1.1976791351832128e-05,
+      "loss": 1.731,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4537497854813798,
+      "grad_norm": 3.561561852429182,
+      "learning_rate": 1.1971342357935491e-05,
+      "loss": 1.6676,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4539214003775528,
+      "grad_norm": 3.462669755214249,
+      "learning_rate": 1.1965892754982502e-05,
+      "loss": 1.747,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45409301527372575,
+      "grad_norm": 2.9375413525713037,
+      "learning_rate": 1.1960442544656832e-05,
+      "loss": 1.6186,
+      "step": 2646
+    },
+    {
+      "epoch": 0.45426463016989876,
+      "grad_norm": 2.7526503418153494,
+      "learning_rate": 1.1954991728642358e-05,
+      "loss": 1.6154,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4544362450660717,
+      "grad_norm": 3.684326053700525,
+      "learning_rate": 1.1949540308623138e-05,
+      "loss": 1.7381,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45460785996224473,
+      "grad_norm": 3.3167738095803356,
+      "learning_rate": 1.1944088286283414e-05,
+      "loss": 1.6568,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4547794748584177,
+      "grad_norm": 3.0301356368030583,
+      "learning_rate": 1.1938635663307617e-05,
+      "loss": 1.7196,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4549510897545907,
+      "grad_norm": 2.3088339456853477,
+      "learning_rate": 1.1933182441380362e-05,
+      "loss": 1.5877,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4551227046507637,
+      "grad_norm": 2.7664559466035787,
+      "learning_rate": 1.1927728622186455e-05,
+      "loss": 1.7043,
+      "step": 2652
+    },
+    {
+      "epoch": 0.45529431954693667,
+      "grad_norm": 2.8404715224080244,
+      "learning_rate": 1.1922274207410874e-05,
+      "loss": 1.598,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4554659344431097,
+      "grad_norm": 3.6439976766696143,
+      "learning_rate": 1.1916819198738794e-05,
+      "loss": 1.6171,
+      "step": 2654
+    },
+    {
+      "epoch": 0.45563754933928263,
+      "grad_norm": 3.7464085059845873,
+      "learning_rate": 1.1911363597855563e-05,
+      "loss": 1.5376,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45580916423545564,
+      "grad_norm": 2.753182859305328,
+      "learning_rate": 1.190590740644672e-05,
+      "loss": 1.6349,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4559807791316286,
+      "grad_norm": 3.0041484309612083,
+      "learning_rate": 1.1900450626197987e-05,
+      "loss": 1.7227,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4561523940278016,
+      "grad_norm": 2.7284218874273454,
+      "learning_rate": 1.1894993258795258e-05,
+      "loss": 1.7016,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4563240089239746,
+      "grad_norm": 3.317737226332948,
+      "learning_rate": 1.1889535305924619e-05,
+      "loss": 1.821,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4564956238201476,
+      "grad_norm": 3.025170055300962,
+      "learning_rate": 1.188407676927233e-05,
+      "loss": 1.5479,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4566672387163206,
+      "grad_norm": 2.777247628260142,
+      "learning_rate": 1.1878617650524832e-05,
+      "loss": 1.7292,
+      "step": 2661
+    },
+    {
+      "epoch": 0.45683885361249355,
+      "grad_norm": 3.3885661872740664,
+      "learning_rate": 1.1873157951368757e-05,
+      "loss": 1.7317,
+      "step": 2662
+    },
+    {
+      "epoch": 0.45701046850866656,
+      "grad_norm": 2.6438284387144746,
+      "learning_rate": 1.1867697673490897e-05,
+      "loss": 1.7308,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45718208340483957,
+      "grad_norm": 3.507458287421913,
+      "learning_rate": 1.1862236818578239e-05,
+      "loss": 1.598,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4573536983010125,
+      "grad_norm": 2.910386059235125,
+      "learning_rate": 1.1856775388317936e-05,
+      "loss": 1.7031,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45752531319718553,
+      "grad_norm": 2.9085010167200527,
+      "learning_rate": 1.1851313384397334e-05,
+      "loss": 1.6195,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4576969280933585,
+      "grad_norm": 3.1498527021788005,
+      "learning_rate": 1.1845850808503939e-05,
+      "loss": 1.7393,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4578685429895315,
+      "grad_norm": 3.2923162633083547,
+      "learning_rate": 1.1840387662325443e-05,
+      "loss": 1.7012,
+      "step": 2668
+    },
+    {
+      "epoch": 0.45804015788570446,
+      "grad_norm": 2.9412764936047098,
+      "learning_rate": 1.183492394754972e-05,
+      "loss": 1.6338,
+      "step": 2669
+    },
+    {
+      "epoch": 0.45821177278187747,
+      "grad_norm": 3.1298385869917182,
+      "learning_rate": 1.1829459665864806e-05,
+      "loss": 1.7253,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4583833876780505,
+      "grad_norm": 3.351910431765254,
+      "learning_rate": 1.1823994818958922e-05,
+      "loss": 1.6607,
+      "step": 2671
+    },
+    {
+      "epoch": 0.45855500257422344,
+      "grad_norm": 2.8272532550704863,
+      "learning_rate": 1.1818529408520458e-05,
+      "loss": 1.691,
+      "step": 2672
+    },
+    {
+      "epoch": 0.45872661747039645,
+      "grad_norm": 2.7253444714846187,
+      "learning_rate": 1.1813063436237981e-05,
+      "loss": 1.5327,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4588982323665694,
+      "grad_norm": 2.669128107949888,
+      "learning_rate": 1.1807596903800234e-05,
+      "loss": 1.6484,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4590698472627424,
+      "grad_norm": 2.777250932196786,
+      "learning_rate": 1.1802129812896128e-05,
+      "loss": 1.5595,
+      "step": 2675
+    },
+    {
+      "epoch": 0.45924146215891537,
+      "grad_norm": 2.827236054270125,
+      "learning_rate": 1.1796662165214752e-05,
+      "loss": 1.5394,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4594130770550884,
+      "grad_norm": 4.2681465656367195,
+      "learning_rate": 1.1791193962445358e-05,
+      "loss": 1.7712,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4595846919512614,
+      "grad_norm": 2.963496890023793,
+      "learning_rate": 1.178572520627738e-05,
+      "loss": 1.851,
+      "step": 2678
+    },
+    {
+      "epoch": 0.45975630684743435,
+      "grad_norm": 3.074764507272029,
+      "learning_rate": 1.1780255898400417e-05,
+      "loss": 1.696,
+      "step": 2679
+    },
+    {
+      "epoch": 0.45992792174360736,
+      "grad_norm": 3.3183106790862897,
+      "learning_rate": 1.1774786040504238e-05,
+      "loss": 1.7456,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4600995366397803,
+      "grad_norm": 2.7424655175457757,
+      "learning_rate": 1.1769315634278791e-05,
+      "loss": 1.4808,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4602711515359533,
+      "grad_norm": 2.823533839590037,
+      "learning_rate": 1.1763844681414177e-05,
+      "loss": 1.7836,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4604427664321263,
+      "grad_norm": 3.590904953923604,
+      "learning_rate": 1.1758373183600678e-05,
+      "loss": 1.7765,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4606143813282993,
+      "grad_norm": 7.192544980367173,
+      "learning_rate": 1.1752901142528743e-05,
+      "loss": 1.7518,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4607859962244723,
+      "grad_norm": 7.8031651625040785,
+      "learning_rate": 1.1747428559888985e-05,
+      "loss": 1.7665,
+      "step": 2685
+    },
+    {
+      "epoch": 0.46095761112064526,
+      "grad_norm": 3.340410247520449,
+      "learning_rate": 1.174195543737219e-05,
+      "loss": 1.7036,
+      "step": 2686
+    },
+    {
+      "epoch": 0.46112922601681827,
+      "grad_norm": 3.3075338998799113,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 1.6083,
+      "step": 2687
+    },
+    {
+      "epoch": 0.46130084091299123,
+      "grad_norm": 3.8067125939464765,
+      "learning_rate": 1.1731007579471445e-05,
+      "loss": 1.8282,
+      "step": 2688
+    },
+    {
+      "epoch": 0.46147245580916424,
+      "grad_norm": 5.399574711714645,
+      "learning_rate": 1.1725532847469892e-05,
+      "loss": 1.4582,
+      "step": 2689
+    },
+    {
+      "epoch": 0.46164407070533725,
+      "grad_norm": 2.9299575208365343,
+      "learning_rate": 1.1720057582356096e-05,
+      "loss": 1.782,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4618156856015102,
+      "grad_norm": 4.5794181157037785,
+      "learning_rate": 1.1714581785821662e-05,
+      "loss": 1.7492,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4619873004976832,
+      "grad_norm": 2.8746188121776295,
+      "learning_rate": 1.1709105459558367e-05,
+      "loss": 1.6018,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4621589153938562,
+      "grad_norm": 2.5910494670148787,
+      "learning_rate": 1.1703628605258157e-05,
+      "loss": 1.5519,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4623305302900292,
+      "grad_norm": 2.765781281656966,
+      "learning_rate": 1.1698151224613126e-05,
+      "loss": 1.6565,
+      "step": 2694
+    },
+    {
+      "epoch": 0.46250214518620214,
+      "grad_norm": 3.258243752870418,
+      "learning_rate": 1.1692673319315541e-05,
+      "loss": 1.7346,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46267376008237515,
+      "grad_norm": 2.659443629422332,
+      "learning_rate": 1.1687194891057834e-05,
+      "loss": 1.6452,
+      "step": 2696
+    },
+    {
+      "epoch": 0.46284537497854816,
+      "grad_norm": 2.829519925969302,
+      "learning_rate": 1.1681715941532583e-05,
+      "loss": 1.7583,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4630169898747211,
+      "grad_norm": 2.926030715759295,
+      "learning_rate": 1.1676236472432549e-05,
+      "loss": 1.5648,
+      "step": 2698
+    },
+    {
+      "epoch": 0.46318860477089413,
+      "grad_norm": 3.6639245475998012,
+      "learning_rate": 1.1670756485450638e-05,
+      "loss": 1.6513,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4633602196670671,
+      "grad_norm": 2.408512108128582,
+      "learning_rate": 1.1665275982279918e-05,
+      "loss": 1.403,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4635318345632401,
+      "grad_norm": 2.763291247900218,
+      "learning_rate": 1.1659794964613619e-05,
+      "loss": 1.4791,
+      "step": 2701
+    },
+    {
+      "epoch": 0.46370344945941305,
+      "grad_norm": 2.682938418598613,
+      "learning_rate": 1.1654313434145133e-05,
+      "loss": 1.7929,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46387506435558606,
+      "grad_norm": 3.0475204545208685,
+      "learning_rate": 1.1648831392568005e-05,
+      "loss": 1.5719,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4640466792517591,
+      "grad_norm": 3.3354148124435863,
+      "learning_rate": 1.1643348841575936e-05,
+      "loss": 1.7212,
+      "step": 2704
+    },
+    {
+      "epoch": 0.46421829414793203,
+      "grad_norm": 2.757606244850724,
+      "learning_rate": 1.1637865782862799e-05,
+      "loss": 1.6638,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46438990904410504,
+      "grad_norm": 3.342908392786111,
+      "learning_rate": 1.16323822181226e-05,
+      "loss": 1.7582,
+      "step": 2706
+    },
+    {
+      "epoch": 0.464561523940278,
+      "grad_norm": 2.443113216571998,
+      "learning_rate": 1.1626898149049523e-05,
+      "loss": 1.5732,
+      "step": 2707
+    },
+    {
+      "epoch": 0.464733138836451,
+      "grad_norm": 3.0684047175990234,
+      "learning_rate": 1.1621413577337897e-05,
+      "loss": 1.5908,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46490475373262397,
+      "grad_norm": 3.1924423834773026,
+      "learning_rate": 1.1615928504682208e-05,
+      "loss": 1.8335,
+      "step": 2709
+    },
+    {
+      "epoch": 0.465076368628797,
+      "grad_norm": 3.8904689495166744,
+      "learning_rate": 1.1610442932777098e-05,
+      "loss": 1.9934,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46524798352497,
+      "grad_norm": 3.089336450071173,
+      "learning_rate": 1.1604956863317365e-05,
+      "loss": 1.6054,
+      "step": 2711
+    },
+    {
+      "epoch": 0.46541959842114294,
+      "grad_norm": 3.2967036052136587,
+      "learning_rate": 1.1599470297997955e-05,
+      "loss": 1.5544,
+      "step": 2712
+    },
+    {
+      "epoch": 0.46559121331731596,
+      "grad_norm": 2.5759711359977615,
+      "learning_rate": 1.1593983238513971e-05,
+      "loss": 1.4312,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4657628282134889,
+      "grad_norm": 2.822059112800912,
+      "learning_rate": 1.1588495686560667e-05,
+      "loss": 1.5536,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4659344431096619,
+      "grad_norm": 2.6844670507249067,
+      "learning_rate": 1.1583007643833456e-05,
+      "loss": 1.6419,
+      "step": 2715
+    },
+    {
+      "epoch": 0.46610605800583493,
+      "grad_norm": 3.257506718681694,
+      "learning_rate": 1.157751911202789e-05,
+      "loss": 1.7382,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4662776729020079,
+      "grad_norm": 3.2333817892731873,
+      "learning_rate": 1.157203009283969e-05,
+      "loss": 1.6799,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4664492877981809,
+      "grad_norm": 2.974785354171492,
+      "learning_rate": 1.1566540587964701e-05,
+      "loss": 1.7748,
+      "step": 2718
+    },
+    {
+      "epoch": 0.46662090269435386,
+      "grad_norm": 3.3615610331804047,
+      "learning_rate": 1.1561050599098946e-05,
+      "loss": 1.585,
+      "step": 2719
+    },
+    {
+      "epoch": 0.46679251759052687,
+      "grad_norm": 3.4648143164697305,
+      "learning_rate": 1.1555560127938581e-05,
+      "loss": 1.6071,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4669641324866998,
+      "grad_norm": 3.0709026539349082,
+      "learning_rate": 1.1550069176179912e-05,
+      "loss": 2.0446,
+      "step": 2721
+    },
+    {
+      "epoch": 0.46713574738287283,
+      "grad_norm": 2.924799060020064,
+      "learning_rate": 1.1544577745519404e-05,
+      "loss": 1.5837,
+      "step": 2722
+    },
+    {
+      "epoch": 0.46730736227904585,
+      "grad_norm": 2.8733459227228977,
+      "learning_rate": 1.1539085837653658e-05,
+      "loss": 1.5945,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4674789771752188,
+      "grad_norm": 2.957541030174589,
+      "learning_rate": 1.153359345427943e-05,
+      "loss": 1.6443,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4676505920713918,
+      "grad_norm": 4.200287007029637,
+      "learning_rate": 1.1528100597093617e-05,
+      "loss": 1.7809,
+      "step": 2725
+    },
+    {
+      "epoch": 0.46782220696756477,
+      "grad_norm": 3.0665873185535135,
+      "learning_rate": 1.1522607267793267e-05,
+      "loss": 1.8688,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4679938218637378,
+      "grad_norm": 2.649840227315986,
+      "learning_rate": 1.1517113468075571e-05,
+      "loss": 1.5216,
+      "step": 2727
+    },
+    {
+      "epoch": 0.46816543675991074,
+      "grad_norm": 3.1182687055183287,
+      "learning_rate": 1.1511619199637867e-05,
+      "loss": 1.801,
+      "step": 2728
+    },
+    {
+      "epoch": 0.46833705165608375,
+      "grad_norm": 3.722326583440218,
+      "learning_rate": 1.1506124464177644e-05,
+      "loss": 1.7242,
+      "step": 2729
+    },
+    {
+      "epoch": 0.46850866655225676,
+      "grad_norm": 3.418809906180343,
+      "learning_rate": 1.1500629263392515e-05,
+      "loss": 1.5194,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4686802814484297,
+      "grad_norm": 2.8030386222127626,
+      "learning_rate": 1.1495133598980263e-05,
+      "loss": 1.669,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4688518963446027,
+      "grad_norm": 2.8254178012646873,
+      "learning_rate": 1.14896374726388e-05,
+      "loss": 1.5963,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4690235112407757,
+      "grad_norm": 3.268378773235964,
+      "learning_rate": 1.1484140886066175e-05,
+      "loss": 1.7468,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4691951261369487,
+      "grad_norm": 3.2343858464434643,
+      "learning_rate": 1.1478643840960593e-05,
+      "loss": 1.5206,
+      "step": 2734
+    },
+    {
+      "epoch": 0.46936674103312165,
+      "grad_norm": 3.101633034688997,
+      "learning_rate": 1.1473146339020395e-05,
+      "loss": 1.7645,
+      "step": 2735
+    },
+    {
+      "epoch": 0.46953835592929466,
+      "grad_norm": 3.261941115428923,
+      "learning_rate": 1.1467648381944059e-05,
+      "loss": 1.5458,
+      "step": 2736
+    },
+    {
+      "epoch": 0.46970997082546767,
+      "grad_norm": 3.2000324926959536,
+      "learning_rate": 1.1462149971430207e-05,
+      "loss": 1.6502,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4698815857216406,
+      "grad_norm": 2.683811909228767,
+      "learning_rate": 1.1456651109177606e-05,
+      "loss": 1.785,
+      "step": 2738
+    },
+    {
+      "epoch": 0.47005320061781364,
+      "grad_norm": 3.473137594881375,
+      "learning_rate": 1.1451151796885152e-05,
+      "loss": 1.9541,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4702248155139866,
+      "grad_norm": 3.89951515926513,
+      "learning_rate": 1.1445652036251889e-05,
+      "loss": 1.7742,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4703964304101596,
+      "grad_norm": 2.8914918340040403,
+      "learning_rate": 1.1440151828976997e-05,
+      "loss": 1.6492,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4705680453063326,
+      "grad_norm": 3.8691061908734024,
+      "learning_rate": 1.1434651176759788e-05,
+      "loss": 1.5376,
+      "step": 2742
+    },
+    {
+      "epoch": 0.47073966020250557,
+      "grad_norm": 2.554098204939939,
+      "learning_rate": 1.1429150081299725e-05,
+      "loss": 1.4747,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4709112750986786,
+      "grad_norm": 2.9892517455549465,
+      "learning_rate": 1.1423648544296396e-05,
+      "loss": 1.6242,
+      "step": 2744
+    },
+    {
+      "epoch": 0.47108288999485154,
+      "grad_norm": 4.264472527148404,
+      "learning_rate": 1.1418146567449524e-05,
+      "loss": 1.5582,
+      "step": 2745
+    },
+    {
+      "epoch": 0.47125450489102455,
+      "grad_norm": 2.5160757051279337,
+      "learning_rate": 1.141264415245898e-05,
+      "loss": 1.4321,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4714261197871975,
+      "grad_norm": 3.6144603903357413,
+      "learning_rate": 1.1407141301024762e-05,
+      "loss": 1.8127,
+      "step": 2747
+    },
+    {
+      "epoch": 0.4715977346833705,
+      "grad_norm": 3.50259899934638,
+      "learning_rate": 1.1401638014847004e-05,
+      "loss": 1.6851,
+      "step": 2748
+    },
+    {
+      "epoch": 0.47176934957954353,
+      "grad_norm": 3.268806407081182,
+      "learning_rate": 1.1396134295625971e-05,
+      "loss": 1.7497,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4719409644757165,
+      "grad_norm": 2.6039966305613738,
+      "learning_rate": 1.1390630145062069e-05,
+      "loss": 1.4863,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4721125793718895,
+      "grad_norm": 3.18163953916345,
+      "learning_rate": 1.1385125564855834e-05,
+      "loss": 1.6147,
+      "step": 2751
+    },
+    {
+      "epoch": 0.47228419426806245,
+      "grad_norm": 2.5101807829812333,
+      "learning_rate": 1.1379620556707934e-05,
+      "loss": 1.4585,
+      "step": 2752
+    },
+    {
+      "epoch": 0.47245580916423546,
+      "grad_norm": 3.116412550758941,
+      "learning_rate": 1.1374115122319168e-05,
+      "loss": 1.837,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4726274240604084,
+      "grad_norm": 2.6387781732206705,
+      "learning_rate": 1.136860926339047e-05,
+      "loss": 1.4391,
+      "step": 2754
+    },
+    {
+      "epoch": 0.47279903895658143,
+      "grad_norm": 3.271791696632264,
+      "learning_rate": 1.13631029816229e-05,
+      "loss": 1.7422,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47297065385275444,
+      "grad_norm": 3.0034510366163345,
+      "learning_rate": 1.1357596278717655e-05,
+      "loss": 1.7334,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4731422687489274,
+      "grad_norm": 6.813218808858523,
+      "learning_rate": 1.1352089156376059e-05,
+      "loss": 1.7685,
+      "step": 2757
+    },
+    {
+      "epoch": 0.4733138836451004,
+      "grad_norm": 3.3620016308789182,
+      "learning_rate": 1.1346581616299565e-05,
+      "loss": 1.648,
+      "step": 2758
+    },
+    {
+      "epoch": 0.47348549854127336,
+      "grad_norm": 3.2204626114586055,
+      "learning_rate": 1.1341073660189761e-05,
+      "loss": 1.5946,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4736571134374464,
+      "grad_norm": 3.23109584682806,
+      "learning_rate": 1.133556528974835e-05,
+      "loss": 1.4936,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47382872833361933,
+      "grad_norm": 3.04568311245728,
+      "learning_rate": 1.1330056506677174e-05,
+      "loss": 1.605,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47400034322979234,
+      "grad_norm": 3.4639504586409586,
+      "learning_rate": 1.1324547312678203e-05,
+      "loss": 1.9035,
+      "step": 2762
+    },
+    {
+      "epoch": 0.47417195812596535,
+      "grad_norm": 4.350620942665327,
+      "learning_rate": 1.1319037709453527e-05,
+      "loss": 1.523,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4743435730221383,
+      "grad_norm": 3.2043148680157008,
+      "learning_rate": 1.1313527698705367e-05,
+      "loss": 1.4964,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4745151879183113,
+      "grad_norm": 3.3011732397718747,
+      "learning_rate": 1.1308017282136074e-05,
+      "loss": 1.7793,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4746868028144843,
+      "grad_norm": 2.823107797702438,
+      "learning_rate": 1.1302506461448112e-05,
+      "loss": 1.5238,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4748584177106573,
+      "grad_norm": 3.1758629735880897,
+      "learning_rate": 1.1296995238344084e-05,
+      "loss": 1.6741,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4750300326068303,
+      "grad_norm": 3.4470923615266544,
+      "learning_rate": 1.1291483614526707e-05,
+      "loss": 1.7997,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47520164750300325,
+      "grad_norm": 4.848256642634677,
+      "learning_rate": 1.1285971591698825e-05,
+      "loss": 1.6794,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47537326239917627,
+      "grad_norm": 3.325807318614395,
+      "learning_rate": 1.1280459171563409e-05,
+      "loss": 1.6517,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4755448772953492,
+      "grad_norm": 2.4559976190038086,
+      "learning_rate": 1.1274946355823544e-05,
+      "loss": 1.519,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47571649219152223,
+      "grad_norm": 3.0289664975297685,
+      "learning_rate": 1.1269433146182452e-05,
+      "loss": 1.6459,
+      "step": 2772
+    },
+    {
+      "epoch": 0.4758881070876952,
+      "grad_norm": 2.744270545965551,
+      "learning_rate": 1.1263919544343458e-05,
+      "loss": 1.4744,
+      "step": 2773
+    },
+    {
+      "epoch": 0.4760597219838682,
+      "grad_norm": 3.400099484689994,
+      "learning_rate": 1.1258405552010024e-05,
+      "loss": 1.6912,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4762313368800412,
+      "grad_norm": 2.626946394496932,
+      "learning_rate": 1.125289117088572e-05,
+      "loss": 1.5595,
+      "step": 2775
+    },
+    {
+      "epoch": 0.47640295177621417,
+      "grad_norm": 2.5841259694104064,
+      "learning_rate": 1.1247376402674248e-05,
+      "loss": 1.7375,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4765745666723872,
+      "grad_norm": 3.1479848933931387,
+      "learning_rate": 1.1241861249079421e-05,
+      "loss": 1.6634,
+      "step": 2777
+    },
+    {
+      "epoch": 0.47674618156856013,
+      "grad_norm": 2.898102679925583,
+      "learning_rate": 1.1236345711805178e-05,
+      "loss": 1.553,
+      "step": 2778
+    },
+    {
+      "epoch": 0.47691779646473315,
+      "grad_norm": 3.0310873410122596,
+      "learning_rate": 1.1230829792555571e-05,
+      "loss": 1.92,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4770894113609061,
+      "grad_norm": 3.407614338605185,
+      "learning_rate": 1.1225313493034766e-05,
+      "loss": 1.9593,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4772610262570791,
+      "grad_norm": 3.1261188715414754,
+      "learning_rate": 1.121979681494706e-05,
+      "loss": 1.7933,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4774326411532521,
+      "grad_norm": 3.4222389957318073,
+      "learning_rate": 1.1214279759996856e-05,
+      "loss": 1.5977,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4776042560494251,
+      "grad_norm": 2.918431014030688,
+      "learning_rate": 1.1208762329888672e-05,
+      "loss": 1.676,
+      "step": 2783
+    },
+    {
+      "epoch": 0.4777758709455981,
+      "grad_norm": 3.661972980629568,
+      "learning_rate": 1.1203244526327154e-05,
+      "loss": 1.8453,
+      "step": 2784
+    },
+    {
+      "epoch": 0.47794748584177105,
+      "grad_norm": 2.916482627882997,
+      "learning_rate": 1.1197726351017052e-05,
+      "loss": 1.5532,
+      "step": 2785
+    },
+    {
+      "epoch": 0.47811910073794406,
+      "grad_norm": 2.925320552586996,
+      "learning_rate": 1.1192207805663235e-05,
+      "loss": 1.6373,
+      "step": 2786
+    },
+    {
+      "epoch": 0.478290715634117,
+      "grad_norm": 4.004958318213895,
+      "learning_rate": 1.1186688891970686e-05,
+      "loss": 1.8329,
+      "step": 2787
+    },
+    {
+      "epoch": 0.47846233053029,
+      "grad_norm": 3.3370594836402296,
+      "learning_rate": 1.11811696116445e-05,
+      "loss": 1.7443,
+      "step": 2788
+    },
+    {
+      "epoch": 0.47863394542646304,
+      "grad_norm": 4.24103521437125,
+      "learning_rate": 1.117564996638989e-05,
+      "loss": 1.6517,
+      "step": 2789
+    },
+    {
+      "epoch": 0.478805560322636,
+      "grad_norm": 3.502691871151845,
+      "learning_rate": 1.1170129957912178e-05,
+      "loss": 1.748,
+      "step": 2790
+    },
+    {
+      "epoch": 0.478977175218809,
+      "grad_norm": 2.8419903616400917,
+      "learning_rate": 1.11646095879168e-05,
+      "loss": 1.2875,
+      "step": 2791
+    },
+    {
+      "epoch": 0.47914879011498196,
+      "grad_norm": 3.1084119414899987,
+      "learning_rate": 1.1159088858109297e-05,
+      "loss": 1.5118,
+      "step": 2792
+    },
+    {
+      "epoch": 0.47932040501115497,
+      "grad_norm": 3.2895469202610963,
+      "learning_rate": 1.1153567770195333e-05,
+      "loss": 1.5436,
+      "step": 2793
+    },
+    {
+      "epoch": 0.479492019907328,
+      "grad_norm": 2.9641494357348446,
+      "learning_rate": 1.1148046325880675e-05,
+      "loss": 1.568,
+      "step": 2794
+    },
+    {
+      "epoch": 0.47966363480350094,
+      "grad_norm": 3.1339999930629623,
+      "learning_rate": 1.1142524526871201e-05,
+      "loss": 1.6977,
+      "step": 2795
+    },
+    {
+      "epoch": 0.47983524969967395,
+      "grad_norm": 3.1014251288204995,
+      "learning_rate": 1.11370023748729e-05,
+      "loss": 1.4696,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4800068645958469,
+      "grad_norm": 2.839720240833112,
+      "learning_rate": 1.1131479871591869e-05,
+      "loss": 1.6802,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4801784794920199,
+      "grad_norm": 3.521269095352219,
+      "learning_rate": 1.1125957018734313e-05,
+      "loss": 1.637,
+      "step": 2798
+    },
+    {
+      "epoch": 0.48035009438819287,
+      "grad_norm": 2.81676392013561,
+      "learning_rate": 1.1120433818006546e-05,
+      "loss": 1.7803,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4805217092843659,
+      "grad_norm": 3.301384589818475,
+      "learning_rate": 1.1114910271114987e-05,
+      "loss": 1.6878,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4806933241805389,
+      "grad_norm": 2.9090433705044947,
+      "learning_rate": 1.1109386379766169e-05,
+      "loss": 1.5992,
+      "step": 2801
+    },
+    {
+      "epoch": 0.48086493907671185,
+      "grad_norm": 2.616402299100735,
+      "learning_rate": 1.1103862145666726e-05,
+      "loss": 1.4713,
+      "step": 2802
+    },
+    {
+      "epoch": 0.48103655397288486,
+      "grad_norm": 3.5448041579770435,
+      "learning_rate": 1.1098337570523397e-05,
+      "loss": 1.7569,
+      "step": 2803
+    },
+    {
+      "epoch": 0.4812081688690578,
+      "grad_norm": 4.259451430634504,
+      "learning_rate": 1.1092812656043024e-05,
+      "loss": 1.8535,
+      "step": 2804
+    },
+    {
+      "epoch": 0.48137978376523083,
+      "grad_norm": 2.9532808526512744,
+      "learning_rate": 1.1087287403932563e-05,
+      "loss": 1.4457,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4815513986614038,
+      "grad_norm": 2.9157893704847395,
+      "learning_rate": 1.108176181589907e-05,
+      "loss": 1.7901,
+      "step": 2806
+    },
+    {
+      "epoch": 0.4817230135575768,
+      "grad_norm": 2.894146662843656,
+      "learning_rate": 1.1076235893649697e-05,
+      "loss": 1.6124,
+      "step": 2807
+    },
+    {
+      "epoch": 0.4818946284537498,
+      "grad_norm": 3.7018813569207785,
+      "learning_rate": 1.1070709638891718e-05,
+      "loss": 1.5624,
+      "step": 2808
+    },
+    {
+      "epoch": 0.48206624334992276,
+      "grad_norm": 4.5730868751411915,
+      "learning_rate": 1.1065183053332481e-05,
+      "loss": 1.6043,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4822378582460958,
+      "grad_norm": 2.80262190571527,
+      "learning_rate": 1.1059656138679468e-05,
+      "loss": 1.664,
+      "step": 2810
+    },
+    {
+      "epoch": 0.48240947314226873,
+      "grad_norm": 2.873055551575474,
+      "learning_rate": 1.1054128896640241e-05,
+      "loss": 1.7196,
+      "step": 2811
+    },
+    {
+      "epoch": 0.48258108803844174,
+      "grad_norm": 2.7503699672206543,
+      "learning_rate": 1.104860132892247e-05,
+      "loss": 1.4872,
+      "step": 2812
+    },
+    {
+      "epoch": 0.4827527029346147,
+      "grad_norm": 3.0155279658603624,
+      "learning_rate": 1.1043073437233927e-05,
+      "loss": 1.8158,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4829243178307877,
+      "grad_norm": 2.7717221391124522,
+      "learning_rate": 1.1037545223282481e-05,
+      "loss": 1.518,
+      "step": 2814
+    },
+    {
+      "epoch": 0.4830959327269607,
+      "grad_norm": 4.218176163062874,
+      "learning_rate": 1.1032016688776106e-05,
+      "loss": 1.6093,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4832675476231337,
+      "grad_norm": 3.482201585177803,
+      "learning_rate": 1.1026487835422866e-05,
+      "loss": 1.7539,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4834391625193067,
+      "grad_norm": 2.989619157050514,
+      "learning_rate": 1.1020958664930932e-05,
+      "loss": 1.9388,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48361077741547964,
+      "grad_norm": 2.669275749442373,
+      "learning_rate": 1.1015429179008567e-05,
+      "loss": 1.6032,
+      "step": 2818
+    },
+    {
+      "epoch": 0.48378239231165265,
+      "grad_norm": 3.621288304107407,
+      "learning_rate": 1.1009899379364136e-05,
+      "loss": 1.7854,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48395400720782566,
+      "grad_norm": 3.704668257147288,
+      "learning_rate": 1.10043692677061e-05,
+      "loss": 1.6491,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4841256221039986,
+      "grad_norm": 2.9470841067166584,
+      "learning_rate": 1.0998838845743012e-05,
+      "loss": 1.5437,
+      "step": 2821
+    },
+    {
+      "epoch": 0.48429723700017163,
+      "grad_norm": 3.2612092655267655,
+      "learning_rate": 1.099330811518353e-05,
+      "loss": 1.6303,
+      "step": 2822
+    },
+    {
+      "epoch": 0.4844688518963446,
+      "grad_norm": 2.9523898895713243,
+      "learning_rate": 1.0987777077736395e-05,
+      "loss": 1.7984,
+      "step": 2823
+    },
+    {
+      "epoch": 0.4846404667925176,
+      "grad_norm": 3.3671597516416747,
+      "learning_rate": 1.0982245735110453e-05,
+      "loss": 1.6399,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48481208168869055,
+      "grad_norm": 3.5269472296949957,
+      "learning_rate": 1.097671408901464e-05,
+      "loss": 1.725,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48498369658486357,
+      "grad_norm": 3.2127318967725658,
+      "learning_rate": 1.0971182141157988e-05,
+      "loss": 1.4497,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4851553114810366,
+      "grad_norm": 2.920910921567526,
+      "learning_rate": 1.0965649893249619e-05,
+      "loss": 1.6653,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48532692637720953,
+      "grad_norm": 2.6731906276150097,
+      "learning_rate": 1.0960117346998748e-05,
+      "loss": 1.6553,
+      "step": 2828
+    },
+    {
+      "epoch": 0.48549854127338254,
+      "grad_norm": 3.689974284079219,
+      "learning_rate": 1.0954584504114689e-05,
+      "loss": 1.7421,
+      "step": 2829
+    },
+    {
+      "epoch": 0.4856701561695555,
+      "grad_norm": 2.924402660432866,
+      "learning_rate": 1.0949051366306838e-05,
+      "loss": 1.7086,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4858417710657285,
+      "grad_norm": 3.734333067273493,
+      "learning_rate": 1.0943517935284684e-05,
+      "loss": 1.7158,
+      "step": 2831
+    },
+    {
+      "epoch": 0.48601338596190147,
+      "grad_norm": 2.8206577575119285,
+      "learning_rate": 1.093798421275782e-05,
+      "loss": 1.5716,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4861850008580745,
+      "grad_norm": 3.027368031065995,
+      "learning_rate": 1.0932450200435902e-05,
+      "loss": 1.5311,
+      "step": 2833
+    },
+    {
+      "epoch": 0.4863566157542475,
+      "grad_norm": 3.086671055753802,
+      "learning_rate": 1.0926915900028703e-05,
+      "loss": 1.6866,
+      "step": 2834
+    },
+    {
+      "epoch": 0.48652823065042045,
+      "grad_norm": 4.204208530057646,
+      "learning_rate": 1.092138131324607e-05,
+      "loss": 1.9312,
+      "step": 2835
+    },
+    {
+      "epoch": 0.48669984554659346,
+      "grad_norm": 2.63895075432221,
+      "learning_rate": 1.0915846441797942e-05,
+      "loss": 1.5832,
+      "step": 2836
+    },
+    {
+      "epoch": 0.4868714604427664,
+      "grad_norm": 2.9889238006434073,
+      "learning_rate": 1.0910311287394348e-05,
+      "loss": 1.634,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4870430753389394,
+      "grad_norm": 4.811711443546964,
+      "learning_rate": 1.0904775851745399e-05,
+      "loss": 1.5856,
+      "step": 2838
+    },
+    {
+      "epoch": 0.48721469023511244,
+      "grad_norm": 5.350197211889695,
+      "learning_rate": 1.08992401365613e-05,
+      "loss": 1.6645,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4873863051312854,
+      "grad_norm": 3.4733803399677106,
+      "learning_rate": 1.0893704143552335e-05,
+      "loss": 1.7122,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4875579200274584,
+      "grad_norm": 3.633986656472618,
+      "learning_rate": 1.0888167874428881e-05,
+      "loss": 1.7649,
+      "step": 2841
+    },
+    {
+      "epoch": 0.48772953492363136,
+      "grad_norm": 3.4331663012659224,
+      "learning_rate": 1.0882631330901394e-05,
+      "loss": 1.7527,
+      "step": 2842
+    },
+    {
+      "epoch": 0.48790114981980437,
+      "grad_norm": 3.2625999967690054,
+      "learning_rate": 1.0877094514680417e-05,
+      "loss": 1.4977,
+      "step": 2843
+    },
+    {
+      "epoch": 0.4880727647159773,
+      "grad_norm": 5.037079012270086,
+      "learning_rate": 1.0871557427476585e-05,
+      "loss": 1.4589,
+      "step": 2844
+    },
+    {
+      "epoch": 0.48824437961215034,
+      "grad_norm": 2.858837902994473,
+      "learning_rate": 1.0866020071000597e-05,
+      "loss": 1.7217,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48841599450832335,
+      "grad_norm": 2.685082105907214,
+      "learning_rate": 1.0860482446963262e-05,
+      "loss": 1.6504,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4885876094044963,
+      "grad_norm": 3.2473865101242474,
+      "learning_rate": 1.0854944557075447e-05,
+      "loss": 1.6097,
+      "step": 2847
+    },
+    {
+      "epoch": 0.4887592243006693,
+      "grad_norm": 2.7403191919040615,
+      "learning_rate": 1.084940640304811e-05,
+      "loss": 1.8809,
+      "step": 2848
+    },
+    {
+      "epoch": 0.48893083919684227,
+      "grad_norm": 2.465539444705478,
+      "learning_rate": 1.0843867986592302e-05,
+      "loss": 1.4832,
+      "step": 2849
+    },
+    {
+      "epoch": 0.4891024540930153,
+      "grad_norm": 2.815987875335105,
+      "learning_rate": 1.0838329309419142e-05,
+      "loss": 1.3072,
+      "step": 2850
+    },
+    {
+      "epoch": 0.48927406898918824,
+      "grad_norm": 3.031167819980559,
+      "learning_rate": 1.083279037323983e-05,
+      "loss": 1.5897,
+      "step": 2851
+    },
+    {
+      "epoch": 0.48944568388536125,
+      "grad_norm": 2.892950854750241,
+      "learning_rate": 1.0827251179765645e-05,
+      "loss": 1.6187,
+      "step": 2852
+    },
+    {
+      "epoch": 0.48961729878153426,
+      "grad_norm": 3.1684669236398952,
+      "learning_rate": 1.082171173070796e-05,
+      "loss": 1.9,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4897889136777072,
+      "grad_norm": 2.508664755082155,
+      "learning_rate": 1.0816172027778208e-05,
+      "loss": 1.4261,
+      "step": 2854
+    },
+    {
+      "epoch": 0.4899605285738802,
+      "grad_norm": 2.957792136495784,
+      "learning_rate": 1.0810632072687909e-05,
+      "loss": 1.8966,
+      "step": 2855
+    },
+    {
+      "epoch": 0.4901321434700532,
+      "grad_norm": 2.4862657280334433,
+      "learning_rate": 1.080509186714867e-05,
+      "loss": 1.5494,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4903037583662262,
+      "grad_norm": 2.8836217406058227,
+      "learning_rate": 1.0799551412872151e-05,
+      "loss": 1.9226,
+      "step": 2857
+    },
+    {
+      "epoch": 0.49047537326239915,
+      "grad_norm": 7.2204784614255155,
+      "learning_rate": 1.0794010711570115e-05,
+      "loss": 1.624,
+      "step": 2858
+    },
+    {
+      "epoch": 0.49064698815857216,
+      "grad_norm": 3.3855069852818134,
+      "learning_rate": 1.0788469764954388e-05,
+      "loss": 1.5452,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4908186030547452,
+      "grad_norm": 3.2140446072472373,
+      "learning_rate": 1.0782928574736869e-05,
+      "loss": 1.7694,
+      "step": 2860
+    },
+    {
+      "epoch": 0.49099021795091813,
+      "grad_norm": 2.5770444220208186,
+      "learning_rate": 1.0777387142629543e-05,
+      "loss": 1.4273,
+      "step": 2861
+    },
+    {
+      "epoch": 0.49116183284709114,
+      "grad_norm": 2.63682652341496,
+      "learning_rate": 1.0771845470344462e-05,
+      "loss": 1.6189,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4913334477432641,
+      "grad_norm": 2.7074590769967743,
+      "learning_rate": 1.0766303559593754e-05,
+      "loss": 1.6293,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4915050626394371,
+      "grad_norm": 3.4775218455695356,
+      "learning_rate": 1.076076141208962e-05,
+      "loss": 1.7053,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4916766775356101,
+      "grad_norm": 3.1410804742670995,
+      "learning_rate": 1.0755219029544337e-05,
+      "loss": 1.8371,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4918482924317831,
+      "grad_norm": 2.726681343008733,
+      "learning_rate": 1.0749676413670252e-05,
+      "loss": 1.6201,
+      "step": 2866
+    },
+    {
+      "epoch": 0.4920199073279561,
+      "grad_norm": 3.0176802644912035,
+      "learning_rate": 1.074413356617978e-05,
+      "loss": 1.8257,
+      "step": 2867
+    },
+    {
+      "epoch": 0.49219152222412904,
+      "grad_norm": 2.6827995439150203,
+      "learning_rate": 1.073859048878542e-05,
+      "loss": 1.5066,
+      "step": 2868
+    },
+    {
+      "epoch": 0.49236313712030205,
+      "grad_norm": 3.2900236891010572,
+      "learning_rate": 1.0733047183199729e-05,
+      "loss": 1.7236,
+      "step": 2869
+    },
+    {
+      "epoch": 0.492534752016475,
+      "grad_norm": 3.0106033897329616,
+      "learning_rate": 1.0727503651135343e-05,
+      "loss": 1.669,
+      "step": 2870
+    },
+    {
+      "epoch": 0.492706366912648,
+      "grad_norm": 3.1550179081101475,
+      "learning_rate": 1.0721959894304963e-05,
+      "loss": 1.5022,
+      "step": 2871
+    },
+    {
+      "epoch": 0.49287798180882103,
+      "grad_norm": 2.81830665913476,
+      "learning_rate": 1.071641591442136e-05,
+      "loss": 1.6757,
+      "step": 2872
+    },
+    {
+      "epoch": 0.493049596704994,
+      "grad_norm": 6.1658539954755405,
+      "learning_rate": 1.0710871713197374e-05,
+      "loss": 1.8812,
+      "step": 2873
+    },
+    {
+      "epoch": 0.493221211601167,
+      "grad_norm": 2.8799999685540225,
+      "learning_rate": 1.0705327292345922e-05,
+      "loss": 1.8775,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49339282649733995,
+      "grad_norm": 2.9679981499183783,
+      "learning_rate": 1.0699782653579973e-05,
+      "loss": 1.8642,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49356444139351296,
+      "grad_norm": 3.2672909629974956,
+      "learning_rate": 1.0694237798612574e-05,
+      "loss": 1.3912,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4937360562896859,
+      "grad_norm": 3.11212245892745,
+      "learning_rate": 1.068869272915684e-05,
+      "loss": 1.6626,
+      "step": 2877
+    },
+    {
+      "epoch": 0.49390767118585893,
+      "grad_norm": 3.3996329143275315,
+      "learning_rate": 1.0683147446925943e-05,
+      "loss": 1.8161,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49407928608203194,
+      "grad_norm": 2.8148004124344173,
+      "learning_rate": 1.067760195363313e-05,
+      "loss": 1.3539,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4942509009782049,
+      "grad_norm": 3.3638402236750844,
+      "learning_rate": 1.0672056250991714e-05,
+      "loss": 1.7545,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4944225158743779,
+      "grad_norm": 3.221325042639084,
+      "learning_rate": 1.0666510340715057e-05,
+      "loss": 1.7475,
+      "step": 2881
+    },
+    {
+      "epoch": 0.49459413077055087,
+      "grad_norm": 3.763264646160915,
+      "learning_rate": 1.0660964224516606e-05,
+      "loss": 1.7146,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4947657456667239,
+      "grad_norm": 2.9305646212777456,
+      "learning_rate": 1.0655417904109859e-05,
+      "loss": 1.7116,
+      "step": 2883
+    },
+    {
+      "epoch": 0.49493736056289683,
+      "grad_norm": 3.9408742167039073,
+      "learning_rate": 1.0649871381208379e-05,
+      "loss": 1.5674,
+      "step": 2884
+    },
+    {
+      "epoch": 0.49510897545906984,
+      "grad_norm": 4.100534497247508,
+      "learning_rate": 1.0644324657525795e-05,
+      "loss": 1.9224,
+      "step": 2885
+    },
+    {
+      "epoch": 0.49528059035524286,
+      "grad_norm": 2.787121561972098,
+      "learning_rate": 1.0638777734775795e-05,
+      "loss": 1.7637,
+      "step": 2886
+    },
+    {
+      "epoch": 0.4954522052514158,
+      "grad_norm": 3.6700902910857347,
+      "learning_rate": 1.063323061467213e-05,
+      "loss": 1.6383,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4956238201475888,
+      "grad_norm": 2.963964690375311,
+      "learning_rate": 1.0627683298928607e-05,
+      "loss": 1.6797,
+      "step": 2888
+    },
+    {
+      "epoch": 0.4957954350437618,
+      "grad_norm": 3.545970293150631,
+      "learning_rate": 1.0622135789259105e-05,
+      "loss": 1.6165,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4959670499399348,
+      "grad_norm": 2.7483820087286284,
+      "learning_rate": 1.061658808737755e-05,
+      "loss": 1.6622,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4961386648361078,
+      "grad_norm": 3.1393123758316546,
+      "learning_rate": 1.0611040194997933e-05,
+      "loss": 1.6765,
+      "step": 2891
+    },
+    {
+      "epoch": 0.49631027973228076,
+      "grad_norm": 3.0805616736243433,
+      "learning_rate": 1.060549211383431e-05,
+      "loss": 1.4133,
+      "step": 2892
+    },
+    {
+      "epoch": 0.49648189462845377,
+      "grad_norm": 4.19817350974281,
+      "learning_rate": 1.0599943845600781e-05,
+      "loss": 1.8748,
+      "step": 2893
+    },
+    {
+      "epoch": 0.4966535095246267,
+      "grad_norm": 5.320784312248385,
+      "learning_rate": 1.0594395392011518e-05,
+      "loss": 1.7341,
+      "step": 2894
+    },
+    {
+      "epoch": 0.49682512442079974,
+      "grad_norm": 3.4793049245072902,
+      "learning_rate": 1.0588846754780744e-05,
+      "loss": 1.5339,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4969967393169727,
+      "grad_norm": 3.4858430134900047,
+      "learning_rate": 1.0583297935622732e-05,
+      "loss": 1.5074,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4971683542131457,
+      "grad_norm": 2.619043149279237,
+      "learning_rate": 1.0577748936251829e-05,
+      "loss": 1.6884,
+      "step": 2897
+    },
+    {
+      "epoch": 0.4973399691093187,
+      "grad_norm": 3.562472813773365,
+      "learning_rate": 1.0572199758382417e-05,
+      "loss": 1.5751,
+      "step": 2898
+    },
+    {
+      "epoch": 0.49751158400549167,
+      "grad_norm": 3.264365661674337,
+      "learning_rate": 1.056665040372895e-05,
+      "loss": 1.6157,
+      "step": 2899
+    },
+    {
+      "epoch": 0.4976831989016647,
+      "grad_norm": 3.2884989911815716,
+      "learning_rate": 1.0561100874005925e-05,
+      "loss": 1.5677,
+      "step": 2900
+    },
+    {
+      "epoch": 0.49785481379783764,
+      "grad_norm": 2.9259898359026795,
+      "learning_rate": 1.05555511709279e-05,
+      "loss": 1.644,
+      "step": 2901
+    },
+    {
+      "epoch": 0.49802642869401065,
+      "grad_norm": 3.2469976450023847,
+      "learning_rate": 1.0550001296209487e-05,
+      "loss": 1.6406,
+      "step": 2902
+    },
+    {
+      "epoch": 0.4981980435901836,
+      "grad_norm": 3.056174535833804,
+      "learning_rate": 1.054445125156534e-05,
+      "loss": 1.619,
+      "step": 2903
+    },
+    {
+      "epoch": 0.4983696584863566,
+      "grad_norm": 2.984069357007831,
+      "learning_rate": 1.0538901038710186e-05,
+      "loss": 1.4725,
+      "step": 2904
+    },
+    {
+      "epoch": 0.4985412733825296,
+      "grad_norm": 3.053950465963755,
+      "learning_rate": 1.0533350659358779e-05,
+      "loss": 1.5265,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4987128882787026,
+      "grad_norm": 3.4651192292977586,
+      "learning_rate": 1.0527800115225945e-05,
+      "loss": 1.6516,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4988845031748756,
+      "grad_norm": 3.057599146523825,
+      "learning_rate": 1.0522249408026553e-05,
+      "loss": 1.6218,
+      "step": 2907
+    },
+    {
+      "epoch": 0.49905611807104855,
+      "grad_norm": 3.129506370889882,
+      "learning_rate": 1.0516698539475518e-05,
+      "loss": 1.6617,
+      "step": 2908
+    },
+    {
+      "epoch": 0.49922773296722156,
+      "grad_norm": 6.273617084517068,
+      "learning_rate": 1.0511147511287813e-05,
+      "loss": 1.7429,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4993993478633945,
+      "grad_norm": 14.76701678271327,
+      "learning_rate": 1.0505596325178458e-05,
+      "loss": 1.8852,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4995709627595675,
+      "grad_norm": 3.211466359060046,
+      "learning_rate": 1.0500044982862519e-05,
+      "loss": 1.7959,
+      "step": 2911
+    },
+    {
+      "epoch": 0.49974257765574054,
+      "grad_norm": 3.223250055081025,
+      "learning_rate": 1.0494493486055108e-05,
+      "loss": 1.5882,
+      "step": 2912
+    },
+    {
+      "epoch": 0.4999141925519135,
+      "grad_norm": 3.2228820051245752,
+      "learning_rate": 1.0488941836471393e-05,
+      "loss": 1.4302,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5000858074480865,
+      "grad_norm": 3.7748502855531703,
+      "learning_rate": 1.0483390035826586e-05,
+      "loss": 1.6462,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5002574223442595,
+      "grad_norm": 2.662677804872833,
+      "learning_rate": 1.0477838085835936e-05,
+      "loss": 1.5748,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5004290372404324,
+      "grad_norm": 3.3593050995196876,
+      "learning_rate": 1.047228598821476e-05,
+      "loss": 1.8182,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5006006521366054,
+      "grad_norm": 2.762102391538692,
+      "learning_rate": 1.0466733744678392e-05,
+      "loss": 1.6905,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5007722670327784,
+      "grad_norm": 3.354254556898484,
+      "learning_rate": 1.0461181356942239e-05,
+      "loss": 1.5207,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5009438819289515,
+      "grad_norm": 3.9023037760207897,
+      "learning_rate": 1.0455628826721733e-05,
+      "loss": 1.6755,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5011154968251245,
+      "grad_norm": 2.763682270620225,
+      "learning_rate": 1.0450076155732357e-05,
+      "loss": 1.6945,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5012871117212974,
+      "grad_norm": 3.2913572941270357,
+      "learning_rate": 1.0444523345689637e-05,
+      "loss": 1.5213,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5014587266174704,
+      "grad_norm": 4.571939576018699,
+      "learning_rate": 1.0438970398309145e-05,
+      "loss": 1.6143,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5016303415136434,
+      "grad_norm": 3.246076186091857,
+      "learning_rate": 1.0433417315306494e-05,
+      "loss": 1.7915,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5018019564098164,
+      "grad_norm": 4.1703491706708,
+      "learning_rate": 1.042786409839733e-05,
+      "loss": 1.8054,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5019735713059894,
+      "grad_norm": 2.7798726698606573,
+      "learning_rate": 1.0422310749297355e-05,
+      "loss": 1.657,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5021451862021623,
+      "grad_norm": 3.277794573925155,
+      "learning_rate": 1.0416757269722301e-05,
+      "loss": 1.7684,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5023168010983353,
+      "grad_norm": 3.198920525110606,
+      "learning_rate": 1.0411203661387944e-05,
+      "loss": 1.796,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5024884159945083,
+      "grad_norm": 3.258831816950959,
+      "learning_rate": 1.0405649926010108e-05,
+      "loss": 1.4681,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5026600308906813,
+      "grad_norm": 3.440171045797029,
+      "learning_rate": 1.0400096065304637e-05,
+      "loss": 1.5857,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5028316457868544,
+      "grad_norm": 3.4265772411020223,
+      "learning_rate": 1.0394542080987429e-05,
+      "loss": 1.6542,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5030032606830273,
+      "grad_norm": 2.975751179522291,
+      "learning_rate": 1.0388987974774417e-05,
+      "loss": 1.6956,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5031748755792003,
+      "grad_norm": 3.117335313176168,
+      "learning_rate": 1.0383433748381571e-05,
+      "loss": 1.6947,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5033464904753733,
+      "grad_norm": 2.5805978855208815,
+      "learning_rate": 1.0377879403524895e-05,
+      "loss": 1.1251,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5035181053715463,
+      "grad_norm": 3.2249294499925374,
+      "learning_rate": 1.0372324941920435e-05,
+      "loss": 1.6929,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5036897202677192,
+      "grad_norm": 3.0081875451924986,
+      "learning_rate": 1.0366770365284271e-05,
+      "loss": 1.7803,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5038613351638922,
+      "grad_norm": 3.249011312384148,
+      "learning_rate": 1.0361215675332512e-05,
+      "loss": 1.4356,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5040329500600652,
+      "grad_norm": 3.585741225233707,
+      "learning_rate": 1.0355660873781316e-05,
+      "loss": 1.4232,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5042045649562382,
+      "grad_norm": 3.4699371079678776,
+      "learning_rate": 1.0350105962346866e-05,
+      "loss": 1.9963,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5043761798524112,
+      "grad_norm": 3.086066657091744,
+      "learning_rate": 1.0344550942745376e-05,
+      "loss": 1.6971,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5045477947485841,
+      "grad_norm": 3.962657677326349,
+      "learning_rate": 1.0338995816693103e-05,
+      "loss": 1.7095,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5047194096447571,
+      "grad_norm": 3.572988293252918,
+      "learning_rate": 1.0333440585906326e-05,
+      "loss": 1.6938,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5048910245409302,
+      "grad_norm": 3.2501704646702843,
+      "learning_rate": 1.0327885252101368e-05,
+      "loss": 1.6038,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5050626394371032,
+      "grad_norm": 3.0890353322432333,
+      "learning_rate": 1.0322329816994574e-05,
+      "loss": 1.564,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5052342543332762,
+      "grad_norm": 2.9619208484997035,
+      "learning_rate": 1.0316774282302324e-05,
+      "loss": 1.5862,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5054058692294491,
+      "grad_norm": 3.72995209511458,
+      "learning_rate": 1.0311218649741033e-05,
+      "loss": 1.5326,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5055774841256221,
+      "grad_norm": 2.9784824088073676,
+      "learning_rate": 1.030566292102714e-05,
+      "loss": 1.7438,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5057490990217951,
+      "grad_norm": 4.106715634550961,
+      "learning_rate": 1.0300107097877114e-05,
+      "loss": 1.7303,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5059207139179681,
+      "grad_norm": 3.1130712411699815,
+      "learning_rate": 1.0294551182007457e-05,
+      "loss": 1.6983,
+      "step": 2948
+    },
+    {
+      "epoch": 0.506092328814141,
+      "grad_norm": 3.218880328414794,
+      "learning_rate": 1.02889951751347e-05,
+      "loss": 1.746,
+      "step": 2949
+    },
+    {
+      "epoch": 0.506263943710314,
+      "grad_norm": 3.6755946075999635,
+      "learning_rate": 1.0283439078975398e-05,
+      "loss": 1.4895,
+      "step": 2950
+    },
+    {
+      "epoch": 0.506435558606487,
+      "grad_norm": 3.750564088714565,
+      "learning_rate": 1.027788289524613e-05,
+      "loss": 1.6799,
+      "step": 2951
+    },
+    {
+      "epoch": 0.50660717350266,
+      "grad_norm": 3.1956894531704774,
+      "learning_rate": 1.0272326625663518e-05,
+      "loss": 1.5424,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5067787883988331,
+      "grad_norm": 3.2637943492249546,
+      "learning_rate": 1.0266770271944191e-05,
+      "loss": 1.6941,
+      "step": 2953
+    },
+    {
+      "epoch": 0.506950403295006,
+      "grad_norm": 3.354373593989549,
+      "learning_rate": 1.0261213835804819e-05,
+      "loss": 1.8451,
+      "step": 2954
+    },
+    {
+      "epoch": 0.507122018191179,
+      "grad_norm": 3.6255017169228205,
+      "learning_rate": 1.0255657318962089e-05,
+      "loss": 1.5691,
+      "step": 2955
+    },
+    {
+      "epoch": 0.507293633087352,
+      "grad_norm": 3.075331238295357,
+      "learning_rate": 1.0250100723132714e-05,
+      "loss": 1.7295,
+      "step": 2956
+    },
+    {
+      "epoch": 0.507465247983525,
+      "grad_norm": 3.4729559836605866,
+      "learning_rate": 1.0244544050033435e-05,
+      "loss": 1.6283,
+      "step": 2957
+    },
+    {
+      "epoch": 0.507636862879698,
+      "grad_norm": 3.3293556911998152,
+      "learning_rate": 1.0238987301381017e-05,
+      "loss": 1.8535,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5078084777758709,
+      "grad_norm": 2.894293906237257,
+      "learning_rate": 1.0233430478892242e-05,
+      "loss": 1.4294,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5079800926720439,
+      "grad_norm": 2.9224548092868163,
+      "learning_rate": 1.0227873584283918e-05,
+      "loss": 1.692,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5081517075682169,
+      "grad_norm": 3.6853315435030893,
+      "learning_rate": 1.022231661927288e-05,
+      "loss": 1.5805,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5083233224643899,
+      "grad_norm": 3.402319340749703,
+      "learning_rate": 1.0216759585575979e-05,
+      "loss": 1.8483,
+      "step": 2962
+    },
+    {
+      "epoch": 0.508494937360563,
+      "grad_norm": 3.3972757193184138,
+      "learning_rate": 1.0211202484910088e-05,
+      "loss": 1.6312,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5086665522567358,
+      "grad_norm": 2.9433187632711126,
+      "learning_rate": 1.0205645318992107e-05,
+      "loss": 1.6051,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5088381671529089,
+      "grad_norm": 3.828306753602516,
+      "learning_rate": 1.0200088089538944e-05,
+      "loss": 1.7072,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5090097820490819,
+      "grad_norm": 3.7850638048350733,
+      "learning_rate": 1.0194530798267538e-05,
+      "loss": 1.7303,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5091813969452549,
+      "grad_norm": 2.862372877817793,
+      "learning_rate": 1.0188973446894844e-05,
+      "loss": 1.4762,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5093530118414278,
+      "grad_norm": 3.0941524300321763,
+      "learning_rate": 1.0183416037137828e-05,
+      "loss": 1.59,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5095246267376008,
+      "grad_norm": 3.265643471826439,
+      "learning_rate": 1.0177858570713486e-05,
+      "loss": 1.8967,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5096962416337738,
+      "grad_norm": 3.2477090029599367,
+      "learning_rate": 1.0172301049338822e-05,
+      "loss": 1.6869,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5098678565299468,
+      "grad_norm": 4.110351359031347,
+      "learning_rate": 1.0166743474730865e-05,
+      "loss": 1.6598,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5100394714261198,
+      "grad_norm": 3.6985490106066257,
+      "learning_rate": 1.0161185848606652e-05,
+      "loss": 1.8339,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5102110863222927,
+      "grad_norm": 3.1256289600190095,
+      "learning_rate": 1.0155628172683245e-05,
+      "loss": 1.5702,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5103827012184657,
+      "grad_norm": 3.1956141812260808,
+      "learning_rate": 1.0150070448677714e-05,
+      "loss": 1.6593,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5105543161146388,
+      "grad_norm": 3.9260590753888787,
+      "learning_rate": 1.014451267830714e-05,
+      "loss": 1.5328,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5107259310108118,
+      "grad_norm": 6.240302313851187,
+      "learning_rate": 1.013895486328864e-05,
+      "loss": 1.6802,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5108975459069848,
+      "grad_norm": 2.679494721565575,
+      "learning_rate": 1.0133397005339313e-05,
+      "loss": 1.5386,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5110691608031577,
+      "grad_norm": 3.030778005303316,
+      "learning_rate": 1.01278391061763e-05,
+      "loss": 1.578,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5112407756993307,
+      "grad_norm": 4.072516500511933,
+      "learning_rate": 1.0122281167516736e-05,
+      "loss": 1.5942,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5114123905955037,
+      "grad_norm": 3.3160383021214788,
+      "learning_rate": 1.0116723191077775e-05,
+      "loss": 1.6524,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5115840054916767,
+      "grad_norm": 3.348323569115134,
+      "learning_rate": 1.0111165178576587e-05,
+      "loss": 1.7913,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5117556203878497,
+      "grad_norm": 3.213964036579384,
+      "learning_rate": 1.0105607131730343e-05,
+      "loss": 1.6197,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5119272352840226,
+      "grad_norm": 2.8553108231672235,
+      "learning_rate": 1.0100049052256236e-05,
+      "loss": 1.489,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5120988501801956,
+      "grad_norm": 3.15221880149023,
+      "learning_rate": 1.0094490941871456e-05,
+      "loss": 1.7178,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5122704650763686,
+      "grad_norm": 2.877827086883412,
+      "learning_rate": 1.0088932802293214e-05,
+      "loss": 1.566,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5124420799725417,
+      "grad_norm": 4.031015391832658,
+      "learning_rate": 1.0083374635238728e-05,
+      "loss": 1.6965,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5126136948687146,
+      "grad_norm": 2.7359894328298635,
+      "learning_rate": 1.0077816442425216e-05,
+      "loss": 1.313,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5127853097648876,
+      "grad_norm": 3.151592104093204,
+      "learning_rate": 1.0072258225569919e-05,
+      "loss": 1.8535,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5129569246610606,
+      "grad_norm": 2.379391053213078,
+      "learning_rate": 1.0066699986390067e-05,
+      "loss": 1.3617,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5131285395572336,
+      "grad_norm": 4.436334611424934,
+      "learning_rate": 1.0061141726602912e-05,
+      "loss": 1.7935,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5133001544534066,
+      "grad_norm": 3.712016639280695,
+      "learning_rate": 1.005558344792571e-05,
+      "loss": 1.7661,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5134717693495795,
+      "grad_norm": 3.1979438906163433,
+      "learning_rate": 1.0050025152075711e-05,
+      "loss": 1.5041,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5136433842457525,
+      "grad_norm": 3.5715845810873246,
+      "learning_rate": 1.0044466840770186e-05,
+      "loss": 1.6445,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5138149991419255,
+      "grad_norm": 3.3884576562126116,
+      "learning_rate": 1.0038908515726403e-05,
+      "loss": 1.4902,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5139866140380985,
+      "grad_norm": 3.544862197998934,
+      "learning_rate": 1.0033350178661633e-05,
+      "loss": 1.6072,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5141582289342715,
+      "grad_norm": 2.714391497465356,
+      "learning_rate": 1.0027791831293152e-05,
+      "loss": 1.5935,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5143298438304444,
+      "grad_norm": 2.626893230202131,
+      "learning_rate": 1.0022233475338245e-05,
+      "loss": 1.681,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5145014587266175,
+      "grad_norm": 3.107941559695089,
+      "learning_rate": 1.0016675112514192e-05,
+      "loss": 1.6247,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5146730736227905,
+      "grad_norm": 3.212739318253138,
+      "learning_rate": 1.0011116744538271e-05,
+      "loss": 1.5936,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5148446885189635,
+      "grad_norm": 2.8151035143881327,
+      "learning_rate": 1.0005558373127784e-05,
+      "loss": 1.4084,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5150163034151364,
+      "grad_norm": 3.386392327409711,
+      "learning_rate": 1e-05,
+      "loss": 1.5833,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5151879183113094,
+      "grad_norm": 2.9561337237382714,
+      "learning_rate": 9.99444162687222e-06,
+      "loss": 1.5548,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5153595332074824,
+      "grad_norm": 3.2153208004661926,
+      "learning_rate": 9.988883255461729e-06,
+      "loss": 1.4379,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5155311481036554,
+      "grad_norm": 2.898719742125427,
+      "learning_rate": 9.98332488748581e-06,
+      "loss": 1.6423,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5157027629998284,
+      "grad_norm": 3.9387096897170597,
+      "learning_rate": 9.97776652466176e-06,
+      "loss": 1.7711,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5158743778960013,
+      "grad_norm": 4.105848117113761,
+      "learning_rate": 9.97220816870685e-06,
+      "loss": 1.7484,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5160459927921743,
+      "grad_norm": 2.834630096278617,
+      "learning_rate": 9.966649821338368e-06,
+      "loss": 1.7465,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5162176076883473,
+      "grad_norm": 3.064264649089282,
+      "learning_rate": 9.961091484273599e-06,
+      "loss": 1.8399,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5163892225845204,
+      "grad_norm": 3.693562209048199,
+      "learning_rate": 9.955533159229814e-06,
+      "loss": 1.5476,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5165608374806934,
+      "grad_norm": 3.4280264854946267,
+      "learning_rate": 9.94997484792429e-06,
+      "loss": 1.5403,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5167324523768663,
+      "grad_norm": 4.573582908609637,
+      "learning_rate": 9.944416552074296e-06,
+      "loss": 1.6736,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5169040672730393,
+      "grad_norm": 3.5798959139392132,
+      "learning_rate": 9.938858273397091e-06,
+      "loss": 1.5538,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5170756821692123,
+      "grad_norm": 3.030795910708309,
+      "learning_rate": 9.933300013609936e-06,
+      "loss": 1.6508,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5172472970653853,
+      "grad_norm": 2.880406233014958,
+      "learning_rate": 9.927741774430084e-06,
+      "loss": 1.6224,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5174189119615583,
+      "grad_norm": 3.113151783626446,
+      "learning_rate": 9.922183557574786e-06,
+      "loss": 1.688,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5175905268577312,
+      "grad_norm": 3.3465255811606074,
+      "learning_rate": 9.916625364761274e-06,
+      "loss": 1.5959,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5177621417539042,
+      "grad_norm": 2.9595810525444146,
+      "learning_rate": 9.91106719770679e-06,
+      "loss": 1.4398,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5179337566500772,
+      "grad_norm": 2.6469513817153705,
+      "learning_rate": 9.905509058128547e-06,
+      "loss": 1.2602,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5181053715462502,
+      "grad_norm": 2.8632946620707704,
+      "learning_rate": 9.899950947743767e-06,
+      "loss": 1.6541,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5182769864424231,
+      "grad_norm": 3.882779895762631,
+      "learning_rate": 9.894392868269658e-06,
+      "loss": 1.7289,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5184486013385962,
+      "grad_norm": 3.223046716096657,
+      "learning_rate": 9.888834821423415e-06,
+      "loss": 1.7866,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5186202162347692,
+      "grad_norm": 2.920297766415385,
+      "learning_rate": 9.883276808922225e-06,
+      "loss": 1.597,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5187918311309422,
+      "grad_norm": 3.4869738402894552,
+      "learning_rate": 9.877718832483269e-06,
+      "loss": 1.8228,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5189634460271152,
+      "grad_norm": 3.0568271775035143,
+      "learning_rate": 9.872160893823702e-06,
+      "loss": 1.8056,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5191350609232881,
+      "grad_norm": 2.66876404811824,
+      "learning_rate": 9.866602994660688e-06,
+      "loss": 1.6332,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5193066758194611,
+      "grad_norm": 3.203141537484068,
+      "learning_rate": 9.861045136711362e-06,
+      "loss": 1.7426,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5194782907156341,
+      "grad_norm": 3.0591674814064147,
+      "learning_rate": 9.85548732169286e-06,
+      "loss": 1.3971,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5196499056118071,
+      "grad_norm": 3.2259742354114223,
+      "learning_rate": 9.849929551322288e-06,
+      "loss": 1.511,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5198215205079801,
+      "grad_norm": 2.9362744614163336,
+      "learning_rate": 9.84437182731676e-06,
+      "loss": 1.5222,
+      "step": 3029
+    },
+    {
+      "epoch": 0.519993135404153,
+      "grad_norm": 3.486300445761983,
+      "learning_rate": 9.83881415139335e-06,
+      "loss": 1.7279,
+      "step": 3030
+    },
+    {
+      "epoch": 0.520164750300326,
+      "grad_norm": 2.807954713021253,
+      "learning_rate": 9.833256525269136e-06,
+      "loss": 1.4414,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5203363651964991,
+      "grad_norm": 3.3137180490243545,
+      "learning_rate": 9.82769895066118e-06,
+      "loss": 1.5196,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5205079800926721,
+      "grad_norm": 2.645204340432645,
+      "learning_rate": 9.822141429286515e-06,
+      "loss": 1.6183,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5206795949888451,
+      "grad_norm": 3.62514299874206,
+      "learning_rate": 9.816583962862174e-06,
+      "loss": 1.7005,
+      "step": 3034
+    },
+    {
+      "epoch": 0.520851209885018,
+      "grad_norm": 3.586347181395042,
+      "learning_rate": 9.811026553105162e-06,
+      "loss": 1.6668,
+      "step": 3035
+    },
+    {
+      "epoch": 0.521022824781191,
+      "grad_norm": 2.8056090701722347,
+      "learning_rate": 9.805469201732463e-06,
+      "loss": 1.4765,
+      "step": 3036
+    },
+    {
+      "epoch": 0.521194439677364,
+      "grad_norm": 2.7152421285079256,
+      "learning_rate": 9.799911910461059e-06,
+      "loss": 1.5281,
+      "step": 3037
+    },
+    {
+      "epoch": 0.521366054573537,
+      "grad_norm": 2.8691051669555776,
+      "learning_rate": 9.794354681007894e-06,
+      "loss": 1.5387,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5215376694697099,
+      "grad_norm": 2.9418932340943575,
+      "learning_rate": 9.788797515089912e-06,
+      "loss": 1.1927,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5217092843658829,
+      "grad_norm": 3.2460695971899836,
+      "learning_rate": 9.783240414424021e-06,
+      "loss": 1.7296,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5218808992620559,
+      "grad_norm": 3.3727553693915806,
+      "learning_rate": 9.777683380727123e-06,
+      "loss": 1.8422,
+      "step": 3041
+    },
+    {
+      "epoch": 0.522052514158229,
+      "grad_norm": 3.846914794687572,
+      "learning_rate": 9.772126415716085e-06,
+      "loss": 1.4607,
+      "step": 3042
+    },
+    {
+      "epoch": 0.522224129054402,
+      "grad_norm": 3.209946771691414,
+      "learning_rate": 9.766569521107762e-06,
+      "loss": 1.5448,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5223957439505749,
+      "grad_norm": 3.486569670983471,
+      "learning_rate": 9.761012698618985e-06,
+      "loss": 1.672,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5225673588467479,
+      "grad_norm": 3.2763119287848013,
+      "learning_rate": 9.755455949966565e-06,
+      "loss": 1.6314,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5227389737429209,
+      "grad_norm": 4.3359332117927245,
+      "learning_rate": 9.749899276867286e-06,
+      "loss": 1.8492,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5229105886390939,
+      "grad_norm": 3.058358055627463,
+      "learning_rate": 9.744342681037916e-06,
+      "loss": 1.5919,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5230822035352669,
+      "grad_norm": 2.8332424891756327,
+      "learning_rate": 9.738786164195183e-06,
+      "loss": 1.7999,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5232538184314398,
+      "grad_norm": 3.2866534189140126,
+      "learning_rate": 9.73322972805581e-06,
+      "loss": 1.6018,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5234254333276128,
+      "grad_norm": 3.195891653244995,
+      "learning_rate": 9.727673374336483e-06,
+      "loss": 1.7746,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5235970482237858,
+      "grad_norm": 3.6342109755862966,
+      "learning_rate": 9.72211710475387e-06,
+      "loss": 1.578,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5237686631199588,
+      "grad_norm": 3.003703295219066,
+      "learning_rate": 9.716560921024604e-06,
+      "loss": 1.7425,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5239402780161319,
+      "grad_norm": 3.2432828461897687,
+      "learning_rate": 9.711004824865304e-06,
+      "loss": 1.5037,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5241118929123048,
+      "grad_norm": 2.692755791278498,
+      "learning_rate": 9.705448817992546e-06,
+      "loss": 1.3273,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5242835078084778,
+      "grad_norm": 3.609256015040475,
+      "learning_rate": 9.699892902122887e-06,
+      "loss": 1.714,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5244551227046508,
+      "grad_norm": 3.1870651312966904,
+      "learning_rate": 9.694337078972864e-06,
+      "loss": 1.6096,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5246267376008238,
+      "grad_norm": 3.235420185447154,
+      "learning_rate": 9.688781350258968e-06,
+      "loss": 1.6898,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5247983524969967,
+      "grad_norm": 3.6810319709805546,
+      "learning_rate": 9.683225717697676e-06,
+      "loss": 1.4189,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5249699673931697,
+      "grad_norm": 3.25925483502963,
+      "learning_rate": 9.67767018300543e-06,
+      "loss": 1.8755,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5251415822893427,
+      "grad_norm": 3.8320906280394906,
+      "learning_rate": 9.672114747898635e-06,
+      "loss": 1.6695,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5253131971855157,
+      "grad_norm": 2.8587355719532934,
+      "learning_rate": 9.666559414093677e-06,
+      "loss": 1.6293,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5254848120816887,
+      "grad_norm": 2.6877686928836897,
+      "learning_rate": 9.6610041833069e-06,
+      "loss": 1.6911,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5256564269778616,
+      "grad_norm": 3.871341156583771,
+      "learning_rate": 9.655449057254626e-06,
+      "loss": 1.6207,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5258280418740346,
+      "grad_norm": 2.9626490480956504,
+      "learning_rate": 9.649894037653134e-06,
+      "loss": 1.5107,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5259996567702077,
+      "grad_norm": 2.9580495649736647,
+      "learning_rate": 9.644339126218685e-06,
+      "loss": 1.629,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5261712716663807,
+      "grad_norm": 3.0515758048142048,
+      "learning_rate": 9.63878432466749e-06,
+      "loss": 1.4563,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5263428865625537,
+      "grad_norm": 2.8631908037679143,
+      "learning_rate": 9.633229634715734e-06,
+      "loss": 1.6682,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5265145014587266,
+      "grad_norm": 3.8304617879037144,
+      "learning_rate": 9.627675058079568e-06,
+      "loss": 1.7055,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5266861163548996,
+      "grad_norm": 3.94461020436735,
+      "learning_rate": 9.622120596475106e-06,
+      "loss": 1.778,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5268577312510726,
+      "grad_norm": 3.1485992130098026,
+      "learning_rate": 9.61656625161843e-06,
+      "loss": 1.5378,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5270293461472456,
+      "grad_norm": 3.4000883307667227,
+      "learning_rate": 9.611012025225586e-06,
+      "loss": 1.782,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5272009610434185,
+      "grad_norm": 3.4845712419648374,
+      "learning_rate": 9.605457919012573e-06,
+      "loss": 1.5906,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5273725759395915,
+      "grad_norm": 2.997439696912847,
+      "learning_rate": 9.599903934695366e-06,
+      "loss": 1.5391,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5275441908357645,
+      "grad_norm": 3.2332163603639534,
+      "learning_rate": 9.594350073989894e-06,
+      "loss": 1.724,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5277158057319375,
+      "grad_norm": 3.0070930550206905,
+      "learning_rate": 9.588796338612054e-06,
+      "loss": 1.5411,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5278874206281106,
+      "grad_norm": 2.8369221225138164,
+      "learning_rate": 9.583242730277699e-06,
+      "loss": 1.4523,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5280590355242835,
+      "grad_norm": 3.0637510898392435,
+      "learning_rate": 9.577689250702648e-06,
+      "loss": 1.4609,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5282306504204565,
+      "grad_norm": 3.1841349008713564,
+      "learning_rate": 9.572135901602674e-06,
+      "loss": 1.5989,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5284022653166295,
+      "grad_norm": 2.4662320141750036,
+      "learning_rate": 9.566582684693511e-06,
+      "loss": 1.555,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5285738802128025,
+      "grad_norm": 4.4480395594401285,
+      "learning_rate": 9.561029601690857e-06,
+      "loss": 1.5729,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5287454951089755,
+      "grad_norm": 3.348426964930606,
+      "learning_rate": 9.555476654310364e-06,
+      "loss": 1.6402,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5289171100051484,
+      "grad_norm": 2.646244640858748,
+      "learning_rate": 9.549923844267645e-06,
+      "loss": 1.5573,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5290887249013214,
+      "grad_norm": 3.807154859508447,
+      "learning_rate": 9.544371173278272e-06,
+      "loss": 1.6559,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5292603397974944,
+      "grad_norm": 2.5699535991811344,
+      "learning_rate": 9.538818643057764e-06,
+      "loss": 1.4321,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5294319546936674,
+      "grad_norm": 2.7079988731211966,
+      "learning_rate": 9.53326625532161e-06,
+      "loss": 1.5292,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5296035695898405,
+      "grad_norm": 3.079291556509928,
+      "learning_rate": 9.527714011785242e-06,
+      "loss": 1.5964,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5297751844860134,
+      "grad_norm": 3.5112551476220415,
+      "learning_rate": 9.522161914164062e-06,
+      "loss": 1.6439,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5299467993821864,
+      "grad_norm": 2.6316422699071955,
+      "learning_rate": 9.516609964173415e-06,
+      "loss": 1.5045,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5301184142783594,
+      "grad_norm": 2.8356694587918763,
+      "learning_rate": 9.511058163528609e-06,
+      "loss": 1.8322,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5302900291745324,
+      "grad_norm": 3.0095093652520353,
+      "learning_rate": 9.505506513944895e-06,
+      "loss": 1.7701,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5304616440707053,
+      "grad_norm": 3.7078079876066483,
+      "learning_rate": 9.499955017137485e-06,
+      "loss": 1.8798,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5306332589668783,
+      "grad_norm": 4.049965696696388,
+      "learning_rate": 9.494403674821544e-06,
+      "loss": 1.8037,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5308048738630513,
+      "grad_norm": 3.3324218702154194,
+      "learning_rate": 9.488852488712186e-06,
+      "loss": 1.6893,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5309764887592243,
+      "grad_norm": 2.8581293827071406,
+      "learning_rate": 9.483301460524484e-06,
+      "loss": 1.8557,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5311481036553973,
+      "grad_norm": 3.473225541807752,
+      "learning_rate": 9.477750591973452e-06,
+      "loss": 1.6691,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5313197185515702,
+      "grad_norm": 3.5794322951330813,
+      "learning_rate": 9.472199884774057e-06,
+      "loss": 1.7149,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5314913334477432,
+      "grad_norm": 2.7382333920686834,
+      "learning_rate": 9.466649340641224e-06,
+      "loss": 1.6974,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5316629483439163,
+      "grad_norm": 2.887841404324223,
+      "learning_rate": 9.461098961289819e-06,
+      "loss": 1.8367,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5318345632400893,
+      "grad_norm": 3.199515880275442,
+      "learning_rate": 9.45554874843466e-06,
+      "loss": 1.6463,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5320061781362623,
+      "grad_norm": 2.7367523527859325,
+      "learning_rate": 9.449998703790515e-06,
+      "loss": 1.4332,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5321777930324352,
+      "grad_norm": 3.1339916092169706,
+      "learning_rate": 9.444448829072103e-06,
+      "loss": 1.8334,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5323494079286082,
+      "grad_norm": 3.1675956152228326,
+      "learning_rate": 9.438899125994078e-06,
+      "loss": 1.5957,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5325210228247812,
+      "grad_norm": 2.964338482506634,
+      "learning_rate": 9.433349596271053e-06,
+      "loss": 1.6785,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5326926377209542,
+      "grad_norm": 3.8456787896391362,
+      "learning_rate": 9.427800241617585e-06,
+      "loss": 1.6171,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5328642526171272,
+      "grad_norm": 2.8051100245106744,
+      "learning_rate": 9.422251063748173e-06,
+      "loss": 1.5772,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5330358675133001,
+      "grad_norm": 3.224182021090952,
+      "learning_rate": 9.416702064377268e-06,
+      "loss": 1.7539,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5332074824094731,
+      "grad_norm": 3.406151133598625,
+      "learning_rate": 9.411153245219262e-06,
+      "loss": 1.8219,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5333790973056461,
+      "grad_norm": 2.8712927931766603,
+      "learning_rate": 9.405604607988484e-06,
+      "loss": 1.6901,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5335507122018192,
+      "grad_norm": 2.7512510766676535,
+      "learning_rate": 9.40005615439922e-06,
+      "loss": 1.447,
+      "step": 3109
+    },
+    {
+      "epoch": 0.533722327097992,
+      "grad_norm": 2.9091697168803616,
+      "learning_rate": 9.394507886165692e-06,
+      "loss": 1.4168,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5338939419941651,
+      "grad_norm": 3.4704008075401305,
+      "learning_rate": 9.388959805002067e-06,
+      "loss": 1.7488,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5340655568903381,
+      "grad_norm": 3.5895672566890706,
+      "learning_rate": 9.383411912622451e-06,
+      "loss": 1.7046,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5342371717865111,
+      "grad_norm": 2.9292126485329453,
+      "learning_rate": 9.3778642107409e-06,
+      "loss": 1.6853,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5344087866826841,
+      "grad_norm": 3.568739642159117,
+      "learning_rate": 9.372316701071395e-06,
+      "loss": 1.5751,
+      "step": 3114
+    },
+    {
+      "epoch": 0.534580401578857,
+      "grad_norm": 3.1154170172613482,
+      "learning_rate": 9.366769385327875e-06,
+      "loss": 1.7357,
+      "step": 3115
+    },
+    {
+      "epoch": 0.53475201647503,
+      "grad_norm": 2.539015831059933,
+      "learning_rate": 9.361222265224208e-06,
+      "loss": 1.3283,
+      "step": 3116
+    },
+    {
+      "epoch": 0.534923631371203,
+      "grad_norm": 3.3481529316957417,
+      "learning_rate": 9.355675342474207e-06,
+      "loss": 1.42,
+      "step": 3117
+    },
+    {
+      "epoch": 0.535095246267376,
+      "grad_norm": 3.664106804156913,
+      "learning_rate": 9.350128618791622e-06,
+      "loss": 1.8234,
+      "step": 3118
+    },
+    {
+      "epoch": 0.535266861163549,
+      "grad_norm": 3.1141428945684364,
+      "learning_rate": 9.344582095890146e-06,
+      "loss": 1.7784,
+      "step": 3119
+    },
+    {
+      "epoch": 0.535438476059722,
+      "grad_norm": 3.2260373550648382,
+      "learning_rate": 9.339035775483397e-06,
+      "loss": 1.849,
+      "step": 3120
+    },
+    {
+      "epoch": 0.535610090955895,
+      "grad_norm": 3.4806057930306125,
+      "learning_rate": 9.333489659284945e-06,
+      "loss": 1.769,
+      "step": 3121
+    },
+    {
+      "epoch": 0.535781705852068,
+      "grad_norm": 3.0437960640118646,
+      "learning_rate": 9.327943749008291e-06,
+      "loss": 1.7897,
+      "step": 3122
+    },
+    {
+      "epoch": 0.535953320748241,
+      "grad_norm": 3.322013640504874,
+      "learning_rate": 9.32239804636687e-06,
+      "loss": 1.6058,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5361249356444139,
+      "grad_norm": 2.8260356895638363,
+      "learning_rate": 9.316852553074057e-06,
+      "loss": 1.6125,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5362965505405869,
+      "grad_norm": 6.503217832256964,
+      "learning_rate": 9.311307270843163e-06,
+      "loss": 1.5118,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5364681654367599,
+      "grad_norm": 2.82857335316144,
+      "learning_rate": 9.305762201387429e-06,
+      "loss": 1.5668,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5366397803329329,
+      "grad_norm": 2.9669791648103843,
+      "learning_rate": 9.300217346420029e-06,
+      "loss": 1.4485,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5368113952291059,
+      "grad_norm": 3.367680783024022,
+      "learning_rate": 9.294672707654082e-06,
+      "loss": 1.5547,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5369830101252788,
+      "grad_norm": 3.4903317504869467,
+      "learning_rate": 9.289128286802626e-06,
+      "loss": 1.5755,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5371546250214518,
+      "grad_norm": 3.1530723549601527,
+      "learning_rate": 9.283584085578642e-06,
+      "loss": 1.7715,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5373262399176248,
+      "grad_norm": 3.124468739546777,
+      "learning_rate": 9.278040105695044e-06,
+      "loss": 1.5466,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5374978548137979,
+      "grad_norm": 3.945707095709437,
+      "learning_rate": 9.27249634886466e-06,
+      "loss": 1.4369,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5376694697099709,
+      "grad_norm": 3.831407487967149,
+      "learning_rate": 9.266952816800273e-06,
+      "loss": 1.6932,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5378410846061438,
+      "grad_norm": 2.749762767683833,
+      "learning_rate": 9.261409511214583e-06,
+      "loss": 1.4992,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5380126995023168,
+      "grad_norm": 3.2977005493599743,
+      "learning_rate": 9.25586643382022e-06,
+      "loss": 1.6776,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5381843143984898,
+      "grad_norm": 4.067565282481566,
+      "learning_rate": 9.25032358632975e-06,
+      "loss": 1.4197,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5383559292946628,
+      "grad_norm": 3.7995147190027736,
+      "learning_rate": 9.244780970455667e-06,
+      "loss": 1.7455,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5385275441908358,
+      "grad_norm": 3.103530138190673,
+      "learning_rate": 9.239238587910383e-06,
+      "loss": 1.648,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5386991590870087,
+      "grad_norm": 2.685026120240778,
+      "learning_rate": 9.233696440406248e-06,
+      "loss": 1.4128,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5388707739831817,
+      "grad_norm": 3.435715831410238,
+      "learning_rate": 9.228154529655541e-06,
+      "loss": 1.5112,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5390423888793547,
+      "grad_norm": 2.9651458083613584,
+      "learning_rate": 9.222612857370457e-06,
+      "loss": 1.6358,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5392140037755278,
+      "grad_norm": 3.8790807136137473,
+      "learning_rate": 9.217071425263133e-06,
+      "loss": 1.8737,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5393856186717007,
+      "grad_norm": 2.98994425488071,
+      "learning_rate": 9.211530235045619e-06,
+      "loss": 1.4321,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5395572335678737,
+      "grad_norm": 2.7756326011718397,
+      "learning_rate": 9.205989288429887e-06,
+      "loss": 1.5311,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5397288484640467,
+      "grad_norm": 2.7959937854848054,
+      "learning_rate": 9.200448587127852e-06,
+      "loss": 1.474,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5399004633602197,
+      "grad_norm": 3.159795372265623,
+      "learning_rate": 9.194908132851335e-06,
+      "loss": 1.7627,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5400720782563927,
+      "grad_norm": 4.014115503951635,
+      "learning_rate": 9.18936792731209e-06,
+      "loss": 1.5017,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5402436931525656,
+      "grad_norm": 4.094963679713884,
+      "learning_rate": 9.183827972221793e-06,
+      "loss": 1.8115,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5404153080487386,
+      "grad_norm": 3.801222655142659,
+      "learning_rate": 9.178288269292044e-06,
+      "loss": 1.5325,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5405869229449116,
+      "grad_norm": 3.224236046502278,
+      "learning_rate": 9.172748820234357e-06,
+      "loss": 1.5453,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5407585378410846,
+      "grad_norm": 3.3694237934199216,
+      "learning_rate": 9.167209626760173e-06,
+      "loss": 1.6274,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5409301527372576,
+      "grad_norm": 2.6918086154650154,
+      "learning_rate": 9.161670690580861e-06,
+      "loss": 1.3827,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5411017676334305,
+      "grad_norm": 3.564404061384756,
+      "learning_rate": 9.156132013407698e-06,
+      "loss": 1.6679,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5412733825296036,
+      "grad_norm": 3.9773108418994694,
+      "learning_rate": 9.150593596951888e-06,
+      "loss": 1.6985,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5414449974257766,
+      "grad_norm": 3.2267741107650285,
+      "learning_rate": 9.14505544292456e-06,
+      "loss": 1.6322,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5416166123219496,
+      "grad_norm": 2.705750818995077,
+      "learning_rate": 9.139517553036743e-06,
+      "loss": 1.49,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5417882272181226,
+      "grad_norm": 3.7191539059704244,
+      "learning_rate": 9.133979928999404e-06,
+      "loss": 1.914,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5419598421142955,
+      "grad_norm": 3.853330523390222,
+      "learning_rate": 9.128442572523418e-06,
+      "loss": 1.6405,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5421314570104685,
+      "grad_norm": 3.6269421619656335,
+      "learning_rate": 9.122905485319583e-06,
+      "loss": 1.7037,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5423030719066415,
+      "grad_norm": 3.281591999021539,
+      "learning_rate": 9.117368669098606e-06,
+      "loss": 1.7438,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5424746868028145,
+      "grad_norm": 3.0282696455524576,
+      "learning_rate": 9.111832125571122e-06,
+      "loss": 1.3565,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5426463016989874,
+      "grad_norm": 4.260601910693571,
+      "learning_rate": 9.106295856447669e-06,
+      "loss": 1.6524,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5428179165951604,
+      "grad_norm": 3.625124872919927,
+      "learning_rate": 9.100759863438702e-06,
+      "loss": 1.5342,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5429895314913334,
+      "grad_norm": 3.534147315392249,
+      "learning_rate": 9.095224148254603e-06,
+      "loss": 1.7402,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5431611463875065,
+      "grad_norm": 3.26583741777128,
+      "learning_rate": 9.089688712605653e-06,
+      "loss": 1.8655,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5433327612836795,
+      "grad_norm": 4.2500124236058525,
+      "learning_rate": 9.084153558202058e-06,
+      "loss": 1.7281,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5435043761798524,
+      "grad_norm": 2.9720610009778494,
+      "learning_rate": 9.078618686753933e-06,
+      "loss": 1.5478,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5436759910760254,
+      "grad_norm": 2.8446562522949614,
+      "learning_rate": 9.073084099971298e-06,
+      "loss": 1.522,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5438476059721984,
+      "grad_norm": 4.0619232355186545,
+      "learning_rate": 9.067549799564101e-06,
+      "loss": 1.8494,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5440192208683714,
+      "grad_norm": 3.31855036678171,
+      "learning_rate": 9.062015787242184e-06,
+      "loss": 1.4725,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5441908357645444,
+      "grad_norm": 3.519317792347857,
+      "learning_rate": 9.056482064715316e-06,
+      "loss": 1.8027,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5443624506607173,
+      "grad_norm": 3.651502367758153,
+      "learning_rate": 9.050948633693162e-06,
+      "loss": 1.6547,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5445340655568903,
+      "grad_norm": 3.224645257590756,
+      "learning_rate": 9.045415495885313e-06,
+      "loss": 1.6618,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5447056804530633,
+      "grad_norm": 2.6408626767559285,
+      "learning_rate": 9.039882653001254e-06,
+      "loss": 1.2583,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5448772953492363,
+      "grad_norm": 2.909202745582501,
+      "learning_rate": 9.034350106750383e-06,
+      "loss": 1.5819,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5450489102454092,
+      "grad_norm": 3.1864166312010496,
+      "learning_rate": 9.028817858842013e-06,
+      "loss": 1.5887,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5452205251415823,
+      "grad_norm": 3.604941346577423,
+      "learning_rate": 9.02328591098536e-06,
+      "loss": 1.8248,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5453921400377553,
+      "grad_norm": 3.609232795481901,
+      "learning_rate": 9.017754264889548e-06,
+      "loss": 1.7879,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5455637549339283,
+      "grad_norm": 3.3737067530801066,
+      "learning_rate": 9.01222292226361e-06,
+      "loss": 1.6805,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5457353698301013,
+      "grad_norm": 3.3427368799239394,
+      "learning_rate": 9.006691884816475e-06,
+      "loss": 1.4917,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5459069847262742,
+      "grad_norm": 3.2118809871704537,
+      "learning_rate": 9.00116115425699e-06,
+      "loss": 1.4789,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5460785996224472,
+      "grad_norm": 4.171758481561278,
+      "learning_rate": 8.995630732293903e-06,
+      "loss": 1.6471,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5462502145186202,
+      "grad_norm": 2.901604661624939,
+      "learning_rate": 8.990100620635866e-06,
+      "loss": 1.4273,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5464218294147932,
+      "grad_norm": 3.347255413351688,
+      "learning_rate": 8.984570820991433e-06,
+      "loss": 1.4522,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5465934443109662,
+      "grad_norm": 3.003179390147555,
+      "learning_rate": 8.979041335069072e-06,
+      "loss": 1.4093,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5467650592071391,
+      "grad_norm": 3.152519016463287,
+      "learning_rate": 8.973512164577137e-06,
+      "loss": 1.5751,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5469366741033121,
+      "grad_norm": 3.4099552247134874,
+      "learning_rate": 8.967983311223898e-06,
+      "loss": 1.6504,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5471082889994852,
+      "grad_norm": 3.0910370271648766,
+      "learning_rate": 8.96245477671752e-06,
+      "loss": 1.4828,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5472799038956582,
+      "grad_norm": 2.9321717583910067,
+      "learning_rate": 8.956926562766073e-06,
+      "loss": 1.5085,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5474515187918312,
+      "grad_norm": 3.860885464620784,
+      "learning_rate": 8.951398671077531e-06,
+      "loss": 1.8999,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5476231336880041,
+      "grad_norm": 3.920706071802172,
+      "learning_rate": 8.945871103359764e-06,
+      "loss": 1.719,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5477947485841771,
+      "grad_norm": 3.430502026203666,
+      "learning_rate": 8.940343861320535e-06,
+      "loss": 1.5829,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5479663634803501,
+      "grad_norm": 3.4300209910536172,
+      "learning_rate": 8.934816946667522e-06,
+      "loss": 1.8548,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5481379783765231,
+      "grad_norm": 3.9460775544773163,
+      "learning_rate": 8.929290361108288e-06,
+      "loss": 1.6156,
+      "step": 3194
+    },
+    {
+      "epoch": 0.548309593272696,
+      "grad_norm": 4.169692879001639,
+      "learning_rate": 8.923764106350303e-06,
+      "loss": 1.7912,
+      "step": 3195
+    },
+    {
+      "epoch": 0.548481208168869,
+      "grad_norm": 3.7996445962538785,
+      "learning_rate": 8.918238184100932e-06,
+      "loss": 1.5151,
+      "step": 3196
+    },
+    {
+      "epoch": 0.548652823065042,
+      "grad_norm": 3.526572977598933,
+      "learning_rate": 8.912712596067439e-06,
+      "loss": 1.6278,
+      "step": 3197
+    },
+    {
+      "epoch": 0.548824437961215,
+      "grad_norm": 4.004996777401031,
+      "learning_rate": 8.90718734395698e-06,
+      "loss": 1.6496,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5489960528573881,
+      "grad_norm": 3.204871251414189,
+      "learning_rate": 8.901662429476607e-06,
+      "loss": 1.6575,
+      "step": 3199
+    },
+    {
+      "epoch": 0.549167667753561,
+      "grad_norm": 3.48869295977204,
+      "learning_rate": 8.896137854333276e-06,
+      "loss": 1.6429,
+      "step": 3200
+    },
+    {
+      "epoch": 0.549339282649734,
+      "grad_norm": 3.0777132727440244,
+      "learning_rate": 8.89061362023383e-06,
+      "loss": 1.5966,
+      "step": 3201
+    },
+    {
+      "epoch": 0.549510897545907,
+      "grad_norm": 3.130753088186336,
+      "learning_rate": 8.885089728885013e-06,
+      "loss": 1.7012,
+      "step": 3202
+    },
+    {
+      "epoch": 0.54968251244208,
+      "grad_norm": 3.8534710437305253,
+      "learning_rate": 8.87956618199346e-06,
+      "loss": 1.555,
+      "step": 3203
+    },
+    {
+      "epoch": 0.549854127338253,
+      "grad_norm": 4.17777308512548,
+      "learning_rate": 8.87404298126569e-06,
+      "loss": 1.6152,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5500257422344259,
+      "grad_norm": 5.884834937691766,
+      "learning_rate": 8.868520128408134e-06,
+      "loss": 1.6929,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5501973571305989,
+      "grad_norm": 3.3700435520397036,
+      "learning_rate": 8.862997625127101e-06,
+      "loss": 1.6399,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5503689720267719,
+      "grad_norm": 3.2331146774163155,
+      "learning_rate": 8.8574754731288e-06,
+      "loss": 1.6229,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5505405869229449,
+      "grad_norm": 2.8465536817599277,
+      "learning_rate": 8.851953674119325e-06,
+      "loss": 1.5305,
+      "step": 3208
+    },
+    {
+      "epoch": 0.550712201819118,
+      "grad_norm": 3.8719672771421205,
+      "learning_rate": 8.846432229804668e-06,
+      "loss": 1.4278,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5508838167152909,
+      "grad_norm": 3.4713391910977824,
+      "learning_rate": 8.840911141890706e-06,
+      "loss": 1.4845,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5510554316114639,
+      "grad_norm": 2.9990677686654994,
+      "learning_rate": 8.835390412083205e-06,
+      "loss": 1.66,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5512270465076369,
+      "grad_norm": 3.1860284710225675,
+      "learning_rate": 8.829870042087825e-06,
+      "loss": 1.8805,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5513986614038099,
+      "grad_norm": 3.1381371519748638,
+      "learning_rate": 8.824350033610111e-06,
+      "loss": 1.6016,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5515702762999828,
+      "grad_norm": 4.248106537432824,
+      "learning_rate": 8.8188303883555e-06,
+      "loss": 1.7105,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5517418911961558,
+      "grad_norm": 3.3046081324328593,
+      "learning_rate": 8.81331110802932e-06,
+      "loss": 1.7055,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5519135060923288,
+      "grad_norm": 2.813984778004484,
+      "learning_rate": 8.807792194336767e-06,
+      "loss": 1.3709,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5520851209885018,
+      "grad_norm": 3.290961639799164,
+      "learning_rate": 8.802273648982951e-06,
+      "loss": 1.5811,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5522567358846748,
+      "grad_norm": 3.4523665038248357,
+      "learning_rate": 8.796755473672847e-06,
+      "loss": 1.6301,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5524283507808477,
+      "grad_norm": 3.3566016376397108,
+      "learning_rate": 8.791237670111328e-06,
+      "loss": 1.4386,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5525999656770207,
+      "grad_norm": 6.89511767181731,
+      "learning_rate": 8.785720240003146e-06,
+      "loss": 1.6216,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5527715805731938,
+      "grad_norm": 3.941630622197514,
+      "learning_rate": 8.780203185052942e-06,
+      "loss": 1.8027,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5529431954693668,
+      "grad_norm": 3.9660821310197085,
+      "learning_rate": 8.774686506965237e-06,
+      "loss": 1.7754,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5531148103655398,
+      "grad_norm": 4.357534638582533,
+      "learning_rate": 8.769170207444432e-06,
+      "loss": 1.5571,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5532864252617127,
+      "grad_norm": 4.532355692806929,
+      "learning_rate": 8.763654288194823e-06,
+      "loss": 1.7203,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5534580401578857,
+      "grad_norm": 4.529123598027301,
+      "learning_rate": 8.758138750920577e-06,
+      "loss": 1.4378,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5536296550540587,
+      "grad_norm": 3.1250973058988665,
+      "learning_rate": 8.752623597325752e-06,
+      "loss": 1.4661,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5538012699502317,
+      "grad_norm": 3.2530388126661824,
+      "learning_rate": 8.747108829114284e-06,
+      "loss": 1.7626,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5539728848464047,
+      "grad_norm": 4.385300760918265,
+      "learning_rate": 8.741594447989981e-06,
+      "loss": 1.6426,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5541444997425776,
+      "grad_norm": 5.252514122415893,
+      "learning_rate": 8.736080455656545e-06,
+      "loss": 1.6596,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5543161146387506,
+      "grad_norm": 3.227288544320505,
+      "learning_rate": 8.73056685381755e-06,
+      "loss": 1.5413,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5544877295349236,
+      "grad_norm": 3.5311116481839115,
+      "learning_rate": 8.725053644176456e-06,
+      "loss": 1.6974,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5546593444310967,
+      "grad_norm": 3.416505908870069,
+      "learning_rate": 8.719540828436593e-06,
+      "loss": 1.7293,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5548309593272696,
+      "grad_norm": 3.457827629344126,
+      "learning_rate": 8.714028408301178e-06,
+      "loss": 1.7514,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5550025742234426,
+      "grad_norm": 3.145198745548947,
+      "learning_rate": 8.708516385473298e-06,
+      "loss": 1.6965,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5551741891196156,
+      "grad_norm": 2.9496799403656104,
+      "learning_rate": 8.703004761655918e-06,
+      "loss": 1.5921,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5553458040157886,
+      "grad_norm": 3.2901368497468213,
+      "learning_rate": 8.69749353855189e-06,
+      "loss": 1.661,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5555174189119616,
+      "grad_norm": 3.2421031810240613,
+      "learning_rate": 8.691982717863928e-06,
+      "loss": 1.4848,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5556890338081345,
+      "grad_norm": 3.4230864183063145,
+      "learning_rate": 8.686472301294633e-06,
+      "loss": 1.5485,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5558606487043075,
+      "grad_norm": 4.018046537000868,
+      "learning_rate": 8.680962290546477e-06,
+      "loss": 1.5272,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5560322636004805,
+      "grad_norm": 2.9049361300802845,
+      "learning_rate": 8.6754526873218e-06,
+      "loss": 1.3446,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5562038784966535,
+      "grad_norm": 3.341568161747712,
+      "learning_rate": 8.66994349332283e-06,
+      "loss": 1.5054,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5563754933928265,
+      "grad_norm": 3.941577393962539,
+      "learning_rate": 8.664434710251654e-06,
+      "loss": 1.6437,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5565471082889994,
+      "grad_norm": 3.105738291948784,
+      "learning_rate": 8.658926339810242e-06,
+      "loss": 1.5786,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5567187231851725,
+      "grad_norm": 3.074318476874637,
+      "learning_rate": 8.653418383700433e-06,
+      "loss": 1.4315,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5568903380813455,
+      "grad_norm": 3.716842029174973,
+      "learning_rate": 8.647910843623943e-06,
+      "loss": 1.7901,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5570619529775185,
+      "grad_norm": 3.3104958528902366,
+      "learning_rate": 8.642403721282348e-06,
+      "loss": 1.6603,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5572335678736914,
+      "grad_norm": 3.3367371036361493,
+      "learning_rate": 8.636897018377104e-06,
+      "loss": 1.6565,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5574051827698644,
+      "grad_norm": 3.0095529205717897,
+      "learning_rate": 8.631390736609535e-06,
+      "loss": 1.345,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5575767976660374,
+      "grad_norm": 3.2369329555542214,
+      "learning_rate": 8.625884877680835e-06,
+      "loss": 1.5876,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5577484125622104,
+      "grad_norm": 2.926198446720704,
+      "learning_rate": 8.620379443292068e-06,
+      "loss": 1.6695,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5579200274583834,
+      "grad_norm": 3.261907790467902,
+      "learning_rate": 8.61487443514417e-06,
+      "loss": 1.6412,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5580916423545563,
+      "grad_norm": 3.367924205214572,
+      "learning_rate": 8.609369854937934e-06,
+      "loss": 1.8815,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5582632572507293,
+      "grad_norm": 3.034778071252895,
+      "learning_rate": 8.603865704374032e-06,
+      "loss": 1.3463,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5584348721469024,
+      "grad_norm": 3.71398727943537,
+      "learning_rate": 8.598361985153e-06,
+      "loss": 1.8752,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5586064870430754,
+      "grad_norm": 2.901154887769359,
+      "learning_rate": 8.592858698975241e-06,
+      "loss": 1.5837,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5587781019392484,
+      "grad_norm": 4.212818273348361,
+      "learning_rate": 8.58735584754102e-06,
+      "loss": 1.7885,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5589497168354213,
+      "grad_norm": 3.057776985604256,
+      "learning_rate": 8.58185343255048e-06,
+      "loss": 1.4712,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5591213317315943,
+      "grad_norm": 3.2238568437144837,
+      "learning_rate": 8.576351455703611e-06,
+      "loss": 1.5019,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5592929466277673,
+      "grad_norm": 3.7427697564934923,
+      "learning_rate": 8.570849918700278e-06,
+      "loss": 1.505,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5594645615239403,
+      "grad_norm": 3.6361104250789005,
+      "learning_rate": 8.565348823240214e-06,
+      "loss": 1.7453,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5596361764201133,
+      "grad_norm": 3.9375479659544106,
+      "learning_rate": 8.559848171023006e-06,
+      "loss": 1.6328,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5598077913162862,
+      "grad_norm": 3.0915534450178033,
+      "learning_rate": 8.554347963748113e-06,
+      "loss": 1.7329,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5599794062124592,
+      "grad_norm": 2.696302089915293,
+      "learning_rate": 8.548848203114853e-06,
+      "loss": 1.4697,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5601510211086322,
+      "grad_norm": 3.508385094337635,
+      "learning_rate": 8.543348890822399e-06,
+      "loss": 1.7017,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5603226360048053,
+      "grad_norm": 3.297750010295088,
+      "learning_rate": 8.537850028569796e-06,
+      "loss": 1.7364,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5604942509009782,
+      "grad_norm": 3.0568413028715047,
+      "learning_rate": 8.532351618055944e-06,
+      "loss": 1.751,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5606658657971512,
+      "grad_norm": 3.7617092670933165,
+      "learning_rate": 8.526853660979609e-06,
+      "loss": 1.8571,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5608374806933242,
+      "grad_norm": 3.7064590959704242,
+      "learning_rate": 8.521356159039407e-06,
+      "loss": 1.5883,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5610090955894972,
+      "grad_norm": 3.035057604090649,
+      "learning_rate": 8.515859113933828e-06,
+      "loss": 1.4653,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5611807104856702,
+      "grad_norm": 3.4457975765676623,
+      "learning_rate": 8.510362527361205e-06,
+      "loss": 1.4828,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5613523253818431,
+      "grad_norm": 2.87825553417923,
+      "learning_rate": 8.504866401019738e-06,
+      "loss": 1.5598,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5615239402780161,
+      "grad_norm": 3.2036235332688365,
+      "learning_rate": 8.499370736607486e-06,
+      "loss": 1.6237,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5616955551741891,
+      "grad_norm": 3.1444165062997635,
+      "learning_rate": 8.49387553582236e-06,
+      "loss": 1.446,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5618671700703621,
+      "grad_norm": 3.9360624692646167,
+      "learning_rate": 8.488380800362133e-06,
+      "loss": 1.6504,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5620387849665351,
+      "grad_norm": 3.051569389324742,
+      "learning_rate": 8.482886531924434e-06,
+      "loss": 1.7111,
+      "step": 3275
+    },
+    {
+      "epoch": 0.562210399862708,
+      "grad_norm": 3.1815340298388,
+      "learning_rate": 8.477392732206738e-06,
+      "loss": 1.723,
+      "step": 3276
+    },
+    {
+      "epoch": 0.562382014758881,
+      "grad_norm": 3.319420710600917,
+      "learning_rate": 8.471899402906387e-06,
+      "loss": 1.8347,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5625536296550541,
+      "grad_norm": 2.7596499428257726,
+      "learning_rate": 8.466406545720572e-06,
+      "loss": 1.5249,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5627252445512271,
+      "grad_norm": 3.1199004761412286,
+      "learning_rate": 8.460914162346343e-06,
+      "loss": 1.7499,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5628968594474001,
+      "grad_norm": 4.087261083242442,
+      "learning_rate": 8.455422254480597e-06,
+      "loss": 1.6515,
+      "step": 3280
+    },
+    {
+      "epoch": 0.563068474343573,
+      "grad_norm": 4.252625514188772,
+      "learning_rate": 8.44993082382009e-06,
+      "loss": 1.5245,
+      "step": 3281
+    },
+    {
+      "epoch": 0.563240089239746,
+      "grad_norm": 4.070922414145992,
+      "learning_rate": 8.444439872061424e-06,
+      "loss": 1.3844,
+      "step": 3282
+    },
+    {
+      "epoch": 0.563411704135919,
+      "grad_norm": 3.1588998630830547,
+      "learning_rate": 8.438949400901057e-06,
+      "loss": 1.4534,
+      "step": 3283
+    },
+    {
+      "epoch": 0.563583319032092,
+      "grad_norm": 2.9017090307147098,
+      "learning_rate": 8.4334594120353e-06,
+      "loss": 1.4493,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5637549339282649,
+      "grad_norm": 3.2987433412615177,
+      "learning_rate": 8.427969907160314e-06,
+      "loss": 1.6449,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5639265488244379,
+      "grad_norm": 3.2903593388123027,
+      "learning_rate": 8.42248088797211e-06,
+      "loss": 1.6136,
+      "step": 3286
+    },
+    {
+      "epoch": 0.564098163720611,
+      "grad_norm": 3.3124787390674393,
+      "learning_rate": 8.416992356166549e-06,
+      "loss": 1.3647,
+      "step": 3287
+    },
+    {
+      "epoch": 0.564269778616784,
+      "grad_norm": 3.0727836880073385,
+      "learning_rate": 8.411504313439335e-06,
+      "loss": 1.4165,
+      "step": 3288
+    },
+    {
+      "epoch": 0.564441393512957,
+      "grad_norm": 2.58661432877082,
+      "learning_rate": 8.406016761486034e-06,
+      "loss": 1.4452,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5646130084091299,
+      "grad_norm": 3.1164017370539896,
+      "learning_rate": 8.400529702002048e-06,
+      "loss": 1.3702,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5647846233053029,
+      "grad_norm": 3.4437558929796213,
+      "learning_rate": 8.395043136682638e-06,
+      "loss": 1.7273,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5649562382014759,
+      "grad_norm": 3.843491971675913,
+      "learning_rate": 8.389557067222902e-06,
+      "loss": 1.5507,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5651278530976489,
+      "grad_norm": 3.8316096131970228,
+      "learning_rate": 8.384071495317795e-06,
+      "loss": 1.4862,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5652994679938219,
+      "grad_norm": 2.797524656384283,
+      "learning_rate": 8.378586422662107e-06,
+      "loss": 1.5572,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5654710828899948,
+      "grad_norm": 3.212208854813479,
+      "learning_rate": 8.37310185095048e-06,
+      "loss": 1.5172,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5656426977861678,
+      "grad_norm": 3.3748620111847054,
+      "learning_rate": 8.367617781877403e-06,
+      "loss": 1.5082,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5658143126823408,
+      "grad_norm": 5.338641434217973,
+      "learning_rate": 8.362134217137205e-06,
+      "loss": 1.375,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5659859275785138,
+      "grad_norm": 3.439008543231657,
+      "learning_rate": 8.356651158424064e-06,
+      "loss": 1.6181,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5661575424746867,
+      "grad_norm": 3.6946178808177086,
+      "learning_rate": 8.351168607432e-06,
+      "loss": 1.614,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5663291573708598,
+      "grad_norm": 2.619305572926834,
+      "learning_rate": 8.345686565854869e-06,
+      "loss": 1.4646,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5665007722670328,
+      "grad_norm": 6.202803906122431,
+      "learning_rate": 8.340205035386384e-06,
+      "loss": 1.7143,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5666723871632058,
+      "grad_norm": 3.067596412776868,
+      "learning_rate": 8.334724017720084e-06,
+      "loss": 1.3359,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5668440020593788,
+      "grad_norm": 3.055630750770194,
+      "learning_rate": 8.329243514549365e-06,
+      "loss": 1.5949,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5670156169555517,
+      "grad_norm": 5.347367936172074,
+      "learning_rate": 8.323763527567451e-06,
+      "loss": 1.6535,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5671872318517247,
+      "grad_norm": 3.7616366234347898,
+      "learning_rate": 8.318284058467419e-06,
+      "loss": 1.497,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5673588467478977,
+      "grad_norm": 4.113971233573492,
+      "learning_rate": 8.312805108942172e-06,
+      "loss": 1.8644,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5675304616440707,
+      "grad_norm": 3.723546907663054,
+      "learning_rate": 8.30732668068446e-06,
+      "loss": 1.4669,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5677020765402437,
+      "grad_norm": 3.361542419482842,
+      "learning_rate": 8.301848775386878e-06,
+      "loss": 1.8408,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5678736914364166,
+      "grad_norm": 3.8728466920005866,
+      "learning_rate": 8.296371394741846e-06,
+      "loss": 1.7269,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5680453063325897,
+      "grad_norm": 3.6652752949819853,
+      "learning_rate": 8.290894540441633e-06,
+      "loss": 1.5592,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5682169212287627,
+      "grad_norm": 3.2717554817030634,
+      "learning_rate": 8.285418214178343e-06,
+      "loss": 1.7266,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5683885361249357,
+      "grad_norm": 3.2018994610917533,
+      "learning_rate": 8.279942417643908e-06,
+      "loss": 1.6679,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5685601510211087,
+      "grad_norm": 3.2982258673231546,
+      "learning_rate": 8.27446715253011e-06,
+      "loss": 1.6281,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5687317659172816,
+      "grad_norm": 2.8488166862204793,
+      "learning_rate": 8.268992420528556e-06,
+      "loss": 1.5427,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5689033808134546,
+      "grad_norm": 2.840911324814438,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 1.3772,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5690749957096276,
+      "grad_norm": 3.551932063297759,
+      "learning_rate": 8.25804456262781e-06,
+      "loss": 1.436,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5692466106058006,
+      "grad_norm": 3.259133958841487,
+      "learning_rate": 8.252571440111018e-06,
+      "loss": 1.5072,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5694182255019735,
+      "grad_norm": 2.7811652539634477,
+      "learning_rate": 8.247098857471262e-06,
+      "loss": 1.4575,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5695898403981465,
+      "grad_norm": 3.530250755015589,
+      "learning_rate": 8.241626816399325e-06,
+      "loss": 1.4875,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5697614552943195,
+      "grad_norm": 4.098755890991295,
+      "learning_rate": 8.236155318585828e-06,
+      "loss": 1.4714,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5699330701904926,
+      "grad_norm": 3.5724158873326313,
+      "learning_rate": 8.230684365721212e-06,
+      "loss": 1.7572,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5701046850866656,
+      "grad_norm": 5.678077705744792,
+      "learning_rate": 8.225213959495762e-06,
+      "loss": 1.5927,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5702762999828385,
+      "grad_norm": 3.2149107406497612,
+      "learning_rate": 8.219744101599588e-06,
+      "loss": 1.4137,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5704479148790115,
+      "grad_norm": 3.7739368050624384,
+      "learning_rate": 8.214274793722622e-06,
+      "loss": 1.5676,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5706195297751845,
+      "grad_norm": 3.21248624627995,
+      "learning_rate": 8.208806037554645e-06,
+      "loss": 1.5946,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5707911446713575,
+      "grad_norm": 2.9103287476101745,
+      "learning_rate": 8.203337834785252e-06,
+      "loss": 1.5727,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5709627595675305,
+      "grad_norm": 3.0598952927149305,
+      "learning_rate": 8.197870187103874e-06,
+      "loss": 1.607,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5711343744637034,
+      "grad_norm": 3.0999520615062335,
+      "learning_rate": 8.192403096199768e-06,
+      "loss": 1.6799,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5713059893598764,
+      "grad_norm": 3.6469921226967656,
+      "learning_rate": 8.186936563762022e-06,
+      "loss": 1.6337,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5714776042560494,
+      "grad_norm": 3.1784677481318413,
+      "learning_rate": 8.181470591479549e-06,
+      "loss": 1.4632,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5716492191522224,
+      "grad_norm": 3.1199228835907253,
+      "learning_rate": 8.176005181041083e-06,
+      "loss": 1.3285,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5718208340483955,
+      "grad_norm": 3.5270143050902654,
+      "learning_rate": 8.170540334135197e-06,
+      "loss": 1.6874,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5719924489445684,
+      "grad_norm": 2.7322996284646033,
+      "learning_rate": 8.165076052450281e-06,
+      "loss": 1.4953,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5721640638407414,
+      "grad_norm": 3.2270146724477557,
+      "learning_rate": 8.159612337674557e-06,
+      "loss": 1.5549,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5723356787369144,
+      "grad_norm": 3.6229730060018284,
+      "learning_rate": 8.154149191496066e-06,
+      "loss": 1.6531,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5725072936330874,
+      "grad_norm": 2.721936439367243,
+      "learning_rate": 8.14868661560267e-06,
+      "loss": 1.4857,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5726789085292603,
+      "grad_norm": 3.881874850935796,
+      "learning_rate": 8.143224611682067e-06,
+      "loss": 1.7753,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5728505234254333,
+      "grad_norm": 3.4400240710189616,
+      "learning_rate": 8.137763181421764e-06,
+      "loss": 1.6223,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5730221383216063,
+      "grad_norm": 4.0946563294590135,
+      "learning_rate": 8.132302326509104e-06,
+      "loss": 1.6434,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5731937532177793,
+      "grad_norm": 3.5743568567005046,
+      "learning_rate": 8.126842048631244e-06,
+      "loss": 1.6623,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5733653681139523,
+      "grad_norm": 2.7374286604824705,
+      "learning_rate": 8.12138234947517e-06,
+      "loss": 1.4436,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5735369830101252,
+      "grad_norm": 3.1525846004358136,
+      "learning_rate": 8.115923230727674e-06,
+      "loss": 1.5641,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5737085979062982,
+      "grad_norm": 3.1002395847941906,
+      "learning_rate": 8.110464694075383e-06,
+      "loss": 1.6402,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5738802128024713,
+      "grad_norm": 3.8447138292190424,
+      "learning_rate": 8.105006741204745e-06,
+      "loss": 1.623,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5740518276986443,
+      "grad_norm": 3.041107744210225,
+      "learning_rate": 8.099549373802014e-06,
+      "loss": 1.6788,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5742234425948173,
+      "grad_norm": 2.895810481490591,
+      "learning_rate": 8.09409259355328e-06,
+      "loss": 1.4531,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5743950574909902,
+      "grad_norm": 2.8138667535010353,
+      "learning_rate": 8.088636402144442e-06,
+      "loss": 1.6378,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5745666723871632,
+      "grad_norm": 3.9127809164321623,
+      "learning_rate": 8.08318080126121e-06,
+      "loss": 1.6785,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5747382872833362,
+      "grad_norm": 2.9350546619628197,
+      "learning_rate": 8.077725792589131e-06,
+      "loss": 1.5561,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5749099021795092,
+      "grad_norm": 3.093616739657815,
+      "learning_rate": 8.072271377813548e-06,
+      "loss": 1.5094,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5750815170756821,
+      "grad_norm": 3.1588668453140842,
+      "learning_rate": 8.06681755861964e-06,
+      "loss": 1.452,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5752531319718551,
+      "grad_norm": 3.5317660270907942,
+      "learning_rate": 8.061364336692385e-06,
+      "loss": 1.6715,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5754247468680281,
+      "grad_norm": 3.927847784966835,
+      "learning_rate": 8.05591171371659e-06,
+      "loss": 1.6801,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5755963617642011,
+      "grad_norm": 3.932959669075851,
+      "learning_rate": 8.050459691376867e-06,
+      "loss": 1.6584,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5757679766603742,
+      "grad_norm": 3.39440898968657,
+      "learning_rate": 8.045008271357644e-06,
+      "loss": 1.6142,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5759395915565471,
+      "grad_norm": 3.3611797623361914,
+      "learning_rate": 8.039557455343172e-06,
+      "loss": 1.7645,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5761112064527201,
+      "grad_norm": 3.1520198245820747,
+      "learning_rate": 8.034107245017501e-06,
+      "loss": 1.6665,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5762828213488931,
+      "grad_norm": 2.9756959283566897,
+      "learning_rate": 8.028657642064509e-06,
+      "loss": 1.4583,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5764544362450661,
+      "grad_norm": 3.205094669041891,
+      "learning_rate": 8.023208648167877e-06,
+      "loss": 1.5664,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5766260511412391,
+      "grad_norm": 4.457442925545306,
+      "learning_rate": 8.017760265011093e-06,
+      "loss": 1.6954,
+      "step": 3360
+    },
+    {
+      "epoch": 0.576797666037412,
+      "grad_norm": 3.4469412796953236,
+      "learning_rate": 8.012312494277473e-06,
+      "loss": 1.7256,
+      "step": 3361
+    },
+    {
+      "epoch": 0.576969280933585,
+      "grad_norm": 3.288181659572955,
+      "learning_rate": 8.006865337650124e-06,
+      "loss": 1.5143,
+      "step": 3362
+    },
+    {
+      "epoch": 0.577140895829758,
+      "grad_norm": 3.0880996791979474,
+      "learning_rate": 8.001418796811981e-06,
+      "loss": 1.602,
+      "step": 3363
+    },
+    {
+      "epoch": 0.577312510725931,
+      "grad_norm": 3.7844470543699598,
+      "learning_rate": 7.995972873445775e-06,
+      "loss": 1.8454,
+      "step": 3364
+    },
+    {
+      "epoch": 0.577484125622104,
+      "grad_norm": 3.707990830553102,
+      "learning_rate": 7.99052756923406e-06,
+      "loss": 1.6903,
+      "step": 3365
+    },
+    {
+      "epoch": 0.577655740518277,
+      "grad_norm": 3.102049014407579,
+      "learning_rate": 7.985082885859184e-06,
+      "loss": 1.6468,
+      "step": 3366
+    },
+    {
+      "epoch": 0.57782735541445,
+      "grad_norm": 3.059964517361926,
+      "learning_rate": 7.979638825003307e-06,
+      "loss": 1.7285,
+      "step": 3367
+    },
+    {
+      "epoch": 0.577998970310623,
+      "grad_norm": 3.0880975662217813,
+      "learning_rate": 7.974195388348406e-06,
+      "loss": 1.5663,
+      "step": 3368
+    },
+    {
+      "epoch": 0.578170585206796,
+      "grad_norm": 2.8394638206491982,
+      "learning_rate": 7.968752577576257e-06,
+      "loss": 1.5792,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5783422001029689,
+      "grad_norm": 2.811150670390754,
+      "learning_rate": 7.963310394368445e-06,
+      "loss": 1.5458,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5785138149991419,
+      "grad_norm": 2.757524937848895,
+      "learning_rate": 7.957868840406362e-06,
+      "loss": 1.4689,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5786854298953149,
+      "grad_norm": 3.6727272333159813,
+      "learning_rate": 7.952427917371198e-06,
+      "loss": 1.6723,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5788570447914879,
+      "grad_norm": 3.300240266782386,
+      "learning_rate": 7.946987626943958e-06,
+      "loss": 1.6598,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5790286596876609,
+      "grad_norm": 3.5740827690897934,
+      "learning_rate": 7.941547970805447e-06,
+      "loss": 1.6765,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5792002745838338,
+      "grad_norm": 3.322321882478033,
+      "learning_rate": 7.936108950636278e-06,
+      "loss": 1.8564,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5793718894800068,
+      "grad_norm": 3.4869863184006102,
+      "learning_rate": 7.930670568116858e-06,
+      "loss": 1.581,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5795435043761799,
+      "grad_norm": 3.9038771561881522,
+      "learning_rate": 7.925232824927411e-06,
+      "loss": 1.6071,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5797151192723529,
+      "grad_norm": 2.977760482414295,
+      "learning_rate": 7.91979572274795e-06,
+      "loss": 1.6114,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5798867341685259,
+      "grad_norm": 3.5577265723671947,
+      "learning_rate": 7.914359263258295e-06,
+      "loss": 1.8329,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5800583490646988,
+      "grad_norm": 4.310151594305367,
+      "learning_rate": 7.908923448138072e-06,
+      "loss": 1.7581,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5802299639608718,
+      "grad_norm": 3.532035262433165,
+      "learning_rate": 7.9034882790667e-06,
+      "loss": 1.5501,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5804015788570448,
+      "grad_norm": 3.620376166544094,
+      "learning_rate": 7.898053757723409e-06,
+      "loss": 1.793,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5805731937532178,
+      "grad_norm": 3.6336384949403038,
+      "learning_rate": 7.892619885787219e-06,
+      "loss": 1.5072,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5807448086493908,
+      "grad_norm": 3.431998543591075,
+      "learning_rate": 7.887186664936945e-06,
+      "loss": 1.6605,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5809164235455637,
+      "grad_norm": 3.5446480844974495,
+      "learning_rate": 7.88175409685122e-06,
+      "loss": 1.7385,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5810880384417367,
+      "grad_norm": 3.4525634624061174,
+      "learning_rate": 7.876322183208456e-06,
+      "loss": 1.4408,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5812596533379097,
+      "grad_norm": 3.3098614633002734,
+      "learning_rate": 7.870890925686875e-06,
+      "loss": 1.6798,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5814312682340828,
+      "grad_norm": 4.078781283852957,
+      "learning_rate": 7.86546032596449e-06,
+      "loss": 1.5687,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5816028831302557,
+      "grad_norm": 3.3484219590109903,
+      "learning_rate": 7.860030385719119e-06,
+      "loss": 1.6588,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5817744980264287,
+      "grad_norm": 3.6308744165007396,
+      "learning_rate": 7.85460110662836e-06,
+      "loss": 1.4102,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5819461129226017,
+      "grad_norm": 3.19022682099308,
+      "learning_rate": 7.849172490369619e-06,
+      "loss": 1.6959,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5821177278187747,
+      "grad_norm": 3.694857470540594,
+      "learning_rate": 7.8437445386201e-06,
+      "loss": 1.6795,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5822893427149477,
+      "grad_norm": 4.43058051589044,
+      "learning_rate": 7.83831725305679e-06,
+      "loss": 1.5871,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5824609576111206,
+      "grad_norm": 3.769675050349629,
+      "learning_rate": 7.832890635356485e-06,
+      "loss": 1.5692,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5826325725072936,
+      "grad_norm": 3.158537183662451,
+      "learning_rate": 7.827464687195762e-06,
+      "loss": 1.8208,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5828041874034666,
+      "grad_norm": 3.572121332569598,
+      "learning_rate": 7.822039410250993e-06,
+      "loss": 1.8154,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5829758022996396,
+      "grad_norm": 4.854917811473112,
+      "learning_rate": 7.816614806198347e-06,
+      "loss": 1.57,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5831474171958126,
+      "grad_norm": 3.5679181906198196,
+      "learning_rate": 7.811190876713783e-06,
+      "loss": 1.6763,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5833190320919855,
+      "grad_norm": 3.4176152561497655,
+      "learning_rate": 7.805767623473055e-06,
+      "loss": 1.5826,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5834906469881586,
+      "grad_norm": 3.450280563128342,
+      "learning_rate": 7.8003450481517e-06,
+      "loss": 1.5395,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5836622618843316,
+      "grad_norm": 2.9783290606056356,
+      "learning_rate": 7.794923152425058e-06,
+      "loss": 1.4799,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5838338767805046,
+      "grad_norm": 3.420392915531139,
+      "learning_rate": 7.789501937968246e-06,
+      "loss": 1.5855,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5840054916766776,
+      "grad_norm": 3.1721732430698566,
+      "learning_rate": 7.784081406456173e-06,
+      "loss": 1.3177,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5841771065728505,
+      "grad_norm": 3.2696159478152915,
+      "learning_rate": 7.778661559563547e-06,
+      "loss": 1.4352,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5843487214690235,
+      "grad_norm": 3.207275864667234,
+      "learning_rate": 7.773242398964851e-06,
+      "loss": 1.7167,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5845203363651965,
+      "grad_norm": 3.456157376810977,
+      "learning_rate": 7.76782392633437e-06,
+      "loss": 1.6332,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5846919512613695,
+      "grad_norm": 3.739277129896731,
+      "learning_rate": 7.76240614334617e-06,
+      "loss": 1.4852,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5848635661575424,
+      "grad_norm": 3.5831469186254328,
+      "learning_rate": 7.756989051674089e-06,
+      "loss": 1.3861,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5850351810537154,
+      "grad_norm": 3.4377752647673323,
+      "learning_rate": 7.751572652991777e-06,
+      "loss": 1.4595,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5852067959498884,
+      "grad_norm": 4.00322428815339,
+      "learning_rate": 7.746156948972654e-06,
+      "loss": 1.6593,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5853784108460615,
+      "grad_norm": 3.869745920018589,
+      "learning_rate": 7.740741941289936e-06,
+      "loss": 1.5643,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5855500257422345,
+      "grad_norm": 3.7251467927168904,
+      "learning_rate": 7.735327631616606e-06,
+      "loss": 1.6529,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5857216406384074,
+      "grad_norm": 3.4462693201149777,
+      "learning_rate": 7.729914021625458e-06,
+      "loss": 1.57,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5858932555345804,
+      "grad_norm": 3.3640653043018562,
+      "learning_rate": 7.724501112989042e-06,
+      "loss": 1.5265,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5860648704307534,
+      "grad_norm": 3.698114510919301,
+      "learning_rate": 7.719088907379705e-06,
+      "loss": 1.7258,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5862364853269264,
+      "grad_norm": 2.659080367977588,
+      "learning_rate": 7.713677406469582e-06,
+      "loss": 1.5443,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5864081002230994,
+      "grad_norm": 3.5203753304916994,
+      "learning_rate": 7.70826661193058e-06,
+      "loss": 1.7366,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5865797151192723,
+      "grad_norm": 3.229286781162943,
+      "learning_rate": 7.702856525434393e-06,
+      "loss": 1.3902,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5867513300154453,
+      "grad_norm": 4.196326915897818,
+      "learning_rate": 7.697447148652499e-06,
+      "loss": 1.6128,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5869229449116183,
+      "grad_norm": 3.6892541904442786,
+      "learning_rate": 7.692038483256141e-06,
+      "loss": 1.6386,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5870945598077913,
+      "grad_norm": 3.6859846651088453,
+      "learning_rate": 7.686630530916366e-06,
+      "loss": 1.6677,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5872661747039642,
+      "grad_norm": 3.7829443629407864,
+      "learning_rate": 7.68122329330398e-06,
+      "loss": 1.5093,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5874377896001373,
+      "grad_norm": 3.7666023151534427,
+      "learning_rate": 7.675816772089585e-06,
+      "loss": 1.6995,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5876094044963103,
+      "grad_norm": 3.1284984931329074,
+      "learning_rate": 7.670410968943548e-06,
+      "loss": 1.9365,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5877810193924833,
+      "grad_norm": 3.4108951569360997,
+      "learning_rate": 7.665005885536023e-06,
+      "loss": 1.6649,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5879526342886563,
+      "grad_norm": 3.2690888979237256,
+      "learning_rate": 7.659601523536936e-06,
+      "loss": 1.7073,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5881242491848292,
+      "grad_norm": 3.1997678158040963,
+      "learning_rate": 7.654197884615991e-06,
+      "loss": 1.759,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5882958640810022,
+      "grad_norm": 4.062338122480571,
+      "learning_rate": 7.648794970442674e-06,
+      "loss": 1.5379,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5884674789771752,
+      "grad_norm": 3.2508313105601325,
+      "learning_rate": 7.643392782686238e-06,
+      "loss": 1.5833,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5886390938733482,
+      "grad_norm": 3.293012414249073,
+      "learning_rate": 7.637991323015722e-06,
+      "loss": 1.6647,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5888107087695212,
+      "grad_norm": 2.861860803830219,
+      "learning_rate": 7.632590593099935e-06,
+      "loss": 1.5197,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5889823236656941,
+      "grad_norm": 5.260677730440887,
+      "learning_rate": 7.627190594607454e-06,
+      "loss": 1.6689,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5891539385618672,
+      "grad_norm": 3.7718797264873456,
+      "learning_rate": 7.621791329206643e-06,
+      "loss": 1.6436,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5893255534580402,
+      "grad_norm": 3.5756958234984944,
+      "learning_rate": 7.616392798565626e-06,
+      "loss": 1.7052,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5894971683542132,
+      "grad_norm": 3.3746200657836485,
+      "learning_rate": 7.610995004352313e-06,
+      "loss": 1.3705,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5896687832503862,
+      "grad_norm": 4.025671863494862,
+      "learning_rate": 7.605597948234377e-06,
+      "loss": 1.7638,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5898403981465591,
+      "grad_norm": 3.2096506433161225,
+      "learning_rate": 7.60020163187927e-06,
+      "loss": 1.5396,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5900120130427321,
+      "grad_norm": 3.641887705980669,
+      "learning_rate": 7.5948060569542085e-06,
+      "loss": 1.4779,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5901836279389051,
+      "grad_norm": 5.191420711528766,
+      "learning_rate": 7.589411225126181e-06,
+      "loss": 1.4139,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5903552428350781,
+      "grad_norm": 3.541926464478629,
+      "learning_rate": 7.584017138061952e-06,
+      "loss": 1.5846,
+      "step": 3440
+    },
+    {
+      "epoch": 0.590526857731251,
+      "grad_norm": 3.299448765178323,
+      "learning_rate": 7.578623797428051e-06,
+      "loss": 1.7244,
+      "step": 3441
+    },
+    {
+      "epoch": 0.590698472627424,
+      "grad_norm": 3.2904805276877,
+      "learning_rate": 7.57323120489078e-06,
+      "loss": 1.3071,
+      "step": 3442
+    },
+    {
+      "epoch": 0.590870087523597,
+      "grad_norm": 3.4367547328660417,
+      "learning_rate": 7.56783936211621e-06,
+      "loss": 1.4949,
+      "step": 3443
+    },
+    {
+      "epoch": 0.59104170241977,
+      "grad_norm": 9.696886988735994,
+      "learning_rate": 7.56244827077017e-06,
+      "loss": 1.8448,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5912133173159431,
+      "grad_norm": 3.5865939085980187,
+      "learning_rate": 7.557057932518274e-06,
+      "loss": 1.6688,
+      "step": 3445
+    },
+    {
+      "epoch": 0.591384932212116,
+      "grad_norm": 3.200773303057205,
+      "learning_rate": 7.551668349025889e-06,
+      "loss": 1.7322,
+      "step": 3446
+    },
+    {
+      "epoch": 0.591556547108289,
+      "grad_norm": 3.4087785819785448,
+      "learning_rate": 7.54627952195816e-06,
+      "loss": 1.7338,
+      "step": 3447
+    },
+    {
+      "epoch": 0.591728162004462,
+      "grad_norm": 3.1826063988764064,
+      "learning_rate": 7.540891452979988e-06,
+      "loss": 1.6004,
+      "step": 3448
+    },
+    {
+      "epoch": 0.591899776900635,
+      "grad_norm": 3.394885182311321,
+      "learning_rate": 7.53550414375605e-06,
+      "loss": 1.5406,
+      "step": 3449
+    },
+    {
+      "epoch": 0.592071391796808,
+      "grad_norm": 3.7271835868720076,
+      "learning_rate": 7.530117595950779e-06,
+      "loss": 1.5926,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5922430066929809,
+      "grad_norm": 3.2447524485665995,
+      "learning_rate": 7.524731811228374e-06,
+      "loss": 1.4928,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5924146215891539,
+      "grad_norm": 3.727512721432605,
+      "learning_rate": 7.5193467912528065e-06,
+      "loss": 1.6125,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5925862364853269,
+      "grad_norm": 4.388206109404833,
+      "learning_rate": 7.513962537687801e-06,
+      "loss": 1.8134,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5927578513815,
+      "grad_norm": 3.076136529734564,
+      "learning_rate": 7.508579052196856e-06,
+      "loss": 1.4729,
+      "step": 3454
+    },
+    {
+      "epoch": 0.592929466277673,
+      "grad_norm": 3.090441968248904,
+      "learning_rate": 7.503196336443226e-06,
+      "loss": 1.5077,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5931010811738459,
+      "grad_norm": 3.9151173250992533,
+      "learning_rate": 7.497814392089921e-06,
+      "loss": 1.7779,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5932726960700189,
+      "grad_norm": 4.704923856248117,
+      "learning_rate": 7.492433220799731e-06,
+      "loss": 1.7743,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5934443109661919,
+      "grad_norm": 3.2743200284497607,
+      "learning_rate": 7.487052824235188e-06,
+      "loss": 1.4268,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5936159258623649,
+      "grad_norm": 3.090962800060612,
+      "learning_rate": 7.481673204058599e-06,
+      "loss": 1.6117,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5937875407585378,
+      "grad_norm": 3.2700323717936928,
+      "learning_rate": 7.476294361932022e-06,
+      "loss": 1.6494,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5939591556547108,
+      "grad_norm": 3.5217792540387625,
+      "learning_rate": 7.470916299517281e-06,
+      "loss": 1.7862,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5941307705508838,
+      "grad_norm": 3.958546529354251,
+      "learning_rate": 7.465539018475954e-06,
+      "loss": 1.5975,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5943023854470568,
+      "grad_norm": 3.1667882246080703,
+      "learning_rate": 7.460162520469379e-06,
+      "loss": 1.8509,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5944740003432298,
+      "grad_norm": 3.4084259635303886,
+      "learning_rate": 7.454786807158654e-06,
+      "loss": 1.7569,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5946456152394027,
+      "grad_norm": 3.4134509716830643,
+      "learning_rate": 7.449411880204629e-06,
+      "loss": 1.4558,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5948172301355757,
+      "grad_norm": 4.9449335423127065,
+      "learning_rate": 7.444037741267923e-06,
+      "loss": 1.4374,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5949888450317488,
+      "grad_norm": 3.1289839977494096,
+      "learning_rate": 7.438664392008903e-06,
+      "loss": 1.6245,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5951604599279218,
+      "grad_norm": 3.44472725161979,
+      "learning_rate": 7.4332918340876834e-06,
+      "loss": 1.6394,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5953320748240948,
+      "grad_norm": 4.035093330546387,
+      "learning_rate": 7.427920069164154e-06,
+      "loss": 1.657,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5955036897202677,
+      "grad_norm": 3.064410566000389,
+      "learning_rate": 7.422549098897942e-06,
+      "loss": 1.6928,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5956753046164407,
+      "grad_norm": 3.6946001978065284,
+      "learning_rate": 7.417178924948442e-06,
+      "loss": 1.6947,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5958469195126137,
+      "grad_norm": 3.364638705208248,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 1.4862,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5960185344087867,
+      "grad_norm": 3.1498637832725933,
+      "learning_rate": 7.406440972635893e-06,
+      "loss": 1.5784,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5961901493049596,
+      "grad_norm": 3.3343464296396386,
+      "learning_rate": 7.401073197590394e-06,
+      "loss": 1.6015,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5963617642011326,
+      "grad_norm": 3.3871116094083633,
+      "learning_rate": 7.39570622549669e-06,
+      "loss": 1.6587,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5965333790973056,
+      "grad_norm": 2.961481816908742,
+      "learning_rate": 7.390340058012942e-06,
+      "loss": 1.4583,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5967049939934786,
+      "grad_norm": 3.0947770440813476,
+      "learning_rate": 7.384974696797051e-06,
+      "loss": 1.4569,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5968766088896517,
+      "grad_norm": 3.7040062861178584,
+      "learning_rate": 7.3796101435066744e-06,
+      "loss": 1.7992,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5970482237858246,
+      "grad_norm": 3.5236170954512605,
+      "learning_rate": 7.374246399799219e-06,
+      "loss": 1.5144,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5972198386819976,
+      "grad_norm": 3.4929880815605396,
+      "learning_rate": 7.368883467331842e-06,
+      "loss": 1.4671,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5973914535781706,
+      "grad_norm": 2.999185898087423,
+      "learning_rate": 7.3635213477614465e-06,
+      "loss": 1.5863,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5975630684743436,
+      "grad_norm": 3.2343396749847932,
+      "learning_rate": 7.358160042744684e-06,
+      "loss": 1.4851,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5977346833705166,
+      "grad_norm": 3.578878332256923,
+      "learning_rate": 7.352799553937966e-06,
+      "loss": 1.6599,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5979062982666895,
+      "grad_norm": 2.8562673552416773,
+      "learning_rate": 7.347439882997434e-06,
+      "loss": 1.5198,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5980779131628625,
+      "grad_norm": 3.370512546047015,
+      "learning_rate": 7.342081031578992e-06,
+      "loss": 1.4607,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5982495280590355,
+      "grad_norm": 3.6359043106964695,
+      "learning_rate": 7.336723001338286e-06,
+      "loss": 1.8209,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5984211429552085,
+      "grad_norm": 3.9482934730495627,
+      "learning_rate": 7.331365793930698e-06,
+      "loss": 1.6941,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5985927578513816,
+      "grad_norm": 2.9820353600099168,
+      "learning_rate": 7.326009411011373e-06,
+      "loss": 1.6194,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5987643727475545,
+      "grad_norm": 3.317409323950506,
+      "learning_rate": 7.3206538542351895e-06,
+      "loss": 1.6654,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5989359876437275,
+      "grad_norm": 3.845945578759833,
+      "learning_rate": 7.3152991252567765e-06,
+      "loss": 1.6838,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5991076025399005,
+      "grad_norm": 3.301791598658535,
+      "learning_rate": 7.309945225730502e-06,
+      "loss": 1.2897,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5992792174360735,
+      "grad_norm": 3.268855205397025,
+      "learning_rate": 7.3045921573104864e-06,
+      "loss": 1.6067,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5994508323322464,
+      "grad_norm": 2.86378842637836,
+      "learning_rate": 7.299239921650583e-06,
+      "loss": 1.3309,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5996224472284194,
+      "grad_norm": 3.3354861696207196,
+      "learning_rate": 7.29388852040439e-06,
+      "loss": 1.6204,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5997940621245924,
+      "grad_norm": 3.3282166085597242,
+      "learning_rate": 7.288537955225257e-06,
+      "loss": 1.5746,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5999656770207654,
+      "grad_norm": 3.246799848306246,
+      "learning_rate": 7.283188227766262e-06,
+      "loss": 1.5323,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6001372919169384,
+      "grad_norm": 4.2661936554700395,
+      "learning_rate": 7.2778393396802374e-06,
+      "loss": 1.6849,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6003089068131113,
+      "grad_norm": 3.5978067353137058,
+      "learning_rate": 7.272491292619747e-06,
+      "loss": 1.7872,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6004805217092843,
+      "grad_norm": 2.8790891648592285,
+      "learning_rate": 7.267144088237095e-06,
+      "loss": 1.5264,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6006521366054574,
+      "grad_norm": 3.120079277745528,
+      "learning_rate": 7.261797728184331e-06,
+      "loss": 1.5583,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6008237515016304,
+      "grad_norm": 2.844945630046783,
+      "learning_rate": 7.256452214113237e-06,
+      "loss": 1.4873,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6009953663978034,
+      "grad_norm": 2.902828187514857,
+      "learning_rate": 7.251107547675341e-06,
+      "loss": 1.6206,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6011669812939763,
+      "grad_norm": 3.673835737909017,
+      "learning_rate": 7.2457637305219016e-06,
+      "loss": 1.7337,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6013385961901493,
+      "grad_norm": 3.394777346119734,
+      "learning_rate": 7.2404207643039236e-06,
+      "loss": 1.7251,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6015102110863223,
+      "grad_norm": 3.4967505570620254,
+      "learning_rate": 7.235078650672141e-06,
+      "loss": 1.6618,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6016818259824953,
+      "grad_norm": 3.5981557137725413,
+      "learning_rate": 7.2297373912770215e-06,
+      "loss": 1.4374,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6018534408786683,
+      "grad_norm": 3.011963170535395,
+      "learning_rate": 7.224396987768785e-06,
+      "loss": 1.4514,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6020250557748412,
+      "grad_norm": 3.018142268614947,
+      "learning_rate": 7.219057441797368e-06,
+      "loss": 1.4997,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6021966706710142,
+      "grad_norm": 3.590709438173592,
+      "learning_rate": 7.213718755012457e-06,
+      "loss": 1.6282,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6023682855671872,
+      "grad_norm": 3.2080284817637956,
+      "learning_rate": 7.208380929063466e-06,
+      "loss": 1.6915,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6025399004633603,
+      "grad_norm": 3.181286690715623,
+      "learning_rate": 7.203043965599536e-06,
+      "loss": 1.693,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6027115153595332,
+      "grad_norm": 3.7303874760906677,
+      "learning_rate": 7.19770786626956e-06,
+      "loss": 1.7308,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6028831302557062,
+      "grad_norm": 3.042777443835479,
+      "learning_rate": 7.1923726327221445e-06,
+      "loss": 1.433,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6030547451518792,
+      "grad_norm": 2.9409896196468863,
+      "learning_rate": 7.1870382666056435e-06,
+      "loss": 1.5407,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6032263600480522,
+      "grad_norm": 3.014341162078118,
+      "learning_rate": 7.181704769568134e-06,
+      "loss": 1.6793,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6033979749442252,
+      "grad_norm": 3.1991872243003296,
+      "learning_rate": 7.176372143257432e-06,
+      "loss": 1.5248,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6035695898403981,
+      "grad_norm": 3.316494157050678,
+      "learning_rate": 7.171040389321073e-06,
+      "loss": 1.6392,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6037412047365711,
+      "grad_norm": 3.1072107684826062,
+      "learning_rate": 7.165709509406331e-06,
+      "loss": 1.4747,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6039128196327441,
+      "grad_norm": 2.9216102514259785,
+      "learning_rate": 7.160379505160211e-06,
+      "loss": 1.56,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6040844345289171,
+      "grad_norm": 3.7776870278156567,
+      "learning_rate": 7.155050378229445e-06,
+      "loss": 1.6509,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6042560494250901,
+      "grad_norm": 3.584135123962141,
+      "learning_rate": 7.149722130260495e-06,
+      "loss": 1.634,
+      "step": 3521
+    },
+    {
+      "epoch": 0.604427664321263,
+      "grad_norm": 3.54520310095583,
+      "learning_rate": 7.144394762899549e-06,
+      "loss": 1.7492,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6045992792174361,
+      "grad_norm": 3.636361126058969,
+      "learning_rate": 7.139068277792524e-06,
+      "loss": 1.4987,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6047708941136091,
+      "grad_norm": 3.3488301752712246,
+      "learning_rate": 7.133742676585067e-06,
+      "loss": 1.6574,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6049425090097821,
+      "grad_norm": 3.5978190190308705,
+      "learning_rate": 7.128417960922546e-06,
+      "loss": 1.5293,
+      "step": 3525
+    },
+    {
+      "epoch": 0.605114123905955,
+      "grad_norm": 3.459716421354361,
+      "learning_rate": 7.123094132450063e-06,
+      "loss": 1.4582,
+      "step": 3526
+    },
+    {
+      "epoch": 0.605285738802128,
+      "grad_norm": 3.5727578192190648,
+      "learning_rate": 7.117771192812441e-06,
+      "loss": 1.6297,
+      "step": 3527
+    },
+    {
+      "epoch": 0.605457353698301,
+      "grad_norm": 3.498299317220624,
+      "learning_rate": 7.112449143654235e-06,
+      "loss": 1.6427,
+      "step": 3528
+    },
+    {
+      "epoch": 0.605628968594474,
+      "grad_norm": 2.9682993796259725,
+      "learning_rate": 7.107127986619711e-06,
+      "loss": 1.4381,
+      "step": 3529
+    },
+    {
+      "epoch": 0.605800583490647,
+      "grad_norm": 3.4931398822266395,
+      "learning_rate": 7.1018077233528685e-06,
+      "loss": 1.5446,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6059721983868199,
+      "grad_norm": 3.449594681051599,
+      "learning_rate": 7.096488355497435e-06,
+      "loss": 1.6931,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6061438132829929,
+      "grad_norm": 3.158346584260765,
+      "learning_rate": 7.091169884696853e-06,
+      "loss": 1.4981,
+      "step": 3532
+    },
+    {
+      "epoch": 0.606315428179166,
+      "grad_norm": 4.636648985828793,
+      "learning_rate": 7.0858523125942944e-06,
+      "loss": 1.6703,
+      "step": 3533
+    },
+    {
+      "epoch": 0.606487043075339,
+      "grad_norm": 4.466582317728687,
+      "learning_rate": 7.080535640832651e-06,
+      "loss": 1.8682,
+      "step": 3534
+    },
+    {
+      "epoch": 0.606658657971512,
+      "grad_norm": 3.745305060606846,
+      "learning_rate": 7.075219871054528e-06,
+      "loss": 1.3963,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6068302728676849,
+      "grad_norm": 4.193770175075539,
+      "learning_rate": 7.069905004902269e-06,
+      "loss": 1.8452,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6070018877638579,
+      "grad_norm": 3.207187857707066,
+      "learning_rate": 7.0645910440179234e-06,
+      "loss": 1.5548,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6071735026600309,
+      "grad_norm": 3.7115883826094227,
+      "learning_rate": 7.059277990043272e-06,
+      "loss": 1.4893,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6073451175562039,
+      "grad_norm": 3.1517930045316978,
+      "learning_rate": 7.0539658446198036e-06,
+      "loss": 1.597,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6075167324523769,
+      "grad_norm": 3.251598528124382,
+      "learning_rate": 7.048654609388741e-06,
+      "loss": 1.5905,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6076883473485498,
+      "grad_norm": 3.193389143061349,
+      "learning_rate": 7.043344285991012e-06,
+      "loss": 1.4434,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6078599622447228,
+      "grad_norm": 3.7575111217500345,
+      "learning_rate": 7.0380348760672655e-06,
+      "loss": 1.75,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6080315771408958,
+      "grad_norm": 3.278457805476395,
+      "learning_rate": 7.032726381257879e-06,
+      "loss": 1.5966,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6082031920370689,
+      "grad_norm": 3.1176551026544987,
+      "learning_rate": 7.027418803202931e-06,
+      "loss": 1.6025,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6083748069332418,
+      "grad_norm": 3.148546222023553,
+      "learning_rate": 7.022112143542232e-06,
+      "loss": 1.5148,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6085464218294148,
+      "grad_norm": 3.7597591439723845,
+      "learning_rate": 7.016806403915302e-06,
+      "loss": 1.667,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6087180367255878,
+      "grad_norm": 3.1000399357350132,
+      "learning_rate": 7.011501585961369e-06,
+      "loss": 1.4906,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6088896516217608,
+      "grad_norm": 3.678846588145142,
+      "learning_rate": 7.00619769131939e-06,
+      "loss": 1.5873,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6090612665179338,
+      "grad_norm": 3.5467011930611023,
+      "learning_rate": 7.000894721628027e-06,
+      "loss": 1.4743,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6092328814141067,
+      "grad_norm": 4.573899462988421,
+      "learning_rate": 6.995592678525662e-06,
+      "loss": 1.5757,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6094044963102797,
+      "grad_norm": 3.853856004027449,
+      "learning_rate": 6.990291563650387e-06,
+      "loss": 1.618,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6095761112064527,
+      "grad_norm": 4.770867604145773,
+      "learning_rate": 6.9849913786400154e-06,
+      "loss": 1.5727,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6097477261026257,
+      "grad_norm": 3.3813302617365864,
+      "learning_rate": 6.979692125132058e-06,
+      "loss": 1.6511,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6099193409987987,
+      "grad_norm": 3.4948128644776006,
+      "learning_rate": 6.974393804763746e-06,
+      "loss": 1.5985,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6100909558949716,
+      "grad_norm": 3.919573765566504,
+      "learning_rate": 6.96909641917203e-06,
+      "loss": 1.699,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6102625707911447,
+      "grad_norm": 3.588520580473983,
+      "learning_rate": 6.963799969993557e-06,
+      "loss": 1.6282,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6104341856873177,
+      "grad_norm": 3.6443422963335164,
+      "learning_rate": 6.958504458864699e-06,
+      "loss": 1.4209,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6106058005834907,
+      "grad_norm": 3.5832394828873095,
+      "learning_rate": 6.953209887421531e-06,
+      "loss": 1.434,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6107774154796637,
+      "grad_norm": 3.0194882536944556,
+      "learning_rate": 6.947916257299829e-06,
+      "loss": 1.329,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6109490303758366,
+      "grad_norm": 2.8369981346213526,
+      "learning_rate": 6.9426235701350975e-06,
+      "loss": 1.3983,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6111206452720096,
+      "grad_norm": 4.3766950840191186,
+      "learning_rate": 6.937331827562532e-06,
+      "loss": 1.4592,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6112922601681826,
+      "grad_norm": 3.575947492900013,
+      "learning_rate": 6.932041031217051e-06,
+      "loss": 1.5822,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6114638750643556,
+      "grad_norm": 3.8254188579891353,
+      "learning_rate": 6.926751182733264e-06,
+      "loss": 1.4851,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6116354899605285,
+      "grad_norm": 3.770091050093029,
+      "learning_rate": 6.92146228374551e-06,
+      "loss": 1.6089,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6118071048567015,
+      "grad_norm": 3.972773931503509,
+      "learning_rate": 6.91617433588781e-06,
+      "loss": 1.5105,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6119787197528745,
+      "grad_norm": 3.071128027261779,
+      "learning_rate": 6.910887340793904e-06,
+      "loss": 1.4183,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6121503346490476,
+      "grad_norm": 4.232389848123197,
+      "learning_rate": 6.905601300097241e-06,
+      "loss": 1.5972,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6123219495452206,
+      "grad_norm": 3.7256258521185432,
+      "learning_rate": 6.900316215430967e-06,
+      "loss": 1.7475,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6124935644413935,
+      "grad_norm": 3.576751242900159,
+      "learning_rate": 6.8950320884279375e-06,
+      "loss": 1.733,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6126651793375665,
+      "grad_norm": 3.366578413225391,
+      "learning_rate": 6.889748920720714e-06,
+      "loss": 1.7823,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6128367942337395,
+      "grad_norm": 3.144677461459144,
+      "learning_rate": 6.88446671394155e-06,
+      "loss": 1.4366,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6130084091299125,
+      "grad_norm": 3.327116844473996,
+      "learning_rate": 6.879185469722417e-06,
+      "loss": 1.4711,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6131800240260855,
+      "grad_norm": 2.873522966046271,
+      "learning_rate": 6.873905189694979e-06,
+      "loss": 1.4945,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6133516389222584,
+      "grad_norm": 3.385729227599632,
+      "learning_rate": 6.86862587549061e-06,
+      "loss": 1.7169,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6135232538184314,
+      "grad_norm": 3.659100526585139,
+      "learning_rate": 6.863347528740376e-06,
+      "loss": 1.4637,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6136948687146044,
+      "grad_norm": 3.5871880291928337,
+      "learning_rate": 6.858070151075057e-06,
+      "loss": 1.5778,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6138664836107774,
+      "grad_norm": 3.4965527291808742,
+      "learning_rate": 6.8527937441251195e-06,
+      "loss": 1.6727,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6140380985069505,
+      "grad_norm": 3.180628062774472,
+      "learning_rate": 6.847518309520734e-06,
+      "loss": 1.55,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6142097134031234,
+      "grad_norm": 4.125068719067522,
+      "learning_rate": 6.842243848891781e-06,
+      "loss": 1.8915,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6143813282992964,
+      "grad_norm": 2.822402815553282,
+      "learning_rate": 6.836970363867825e-06,
+      "loss": 1.4799,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6145529431954694,
+      "grad_norm": 4.91866492123016,
+      "learning_rate": 6.831697856078143e-06,
+      "loss": 1.6284,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6147245580916424,
+      "grad_norm": 4.034526444138595,
+      "learning_rate": 6.826426327151703e-06,
+      "loss": 1.4852,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6148961729878153,
+      "grad_norm": 3.9752131268320428,
+      "learning_rate": 6.821155778717164e-06,
+      "loss": 1.7132,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6150677878839883,
+      "grad_norm": 3.7212114180900366,
+      "learning_rate": 6.8158862124028935e-06,
+      "loss": 1.6357,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6152394027801613,
+      "grad_norm": 3.3607635813613133,
+      "learning_rate": 6.81061762983695e-06,
+      "loss": 1.5627,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6154110176763343,
+      "grad_norm": 3.102328408443979,
+      "learning_rate": 6.805350032647091e-06,
+      "loss": 1.1952,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6155826325725073,
+      "grad_norm": 3.3889154530800862,
+      "learning_rate": 6.800083422460766e-06,
+      "loss": 1.5259,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6157542474686802,
+      "grad_norm": 4.278932537075945,
+      "learning_rate": 6.794817800905126e-06,
+      "loss": 1.5404,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6159258623648532,
+      "grad_norm": 4.498275024820423,
+      "learning_rate": 6.789553169607005e-06,
+      "loss": 1.904,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6160974772610263,
+      "grad_norm": 3.6648361183917317,
+      "learning_rate": 6.78428953019294e-06,
+      "loss": 1.4874,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6162690921571993,
+      "grad_norm": 4.730235381280263,
+      "learning_rate": 6.779026884289161e-06,
+      "loss": 1.5854,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6164407070533723,
+      "grad_norm": 4.039251415789363,
+      "learning_rate": 6.7737652335215875e-06,
+      "loss": 1.6877,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6166123219495452,
+      "grad_norm": 2.949797540673822,
+      "learning_rate": 6.7685045795158366e-06,
+      "loss": 1.5218,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6167839368457182,
+      "grad_norm": 3.1462755989389546,
+      "learning_rate": 6.763244923897214e-06,
+      "loss": 1.66,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6169555517418912,
+      "grad_norm": 8.790351768451915,
+      "learning_rate": 6.757986268290713e-06,
+      "loss": 1.6265,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6171271666380642,
+      "grad_norm": 3.625955100783232,
+      "learning_rate": 6.752728614321027e-06,
+      "loss": 1.7215,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6172987815342371,
+      "grad_norm": 3.7287328863114855,
+      "learning_rate": 6.747471963612533e-06,
+      "loss": 1.5203,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6174703964304101,
+      "grad_norm": 3.0202138281623854,
+      "learning_rate": 6.742216317789301e-06,
+      "loss": 1.378,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6176420113265831,
+      "grad_norm": 4.725003682769307,
+      "learning_rate": 6.7369616784750865e-06,
+      "loss": 1.8272,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6178136262227562,
+      "grad_norm": 3.6525662232026535,
+      "learning_rate": 6.731708047293347e-06,
+      "loss": 1.4771,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6179852411189292,
+      "grad_norm": 3.0653076577483764,
+      "learning_rate": 6.726455425867209e-06,
+      "loss": 1.6039,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6181568560151021,
+      "grad_norm": 3.546807468213588,
+      "learning_rate": 6.721203815819499e-06,
+      "loss": 1.4522,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6183284709112751,
+      "grad_norm": 3.231561608886287,
+      "learning_rate": 6.715953218772731e-06,
+      "loss": 1.5448,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6185000858074481,
+      "grad_norm": 4.255214525889074,
+      "learning_rate": 6.710703636349102e-06,
+      "loss": 1.4211,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6186717007036211,
+      "grad_norm": 3.5580390080097235,
+      "learning_rate": 6.7054550701705e-06,
+      "loss": 1.8038,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6188433155997941,
+      "grad_norm": 3.980774314848651,
+      "learning_rate": 6.700207521858497e-06,
+      "loss": 1.7348,
+      "step": 3606
+    },
+    {
+      "epoch": 0.619014930495967,
+      "grad_norm": 3.838505218644239,
+      "learning_rate": 6.694960993034345e-06,
+      "loss": 1.8029,
+      "step": 3607
+    },
+    {
+      "epoch": 0.61918654539214,
+      "grad_norm": 3.55897323339006,
+      "learning_rate": 6.689715485318991e-06,
+      "loss": 1.6498,
+      "step": 3608
+    },
+    {
+      "epoch": 0.619358160288313,
+      "grad_norm": 3.6483570376923558,
+      "learning_rate": 6.684471000333057e-06,
+      "loss": 1.7088,
+      "step": 3609
+    },
+    {
+      "epoch": 0.619529775184486,
+      "grad_norm": 3.286987247453582,
+      "learning_rate": 6.679227539696859e-06,
+      "loss": 1.6491,
+      "step": 3610
+    },
+    {
+      "epoch": 0.619701390080659,
+      "grad_norm": 3.6550228488817593,
+      "learning_rate": 6.673985105030386e-06,
+      "loss": 1.6154,
+      "step": 3611
+    },
+    {
+      "epoch": 0.619873004976832,
+      "grad_norm": 3.6965410509474026,
+      "learning_rate": 6.668743697953322e-06,
+      "loss": 1.4387,
+      "step": 3612
+    },
+    {
+      "epoch": 0.620044619873005,
+      "grad_norm": 3.8054483669238683,
+      "learning_rate": 6.663503320085021e-06,
+      "loss": 1.7772,
+      "step": 3613
+    },
+    {
+      "epoch": 0.620216234769178,
+      "grad_norm": 4.100195875464454,
+      "learning_rate": 6.658263973044522e-06,
+      "loss": 1.8859,
+      "step": 3614
+    },
+    {
+      "epoch": 0.620387849665351,
+      "grad_norm": 2.8411303765212006,
+      "learning_rate": 6.653025658450556e-06,
+      "loss": 1.6475,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6205594645615239,
+      "grad_norm": 3.738410759369591,
+      "learning_rate": 6.64778837792152e-06,
+      "loss": 1.5985,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6207310794576969,
+      "grad_norm": 4.222578233471034,
+      "learning_rate": 6.642552133075506e-06,
+      "loss": 1.4064,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6209026943538699,
+      "grad_norm": 3.2683299982967164,
+      "learning_rate": 6.637316925530275e-06,
+      "loss": 1.442,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6210743092500429,
+      "grad_norm": 3.5968973542926514,
+      "learning_rate": 6.6320827569032684e-06,
+      "loss": 1.7,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6212459241462159,
+      "grad_norm": 3.431879397853006,
+      "learning_rate": 6.626849628811613e-06,
+      "loss": 1.7326,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6214175390423888,
+      "grad_norm": 3.3466457830514402,
+      "learning_rate": 6.62161754287211e-06,
+      "loss": 1.6379,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6215891539385618,
+      "grad_norm": 3.585286521571043,
+      "learning_rate": 6.61638650070124e-06,
+      "loss": 1.6252,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6217607688347349,
+      "grad_norm": 3.441388396163768,
+      "learning_rate": 6.611156503915158e-06,
+      "loss": 1.5805,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6219323837309079,
+      "grad_norm": 2.8788988398472104,
+      "learning_rate": 6.605927554129705e-06,
+      "loss": 1.4655,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6221039986270809,
+      "grad_norm": 4.097672213672383,
+      "learning_rate": 6.600699652960383e-06,
+      "loss": 1.704,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6222756135232538,
+      "grad_norm": 3.8053155913992014,
+      "learning_rate": 6.595472802022382e-06,
+      "loss": 1.6123,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6224472284194268,
+      "grad_norm": 2.9413464367244906,
+      "learning_rate": 6.590247002930567e-06,
+      "loss": 1.5142,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6226188433155998,
+      "grad_norm": 3.7481112593904973,
+      "learning_rate": 6.585022257299473e-06,
+      "loss": 1.552,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6227904582117728,
+      "grad_norm": 3.9403199597742726,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 1.3722,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6229620731079458,
+      "grad_norm": 3.5741659718144856,
+      "learning_rate": 6.574575932875977e-06,
+      "loss": 1.6998,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6231336880041187,
+      "grad_norm": 3.2833445182141214,
+      "learning_rate": 6.569354357311015e-06,
+      "loss": 1.625,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6233053029002917,
+      "grad_norm": 3.3823846836668863,
+      "learning_rate": 6.564133841661667e-06,
+      "loss": 1.5147,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6234769177964647,
+      "grad_norm": 3.722953347222163,
+      "learning_rate": 6.558914387540835e-06,
+      "loss": 1.8322,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6236485326926378,
+      "grad_norm": 3.2166245726465883,
+      "learning_rate": 6.5536959965611e-06,
+      "loss": 1.68,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6238201475888107,
+      "grad_norm": 3.616491143354469,
+      "learning_rate": 6.548478670334706e-06,
+      "loss": 1.8139,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6239917624849837,
+      "grad_norm": 3.7048168605553258,
+      "learning_rate": 6.543262410473579e-06,
+      "loss": 1.7876,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6241633773811567,
+      "grad_norm": 3.842558187299423,
+      "learning_rate": 6.538047218589303e-06,
+      "loss": 1.4199,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6243349922773297,
+      "grad_norm": 3.494201520092776,
+      "learning_rate": 6.532833096293139e-06,
+      "loss": 1.5375,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6245066071735027,
+      "grad_norm": 3.217061689816585,
+      "learning_rate": 6.527620045196019e-06,
+      "loss": 1.6631,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6246782220696756,
+      "grad_norm": 3.6928714876781243,
+      "learning_rate": 6.522408066908541e-06,
+      "loss": 1.408,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6248498369658486,
+      "grad_norm": 3.952689029118906,
+      "learning_rate": 6.517197163040974e-06,
+      "loss": 1.5684,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6250214518620216,
+      "grad_norm": 3.6043387599734373,
+      "learning_rate": 6.511987335203256e-06,
+      "loss": 1.6327,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6251930667581946,
+      "grad_norm": 3.7043748833197077,
+      "learning_rate": 6.506778585004982e-06,
+      "loss": 1.4229,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6253646816543676,
+      "grad_norm": 3.261278120490368,
+      "learning_rate": 6.501570914055428e-06,
+      "loss": 1.6165,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6255362965505405,
+      "grad_norm": 3.669892061644329,
+      "learning_rate": 6.496364323963528e-06,
+      "loss": 1.8881,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6257079114467136,
+      "grad_norm": 3.465479602926153,
+      "learning_rate": 6.491158816337889e-06,
+      "loss": 1.3925,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6258795263428866,
+      "grad_norm": 3.4571562419806408,
+      "learning_rate": 6.485954392786775e-06,
+      "loss": 1.4355,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6260511412390596,
+      "grad_norm": 3.869149989507204,
+      "learning_rate": 6.480751054918124e-06,
+      "loss": 1.7249,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6262227561352325,
+      "grad_norm": 4.523040760674294,
+      "learning_rate": 6.475548804339529e-06,
+      "loss": 1.6034,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6263943710314055,
+      "grad_norm": 3.67809048459109,
+      "learning_rate": 6.4703476426582525e-06,
+      "loss": 1.61,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6265659859275785,
+      "grad_norm": 3.7439101481056194,
+      "learning_rate": 6.465147571481223e-06,
+      "loss": 1.4205,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6267376008237515,
+      "grad_norm": 3.823136874827558,
+      "learning_rate": 6.459948592415027e-06,
+      "loss": 1.622,
+      "step": 3652
+    },
+    {
+      "epoch": 0.6269092157199245,
+      "grad_norm": 3.518818248937889,
+      "learning_rate": 6.454750707065919e-06,
+      "loss": 1.3603,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6270808306160974,
+      "grad_norm": 3.263262581086908,
+      "learning_rate": 6.449553917039811e-06,
+      "loss": 1.4911,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6272524455122704,
+      "grad_norm": 4.004275786512978,
+      "learning_rate": 6.4443582239422744e-06,
+      "loss": 1.6386,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6274240604084435,
+      "grad_norm": 3.5166375079723893,
+      "learning_rate": 6.439163629378549e-06,
+      "loss": 1.6907,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6275956753046165,
+      "grad_norm": 3.1762171216125656,
+      "learning_rate": 6.4339701349535296e-06,
+      "loss": 1.444,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6277672902007895,
+      "grad_norm": 3.564924980964853,
+      "learning_rate": 6.428777742271776e-06,
+      "loss": 1.7593,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6279389050969624,
+      "grad_norm": 4.009595801412592,
+      "learning_rate": 6.4235864529375005e-06,
+      "loss": 1.639,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6281105199931354,
+      "grad_norm": 3.382801063046933,
+      "learning_rate": 6.4183962685545845e-06,
+      "loss": 1.3942,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6282821348893084,
+      "grad_norm": 3.0578832312616093,
+      "learning_rate": 6.413207190726556e-06,
+      "loss": 1.555,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6284537497854814,
+      "grad_norm": 3.462684674440041,
+      "learning_rate": 6.4080192210566085e-06,
+      "loss": 1.6732,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6286253646816544,
+      "grad_norm": 3.0427530879325726,
+      "learning_rate": 6.402832361147595e-06,
+      "loss": 1.429,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6287969795778273,
+      "grad_norm": 3.3477560926201955,
+      "learning_rate": 6.397646612602017e-06,
+      "loss": 1.5817,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6289685944740003,
+      "grad_norm": 3.170673179673043,
+      "learning_rate": 6.392461977022044e-06,
+      "loss": 1.585,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6291402093701733,
+      "grad_norm": 3.7790798834673214,
+      "learning_rate": 6.3872784560095e-06,
+      "loss": 1.4983,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6293118242663464,
+      "grad_norm": 3.126668507127899,
+      "learning_rate": 6.382096051165847e-06,
+      "loss": 1.6057,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6294834391625193,
+      "grad_norm": 3.2705280565880854,
+      "learning_rate": 6.3769147640922254e-06,
+      "loss": 1.2736,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6296550540586923,
+      "grad_norm": 3.225755690959349,
+      "learning_rate": 6.371734596389415e-06,
+      "loss": 1.4795,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6298266689548653,
+      "grad_norm": 4.128931141771984,
+      "learning_rate": 6.36655554965786e-06,
+      "loss": 1.4248,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6299982838510383,
+      "grad_norm": 2.8665491318890024,
+      "learning_rate": 6.361377625497651e-06,
+      "loss": 1.5379,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6301698987472113,
+      "grad_norm": 3.391287368139855,
+      "learning_rate": 6.356200825508538e-06,
+      "loss": 1.5778,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6303415136433842,
+      "grad_norm": 3.2451638968289136,
+      "learning_rate": 6.351025151289915e-06,
+      "loss": 1.5252,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6305131285395572,
+      "grad_norm": 3.437817930831985,
+      "learning_rate": 6.345850604440833e-06,
+      "loss": 1.587,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6306847434357302,
+      "grad_norm": 4.4183841791637954,
+      "learning_rate": 6.340677186559998e-06,
+      "loss": 1.5805,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6308563583319032,
+      "grad_norm": 4.337792287117245,
+      "learning_rate": 6.335504899245761e-06,
+      "loss": 1.5193,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6310279732280762,
+      "grad_norm": 3.5091697528556445,
+      "learning_rate": 6.330333744096131e-06,
+      "loss": 1.6965,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6311995881242491,
+      "grad_norm": 3.6592210897623647,
+      "learning_rate": 6.3251637227087645e-06,
+      "loss": 1.3187,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6313712030204222,
+      "grad_norm": 3.4214428461130377,
+      "learning_rate": 6.319994836680957e-06,
+      "loss": 1.5261,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6315428179165952,
+      "grad_norm": 3.9302958773314693,
+      "learning_rate": 6.3148270876096695e-06,
+      "loss": 1.6667,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6317144328127682,
+      "grad_norm": 3.881326719868973,
+      "learning_rate": 6.309660477091501e-06,
+      "loss": 1.5373,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6318860477089412,
+      "grad_norm": 4.872974155548524,
+      "learning_rate": 6.3044950067227066e-06,
+      "loss": 1.7324,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6320576626051141,
+      "grad_norm": 5.07238964319186,
+      "learning_rate": 6.299330678099182e-06,
+      "loss": 1.7323,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6322292775012871,
+      "grad_norm": 3.9692110166764167,
+      "learning_rate": 6.294167492816476e-06,
+      "loss": 1.6193,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6324008923974601,
+      "grad_norm": 3.1700893639605647,
+      "learning_rate": 6.289005452469778e-06,
+      "loss": 1.3325,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6325725072936331,
+      "grad_norm": 3.416355454492311,
+      "learning_rate": 6.2838445586539255e-06,
+      "loss": 1.5966,
+      "step": 3686
+    },
+    {
+      "epoch": 0.632744122189806,
+      "grad_norm": 3.791320115429006,
+      "learning_rate": 6.278684812963407e-06,
+      "loss": 1.3722,
+      "step": 3687
+    },
+    {
+      "epoch": 0.632915737085979,
+      "grad_norm": 3.1000746339948617,
+      "learning_rate": 6.2735262169923495e-06,
+      "loss": 1.5722,
+      "step": 3688
+    },
+    {
+      "epoch": 0.633087351982152,
+      "grad_norm": 3.6052364477817123,
+      "learning_rate": 6.2683687723345295e-06,
+      "loss": 1.6687,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6332589668783251,
+      "grad_norm": 4.18193478632042,
+      "learning_rate": 6.263212480583372e-06,
+      "loss": 1.7141,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6334305817744981,
+      "grad_norm": 3.6988980422128104,
+      "learning_rate": 6.258057343331927e-06,
+      "loss": 1.5213,
+      "step": 3691
+    },
+    {
+      "epoch": 0.633602196670671,
+      "grad_norm": 3.7801072635505193,
+      "learning_rate": 6.252903362172911e-06,
+      "loss": 1.7459,
+      "step": 3692
+    },
+    {
+      "epoch": 0.633773811566844,
+      "grad_norm": 3.748680941484881,
+      "learning_rate": 6.2477505386986655e-06,
+      "loss": 1.9347,
+      "step": 3693
+    },
+    {
+      "epoch": 0.633945426463017,
+      "grad_norm": 3.863835905110704,
+      "learning_rate": 6.242598874501187e-06,
+      "loss": 1.6168,
+      "step": 3694
+    },
+    {
+      "epoch": 0.63411704135919,
+      "grad_norm": 3.7547104864174883,
+      "learning_rate": 6.237448371172108e-06,
+      "loss": 1.7518,
+      "step": 3695
+    },
+    {
+      "epoch": 0.634288656255363,
+      "grad_norm": 3.7202810068051186,
+      "learning_rate": 6.2322990303027055e-06,
+      "loss": 1.4998,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6344602711515359,
+      "grad_norm": 5.1686129810012735,
+      "learning_rate": 6.2271508534838885e-06,
+      "loss": 1.7796,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6346318860477089,
+      "grad_norm": 3.8366474035647844,
+      "learning_rate": 6.222003842306214e-06,
+      "loss": 1.7047,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6348035009438819,
+      "grad_norm": 4.00011838855578,
+      "learning_rate": 6.21685799835988e-06,
+      "loss": 1.5354,
+      "step": 3699
+    },
+    {
+      "epoch": 0.634975115840055,
+      "grad_norm": 3.8097320904639966,
+      "learning_rate": 6.2117133232347225e-06,
+      "loss": 1.5618,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6351467307362278,
+      "grad_norm": 3.3871116041841103,
+      "learning_rate": 6.206569818520212e-06,
+      "loss": 1.6155,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6353183456324009,
+      "grad_norm": 3.613384562896955,
+      "learning_rate": 6.201427485805469e-06,
+      "loss": 1.3645,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6354899605285739,
+      "grad_norm": 3.3617147891575136,
+      "learning_rate": 6.196286326679231e-06,
+      "loss": 1.487,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6356615754247469,
+      "grad_norm": 3.9168910731786566,
+      "learning_rate": 6.191146342729892e-06,
+      "loss": 1.5568,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6358331903209199,
+      "grad_norm": 3.931999100186654,
+      "learning_rate": 6.186007535545475e-06,
+      "loss": 1.5467,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6360048052170928,
+      "grad_norm": 3.252701382926673,
+      "learning_rate": 6.1808699067136425e-06,
+      "loss": 1.5966,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6361764201132658,
+      "grad_norm": 3.983194022487637,
+      "learning_rate": 6.175733457821691e-06,
+      "loss": 1.7826,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6363480350094388,
+      "grad_norm": 4.114251383059512,
+      "learning_rate": 6.170598190456556e-06,
+      "loss": 1.3341,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6365196499056118,
+      "grad_norm": 3.9162467459067525,
+      "learning_rate": 6.1654641062047996e-06,
+      "loss": 1.7739,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6366912648017848,
+      "grad_norm": 3.218901458927811,
+      "learning_rate": 6.160331206652624e-06,
+      "loss": 1.6653,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6368628796979577,
+      "grad_norm": 3.5083626587042596,
+      "learning_rate": 6.155199493385864e-06,
+      "loss": 1.6322,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6370344945941308,
+      "grad_norm": 3.6412562653893294,
+      "learning_rate": 6.150068967989995e-06,
+      "loss": 1.6263,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6372061094903038,
+      "grad_norm": 4.315141381195538,
+      "learning_rate": 6.144939632050112e-06,
+      "loss": 1.6946,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6373777243864768,
+      "grad_norm": 3.874185437103271,
+      "learning_rate": 6.13981148715096e-06,
+      "loss": 1.7088,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6375493392826498,
+      "grad_norm": 3.076904863041038,
+      "learning_rate": 6.134684534876892e-06,
+      "loss": 1.4948,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6377209541788227,
+      "grad_norm": 3.911491755708042,
+      "learning_rate": 6.129558776811915e-06,
+      "loss": 1.3543,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6378925690749957,
+      "grad_norm": 4.320289096329126,
+      "learning_rate": 6.124434214539654e-06,
+      "loss": 1.5979,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6380641839711687,
+      "grad_norm": 3.5407349273503086,
+      "learning_rate": 6.119310849643371e-06,
+      "loss": 1.7499,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6382357988673417,
+      "grad_norm": 3.4988565958795466,
+      "learning_rate": 6.114188683705959e-06,
+      "loss": 1.6295,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6384074137635146,
+      "grad_norm": 3.2751617424176196,
+      "learning_rate": 6.109067718309936e-06,
+      "loss": 1.3878,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6385790286596876,
+      "grad_norm": 4.0274202279877835,
+      "learning_rate": 6.103947955037447e-06,
+      "loss": 1.7432,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6387506435558606,
+      "grad_norm": 3.412700859198055,
+      "learning_rate": 6.098829395470269e-06,
+      "loss": 1.7594,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6389222584520337,
+      "grad_norm": 3.6947286015485927,
+      "learning_rate": 6.09371204118981e-06,
+      "loss": 1.611,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6390938733482067,
+      "grad_norm": 3.4723923501632163,
+      "learning_rate": 6.088595893777105e-06,
+      "loss": 1.5917,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6392654882443796,
+      "grad_norm": 3.810015283025792,
+      "learning_rate": 6.083480954812809e-06,
+      "loss": 1.6391,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6394371031405526,
+      "grad_norm": 4.316265361954921,
+      "learning_rate": 6.078367225877217e-06,
+      "loss": 1.432,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6396087180367256,
+      "grad_norm": 3.6392790149117564,
+      "learning_rate": 6.073254708550231e-06,
+      "loss": 1.5334,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6397803329328986,
+      "grad_norm": 4.5333545093331775,
+      "learning_rate": 6.068143404411395e-06,
+      "loss": 1.6917,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6399519478290716,
+      "grad_norm": 3.6116840105698556,
+      "learning_rate": 6.063033315039873e-06,
+      "loss": 1.6131,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6401235627252445,
+      "grad_norm": 3.997257841992301,
+      "learning_rate": 6.057924442014451e-06,
+      "loss": 1.6189,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6402951776214175,
+      "grad_norm": 3.0261820059168856,
+      "learning_rate": 6.052816786913547e-06,
+      "loss": 1.4005,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6404667925175905,
+      "grad_norm": 3.735924258990418,
+      "learning_rate": 6.047710351315195e-06,
+      "loss": 1.5999,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6406384074137635,
+      "grad_norm": 3.871643151320293,
+      "learning_rate": 6.042605136797051e-06,
+      "loss": 1.8283,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6408100223099366,
+      "grad_norm": 3.91113298025619,
+      "learning_rate": 6.037501144936395e-06,
+      "loss": 1.4033,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6409816372061095,
+      "grad_norm": 3.1636064280197838,
+      "learning_rate": 6.032398377310139e-06,
+      "loss": 1.4198,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6411532521022825,
+      "grad_norm": 3.3514879313978954,
+      "learning_rate": 6.0272968354948065e-06,
+      "loss": 1.4831,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6413248669984555,
+      "grad_norm": 3.3447151992771893,
+      "learning_rate": 6.022196521066545e-06,
+      "loss": 1.2749,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6414964818946285,
+      "grad_norm": 5.139484632492181,
+      "learning_rate": 6.017097435601127e-06,
+      "loss": 1.6103,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6416680967908014,
+      "grad_norm": 3.6517191457366662,
+      "learning_rate": 6.0119995806739316e-06,
+      "loss": 1.5468,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6418397116869744,
+      "grad_norm": 3.2394809338420085,
+      "learning_rate": 6.006902957859973e-06,
+      "loss": 1.5817,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6420113265831474,
+      "grad_norm": 3.098384107595828,
+      "learning_rate": 6.00180756873388e-06,
+      "loss": 1.411,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6421829414793204,
+      "grad_norm": 3.489325450387917,
+      "learning_rate": 5.996713414869895e-06,
+      "loss": 1.6147,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6423545563754934,
+      "grad_norm": 2.6642365943225137,
+      "learning_rate": 5.991620497841889e-06,
+      "loss": 1.2264,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6425261712716663,
+      "grad_norm": 3.6552380419471473,
+      "learning_rate": 5.986528819223344e-06,
+      "loss": 1.4133,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6426977861678393,
+      "grad_norm": 4.197809834211433,
+      "learning_rate": 5.981438380587355e-06,
+      "loss": 1.6674,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6428694010640124,
+      "grad_norm": 3.418750052935898,
+      "learning_rate": 5.976349183506643e-06,
+      "loss": 1.5798,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6430410159601854,
+      "grad_norm": 3.813957763526605,
+      "learning_rate": 5.9712612295535375e-06,
+      "loss": 1.5,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6432126308563584,
+      "grad_norm": 4.043553827055629,
+      "learning_rate": 5.966174520299996e-06,
+      "loss": 1.5963,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6433842457525313,
+      "grad_norm": 3.1059224258429685,
+      "learning_rate": 5.961089057317575e-06,
+      "loss": 1.4502,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6435558606487043,
+      "grad_norm": 3.254189456622494,
+      "learning_rate": 5.956004842177466e-06,
+      "loss": 1.6232,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6437274755448773,
+      "grad_norm": 4.343900909263778,
+      "learning_rate": 5.95092187645045e-06,
+      "loss": 1.7305,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6438990904410503,
+      "grad_norm": 3.83175622660869,
+      "learning_rate": 5.94584016170694e-06,
+      "loss": 1.5942,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6440707053372233,
+      "grad_norm": 3.828478919426302,
+      "learning_rate": 5.9407596995169615e-06,
+      "loss": 1.422,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6442423202333962,
+      "grad_norm": 3.60135097849645,
+      "learning_rate": 5.935680491450144e-06,
+      "loss": 1.7458,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6444139351295692,
+      "grad_norm": 3.7429744387296577,
+      "learning_rate": 5.930602539075741e-06,
+      "loss": 1.5854,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6445855500257422,
+      "grad_norm": 4.34838921481234,
+      "learning_rate": 5.9255258439626115e-06,
+      "loss": 1.6082,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6447571649219153,
+      "grad_norm": 3.4027586087568307,
+      "learning_rate": 5.920450407679219e-06,
+      "loss": 1.3763,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6449287798180882,
+      "grad_norm": 3.348098033984385,
+      "learning_rate": 5.915376231793654e-06,
+      "loss": 1.4189,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6451003947142612,
+      "grad_norm": 3.7797218768853753,
+      "learning_rate": 5.910303317873603e-06,
+      "loss": 1.6864,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6452720096104342,
+      "grad_norm": 3.573339340544784,
+      "learning_rate": 5.905231667486373e-06,
+      "loss": 1.6512,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6454436245066072,
+      "grad_norm": 3.4938775035989518,
+      "learning_rate": 5.900161282198874e-06,
+      "loss": 1.4509,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6456152394027802,
+      "grad_norm": 3.7614261762484764,
+      "learning_rate": 5.8950921635776334e-06,
+      "loss": 1.5003,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6457868542989531,
+      "grad_norm": 3.32744800427139,
+      "learning_rate": 5.890024313188774e-06,
+      "loss": 1.4281,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6459584691951261,
+      "grad_norm": 4.215982281593581,
+      "learning_rate": 5.884957732598036e-06,
+      "loss": 1.836,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6461300840912991,
+      "grad_norm": 3.554349093520539,
+      "learning_rate": 5.879892423370768e-06,
+      "loss": 1.5159,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6463016989874721,
+      "grad_norm": 3.170941512623266,
+      "learning_rate": 5.874828387071919e-06,
+      "loss": 1.3565,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6464733138836452,
+      "grad_norm": 4.046913398441757,
+      "learning_rate": 5.869765625266055e-06,
+      "loss": 1.5758,
+      "step": 3767
+    },
+    {
+      "epoch": 0.646644928779818,
+      "grad_norm": 4.004195100902059,
+      "learning_rate": 5.864704139517341e-06,
+      "loss": 1.7527,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6468165436759911,
+      "grad_norm": 3.8295756915763732,
+      "learning_rate": 5.8596439313895445e-06,
+      "loss": 1.8035,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6469881585721641,
+      "grad_norm": 3.294116541785648,
+      "learning_rate": 5.854585002446047e-06,
+      "loss": 1.6663,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6471597734683371,
+      "grad_norm": 4.0980814948586355,
+      "learning_rate": 5.849527354249827e-06,
+      "loss": 1.6069,
+      "step": 3771
+    },
+    {
+      "epoch": 0.64733138836451,
+      "grad_norm": 3.378433148526006,
+      "learning_rate": 5.844470988363475e-06,
+      "loss": 1.4084,
+      "step": 3772
+    },
+    {
+      "epoch": 0.647503003260683,
+      "grad_norm": 3.9355387344465473,
+      "learning_rate": 5.839415906349178e-06,
+      "loss": 1.4713,
+      "step": 3773
+    },
+    {
+      "epoch": 0.647674618156856,
+      "grad_norm": 4.192212967061259,
+      "learning_rate": 5.834362109768737e-06,
+      "loss": 1.5389,
+      "step": 3774
+    },
+    {
+      "epoch": 0.647846233053029,
+      "grad_norm": 3.305185233657875,
+      "learning_rate": 5.829309600183536e-06,
+      "loss": 1.589,
+      "step": 3775
+    },
+    {
+      "epoch": 0.648017847949202,
+      "grad_norm": 3.133777330040181,
+      "learning_rate": 5.82425837915458e-06,
+      "loss": 1.5038,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6481894628453749,
+      "grad_norm": 3.9107724160631836,
+      "learning_rate": 5.81920844824247e-06,
+      "loss": 1.375,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6483610777415479,
+      "grad_norm": 3.8374469638962587,
+      "learning_rate": 5.814159809007414e-06,
+      "loss": 1.2163,
+      "step": 3778
+    },
+    {
+      "epoch": 0.648532692637721,
+      "grad_norm": 3.9245893513525694,
+      "learning_rate": 5.809112463009203e-06,
+      "loss": 1.6612,
+      "step": 3779
+    },
+    {
+      "epoch": 0.648704307533894,
+      "grad_norm": 3.3793029710745315,
+      "learning_rate": 5.804066411807252e-06,
+      "loss": 1.4323,
+      "step": 3780
+    },
+    {
+      "epoch": 0.648875922430067,
+      "grad_norm": 5.7499704582100755,
+      "learning_rate": 5.799021656960555e-06,
+      "loss": 1.9008,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6490475373262399,
+      "grad_norm": 4.145319289603786,
+      "learning_rate": 5.793978200027719e-06,
+      "loss": 1.5552,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6492191522224129,
+      "grad_norm": 3.5239460063024635,
+      "learning_rate": 5.78893604256695e-06,
+      "loss": 1.7757,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6493907671185859,
+      "grad_norm": 4.241472784053818,
+      "learning_rate": 5.78389518613604e-06,
+      "loss": 1.5683,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6495623820147589,
+      "grad_norm": 3.5956903412864656,
+      "learning_rate": 5.778855632292393e-06,
+      "loss": 1.4185,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6497339969109319,
+      "grad_norm": 3.34291404104676,
+      "learning_rate": 5.773817382593008e-06,
+      "loss": 1.7483,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6499056118071048,
+      "grad_norm": 3.576626686486406,
+      "learning_rate": 5.768780438594469e-06,
+      "loss": 1.5932,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6500772267032778,
+      "grad_norm": 3.660442575443035,
+      "learning_rate": 5.763744801852976e-06,
+      "loss": 1.3911,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6502488415994508,
+      "grad_norm": 4.271863341690556,
+      "learning_rate": 5.758710473924305e-06,
+      "loss": 1.6114,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6504204564956239,
+      "grad_norm": 4.1106358150249935,
+      "learning_rate": 5.753677456363844e-06,
+      "loss": 1.5902,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6505920713917968,
+      "grad_norm": 4.054778537435686,
+      "learning_rate": 5.748645750726567e-06,
+      "loss": 1.4673,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6507636862879698,
+      "grad_norm": 3.8012734490915245,
+      "learning_rate": 5.743615358567051e-06,
+      "loss": 1.7016,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6509353011841428,
+      "grad_norm": 3.606284344280702,
+      "learning_rate": 5.738586281439455e-06,
+      "loss": 1.569,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6511069160803158,
+      "grad_norm": 3.507653202505188,
+      "learning_rate": 5.733558520897539e-06,
+      "loss": 1.7072,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6512785309764888,
+      "grad_norm": 4.191460744400435,
+      "learning_rate": 5.728532078494655e-06,
+      "loss": 1.6124,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6514501458726617,
+      "grad_norm": 4.41163625688169,
+      "learning_rate": 5.72350695578375e-06,
+      "loss": 1.6001,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6516217607688347,
+      "grad_norm": 3.643930792663315,
+      "learning_rate": 5.718483154317365e-06,
+      "loss": 1.5156,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6517933756650077,
+      "grad_norm": 4.110379256687413,
+      "learning_rate": 5.713460675647626e-06,
+      "loss": 1.6141,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6519649905611807,
+      "grad_norm": 3.5174963074233845,
+      "learning_rate": 5.70843952132625e-06,
+      "loss": 1.6481,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6521366054573537,
+      "grad_norm": 3.995495285273367,
+      "learning_rate": 5.70341969290455e-06,
+      "loss": 1.5012,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6523082203535266,
+      "grad_norm": 3.354203830219037,
+      "learning_rate": 5.698401191933431e-06,
+      "loss": 1.5803,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6524798352496997,
+      "grad_norm": 4.361345608317878,
+      "learning_rate": 5.693384019963386e-06,
+      "loss": 1.6315,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6526514501458727,
+      "grad_norm": 3.5559420376132778,
+      "learning_rate": 5.68836817854449e-06,
+      "loss": 1.4097,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6528230650420457,
+      "grad_norm": 8.442507747433941,
+      "learning_rate": 5.6833536692264204e-06,
+      "loss": 1.571,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6529946799382187,
+      "grad_norm": 3.3647883521484743,
+      "learning_rate": 5.678340493558427e-06,
+      "loss": 1.5608,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6531662948343916,
+      "grad_norm": 3.861798907269382,
+      "learning_rate": 5.67332865308936e-06,
+      "loss": 1.4725,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6533379097305646,
+      "grad_norm": 3.2557878562209672,
+      "learning_rate": 5.66831814936766e-06,
+      "loss": 1.4398,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6535095246267376,
+      "grad_norm": 3.2104530763784234,
+      "learning_rate": 5.663308983941335e-06,
+      "loss": 1.4769,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6536811395229106,
+      "grad_norm": 3.2326051216391236,
+      "learning_rate": 5.6583011583580026e-06,
+      "loss": 1.3695,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6538527544190835,
+      "grad_norm": 3.979681837676408,
+      "learning_rate": 5.653294674164855e-06,
+      "loss": 1.5013,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6540243693152565,
+      "grad_norm": 3.8853775755973565,
+      "learning_rate": 5.648289532908666e-06,
+      "loss": 1.5082,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6541959842114295,
+      "grad_norm": 4.6946791429565256,
+      "learning_rate": 5.643285736135806e-06,
+      "loss": 1.8142,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6543675991076026,
+      "grad_norm": 3.808908044611442,
+      "learning_rate": 5.638283285392218e-06,
+      "loss": 1.4008,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6545392140037756,
+      "grad_norm": 3.645256850711633,
+      "learning_rate": 5.633282182223436e-06,
+      "loss": 1.6268,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6547108288999485,
+      "grad_norm": 3.7812254506882663,
+      "learning_rate": 5.628282428174579e-06,
+      "loss": 1.6314,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6548824437961215,
+      "grad_norm": 3.6569731191614476,
+      "learning_rate": 5.62328402479035e-06,
+      "loss": 1.585,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6550540586922945,
+      "grad_norm": 3.5613744705609744,
+      "learning_rate": 5.618286973615026e-06,
+      "loss": 1.5652,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6552256735884675,
+      "grad_norm": 4.0761604639075495,
+      "learning_rate": 5.61329127619247e-06,
+      "loss": 1.7782,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6553972884846405,
+      "grad_norm": 3.7784871877702244,
+      "learning_rate": 5.608296934066131e-06,
+      "loss": 1.4937,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6555689033808134,
+      "grad_norm": 3.245147096373824,
+      "learning_rate": 5.603303948779036e-06,
+      "loss": 1.4278,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6557405182769864,
+      "grad_norm": 4.863597632653321,
+      "learning_rate": 5.5983123218738e-06,
+      "loss": 1.7635,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6559121331731594,
+      "grad_norm": 3.667927899453101,
+      "learning_rate": 5.593322054892605e-06,
+      "loss": 1.7334,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6560837480693325,
+      "grad_norm": 3.405052814879707,
+      "learning_rate": 5.588333149377218e-06,
+      "loss": 1.491,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6562553629655054,
+      "grad_norm": 3.827261534813475,
+      "learning_rate": 5.583345606868988e-06,
+      "loss": 1.5387,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6564269778616784,
+      "grad_norm": 3.393721533454186,
+      "learning_rate": 5.5783594289088465e-06,
+      "loss": 1.6266,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6565985927578514,
+      "grad_norm": 3.3119573433315592,
+      "learning_rate": 5.573374617037298e-06,
+      "loss": 1.5775,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6567702076540244,
+      "grad_norm": 3.475392878069581,
+      "learning_rate": 5.568391172794421e-06,
+      "loss": 1.7677,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6569418225501974,
+      "grad_norm": 3.947711519121117,
+      "learning_rate": 5.5634090977198825e-06,
+      "loss": 1.5809,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6571134374463703,
+      "grad_norm": 3.215947447834526,
+      "learning_rate": 5.558428393352914e-06,
+      "loss": 1.3401,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6572850523425433,
+      "grad_norm": 3.29675215204399,
+      "learning_rate": 5.553449061232331e-06,
+      "loss": 1.4942,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6574566672387163,
+      "grad_norm": 3.5809315580188,
+      "learning_rate": 5.548471102896531e-06,
+      "loss": 1.6686,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6576282821348893,
+      "grad_norm": 3.4423779321189536,
+      "learning_rate": 5.543494519883468e-06,
+      "loss": 1.5861,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6577998970310623,
+      "grad_norm": 3.5034463744390956,
+      "learning_rate": 5.53851931373069e-06,
+      "loss": 1.4006,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6579715119272352,
+      "grad_norm": 3.2764241047334086,
+      "learning_rate": 5.5335454859753154e-06,
+      "loss": 1.6529,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6581431268234083,
+      "grad_norm": 3.3531414852488126,
+      "learning_rate": 5.528573038154028e-06,
+      "loss": 1.3864,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6583147417195813,
+      "grad_norm": 3.3001063078621087,
+      "learning_rate": 5.523601971803094e-06,
+      "loss": 1.1959,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6584863566157543,
+      "grad_norm": 3.388122985617715,
+      "learning_rate": 5.5186322884583475e-06,
+      "loss": 1.4468,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6586579715119273,
+      "grad_norm": 3.4799586327912655,
+      "learning_rate": 5.513663989655198e-06,
+      "loss": 1.7559,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6588295864081002,
+      "grad_norm": 3.7502500757389354,
+      "learning_rate": 5.50869707692863e-06,
+      "loss": 1.6364,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6590012013042732,
+      "grad_norm": 4.1225471729109975,
+      "learning_rate": 5.5037315518131975e-06,
+      "loss": 1.7462,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6591728162004462,
+      "grad_norm": 3.2344214171439796,
+      "learning_rate": 5.498767415843023e-06,
+      "loss": 1.3246,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6593444310966192,
+      "grad_norm": 3.356319978431104,
+      "learning_rate": 5.493804670551796e-06,
+      "loss": 1.5977,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6595160459927921,
+      "grad_norm": 4.315759669720856,
+      "learning_rate": 5.488843317472789e-06,
+      "loss": 1.5148,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6596876608889651,
+      "grad_norm": 3.669430836405578,
+      "learning_rate": 5.483883358138835e-06,
+      "loss": 1.7643,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6598592757851381,
+      "grad_norm": 3.589622468686828,
+      "learning_rate": 5.478924794082345e-06,
+      "loss": 1.7152,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6600308906813112,
+      "grad_norm": 3.1092434015412986,
+      "learning_rate": 5.4739676268352885e-06,
+      "loss": 1.6018,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6602025055774842,
+      "grad_norm": 4.378320691719215,
+      "learning_rate": 5.469011857929202e-06,
+      "loss": 1.5139,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6603741204736571,
+      "grad_norm": 3.258108429552527,
+      "learning_rate": 5.4640574888952e-06,
+      "loss": 1.5399,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6605457353698301,
+      "grad_norm": 4.24650783765689,
+      "learning_rate": 5.459104521263962e-06,
+      "loss": 1.7231,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6607173502660031,
+      "grad_norm": 3.3134825337100677,
+      "learning_rate": 5.454152956565737e-06,
+      "loss": 1.431,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6608889651621761,
+      "grad_norm": 3.654304765614279,
+      "learning_rate": 5.449202796330328e-06,
+      "loss": 1.5113,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6610605800583491,
+      "grad_norm": 3.1029310907071506,
+      "learning_rate": 5.444254042087118e-06,
+      "loss": 1.5245,
+      "step": 3852
+    },
+    {
+      "epoch": 0.661232194954522,
+      "grad_norm": 3.9369182906762337,
+      "learning_rate": 5.439306695365045e-06,
+      "loss": 1.7089,
+      "step": 3853
+    },
+    {
+      "epoch": 0.661403809850695,
+      "grad_norm": 3.762041391146191,
+      "learning_rate": 5.43436075769262e-06,
+      "loss": 1.7283,
+      "step": 3854
+    },
+    {
+      "epoch": 0.661575424746868,
+      "grad_norm": 3.3277086108831666,
+      "learning_rate": 5.42941623059792e-06,
+      "loss": 1.4766,
+      "step": 3855
+    },
+    {
+      "epoch": 0.661747039643041,
+      "grad_norm": 3.5650734591560647,
+      "learning_rate": 5.424473115608575e-06,
+      "loss": 1.564,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6619186545392141,
+      "grad_norm": 4.164504200892568,
+      "learning_rate": 5.419531414251788e-06,
+      "loss": 1.6105,
+      "step": 3857
+    },
+    {
+      "epoch": 0.662090269435387,
+      "grad_norm": 3.4987612862190285,
+      "learning_rate": 5.414591128054328e-06,
+      "loss": 1.5773,
+      "step": 3858
+    },
+    {
+      "epoch": 0.66226188433156,
+      "grad_norm": 3.0266802239828956,
+      "learning_rate": 5.409652258542514e-06,
+      "loss": 1.4863,
+      "step": 3859
+    },
+    {
+      "epoch": 0.662433499227733,
+      "grad_norm": 4.254773157911801,
+      "learning_rate": 5.404714807242243e-06,
+      "loss": 1.5582,
+      "step": 3860
+    },
+    {
+      "epoch": 0.662605114123906,
+      "grad_norm": 3.685905771110845,
+      "learning_rate": 5.399778775678958e-06,
+      "loss": 1.6023,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6627767290200789,
+      "grad_norm": 3.728081565392676,
+      "learning_rate": 5.3948441653776755e-06,
+      "loss": 1.5571,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6629483439162519,
+      "grad_norm": 3.509278680247271,
+      "learning_rate": 5.389910977862967e-06,
+      "loss": 1.4998,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6631199588124249,
+      "grad_norm": 3.709286049239896,
+      "learning_rate": 5.384979214658972e-06,
+      "loss": 1.4901,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6632915737085979,
+      "grad_norm": 3.2930265761792805,
+      "learning_rate": 5.380048877289381e-06,
+      "loss": 1.4237,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6634631886047709,
+      "grad_norm": 3.114168467724853,
+      "learning_rate": 5.37511996727744e-06,
+      "loss": 1.5126,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6636348035009438,
+      "grad_norm": 3.8053023646331616,
+      "learning_rate": 5.370192486145968e-06,
+      "loss": 1.6188,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6638064183971168,
+      "grad_norm": 3.7342387248357003,
+      "learning_rate": 5.365266435417335e-06,
+      "loss": 1.4849,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6639780332932899,
+      "grad_norm": 4.255470305909646,
+      "learning_rate": 5.360341816613472e-06,
+      "loss": 1.7316,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6641496481894629,
+      "grad_norm": 3.234178012456068,
+      "learning_rate": 5.355418631255864e-06,
+      "loss": 1.5925,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6643212630856359,
+      "grad_norm": 3.368678076546681,
+      "learning_rate": 5.350496880865548e-06,
+      "loss": 1.3303,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6644928779818088,
+      "grad_norm": 3.755630333185734,
+      "learning_rate": 5.34557656696313e-06,
+      "loss": 1.7208,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6646644928779818,
+      "grad_norm": 3.7523038131565043,
+      "learning_rate": 5.340657691068765e-06,
+      "loss": 1.5299,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6648361077741548,
+      "grad_norm": 4.06936890653067,
+      "learning_rate": 5.335740254702168e-06,
+      "loss": 1.4689,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6650077226703278,
+      "grad_norm": 3.4581041710460014,
+      "learning_rate": 5.330824259382601e-06,
+      "loss": 1.4403,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6651793375665007,
+      "grad_norm": 3.703610318006061,
+      "learning_rate": 5.325909706628891e-06,
+      "loss": 1.4609,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6653509524626737,
+      "grad_norm": 3.6637388270178706,
+      "learning_rate": 5.32099659795941e-06,
+      "loss": 1.5536,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6655225673588467,
+      "grad_norm": 3.356576624946815,
+      "learning_rate": 5.31608493489209e-06,
+      "loss": 1.6712,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6656941822550198,
+      "grad_norm": 3.8669640864273096,
+      "learning_rate": 5.311174718944417e-06,
+      "loss": 1.7317,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6658657971511928,
+      "grad_norm": 3.398308623956015,
+      "learning_rate": 5.3062659516334224e-06,
+      "loss": 1.6039,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6660374120473657,
+      "grad_norm": 3.6410260692023466,
+      "learning_rate": 5.301358634475697e-06,
+      "loss": 1.3994,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6662090269435387,
+      "grad_norm": 2.854901084902104,
+      "learning_rate": 5.296452768987388e-06,
+      "loss": 1.3275,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6663806418397117,
+      "grad_norm": 3.2308176254722323,
+      "learning_rate": 5.291548356684177e-06,
+      "loss": 1.6324,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6665522567358847,
+      "grad_norm": 3.892612168344562,
+      "learning_rate": 5.286645399081317e-06,
+      "loss": 1.5143,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6667238716320577,
+      "grad_norm": 3.730994175188248,
+      "learning_rate": 5.281743897693594e-06,
+      "loss": 1.7345,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6668954865282306,
+      "grad_norm": 3.69902992748495,
+      "learning_rate": 5.276843854035356e-06,
+      "loss": 1.7559,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6670671014244036,
+      "grad_norm": 2.9942314219098405,
+      "learning_rate": 5.271945269620496e-06,
+      "loss": 1.4323,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6672387163205766,
+      "grad_norm": 3.3150870539401334,
+      "learning_rate": 5.267048145962461e-06,
+      "loss": 1.5008,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6674103312167496,
+      "grad_norm": 3.5851329871551605,
+      "learning_rate": 5.2621524845742386e-06,
+      "loss": 1.252,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6675819461129227,
+      "grad_norm": 4.6445493644629,
+      "learning_rate": 5.2572582869683655e-06,
+      "loss": 1.7305,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6677535610090956,
+      "grad_norm": 3.871220324859534,
+      "learning_rate": 5.252365554656933e-06,
+      "loss": 1.657,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6679251759052686,
+      "grad_norm": 3.4778828998601305,
+      "learning_rate": 5.247474289151574e-06,
+      "loss": 1.5563,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6680967908014416,
+      "grad_norm": 3.8592138961417977,
+      "learning_rate": 5.2425844919634764e-06,
+      "loss": 1.4394,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6682684056976146,
+      "grad_norm": 3.8650916338475887,
+      "learning_rate": 5.237696164603363e-06,
+      "loss": 1.5094,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6684400205937875,
+      "grad_norm": 3.309052291731976,
+      "learning_rate": 5.232809308581504e-06,
+      "loss": 1.6879,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6686116354899605,
+      "grad_norm": 3.2356204945325384,
+      "learning_rate": 5.227923925407723e-06,
+      "loss": 1.4779,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6687832503861335,
+      "grad_norm": 3.512022359886158,
+      "learning_rate": 5.223040016591383e-06,
+      "loss": 1.6374,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6689548652823065,
+      "grad_norm": 3.2002671797979407,
+      "learning_rate": 5.218157583641399e-06,
+      "loss": 1.3384,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6691264801784795,
+      "grad_norm": 3.7582122403995695,
+      "learning_rate": 5.213276628066213e-06,
+      "loss": 1.6508,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6692980950746524,
+      "grad_norm": 3.9795046424670812,
+      "learning_rate": 5.2083971513738315e-06,
+      "loss": 1.7793,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6694697099708254,
+      "grad_norm": 3.475347264737436,
+      "learning_rate": 5.2035191550717856e-06,
+      "loss": 1.6378,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6696413248669985,
+      "grad_norm": 3.6550460250906895,
+      "learning_rate": 5.19864264066716e-06,
+      "loss": 1.5344,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6698129397631715,
+      "grad_norm": 3.591239813010484,
+      "learning_rate": 5.193767609666584e-06,
+      "loss": 1.4901,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6699845546593445,
+      "grad_norm": 3.958480771878438,
+      "learning_rate": 5.188894063576217e-06,
+      "loss": 1.4217,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6701561695555174,
+      "grad_norm": 3.363445769887249,
+      "learning_rate": 5.184022003901767e-06,
+      "loss": 1.3202,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6703277844516904,
+      "grad_norm": 3.5026037587256593,
+      "learning_rate": 5.179151432148486e-06,
+      "loss": 1.5207,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6704993993478634,
+      "grad_norm": 4.397067444312862,
+      "learning_rate": 5.174282349821158e-06,
+      "loss": 1.6645,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6706710142440364,
+      "grad_norm": 3.4970869102819178,
+      "learning_rate": 5.169414758424115e-06,
+      "loss": 1.5136,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6708426291402094,
+      "grad_norm": 2.9771055986395134,
+      "learning_rate": 5.164548659461219e-06,
+      "loss": 1.3559,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6710142440363823,
+      "grad_norm": 2.8996405171780557,
+      "learning_rate": 5.159684054435879e-06,
+      "loss": 1.5426,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6711858589325553,
+      "grad_norm": 4.678963799671272,
+      "learning_rate": 5.154820944851038e-06,
+      "loss": 1.5147,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6713574738287283,
+      "grad_norm": 2.9234519732780657,
+      "learning_rate": 5.149959332209187e-06,
+      "loss": 1.3999,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6715290887249014,
+      "grad_norm": 3.885121016047288,
+      "learning_rate": 5.145099218012339e-06,
+      "loss": 1.8163,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6717007036210743,
+      "grad_norm": 3.312818127045676,
+      "learning_rate": 5.140240603762048e-06,
+      "loss": 1.701,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6718723185172473,
+      "grad_norm": 3.219063271686853,
+      "learning_rate": 5.135383490959412e-06,
+      "loss": 1.4818,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6720439334134203,
+      "grad_norm": 3.243975262572838,
+      "learning_rate": 5.130527881105062e-06,
+      "loss": 1.5344,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6722155483095933,
+      "grad_norm": 3.5153863737130764,
+      "learning_rate": 5.125673775699165e-06,
+      "loss": 1.5154,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6723871632057663,
+      "grad_norm": 3.2679862701220874,
+      "learning_rate": 5.12082117624142e-06,
+      "loss": 1.6109,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6725587781019392,
+      "grad_norm": 4.416280155724237,
+      "learning_rate": 5.115970084231059e-06,
+      "loss": 1.5567,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6727303929981122,
+      "grad_norm": 3.697086990633611,
+      "learning_rate": 5.111120501166855e-06,
+      "loss": 1.5335,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6729020078942852,
+      "grad_norm": 3.265268161005393,
+      "learning_rate": 5.10627242854711e-06,
+      "loss": 1.4798,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6730736227904582,
+      "grad_norm": 3.4879151569492337,
+      "learning_rate": 5.101425867869667e-06,
+      "loss": 1.3589,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6732452376866312,
+      "grad_norm": 3.497302152468901,
+      "learning_rate": 5.096580820631888e-06,
+      "loss": 1.5112,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6734168525828041,
+      "grad_norm": 4.016720530507737,
+      "learning_rate": 5.091737288330683e-06,
+      "loss": 1.3377,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6735884674789772,
+      "grad_norm": 3.258448118671295,
+      "learning_rate": 5.086895272462475e-06,
+      "loss": 1.507,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6737600823751502,
+      "grad_norm": 3.453691002203017,
+      "learning_rate": 5.08205477452324e-06,
+      "loss": 1.6845,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6739316972713232,
+      "grad_norm": 3.890507740456927,
+      "learning_rate": 5.0772157960084724e-06,
+      "loss": 1.6099,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6741033121674962,
+      "grad_norm": 4.423308475343997,
+      "learning_rate": 5.072378338413195e-06,
+      "loss": 1.5979,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6742749270636691,
+      "grad_norm": 3.379399039876328,
+      "learning_rate": 5.067542403231968e-06,
+      "loss": 1.3022,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6744465419598421,
+      "grad_norm": 3.321410218148661,
+      "learning_rate": 5.062707991958884e-06,
+      "loss": 1.5074,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6746181568560151,
+      "grad_norm": 3.5751191505961994,
+      "learning_rate": 5.05787510608755e-06,
+      "loss": 1.5229,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6747897717521881,
+      "grad_norm": 3.421852501349972,
+      "learning_rate": 5.053043747111119e-06,
+      "loss": 1.6119,
+      "step": 3932
+    },
+    {
+      "epoch": 0.674961386648361,
+      "grad_norm": 4.341804251061823,
+      "learning_rate": 5.048213916522255e-06,
+      "loss": 1.7233,
+      "step": 3933
+    },
+    {
+      "epoch": 0.675133001544534,
+      "grad_norm": 3.7237629610082505,
+      "learning_rate": 5.043385615813166e-06,
+      "loss": 1.5585,
+      "step": 3934
+    },
+    {
+      "epoch": 0.675304616440707,
+      "grad_norm": 4.215713304774691,
+      "learning_rate": 5.038558846475579e-06,
+      "loss": 1.582,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6754762313368801,
+      "grad_norm": 3.782032656063582,
+      "learning_rate": 5.033733610000751e-06,
+      "loss": 1.5575,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6756478462330531,
+      "grad_norm": 3.5289866908044862,
+      "learning_rate": 5.028909907879462e-06,
+      "loss": 1.5646,
+      "step": 3937
+    },
+    {
+      "epoch": 0.675819461129226,
+      "grad_norm": 4.2239655534511655,
+      "learning_rate": 5.024087741602015e-06,
+      "loss": 1.7533,
+      "step": 3938
+    },
+    {
+      "epoch": 0.675991076025399,
+      "grad_norm": 3.291419434747381,
+      "learning_rate": 5.019267112658247e-06,
+      "loss": 1.5589,
+      "step": 3939
+    },
+    {
+      "epoch": 0.676162690921572,
+      "grad_norm": 3.310097163665421,
+      "learning_rate": 5.014448022537517e-06,
+      "loss": 1.6601,
+      "step": 3940
+    },
+    {
+      "epoch": 0.676334305817745,
+      "grad_norm": 4.093098308746624,
+      "learning_rate": 5.009630472728711e-06,
+      "loss": 1.6844,
+      "step": 3941
+    },
+    {
+      "epoch": 0.676505920713918,
+      "grad_norm": 3.561701002419818,
+      "learning_rate": 5.0048144647202294e-06,
+      "loss": 1.5865,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6766775356100909,
+      "grad_norm": 3.7609693232779935,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 1.5708,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6768491505062639,
+      "grad_norm": 3.1049425636013828,
+      "learning_rate": 4.9951870800554855e-06,
+      "loss": 1.6051,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6770207654024369,
+      "grad_norm": 3.8483992368754807,
+      "learning_rate": 4.990375706373657e-06,
+      "loss": 1.4232,
+      "step": 3945
+    },
+    {
+      "epoch": 0.67719238029861,
+      "grad_norm": 3.5360274189874366,
+      "learning_rate": 4.985565880441017e-06,
+      "loss": 1.5147,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6773639951947829,
+      "grad_norm": 3.94111134637556,
+      "learning_rate": 4.98075760374358e-06,
+      "loss": 1.7055,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6775356100909559,
+      "grad_norm": 3.8728257073790746,
+      "learning_rate": 4.97595087776689e-06,
+      "loss": 1.4102,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6777072249871289,
+      "grad_norm": 3.8172761566025573,
+      "learning_rate": 4.971145703996015e-06,
+      "loss": 1.4536,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6778788398833019,
+      "grad_norm": 3.9262082706288686,
+      "learning_rate": 4.9663420839155305e-06,
+      "loss": 1.435,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6780504547794749,
+      "grad_norm": 3.6703704320806385,
+      "learning_rate": 4.961540019009546e-06,
+      "loss": 1.3904,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6782220696756478,
+      "grad_norm": 3.2000342228278797,
+      "learning_rate": 4.956739510761678e-06,
+      "loss": 1.4211,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6783936845718208,
+      "grad_norm": 3.635430532305922,
+      "learning_rate": 4.9519405606550704e-06,
+      "loss": 1.6128,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6785652994679938,
+      "grad_norm": 3.570126988570555,
+      "learning_rate": 4.947143170172385e-06,
+      "loss": 1.5894,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6787369143641668,
+      "grad_norm": 3.420246295829541,
+      "learning_rate": 4.942347340795803e-06,
+      "loss": 1.2383,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6789085292603398,
+      "grad_norm": 3.1694378632431643,
+      "learning_rate": 4.9375530740070174e-06,
+      "loss": 1.4508,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6790801441565127,
+      "grad_norm": 3.2618957057509683,
+      "learning_rate": 4.932760371287238e-06,
+      "loss": 1.5939,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6792517590526858,
+      "grad_norm": 4.197017022057773,
+      "learning_rate": 4.927969234117199e-06,
+      "loss": 1.7631,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6794233739488588,
+      "grad_norm": 3.9309922658244183,
+      "learning_rate": 4.923179663977144e-06,
+      "loss": 1.6707,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6795949888450318,
+      "grad_norm": 3.631683877755284,
+      "learning_rate": 4.9183916623468446e-06,
+      "loss": 1.6565,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6797666037412048,
+      "grad_norm": 3.858389880007458,
+      "learning_rate": 4.913605230705568e-06,
+      "loss": 1.4327,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6799382186373777,
+      "grad_norm": 3.12840387065218,
+      "learning_rate": 4.908820370532109e-06,
+      "loss": 1.3767,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6801098335335507,
+      "grad_norm": 4.593703744738146,
+      "learning_rate": 4.904037083304774e-06,
+      "loss": 1.5566,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6802814484297237,
+      "grad_norm": 3.6983099720446493,
+      "learning_rate": 4.899255370501388e-06,
+      "loss": 1.608,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6804530633258967,
+      "grad_norm": 3.2112995316900967,
+      "learning_rate": 4.894475233599287e-06,
+      "loss": 1.3857,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6806246782220696,
+      "grad_norm": 3.5235278000118737,
+      "learning_rate": 4.889696674075313e-06,
+      "loss": 1.5892,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6807962931182426,
+      "grad_norm": 4.175060829090475,
+      "learning_rate": 4.884919693405833e-06,
+      "loss": 1.362,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6809679080144156,
+      "grad_norm": 4.186048426240538,
+      "learning_rate": 4.880144293066711e-06,
+      "loss": 1.538,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6811395229105887,
+      "grad_norm": 3.591762553969462,
+      "learning_rate": 4.875370474533339e-06,
+      "loss": 1.4474,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6813111378067617,
+      "grad_norm": 3.01527327008247,
+      "learning_rate": 4.8705982392806115e-06,
+      "loss": 1.4743,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6814827527029346,
+      "grad_norm": 4.530116107800682,
+      "learning_rate": 4.865827588782932e-06,
+      "loss": 1.8223,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6816543675991076,
+      "grad_norm": 3.2203214313951447,
+      "learning_rate": 4.861058524514219e-06,
+      "loss": 1.4948,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6818259824952806,
+      "grad_norm": 4.499934906807462,
+      "learning_rate": 4.856291047947904e-06,
+      "loss": 1.6804,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6819975973914536,
+      "grad_norm": 3.84182167045221,
+      "learning_rate": 4.8515251605569155e-06,
+      "loss": 1.508,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6821692122876266,
+      "grad_norm": 3.679629477015102,
+      "learning_rate": 4.846760863813707e-06,
+      "loss": 1.4963,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6823408271837995,
+      "grad_norm": 3.2077864245371317,
+      "learning_rate": 4.841998159190224e-06,
+      "loss": 1.3114,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6825124420799725,
+      "grad_norm": 3.2994927698894694,
+      "learning_rate": 4.837237048157933e-06,
+      "loss": 1.4663,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6826840569761455,
+      "grad_norm": 5.101126384386216,
+      "learning_rate": 4.832477532187805e-06,
+      "loss": 1.5771,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6828556718723185,
+      "grad_norm": 3.5809334318824257,
+      "learning_rate": 4.827719612750318e-06,
+      "loss": 1.4568,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6830272867684916,
+      "grad_norm": 3.941324145961898,
+      "learning_rate": 4.822963291315453e-06,
+      "loss": 1.478,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6831989016646645,
+      "grad_norm": 3.9843793250742983,
+      "learning_rate": 4.818208569352697e-06,
+      "loss": 1.4362,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6833705165608375,
+      "grad_norm": 3.407457794298381,
+      "learning_rate": 4.813455448331048e-06,
+      "loss": 1.5836,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6835421314570105,
+      "grad_norm": 3.7493825930728195,
+      "learning_rate": 4.808703929719009e-06,
+      "loss": 1.5222,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6837137463531835,
+      "grad_norm": 4.707694698208893,
+      "learning_rate": 4.803954014984589e-06,
+      "loss": 1.5301,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6838853612493564,
+      "grad_norm": 3.6196378584014792,
+      "learning_rate": 4.799205705595294e-06,
+      "loss": 1.7968,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6840569761455294,
+      "grad_norm": 3.692811793847762,
+      "learning_rate": 4.794459003018137e-06,
+      "loss": 1.7723,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6842285910417024,
+      "grad_norm": 3.4371220748045483,
+      "learning_rate": 4.7897139087196385e-06,
+      "loss": 1.3896,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6844002059378754,
+      "grad_norm": 4.620643225754141,
+      "learning_rate": 4.7849704241658205e-06,
+      "loss": 1.6415,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6845718208340484,
+      "grad_norm": 7.274883933376554,
+      "learning_rate": 4.780228550822209e-06,
+      "loss": 1.4716,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6847434357302213,
+      "grad_norm": 3.6952095520194157,
+      "learning_rate": 4.775488290153824e-06,
+      "loss": 1.7042,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6849150506263944,
+      "grad_norm": 3.590090572326343,
+      "learning_rate": 4.7707496436252e-06,
+      "loss": 1.7137,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6850866655225674,
+      "grad_norm": 3.3692361967380475,
+      "learning_rate": 4.76601261270036e-06,
+      "loss": 1.7711,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6852582804187404,
+      "grad_norm": 3.691141304216351,
+      "learning_rate": 4.761277198842838e-06,
+      "loss": 1.546,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6854298953149134,
+      "grad_norm": 3.460612458724902,
+      "learning_rate": 4.756543403515666e-06,
+      "loss": 1.5853,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6856015102110863,
+      "grad_norm": 3.5285664382028434,
+      "learning_rate": 4.751811228181369e-06,
+      "loss": 1.6726,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6857731251072593,
+      "grad_norm": 3.4892684150041466,
+      "learning_rate": 4.747080674301981e-06,
+      "loss": 1.5143,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6859447400034323,
+      "grad_norm": 3.9705401165250587,
+      "learning_rate": 4.742351743339032e-06,
+      "loss": 1.8011,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6861163548996053,
+      "grad_norm": 3.7408178732131163,
+      "learning_rate": 4.737624436753544e-06,
+      "loss": 1.5359,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6862879697957782,
+      "grad_norm": 3.604401902013977,
+      "learning_rate": 4.7328987560060505e-06,
+      "loss": 1.5522,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6864595846919512,
+      "grad_norm": 4.525970195286252,
+      "learning_rate": 4.728174702556566e-06,
+      "loss": 1.7636,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6866311995881242,
+      "grad_norm": 3.9208114400596816,
+      "learning_rate": 4.723452277864617e-06,
+      "loss": 1.7299,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6868028144842973,
+      "grad_norm": 3.846729956146148,
+      "learning_rate": 4.718731483389217e-06,
+      "loss": 1.659,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6869744293804703,
+      "grad_norm": 3.4020880836863716,
+      "learning_rate": 4.7140123205888885e-06,
+      "loss": 1.7365,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6871460442766432,
+      "grad_norm": 4.3065910455440575,
+      "learning_rate": 4.709294790921634e-06,
+      "loss": 1.5898,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6873176591728162,
+      "grad_norm": 4.421336171931853,
+      "learning_rate": 4.704578895844956e-06,
+      "loss": 1.5378,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6874892740689892,
+      "grad_norm": 3.6931190135132277,
+      "learning_rate": 4.699864636815856e-06,
+      "loss": 1.5436,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6876608889651622,
+      "grad_norm": 3.7940630088136094,
+      "learning_rate": 4.695152015290833e-06,
+      "loss": 1.5711,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6878325038613352,
+      "grad_norm": 3.358202511057543,
+      "learning_rate": 4.690441032725876e-06,
+      "loss": 1.5664,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6880041187575081,
+      "grad_norm": 3.9204318584868174,
+      "learning_rate": 4.685731690576464e-06,
+      "loss": 1.677,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6881757336536811,
+      "grad_norm": 3.5628801706867734,
+      "learning_rate": 4.681023990297568e-06,
+      "loss": 1.3552,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6883473485498541,
+      "grad_norm": 4.575139356614996,
+      "learning_rate": 4.6763179333436635e-06,
+      "loss": 1.4469,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6885189634460271,
+      "grad_norm": 3.963644885046186,
+      "learning_rate": 4.6716135211687075e-06,
+      "loss": 1.2191,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6886905783422002,
+      "grad_norm": 3.696514661951681,
+      "learning_rate": 4.666910755226157e-06,
+      "loss": 1.4355,
+      "step": 4013
+    },
+    {
+      "epoch": 0.688862193238373,
+      "grad_norm": 3.513993535842884,
+      "learning_rate": 4.66220963696895e-06,
+      "loss": 1.5328,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6890338081345461,
+      "grad_norm": 3.4321143564786074,
+      "learning_rate": 4.657510167849525e-06,
+      "loss": 1.491,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6892054230307191,
+      "grad_norm": 3.929852951071529,
+      "learning_rate": 4.652812349319803e-06,
+      "loss": 1.3728,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6893770379268921,
+      "grad_norm": 3.3265771772208454,
+      "learning_rate": 4.6481161828312e-06,
+      "loss": 1.6598,
+      "step": 4017
+    },
+    {
+      "epoch": 0.689548652823065,
+      "grad_norm": 3.306244487602365,
+      "learning_rate": 4.643421669834628e-06,
+      "loss": 1.4037,
+      "step": 4018
+    },
+    {
+      "epoch": 0.689720267719238,
+      "grad_norm": 4.06921468185158,
+      "learning_rate": 4.638728811780471e-06,
+      "loss": 1.6693,
+      "step": 4019
+    },
+    {
+      "epoch": 0.689891882615411,
+      "grad_norm": 3.593139803176703,
+      "learning_rate": 4.634037610118616e-06,
+      "loss": 1.2728,
+      "step": 4020
+    },
+    {
+      "epoch": 0.690063497511584,
+      "grad_norm": 4.132664894066803,
+      "learning_rate": 4.629348066298438e-06,
+      "loss": 1.5557,
+      "step": 4021
+    },
+    {
+      "epoch": 0.690235112407757,
+      "grad_norm": 4.290982804283574,
+      "learning_rate": 4.624660181768788e-06,
+      "loss": 1.549,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6904067273039299,
+      "grad_norm": 3.3989843088938216,
+      "learning_rate": 4.619973957978018e-06,
+      "loss": 1.3847,
+      "step": 4023
+    },
+    {
+      "epoch": 0.690578342200103,
+      "grad_norm": 3.3372460207993027,
+      "learning_rate": 4.6152893963739555e-06,
+      "loss": 1.5603,
+      "step": 4024
+    },
+    {
+      "epoch": 0.690749957096276,
+      "grad_norm": 3.7083933807875407,
+      "learning_rate": 4.610606498403924e-06,
+      "loss": 1.5215,
+      "step": 4025
+    },
+    {
+      "epoch": 0.690921571992449,
+      "grad_norm": 3.56722328582263,
+      "learning_rate": 4.605925265514726e-06,
+      "loss": 1.4015,
+      "step": 4026
+    },
+    {
+      "epoch": 0.691093186888622,
+      "grad_norm": 3.7741877114155358,
+      "learning_rate": 4.601245699152659e-06,
+      "loss": 1.5962,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6912648017847949,
+      "grad_norm": 3.3260800052987745,
+      "learning_rate": 4.596567800763494e-06,
+      "loss": 1.7645,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6914364166809679,
+      "grad_norm": 4.328836835025154,
+      "learning_rate": 4.591891571792489e-06,
+      "loss": 1.8139,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6916080315771409,
+      "grad_norm": 3.828612340224736,
+      "learning_rate": 4.587217013684392e-06,
+      "loss": 1.5977,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6917796464733139,
+      "grad_norm": 3.9756673634923003,
+      "learning_rate": 4.582544127883432e-06,
+      "loss": 1.7308,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6919512613694869,
+      "grad_norm": 3.464498078097296,
+      "learning_rate": 4.5778729158333245e-06,
+      "loss": 1.5444,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6921228762656598,
+      "grad_norm": 3.843080611596852,
+      "learning_rate": 4.57320337897726e-06,
+      "loss": 1.5149,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6922944911618328,
+      "grad_norm": 3.7335864572779074,
+      "learning_rate": 4.568535518757915e-06,
+      "loss": 1.767,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6924661060580058,
+      "grad_norm": 3.693415003019225,
+      "learning_rate": 4.56386933661745e-06,
+      "loss": 1.5709,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6926377209541789,
+      "grad_norm": 3.7245490024756664,
+      "learning_rate": 4.559204833997507e-06,
+      "loss": 1.5156,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6928093358503518,
+      "grad_norm": 3.9924995921847755,
+      "learning_rate": 4.554542012339212e-06,
+      "loss": 1.5499,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6929809507465248,
+      "grad_norm": 3.973291131857169,
+      "learning_rate": 4.549880873083158e-06,
+      "loss": 1.8463,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6931525656426978,
+      "grad_norm": 3.433478402835639,
+      "learning_rate": 4.545221417669438e-06,
+      "loss": 1.5541,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6933241805388708,
+      "grad_norm": 4.639568255700584,
+      "learning_rate": 4.540563647537608e-06,
+      "loss": 1.6661,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6934957954350438,
+      "grad_norm": 3.450959263034471,
+      "learning_rate": 4.535907564126711e-06,
+      "loss": 1.3996,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6936674103312167,
+      "grad_norm": 3.654764062340256,
+      "learning_rate": 4.5312531688752735e-06,
+      "loss": 1.5362,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6938390252273897,
+      "grad_norm": 3.4076596318593326,
+      "learning_rate": 4.526600463221287e-06,
+      "loss": 1.42,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6940106401235627,
+      "grad_norm": 3.939674494036258,
+      "learning_rate": 4.521949448602231e-06,
+      "loss": 1.2365,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6941822550197357,
+      "grad_norm": 3.7290846451393733,
+      "learning_rate": 4.5173001264550665e-06,
+      "loss": 1.6374,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6943538699159088,
+      "grad_norm": 3.434813455322173,
+      "learning_rate": 4.5126524982162165e-06,
+      "loss": 1.6228,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6945254848120817,
+      "grad_norm": 3.720045428205574,
+      "learning_rate": 4.508006565321599e-06,
+      "loss": 1.7048,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6946970997082547,
+      "grad_norm": 3.7676548823936837,
+      "learning_rate": 4.503362329206589e-06,
+      "loss": 1.4129,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6948687146044277,
+      "grad_norm": 3.508726783470997,
+      "learning_rate": 4.498719791306051e-06,
+      "loss": 1.6265,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6950403295006007,
+      "grad_norm": 3.373178187930436,
+      "learning_rate": 4.494078953054323e-06,
+      "loss": 1.5656,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6952119443967736,
+      "grad_norm": 3.9232389110692605,
+      "learning_rate": 4.489439815885218e-06,
+      "loss": 1.4648,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6953835592929466,
+      "grad_norm": 3.6837886057667797,
+      "learning_rate": 4.484802381232018e-06,
+      "loss": 1.4707,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6955551741891196,
+      "grad_norm": 3.9613633675761895,
+      "learning_rate": 4.480166650527479e-06,
+      "loss": 1.779,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6957267890852926,
+      "grad_norm": 3.663444023305081,
+      "learning_rate": 4.475532625203837e-06,
+      "loss": 1.5855,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6958984039814656,
+      "grad_norm": 3.3590646488139084,
+      "learning_rate": 4.470900306692797e-06,
+      "loss": 1.5041,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6960700188776385,
+      "grad_norm": 3.403868381805293,
+      "learning_rate": 4.466269696425543e-06,
+      "loss": 1.4067,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6962416337738115,
+      "grad_norm": 3.4072122144410657,
+      "learning_rate": 4.461640795832721e-06,
+      "loss": 1.5374,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6964132486699846,
+      "grad_norm": 4.699780128207795,
+      "learning_rate": 4.457013606344451e-06,
+      "loss": 1.4838,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6965848635661576,
+      "grad_norm": 3.086063252298173,
+      "learning_rate": 4.45238812939033e-06,
+      "loss": 1.4686,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6967564784623306,
+      "grad_norm": 3.8135958341538974,
+      "learning_rate": 4.447764366399424e-06,
+      "loss": 1.6902,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6969280933585035,
+      "grad_norm": 3.932678189265142,
+      "learning_rate": 4.44314231880027e-06,
+      "loss": 1.6853,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6970997082546765,
+      "grad_norm": 3.733872867187891,
+      "learning_rate": 4.438521988020867e-06,
+      "loss": 1.5015,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6972713231508495,
+      "grad_norm": 6.375278236006696,
+      "learning_rate": 4.4339033754886974e-06,
+      "loss": 1.602,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6974429380470225,
+      "grad_norm": 5.154963652048058,
+      "learning_rate": 4.429286482630698e-06,
+      "loss": 1.6382,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6976145529431955,
+      "grad_norm": 3.672128824074522,
+      "learning_rate": 4.424671310873286e-06,
+      "loss": 1.5292,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6977861678393684,
+      "grad_norm": 3.744522790354073,
+      "learning_rate": 4.420057861642344e-06,
+      "loss": 1.7362,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6979577827355414,
+      "grad_norm": 3.1371099955769894,
+      "learning_rate": 4.415446136363216e-06,
+      "loss": 1.3375,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6981293976317144,
+      "grad_norm": 3.945222515907118,
+      "learning_rate": 4.41083613646072e-06,
+      "loss": 1.7616,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6983010125278875,
+      "grad_norm": 3.91561660874161,
+      "learning_rate": 4.406227863359144e-06,
+      "loss": 1.9275,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6984726274240604,
+      "grad_norm": 3.44762149640812,
+      "learning_rate": 4.40162131848223e-06,
+      "loss": 1.4889,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6986442423202334,
+      "grad_norm": 3.4902156664961175,
+      "learning_rate": 4.3970165032532e-06,
+      "loss": 1.4691,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6988158572164064,
+      "grad_norm": 3.231608556726426,
+      "learning_rate": 4.39241341909473e-06,
+      "loss": 1.4245,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6989874721125794,
+      "grad_norm": 3.616164952747308,
+      "learning_rate": 4.387812067428968e-06,
+      "loss": 1.4924,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6991590870087524,
+      "grad_norm": 4.397638224480771,
+      "learning_rate": 4.383212449677528e-06,
+      "loss": 1.3925,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6993307019049253,
+      "grad_norm": 3.603148272767758,
+      "learning_rate": 4.378614567261487e-06,
+      "loss": 1.4656,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6995023168010983,
+      "grad_norm": 3.9208723980667224,
+      "learning_rate": 4.3740184216013805e-06,
+      "loss": 1.5937,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6996739316972713,
+      "grad_norm": 3.399511519988169,
+      "learning_rate": 4.369424014117209e-06,
+      "loss": 1.6229,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6998455465934443,
+      "grad_norm": 3.7397111407194417,
+      "learning_rate": 4.364831346228442e-06,
+      "loss": 1.535,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7000171614896173,
+      "grad_norm": 3.52731546507338,
+      "learning_rate": 4.360240419354007e-06,
+      "loss": 1.4624,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7001887763857902,
+      "grad_norm": 4.43507463136684,
+      "learning_rate": 4.355651234912299e-06,
+      "loss": 1.6445,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7003603912819633,
+      "grad_norm": 3.7480208567212716,
+      "learning_rate": 4.3510637943211656e-06,
+      "loss": 1.5341,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7005320061781363,
+      "grad_norm": 3.694172971813163,
+      "learning_rate": 4.346478098997917e-06,
+      "loss": 1.6232,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7007036210743093,
+      "grad_norm": 3.266704945075871,
+      "learning_rate": 4.341894150359332e-06,
+      "loss": 1.5689,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7008752359704823,
+      "grad_norm": 3.455235367572192,
+      "learning_rate": 4.337311949821642e-06,
+      "loss": 1.6549,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7010468508666552,
+      "grad_norm": 3.7691337026217604,
+      "learning_rate": 4.3327314988005475e-06,
+      "loss": 1.4269,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7012184657628282,
+      "grad_norm": 3.7565675013125857,
+      "learning_rate": 4.328152798711196e-06,
+      "loss": 1.5472,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7013900806590012,
+      "grad_norm": 3.3930873798837067,
+      "learning_rate": 4.323575850968206e-06,
+      "loss": 1.5944,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7015616955551742,
+      "grad_norm": 3.6312343951114032,
+      "learning_rate": 4.319000656985641e-06,
+      "loss": 1.5139,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7017333104513471,
+      "grad_norm": 4.0170682433801765,
+      "learning_rate": 4.314427218177037e-06,
+      "loss": 1.4584,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7019049253475201,
+      "grad_norm": 3.3218281178629168,
+      "learning_rate": 4.309855535955383e-06,
+      "loss": 1.3446,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7020765402436931,
+      "grad_norm": 3.63206462801692,
+      "learning_rate": 4.305285611733117e-06,
+      "loss": 1.5769,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7022481551398662,
+      "grad_norm": 3.8142679342141164,
+      "learning_rate": 4.300717446922144e-06,
+      "loss": 1.7094,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7024197700360392,
+      "grad_norm": 3.64003018876382,
+      "learning_rate": 4.296151042933826e-06,
+      "loss": 1.272,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7025913849322121,
+      "grad_norm": 3.604322506991338,
+      "learning_rate": 4.29158640117897e-06,
+      "loss": 1.6433,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7027629998283851,
+      "grad_norm": 3.2995988187359124,
+      "learning_rate": 4.287023523067853e-06,
+      "loss": 1.5728,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7029346147245581,
+      "grad_norm": 3.334511565697689,
+      "learning_rate": 4.28246241001019e-06,
+      "loss": 1.564,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7031062296207311,
+      "grad_norm": 4.081831995536258,
+      "learning_rate": 4.277903063415167e-06,
+      "loss": 1.474,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7032778445169041,
+      "grad_norm": 3.578665855917808,
+      "learning_rate": 4.273345484691416e-06,
+      "loss": 1.5292,
+      "step": 4098
+    },
+    {
+      "epoch": 0.703449459413077,
+      "grad_norm": 3.5301537985466687,
+      "learning_rate": 4.268789675247029e-06,
+      "loss": 1.4884,
+      "step": 4099
+    },
+    {
+      "epoch": 0.70362107430925,
+      "grad_norm": 3.9266155593547385,
+      "learning_rate": 4.264235636489542e-06,
+      "loss": 1.5206,
+      "step": 4100
+    },
+    {
+      "epoch": 0.703792689205423,
+      "grad_norm": 3.6417470684530726,
+      "learning_rate": 4.259683369825945e-06,
+      "loss": 1.6743,
+      "step": 4101
+    },
+    {
+      "epoch": 0.703964304101596,
+      "grad_norm": 4.112793970090927,
+      "learning_rate": 4.25513287666269e-06,
+      "loss": 1.5537,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7041359189977691,
+      "grad_norm": 3.8793763042043365,
+      "learning_rate": 4.250584158405673e-06,
+      "loss": 1.544,
+      "step": 4103
+    },
+    {
+      "epoch": 0.704307533893942,
+      "grad_norm": 4.2882187494149235,
+      "learning_rate": 4.246037216460248e-06,
+      "loss": 1.6455,
+      "step": 4104
+    },
+    {
+      "epoch": 0.704479148790115,
+      "grad_norm": 4.237907550307505,
+      "learning_rate": 4.241492052231213e-06,
+      "loss": 1.4795,
+      "step": 4105
+    },
+    {
+      "epoch": 0.704650763686288,
+      "grad_norm": 4.150333992739989,
+      "learning_rate": 4.236948667122815e-06,
+      "loss": 1.5497,
+      "step": 4106
+    },
+    {
+      "epoch": 0.704822378582461,
+      "grad_norm": 3.6604784973274636,
+      "learning_rate": 4.232407062538761e-06,
+      "loss": 1.4729,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7049939934786339,
+      "grad_norm": 3.89555778492882,
+      "learning_rate": 4.227867239882203e-06,
+      "loss": 1.5899,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7051656083748069,
+      "grad_norm": 3.8994460047266943,
+      "learning_rate": 4.223329200555747e-06,
+      "loss": 1.6008,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7053372232709799,
+      "grad_norm": 3.2922178156879975,
+      "learning_rate": 4.218792945961434e-06,
+      "loss": 1.5273,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7055088381671529,
+      "grad_norm": 3.2643824493793194,
+      "learning_rate": 4.214258477500772e-06,
+      "loss": 1.5119,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7056804530633259,
+      "grad_norm": 3.6014945434698267,
+      "learning_rate": 4.2097257965747005e-06,
+      "loss": 1.4116,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7058520679594988,
+      "grad_norm": 3.850852055560369,
+      "learning_rate": 4.205194904583619e-06,
+      "loss": 1.5465,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7060236828556719,
+      "grad_norm": 4.299358524114042,
+      "learning_rate": 4.200665802927373e-06,
+      "loss": 1.4715,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7061952977518449,
+      "grad_norm": 3.472152201839758,
+      "learning_rate": 4.196138493005245e-06,
+      "loss": 1.5562,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7063669126480179,
+      "grad_norm": 4.465213204247648,
+      "learning_rate": 4.191612976215973e-06,
+      "loss": 1.722,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7065385275441909,
+      "grad_norm": 3.983595751429719,
+      "learning_rate": 4.1870892539577435e-06,
+      "loss": 1.5408,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7067101424403638,
+      "grad_norm": 3.964024583986105,
+      "learning_rate": 4.182567327628177e-06,
+      "loss": 1.6629,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7068817573365368,
+      "grad_norm": 3.658129939311069,
+      "learning_rate": 4.178047198624351e-06,
+      "loss": 1.3729,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7070533722327098,
+      "grad_norm": 3.9613033011493473,
+      "learning_rate": 4.173528868342777e-06,
+      "loss": 1.5596,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7072249871288828,
+      "grad_norm": 3.8130375825594163,
+      "learning_rate": 4.16901233817942e-06,
+      "loss": 1.47,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7073966020250557,
+      "grad_norm": 3.4381252294591302,
+      "learning_rate": 4.164497609529685e-06,
+      "loss": 1.6529,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7075682169212287,
+      "grad_norm": 3.2021211782658074,
+      "learning_rate": 4.159984683788424e-06,
+      "loss": 1.4202,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7077398318174017,
+      "grad_norm": 4.073196476611487,
+      "learning_rate": 4.1554735623499244e-06,
+      "loss": 1.6303,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7079114467135748,
+      "grad_norm": 3.364889360925374,
+      "learning_rate": 4.150964246607918e-06,
+      "loss": 1.5816,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7080830616097478,
+      "grad_norm": 3.545722653185063,
+      "learning_rate": 4.1464567379555846e-06,
+      "loss": 1.4743,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7082546765059207,
+      "grad_norm": 3.8830091650422114,
+      "learning_rate": 4.141951037785541e-06,
+      "loss": 1.8062,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7084262914020937,
+      "grad_norm": 4.245335656554135,
+      "learning_rate": 4.1374471474898514e-06,
+      "loss": 1.4273,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7085979062982667,
+      "grad_norm": 3.958793628954796,
+      "learning_rate": 4.132945068460011e-06,
+      "loss": 1.6064,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7087695211944397,
+      "grad_norm": 4.658720578637188,
+      "learning_rate": 4.128444802086958e-06,
+      "loss": 1.6416,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7089411360906127,
+      "grad_norm": 4.419496815867452,
+      "learning_rate": 4.123946349761075e-06,
+      "loss": 1.5202,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7091127509867856,
+      "grad_norm": 3.9191607510646786,
+      "learning_rate": 4.1194497128721846e-06,
+      "loss": 1.6093,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7092843658829586,
+      "grad_norm": 3.802835223864519,
+      "learning_rate": 4.114954892809545e-06,
+      "loss": 1.3616,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7094559807791316,
+      "grad_norm": 4.804130466472091,
+      "learning_rate": 4.110461890961851e-06,
+      "loss": 1.5285,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7096275956753046,
+      "grad_norm": 3.5210648795658868,
+      "learning_rate": 4.105970708717244e-06,
+      "loss": 1.7624,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7097992105714777,
+      "grad_norm": 4.190089036123137,
+      "learning_rate": 4.101481347463291e-06,
+      "loss": 1.6321,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7099708254676506,
+      "grad_norm": 3.844657666967626,
+      "learning_rate": 4.096993808587007e-06,
+      "loss": 1.5408,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7101424403638236,
+      "grad_norm": 3.3821232935527044,
+      "learning_rate": 4.092508093474841e-06,
+      "loss": 1.6364,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7103140552599966,
+      "grad_norm": 3.8616701090670578,
+      "learning_rate": 4.088024203512676e-06,
+      "loss": 1.7954,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7104856701561696,
+      "grad_norm": 3.9356325259720935,
+      "learning_rate": 4.083542140085832e-06,
+      "loss": 1.8199,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7106572850523425,
+      "grad_norm": 3.237348938620819,
+      "learning_rate": 4.079061904579071e-06,
+      "loss": 1.2986,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7108288999485155,
+      "grad_norm": 3.7419993517779364,
+      "learning_rate": 4.074583498376577e-06,
+      "loss": 1.5519,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7110005148446885,
+      "grad_norm": 3.354023949317062,
+      "learning_rate": 4.070106922861985e-06,
+      "loss": 1.6365,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7111721297408615,
+      "grad_norm": 4.031039241130169,
+      "learning_rate": 4.065632179418348e-06,
+      "loss": 1.6337,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7113437446370345,
+      "grad_norm": 4.066715301336402,
+      "learning_rate": 4.061159269428165e-06,
+      "loss": 1.728,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7115153595332074,
+      "grad_norm": 3.897809089063765,
+      "learning_rate": 4.056688194273362e-06,
+      "loss": 1.6542,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7116869744293804,
+      "grad_norm": 4.4778195751932826,
+      "learning_rate": 4.0522189553353075e-06,
+      "loss": 1.4832,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7118585893255535,
+      "grad_norm": 3.181424112780336,
+      "learning_rate": 4.047751553994791e-06,
+      "loss": 1.3256,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7120302042217265,
+      "grad_norm": 3.3504020389732947,
+      "learning_rate": 4.043285991632035e-06,
+      "loss": 1.4827,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7122018191178995,
+      "grad_norm": 3.3747065996638423,
+      "learning_rate": 4.038822269626703e-06,
+      "loss": 1.3242,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7123734340140724,
+      "grad_norm": 4.009696666042803,
+      "learning_rate": 4.034360389357883e-06,
+      "loss": 1.4059,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7125450489102454,
+      "grad_norm": 4.102033684209833,
+      "learning_rate": 4.029900352204102e-06,
+      "loss": 1.5107,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7127166638064184,
+      "grad_norm": 3.00024042828498,
+      "learning_rate": 4.0254421595433045e-06,
+      "loss": 1.2681,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7128882787025914,
+      "grad_norm": 3.216592227468947,
+      "learning_rate": 4.020985812752871e-06,
+      "loss": 1.6086,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7130598935987644,
+      "grad_norm": 3.654491742778158,
+      "learning_rate": 4.016531313209616e-06,
+      "loss": 1.6507,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7132315084949373,
+      "grad_norm": 3.2104212583477416,
+      "learning_rate": 4.01207866228978e-06,
+      "loss": 1.3259,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7134031233911103,
+      "grad_norm": 4.413950328751978,
+      "learning_rate": 4.007627861369034e-06,
+      "loss": 1.5079,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7135747382872833,
+      "grad_norm": 3.9291451747795056,
+      "learning_rate": 4.003178911822472e-06,
+      "loss": 1.6011,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7137463531834564,
+      "grad_norm": 3.6307195680910533,
+      "learning_rate": 3.9987318150246235e-06,
+      "loss": 1.5678,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7139179680796293,
+      "grad_norm": 3.649961139626407,
+      "learning_rate": 3.994286572349438e-06,
+      "loss": 1.4753,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7140895829758023,
+      "grad_norm": 3.905880519082177,
+      "learning_rate": 3.989843185170298e-06,
+      "loss": 1.7013,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7142611978719753,
+      "grad_norm": 4.337249873636874,
+      "learning_rate": 3.985401654860014e-06,
+      "loss": 1.4474,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7144328127681483,
+      "grad_norm": 3.7770609259127053,
+      "learning_rate": 3.980961982790813e-06,
+      "loss": 1.4295,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7146044276643213,
+      "grad_norm": 3.747623510264429,
+      "learning_rate": 3.976524170334358e-06,
+      "loss": 1.4052,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7147760425604942,
+      "grad_norm": 3.544264003828599,
+      "learning_rate": 3.972088218861738e-06,
+      "loss": 1.347,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7149476574566672,
+      "grad_norm": 3.3736560903106207,
+      "learning_rate": 3.967654129743457e-06,
+      "loss": 1.7092,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7151192723528402,
+      "grad_norm": 3.5771774790171027,
+      "learning_rate": 3.963221904349454e-06,
+      "loss": 1.5427,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7152908872490132,
+      "grad_norm": 3.392893352972987,
+      "learning_rate": 3.958791544049082e-06,
+      "loss": 1.4877,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7154625021451863,
+      "grad_norm": 3.290042003772946,
+      "learning_rate": 3.9543630502111266e-06,
+      "loss": 1.3702,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7156341170413592,
+      "grad_norm": 3.692144051494739,
+      "learning_rate": 3.949936424203795e-06,
+      "loss": 1.5409,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7158057319375322,
+      "grad_norm": 3.552685893909201,
+      "learning_rate": 3.945511667394719e-06,
+      "loss": 1.5695,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7159773468337052,
+      "grad_norm": 4.186359995509498,
+      "learning_rate": 3.9410887811509465e-06,
+      "loss": 1.6776,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7161489617298782,
+      "grad_norm": 3.863766662151149,
+      "learning_rate": 3.936667766838947e-06,
+      "loss": 1.8969,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7163205766260511,
+      "grad_norm": 4.012929081461003,
+      "learning_rate": 3.9322486258246195e-06,
+      "loss": 1.6432,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7164921915222241,
+      "grad_norm": 3.715953431646086,
+      "learning_rate": 3.927831359473281e-06,
+      "loss": 1.4083,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7166638064183971,
+      "grad_norm": 2.982423299890264,
+      "learning_rate": 3.923415969149671e-06,
+      "loss": 1.666,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7168354213145701,
+      "grad_norm": 3.997966294065927,
+      "learning_rate": 3.919002456217944e-06,
+      "loss": 1.5069,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7170070362107431,
+      "grad_norm": 3.5570816577556554,
+      "learning_rate": 3.914590822041673e-06,
+      "loss": 1.6799,
+      "step": 4178
+    },
+    {
+      "epoch": 0.717178651106916,
+      "grad_norm": 3.441927183853842,
+      "learning_rate": 3.91018106798386e-06,
+      "loss": 1.4168,
+      "step": 4179
+    },
+    {
+      "epoch": 0.717350266003089,
+      "grad_norm": 3.886653666454556,
+      "learning_rate": 3.9057731954069225e-06,
+      "loss": 1.6133,
+      "step": 4180
+    },
+    {
+      "epoch": 0.717521880899262,
+      "grad_norm": 3.470311096561256,
+      "learning_rate": 3.901367205672695e-06,
+      "loss": 1.5119,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7176934957954351,
+      "grad_norm": 4.073690221848995,
+      "learning_rate": 3.896963100142427e-06,
+      "loss": 1.4679,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7178651106916081,
+      "grad_norm": 3.8517048673548637,
+      "learning_rate": 3.892560880176795e-06,
+      "loss": 1.4875,
+      "step": 4183
+    },
+    {
+      "epoch": 0.718036725587781,
+      "grad_norm": 3.794484124652863,
+      "learning_rate": 3.888160547135881e-06,
+      "loss": 1.5112,
+      "step": 4184
+    },
+    {
+      "epoch": 0.718208340483954,
+      "grad_norm": 3.752988556760734,
+      "learning_rate": 3.883762102379193e-06,
+      "loss": 1.5393,
+      "step": 4185
+    },
+    {
+      "epoch": 0.718379955380127,
+      "grad_norm": 3.9074788540079135,
+      "learning_rate": 3.8793655472656585e-06,
+      "loss": 1.3333,
+      "step": 4186
+    },
+    {
+      "epoch": 0.7185515702763,
+      "grad_norm": 3.9272319866250984,
+      "learning_rate": 3.874970883153607e-06,
+      "loss": 1.4151,
+      "step": 4187
+    },
+    {
+      "epoch": 0.718723185172473,
+      "grad_norm": 3.857044572606296,
+      "learning_rate": 3.870578111400798e-06,
+      "loss": 1.6884,
+      "step": 4188
+    },
+    {
+      "epoch": 0.7188948000686459,
+      "grad_norm": 3.8667492396619743,
+      "learning_rate": 3.866187233364402e-06,
+      "loss": 1.7396,
+      "step": 4189
+    },
+    {
+      "epoch": 0.7190664149648189,
+      "grad_norm": 4.05232268287602,
+      "learning_rate": 3.861798250400998e-06,
+      "loss": 1.5502,
+      "step": 4190
+    },
+    {
+      "epoch": 0.719238029860992,
+      "grad_norm": 4.056784155266631,
+      "learning_rate": 3.857411163866593e-06,
+      "loss": 1.765,
+      "step": 4191
+    },
+    {
+      "epoch": 0.719409644757165,
+      "grad_norm": 4.2670263815396,
+      "learning_rate": 3.853025975116591e-06,
+      "loss": 1.6919,
+      "step": 4192
+    },
+    {
+      "epoch": 0.7195812596533379,
+      "grad_norm": 4.344422849915019,
+      "learning_rate": 3.8486426855058214e-06,
+      "loss": 1.79,
+      "step": 4193
+    },
+    {
+      "epoch": 0.7197528745495109,
+      "grad_norm": 3.6431004901370896,
+      "learning_rate": 3.844261296388526e-06,
+      "loss": 1.5082,
+      "step": 4194
+    },
+    {
+      "epoch": 0.7199244894456839,
+      "grad_norm": 4.427966420765549,
+      "learning_rate": 3.83988180911836e-06,
+      "loss": 1.4694,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7200961043418569,
+      "grad_norm": 4.580751526339576,
+      "learning_rate": 3.835504225048385e-06,
+      "loss": 1.4903,
+      "step": 4196
+    },
+    {
+      "epoch": 0.7202677192380299,
+      "grad_norm": 3.718492663068725,
+      "learning_rate": 3.831128545531076e-06,
+      "loss": 1.4353,
+      "step": 4197
+    },
+    {
+      "epoch": 0.7204393341342028,
+      "grad_norm": 3.7623360213805883,
+      "learning_rate": 3.8267547719183225e-06,
+      "loss": 1.4407,
+      "step": 4198
+    },
+    {
+      "epoch": 0.7206109490303758,
+      "grad_norm": 2.8888406331867293,
+      "learning_rate": 3.822382905561428e-06,
+      "loss": 1.3856,
+      "step": 4199
+    },
+    {
+      "epoch": 0.7207825639265488,
+      "grad_norm": 3.9904595803750644,
+      "learning_rate": 3.818012947811103e-06,
+      "loss": 1.6361,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7209541788227218,
+      "grad_norm": 4.40424436074814,
+      "learning_rate": 3.8136449000174667e-06,
+      "loss": 1.5679,
+      "step": 4201
+    },
+    {
+      "epoch": 0.7211257937188948,
+      "grad_norm": 4.3122044623779985,
+      "learning_rate": 3.8092787635300455e-06,
+      "loss": 1.6252,
+      "step": 4202
+    },
+    {
+      "epoch": 0.7212974086150677,
+      "grad_norm": 3.0944638527457875,
+      "learning_rate": 3.804914539697785e-06,
+      "loss": 1.2539,
+      "step": 4203
+    },
+    {
+      "epoch": 0.7214690235112408,
+      "grad_norm": 3.4475272908740187,
+      "learning_rate": 3.8005522298690333e-06,
+      "loss": 1.3981,
+      "step": 4204
+    },
+    {
+      "epoch": 0.7216406384074138,
+      "grad_norm": 3.553179655067805,
+      "learning_rate": 3.796191835391552e-06,
+      "loss": 1.5454,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7218122533035868,
+      "grad_norm": 3.711527962389041,
+      "learning_rate": 3.7918333576124997e-06,
+      "loss": 1.5268,
+      "step": 4206
+    },
+    {
+      "epoch": 0.7219838681997598,
+      "grad_norm": 4.226263184512579,
+      "learning_rate": 3.787476797878459e-06,
+      "loss": 1.6009,
+      "step": 4207
+    },
+    {
+      "epoch": 0.7221554830959327,
+      "grad_norm": 3.4436668630869995,
+      "learning_rate": 3.7831221575354036e-06,
+      "loss": 1.6827,
+      "step": 4208
+    },
+    {
+      "epoch": 0.7223270979921057,
+      "grad_norm": 3.699911617745991,
+      "learning_rate": 3.7787694379287256e-06,
+      "loss": 1.5477,
+      "step": 4209
+    },
+    {
+      "epoch": 0.7224987128882787,
+      "grad_norm": 3.8847746616409933,
+      "learning_rate": 3.7744186404032233e-06,
+      "loss": 1.7099,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7226703277844517,
+      "grad_norm": 4.239215143982443,
+      "learning_rate": 3.7700697663030916e-06,
+      "loss": 1.5431,
+      "step": 4211
+    },
+    {
+      "epoch": 0.7228419426806246,
+      "grad_norm": 3.8365329567483775,
+      "learning_rate": 3.765722816971942e-06,
+      "loss": 1.4774,
+      "step": 4212
+    },
+    {
+      "epoch": 0.7230135575767976,
+      "grad_norm": 3.444829700041154,
+      "learning_rate": 3.7613777937527873e-06,
+      "loss": 1.3581,
+      "step": 4213
+    },
+    {
+      "epoch": 0.7231851724729706,
+      "grad_norm": 4.17790174750879,
+      "learning_rate": 3.7570346979880412e-06,
+      "loss": 1.4967,
+      "step": 4214
+    },
+    {
+      "epoch": 0.7233567873691437,
+      "grad_norm": 5.598203431995362,
+      "learning_rate": 3.75269353101953e-06,
+      "loss": 1.5852,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7235284022653167,
+      "grad_norm": 3.7113210566465127,
+      "learning_rate": 3.7483542941884733e-06,
+      "loss": 1.3141,
+      "step": 4216
+    },
+    {
+      "epoch": 0.7237000171614896,
+      "grad_norm": 3.694554808473566,
+      "learning_rate": 3.744016988835504e-06,
+      "loss": 1.4322,
+      "step": 4217
+    },
+    {
+      "epoch": 0.7238716320576626,
+      "grad_norm": 4.118538554416828,
+      "learning_rate": 3.7396816163006554e-06,
+      "loss": 1.615,
+      "step": 4218
+    },
+    {
+      "epoch": 0.7240432469538356,
+      "grad_norm": 3.687426293700273,
+      "learning_rate": 3.735348177923367e-06,
+      "loss": 1.4754,
+      "step": 4219
+    },
+    {
+      "epoch": 0.7242148618500086,
+      "grad_norm": 3.246083589221187,
+      "learning_rate": 3.7310166750424727e-06,
+      "loss": 1.5051,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7243864767461816,
+      "grad_norm": 3.6638808136753775,
+      "learning_rate": 3.7266871089962086e-06,
+      "loss": 1.5304,
+      "step": 4221
+    },
+    {
+      "epoch": 0.7245580916423545,
+      "grad_norm": 3.6315075448144403,
+      "learning_rate": 3.7223594811222197e-06,
+      "loss": 1.4357,
+      "step": 4222
+    },
+    {
+      "epoch": 0.7247297065385275,
+      "grad_norm": 3.843547297793905,
+      "learning_rate": 3.7180337927575507e-06,
+      "loss": 1.5876,
+      "step": 4223
+    },
+    {
+      "epoch": 0.7249013214347005,
+      "grad_norm": 3.582255070343444,
+      "learning_rate": 3.7137100452386478e-06,
+      "loss": 1.4231,
+      "step": 4224
+    },
+    {
+      "epoch": 0.7250729363308736,
+      "grad_norm": 4.068006016681625,
+      "learning_rate": 3.7093882399013504e-06,
+      "loss": 1.615,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7252445512270465,
+      "grad_norm": 3.6184958776841585,
+      "learning_rate": 3.7050683780809016e-06,
+      "loss": 1.501,
+      "step": 4226
+    },
+    {
+      "epoch": 0.7254161661232195,
+      "grad_norm": 3.488722342137974,
+      "learning_rate": 3.700750461111945e-06,
+      "loss": 1.5459,
+      "step": 4227
+    },
+    {
+      "epoch": 0.7255877810193925,
+      "grad_norm": 3.9436530743783336,
+      "learning_rate": 3.6964344903285267e-06,
+      "loss": 1.4607,
+      "step": 4228
+    },
+    {
+      "epoch": 0.7257593959155655,
+      "grad_norm": 4.540621473782923,
+      "learning_rate": 3.6921204670640886e-06,
+      "loss": 1.6634,
+      "step": 4229
+    },
+    {
+      "epoch": 0.7259310108117385,
+      "grad_norm": 3.727330336649548,
+      "learning_rate": 3.6878083926514654e-06,
+      "loss": 1.6113,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7261026257079114,
+      "grad_norm": 3.694818917390779,
+      "learning_rate": 3.6834982684229003e-06,
+      "loss": 1.5466,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7262742406040844,
+      "grad_norm": 3.8740477659751558,
+      "learning_rate": 3.6791900957100224e-06,
+      "loss": 1.5952,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7264458555002574,
+      "grad_norm": 4.374677312714116,
+      "learning_rate": 3.674883875843868e-06,
+      "loss": 1.4432,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7266174703964304,
+      "grad_norm": 4.811934275320881,
+      "learning_rate": 3.6705796101548664e-06,
+      "loss": 1.527,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7267890852926034,
+      "grad_norm": 3.9073849723155996,
+      "learning_rate": 3.6662772999728392e-06,
+      "loss": 1.4628,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7269607001887763,
+      "grad_norm": 4.108912474902295,
+      "learning_rate": 3.661976946627007e-06,
+      "loss": 1.5339,
+      "step": 4236
+    },
+    {
+      "epoch": 0.7271323150849494,
+      "grad_norm": 4.476139117236807,
+      "learning_rate": 3.6576785514459933e-06,
+      "loss": 1.7686,
+      "step": 4237
+    },
+    {
+      "epoch": 0.7273039299811224,
+      "grad_norm": 3.832367535708342,
+      "learning_rate": 3.653382115757801e-06,
+      "loss": 1.6848,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7274755448772954,
+      "grad_norm": 3.9488026516957175,
+      "learning_rate": 3.649087640889841e-06,
+      "loss": 1.4907,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7276471597734684,
+      "grad_norm": 4.074351772137394,
+      "learning_rate": 3.6447951281689097e-06,
+      "loss": 1.3454,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7278187746696413,
+      "grad_norm": 3.6569211851443217,
+      "learning_rate": 3.6405045789212025e-06,
+      "loss": 1.5901,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7279903895658143,
+      "grad_norm": 3.545716979705867,
+      "learning_rate": 3.6362159944723075e-06,
+      "loss": 1.4607,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7281620044619873,
+      "grad_norm": 4.375404581666328,
+      "learning_rate": 3.6319293761472073e-06,
+      "loss": 1.5912,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7283336193581603,
+      "grad_norm": 4.0792360801882435,
+      "learning_rate": 3.627644725270272e-06,
+      "loss": 1.7369,
+      "step": 4244
+    },
+    {
+      "epoch": 0.7285052342543332,
+      "grad_norm": 4.326875157272401,
+      "learning_rate": 3.623362043165263e-06,
+      "loss": 1.6896,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7286768491505062,
+      "grad_norm": 4.236716240356844,
+      "learning_rate": 3.61908133115534e-06,
+      "loss": 1.6729,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7288484640466792,
+      "grad_norm": 3.930875256474182,
+      "learning_rate": 3.614802590563051e-06,
+      "loss": 1.606,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7290200789428523,
+      "grad_norm": 3.9124333770742235,
+      "learning_rate": 3.610525822710339e-06,
+      "loss": 1.5453,
+      "step": 4248
+    },
+    {
+      "epoch": 0.7291916938390253,
+      "grad_norm": 3.775756731600947,
+      "learning_rate": 3.6062510289185305e-06,
+      "loss": 1.3937,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7293633087351982,
+      "grad_norm": 4.461266934920697,
+      "learning_rate": 3.601978210508339e-06,
+      "loss": 1.6767,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7295349236313712,
+      "grad_norm": 4.162555711148938,
+      "learning_rate": 3.597707368799881e-06,
+      "loss": 1.6228,
+      "step": 4251
+    },
+    {
+      "epoch": 0.7297065385275442,
+      "grad_norm": 3.467984933239619,
+      "learning_rate": 3.5934385051126507e-06,
+      "loss": 1.4967,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7298781534237172,
+      "grad_norm": 3.1445468777807255,
+      "learning_rate": 3.5891716207655423e-06,
+      "loss": 1.4428,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7300497683198902,
+      "grad_norm": 4.31484890176354,
+      "learning_rate": 3.5849067170768237e-06,
+      "loss": 1.6264,
+      "step": 4254
+    },
+    {
+      "epoch": 0.7302213832160631,
+      "grad_norm": 5.075828693659018,
+      "learning_rate": 3.580643795364166e-06,
+      "loss": 1.4281,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7303929981122361,
+      "grad_norm": 4.245735053069527,
+      "learning_rate": 3.5763828569446122e-06,
+      "loss": 1.447,
+      "step": 4256
+    },
+    {
+      "epoch": 0.7305646130084091,
+      "grad_norm": 3.5932678574319006,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 1.559,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7307362279045821,
+      "grad_norm": 4.0156421473151855,
+      "learning_rate": 3.5678669352499783e-06,
+      "loss": 1.5363,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7309078428007552,
+      "grad_norm": 3.9641798252776175,
+      "learning_rate": 3.563611954605931e-06,
+      "loss": 1.5863,
+      "step": 4259
+    },
+    {
+      "epoch": 0.7310794576969281,
+      "grad_norm": 3.6682740174340145,
+      "learning_rate": 3.5593589625170668e-06,
+      "loss": 1.5641,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7312510725931011,
+      "grad_norm": 3.700370241280364,
+      "learning_rate": 3.5551079602973735e-06,
+      "loss": 1.4792,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7314226874892741,
+      "grad_norm": 3.6183367563133952,
+      "learning_rate": 3.5508589492602143e-06,
+      "loss": 1.522,
+      "step": 4262
+    },
+    {
+      "epoch": 0.7315943023854471,
+      "grad_norm": 3.226406926106017,
+      "learning_rate": 3.5466119307183465e-06,
+      "loss": 1.3982,
+      "step": 4263
+    },
+    {
+      "epoch": 0.73176591728162,
+      "grad_norm": 3.7197236361283026,
+      "learning_rate": 3.542366905983905e-06,
+      "loss": 1.3868,
+      "step": 4264
+    },
+    {
+      "epoch": 0.731937532177793,
+      "grad_norm": 4.219786376061305,
+      "learning_rate": 3.5381238763684134e-06,
+      "loss": 1.5433,
+      "step": 4265
+    },
+    {
+      "epoch": 0.732109147073966,
+      "grad_norm": 4.20728905880942,
+      "learning_rate": 3.5338828431827777e-06,
+      "loss": 1.5269,
+      "step": 4266
+    },
+    {
+      "epoch": 0.732280761970139,
+      "grad_norm": 3.7356235183249686,
+      "learning_rate": 3.5296438077372908e-06,
+      "loss": 1.4627,
+      "step": 4267
+    },
+    {
+      "epoch": 0.732452376866312,
+      "grad_norm": 3.9121336191067004,
+      "learning_rate": 3.5254067713416218e-06,
+      "loss": 1.4723,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7326239917624849,
+      "grad_norm": 3.9774343987033003,
+      "learning_rate": 3.5211717353048204e-06,
+      "loss": 1.3861,
+      "step": 4269
+    },
+    {
+      "epoch": 0.732795606658658,
+      "grad_norm": 3.683169633164081,
+      "learning_rate": 3.516938700935325e-06,
+      "loss": 1.3967,
+      "step": 4270
+    },
+    {
+      "epoch": 0.732967221554831,
+      "grad_norm": 3.8669803013588315,
+      "learning_rate": 3.5127076695409568e-06,
+      "loss": 1.3631,
+      "step": 4271
+    },
+    {
+      "epoch": 0.733138836451004,
+      "grad_norm": 4.060345716869961,
+      "learning_rate": 3.508478642428914e-06,
+      "loss": 1.6325,
+      "step": 4272
+    },
+    {
+      "epoch": 0.733310451347177,
+      "grad_norm": 3.3206224113874425,
+      "learning_rate": 3.5042516209057763e-06,
+      "loss": 1.3753,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7334820662433499,
+      "grad_norm": 3.2209484177252454,
+      "learning_rate": 3.500026606277499e-06,
+      "loss": 1.3568,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7336536811395229,
+      "grad_norm": 3.8786046629031325,
+      "learning_rate": 3.4958035998494255e-06,
+      "loss": 1.5166,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7338252960356959,
+      "grad_norm": 3.8555135318509115,
+      "learning_rate": 3.4915826029262745e-06,
+      "loss": 1.6155,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7339969109318689,
+      "grad_norm": 3.979545349046841,
+      "learning_rate": 3.4873636168121496e-06,
+      "loss": 1.5755,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7341685258280419,
+      "grad_norm": 3.80412037767361,
+      "learning_rate": 3.4831466428105196e-06,
+      "loss": 1.5559,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7343401407242148,
+      "grad_norm": 3.6481101800033624,
+      "learning_rate": 3.47893168222425e-06,
+      "loss": 1.4232,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7345117556203878,
+      "grad_norm": 4.380020561943973,
+      "learning_rate": 3.4747187363555655e-06,
+      "loss": 1.5165,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7346833705165609,
+      "grad_norm": 4.2090574363288535,
+      "learning_rate": 3.4705078065060804e-06,
+      "loss": 1.5358,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7348549854127339,
+      "grad_norm": 4.2454950701074,
+      "learning_rate": 3.4662988939767885e-06,
+      "loss": 1.7111,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7350266003089068,
+      "grad_norm": 4.201386951428244,
+      "learning_rate": 3.462092000068047e-06,
+      "loss": 1.7071,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7351982152050798,
+      "grad_norm": 3.9244992096232396,
+      "learning_rate": 3.4578871260796e-06,
+      "loss": 1.5966,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7353698301012528,
+      "grad_norm": 3.68781316870011,
+      "learning_rate": 3.4536842733105702e-06,
+      "loss": 1.4676,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7355414449974258,
+      "grad_norm": 4.1067657843468846,
+      "learning_rate": 3.4494834430594426e-06,
+      "loss": 1.3194,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7357130598935988,
+      "grad_norm": 3.6377749201365344,
+      "learning_rate": 3.4452846366240922e-06,
+      "loss": 1.5585,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7358846747897717,
+      "grad_norm": 4.355868122475021,
+      "learning_rate": 3.4410878553017547e-06,
+      "loss": 1.6817,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7360562896859447,
+      "grad_norm": 3.383902265050565,
+      "learning_rate": 3.4368931003890516e-06,
+      "loss": 1.522,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7362279045821177,
+      "grad_norm": 4.163455615183541,
+      "learning_rate": 3.432700373181974e-06,
+      "loss": 1.5336,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7363995194782907,
+      "grad_norm": 3.7482452895417167,
+      "learning_rate": 3.4285096749758896e-06,
+      "loss": 1.4256,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7365711343744638,
+      "grad_norm": 4.102699385731431,
+      "learning_rate": 3.424321007065532e-06,
+      "loss": 1.5015,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7367427492706367,
+      "grad_norm": 3.5676477686557293,
+      "learning_rate": 3.4201343707450087e-06,
+      "loss": 1.535,
+      "step": 4293
+    },
+    {
+      "epoch": 0.7369143641668097,
+      "grad_norm": 4.3390912734984655,
+      "learning_rate": 3.4159497673078067e-06,
+      "loss": 1.6811,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7370859790629827,
+      "grad_norm": 3.9100003490084267,
+      "learning_rate": 3.4117671980467794e-06,
+      "loss": 1.4424,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7372575939591557,
+      "grad_norm": 3.8109020729214955,
+      "learning_rate": 3.407586664254158e-06,
+      "loss": 1.764,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7374292088553286,
+      "grad_norm": 3.7640562262112667,
+      "learning_rate": 3.403408167221536e-06,
+      "loss": 1.6261,
+      "step": 4297
+    },
+    {
+      "epoch": 0.7376008237515016,
+      "grad_norm": 4.428321995632007,
+      "learning_rate": 3.3992317082398774e-06,
+      "loss": 1.3457,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7377724386476746,
+      "grad_norm": 4.0475446708955705,
+      "learning_rate": 3.395057288599525e-06,
+      "loss": 1.5551,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7379440535438476,
+      "grad_norm": 3.882998035471941,
+      "learning_rate": 3.3908849095901865e-06,
+      "loss": 1.6307,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7381156684400206,
+      "grad_norm": 3.5938893722636496,
+      "learning_rate": 3.386714572500943e-06,
+      "loss": 1.6644,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7382872833361935,
+      "grad_norm": 4.156852828694422,
+      "learning_rate": 3.3825462786202367e-06,
+      "loss": 1.5669,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7384588982323665,
+      "grad_norm": 4.9841980895273705,
+      "learning_rate": 3.378380029235888e-06,
+      "loss": 1.8169,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7386305131285396,
+      "grad_norm": 4.336164860735623,
+      "learning_rate": 3.3742158256350756e-06,
+      "loss": 1.7264,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7388021280247126,
+      "grad_norm": 3.548662828843063,
+      "learning_rate": 3.370053669104354e-06,
+      "loss": 1.6414,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7389737429208856,
+      "grad_norm": 4.9169637546645335,
+      "learning_rate": 3.3658935609296472e-06,
+      "loss": 1.5945,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7391453578170585,
+      "grad_norm": 4.520864212645562,
+      "learning_rate": 3.361735502396235e-06,
+      "loss": 1.6859,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7393169727132315,
+      "grad_norm": 4.629052609733307,
+      "learning_rate": 3.3575794947887742e-06,
+      "loss": 1.6373,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7394885876094045,
+      "grad_norm": 3.959878356321814,
+      "learning_rate": 3.3534255393912885e-06,
+      "loss": 1.4297,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7396602025055775,
+      "grad_norm": 3.8793786961428554,
+      "learning_rate": 3.349273637487156e-06,
+      "loss": 1.6751,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7398318174017505,
+      "grad_norm": 3.7592403632021174,
+      "learning_rate": 3.345123790359135e-06,
+      "loss": 1.7202,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7400034322979234,
+      "grad_norm": 3.8167671447340137,
+      "learning_rate": 3.3409759992893377e-06,
+      "loss": 1.57,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7401750471940964,
+      "grad_norm": 3.4155260456210894,
+      "learning_rate": 3.3368302655592454e-06,
+      "loss": 1.6994,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7403466620902694,
+      "grad_norm": 4.286894002448902,
+      "learning_rate": 3.3326865904497065e-06,
+      "loss": 1.8381,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7405182769864425,
+      "grad_norm": 3.936691826143307,
+      "learning_rate": 3.3285449752409315e-06,
+      "loss": 1.4833,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7406898918826154,
+      "grad_norm": 4.768572173515226,
+      "learning_rate": 3.3244054212124934e-06,
+      "loss": 1.5982,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7408615067787884,
+      "grad_norm": 3.7730004473871848,
+      "learning_rate": 3.320267929643324e-06,
+      "loss": 1.3591,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7410331216749614,
+      "grad_norm": 3.838244071972954,
+      "learning_rate": 3.316132501811725e-06,
+      "loss": 1.4783,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7412047365711344,
+      "grad_norm": 4.632103751384495,
+      "learning_rate": 3.3119991389953597e-06,
+      "loss": 1.5828,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7413763514673074,
+      "grad_norm": 3.7475944488748567,
+      "learning_rate": 3.307867842471254e-06,
+      "loss": 1.4898,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7415479663634803,
+      "grad_norm": 4.107224198793785,
+      "learning_rate": 3.30373861351579e-06,
+      "loss": 1.4434,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7417195812596533,
+      "grad_norm": 4.043393490083267,
+      "learning_rate": 3.299611453404712e-06,
+      "loss": 1.4274,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7418911961558263,
+      "grad_norm": 3.912694064358962,
+      "learning_rate": 3.2954863634131295e-06,
+      "loss": 1.4371,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7420628110519993,
+      "grad_norm": 3.7840593343484916,
+      "learning_rate": 3.29136334481551e-06,
+      "loss": 1.3687,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7422344259481723,
+      "grad_norm": 3.609273572627841,
+      "learning_rate": 3.287242398885685e-06,
+      "loss": 1.5283,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7424060408443452,
+      "grad_norm": 4.06308237770723,
+      "learning_rate": 3.2831235268968363e-06,
+      "loss": 1.4492,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7425776557405183,
+      "grad_norm": 4.048618922071043,
+      "learning_rate": 3.2790067301215165e-06,
+      "loss": 1.5934,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7427492706366913,
+      "grad_norm": 4.444850487302984,
+      "learning_rate": 3.274892009831624e-06,
+      "loss": 1.6145,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7429208855328643,
+      "grad_norm": 3.3165885122042553,
+      "learning_rate": 3.2707793672984265e-06,
+      "loss": 1.319,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7430925004290373,
+      "grad_norm": 3.427003278499009,
+      "learning_rate": 3.266668803792549e-06,
+      "loss": 1.3144,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7432641153252102,
+      "grad_norm": 4.09055139220498,
+      "learning_rate": 3.2625603205839652e-06,
+      "loss": 1.4437,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7434357302213832,
+      "grad_norm": 4.422786009591833,
+      "learning_rate": 3.258453918942016e-06,
+      "loss": 1.4695,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7436073451175562,
+      "grad_norm": 4.19317083241391,
+      "learning_rate": 3.2543496001353968e-06,
+      "loss": 1.6418,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7437789600137292,
+      "grad_norm": 4.360149015846294,
+      "learning_rate": 3.2502473654321532e-06,
+      "loss": 1.5356,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7439505749099021,
+      "grad_norm": 4.0686903831257135,
+      "learning_rate": 3.2461472160996965e-06,
+      "loss": 1.512,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7441221898060751,
+      "grad_norm": 4.835696832795532,
+      "learning_rate": 3.242049153404784e-06,
+      "loss": 1.521,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7442938047022482,
+      "grad_norm": 4.006478001591315,
+      "learning_rate": 3.2379531786135364e-06,
+      "loss": 1.7772,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7444654195984212,
+      "grad_norm": 3.630612051867948,
+      "learning_rate": 3.2338592929914247e-06,
+      "loss": 1.6493,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7446370344945942,
+      "grad_norm": 3.9455832403832547,
+      "learning_rate": 3.2297674978032813e-06,
+      "loss": 1.6707,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7448086493907671,
+      "grad_norm": 4.72664428829333,
+      "learning_rate": 3.225677794313281e-06,
+      "loss": 1.5191,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7449802642869401,
+      "grad_norm": 4.494412940502362,
+      "learning_rate": 3.2215901837849593e-06,
+      "loss": 1.5572,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7451518791831131,
+      "grad_norm": 3.7054500488679567,
+      "learning_rate": 3.217504667481205e-06,
+      "loss": 1.5167,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7453234940792861,
+      "grad_norm": 3.6406811183637937,
+      "learning_rate": 3.21342124666426e-06,
+      "loss": 1.5049,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7454951089754591,
+      "grad_norm": 3.407716422457096,
+      "learning_rate": 3.209339922595722e-06,
+      "loss": 1.3556,
+      "step": 4344
+    },
+    {
+      "epoch": 0.745666723871632,
+      "grad_norm": 3.329594279822077,
+      "learning_rate": 3.205260696536534e-06,
+      "loss": 1.4475,
+      "step": 4345
+    },
+    {
+      "epoch": 0.745838338767805,
+      "grad_norm": 3.499493027641549,
+      "learning_rate": 3.2011835697469894e-06,
+      "loss": 1.424,
+      "step": 4346
+    },
+    {
+      "epoch": 0.746009953663978,
+      "grad_norm": 4.147342968307836,
+      "learning_rate": 3.197108543486741e-06,
+      "loss": 1.5368,
+      "step": 4347
+    },
+    {
+      "epoch": 0.746181568560151,
+      "grad_norm": 5.174070905740056,
+      "learning_rate": 3.1930356190147904e-06,
+      "loss": 1.727,
+      "step": 4348
+    },
+    {
+      "epoch": 0.746353183456324,
+      "grad_norm": 4.350654957498894,
+      "learning_rate": 3.18896479758949e-06,
+      "loss": 1.6562,
+      "step": 4349
+    },
+    {
+      "epoch": 0.746524798352497,
+      "grad_norm": 4.216465648780869,
+      "learning_rate": 3.184896080468536e-06,
+      "loss": 1.3828,
+      "step": 4350
+    },
+    {
+      "epoch": 0.74669641324867,
+      "grad_norm": 4.131334440721682,
+      "learning_rate": 3.1808294689089856e-06,
+      "loss": 1.6477,
+      "step": 4351
+    },
+    {
+      "epoch": 0.746868028144843,
+      "grad_norm": 3.780379139401222,
+      "learning_rate": 3.176764964167233e-06,
+      "loss": 1.4671,
+      "step": 4352
+    },
+    {
+      "epoch": 0.747039643041016,
+      "grad_norm": 4.035466042971817,
+      "learning_rate": 3.1727025674990297e-06,
+      "loss": 1.6253,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7472112579371889,
+      "grad_norm": 4.523503512640381,
+      "learning_rate": 3.1686422801594786e-06,
+      "loss": 1.5261,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7473828728333619,
+      "grad_norm": 4.31883278979497,
+      "learning_rate": 3.16458410340302e-06,
+      "loss": 1.4649,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7475544877295349,
+      "grad_norm": 4.2683762988933855,
+      "learning_rate": 3.160528038483451e-06,
+      "loss": 1.6043,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7477261026257079,
+      "grad_norm": 3.840626024511075,
+      "learning_rate": 3.1564740866539166e-06,
+      "loss": 1.643,
+      "step": 4357
+    },
+    {
+      "epoch": 0.747897717521881,
+      "grad_norm": 4.781855580486221,
+      "learning_rate": 3.152422249166902e-06,
+      "loss": 1.6599,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7480693324180538,
+      "grad_norm": 3.7472182495196686,
+      "learning_rate": 3.148372527274246e-06,
+      "loss": 1.614,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7482409473142269,
+      "grad_norm": 4.200890232144251,
+      "learning_rate": 3.1443249222271278e-06,
+      "loss": 1.7006,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7484125622103999,
+      "grad_norm": 3.8148884685968665,
+      "learning_rate": 3.140279435276078e-06,
+      "loss": 1.5304,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7485841771065729,
+      "grad_norm": 3.821136573563433,
+      "learning_rate": 3.1362360676709693e-06,
+      "loss": 1.5777,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7487557920027459,
+      "grad_norm": 3.579578321326749,
+      "learning_rate": 3.1321948206610255e-06,
+      "loss": 1.8361,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7489274068989188,
+      "grad_norm": 3.715706697950207,
+      "learning_rate": 3.128155695494807e-06,
+      "loss": 1.3442,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7490990217950918,
+      "grad_norm": 4.190772879519168,
+      "learning_rate": 3.1241186934202196e-06,
+      "loss": 1.6312,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7492706366912648,
+      "grad_norm": 3.7183261313466676,
+      "learning_rate": 3.120083815684518e-06,
+      "loss": 1.5766,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7494422515874378,
+      "grad_norm": 4.355326002073972,
+      "learning_rate": 3.116051063534299e-06,
+      "loss": 1.4752,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7496138664836107,
+      "grad_norm": 3.8506539861445326,
+      "learning_rate": 3.112020438215506e-06,
+      "loss": 1.6699,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7497854813797837,
+      "grad_norm": 3.613591564230547,
+      "learning_rate": 3.107991940973417e-06,
+      "loss": 1.3734,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7499570962759567,
+      "grad_norm": 3.4679787563472653,
+      "learning_rate": 3.1039655730526543e-06,
+      "loss": 1.6845,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7501287111721298,
+      "grad_norm": 4.371918251736763,
+      "learning_rate": 3.0999413356971885e-06,
+      "loss": 1.6276,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7503003260683028,
+      "grad_norm": 4.106581842373159,
+      "learning_rate": 3.0959192301503284e-06,
+      "loss": 1.6236,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7504719409644757,
+      "grad_norm": 3.6726328270673836,
+      "learning_rate": 3.0918992576547257e-06,
+      "loss": 1.4985,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7506435558606487,
+      "grad_norm": 4.093395687833719,
+      "learning_rate": 3.087881419452368e-06,
+      "loss": 1.3704,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7508151707568217,
+      "grad_norm": 3.649386776645276,
+      "learning_rate": 3.083865716784592e-06,
+      "loss": 1.5872,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7509867856529947,
+      "grad_norm": 4.541331779282653,
+      "learning_rate": 3.079852150892063e-06,
+      "loss": 1.7192,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7511584005491677,
+      "grad_norm": 4.080763925371064,
+      "learning_rate": 3.075840723014796e-06,
+      "loss": 1.4221,
+      "step": 4377
+    },
+    {
+      "epoch": 0.7513300154453406,
+      "grad_norm": 3.6730592754379443,
+      "learning_rate": 3.0718314343921465e-06,
+      "loss": 1.6523,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7515016303415136,
+      "grad_norm": 4.31753251013167,
+      "learning_rate": 3.0678242862627974e-06,
+      "loss": 1.7085,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7516732452376866,
+      "grad_norm": 4.075099729353553,
+      "learning_rate": 3.0638192798647804e-06,
+      "loss": 1.5867,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7518448601338596,
+      "grad_norm": 3.7788045181733523,
+      "learning_rate": 3.0598164164354683e-06,
+      "loss": 1.632,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7520164750300327,
+      "grad_norm": 4.371404573389474,
+      "learning_rate": 3.055815697211557e-06,
+      "loss": 1.479,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7521880899262056,
+      "grad_norm": 3.969994890521598,
+      "learning_rate": 3.051817123429098e-06,
+      "loss": 1.5711,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7523597048223786,
+      "grad_norm": 3.9423259909841284,
+      "learning_rate": 3.0478206963234645e-06,
+      "loss": 1.4418,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7525313197185516,
+      "grad_norm": 5.099522267010024,
+      "learning_rate": 3.0438264171293762e-06,
+      "loss": 1.6361,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7527029346147246,
+      "grad_norm": 4.29007202872139,
+      "learning_rate": 3.039834287080885e-06,
+      "loss": 1.5624,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7528745495108975,
+      "grad_norm": 5.650833280005532,
+      "learning_rate": 3.035844307411384e-06,
+      "loss": 1.4914,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7530461644070705,
+      "grad_norm": 3.7929070758176024,
+      "learning_rate": 3.0318564793535964e-06,
+      "loss": 1.5812,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7532177793032435,
+      "grad_norm": 4.953814515321925,
+      "learning_rate": 3.0278708041395764e-06,
+      "loss": 1.5046,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7533893941994165,
+      "grad_norm": 3.9677165722369194,
+      "learning_rate": 3.0238872830007226e-06,
+      "loss": 1.4114,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7535610090955895,
+      "grad_norm": 4.47966330733451,
+      "learning_rate": 3.0199059171677657e-06,
+      "loss": 1.568,
+      "step": 4391
+    },
+    {
+      "epoch": 0.7537326239917624,
+      "grad_norm": 3.769070956683083,
+      "learning_rate": 3.015926707870768e-06,
+      "loss": 1.5562,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7539042388879355,
+      "grad_norm": 4.087450790596648,
+      "learning_rate": 3.011949656339134e-06,
+      "loss": 1.4098,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7540758537841085,
+      "grad_norm": 4.5315360723002005,
+      "learning_rate": 3.0079747638015798e-06,
+      "loss": 1.7902,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7542474686802815,
+      "grad_norm": 3.950329779940293,
+      "learning_rate": 3.0040020314861772e-06,
+      "loss": 1.433,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7544190835764545,
+      "grad_norm": 3.7349475438899633,
+      "learning_rate": 3.0000314606203206e-06,
+      "loss": 1.3761,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7545906984726274,
+      "grad_norm": 3.894904685386614,
+      "learning_rate": 2.996063052430739e-06,
+      "loss": 1.5998,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7547623133688004,
+      "grad_norm": 5.066690052255709,
+      "learning_rate": 2.9920968081434962e-06,
+      "loss": 1.7296,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7549339282649734,
+      "grad_norm": 3.8931630710000964,
+      "learning_rate": 2.98813272898398e-06,
+      "loss": 1.5806,
+      "step": 4399
+    },
+    {
+      "epoch": 0.7551055431611464,
+      "grad_norm": 3.1975501937477433,
+      "learning_rate": 2.98417081617691e-06,
+      "loss": 1.418,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7552771580573193,
+      "grad_norm": 3.885865392167385,
+      "learning_rate": 2.980211070946343e-06,
+      "loss": 1.3606,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7554487729534923,
+      "grad_norm": 4.114796699854625,
+      "learning_rate": 2.976253494515665e-06,
+      "loss": 1.557,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7556203878496653,
+      "grad_norm": 3.387230949391646,
+      "learning_rate": 2.9722980881075848e-06,
+      "loss": 1.5567,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7557920027458384,
+      "grad_norm": 3.871387707984601,
+      "learning_rate": 2.968344852944147e-06,
+      "loss": 1.4257,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7559636176420114,
+      "grad_norm": 4.763593299447595,
+      "learning_rate": 2.964393790246728e-06,
+      "loss": 1.5408,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7561352325381843,
+      "grad_norm": 3.8146427813491894,
+      "learning_rate": 2.9604449012360213e-06,
+      "loss": 1.3151,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7563068474343573,
+      "grad_norm": 3.7267863419570673,
+      "learning_rate": 2.9564981871320654e-06,
+      "loss": 1.5651,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7564784623305303,
+      "grad_norm": 4.1067316997557795,
+      "learning_rate": 2.9525536491542093e-06,
+      "loss": 1.4934,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7566500772267033,
+      "grad_norm": 3.9447501263304465,
+      "learning_rate": 2.948611288521143e-06,
+      "loss": 1.4834,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7568216921228763,
+      "grad_norm": 3.212018776639665,
+      "learning_rate": 2.944671106450877e-06,
+      "loss": 1.2799,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7569933070190492,
+      "grad_norm": 4.073104554948966,
+      "learning_rate": 2.9407331041607566e-06,
+      "loss": 1.3709,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7571649219152222,
+      "grad_norm": 4.748784333190554,
+      "learning_rate": 2.936797282867443e-06,
+      "loss": 1.6164,
+      "step": 4412
+    },
+    {
+      "epoch": 0.7573365368113952,
+      "grad_norm": 3.704864660949902,
+      "learning_rate": 2.932863643786925e-06,
+      "loss": 1.46,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7575081517075682,
+      "grad_norm": 4.048253645903471,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.57,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7576797666037413,
+      "grad_norm": 3.635895460198856,
+      "learning_rate": 2.925002917124886e-06,
+      "loss": 1.5115,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7578513814999142,
+      "grad_norm": 4.108851723343436,
+      "learning_rate": 2.9210758319719766e-06,
+      "loss": 1.4181,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7580229963960872,
+      "grad_norm": 5.543846090173724,
+      "learning_rate": 2.9171509338890924e-06,
+      "loss": 1.6657,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7581946112922602,
+      "grad_norm": 4.202833507114869,
+      "learning_rate": 2.913228224088849e-06,
+      "loss": 1.6016,
+      "step": 4418
+    },
+    {
+      "epoch": 0.7583662261884332,
+      "grad_norm": 3.299271885011321,
+      "learning_rate": 2.909307703783183e-06,
+      "loss": 1.3312,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7585378410846061,
+      "grad_norm": 3.801345031125703,
+      "learning_rate": 2.905389374183364e-06,
+      "loss": 1.6204,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7587094559807791,
+      "grad_norm": 3.6499202018479533,
+      "learning_rate": 2.9014732364999785e-06,
+      "loss": 1.4969,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7588810708769521,
+      "grad_norm": 3.926759161067827,
+      "learning_rate": 2.897559291942942e-06,
+      "loss": 1.4524,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7590526857731251,
+      "grad_norm": 3.546063765997647,
+      "learning_rate": 2.8936475417214795e-06,
+      "loss": 1.6617,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7592243006692981,
+      "grad_norm": 3.905646401381761,
+      "learning_rate": 2.8897379870441557e-06,
+      "loss": 1.3967,
+      "step": 4424
+    },
+    {
+      "epoch": 0.759395915565471,
+      "grad_norm": 3.425529299989414,
+      "learning_rate": 2.8858306291188387e-06,
+      "loss": 1.3189,
+      "step": 4425
+    },
+    {
+      "epoch": 0.759567530461644,
+      "grad_norm": 3.6579073263437913,
+      "learning_rate": 2.881925469152731e-06,
+      "loss": 1.5023,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7597391453578171,
+      "grad_norm": 4.092137941152871,
+      "learning_rate": 2.878022508352355e-06,
+      "loss": 1.7651,
+      "step": 4427
+    },
+    {
+      "epoch": 0.7599107602539901,
+      "grad_norm": 3.924362930139786,
+      "learning_rate": 2.874121747923544e-06,
+      "loss": 1.7015,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7600823751501631,
+      "grad_norm": 3.8899863715485843,
+      "learning_rate": 2.8702231890714606e-06,
+      "loss": 1.6186,
+      "step": 4429
+    },
+    {
+      "epoch": 0.760253990046336,
+      "grad_norm": 4.457015942288991,
+      "learning_rate": 2.866326833000589e-06,
+      "loss": 1.8369,
+      "step": 4430
+    },
+    {
+      "epoch": 0.760425604942509,
+      "grad_norm": 3.6213059778529795,
+      "learning_rate": 2.862432680914725e-06,
+      "loss": 1.4306,
+      "step": 4431
+    },
+    {
+      "epoch": 0.760597219838682,
+      "grad_norm": 3.2567857130058506,
+      "learning_rate": 2.8585407340169834e-06,
+      "loss": 1.5562,
+      "step": 4432
+    },
+    {
+      "epoch": 0.760768834734855,
+      "grad_norm": 3.9008085925184717,
+      "learning_rate": 2.8546509935098045e-06,
+      "loss": 1.6365,
+      "step": 4433
+    },
+    {
+      "epoch": 0.760940449631028,
+      "grad_norm": 3.839764394599854,
+      "learning_rate": 2.850763460594943e-06,
+      "loss": 1.4702,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7611120645272009,
+      "grad_norm": 3.9404823558506603,
+      "learning_rate": 2.846878136473472e-06,
+      "loss": 1.5432,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7612836794233739,
+      "grad_norm": 3.813111508049283,
+      "learning_rate": 2.8429950223457903e-06,
+      "loss": 1.4226,
+      "step": 4436
+    },
+    {
+      "epoch": 0.761455294319547,
+      "grad_norm": 4.784497204958719,
+      "learning_rate": 2.8391141194115933e-06,
+      "loss": 1.3172,
+      "step": 4437
+    },
+    {
+      "epoch": 0.76162690921572,
+      "grad_norm": 4.32480829849716,
+      "learning_rate": 2.83523542886991e-06,
+      "loss": 1.5136,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7617985241118929,
+      "grad_norm": 5.151200374053781,
+      "learning_rate": 2.831358951919084e-06,
+      "loss": 1.7607,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7619701390080659,
+      "grad_norm": 3.7860110227396246,
+      "learning_rate": 2.827484689756772e-06,
+      "loss": 1.4209,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7621417539042389,
+      "grad_norm": 5.223063833393038,
+      "learning_rate": 2.8236126435799492e-06,
+      "loss": 1.4223,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7623133688004119,
+      "grad_norm": 3.6918572937442473,
+      "learning_rate": 2.8197428145849037e-06,
+      "loss": 1.617,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7624849836965849,
+      "grad_norm": 5.373172443424373,
+      "learning_rate": 2.8158752039672332e-06,
+      "loss": 1.7911,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7626565985927578,
+      "grad_norm": 4.186438387440009,
+      "learning_rate": 2.812009812921861e-06,
+      "loss": 1.5638,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7628282134889308,
+      "grad_norm": 3.423061255454521,
+      "learning_rate": 2.808146642643018e-06,
+      "loss": 1.385,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7629998283851038,
+      "grad_norm": 4.380412032029947,
+      "learning_rate": 2.8042856943242548e-06,
+      "loss": 1.5834,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7631714432812768,
+      "grad_norm": 3.4097907038655815,
+      "learning_rate": 2.800426969158425e-06,
+      "loss": 1.5104,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7633430581774499,
+      "grad_norm": 4.465513251242325,
+      "learning_rate": 2.796570468337707e-06,
+      "loss": 1.7648,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7635146730736228,
+      "grad_norm": 3.474228806745961,
+      "learning_rate": 2.7927161930535805e-06,
+      "loss": 1.4709,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7636862879697958,
+      "grad_norm": 4.658185522178163,
+      "learning_rate": 2.788864144496848e-06,
+      "loss": 1.5629,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7638579028659688,
+      "grad_norm": 3.6260411182022976,
+      "learning_rate": 2.78501432385762e-06,
+      "loss": 1.5911,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7640295177621418,
+      "grad_norm": 3.835475896891879,
+      "learning_rate": 2.7811667323253155e-06,
+      "loss": 1.5582,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7642011326583148,
+      "grad_norm": 4.185508149101832,
+      "learning_rate": 2.7773213710886693e-06,
+      "loss": 1.3685,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7643727475544877,
+      "grad_norm": 3.7827348196473,
+      "learning_rate": 2.7734782413357274e-06,
+      "loss": 1.5666,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7645443624506607,
+      "grad_norm": 3.9627325602002093,
+      "learning_rate": 2.7696373442538406e-06,
+      "loss": 1.3867,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7647159773468337,
+      "grad_norm": 4.017457544356005,
+      "learning_rate": 2.765798681029679e-06,
+      "loss": 1.5754,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7648875922430067,
+      "grad_norm": 4.432502003093646,
+      "learning_rate": 2.7619622528492105e-06,
+      "loss": 1.6707,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7650592071391796,
+      "grad_norm": 4.14798269207826,
+      "learning_rate": 2.7581280608977233e-06,
+      "loss": 1.6452,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7652308220353526,
+      "grad_norm": 3.7581899694010876,
+      "learning_rate": 2.754296106359811e-06,
+      "loss": 1.4507,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7654024369315257,
+      "grad_norm": 3.8054250694000737,
+      "learning_rate": 2.750466390419376e-06,
+      "loss": 1.4753,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7655740518276987,
+      "grad_norm": 3.968374976960048,
+      "learning_rate": 2.7466389142596296e-06,
+      "loss": 1.7675,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7657456667238717,
+      "grad_norm": 3.6452133254946535,
+      "learning_rate": 2.7428136790630853e-06,
+      "loss": 1.3846,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7659172816200446,
+      "grad_norm": 4.323195481441205,
+      "learning_rate": 2.7389906860115724e-06,
+      "loss": 1.5019,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7660888965162176,
+      "grad_norm": 4.163743686034393,
+      "learning_rate": 2.7351699362862237e-06,
+      "loss": 1.4608,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7662605114123906,
+      "grad_norm": 3.268240556746148,
+      "learning_rate": 2.7313514310674826e-06,
+      "loss": 1.4187,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7664321263085636,
+      "grad_norm": 4.316649118879265,
+      "learning_rate": 2.727535171535094e-06,
+      "loss": 1.6098,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7666037412047366,
+      "grad_norm": 3.8992424533540464,
+      "learning_rate": 2.723721158868107e-06,
+      "loss": 1.6334,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7667753561009095,
+      "grad_norm": 6.2022856005080875,
+      "learning_rate": 2.719909394244884e-06,
+      "loss": 1.7317,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7669469709970825,
+      "grad_norm": 3.933953541630205,
+      "learning_rate": 2.7160998788430882e-06,
+      "loss": 1.3775,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7671185858932555,
+      "grad_norm": 3.7712211752873164,
+      "learning_rate": 2.712292613839692e-06,
+      "loss": 1.4645,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7672902007894286,
+      "grad_norm": 4.375401432276708,
+      "learning_rate": 2.7084876004109663e-06,
+      "loss": 1.4993,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7674618156856015,
+      "grad_norm": 4.273813274309977,
+      "learning_rate": 2.704684839732492e-06,
+      "loss": 1.399,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7676334305817745,
+      "grad_norm": 3.3765323403232625,
+      "learning_rate": 2.700884332979147e-06,
+      "loss": 1.4704,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7678050454779475,
+      "grad_norm": 3.674698038908049,
+      "learning_rate": 2.69708608132512e-06,
+      "loss": 1.4316,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7679766603741205,
+      "grad_norm": 3.8614394928487186,
+      "learning_rate": 2.6932900859439048e-06,
+      "loss": 1.4047,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7681482752702935,
+      "grad_norm": 3.2861669001575526,
+      "learning_rate": 2.689496348008286e-06,
+      "loss": 1.3076,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7683198901664664,
+      "grad_norm": 4.084108463315017,
+      "learning_rate": 2.6857048686903618e-06,
+      "loss": 1.5516,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7684915050626394,
+      "grad_norm": 3.568100865348519,
+      "learning_rate": 2.681915649161534e-06,
+      "loss": 1.2627,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7686631199588124,
+      "grad_norm": 3.4677180059123,
+      "learning_rate": 2.678128690592493e-06,
+      "loss": 1.391,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7688347348549854,
+      "grad_norm": 3.6824598718817203,
+      "learning_rate": 2.6743439941532457e-06,
+      "loss": 1.401,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7690063497511584,
+      "grad_norm": 3.8041152582511075,
+      "learning_rate": 2.6705615610130885e-06,
+      "loss": 1.488,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7691779646473313,
+      "grad_norm": 3.705435111785703,
+      "learning_rate": 2.666781392340626e-06,
+      "loss": 1.64,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7693495795435044,
+      "grad_norm": 4.24375975458298,
+      "learning_rate": 2.6630034893037617e-06,
+      "loss": 1.4309,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7695211944396774,
+      "grad_norm": 3.8504794019221187,
+      "learning_rate": 2.659227853069699e-06,
+      "loss": 1.4305,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7696928093358504,
+      "grad_norm": 3.75359595511843,
+      "learning_rate": 2.655454484804938e-06,
+      "loss": 1.5275,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7698644242320234,
+      "grad_norm": 3.914110370683981,
+      "learning_rate": 2.651683385675279e-06,
+      "loss": 1.6047,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7700360391281963,
+      "grad_norm": 3.840239834633645,
+      "learning_rate": 2.647914556845823e-06,
+      "loss": 1.7163,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7702076540243693,
+      "grad_norm": 4.857834316263262,
+      "learning_rate": 2.6441479994809703e-06,
+      "loss": 1.5096,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7703792689205423,
+      "grad_norm": 4.100397469726632,
+      "learning_rate": 2.640383714744421e-06,
+      "loss": 1.507,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7705508838167153,
+      "grad_norm": 4.548457883960667,
+      "learning_rate": 2.636621703799167e-06,
+      "loss": 1.7107,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7707224987128882,
+      "grad_norm": 3.67459602559332,
+      "learning_rate": 2.6328619678074986e-06,
+      "loss": 1.5633,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7708941136090612,
+      "grad_norm": 4.101174492648497,
+      "learning_rate": 2.6291045079310085e-06,
+      "loss": 1.3496,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7710657285052342,
+      "grad_norm": 4.056990905886543,
+      "learning_rate": 2.625349325330582e-06,
+      "loss": 1.5651,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7712373434014073,
+      "grad_norm": 3.733589344347334,
+      "learning_rate": 2.621596421166407e-06,
+      "loss": 1.5497,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7714089582975803,
+      "grad_norm": 3.783740714792206,
+      "learning_rate": 2.6178457965979543e-06,
+      "loss": 1.5104,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7715805731937532,
+      "grad_norm": 3.7677226534491903,
+      "learning_rate": 2.614097452784006e-06,
+      "loss": 1.3819,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7717521880899262,
+      "grad_norm": 4.423261448478865,
+      "learning_rate": 2.610351390882626e-06,
+      "loss": 1.4501,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7719238029860992,
+      "grad_norm": 3.9400255608160673,
+      "learning_rate": 2.606607612051183e-06,
+      "loss": 1.6644,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7720954178822722,
+      "grad_norm": 4.053939427153468,
+      "learning_rate": 2.602866117446339e-06,
+      "loss": 1.6112,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7722670327784452,
+      "grad_norm": 5.174646121012897,
+      "learning_rate": 2.599126908224041e-06,
+      "loss": 1.4842,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7724386476746181,
+      "grad_norm": 3.352561240146486,
+      "learning_rate": 2.5953899855395428e-06,
+      "loss": 1.296,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7726102625707911,
+      "grad_norm": 4.188214515774507,
+      "learning_rate": 2.591655350547385e-06,
+      "loss": 1.792,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7727818774669641,
+      "grad_norm": 4.6857347627633175,
+      "learning_rate": 2.5879230044014003e-06,
+      "loss": 1.4828,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7729534923631372,
+      "grad_norm": 3.8344725531467905,
+      "learning_rate": 2.584192948254719e-06,
+      "loss": 1.3325,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7731251072593102,
+      "grad_norm": 3.783532423223033,
+      "learning_rate": 2.580465183259757e-06,
+      "loss": 1.5223,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7732967221554831,
+      "grad_norm": 3.98052990658642,
+      "learning_rate": 2.576739710568229e-06,
+      "loss": 1.5266,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7734683370516561,
+      "grad_norm": 3.8682766203140537,
+      "learning_rate": 2.5730165313311396e-06,
+      "loss": 1.5732,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7736399519478291,
+      "grad_norm": 4.123862917529191,
+      "learning_rate": 2.569295646698785e-06,
+      "loss": 1.591,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7738115668440021,
+      "grad_norm": 4.098189577086668,
+      "learning_rate": 2.5655770578207505e-06,
+      "loss": 1.557,
+      "step": 4509
+    },
+    {
+      "epoch": 0.773983181740175,
+      "grad_norm": 6.481973805275883,
+      "learning_rate": 2.5618607658459105e-06,
+      "loss": 1.3156,
+      "step": 4510
+    },
+    {
+      "epoch": 0.774154796636348,
+      "grad_norm": 3.9290315116549888,
+      "learning_rate": 2.558146771922434e-06,
+      "loss": 1.704,
+      "step": 4511
+    },
+    {
+      "epoch": 0.774326411532521,
+      "grad_norm": 4.528762313524045,
+      "learning_rate": 2.55443507719778e-06,
+      "loss": 1.4495,
+      "step": 4512
+    },
+    {
+      "epoch": 0.774498026428694,
+      "grad_norm": 4.286421310674162,
+      "learning_rate": 2.5507256828186978e-06,
+      "loss": 1.5507,
+      "step": 4513
+    },
+    {
+      "epoch": 0.774669641324867,
+      "grad_norm": 3.7519312087952454,
+      "learning_rate": 2.5470185899312207e-06,
+      "loss": 1.4529,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7748412562210399,
+      "grad_norm": 4.241760455792032,
+      "learning_rate": 2.5433137996806723e-06,
+      "loss": 1.4337,
+      "step": 4515
+    },
+    {
+      "epoch": 0.775012871117213,
+      "grad_norm": 3.7873020720678174,
+      "learning_rate": 2.5396113132116685e-06,
+      "loss": 1.2707,
+      "step": 4516
+    },
+    {
+      "epoch": 0.775184486013386,
+      "grad_norm": 4.1714384224692616,
+      "learning_rate": 2.5359111316681116e-06,
+      "loss": 1.5916,
+      "step": 4517
+    },
+    {
+      "epoch": 0.775356100909559,
+      "grad_norm": 3.652784792387227,
+      "learning_rate": 2.5322132561931946e-06,
+      "loss": 1.5952,
+      "step": 4518
+    },
+    {
+      "epoch": 0.775527715805732,
+      "grad_norm": 4.953929865625447,
+      "learning_rate": 2.5285176879293903e-06,
+      "loss": 1.5933,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7756993307019049,
+      "grad_norm": 5.110724693408642,
+      "learning_rate": 2.5248244280184688e-06,
+      "loss": 1.5387,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7758709455980779,
+      "grad_norm": 4.646757485427627,
+      "learning_rate": 2.5211334776014752e-06,
+      "loss": 1.6548,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7760425604942509,
+      "grad_norm": 3.950822310675609,
+      "learning_rate": 2.517444837818751e-06,
+      "loss": 1.6421,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7762141753904239,
+      "grad_norm": 3.664188226279325,
+      "learning_rate": 2.5137585098099237e-06,
+      "loss": 1.5203,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7763857902865968,
+      "grad_norm": 3.7627221089697738,
+      "learning_rate": 2.5100744947138967e-06,
+      "loss": 1.4911,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7765574051827698,
+      "grad_norm": 3.7254328235045406,
+      "learning_rate": 2.506392793668869e-06,
+      "loss": 1.4806,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7767290200789428,
+      "grad_norm": 4.350941848753395,
+      "learning_rate": 2.502713407812324e-06,
+      "loss": 1.4898,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7769006349751159,
+      "grad_norm": 3.869375522392363,
+      "learning_rate": 2.4990363382810214e-06,
+      "loss": 1.3624,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7770722498712889,
+      "grad_norm": 4.128531803296474,
+      "learning_rate": 2.4953615862110146e-06,
+      "loss": 1.5084,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7772438647674618,
+      "grad_norm": 3.3730510984713513,
+      "learning_rate": 2.4916891527376342e-06,
+      "loss": 1.5021,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7774154796636348,
+      "grad_norm": 3.501713048941487,
+      "learning_rate": 2.488019038995497e-06,
+      "loss": 1.5368,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7775870945598078,
+      "grad_norm": 4.037353076532261,
+      "learning_rate": 2.484351246118507e-06,
+      "loss": 1.5601,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7777587094559808,
+      "grad_norm": 8.097110207268901,
+      "learning_rate": 2.4806857752398484e-06,
+      "loss": 1.5914,
+      "step": 4532
+    },
+    {
+      "epoch": 0.7779303243521538,
+      "grad_norm": 3.536299649948467,
+      "learning_rate": 2.477022627491985e-06,
+      "loss": 1.4101,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7781019392483267,
+      "grad_norm": 3.2000364500466705,
+      "learning_rate": 2.473361804006662e-06,
+      "loss": 1.3659,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7782735541444997,
+      "grad_norm": 5.089967887166785,
+      "learning_rate": 2.469703305914913e-06,
+      "loss": 1.6735,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7784451690406727,
+      "grad_norm": 4.007247000947681,
+      "learning_rate": 2.466047134347049e-06,
+      "loss": 1.6003,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7786167839368457,
+      "grad_norm": 4.204910629456333,
+      "learning_rate": 2.4623932904326665e-06,
+      "loss": 1.3892,
+      "step": 4537
+    },
+    {
+      "epoch": 0.7787883988330188,
+      "grad_norm": 3.987292947688818,
+      "learning_rate": 2.458741775300637e-06,
+      "loss": 1.433,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7789600137291917,
+      "grad_norm": 4.360384539749673,
+      "learning_rate": 2.45509259007911e-06,
+      "loss": 1.406,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7791316286253647,
+      "grad_norm": 4.522888518202757,
+      "learning_rate": 2.4514457358955255e-06,
+      "loss": 1.4886,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7793032435215377,
+      "grad_norm": 4.020788045386832,
+      "learning_rate": 2.4478012138765962e-06,
+      "loss": 1.3645,
+      "step": 4541
+    },
+    {
+      "epoch": 0.7794748584177107,
+      "grad_norm": 3.5694199406070832,
+      "learning_rate": 2.4441590251483194e-06,
+      "loss": 1.3175,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7796464733138836,
+      "grad_norm": 4.019233931035732,
+      "learning_rate": 2.4405191708359623e-06,
+      "loss": 1.4375,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7798180882100566,
+      "grad_norm": 4.091775692677808,
+      "learning_rate": 2.4368816520640813e-06,
+      "loss": 1.4375,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7799897031062296,
+      "grad_norm": 4.270409007492003,
+      "learning_rate": 2.4332464699565015e-06,
+      "loss": 1.6578,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7801613180024026,
+      "grad_norm": 4.134044993120238,
+      "learning_rate": 2.4296136256363345e-06,
+      "loss": 1.6776,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7803329328985756,
+      "grad_norm": 3.975188865992825,
+      "learning_rate": 2.4259831202259697e-06,
+      "loss": 1.5708,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7805045477947485,
+      "grad_norm": 3.8247261437705933,
+      "learning_rate": 2.4223549548470628e-06,
+      "loss": 1.5583,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7806761626909215,
+      "grad_norm": 3.9786166538535,
+      "learning_rate": 2.4187291306205573e-06,
+      "loss": 1.3402,
+      "step": 4549
+    },
+    {
+      "epoch": 0.7808477775870946,
+      "grad_norm": 3.844275444933562,
+      "learning_rate": 2.4151056486666748e-06,
+      "loss": 1.6355,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7810193924832676,
+      "grad_norm": 4.273336345579903,
+      "learning_rate": 2.4114845101049013e-06,
+      "loss": 1.8647,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7811910073794406,
+      "grad_norm": 4.087677119858582,
+      "learning_rate": 2.407865716054013e-06,
+      "loss": 1.4818,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7813626222756135,
+      "grad_norm": 4.182573259912535,
+      "learning_rate": 2.4042492676320482e-06,
+      "loss": 1.3974,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7815342371717865,
+      "grad_norm": 4.125214870502783,
+      "learning_rate": 2.400635165956331e-06,
+      "loss": 1.6131,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7817058520679595,
+      "grad_norm": 3.661800085485567,
+      "learning_rate": 2.3970234121434555e-06,
+      "loss": 1.5482,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7818774669641325,
+      "grad_norm": 4.145476093215818,
+      "learning_rate": 2.393414007309296e-06,
+      "loss": 1.7584,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7820490818603055,
+      "grad_norm": 3.8974989480822524,
+      "learning_rate": 2.3898069525689925e-06,
+      "loss": 1.5632,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7822206967564784,
+      "grad_norm": 4.605563261774743,
+      "learning_rate": 2.386202249036961e-06,
+      "loss": 1.4757,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7823923116526514,
+      "grad_norm": 4.066695993715914,
+      "learning_rate": 2.3825998978268962e-06,
+      "loss": 1.6209,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7825639265488245,
+      "grad_norm": 4.514725466414961,
+      "learning_rate": 2.3789999000517627e-06,
+      "loss": 1.5252,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7827355414449975,
+      "grad_norm": 3.8552784172237367,
+      "learning_rate": 2.3754022568238e-06,
+      "loss": 1.4558,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7829071563411704,
+      "grad_norm": 3.669796767190878,
+      "learning_rate": 2.3718069692545187e-06,
+      "loss": 1.6467,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7830787712373434,
+      "grad_norm": 3.9705014700658086,
+      "learning_rate": 2.368214038454695e-06,
+      "loss": 1.4688,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7832503861335164,
+      "grad_norm": 4.039811873462365,
+      "learning_rate": 2.3646234655343894e-06,
+      "loss": 1.4324,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7834220010296894,
+      "grad_norm": 4.046197720990133,
+      "learning_rate": 2.3610352516029244e-06,
+      "loss": 1.417,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7835936159258624,
+      "grad_norm": 4.414451727332304,
+      "learning_rate": 2.3574493977689026e-06,
+      "loss": 1.391,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7837652308220353,
+      "grad_norm": 3.5724871672587692,
+      "learning_rate": 2.353865905140187e-06,
+      "loss": 1.2887,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7839368457182083,
+      "grad_norm": 4.118855242545041,
+      "learning_rate": 2.350284774823919e-06,
+      "loss": 1.3697,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7841084606143813,
+      "grad_norm": 5.06417793779705,
+      "learning_rate": 2.3467060079265026e-06,
+      "loss": 1.5435,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7842800755105543,
+      "grad_norm": 3.6181555666804472,
+      "learning_rate": 2.34312960555362e-06,
+      "loss": 1.3425,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7844516904067274,
+      "grad_norm": 4.451075665002024,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 1.3767,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7846233053029003,
+      "grad_norm": 4.134181542467542,
+      "learning_rate": 2.3359838988005177e-06,
+      "loss": 1.3222,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7847949201990733,
+      "grad_norm": 4.073668045984023,
+      "learning_rate": 2.3324145966279978e-06,
+      "loss": 1.519,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7849665350952463,
+      "grad_norm": 4.530114921473998,
+      "learning_rate": 2.3288476633954194e-06,
+      "loss": 1.7735,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7851381499914193,
+      "grad_norm": 4.121616005573902,
+      "learning_rate": 2.3252831002047983e-06,
+      "loss": 1.4971,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7853097648875922,
+      "grad_norm": 4.323196090661877,
+      "learning_rate": 2.321720908157431e-06,
+      "loss": 1.5966,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7854813797837652,
+      "grad_norm": 3.6960822152095045,
+      "learning_rate": 2.3181610883538687e-06,
+      "loss": 1.5503,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7856529946799382,
+      "grad_norm": 4.422497123418933,
+      "learning_rate": 2.314603641893939e-06,
+      "loss": 1.2902,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7858246095761112,
+      "grad_norm": 4.000794942195877,
+      "learning_rate": 2.3110485698767336e-06,
+      "loss": 1.4545,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7859962244722842,
+      "grad_norm": 3.714867332806828,
+      "learning_rate": 2.3074958734006136e-06,
+      "loss": 1.5196,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7861678393684571,
+      "grad_norm": 4.122749732512584,
+      "learning_rate": 2.3039455535631995e-06,
+      "loss": 1.6527,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7863394542646301,
+      "grad_norm": 3.5470760552100566,
+      "learning_rate": 2.300397611461378e-06,
+      "loss": 1.3618,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7865110691608032,
+      "grad_norm": 3.660475328826211,
+      "learning_rate": 2.296852048191306e-06,
+      "loss": 1.548,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7866826840569762,
+      "grad_norm": 4.379460706114087,
+      "learning_rate": 2.293308864848405e-06,
+      "loss": 1.3725,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7868542989531492,
+      "grad_norm": 4.110023104077812,
+      "learning_rate": 2.2897680625273623e-06,
+      "loss": 1.4684,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7870259138493221,
+      "grad_norm": 3.8559896212443614,
+      "learning_rate": 2.2862296423221232e-06,
+      "loss": 1.3079,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7871975287454951,
+      "grad_norm": 4.456887935980542,
+      "learning_rate": 2.2826936053258986e-06,
+      "loss": 1.5977,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7873691436416681,
+      "grad_norm": 4.3797378698923595,
+      "learning_rate": 2.279159952631168e-06,
+      "loss": 1.6759,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7875407585378411,
+      "grad_norm": 4.895653407277305,
+      "learning_rate": 2.2756286853296716e-06,
+      "loss": 1.5717,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7877123734340141,
+      "grad_norm": 4.167115803492636,
+      "learning_rate": 2.2720998045124155e-06,
+      "loss": 1.447,
+      "step": 4590
+    },
+    {
+      "epoch": 0.787883988330187,
+      "grad_norm": 4.174668458199423,
+      "learning_rate": 2.2685733112696606e-06,
+      "loss": 1.5835,
+      "step": 4591
+    },
+    {
+      "epoch": 0.78805560322636,
+      "grad_norm": 4.9347806304797714,
+      "learning_rate": 2.265049206690939e-06,
+      "loss": 1.5882,
+      "step": 4592
+    },
+    {
+      "epoch": 0.788227218122533,
+      "grad_norm": 4.545062690187829,
+      "learning_rate": 2.2615274918650376e-06,
+      "loss": 1.57,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7883988330187061,
+      "grad_norm": 4.7718352818444325,
+      "learning_rate": 2.258008167880009e-06,
+      "loss": 1.3868,
+      "step": 4594
+    },
+    {
+      "epoch": 0.788570447914879,
+      "grad_norm": 4.144523611947049,
+      "learning_rate": 2.2544912358231706e-06,
+      "loss": 1.3997,
+      "step": 4595
+    },
+    {
+      "epoch": 0.788742062811052,
+      "grad_norm": 4.534395590692036,
+      "learning_rate": 2.2509766967810896e-06,
+      "loss": 1.6358,
+      "step": 4596
+    },
+    {
+      "epoch": 0.788913677707225,
+      "grad_norm": 3.398494400720974,
+      "learning_rate": 2.2474645518396065e-06,
+      "loss": 1.5059,
+      "step": 4597
+    },
+    {
+      "epoch": 0.789085292603398,
+      "grad_norm": 4.805576840139967,
+      "learning_rate": 2.2439548020838166e-06,
+      "loss": 1.7907,
+      "step": 4598
+    },
+    {
+      "epoch": 0.789256907499571,
+      "grad_norm": 3.950006902277852,
+      "learning_rate": 2.24044744859807e-06,
+      "loss": 1.5026,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7894285223957439,
+      "grad_norm": 5.691220905255036,
+      "learning_rate": 2.2369424924659877e-06,
+      "loss": 1.5733,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7896001372919169,
+      "grad_norm": 4.143064291643852,
+      "learning_rate": 2.233439934770438e-06,
+      "loss": 1.4565,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7897717521880899,
+      "grad_norm": 4.1962305872311605,
+      "learning_rate": 2.2299397765935582e-06,
+      "loss": 1.6948,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7899433670842629,
+      "grad_norm": 4.136587175410183,
+      "learning_rate": 2.226442019016739e-06,
+      "loss": 1.5132,
+      "step": 4603
+    },
+    {
+      "epoch": 0.790114981980436,
+      "grad_norm": 3.676086791613163,
+      "learning_rate": 2.222946663120633e-06,
+      "loss": 1.4024,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7902865968766088,
+      "grad_norm": 3.744341637558946,
+      "learning_rate": 2.2194537099851465e-06,
+      "loss": 1.4473,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7904582117727819,
+      "grad_norm": 4.4355070229087605,
+      "learning_rate": 2.2159631606894427e-06,
+      "loss": 1.4113,
+      "step": 4606
+    },
+    {
+      "epoch": 0.7906298266689549,
+      "grad_norm": 3.5095623292619353,
+      "learning_rate": 2.2124750163119457e-06,
+      "loss": 1.4368,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7908014415651279,
+      "grad_norm": 3.7407516024751404,
+      "learning_rate": 2.208989277930338e-06,
+      "loss": 1.4771,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7909730564613009,
+      "grad_norm": 4.007317668639387,
+      "learning_rate": 2.205505946621558e-06,
+      "loss": 1.5252,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7911446713574738,
+      "grad_norm": 3.7571818491464426,
+      "learning_rate": 2.2020250234617957e-06,
+      "loss": 1.5841,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7913162862536468,
+      "grad_norm": 4.612550439977165,
+      "learning_rate": 2.198546509526498e-06,
+      "loss": 1.5319,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7914879011498198,
+      "grad_norm": 4.005278833794611,
+      "learning_rate": 2.1950704058903727e-06,
+      "loss": 1.5433,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7916595160459928,
+      "grad_norm": 3.8769935412426224,
+      "learning_rate": 2.1915967136273797e-06,
+      "loss": 1.5392,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7918311309421657,
+      "grad_norm": 4.090312146460333,
+      "learning_rate": 2.188125433810736e-06,
+      "loss": 1.5156,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7920027458383387,
+      "grad_norm": 3.5700381420604255,
+      "learning_rate": 2.1846565675129074e-06,
+      "loss": 1.2586,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7921743607345118,
+      "grad_norm": 3.7294634405637885,
+      "learning_rate": 2.1811901158056213e-06,
+      "loss": 1.3121,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7923459756306848,
+      "grad_norm": 3.2927047845992194,
+      "learning_rate": 2.1777260797598523e-06,
+      "loss": 1.5085,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7925175905268578,
+      "grad_norm": 3.6449002749128354,
+      "learning_rate": 2.174264460445834e-06,
+      "loss": 1.385,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7926892054230307,
+      "grad_norm": 3.7236303330209855,
+      "learning_rate": 2.1708052589330553e-06,
+      "loss": 1.6128,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7928608203192037,
+      "grad_norm": 4.125968023663776,
+      "learning_rate": 2.1673484762902473e-06,
+      "loss": 1.6141,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7930324352153767,
+      "grad_norm": 3.517979458271183,
+      "learning_rate": 2.1638941135854042e-06,
+      "loss": 1.259,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7932040501115497,
+      "grad_norm": 4.18192251590275,
+      "learning_rate": 2.1604421718857714e-06,
+      "loss": 1.4995,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7933756650077227,
+      "grad_norm": 3.799625812860469,
+      "learning_rate": 2.156992652257839e-06,
+      "loss": 1.6501,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7935472799038956,
+      "grad_norm": 4.184488734604103,
+      "learning_rate": 2.153545555767359e-06,
+      "loss": 1.5952,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7937188948000686,
+      "grad_norm": 3.724464336165782,
+      "learning_rate": 2.1501008834793257e-06,
+      "loss": 1.4416,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7938905096962416,
+      "grad_norm": 3.9124088421508563,
+      "learning_rate": 2.146658636457989e-06,
+      "loss": 1.498,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7940621245924147,
+      "grad_norm": 4.035782081627623,
+      "learning_rate": 2.143218815766849e-06,
+      "loss": 1.3016,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7942337394885877,
+      "grad_norm": 4.47954400839062,
+      "learning_rate": 2.13978142246866e-06,
+      "loss": 1.4967,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7944053543847606,
+      "grad_norm": 3.803009861546744,
+      "learning_rate": 2.136346457625418e-06,
+      "loss": 1.613,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7945769692809336,
+      "grad_norm": 4.004745018724956,
+      "learning_rate": 2.132913922298372e-06,
+      "loss": 1.5802,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7947485841771066,
+      "grad_norm": 6.184376835111989,
+      "learning_rate": 2.1294838175480237e-06,
+      "loss": 1.6301,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7949201990732796,
+      "grad_norm": 4.585837532680891,
+      "learning_rate": 2.12605614443412e-06,
+      "loss": 1.489,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7950918139694525,
+      "grad_norm": 3.9897481821223018,
+      "learning_rate": 2.1226309040156634e-06,
+      "loss": 1.5182,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7952634288656255,
+      "grad_norm": 4.662142462027689,
+      "learning_rate": 2.1192080973508944e-06,
+      "loss": 1.6099,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7954350437617985,
+      "grad_norm": 3.9027942977554955,
+      "learning_rate": 2.1157877254973057e-06,
+      "loss": 1.4579,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7956066586579715,
+      "grad_norm": 3.500944414217306,
+      "learning_rate": 2.1123697895116412e-06,
+      "loss": 1.4072,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7957782735541445,
+      "grad_norm": 3.9063468670755426,
+      "learning_rate": 2.1089542904498893e-06,
+      "loss": 1.5136,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7959498884503174,
+      "grad_norm": 3.51807468341857,
+      "learning_rate": 2.1055412293672883e-06,
+      "loss": 1.3795,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7961215033464905,
+      "grad_norm": 3.7092350093674487,
+      "learning_rate": 2.1021306073183166e-06,
+      "loss": 1.6764,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7962931182426635,
+      "grad_norm": 4.355211550983136,
+      "learning_rate": 2.0987224253567096e-06,
+      "loss": 1.5256,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7964647331388365,
+      "grad_norm": 4.084868416364256,
+      "learning_rate": 2.0953166845354346e-06,
+      "loss": 1.6207,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7966363480350095,
+      "grad_norm": 3.4407769552269603,
+      "learning_rate": 2.0919133859067174e-06,
+      "loss": 1.449,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7968079629311824,
+      "grad_norm": 4.834590952221245,
+      "learning_rate": 2.0885125305220267e-06,
+      "loss": 1.5994,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7969795778273554,
+      "grad_norm": 3.5615286063482854,
+      "learning_rate": 2.0851141194320688e-06,
+      "loss": 1.4866,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7971511927235284,
+      "grad_norm": 4.019456861440628,
+      "learning_rate": 2.0817181536868035e-06,
+      "loss": 1.3091,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7973228076197014,
+      "grad_norm": 3.59124948231215,
+      "learning_rate": 2.0783246343354334e-06,
+      "loss": 1.49,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7974944225158743,
+      "grad_norm": 6.121992832398803,
+      "learning_rate": 2.0749335624264e-06,
+      "loss": 1.5963,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7976660374120473,
+      "grad_norm": 4.5099743319393415,
+      "learning_rate": 2.071544939007397e-06,
+      "loss": 1.5077,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7978376523082203,
+      "grad_norm": 3.673607780839353,
+      "learning_rate": 2.068158765125352e-06,
+      "loss": 1.3995,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7980092672043934,
+      "grad_norm": 4.133335034633506,
+      "learning_rate": 2.064775041826442e-06,
+      "loss": 1.4355,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7981808821005664,
+      "grad_norm": 3.9052805724034685,
+      "learning_rate": 2.061393770156088e-06,
+      "loss": 1.7653,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7983524969967393,
+      "grad_norm": 4.322315352817901,
+      "learning_rate": 2.058014951158952e-06,
+      "loss": 1.6881,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7985241118929123,
+      "grad_norm": 3.0129296469013007,
+      "learning_rate": 2.0546385858789363e-06,
+      "loss": 1.1219,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7986957267890853,
+      "grad_norm": 5.431230361859624,
+      "learning_rate": 2.0512646753591827e-06,
+      "loss": 1.6602,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7988673416852583,
+      "grad_norm": 3.5238135179744012,
+      "learning_rate": 2.047893220642081e-06,
+      "loss": 1.4516,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7990389565814313,
+      "grad_norm": 4.051884364875317,
+      "learning_rate": 2.0445242227692597e-06,
+      "loss": 1.4125,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7992105714776042,
+      "grad_norm": 4.0383029252118385,
+      "learning_rate": 2.0411576827815904e-06,
+      "loss": 1.7096,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7993821863737772,
+      "grad_norm": 4.187493705477895,
+      "learning_rate": 2.0377936017191803e-06,
+      "loss": 1.492,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7995538012699502,
+      "grad_norm": 4.456936084030553,
+      "learning_rate": 2.0344319806213774e-06,
+      "loss": 1.5869,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7997254161661232,
+      "grad_norm": 4.246077510198194,
+      "learning_rate": 2.031072820526775e-06,
+      "loss": 1.3872,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7998970310622963,
+      "grad_norm": 3.4948357368580245,
+      "learning_rate": 2.0277161224732012e-06,
+      "loss": 1.4101,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8000686459584692,
+      "grad_norm": 4.2907802548054335,
+      "learning_rate": 2.024361887497728e-06,
+      "loss": 1.4822,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8002402608546422,
+      "grad_norm": 4.0944230094276275,
+      "learning_rate": 2.02101011663666e-06,
+      "loss": 1.5002,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8004118757508152,
+      "grad_norm": 3.90006943843002,
+      "learning_rate": 2.017660810925548e-06,
+      "loss": 1.6627,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8005834906469882,
+      "grad_norm": 4.357170239633029,
+      "learning_rate": 2.014313971399171e-06,
+      "loss": 1.7669,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8007551055431611,
+      "grad_norm": 3.982979722074304,
+      "learning_rate": 2.010969599091557e-06,
+      "loss": 1.4318,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8009267204393341,
+      "grad_norm": 3.768009542140195,
+      "learning_rate": 2.0076276950359687e-06,
+      "loss": 1.5569,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8010983353355071,
+      "grad_norm": 4.126049283758045,
+      "learning_rate": 2.0042882602648996e-06,
+      "loss": 1.6959,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8012699502316801,
+      "grad_norm": 4.634676631020037,
+      "learning_rate": 2.0009512958100862e-06,
+      "loss": 1.7168,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8014415651278531,
+      "grad_norm": 3.6905406160230276,
+      "learning_rate": 1.9976168027025066e-06,
+      "loss": 1.4887,
+      "step": 4670
+    },
+    {
+      "epoch": 0.801613180024026,
+      "grad_norm": 4.991378128678878,
+      "learning_rate": 1.994284781972361e-06,
+      "loss": 1.6904,
+      "step": 4671
+    },
+    {
+      "epoch": 0.801784794920199,
+      "grad_norm": 3.8939800226272063,
+      "learning_rate": 1.990955234649102e-06,
+      "loss": 1.4813,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8019564098163721,
+      "grad_norm": 4.640334195680962,
+      "learning_rate": 1.9876281617614045e-06,
+      "loss": 1.536,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8021280247125451,
+      "grad_norm": 5.02677592844051,
+      "learning_rate": 1.9843035643371865e-06,
+      "loss": 1.5338,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8022996396087181,
+      "grad_norm": 3.88387810281538,
+      "learning_rate": 1.9809814434036e-06,
+      "loss": 1.4312,
+      "step": 4675
+    },
+    {
+      "epoch": 0.802471254504891,
+      "grad_norm": 5.085624113110938,
+      "learning_rate": 1.977661799987035e-06,
+      "loss": 1.5967,
+      "step": 4676
+    },
+    {
+      "epoch": 0.802642869401064,
+      "grad_norm": 4.10904531677117,
+      "learning_rate": 1.974344635113108e-06,
+      "loss": 1.6195,
+      "step": 4677
+    },
+    {
+      "epoch": 0.802814484297237,
+      "grad_norm": 4.139654311513187,
+      "learning_rate": 1.971029949806673e-06,
+      "loss": 1.4278,
+      "step": 4678
+    },
+    {
+      "epoch": 0.80298609919341,
+      "grad_norm": 4.133107737175155,
+      "learning_rate": 1.9677177450918206e-06,
+      "loss": 1.5947,
+      "step": 4679
+    },
+    {
+      "epoch": 0.803157714089583,
+      "grad_norm": 4.337713520500158,
+      "learning_rate": 1.964408021991874e-06,
+      "loss": 1.6304,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8033293289857559,
+      "grad_norm": 4.482565791909271,
+      "learning_rate": 1.9611007815293926e-06,
+      "loss": 1.4749,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8035009438819289,
+      "grad_norm": 5.0380616373243985,
+      "learning_rate": 1.9577960247261606e-06,
+      "loss": 1.6686,
+      "step": 4682
+    },
+    {
+      "epoch": 0.803672558778102,
+      "grad_norm": 3.8579526563832633,
+      "learning_rate": 1.954493752603199e-06,
+      "loss": 1.491,
+      "step": 4683
+    },
+    {
+      "epoch": 0.803844173674275,
+      "grad_norm": 4.5603977193272875,
+      "learning_rate": 1.951193966180763e-06,
+      "loss": 1.614,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8040157885704479,
+      "grad_norm": 4.913323652369876,
+      "learning_rate": 1.9478966664783407e-06,
+      "loss": 1.486,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8041874034666209,
+      "grad_norm": 4.069079882651872,
+      "learning_rate": 1.9446018545146493e-06,
+      "loss": 1.5907,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8043590183627939,
+      "grad_norm": 3.7424499704879,
+      "learning_rate": 1.941309531307636e-06,
+      "loss": 1.4458,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8045306332589669,
+      "grad_norm": 4.598478927516427,
+      "learning_rate": 1.9380196978744824e-06,
+      "loss": 1.7173,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8047022481551399,
+      "grad_norm": 3.6103840107552085,
+      "learning_rate": 1.934732355231598e-06,
+      "loss": 1.4013,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8048738630513128,
+      "grad_norm": 4.34924179381626,
+      "learning_rate": 1.9314475043946244e-06,
+      "loss": 1.5084,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8050454779474858,
+      "grad_norm": 4.15588545836338,
+      "learning_rate": 1.928165146378437e-06,
+      "loss": 1.5732,
+      "step": 4691
+    },
+    {
+      "epoch": 0.8052170928436588,
+      "grad_norm": 3.692125388138388,
+      "learning_rate": 1.924885282197132e-06,
+      "loss": 1.3896,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8053887077398318,
+      "grad_norm": 4.074850141367472,
+      "learning_rate": 1.921607912864042e-06,
+      "loss": 1.7314,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8055603226360049,
+      "grad_norm": 3.229449380375219,
+      "learning_rate": 1.91833303939173e-06,
+      "loss": 1.4361,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8057319375321778,
+      "grad_norm": 4.390553008453175,
+      "learning_rate": 1.91506066279198e-06,
+      "loss": 1.6182,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8059035524283508,
+      "grad_norm": 4.661397986958235,
+      "learning_rate": 1.911790784075814e-06,
+      "loss": 1.4064,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8060751673245238,
+      "grad_norm": 4.973239896643652,
+      "learning_rate": 1.908523404253474e-06,
+      "loss": 1.3861,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8062467822206968,
+      "grad_norm": 4.3405868232545295,
+      "learning_rate": 1.9052585243344356e-06,
+      "loss": 1.5612,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8064183971168697,
+      "grad_norm": 6.351442865826272,
+      "learning_rate": 1.9019961453274005e-06,
+      "loss": 1.5969,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8065900120130427,
+      "grad_norm": 4.471195740534678,
+      "learning_rate": 1.8987362682402998e-06,
+      "loss": 1.5163,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8067616269092157,
+      "grad_norm": 3.9708576317752065,
+      "learning_rate": 1.895478894080287e-06,
+      "loss": 1.4619,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8069332418053887,
+      "grad_norm": 4.302066986283456,
+      "learning_rate": 1.8922240238537426e-06,
+      "loss": 1.506,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8071048567015617,
+      "grad_norm": 4.100824876716006,
+      "learning_rate": 1.8889716585662766e-06,
+      "loss": 1.545,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8072764715977346,
+      "grad_norm": 4.686480876882765,
+      "learning_rate": 1.885721799222725e-06,
+      "loss": 1.6029,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8074480864939076,
+      "grad_norm": 3.9824909649244393,
+      "learning_rate": 1.8824744468271506e-06,
+      "loss": 1.4758,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8076197013900807,
+      "grad_norm": 4.357171334647611,
+      "learning_rate": 1.879229602382837e-06,
+      "loss": 1.6839,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8077913162862537,
+      "grad_norm": 3.955053416751871,
+      "learning_rate": 1.875987266892294e-06,
+      "loss": 1.5745,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8079629311824267,
+      "grad_norm": 4.724695923305437,
+      "learning_rate": 1.872747441357259e-06,
+      "loss": 1.5799,
+      "step": 4708
+    },
+    {
+      "epoch": 0.8081345460785996,
+      "grad_norm": 4.232976906985312,
+      "learning_rate": 1.869510126778694e-06,
+      "loss": 1.5525,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8083061609747726,
+      "grad_norm": 5.048713374355663,
+      "learning_rate": 1.8662753241567855e-06,
+      "loss": 1.5795,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8084777758709456,
+      "grad_norm": 3.725181124131211,
+      "learning_rate": 1.8630430344909378e-06,
+      "loss": 1.6428,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8086493907671186,
+      "grad_norm": 4.565425202474584,
+      "learning_rate": 1.8598132587797891e-06,
+      "loss": 1.5042,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8088210056632916,
+      "grad_norm": 4.1232125966384165,
+      "learning_rate": 1.8565859980211898e-06,
+      "loss": 1.4514,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8089926205594645,
+      "grad_norm": 4.727431207816188,
+      "learning_rate": 1.8533612532122204e-06,
+      "loss": 1.6301,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8091642354556375,
+      "grad_norm": 4.875678069537726,
+      "learning_rate": 1.8501390253491858e-06,
+      "loss": 1.7241,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8093358503518105,
+      "grad_norm": 4.4394776014894335,
+      "learning_rate": 1.8469193154276033e-06,
+      "loss": 1.6361,
+      "step": 4716
+    },
+    {
+      "epoch": 0.8095074652479836,
+      "grad_norm": 3.9581792061584764,
+      "learning_rate": 1.8437021244422237e-06,
+      "loss": 1.4852,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8096790801441565,
+      "grad_norm": 4.5138373874870545,
+      "learning_rate": 1.8404874533870165e-06,
+      "loss": 1.6123,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8098506950403295,
+      "grad_norm": 3.9602782813737236,
+      "learning_rate": 1.837275303255165e-06,
+      "loss": 1.6562,
+      "step": 4719
+    },
+    {
+      "epoch": 0.8100223099365025,
+      "grad_norm": 4.6732817448838935,
+      "learning_rate": 1.834065675039084e-06,
+      "loss": 1.544,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8101939248326755,
+      "grad_norm": 3.9327035262853967,
+      "learning_rate": 1.8308585697304015e-06,
+      "loss": 1.5526,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8103655397288485,
+      "grad_norm": 4.416909666806564,
+      "learning_rate": 1.8276539883199706e-06,
+      "loss": 1.5307,
+      "step": 4722
+    },
+    {
+      "epoch": 0.8105371546250214,
+      "grad_norm": 4.253970280231939,
+      "learning_rate": 1.8244519317978638e-06,
+      "loss": 1.6755,
+      "step": 4723
+    },
+    {
+      "epoch": 0.8107087695211944,
+      "grad_norm": 4.217879330674126,
+      "learning_rate": 1.8212524011533738e-06,
+      "loss": 1.5802,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8108803844173674,
+      "grad_norm": 4.031020695680582,
+      "learning_rate": 1.8180553973750125e-06,
+      "loss": 1.4769,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8110519993135404,
+      "grad_norm": 4.257534302144836,
+      "learning_rate": 1.8148609214505053e-06,
+      "loss": 1.5336,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8112236142097135,
+      "grad_norm": 3.7624650914966487,
+      "learning_rate": 1.8116689743668047e-06,
+      "loss": 1.6382,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8113952291058864,
+      "grad_norm": 4.483923293459603,
+      "learning_rate": 1.808479557110081e-06,
+      "loss": 1.6449,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8115668440020594,
+      "grad_norm": 4.462424705003552,
+      "learning_rate": 1.8052926706657226e-06,
+      "loss": 1.4666,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8117384588982324,
+      "grad_norm": 5.586791444242014,
+      "learning_rate": 1.8021083160183317e-06,
+      "loss": 1.6004,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8119100737944054,
+      "grad_norm": 5.2472121785558,
+      "learning_rate": 1.7989264941517292e-06,
+      "loss": 1.5153,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8120816886905784,
+      "grad_norm": 4.033707504392723,
+      "learning_rate": 1.7957472060489568e-06,
+      "loss": 1.3814,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8122533035867513,
+      "grad_norm": 4.5260967878582745,
+      "learning_rate": 1.7925704526922728e-06,
+      "loss": 1.6383,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8124249184829243,
+      "grad_norm": 4.129105455195443,
+      "learning_rate": 1.7893962350631543e-06,
+      "loss": 1.5586,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8125965333790973,
+      "grad_norm": 3.376448378450376,
+      "learning_rate": 1.786224554142285e-06,
+      "loss": 1.3723,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8127681482752703,
+      "grad_norm": 4.13608464373614,
+      "learning_rate": 1.78305541090958e-06,
+      "loss": 1.5269,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8129397631714432,
+      "grad_norm": 4.216916110100886,
+      "learning_rate": 1.7798888063441556e-06,
+      "loss": 1.4988,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8131113780676162,
+      "grad_norm": 4.021696832185047,
+      "learning_rate": 1.776724741424354e-06,
+      "loss": 1.323,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8132829929637893,
+      "grad_norm": 3.4524435224767487,
+      "learning_rate": 1.7735632171277295e-06,
+      "loss": 1.3778,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8134546078599623,
+      "grad_norm": 3.8863716758862914,
+      "learning_rate": 1.770404234431049e-06,
+      "loss": 1.5976,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8136262227561353,
+      "grad_norm": 4.906590054353688,
+      "learning_rate": 1.7672477943102984e-06,
+      "loss": 1.57,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8137978376523082,
+      "grad_norm": 4.204491219090956,
+      "learning_rate": 1.7640938977406786e-06,
+      "loss": 1.5763,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8139694525484812,
+      "grad_norm": 4.640025319060085,
+      "learning_rate": 1.7609425456965957e-06,
+      "loss": 1.5754,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8141410674446542,
+      "grad_norm": 3.970601891040093,
+      "learning_rate": 1.757793739151683e-06,
+      "loss": 1.7266,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8143126823408272,
+      "grad_norm": 4.182392166077962,
+      "learning_rate": 1.7546474790787737e-06,
+      "loss": 1.4951,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8144842972370002,
+      "grad_norm": 3.7449454443494186,
+      "learning_rate": 1.7515037664499257e-06,
+      "loss": 1.3469,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8146559121331731,
+      "grad_norm": 3.8282159102359445,
+      "learning_rate": 1.748362602236403e-06,
+      "loss": 1.4556,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8148275270293461,
+      "grad_norm": 3.815868293526736,
+      "learning_rate": 1.7452239874086884e-06,
+      "loss": 1.2008,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8149991419255191,
+      "grad_norm": 4.333051862056449,
+      "learning_rate": 1.7420879229364695e-06,
+      "loss": 1.6515,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8151707568216922,
+      "grad_norm": 3.1891068396424935,
+      "learning_rate": 1.738954409788648e-06,
+      "loss": 1.2357,
+      "step": 4750
+    },
+    {
+      "epoch": 0.815342371717865,
+      "grad_norm": 4.017193360766739,
+      "learning_rate": 1.7358234489333415e-06,
+      "loss": 1.5298,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8155139866140381,
+      "grad_norm": 3.964710874867819,
+      "learning_rate": 1.7326950413378752e-06,
+      "loss": 1.2864,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8156856015102111,
+      "grad_norm": 4.362086560697036,
+      "learning_rate": 1.7295691879687904e-06,
+      "loss": 1.755,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8158572164063841,
+      "grad_norm": 3.8706716414158744,
+      "learning_rate": 1.7264458897918335e-06,
+      "loss": 1.443,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8160288313025571,
+      "grad_norm": 4.1645769703681506,
+      "learning_rate": 1.7233251477719593e-06,
+      "loss": 1.6262,
+      "step": 4755
+    },
+    {
+      "epoch": 0.81620044619873,
+      "grad_norm": 4.740649624803317,
+      "learning_rate": 1.7202069628733398e-06,
+      "loss": 1.7114,
+      "step": 4756
+    },
+    {
+      "epoch": 0.816372061094903,
+      "grad_norm": 4.422538843552403,
+      "learning_rate": 1.7170913360593566e-06,
+      "loss": 1.4931,
+      "step": 4757
+    },
+    {
+      "epoch": 0.816543675991076,
+      "grad_norm": 3.7528528981885745,
+      "learning_rate": 1.7139782682925988e-06,
+      "loss": 1.5633,
+      "step": 4758
+    },
+    {
+      "epoch": 0.816715290887249,
+      "grad_norm": 3.7275410790321786,
+      "learning_rate": 1.71086776053486e-06,
+      "loss": 1.5592,
+      "step": 4759
+    },
+    {
+      "epoch": 0.816886905783422,
+      "grad_norm": 4.2392713519734535,
+      "learning_rate": 1.707759813747153e-06,
+      "loss": 1.3791,
+      "step": 4760
+    },
+    {
+      "epoch": 0.817058520679595,
+      "grad_norm": 4.133337514841396,
+      "learning_rate": 1.7046544288896893e-06,
+      "loss": 1.3988,
+      "step": 4761
+    },
+    {
+      "epoch": 0.817230135575768,
+      "grad_norm": 4.0069726982922385,
+      "learning_rate": 1.7015516069218962e-06,
+      "loss": 1.5614,
+      "step": 4762
+    },
+    {
+      "epoch": 0.817401750471941,
+      "grad_norm": 3.7808619741729825,
+      "learning_rate": 1.6984513488024068e-06,
+      "loss": 1.3589,
+      "step": 4763
+    },
+    {
+      "epoch": 0.817573365368114,
+      "grad_norm": 3.6219637454769065,
+      "learning_rate": 1.6953536554890593e-06,
+      "loss": 1.4129,
+      "step": 4764
+    },
+    {
+      "epoch": 0.817744980264287,
+      "grad_norm": 6.2183632778325855,
+      "learning_rate": 1.6922585279389037e-06,
+      "loss": 1.6128,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8179165951604599,
+      "grad_norm": 4.232690515583152,
+      "learning_rate": 1.6891659671081983e-06,
+      "loss": 1.4493,
+      "step": 4766
+    },
+    {
+      "epoch": 0.8180882100566329,
+      "grad_norm": 4.328589401136215,
+      "learning_rate": 1.6860759739524003e-06,
+      "loss": 1.6701,
+      "step": 4767
+    },
+    {
+      "epoch": 0.8182598249528059,
+      "grad_norm": 3.515764342831475,
+      "learning_rate": 1.6829885494261832e-06,
+      "loss": 1.2537,
+      "step": 4768
+    },
+    {
+      "epoch": 0.8184314398489789,
+      "grad_norm": 4.523942785750915,
+      "learning_rate": 1.6799036944834191e-06,
+      "loss": 1.3621,
+      "step": 4769
+    },
+    {
+      "epoch": 0.8186030547451518,
+      "grad_norm": 4.123277429371624,
+      "learning_rate": 1.6768214100771917e-06,
+      "loss": 1.4811,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8187746696413248,
+      "grad_norm": 4.110616187162248,
+      "learning_rate": 1.6737416971597876e-06,
+      "loss": 1.547,
+      "step": 4771
+    },
+    {
+      "epoch": 0.8189462845374978,
+      "grad_norm": 3.859634526065418,
+      "learning_rate": 1.670664556682705e-06,
+      "loss": 1.6499,
+      "step": 4772
+    },
+    {
+      "epoch": 0.8191178994336709,
+      "grad_norm": 3.869249966716569,
+      "learning_rate": 1.6675899895966374e-06,
+      "loss": 1.3126,
+      "step": 4773
+    },
+    {
+      "epoch": 0.8192895143298439,
+      "grad_norm": 3.946414039882373,
+      "learning_rate": 1.6645179968514858e-06,
+      "loss": 1.4108,
+      "step": 4774
+    },
+    {
+      "epoch": 0.8194611292260168,
+      "grad_norm": 4.381632964889389,
+      "learning_rate": 1.6614485793963619e-06,
+      "loss": 1.5608,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8196327441221898,
+      "grad_norm": 3.692932448305564,
+      "learning_rate": 1.6583817381795775e-06,
+      "loss": 1.5078,
+      "step": 4776
+    },
+    {
+      "epoch": 0.8198043590183628,
+      "grad_norm": 3.5047680662497878,
+      "learning_rate": 1.655317474148652e-06,
+      "loss": 1.3324,
+      "step": 4777
+    },
+    {
+      "epoch": 0.8199759739145358,
+      "grad_norm": 4.581145451349372,
+      "learning_rate": 1.652255788250301e-06,
+      "loss": 1.7512,
+      "step": 4778
+    },
+    {
+      "epoch": 0.8201475888107088,
+      "grad_norm": 4.000911757978101,
+      "learning_rate": 1.6491966814304483e-06,
+      "loss": 1.4089,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8203192037068817,
+      "grad_norm": 4.5360554149582395,
+      "learning_rate": 1.6461401546342214e-06,
+      "loss": 1.5827,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8204908186030547,
+      "grad_norm": 4.479289803549678,
+      "learning_rate": 1.6430862088059497e-06,
+      "loss": 1.5964,
+      "step": 4781
+    },
+    {
+      "epoch": 0.8206624334992277,
+      "grad_norm": 4.987443943667715,
+      "learning_rate": 1.64003484488917e-06,
+      "loss": 1.5473,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8208340483954008,
+      "grad_norm": 4.239196834024588,
+      "learning_rate": 1.63698606382661e-06,
+      "loss": 1.6608,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8210056632915738,
+      "grad_norm": 3.909332322679399,
+      "learning_rate": 1.6339398665602112e-06,
+      "loss": 1.7064,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8211772781877467,
+      "grad_norm": 4.0157885244486256,
+      "learning_rate": 1.630896254031108e-06,
+      "loss": 1.4625,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8213488930839197,
+      "grad_norm": 4.699746618327083,
+      "learning_rate": 1.6278552271796422e-06,
+      "loss": 1.5505,
+      "step": 4786
+    },
+    {
+      "epoch": 0.8215205079800927,
+      "grad_norm": 4.451636242883355,
+      "learning_rate": 1.624816786945358e-06,
+      "loss": 1.3828,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8216921228762657,
+      "grad_norm": 4.109800815390824,
+      "learning_rate": 1.6217809342669898e-06,
+      "loss": 1.4333,
+      "step": 4788
+    },
+    {
+      "epoch": 0.8218637377724386,
+      "grad_norm": 3.6873793313267527,
+      "learning_rate": 1.6187476700824855e-06,
+      "loss": 1.4101,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8220353526686116,
+      "grad_norm": 3.985709483580345,
+      "learning_rate": 1.6157169953289876e-06,
+      "loss": 1.5512,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8222069675647846,
+      "grad_norm": 4.202289954357266,
+      "learning_rate": 1.6126889109428345e-06,
+      "loss": 1.5665,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8223785824609576,
+      "grad_norm": 4.612864097344822,
+      "learning_rate": 1.609663417859575e-06,
+      "loss": 1.3557,
+      "step": 4792
+    },
+    {
+      "epoch": 0.8225501973571306,
+      "grad_norm": 3.8126135954779325,
+      "learning_rate": 1.6066405170139431e-06,
+      "loss": 1.2275,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8227218122533035,
+      "grad_norm": 4.203206346147437,
+      "learning_rate": 1.6036202093398833e-06,
+      "loss": 1.6516,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8228934271494766,
+      "grad_norm": 4.69713864144923,
+      "learning_rate": 1.6006024957705357e-06,
+      "loss": 1.6989,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8230650420456496,
+      "grad_norm": 4.15576583508913,
+      "learning_rate": 1.5975873772382399e-06,
+      "loss": 1.5273,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8232366569418226,
+      "grad_norm": 3.7490580552955493,
+      "learning_rate": 1.594574854674531e-06,
+      "loss": 1.3211,
+      "step": 4797
+    },
+    {
+      "epoch": 0.8234082718379956,
+      "grad_norm": 4.282943635534798,
+      "learning_rate": 1.59156492901014e-06,
+      "loss": 1.3939,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8235798867341685,
+      "grad_norm": 4.76895852168925,
+      "learning_rate": 1.5885576011750004e-06,
+      "loss": 1.4788,
+      "step": 4799
+    },
+    {
+      "epoch": 0.8237515016303415,
+      "grad_norm": 3.8300849614430077,
+      "learning_rate": 1.5855528720982438e-06,
+      "loss": 1.4255,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8239231165265145,
+      "grad_norm": 5.016029711279405,
+      "learning_rate": 1.5825507427081976e-06,
+      "loss": 1.4969,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8240947314226875,
+      "grad_norm": 3.6640126225191882,
+      "learning_rate": 1.5795512139323832e-06,
+      "loss": 1.4868,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8242663463188605,
+      "grad_norm": 4.36041446119268,
+      "learning_rate": 1.5765542866975181e-06,
+      "loss": 1.7193,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8244379612150334,
+      "grad_norm": 4.12413500391701,
+      "learning_rate": 1.5735599619295206e-06,
+      "loss": 1.5277,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8246095761112064,
+      "grad_norm": 3.475721673600634,
+      "learning_rate": 1.5705682405535028e-06,
+      "loss": 1.3113,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8247811910073795,
+      "grad_norm": 3.7523614754537133,
+      "learning_rate": 1.5675791234937753e-06,
+      "loss": 1.5003,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8249528059035525,
+      "grad_norm": 4.476343290594001,
+      "learning_rate": 1.5645926116738353e-06,
+      "loss": 1.4724,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8251244207997254,
+      "grad_norm": 3.658220569229897,
+      "learning_rate": 1.5616087060163865e-06,
+      "loss": 1.4263,
+      "step": 4808
+    },
+    {
+      "epoch": 0.8252960356958984,
+      "grad_norm": 4.016234509455902,
+      "learning_rate": 1.5586274074433171e-06,
+      "loss": 1.4076,
+      "step": 4809
+    },
+    {
+      "epoch": 0.8254676505920714,
+      "grad_norm": 4.508069012498651,
+      "learning_rate": 1.5556487168757161e-06,
+      "loss": 1.6508,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8256392654882444,
+      "grad_norm": 3.976340598480248,
+      "learning_rate": 1.5526726352338695e-06,
+      "loss": 1.5211,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8258108803844174,
+      "grad_norm": 4.101646086204409,
+      "learning_rate": 1.549699163437247e-06,
+      "loss": 1.4162,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8259824952805903,
+      "grad_norm": 3.453704204465394,
+      "learning_rate": 1.54672830240452e-06,
+      "loss": 1.3222,
+      "step": 4813
+    },
+    {
+      "epoch": 0.8261541101767633,
+      "grad_norm": 3.3381820210281647,
+      "learning_rate": 1.5437600530535545e-06,
+      "loss": 1.2713,
+      "step": 4814
+    },
+    {
+      "epoch": 0.8263257250729363,
+      "grad_norm": 4.318833631122905,
+      "learning_rate": 1.540794416301402e-06,
+      "loss": 1.5501,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8264973399691093,
+      "grad_norm": 4.051409876299266,
+      "learning_rate": 1.5378313930643141e-06,
+      "loss": 1.5782,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8266689548652824,
+      "grad_norm": 3.892167570481565,
+      "learning_rate": 1.5348709842577302e-06,
+      "loss": 1.5381,
+      "step": 4817
+    },
+    {
+      "epoch": 0.8268405697614553,
+      "grad_norm": 4.539698615082214,
+      "learning_rate": 1.5319131907962837e-06,
+      "loss": 1.4202,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8270121846576283,
+      "grad_norm": 4.287637444952206,
+      "learning_rate": 1.528958013593801e-06,
+      "loss": 1.4394,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8271837995538013,
+      "grad_norm": 5.100675825487883,
+      "learning_rate": 1.5260054535633018e-06,
+      "loss": 1.4594,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8273554144499743,
+      "grad_norm": 3.6114765169500624,
+      "learning_rate": 1.523055511616992e-06,
+      "loss": 1.4732,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8275270293461472,
+      "grad_norm": 4.5860065102868015,
+      "learning_rate": 1.5201081886662695e-06,
+      "loss": 1.5918,
+      "step": 4822
+    },
+    {
+      "epoch": 0.8276986442423202,
+      "grad_norm": 4.490787211977917,
+      "learning_rate": 1.5171634856217265e-06,
+      "loss": 1.5841,
+      "step": 4823
+    },
+    {
+      "epoch": 0.8278702591384932,
+      "grad_norm": 4.267019101678434,
+      "learning_rate": 1.5142214033931447e-06,
+      "loss": 1.4969,
+      "step": 4824
+    },
+    {
+      "epoch": 0.8280418740346662,
+      "grad_norm": 3.842204597012564,
+      "learning_rate": 1.5112819428894976e-06,
+      "loss": 1.4329,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8282134889308392,
+      "grad_norm": 4.497900159979859,
+      "learning_rate": 1.5083451050189446e-06,
+      "loss": 1.3571,
+      "step": 4826
+    },
+    {
+      "epoch": 0.8283851038270121,
+      "grad_norm": 4.387558894410822,
+      "learning_rate": 1.5054108906888342e-06,
+      "loss": 1.5545,
+      "step": 4827
+    },
+    {
+      "epoch": 0.8285567187231851,
+      "grad_norm": 4.432256994511522,
+      "learning_rate": 1.5024793008057093e-06,
+      "loss": 1.5436,
+      "step": 4828
+    },
+    {
+      "epoch": 0.8287283336193582,
+      "grad_norm": 4.223784179180186,
+      "learning_rate": 1.4995503362753006e-06,
+      "loss": 1.6514,
+      "step": 4829
+    },
+    {
+      "epoch": 0.8288999485155312,
+      "grad_norm": 4.565491457944558,
+      "learning_rate": 1.4966239980025278e-06,
+      "loss": 1.4604,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8290715634117042,
+      "grad_norm": 4.046117453594091,
+      "learning_rate": 1.493700286891494e-06,
+      "loss": 1.3684,
+      "step": 4831
+    },
+    {
+      "epoch": 0.8292431783078771,
+      "grad_norm": 3.900408343149284,
+      "learning_rate": 1.4907792038454994e-06,
+      "loss": 1.5458,
+      "step": 4832
+    },
+    {
+      "epoch": 0.8294147932040501,
+      "grad_norm": 5.6475835022018614,
+      "learning_rate": 1.4878607497670229e-06,
+      "loss": 1.6353,
+      "step": 4833
+    },
+    {
+      "epoch": 0.8295864081002231,
+      "grad_norm": 4.502249600480854,
+      "learning_rate": 1.4849449255577386e-06,
+      "loss": 1.5623,
+      "step": 4834
+    },
+    {
+      "epoch": 0.8297580229963961,
+      "grad_norm": 3.297642927712247,
+      "learning_rate": 1.4820317321185074e-06,
+      "loss": 1.1199,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8299296378925691,
+      "grad_norm": 3.469177220456171,
+      "learning_rate": 1.4791211703493713e-06,
+      "loss": 1.4682,
+      "step": 4836
+    },
+    {
+      "epoch": 0.830101252788742,
+      "grad_norm": 3.687979980120009,
+      "learning_rate": 1.4762132411495644e-06,
+      "loss": 1.4484,
+      "step": 4837
+    },
+    {
+      "epoch": 0.830272867684915,
+      "grad_norm": 4.028501921288012,
+      "learning_rate": 1.4733079454175102e-06,
+      "loss": 1.4705,
+      "step": 4838
+    },
+    {
+      "epoch": 0.830444482581088,
+      "grad_norm": 4.243605256930422,
+      "learning_rate": 1.4704052840508087e-06,
+      "loss": 1.4964,
+      "step": 4839
+    },
+    {
+      "epoch": 0.8306160974772611,
+      "grad_norm": 4.513967087383656,
+      "learning_rate": 1.4675052579462578e-06,
+      "loss": 1.5293,
+      "step": 4840
+    },
+    {
+      "epoch": 0.830787712373434,
+      "grad_norm": 4.109978109307644,
+      "learning_rate": 1.464607867999831e-06,
+      "loss": 1.4468,
+      "step": 4841
+    },
+    {
+      "epoch": 0.830959327269607,
+      "grad_norm": 3.780967028429485,
+      "learning_rate": 1.4617131151066933e-06,
+      "loss": 1.4257,
+      "step": 4842
+    },
+    {
+      "epoch": 0.83113094216578,
+      "grad_norm": 4.9374045544993255,
+      "learning_rate": 1.4588210001611923e-06,
+      "loss": 1.4617,
+      "step": 4843
+    },
+    {
+      "epoch": 0.831302557061953,
+      "grad_norm": 4.334631495387098,
+      "learning_rate": 1.4559315240568672e-06,
+      "loss": 1.5331,
+      "step": 4844
+    },
+    {
+      "epoch": 0.831474171958126,
+      "grad_norm": 4.182110435102105,
+      "learning_rate": 1.4530446876864324e-06,
+      "loss": 1.5767,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8316457868542989,
+      "grad_norm": 4.413169523866981,
+      "learning_rate": 1.4501604919417878e-06,
+      "loss": 1.7041,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8318174017504719,
+      "grad_norm": 4.363013721531667,
+      "learning_rate": 1.447278937714024e-06,
+      "loss": 1.5366,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8319890166466449,
+      "grad_norm": 4.63214397270743,
+      "learning_rate": 1.444400025893412e-06,
+      "loss": 1.6583,
+      "step": 4848
+    },
+    {
+      "epoch": 0.8321606315428179,
+      "grad_norm": 3.958468464272295,
+      "learning_rate": 1.4415237573694086e-06,
+      "loss": 1.5197,
+      "step": 4849
+    },
+    {
+      "epoch": 0.832332246438991,
+      "grad_norm": 3.558106307386517,
+      "learning_rate": 1.438650133030648e-06,
+      "loss": 1.5807,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8325038613351639,
+      "grad_norm": 4.445468644247436,
+      "learning_rate": 1.4357791537649512e-06,
+      "loss": 1.6686,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8326754762313369,
+      "grad_norm": 4.2650448005872805,
+      "learning_rate": 1.4329108204593235e-06,
+      "loss": 1.6139,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8328470911275099,
+      "grad_norm": 4.5603046536703244,
+      "learning_rate": 1.43004513399995e-06,
+      "loss": 1.5786,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8330187060236829,
+      "grad_norm": 3.9397265358772655,
+      "learning_rate": 1.4271820952722038e-06,
+      "loss": 1.5033,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8331903209198559,
+      "grad_norm": 4.146378090835399,
+      "learning_rate": 1.4243217051606285e-06,
+      "loss": 1.4237,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8333619358160288,
+      "grad_norm": 3.4503975903945916,
+      "learning_rate": 1.4214639645489625e-06,
+      "loss": 1.3647,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8335335507122018,
+      "grad_norm": 3.9253066351318475,
+      "learning_rate": 1.4186088743201144e-06,
+      "loss": 1.8014,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8337051656083748,
+      "grad_norm": 4.089072348629632,
+      "learning_rate": 1.4157564353561815e-06,
+      "loss": 1.4447,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8338767805045478,
+      "grad_norm": 5.051106196694771,
+      "learning_rate": 1.4129066485384413e-06,
+      "loss": 1.5858,
+      "step": 4859
+    },
+    {
+      "epoch": 0.8340483954007207,
+      "grad_norm": 5.351557048739141,
+      "learning_rate": 1.410059514747345e-06,
+      "loss": 1.6448,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8342200102968937,
+      "grad_norm": 4.247637833490804,
+      "learning_rate": 1.4072150348625336e-06,
+      "loss": 1.5563,
+      "step": 4861
+    },
+    {
+      "epoch": 0.8343916251930668,
+      "grad_norm": 3.9093476089540244,
+      "learning_rate": 1.4043732097628239e-06,
+      "loss": 1.3501,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8345632400892398,
+      "grad_norm": 3.712540712754464,
+      "learning_rate": 1.401534040326209e-06,
+      "loss": 1.4207,
+      "step": 4863
+    },
+    {
+      "epoch": 0.8347348549854128,
+      "grad_norm": 3.785723576531704,
+      "learning_rate": 1.3986975274298687e-06,
+      "loss": 1.4789,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8349064698815857,
+      "grad_norm": 3.7824480604787802,
+      "learning_rate": 1.395863671950155e-06,
+      "loss": 1.5411,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8350780847777587,
+      "grad_norm": 4.009935612798529,
+      "learning_rate": 1.3930324747626034e-06,
+      "loss": 1.4992,
+      "step": 4866
+    },
+    {
+      "epoch": 0.8352496996739317,
+      "grad_norm": 4.213824742038228,
+      "learning_rate": 1.3902039367419262e-06,
+      "loss": 1.5721,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8354213145701047,
+      "grad_norm": 5.197594748157084,
+      "learning_rate": 1.3873780587620178e-06,
+      "loss": 1.5087,
+      "step": 4868
+    },
+    {
+      "epoch": 0.8355929294662777,
+      "grad_norm": 4.160118945048529,
+      "learning_rate": 1.3845548416959464e-06,
+      "loss": 1.4403,
+      "step": 4869
+    },
+    {
+      "epoch": 0.8357645443624506,
+      "grad_norm": 3.6151630356188655,
+      "learning_rate": 1.3817342864159555e-06,
+      "loss": 1.4555,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8359361592586236,
+      "grad_norm": 3.7273819753883157,
+      "learning_rate": 1.3789163937934735e-06,
+      "loss": 1.4141,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8361077741547966,
+      "grad_norm": 4.351352865307784,
+      "learning_rate": 1.3761011646991019e-06,
+      "loss": 1.5309,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8362793890509697,
+      "grad_norm": 5.068828150582419,
+      "learning_rate": 1.3732886000026235e-06,
+      "loss": 1.5127,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8364510039471426,
+      "grad_norm": 4.903164839468747,
+      "learning_rate": 1.3704787005729914e-06,
+      "loss": 1.4871,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8366226188433156,
+      "grad_norm": 4.000762469224886,
+      "learning_rate": 1.3676714672783376e-06,
+      "loss": 1.7044,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8367942337394886,
+      "grad_norm": 4.224156600248199,
+      "learning_rate": 1.3648669009859716e-06,
+      "loss": 1.292,
+      "step": 4876
+    },
+    {
+      "epoch": 0.8369658486356616,
+      "grad_norm": 3.849538820719491,
+      "learning_rate": 1.3620650025623805e-06,
+      "loss": 1.4644,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8371374635318346,
+      "grad_norm": 4.2560570899830905,
+      "learning_rate": 1.3592657728732262e-06,
+      "loss": 1.8338,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8373090784280075,
+      "grad_norm": 3.586444096943704,
+      "learning_rate": 1.35646921278334e-06,
+      "loss": 1.4924,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8374806933241805,
+      "grad_norm": 4.701462251937326,
+      "learning_rate": 1.35367532315674e-06,
+      "loss": 1.5888,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8376523082203535,
+      "grad_norm": 4.039342999633316,
+      "learning_rate": 1.350884104856608e-06,
+      "loss": 1.5599,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8378239231165265,
+      "grad_norm": 5.105211189598625,
+      "learning_rate": 1.348095558745306e-06,
+      "loss": 1.5338,
+      "step": 4882
+    },
+    {
+      "epoch": 0.8379955380126995,
+      "grad_norm": 4.124177488479943,
+      "learning_rate": 1.3453096856843716e-06,
+      "loss": 1.5136,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8381671529088724,
+      "grad_norm": 4.008081851658497,
+      "learning_rate": 1.3425264865345134e-06,
+      "loss": 1.2128,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8383387678050455,
+      "grad_norm": 4.953653803735307,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 1.5761,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8385103827012185,
+      "grad_norm": 3.6739364663311913,
+      "learning_rate": 1.3369681134067314e-06,
+      "loss": 1.5181,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8386819975973915,
+      "grad_norm": 4.060577047166199,
+      "learning_rate": 1.3341929411460986e-06,
+      "loss": 1.4968,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8388536124935645,
+      "grad_norm": 4.229375995298623,
+      "learning_rate": 1.3314204462311186e-06,
+      "loss": 1.3645,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8390252273897374,
+      "grad_norm": 4.906805488778325,
+      "learning_rate": 1.3286506295183643e-06,
+      "loss": 1.4489,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8391968422859104,
+      "grad_norm": 4.63708358627252,
+      "learning_rate": 1.3258834918635865e-06,
+      "loss": 1.6983,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8393684571820834,
+      "grad_norm": 3.884785921572161,
+      "learning_rate": 1.3231190341217081e-06,
+      "loss": 1.4352,
+      "step": 4891
+    },
+    {
+      "epoch": 0.8395400720782564,
+      "grad_norm": 4.8258368195583765,
+      "learning_rate": 1.3203572571468238e-06,
+      "loss": 1.4187,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8397116869744293,
+      "grad_norm": 4.848819391784695,
+      "learning_rate": 1.3175981617921974e-06,
+      "loss": 1.5742,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8398833018706023,
+      "grad_norm": 5.020448839612587,
+      "learning_rate": 1.3148417489102628e-06,
+      "loss": 1.654,
+      "step": 4894
+    },
+    {
+      "epoch": 0.8400549167667754,
+      "grad_norm": 5.224156094790471,
+      "learning_rate": 1.3120880193526297e-06,
+      "loss": 1.6418,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8402265316629484,
+      "grad_norm": 4.867361771294044,
+      "learning_rate": 1.3093369739700768e-06,
+      "loss": 1.4601,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8403981465591214,
+      "grad_norm": 4.696972094899303,
+      "learning_rate": 1.306588613612557e-06,
+      "loss": 1.5774,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8405697614552943,
+      "grad_norm": 4.833801867960432,
+      "learning_rate": 1.3038429391291862e-06,
+      "loss": 1.6262,
+      "step": 4898
+    },
+    {
+      "epoch": 0.8407413763514673,
+      "grad_norm": 5.050016630207017,
+      "learning_rate": 1.3010999513682586e-06,
+      "loss": 1.4824,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8409129912476403,
+      "grad_norm": 4.516460164717836,
+      "learning_rate": 1.298359651177229e-06,
+      "loss": 1.5337,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8410846061438133,
+      "grad_norm": 4.131405454190068,
+      "learning_rate": 1.2956220394027309e-06,
+      "loss": 1.4123,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8412562210399863,
+      "grad_norm": 4.619958398703142,
+      "learning_rate": 1.2928871168905643e-06,
+      "loss": 1.8099,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8414278359361592,
+      "grad_norm": 4.142511540821578,
+      "learning_rate": 1.2901548844856948e-06,
+      "loss": 1.5955,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8415994508323322,
+      "grad_norm": 4.157974422323813,
+      "learning_rate": 1.2874253430322604e-06,
+      "loss": 1.5653,
+      "step": 4904
+    },
+    {
+      "epoch": 0.8417710657285052,
+      "grad_norm": 4.3324850075218,
+      "learning_rate": 1.2846984933735696e-06,
+      "loss": 1.6068,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8419426806246783,
+      "grad_norm": 3.551207823278202,
+      "learning_rate": 1.281974336352092e-06,
+      "loss": 1.3378,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8421142955208513,
+      "grad_norm": 4.0570514944507305,
+      "learning_rate": 1.2792528728094756e-06,
+      "loss": 1.375,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8422859104170242,
+      "grad_norm": 4.149422610434124,
+      "learning_rate": 1.2765341035865253e-06,
+      "loss": 1.5898,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8424575253131972,
+      "grad_norm": 4.00378279437493,
+      "learning_rate": 1.2738180295232206e-06,
+      "loss": 1.5508,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8426291402093702,
+      "grad_norm": 3.874874270494233,
+      "learning_rate": 1.2711046514587067e-06,
+      "loss": 1.5235,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8428007551055432,
+      "grad_norm": 4.733481077862293,
+      "learning_rate": 1.2683939702312986e-06,
+      "loss": 1.6161,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8429723700017161,
+      "grad_norm": 3.905714841185302,
+      "learning_rate": 1.2656859866784721e-06,
+      "loss": 1.4473,
+      "step": 4912
+    },
+    {
+      "epoch": 0.8431439848978891,
+      "grad_norm": 4.001001903573795,
+      "learning_rate": 1.2629807016368712e-06,
+      "loss": 1.4721,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8433155997940621,
+      "grad_norm": 4.717438227131749,
+      "learning_rate": 1.2602781159423095e-06,
+      "loss": 1.6382,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8434872146902351,
+      "grad_norm": 4.286179263292802,
+      "learning_rate": 1.2575782304297647e-06,
+      "loss": 1.5085,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8436588295864081,
+      "grad_norm": 4.080322141344338,
+      "learning_rate": 1.2548810459333816e-06,
+      "loss": 1.6588,
+      "step": 4916
+    },
+    {
+      "epoch": 0.843830444482581,
+      "grad_norm": 5.021944623988521,
+      "learning_rate": 1.2521865632864694e-06,
+      "loss": 1.5803,
+      "step": 4917
+    },
+    {
+      "epoch": 0.844002059378754,
+      "grad_norm": 4.175139683115828,
+      "learning_rate": 1.2494947833214976e-06,
+      "loss": 1.2728,
+      "step": 4918
+    },
+    {
+      "epoch": 0.8441736742749271,
+      "grad_norm": 4.364286458913085,
+      "learning_rate": 1.24680570687011e-06,
+      "loss": 1.3974,
+      "step": 4919
+    },
+    {
+      "epoch": 0.8443452891711001,
+      "grad_norm": 4.732550637583035,
+      "learning_rate": 1.2441193347631086e-06,
+      "loss": 1.3219,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8445169040672731,
+      "grad_norm": 4.433737436142455,
+      "learning_rate": 1.2414356678304652e-06,
+      "loss": 1.5013,
+      "step": 4921
+    },
+    {
+      "epoch": 0.844688518963446,
+      "grad_norm": 5.3651236085945575,
+      "learning_rate": 1.2387547069013073e-06,
+      "loss": 1.4414,
+      "step": 4922
+    },
+    {
+      "epoch": 0.844860133859619,
+      "grad_norm": 5.52513695681018,
+      "learning_rate": 1.236076452803937e-06,
+      "loss": 1.6948,
+      "step": 4923
+    },
+    {
+      "epoch": 0.845031748755792,
+      "grad_norm": 3.537381656573495,
+      "learning_rate": 1.2334009063658103e-06,
+      "loss": 1.4957,
+      "step": 4924
+    },
+    {
+      "epoch": 0.845203363651965,
+      "grad_norm": 4.492606462761248,
+      "learning_rate": 1.2307280684135514e-06,
+      "loss": 1.3089,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8453749785481379,
+      "grad_norm": 4.3620071337405895,
+      "learning_rate": 1.2280579397729507e-06,
+      "loss": 1.5648,
+      "step": 4926
+    },
+    {
+      "epoch": 0.8455465934443109,
+      "grad_norm": 4.169530162235752,
+      "learning_rate": 1.2253905212689554e-06,
+      "loss": 1.628,
+      "step": 4927
+    },
+    {
+      "epoch": 0.845718208340484,
+      "grad_norm": 4.030526324953615,
+      "learning_rate": 1.2227258137256782e-06,
+      "loss": 1.3884,
+      "step": 4928
+    },
+    {
+      "epoch": 0.845889823236657,
+      "grad_norm": 4.409947364191878,
+      "learning_rate": 1.2200638179663959e-06,
+      "loss": 1.3175,
+      "step": 4929
+    },
+    {
+      "epoch": 0.84606143813283,
+      "grad_norm": 4.162913692704332,
+      "learning_rate": 1.2174045348135433e-06,
+      "loss": 1.4428,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8462330530290029,
+      "grad_norm": 3.6697022807315673,
+      "learning_rate": 1.214747965088723e-06,
+      "loss": 1.5566,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8464046679251759,
+      "grad_norm": 3.9670220174527615,
+      "learning_rate": 1.2120941096126925e-06,
+      "loss": 1.4143,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8465762828213489,
+      "grad_norm": 4.13642767727758,
+      "learning_rate": 1.2094429692053745e-06,
+      "loss": 1.611,
+      "step": 4933
+    },
+    {
+      "epoch": 0.8467478977175219,
+      "grad_norm": 4.2563311703541995,
+      "learning_rate": 1.2067945446858542e-06,
+      "loss": 1.5385,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8469195126136949,
+      "grad_norm": 3.8465462998280637,
+      "learning_rate": 1.2041488368723763e-06,
+      "loss": 1.5377,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8470911275098678,
+      "grad_norm": 3.7779903681907374,
+      "learning_rate": 1.2015058465823458e-06,
+      "loss": 1.4495,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8472627424060408,
+      "grad_norm": 4.153909268677651,
+      "learning_rate": 1.1988655746323242e-06,
+      "loss": 1.6724,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8474343573022138,
+      "grad_norm": 3.820918303074717,
+      "learning_rate": 1.1962280218380406e-06,
+      "loss": 1.513,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8476059721983868,
+      "grad_norm": 4.3417443080505,
+      "learning_rate": 1.1935931890143793e-06,
+      "loss": 1.5835,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8477775870945599,
+      "grad_norm": 4.267216106440781,
+      "learning_rate": 1.1909610769753887e-06,
+      "loss": 1.3173,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8479492019907328,
+      "grad_norm": 4.285624146019127,
+      "learning_rate": 1.188331686534271e-06,
+      "loss": 1.5256,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8481208168869058,
+      "grad_norm": 3.887066518498248,
+      "learning_rate": 1.185705018503387e-06,
+      "loss": 1.4094,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8482924317830788,
+      "grad_norm": 4.103010896551944,
+      "learning_rate": 1.183081073694261e-06,
+      "loss": 1.3719,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8484640466792518,
+      "grad_norm": 4.526756374001578,
+      "learning_rate": 1.180459852917576e-06,
+      "loss": 1.4711,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8486356615754247,
+      "grad_norm": 3.609446816689786,
+      "learning_rate": 1.1778413569831726e-06,
+      "loss": 1.4199,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8488072764715977,
+      "grad_norm": 5.953425237138315,
+      "learning_rate": 1.175225586700045e-06,
+      "loss": 1.6101,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8489788913677707,
+      "grad_norm": 4.882132753359956,
+      "learning_rate": 1.1726125428763523e-06,
+      "loss": 1.4541,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8491505062639437,
+      "grad_norm": 5.112609916836839,
+      "learning_rate": 1.1700022263194043e-06,
+      "loss": 1.579,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8493221211601167,
+      "grad_norm": 4.361799921627873,
+      "learning_rate": 1.1673946378356738e-06,
+      "loss": 1.5055,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8494937360562896,
+      "grad_norm": 4.39054833895474,
+      "learning_rate": 1.1647897782307894e-06,
+      "loss": 1.4197,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8496653509524627,
+      "grad_norm": 4.622661432081524,
+      "learning_rate": 1.162187648309534e-06,
+      "loss": 1.6335,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8498369658486357,
+      "grad_norm": 4.596682299517377,
+      "learning_rate": 1.1595882488758492e-06,
+      "loss": 1.607,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8500085807448087,
+      "grad_norm": 3.653893226363181,
+      "learning_rate": 1.1569915807328358e-06,
+      "loss": 1.2539,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8501801956409817,
+      "grad_norm": 4.791232507713193,
+      "learning_rate": 1.1543976446827444e-06,
+      "loss": 1.3017,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8503518105371546,
+      "grad_norm": 3.961604458265738,
+      "learning_rate": 1.151806441526987e-06,
+      "loss": 1.6042,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8505234254333276,
+      "grad_norm": 4.63276277217101,
+      "learning_rate": 1.1492179720661267e-06,
+      "loss": 1.4828,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8506950403295006,
+      "grad_norm": 4.785406970637545,
+      "learning_rate": 1.146632237099885e-06,
+      "loss": 1.4945,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8508666552256736,
+      "grad_norm": 4.572383268847962,
+      "learning_rate": 1.1440492374271394e-06,
+      "loss": 1.6562,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8510382701218466,
+      "grad_norm": 4.518244031891752,
+      "learning_rate": 1.141468973845923e-06,
+      "loss": 1.8368,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8512098850180195,
+      "grad_norm": 4.444030206365939,
+      "learning_rate": 1.1388914471534184e-06,
+      "loss": 1.6599,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8513814999141925,
+      "grad_norm": 4.855465196737093,
+      "learning_rate": 1.136316658145964e-06,
+      "loss": 1.5328,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8515531148103656,
+      "grad_norm": 4.65434714192929,
+      "learning_rate": 1.1337446076190561e-06,
+      "loss": 1.4113,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8517247297065386,
+      "grad_norm": 3.802153942979901,
+      "learning_rate": 1.1311752963673438e-06,
+      "loss": 1.4784,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8518963446027115,
+      "grad_norm": 4.532892654549196,
+      "learning_rate": 1.12860872518463e-06,
+      "loss": 1.6059,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8520679594988845,
+      "grad_norm": 4.107953320866706,
+      "learning_rate": 1.1260448948638691e-06,
+      "loss": 1.5517,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8522395743950575,
+      "grad_norm": 3.810765754322222,
+      "learning_rate": 1.1234838061971675e-06,
+      "loss": 1.371,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8524111892912305,
+      "grad_norm": 4.140868989260617,
+      "learning_rate": 1.1209254599757869e-06,
+      "loss": 1.4632,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8525828041874035,
+      "grad_norm": 3.8428772865038643,
+      "learning_rate": 1.1183698569901435e-06,
+      "loss": 1.3862,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8527544190835764,
+      "grad_norm": 4.890515004008787,
+      "learning_rate": 1.115816998029805e-06,
+      "loss": 1.5759,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8529260339797494,
+      "grad_norm": 3.795053919538234,
+      "learning_rate": 1.1132668838834882e-06,
+      "loss": 1.5987,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8530976488759224,
+      "grad_norm": 4.1313104099208315,
+      "learning_rate": 1.110719515339065e-06,
+      "loss": 1.5364,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8532692637720954,
+      "grad_norm": 7.544968788076173,
+      "learning_rate": 1.1081748931835556e-06,
+      "loss": 1.4705,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8534408786682685,
+      "grad_norm": 4.5034509544070325,
+      "learning_rate": 1.105633018203136e-06,
+      "loss": 1.6429,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8536124935644414,
+      "grad_norm": 4.654519761720039,
+      "learning_rate": 1.1030938911831334e-06,
+      "loss": 1.4978,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8537841084606144,
+      "grad_norm": 3.808434468753352,
+      "learning_rate": 1.1005575129080203e-06,
+      "loss": 1.5589,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8539557233567874,
+      "grad_norm": 3.9555730128612754,
+      "learning_rate": 1.0980238841614266e-06,
+      "loss": 1.6177,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8541273382529604,
+      "grad_norm": 4.1284567248545985,
+      "learning_rate": 1.0954930057261303e-06,
+      "loss": 1.4358,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8542989531491334,
+      "grad_norm": 3.807985608381048,
+      "learning_rate": 1.0929648783840562e-06,
+      "loss": 1.3752,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8544705680453063,
+      "grad_norm": 5.248060134853908,
+      "learning_rate": 1.0904395029162863e-06,
+      "loss": 1.5525,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8546421829414793,
+      "grad_norm": 4.8533334945416176,
+      "learning_rate": 1.0879168801030437e-06,
+      "loss": 1.3577,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8548137978376523,
+      "grad_norm": 4.015377988435641,
+      "learning_rate": 1.085397010723709e-06,
+      "loss": 1.431,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8549854127338253,
+      "grad_norm": 3.921053711843901,
+      "learning_rate": 1.0828798955568065e-06,
+      "loss": 1.5875,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8551570276299982,
+      "grad_norm": 4.569632444159621,
+      "learning_rate": 1.080365535380017e-06,
+      "loss": 1.2945,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8553286425261712,
+      "grad_norm": 4.345905303645349,
+      "learning_rate": 1.0778539309701609e-06,
+      "loss": 1.5331,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8555002574223443,
+      "grad_norm": 3.996629916715755,
+      "learning_rate": 1.0753450831032087e-06,
+      "loss": 1.4953,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8556718723185173,
+      "grad_norm": 4.238271019743073,
+      "learning_rate": 1.072838992554286e-06,
+      "loss": 1.6939,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8558434872146903,
+      "grad_norm": 4.765757086947947,
+      "learning_rate": 1.0703356600976611e-06,
+      "loss": 1.4338,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8560151021108632,
+      "grad_norm": 3.800372706873989,
+      "learning_rate": 1.0678350865067533e-06,
+      "loss": 1.4469,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8561867170070362,
+      "grad_norm": 4.568999646801855,
+      "learning_rate": 1.0653372725541267e-06,
+      "loss": 1.7575,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8563583319032092,
+      "grad_norm": 4.482562117263046,
+      "learning_rate": 1.0628422190114895e-06,
+      "loss": 1.3495,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8565299467993822,
+      "grad_norm": 4.213145510294035,
+      "learning_rate": 1.0603499266497063e-06,
+      "loss": 1.493,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8567015616955552,
+      "grad_norm": 4.1229405156760235,
+      "learning_rate": 1.0578603962387813e-06,
+      "loss": 1.4297,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8568731765917281,
+      "grad_norm": 4.279019490313811,
+      "learning_rate": 1.055373628547871e-06,
+      "loss": 1.379,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8570447914879011,
+      "grad_norm": 3.929186299149795,
+      "learning_rate": 1.0528896243452713e-06,
+      "loss": 1.329,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8572164063840741,
+      "grad_norm": 3.902466070417452,
+      "learning_rate": 1.0504083843984304e-06,
+      "loss": 1.2828,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8573880212802472,
+      "grad_norm": 4.030466353548272,
+      "learning_rate": 1.0479299094739381e-06,
+      "loss": 1.435,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8575596361764201,
+      "grad_norm": 4.445195079184545,
+      "learning_rate": 1.0454542003375323e-06,
+      "loss": 1.6691,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8577312510725931,
+      "grad_norm": 5.1572071007171925,
+      "learning_rate": 1.0429812577540987e-06,
+      "loss": 1.2057,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8579028659687661,
+      "grad_norm": 3.8568960779106245,
+      "learning_rate": 1.0405110824876619e-06,
+      "loss": 1.4172,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8580744808649391,
+      "grad_norm": 4.248273821501593,
+      "learning_rate": 1.0380436753013978e-06,
+      "loss": 1.4256,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8582460957611121,
+      "grad_norm": 5.1034806290838,
+      "learning_rate": 1.0355790369576257e-06,
+      "loss": 1.6606,
+      "step": 5001
+    },
+    {
+      "epoch": 0.858417710657285,
+      "grad_norm": 4.392581789796283,
+      "learning_rate": 1.0331171682178043e-06,
+      "loss": 1.5253,
+      "step": 5002
+    },
+    {
+      "epoch": 0.858589325553458,
+      "grad_norm": 3.947293078550276,
+      "learning_rate": 1.0306580698425439e-06,
+      "loss": 1.6636,
+      "step": 5003
+    },
+    {
+      "epoch": 0.858760940449631,
+      "grad_norm": 3.9923118216889217,
+      "learning_rate": 1.0282017425915936e-06,
+      "loss": 1.691,
+      "step": 5004
+    },
+    {
+      "epoch": 0.858932555345804,
+      "grad_norm": 4.672225209413681,
+      "learning_rate": 1.0257481872238483e-06,
+      "loss": 1.5384,
+      "step": 5005
+    },
+    {
+      "epoch": 0.859104170241977,
+      "grad_norm": 4.075767945124762,
+      "learning_rate": 1.0232974044973476e-06,
+      "loss": 1.5039,
+      "step": 5006
+    },
+    {
+      "epoch": 0.85927578513815,
+      "grad_norm": 4.695985161143493,
+      "learning_rate": 1.0208493951692755e-06,
+      "loss": 1.5919,
+      "step": 5007
+    },
+    {
+      "epoch": 0.859447400034323,
+      "grad_norm": 4.646407671431997,
+      "learning_rate": 1.018404159995955e-06,
+      "loss": 1.5713,
+      "step": 5008
+    },
+    {
+      "epoch": 0.859619014930496,
+      "grad_norm": 5.681333667663091,
+      "learning_rate": 1.0159616997328514e-06,
+      "loss": 1.5,
+      "step": 5009
+    },
+    {
+      "epoch": 0.859790629826669,
+      "grad_norm": 4.211669878975965,
+      "learning_rate": 1.0135220151345781e-06,
+      "loss": 1.3252,
+      "step": 5010
+    },
+    {
+      "epoch": 0.859962244722842,
+      "grad_norm": 5.012749115569514,
+      "learning_rate": 1.0110851069548887e-06,
+      "loss": 1.3362,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8601338596190149,
+      "grad_norm": 4.799166594280435,
+      "learning_rate": 1.0086509759466789e-06,
+      "loss": 1.5699,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8603054745151879,
+      "grad_norm": 4.11167738160439,
+      "learning_rate": 1.0062196228619848e-06,
+      "loss": 1.2304,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8604770894113609,
+      "grad_norm": 3.85426264548864,
+      "learning_rate": 1.003791048451983e-06,
+      "loss": 1.3575,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8606487043075339,
+      "grad_norm": 4.932616436040077,
+      "learning_rate": 1.0013652534669982e-06,
+      "loss": 1.6809,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8608203192037068,
+      "grad_norm": 4.072350979329002,
+      "learning_rate": 9.989422386564895e-07,
+      "loss": 1.5079,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8609919340998798,
+      "grad_norm": 4.183295675641887,
+      "learning_rate": 9.965220047690626e-07,
+      "loss": 1.5894,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8611635489960529,
+      "grad_norm": 4.699226910725941,
+      "learning_rate": 9.941045525524573e-07,
+      "loss": 1.4409,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8613351638922259,
+      "grad_norm": 5.677388033360187,
+      "learning_rate": 9.916898827535625e-07,
+      "loss": 1.629,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8615067787883989,
+      "grad_norm": 5.537331101016582,
+      "learning_rate": 9.892779961183985e-07,
+      "loss": 1.5652,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8616783936845718,
+      "grad_norm": 3.904377098747771,
+      "learning_rate": 9.868688933921322e-07,
+      "loss": 1.4424,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8618500085807448,
+      "grad_norm": 4.1536312005977125,
+      "learning_rate": 9.844625753190696e-07,
+      "loss": 1.4112,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8620216234769178,
+      "grad_norm": 5.393470848636458,
+      "learning_rate": 9.820590426426523e-07,
+      "loss": 1.3961,
+      "step": 5023
+    },
+    {
+      "epoch": 0.8621932383730908,
+      "grad_norm": 4.294607425487386,
+      "learning_rate": 9.796582961054645e-07,
+      "loss": 1.4801,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8623648532692638,
+      "grad_norm": 4.24026990525951,
+      "learning_rate": 9.772603364492316e-07,
+      "loss": 1.5248,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8625364681654367,
+      "grad_norm": 4.235663044342673,
+      "learning_rate": 9.748651644148132e-07,
+      "loss": 1.2612,
+      "step": 5026
+    },
+    {
+      "epoch": 0.8627080830616097,
+      "grad_norm": 4.108277109157669,
+      "learning_rate": 9.72472780742212e-07,
+      "loss": 1.4762,
+      "step": 5027
+    },
+    {
+      "epoch": 0.8628796979577827,
+      "grad_norm": 4.4272745917650385,
+      "learning_rate": 9.700831861705639e-07,
+      "loss": 1.4225,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8630513128539558,
+      "grad_norm": 3.7793833088170388,
+      "learning_rate": 9.67696381438149e-07,
+      "loss": 1.4758,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8632229277501288,
+      "grad_norm": 5.045069406961064,
+      "learning_rate": 9.653123672823817e-07,
+      "loss": 1.4896,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8633945426463017,
+      "grad_norm": 4.22460255620109,
+      "learning_rate": 9.629311444398193e-07,
+      "loss": 1.4656,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8635661575424747,
+      "grad_norm": 3.774385570847925,
+      "learning_rate": 9.60552713646149e-07,
+      "loss": 1.3369,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8637377724386477,
+      "grad_norm": 4.2488624472773635,
+      "learning_rate": 9.581770756361975e-07,
+      "loss": 1.3296,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8639093873348207,
+      "grad_norm": 4.1280459677980526,
+      "learning_rate": 9.558042311439331e-07,
+      "loss": 1.446,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8640810022309936,
+      "grad_norm": 4.343120172305175,
+      "learning_rate": 9.534341809024583e-07,
+      "loss": 1.6202,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8642526171271666,
+      "grad_norm": 4.372806833612556,
+      "learning_rate": 9.510669256440142e-07,
+      "loss": 1.5609,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8644242320233396,
+      "grad_norm": 4.423582235138371,
+      "learning_rate": 9.487024660999733e-07,
+      "loss": 1.7528,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8645958469195126,
+      "grad_norm": 4.31289752392683,
+      "learning_rate": 9.46340803000847e-07,
+      "loss": 1.5075,
+      "step": 5038
+    },
+    {
+      "epoch": 0.8647674618156856,
+      "grad_norm": 4.66911123731338,
+      "learning_rate": 9.439819370762848e-07,
+      "loss": 1.4164,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8649390767118585,
+      "grad_norm": 4.370870713636476,
+      "learning_rate": 9.416258690550706e-07,
+      "loss": 1.7085,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8651106916080316,
+      "grad_norm": 4.168612314843888,
+      "learning_rate": 9.392725996651253e-07,
+      "loss": 1.3594,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8652823065042046,
+      "grad_norm": 4.336387567101559,
+      "learning_rate": 9.369221296335007e-07,
+      "loss": 1.4699,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8654539214003776,
+      "grad_norm": 4.915770565892056,
+      "learning_rate": 9.345744596863892e-07,
+      "loss": 1.3862,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8656255362965506,
+      "grad_norm": 4.011183926503289,
+      "learning_rate": 9.322295905491119e-07,
+      "loss": 1.3517,
+      "step": 5044
+    },
+    {
+      "epoch": 0.8657971511927235,
+      "grad_norm": 4.4582046103936674,
+      "learning_rate": 9.298875229461301e-07,
+      "loss": 1.6414,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8659687660888965,
+      "grad_norm": 4.534494007187588,
+      "learning_rate": 9.275482576010409e-07,
+      "loss": 1.5888,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8661403809850695,
+      "grad_norm": 4.49282827575606,
+      "learning_rate": 9.252117952365669e-07,
+      "loss": 1.3449,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8663119958812425,
+      "grad_norm": 4.875803471072843,
+      "learning_rate": 9.228781365745721e-07,
+      "loss": 1.6039,
+      "step": 5048
+    },
+    {
+      "epoch": 0.8664836107774154,
+      "grad_norm": 4.599963016744397,
+      "learning_rate": 9.20547282336055e-07,
+      "loss": 1.5185,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8666552256735884,
+      "grad_norm": 4.743972922196142,
+      "learning_rate": 9.18219233241141e-07,
+      "loss": 1.7756,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8668268405697614,
+      "grad_norm": 5.507502229554712,
+      "learning_rate": 9.158939900090968e-07,
+      "loss": 1.4671,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8669984554659345,
+      "grad_norm": 4.854221079348115,
+      "learning_rate": 9.135715533583134e-07,
+      "loss": 1.5826,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8671700703621075,
+      "grad_norm": 4.546582928269538,
+      "learning_rate": 9.112519240063234e-07,
+      "loss": 1.4795,
+      "step": 5053
+    },
+    {
+      "epoch": 0.8673416852582804,
+      "grad_norm": 4.404939154781044,
+      "learning_rate": 9.089351026697868e-07,
+      "loss": 1.4465,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8675133001544534,
+      "grad_norm": 5.352269181698289,
+      "learning_rate": 9.066210900644989e-07,
+      "loss": 1.471,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8676849150506264,
+      "grad_norm": 4.629490754087409,
+      "learning_rate": 9.043098869053846e-07,
+      "loss": 1.6782,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8678565299467994,
+      "grad_norm": 4.479723296306851,
+      "learning_rate": 9.020014939065014e-07,
+      "loss": 1.462,
+      "step": 5057
+    },
+    {
+      "epoch": 0.8680281448429724,
+      "grad_norm": 4.414508127403711,
+      "learning_rate": 8.996959117810388e-07,
+      "loss": 1.4556,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8681997597391453,
+      "grad_norm": 4.249810920876259,
+      "learning_rate": 8.973931412413195e-07,
+      "loss": 1.514,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8683713746353183,
+      "grad_norm": 4.041377062003575,
+      "learning_rate": 8.950931829987985e-07,
+      "loss": 1.5281,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8685429895314913,
+      "grad_norm": 3.9156421493984714,
+      "learning_rate": 8.927960377640565e-07,
+      "loss": 1.3325,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8687146044276643,
+      "grad_norm": 4.652944484996688,
+      "learning_rate": 8.90501706246808e-07,
+      "loss": 1.6896,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8688862193238374,
+      "grad_norm": 4.370289293022309,
+      "learning_rate": 8.882101891558992e-07,
+      "loss": 1.6389,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8690578342200103,
+      "grad_norm": 5.355449990024723,
+      "learning_rate": 8.859214871993071e-07,
+      "loss": 1.4631,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8692294491161833,
+      "grad_norm": 3.645103949548479,
+      "learning_rate": 8.836356010841385e-07,
+      "loss": 1.4555,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8694010640123563,
+      "grad_norm": 4.609444744786666,
+      "learning_rate": 8.813525315166271e-07,
+      "loss": 1.5836,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8695726789085293,
+      "grad_norm": 4.41118749061462,
+      "learning_rate": 8.790722792021422e-07,
+      "loss": 1.403,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8697442938047022,
+      "grad_norm": 6.118946041906018,
+      "learning_rate": 8.767948448451758e-07,
+      "loss": 1.3973,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8699159087008752,
+      "grad_norm": 3.881324435908571,
+      "learning_rate": 8.745202291493548e-07,
+      "loss": 1.282,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8700875235970482,
+      "grad_norm": 4.657596366879938,
+      "learning_rate": 8.722484328174364e-07,
+      "loss": 1.4645,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8702591384932212,
+      "grad_norm": 4.333360027057549,
+      "learning_rate": 8.699794565512976e-07,
+      "loss": 1.4134,
+      "step": 5071
+    },
+    {
+      "epoch": 0.8704307533893942,
+      "grad_norm": 4.349218612199456,
+      "learning_rate": 8.677133010519545e-07,
+      "loss": 1.469,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8706023682855671,
+      "grad_norm": 4.22087238577858,
+      "learning_rate": 8.654499670195482e-07,
+      "loss": 1.661,
+      "step": 5073
+    },
+    {
+      "epoch": 0.8707739831817402,
+      "grad_norm": 4.2643783383128895,
+      "learning_rate": 8.631894551533437e-07,
+      "loss": 1.4659,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8709455980779132,
+      "grad_norm": 3.7801301587071237,
+      "learning_rate": 8.609317661517413e-07,
+      "loss": 1.1628,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8711172129740862,
+      "grad_norm": 4.807454155631998,
+      "learning_rate": 8.586769007122619e-07,
+      "loss": 1.5086,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8712888278702592,
+      "grad_norm": 4.626253892623588,
+      "learning_rate": 8.564248595315616e-07,
+      "loss": 1.5777,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8714604427664321,
+      "grad_norm": 4.163467749778554,
+      "learning_rate": 8.541756433054172e-07,
+      "loss": 1.261,
+      "step": 5078
+    },
+    {
+      "epoch": 0.8716320576626051,
+      "grad_norm": 4.206738252252562,
+      "learning_rate": 8.519292527287393e-07,
+      "loss": 1.6554,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8718036725587781,
+      "grad_norm": 3.9233257823993584,
+      "learning_rate": 8.496856884955585e-07,
+      "loss": 1.3324,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8719752874549511,
+      "grad_norm": 4.352104845463383,
+      "learning_rate": 8.474449512990357e-07,
+      "loss": 1.6384,
+      "step": 5081
+    },
+    {
+      "epoch": 0.8721469023511241,
+      "grad_norm": 4.432618058077031,
+      "learning_rate": 8.452070418314584e-07,
+      "loss": 1.4949,
+      "step": 5082
+    },
+    {
+      "epoch": 0.872318517247297,
+      "grad_norm": 3.9214483831480806,
+      "learning_rate": 8.429719607842413e-07,
+      "loss": 1.397,
+      "step": 5083
+    },
+    {
+      "epoch": 0.87249013214347,
+      "grad_norm": 5.359178294980305,
+      "learning_rate": 8.40739708847923e-07,
+      "loss": 1.6092,
+      "step": 5084
+    },
+    {
+      "epoch": 0.872661747039643,
+      "grad_norm": 3.8919782825954194,
+      "learning_rate": 8.385102867121731e-07,
+      "loss": 1.3887,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8728333619358161,
+      "grad_norm": 4.68876473689935,
+      "learning_rate": 8.362836950657771e-07,
+      "loss": 1.4126,
+      "step": 5086
+    },
+    {
+      "epoch": 0.873004976831989,
+      "grad_norm": 5.023640914761555,
+      "learning_rate": 8.340599345966538e-07,
+      "loss": 1.5846,
+      "step": 5087
+    },
+    {
+      "epoch": 0.873176591728162,
+      "grad_norm": 4.336703494860674,
+      "learning_rate": 8.318390059918468e-07,
+      "loss": 1.5998,
+      "step": 5088
+    },
+    {
+      "epoch": 0.873348206624335,
+      "grad_norm": 4.295241706006463,
+      "learning_rate": 8.296209099375252e-07,
+      "loss": 1.3604,
+      "step": 5089
+    },
+    {
+      "epoch": 0.873519821520508,
+      "grad_norm": 4.055022805214807,
+      "learning_rate": 8.274056471189762e-07,
+      "loss": 1.5656,
+      "step": 5090
+    },
+    {
+      "epoch": 0.873691436416681,
+      "grad_norm": 5.265318877741098,
+      "learning_rate": 8.251932182206213e-07,
+      "loss": 1.6882,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8738630513128539,
+      "grad_norm": 4.213946294428783,
+      "learning_rate": 8.229836239259981e-07,
+      "loss": 1.3338,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8740346662090269,
+      "grad_norm": 5.298311740209048,
+      "learning_rate": 8.207768649177738e-07,
+      "loss": 1.643,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8742062811051999,
+      "grad_norm": 3.861289093619137,
+      "learning_rate": 8.185729418777388e-07,
+      "loss": 1.2493,
+      "step": 5094
+    },
+    {
+      "epoch": 0.874377896001373,
+      "grad_norm": 4.479952892249776,
+      "learning_rate": 8.16371855486805e-07,
+      "loss": 1.5119,
+      "step": 5095
+    },
+    {
+      "epoch": 0.874549510897546,
+      "grad_norm": 4.756410735988385,
+      "learning_rate": 8.141736064250094e-07,
+      "loss": 1.3881,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8747211257937189,
+      "grad_norm": 4.387309480149174,
+      "learning_rate": 8.119781953715145e-07,
+      "loss": 1.4072,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8748927406898919,
+      "grad_norm": 4.019570052494385,
+      "learning_rate": 8.097856230046009e-07,
+      "loss": 1.4999,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8750643555860649,
+      "grad_norm": 4.407039523302897,
+      "learning_rate": 8.075958900016778e-07,
+      "loss": 1.3824,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8752359704822379,
+      "grad_norm": 4.037211823213187,
+      "learning_rate": 8.05408997039272e-07,
+      "loss": 1.3855,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8754075853784108,
+      "grad_norm": 4.01127887262451,
+      "learning_rate": 8.032249447930362e-07,
+      "loss": 1.655,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8755792002745838,
+      "grad_norm": 4.017426422896338,
+      "learning_rate": 8.010437339377441e-07,
+      "loss": 1.2587,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8757508151707568,
+      "grad_norm": 4.092556992762357,
+      "learning_rate": 7.988653651472956e-07,
+      "loss": 1.4627,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8759224300669298,
+      "grad_norm": 4.914417080451019,
+      "learning_rate": 7.966898390947053e-07,
+      "loss": 1.4925,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8760940449631028,
+      "grad_norm": 3.842738580606235,
+      "learning_rate": 7.945171564521126e-07,
+      "loss": 1.3254,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8762656598592757,
+      "grad_norm": 4.958024354974342,
+      "learning_rate": 7.923473178907803e-07,
+      "loss": 1.5749,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8764372747554487,
+      "grad_norm": 4.024535446446941,
+      "learning_rate": 7.901803240810901e-07,
+      "loss": 1.2354,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8766088896516218,
+      "grad_norm": 4.744298179863797,
+      "learning_rate": 7.880161756925487e-07,
+      "loss": 1.5598,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8767805045477948,
+      "grad_norm": 4.406646055381877,
+      "learning_rate": 7.85854873393781e-07,
+      "loss": 1.5606,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8769521194439678,
+      "grad_norm": 3.889735812049366,
+      "learning_rate": 7.836964178525275e-07,
+      "loss": 1.5221,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8771237343401407,
+      "grad_norm": 4.072053408410405,
+      "learning_rate": 7.81540809735658e-07,
+      "loss": 1.4832,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8772953492363137,
+      "grad_norm": 4.653716773949536,
+      "learning_rate": 7.793880497091578e-07,
+      "loss": 1.4553,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8774669641324867,
+      "grad_norm": 4.129292446127243,
+      "learning_rate": 7.772381384381323e-07,
+      "loss": 1.3184,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8776385790286597,
+      "grad_norm": 4.147999022478401,
+      "learning_rate": 7.750910765868113e-07,
+      "loss": 1.496,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8778101939248327,
+      "grad_norm": 4.498973180218164,
+      "learning_rate": 7.729468648185379e-07,
+      "loss": 1.491,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8779818088210056,
+      "grad_norm": 3.918602235664639,
+      "learning_rate": 7.708055037957751e-07,
+      "loss": 1.5765,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8781534237171786,
+      "grad_norm": 4.0886953095453045,
+      "learning_rate": 7.686669941801106e-07,
+      "loss": 1.41,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8783250386133516,
+      "grad_norm": 4.028126965015301,
+      "learning_rate": 7.665313366322458e-07,
+      "loss": 1.4777,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8784966535095247,
+      "grad_norm": 4.591336670412107,
+      "learning_rate": 7.643985318120073e-07,
+      "loss": 1.5425,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8786682684056976,
+      "grad_norm": 4.544103967482204,
+      "learning_rate": 7.622685803783303e-07,
+      "loss": 1.5423,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8788398833018706,
+      "grad_norm": 4.347906453781017,
+      "learning_rate": 7.601414829892795e-07,
+      "loss": 1.553,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8790114981980436,
+      "grad_norm": 4.719296307510465,
+      "learning_rate": 7.580172403020291e-07,
+      "loss": 1.676,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8791831130942166,
+      "grad_norm": 4.5931446291254545,
+      "learning_rate": 7.558958529728766e-07,
+      "loss": 1.4506,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8793547279903896,
+      "grad_norm": 3.875195108008688,
+      "learning_rate": 7.537773216572331e-07,
+      "loss": 1.5131,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8795263428865625,
+      "grad_norm": 3.8027806816798533,
+      "learning_rate": 7.516616470096317e-07,
+      "loss": 1.4165,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8796979577827355,
+      "grad_norm": 4.2045088396829335,
+      "learning_rate": 7.495488296837205e-07,
+      "loss": 1.4846,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8798695726789085,
+      "grad_norm": 4.356659460779504,
+      "learning_rate": 7.474388703322677e-07,
+      "loss": 1.28,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8800411875750815,
+      "grad_norm": 3.924157987508082,
+      "learning_rate": 7.453317696071538e-07,
+      "loss": 1.5187,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8802128024712546,
+      "grad_norm": 4.2471735256036185,
+      "learning_rate": 7.43227528159376e-07,
+      "loss": 1.4178,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8803844173674275,
+      "grad_norm": 3.814645226451688,
+      "learning_rate": 7.41126146639054e-07,
+      "loss": 1.2214,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8805560322636005,
+      "grad_norm": 4.50307800406103,
+      "learning_rate": 7.390276256954188e-07,
+      "loss": 1.4051,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8807276471597735,
+      "grad_norm": 4.22926123412834,
+      "learning_rate": 7.369319659768204e-07,
+      "loss": 1.6091,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8808992620559465,
+      "grad_norm": 4.541228544104218,
+      "learning_rate": 7.348391681307254e-07,
+      "loss": 1.3364,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8810708769521195,
+      "grad_norm": 4.781747943997346,
+      "learning_rate": 7.3274923280371e-07,
+      "loss": 1.7184,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8812424918482924,
+      "grad_norm": 4.763222307588499,
+      "learning_rate": 7.306621606414721e-07,
+      "loss": 1.4953,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8814141067444654,
+      "grad_norm": 4.066786952295638,
+      "learning_rate": 7.285779522888237e-07,
+      "loss": 1.3728,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8815857216406384,
+      "grad_norm": 4.225379332033343,
+      "learning_rate": 7.264966083896918e-07,
+      "loss": 1.6345,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8817573365368114,
+      "grad_norm": 4.184227822797963,
+      "learning_rate": 7.2441812958712e-07,
+      "loss": 1.4949,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8819289514329843,
+      "grad_norm": 4.473972291437994,
+      "learning_rate": 7.223425165232645e-07,
+      "loss": 1.5525,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8821005663291573,
+      "grad_norm": 4.272178325788325,
+      "learning_rate": 7.202697698393923e-07,
+      "loss": 1.504,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8822721812253304,
+      "grad_norm": 4.532353998495571,
+      "learning_rate": 7.181998901758936e-07,
+      "loss": 1.6841,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8824437961215034,
+      "grad_norm": 4.27306585291111,
+      "learning_rate": 7.161328781722665e-07,
+      "loss": 1.5299,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8826154110176764,
+      "grad_norm": 4.360330423878417,
+      "learning_rate": 7.140687344671282e-07,
+      "loss": 1.4884,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8827870259138493,
+      "grad_norm": 3.9986389081246374,
+      "learning_rate": 7.120074596982008e-07,
+      "loss": 1.5262,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8829586408100223,
+      "grad_norm": 4.611848913050924,
+      "learning_rate": 7.099490545023313e-07,
+      "loss": 1.509,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8831302557061953,
+      "grad_norm": 4.901857884901667,
+      "learning_rate": 7.078935195154712e-07,
+      "loss": 1.4917,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8833018706023683,
+      "grad_norm": 5.805379132913955,
+      "learning_rate": 7.058408553726881e-07,
+      "loss": 1.6059,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8834734854985413,
+      "grad_norm": 6.410889962180712,
+      "learning_rate": 7.037910627081678e-07,
+      "loss": 1.5172,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8836451003947142,
+      "grad_norm": 4.732516099682142,
+      "learning_rate": 7.017441421551985e-07,
+      "loss": 1.437,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8838167152908872,
+      "grad_norm": 3.773438903238695,
+      "learning_rate": 6.997000943461895e-07,
+      "loss": 1.4421,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8839883301870602,
+      "grad_norm": 4.356289653010965,
+      "learning_rate": 6.976589199126615e-07,
+      "loss": 1.3342,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8841599450832333,
+      "grad_norm": 4.355173533103943,
+      "learning_rate": 6.956206194852422e-07,
+      "loss": 1.4601,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8843315599794063,
+      "grad_norm": 4.178629863475834,
+      "learning_rate": 6.935851936936788e-07,
+      "loss": 1.4507,
+      "step": 5153
+    },
+    {
+      "epoch": 0.8845031748755792,
+      "grad_norm": 4.336461272694183,
+      "learning_rate": 6.91552643166823e-07,
+      "loss": 1.3607,
+      "step": 5154
+    },
+    {
+      "epoch": 0.8846747897717522,
+      "grad_norm": 4.026678360353491,
+      "learning_rate": 6.895229685326443e-07,
+      "loss": 1.507,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8848464046679252,
+      "grad_norm": 5.873799972532674,
+      "learning_rate": 6.874961704182193e-07,
+      "loss": 1.7137,
+      "step": 5156
+    },
+    {
+      "epoch": 0.8850180195640982,
+      "grad_norm": 4.18488910121758,
+      "learning_rate": 6.854722494497424e-07,
+      "loss": 1.5954,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8851896344602711,
+      "grad_norm": 4.802236049468994,
+      "learning_rate": 6.834512062525067e-07,
+      "loss": 1.4101,
+      "step": 5158
+    },
+    {
+      "epoch": 0.8853612493564441,
+      "grad_norm": 4.344564222834242,
+      "learning_rate": 6.814330414509285e-07,
+      "loss": 1.5013,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8855328642526171,
+      "grad_norm": 3.8215241716912165,
+      "learning_rate": 6.794177556685288e-07,
+      "loss": 1.2992,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8857044791487901,
+      "grad_norm": 5.286115055593522,
+      "learning_rate": 6.77405349527942e-07,
+      "loss": 1.5744,
+      "step": 5161
+    },
+    {
+      "epoch": 0.8858760940449631,
+      "grad_norm": 4.690246953547851,
+      "learning_rate": 6.753958236509117e-07,
+      "loss": 1.6899,
+      "step": 5162
+    },
+    {
+      "epoch": 0.886047708941136,
+      "grad_norm": 4.507851126536749,
+      "learning_rate": 6.733891786582902e-07,
+      "loss": 1.3506,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8862193238373091,
+      "grad_norm": 3.741963759002263,
+      "learning_rate": 6.713854151700383e-07,
+      "loss": 1.3187,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8863909387334821,
+      "grad_norm": 4.108044182202845,
+      "learning_rate": 6.693845338052329e-07,
+      "loss": 1.4622,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8865625536296551,
+      "grad_norm": 4.21407652674191,
+      "learning_rate": 6.673865351820541e-07,
+      "loss": 1.3663,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8867341685258281,
+      "grad_norm": 4.250956023727706,
+      "learning_rate": 6.653914199177968e-07,
+      "loss": 1.5219,
+      "step": 5167
+    },
+    {
+      "epoch": 0.886905783422001,
+      "grad_norm": 4.554237603698288,
+      "learning_rate": 6.633991886288582e-07,
+      "loss": 1.5059,
+      "step": 5168
+    },
+    {
+      "epoch": 0.887077398318174,
+      "grad_norm": 4.695083667332852,
+      "learning_rate": 6.614098419307525e-07,
+      "loss": 1.5965,
+      "step": 5169
+    },
+    {
+      "epoch": 0.887249013214347,
+      "grad_norm": 4.8651393462264565,
+      "learning_rate": 6.594233804380945e-07,
+      "loss": 1.4644,
+      "step": 5170
+    },
+    {
+      "epoch": 0.88742062811052,
+      "grad_norm": 4.417496382127124,
+      "learning_rate": 6.574398047646146e-07,
+      "loss": 1.484,
+      "step": 5171
+    },
+    {
+      "epoch": 0.8875922430066929,
+      "grad_norm": 5.293671134677378,
+      "learning_rate": 6.554591155231482e-07,
+      "loss": 1.4422,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8877638579028659,
+      "grad_norm": 4.210464648271912,
+      "learning_rate": 6.53481313325638e-07,
+      "loss": 1.5188,
+      "step": 5173
+    },
+    {
+      "epoch": 0.887935472799039,
+      "grad_norm": 3.9454332618832777,
+      "learning_rate": 6.515063987831372e-07,
+      "loss": 1.478,
+      "step": 5174
+    },
+    {
+      "epoch": 0.888107087695212,
+      "grad_norm": 4.175447852579143,
+      "learning_rate": 6.495343725058067e-07,
+      "loss": 1.5994,
+      "step": 5175
+    },
+    {
+      "epoch": 0.888278702591385,
+      "grad_norm": 4.702515197029811,
+      "learning_rate": 6.475652351029116e-07,
+      "loss": 1.5272,
+      "step": 5176
+    },
+    {
+      "epoch": 0.8884503174875579,
+      "grad_norm": 4.255356582993296,
+      "learning_rate": 6.455989871828305e-07,
+      "loss": 1.5315,
+      "step": 5177
+    },
+    {
+      "epoch": 0.8886219323837309,
+      "grad_norm": 4.766788121069817,
+      "learning_rate": 6.436356293530433e-07,
+      "loss": 1.411,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8887935472799039,
+      "grad_norm": 4.794631815797826,
+      "learning_rate": 6.416751622201389e-07,
+      "loss": 1.5831,
+      "step": 5179
+    },
+    {
+      "epoch": 0.8889651621760769,
+      "grad_norm": 4.29972368777464,
+      "learning_rate": 6.397175863898153e-07,
+      "loss": 1.663,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8891367770722499,
+      "grad_norm": 3.7789557598739827,
+      "learning_rate": 6.377629024668774e-07,
+      "loss": 1.2915,
+      "step": 5181
+    },
+    {
+      "epoch": 0.8893083919684228,
+      "grad_norm": 3.9090569848790118,
+      "learning_rate": 6.358111110552324e-07,
+      "loss": 1.3081,
+      "step": 5182
+    },
+    {
+      "epoch": 0.8894800068645958,
+      "grad_norm": 4.519018907240786,
+      "learning_rate": 6.338622127578942e-07,
+      "loss": 1.4948,
+      "step": 5183
+    },
+    {
+      "epoch": 0.8896516217607688,
+      "grad_norm": 3.985418638820712,
+      "learning_rate": 6.31916208176988e-07,
+      "loss": 1.4789,
+      "step": 5184
+    },
+    {
+      "epoch": 0.8898232366569419,
+      "grad_norm": 4.391914961738118,
+      "learning_rate": 6.299730979137419e-07,
+      "loss": 1.3071,
+      "step": 5185
+    },
+    {
+      "epoch": 0.8899948515531149,
+      "grad_norm": 4.142436826140557,
+      "learning_rate": 6.2803288256849e-07,
+      "loss": 1.4889,
+      "step": 5186
+    },
+    {
+      "epoch": 0.8901664664492878,
+      "grad_norm": 4.299723211469974,
+      "learning_rate": 6.260955627406706e-07,
+      "loss": 1.2638,
+      "step": 5187
+    },
+    {
+      "epoch": 0.8903380813454608,
+      "grad_norm": 3.5712943736822593,
+      "learning_rate": 6.241611390288283e-07,
+      "loss": 1.3432,
+      "step": 5188
+    },
+    {
+      "epoch": 0.8905096962416338,
+      "grad_norm": 4.171748995329112,
+      "learning_rate": 6.222296120306137e-07,
+      "loss": 1.6976,
+      "step": 5189
+    },
+    {
+      "epoch": 0.8906813111378068,
+      "grad_norm": 4.138958074288165,
+      "learning_rate": 6.20300982342783e-07,
+      "loss": 1.4986,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8908529260339797,
+      "grad_norm": 4.374146664687719,
+      "learning_rate": 6.183752505611962e-07,
+      "loss": 1.501,
+      "step": 5191
+    },
+    {
+      "epoch": 0.8910245409301527,
+      "grad_norm": 4.106357547216749,
+      "learning_rate": 6.164524172808162e-07,
+      "loss": 1.5257,
+      "step": 5192
+    },
+    {
+      "epoch": 0.8911961558263257,
+      "grad_norm": 5.846407060589214,
+      "learning_rate": 6.145324830957144e-07,
+      "loss": 1.5573,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8913677707224987,
+      "grad_norm": 4.5186496629414545,
+      "learning_rate": 6.12615448599062e-07,
+      "loss": 1.6232,
+      "step": 5194
+    },
+    {
+      "epoch": 0.8915393856186717,
+      "grad_norm": 5.0965856982743505,
+      "learning_rate": 6.107013143831375e-07,
+      "loss": 1.4859,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8917110005148446,
+      "grad_norm": 4.744328446437492,
+      "learning_rate": 6.087900810393254e-07,
+      "loss": 1.688,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8918826154110177,
+      "grad_norm": 4.909702260880713,
+      "learning_rate": 6.068817491581069e-07,
+      "loss": 1.4724,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8920542303071907,
+      "grad_norm": 3.7809838046607274,
+      "learning_rate": 6.049763193290725e-07,
+      "loss": 1.3656,
+      "step": 5198
+    },
+    {
+      "epoch": 0.8922258452033637,
+      "grad_norm": 4.387571464690832,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 1.4564,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8923974600995367,
+      "grad_norm": 4.215459866078354,
+      "learning_rate": 6.011741681814309e-07,
+      "loss": 1.4627,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8925690749957096,
+      "grad_norm": 4.912869724481879,
+      "learning_rate": 5.992774480375185e-07,
+      "loss": 1.4891,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8927406898918826,
+      "grad_norm": 4.463581542784137,
+      "learning_rate": 5.973836322951765e-07,
+      "loss": 1.4826,
+      "step": 5202
+    },
+    {
+      "epoch": 0.8929123047880556,
+      "grad_norm": 3.6769585976356787,
+      "learning_rate": 5.954927215395101e-07,
+      "loss": 1.2443,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8930839196842286,
+      "grad_norm": 4.895161395407363,
+      "learning_rate": 5.936047163547276e-07,
+      "loss": 1.4917,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8932555345804016,
+      "grad_norm": 4.040043100111145,
+      "learning_rate": 5.91719617324138e-07,
+      "loss": 1.585,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8934271494765745,
+      "grad_norm": 4.247969886257527,
+      "learning_rate": 5.898374250301508e-07,
+      "loss": 1.5476,
+      "step": 5206
+    },
+    {
+      "epoch": 0.8935987643727475,
+      "grad_norm": 4.298199607245423,
+      "learning_rate": 5.879581400542788e-07,
+      "loss": 1.4175,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8937703792689206,
+      "grad_norm": 4.224879587969809,
+      "learning_rate": 5.860817629771376e-07,
+      "loss": 1.5978,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8939419941650936,
+      "grad_norm": 4.886504545556072,
+      "learning_rate": 5.842082943784433e-07,
+      "loss": 1.5513,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8941136090612665,
+      "grad_norm": 4.787234713050884,
+      "learning_rate": 5.823377348370152e-07,
+      "loss": 1.737,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8942852239574395,
+      "grad_norm": 4.526442662219888,
+      "learning_rate": 5.804700849307699e-07,
+      "loss": 1.3806,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8944568388536125,
+      "grad_norm": 4.204401915207515,
+      "learning_rate": 5.78605345236728e-07,
+      "loss": 1.451,
+      "step": 5212
+    },
+    {
+      "epoch": 0.8946284537497855,
+      "grad_norm": 3.7932028280801005,
+      "learning_rate": 5.767435163310109e-07,
+      "loss": 1.45,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8948000686459585,
+      "grad_norm": 4.813596301989631,
+      "learning_rate": 5.748845987888396e-07,
+      "loss": 1.5563,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8949716835421314,
+      "grad_norm": 4.221783949309949,
+      "learning_rate": 5.730285931845381e-07,
+      "loss": 1.5451,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8951432984383044,
+      "grad_norm": 4.509655846209165,
+      "learning_rate": 5.711755000915265e-07,
+      "loss": 1.4448,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8953149133344774,
+      "grad_norm": 4.428722973126899,
+      "learning_rate": 5.693253200823301e-07,
+      "loss": 1.4418,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8954865282306504,
+      "grad_norm": 4.4419654505271,
+      "learning_rate": 5.674780537285673e-07,
+      "loss": 1.5026,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8956581431268235,
+      "grad_norm": 4.513905473356673,
+      "learning_rate": 5.65633701600965e-07,
+      "loss": 1.59,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8958297580229964,
+      "grad_norm": 3.910233224502691,
+      "learning_rate": 5.637922642693439e-07,
+      "loss": 1.47,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8960013729191694,
+      "grad_norm": 4.49935798951609,
+      "learning_rate": 5.619537423026234e-07,
+      "loss": 1.5129,
+      "step": 5221
+    },
+    {
+      "epoch": 0.8961729878153424,
+      "grad_norm": 4.038674296468667,
+      "learning_rate": 5.601181362688257e-07,
+      "loss": 1.438,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8963446027115154,
+      "grad_norm": 3.7292276019146233,
+      "learning_rate": 5.582854467350729e-07,
+      "loss": 1.5849,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8965162176076883,
+      "grad_norm": 4.3061995042493,
+      "learning_rate": 5.564556742675808e-07,
+      "loss": 1.715,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8966878325038613,
+      "grad_norm": 4.128244697435798,
+      "learning_rate": 5.546288194316684e-07,
+      "loss": 1.4726,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8968594474000343,
+      "grad_norm": 4.671448654437281,
+      "learning_rate": 5.52804882791752e-07,
+      "loss": 1.5278,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8970310622962073,
+      "grad_norm": 3.9520327928409844,
+      "learning_rate": 5.509838649113442e-07,
+      "loss": 1.2045,
+      "step": 5227
+    },
+    {
+      "epoch": 0.8972026771923803,
+      "grad_norm": 4.03523079885433,
+      "learning_rate": 5.491657663530603e-07,
+      "loss": 1.5759,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8973742920885532,
+      "grad_norm": 4.513092002116728,
+      "learning_rate": 5.473505876786123e-07,
+      "loss": 1.6297,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8975459069847262,
+      "grad_norm": 4.988141289982714,
+      "learning_rate": 5.455383294488059e-07,
+      "loss": 1.4124,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8977175218808993,
+      "grad_norm": 4.800390771221265,
+      "learning_rate": 5.437289922235489e-07,
+      "loss": 1.5722,
+      "step": 5231
+    },
+    {
+      "epoch": 0.8978891367770723,
+      "grad_norm": 4.938742163744131,
+      "learning_rate": 5.419225765618442e-07,
+      "loss": 1.4789,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8980607516732453,
+      "grad_norm": 4.325577003534197,
+      "learning_rate": 5.40119083021794e-07,
+      "loss": 1.4945,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8982323665694182,
+      "grad_norm": 4.55228196129667,
+      "learning_rate": 5.383185121605994e-07,
+      "loss": 1.3432,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8984039814655912,
+      "grad_norm": 4.47766500065033,
+      "learning_rate": 5.365208645345532e-07,
+      "loss": 1.4544,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8985755963617642,
+      "grad_norm": 4.5823923539412394,
+      "learning_rate": 5.347261406990468e-07,
+      "loss": 1.3966,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8987472112579372,
+      "grad_norm": 4.487230443599707,
+      "learning_rate": 5.329343412085708e-07,
+      "loss": 1.4559,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8989188261541102,
+      "grad_norm": 5.201942990297392,
+      "learning_rate": 5.311454666167115e-07,
+      "loss": 1.6059,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8990904410502831,
+      "grad_norm": 3.556085086748516,
+      "learning_rate": 5.293595174761523e-07,
+      "loss": 1.224,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8992620559464561,
+      "grad_norm": 4.705590547794815,
+      "learning_rate": 5.275764943386674e-07,
+      "loss": 1.5347,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8994336708426292,
+      "grad_norm": 3.975371850358197,
+      "learning_rate": 5.25796397755135e-07,
+      "loss": 1.4857,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8996052857388022,
+      "grad_norm": 5.261206540896355,
+      "learning_rate": 5.240192282755219e-07,
+      "loss": 1.4698,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8997769006349751,
+      "grad_norm": 4.311722203810764,
+      "learning_rate": 5.222449864488943e-07,
+      "loss": 1.6503,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8999485155311481,
+      "grad_norm": 4.327256117526736,
+      "learning_rate": 5.204736728234161e-07,
+      "loss": 1.5411,
+      "step": 5244
+    },
+    {
+      "epoch": 0.9001201304273211,
+      "grad_norm": 3.8397599977089083,
+      "learning_rate": 5.187052879463394e-07,
+      "loss": 1.453,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9002917453234941,
+      "grad_norm": 4.755520237202437,
+      "learning_rate": 5.169398323640196e-07,
+      "loss": 1.4264,
+      "step": 5246
+    },
+    {
+      "epoch": 0.9004633602196671,
+      "grad_norm": 4.460485017045077,
+      "learning_rate": 5.151773066219024e-07,
+      "loss": 1.4779,
+      "step": 5247
+    },
+    {
+      "epoch": 0.90063497511584,
+      "grad_norm": 4.646715853530595,
+      "learning_rate": 5.134177112645267e-07,
+      "loss": 1.4989,
+      "step": 5248
+    },
+    {
+      "epoch": 0.900806590012013,
+      "grad_norm": 4.008537925585396,
+      "learning_rate": 5.116610468355332e-07,
+      "loss": 1.4514,
+      "step": 5249
+    },
+    {
+      "epoch": 0.900978204908186,
+      "grad_norm": 3.887643337022098,
+      "learning_rate": 5.099073138776467e-07,
+      "loss": 1.3845,
+      "step": 5250
+    },
+    {
+      "epoch": 0.901149819804359,
+      "grad_norm": 3.9385917428235864,
+      "learning_rate": 5.08156512932696e-07,
+      "loss": 1.5364,
+      "step": 5251
+    },
+    {
+      "epoch": 0.901321434700532,
+      "grad_norm": 4.058467145465141,
+      "learning_rate": 5.064086445415983e-07,
+      "loss": 1.5497,
+      "step": 5252
+    },
+    {
+      "epoch": 0.901493049596705,
+      "grad_norm": 5.143232568164296,
+      "learning_rate": 5.046637092443696e-07,
+      "loss": 1.5667,
+      "step": 5253
+    },
+    {
+      "epoch": 0.901664664492878,
+      "grad_norm": 3.969519446132455,
+      "learning_rate": 5.029217075801141e-07,
+      "loss": 1.2721,
+      "step": 5254
+    },
+    {
+      "epoch": 0.901836279389051,
+      "grad_norm": 4.403318344872112,
+      "learning_rate": 5.011826400870301e-07,
+      "loss": 1.5335,
+      "step": 5255
+    },
+    {
+      "epoch": 0.902007894285224,
+      "grad_norm": 4.9982191788793475,
+      "learning_rate": 4.994465073024147e-07,
+      "loss": 1.8223,
+      "step": 5256
+    },
+    {
+      "epoch": 0.902179509181397,
+      "grad_norm": 3.776378028347021,
+      "learning_rate": 4.97713309762653e-07,
+      "loss": 1.4572,
+      "step": 5257
+    },
+    {
+      "epoch": 0.9023511240775699,
+      "grad_norm": 4.455696113057125,
+      "learning_rate": 4.95983048003228e-07,
+      "loss": 1.4217,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9025227389737429,
+      "grad_norm": 4.165256101939233,
+      "learning_rate": 4.942557225587119e-07,
+      "loss": 1.5347,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9026943538699159,
+      "grad_norm": 6.578797851625758,
+      "learning_rate": 4.925313339627679e-07,
+      "loss": 1.5379,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9028659687660889,
+      "grad_norm": 3.8680726495336217,
+      "learning_rate": 4.908098827481578e-07,
+      "loss": 1.479,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9030375836622618,
+      "grad_norm": 4.545699649524925,
+      "learning_rate": 4.890913694467325e-07,
+      "loss": 1.4346,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9032091985584348,
+      "grad_norm": 4.181696160904031,
+      "learning_rate": 4.873757945894352e-07,
+      "loss": 1.3432,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9033808134546079,
+      "grad_norm": 4.582981631392994,
+      "learning_rate": 4.85663158706301e-07,
+      "loss": 1.6212,
+      "step": 5264
+    },
+    {
+      "epoch": 0.9035524283507809,
+      "grad_norm": 4.8575917294020625,
+      "learning_rate": 4.839534623264586e-07,
+      "loss": 1.3309,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9037240432469539,
+      "grad_norm": 5.494250578514054,
+      "learning_rate": 4.822467059781255e-07,
+      "loss": 1.6245,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9038956581431268,
+      "grad_norm": 4.672265268035886,
+      "learning_rate": 4.805428901886145e-07,
+      "loss": 1.6497,
+      "step": 5267
+    },
+    {
+      "epoch": 0.9040672730392998,
+      "grad_norm": 5.219870244895374,
+      "learning_rate": 4.788420154843287e-07,
+      "loss": 1.5394,
+      "step": 5268
+    },
+    {
+      "epoch": 0.9042388879354728,
+      "grad_norm": 5.382312411966969,
+      "learning_rate": 4.771440823907603e-07,
+      "loss": 1.5024,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9044105028316458,
+      "grad_norm": 4.363802487685458,
+      "learning_rate": 4.7544909143249605e-07,
+      "loss": 1.2821,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9045821177278188,
+      "grad_norm": 4.038441773227084,
+      "learning_rate": 4.737570431332128e-07,
+      "loss": 1.2983,
+      "step": 5271
+    },
+    {
+      "epoch": 0.9047537326239917,
+      "grad_norm": 4.3903952681436476,
+      "learning_rate": 4.720679380156745e-07,
+      "loss": 1.6009,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9049253475201647,
+      "grad_norm": 4.1774739609620335,
+      "learning_rate": 4.703817766017427e-07,
+      "loss": 1.4655,
+      "step": 5273
+    },
+    {
+      "epoch": 0.9050969624163377,
+      "grad_norm": 4.788134222975619,
+      "learning_rate": 4.6869855941236165e-07,
+      "loss": 1.4647,
+      "step": 5274
+    },
+    {
+      "epoch": 0.9052685773125108,
+      "grad_norm": 4.117806313593401,
+      "learning_rate": 4.6701828696757213e-07,
+      "loss": 1.68,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9054401922086837,
+      "grad_norm": 3.643260914188979,
+      "learning_rate": 4.653409597865033e-07,
+      "loss": 1.4958,
+      "step": 5276
+    },
+    {
+      "epoch": 0.9056118071048567,
+      "grad_norm": 5.456473938336759,
+      "learning_rate": 4.6366657838737394e-07,
+      "loss": 1.4086,
+      "step": 5277
+    },
+    {
+      "epoch": 0.9057834220010297,
+      "grad_norm": 4.485563589001977,
+      "learning_rate": 4.619951432874936e-07,
+      "loss": 1.1558,
+      "step": 5278
+    },
+    {
+      "epoch": 0.9059550368972027,
+      "grad_norm": 3.949412510101276,
+      "learning_rate": 4.6032665500325704e-07,
+      "loss": 1.5094,
+      "step": 5279
+    },
+    {
+      "epoch": 0.9061266517933757,
+      "grad_norm": 5.767455139586388,
+      "learning_rate": 4.586611140501551e-07,
+      "loss": 1.4061,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9062982666895486,
+      "grad_norm": 4.065568408551128,
+      "learning_rate": 4.569985209427652e-07,
+      "loss": 1.4209,
+      "step": 5281
+    },
+    {
+      "epoch": 0.9064698815857216,
+      "grad_norm": 4.178978782132392,
+      "learning_rate": 4.5533887619475415e-07,
+      "loss": 1.3886,
+      "step": 5282
+    },
+    {
+      "epoch": 0.9066414964818946,
+      "grad_norm": 5.199621160985476,
+      "learning_rate": 4.5368218031887735e-07,
+      "loss": 1.4529,
+      "step": 5283
+    },
+    {
+      "epoch": 0.9068131113780676,
+      "grad_norm": 5.037264222852291,
+      "learning_rate": 4.520284338269787e-07,
+      "loss": 1.5575,
+      "step": 5284
+    },
+    {
+      "epoch": 0.9069847262742406,
+      "grad_norm": 3.7275594194059742,
+      "learning_rate": 4.5037763722999284e-07,
+      "loss": 1.4773,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9071563411704135,
+      "grad_norm": 4.42312004462506,
+      "learning_rate": 4.487297910379407e-07,
+      "loss": 1.6172,
+      "step": 5286
+    },
+    {
+      "epoch": 0.9073279560665866,
+      "grad_norm": 4.201406125086892,
+      "learning_rate": 4.4708489575993496e-07,
+      "loss": 1.3437,
+      "step": 5287
+    },
+    {
+      "epoch": 0.9074995709627596,
+      "grad_norm": 3.998668600183201,
+      "learning_rate": 4.4544295190417253e-07,
+      "loss": 1.5552,
+      "step": 5288
+    },
+    {
+      "epoch": 0.9076711858589326,
+      "grad_norm": 3.961995186279598,
+      "learning_rate": 4.438039599779409e-07,
+      "loss": 1.4371,
+      "step": 5289
+    },
+    {
+      "epoch": 0.9078428007551056,
+      "grad_norm": 4.128514877165511,
+      "learning_rate": 4.42167920487615e-07,
+      "loss": 1.6085,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9080144156512785,
+      "grad_norm": 5.047524015574739,
+      "learning_rate": 4.4053483393865724e-07,
+      "loss": 1.7038,
+      "step": 5291
+    },
+    {
+      "epoch": 0.9081860305474515,
+      "grad_norm": 4.673752678909029,
+      "learning_rate": 4.3890470083562066e-07,
+      "loss": 1.5429,
+      "step": 5292
+    },
+    {
+      "epoch": 0.9083576454436245,
+      "grad_norm": 4.77742482137629,
+      "learning_rate": 4.37277521682139e-07,
+      "loss": 1.6082,
+      "step": 5293
+    },
+    {
+      "epoch": 0.9085292603397975,
+      "grad_norm": 4.630445791702161,
+      "learning_rate": 4.356532969809402e-07,
+      "loss": 1.2921,
+      "step": 5294
+    },
+    {
+      "epoch": 0.9087008752359704,
+      "grad_norm": 4.139412099054128,
+      "learning_rate": 4.340320272338372e-07,
+      "loss": 1.5239,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9088724901321434,
+      "grad_norm": 3.9404021236159936,
+      "learning_rate": 4.324137129417283e-07,
+      "loss": 1.5066,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9090441050283165,
+      "grad_norm": 3.8511515844541666,
+      "learning_rate": 4.307983546046024e-07,
+      "loss": 1.5601,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9092157199244895,
+      "grad_norm": 4.318761606690782,
+      "learning_rate": 4.2918595272152916e-07,
+      "loss": 1.4133,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9093873348206625,
+      "grad_norm": 5.163062131663407,
+      "learning_rate": 4.2757650779067016e-07,
+      "loss": 1.7038,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9095589497168354,
+      "grad_norm": 5.281425754772876,
+      "learning_rate": 4.25970020309272e-07,
+      "loss": 1.5196,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9097305646130084,
+      "grad_norm": 4.202807507081207,
+      "learning_rate": 4.243664907736689e-07,
+      "loss": 1.3688,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9099021795091814,
+      "grad_norm": 4.136558213734609,
+      "learning_rate": 4.2276591967927663e-07,
+      "loss": 1.1983,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9100737944053544,
+      "grad_norm": 4.1092169348474,
+      "learning_rate": 4.211683075206008e-07,
+      "loss": 1.6398,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9102454093015274,
+      "grad_norm": 5.159421862381322,
+      "learning_rate": 4.19573654791231e-07,
+      "loss": 1.6633,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9104170241977003,
+      "grad_norm": 4.620241065129417,
+      "learning_rate": 4.1798196198384545e-07,
+      "loss": 1.4688,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9105886390938733,
+      "grad_norm": 5.464614299595415,
+      "learning_rate": 4.163932295902051e-07,
+      "loss": 1.4655,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9107602539900463,
+      "grad_norm": 3.795028293352429,
+      "learning_rate": 4.148074581011574e-07,
+      "loss": 1.2971,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9109318688862194,
+      "grad_norm": 4.27588811343906,
+      "learning_rate": 4.1322464800663265e-07,
+      "loss": 1.2765,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9111034837823924,
+      "grad_norm": 4.919279668780527,
+      "learning_rate": 4.1164479979564853e-07,
+      "loss": 1.7586,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9112750986785653,
+      "grad_norm": 4.65101060675833,
+      "learning_rate": 4.1006791395630906e-07,
+      "loss": 1.4632,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9114467135747383,
+      "grad_norm": 4.701162057334624,
+      "learning_rate": 4.084939909758012e-07,
+      "loss": 1.5103,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9116183284709113,
+      "grad_norm": 4.360328841283706,
+      "learning_rate": 4.069230313403949e-07,
+      "loss": 1.5297,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9117899433670843,
+      "grad_norm": 4.638740863989734,
+      "learning_rate": 4.053550355354485e-07,
+      "loss": 1.7828,
+      "step": 5313
+    },
+    {
+      "epoch": 0.9119615582632572,
+      "grad_norm": 4.159565266083416,
+      "learning_rate": 4.03790004045399e-07,
+      "loss": 1.4288,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9121331731594302,
+      "grad_norm": 4.232778824601497,
+      "learning_rate": 4.0222793735377295e-07,
+      "loss": 1.6523,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9123047880556032,
+      "grad_norm": 5.449907376759737,
+      "learning_rate": 4.006688359431798e-07,
+      "loss": 1.4393,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9124764029517762,
+      "grad_norm": 4.8068634680623425,
+      "learning_rate": 3.991127002953099e-07,
+      "loss": 1.521,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9126480178479492,
+      "grad_norm": 4.981338601725879,
+      "learning_rate": 3.975595308909397e-07,
+      "loss": 1.3967,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9128196327441221,
+      "grad_norm": 5.441131859753169,
+      "learning_rate": 3.96009328209932e-07,
+      "loss": 1.5284,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9129912476402952,
+      "grad_norm": 4.659763391867403,
+      "learning_rate": 3.944620927312248e-07,
+      "loss": 1.5035,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9131628625364682,
+      "grad_norm": 4.077267955695744,
+      "learning_rate": 3.9291782493284913e-07,
+      "loss": 1.2405,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9133344774326412,
+      "grad_norm": 3.778531105247793,
+      "learning_rate": 3.9137652529191105e-07,
+      "loss": 1.3553,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9135060923288142,
+      "grad_norm": 4.519886256364349,
+      "learning_rate": 3.8983819428460414e-07,
+      "loss": 1.472,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9136777072249871,
+      "grad_norm": 5.036479945569544,
+      "learning_rate": 3.883028323862048e-07,
+      "loss": 1.6447,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9138493221211601,
+      "grad_norm": 4.143650999507501,
+      "learning_rate": 3.867704400710703e-07,
+      "loss": 1.457,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9140209370173331,
+      "grad_norm": 4.151242116647198,
+      "learning_rate": 3.852410178126409e-07,
+      "loss": 1.5368,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9141925519135061,
+      "grad_norm": 4.8733356611955765,
+      "learning_rate": 3.837145660834385e-07,
+      "loss": 1.6504,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9143641668096791,
+      "grad_norm": 4.6171287558702705,
+      "learning_rate": 3.821910853550692e-07,
+      "loss": 1.68,
+      "step": 5328
+    },
+    {
+      "epoch": 0.914535781705852,
+      "grad_norm": 4.195227763871497,
+      "learning_rate": 3.80670576098221e-07,
+      "loss": 1.4752,
+      "step": 5329
+    },
+    {
+      "epoch": 0.914707396602025,
+      "grad_norm": 3.906091652760167,
+      "learning_rate": 3.7915303878266253e-07,
+      "loss": 1.2803,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9148790114981981,
+      "grad_norm": 4.182155883621883,
+      "learning_rate": 3.776384738772443e-07,
+      "loss": 1.4749,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9150506263943711,
+      "grad_norm": 4.718364924466981,
+      "learning_rate": 3.7612688184989865e-07,
+      "loss": 1.5893,
+      "step": 5332
+    },
+    {
+      "epoch": 0.915222241290544,
+      "grad_norm": 4.555852311365258,
+      "learning_rate": 3.7461826316763984e-07,
+      "loss": 1.4063,
+      "step": 5333
+    },
+    {
+      "epoch": 0.915393856186717,
+      "grad_norm": 4.757136065648205,
+      "learning_rate": 3.7311261829656276e-07,
+      "loss": 1.4902,
+      "step": 5334
+    },
+    {
+      "epoch": 0.91556547108289,
+      "grad_norm": 4.914420316871826,
+      "learning_rate": 3.716099477018475e-07,
+      "loss": 1.6006,
+      "step": 5335
+    },
+    {
+      "epoch": 0.915737085979063,
+      "grad_norm": 5.553957609853261,
+      "learning_rate": 3.701102518477473e-07,
+      "loss": 1.547,
+      "step": 5336
+    },
+    {
+      "epoch": 0.915908700875236,
+      "grad_norm": 4.256268050121203,
+      "learning_rate": 3.686135311976047e-07,
+      "loss": 1.612,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9160803157714089,
+      "grad_norm": 4.245408389763915,
+      "learning_rate": 3.6711978621383647e-07,
+      "loss": 1.5116,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9162519306675819,
+      "grad_norm": 5.081687820531221,
+      "learning_rate": 3.6562901735794464e-07,
+      "loss": 1.7774,
+      "step": 5339
+    },
+    {
+      "epoch": 0.9164235455637549,
+      "grad_norm": 5.424587713412453,
+      "learning_rate": 3.6414122509050963e-07,
+      "loss": 1.7561,
+      "step": 5340
+    },
+    {
+      "epoch": 0.916595160459928,
+      "grad_norm": 4.0159394527488,
+      "learning_rate": 3.626564098711904e-07,
+      "loss": 1.2071,
+      "step": 5341
+    },
+    {
+      "epoch": 0.916766775356101,
+      "grad_norm": 3.653732629353631,
+      "learning_rate": 3.6117457215873007e-07,
+      "loss": 1.31,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9169383902522739,
+      "grad_norm": 4.417356671719247,
+      "learning_rate": 3.5969571241095125e-07,
+      "loss": 1.491,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9171100051484469,
+      "grad_norm": 4.205177223904458,
+      "learning_rate": 3.582198310847529e-07,
+      "loss": 1.3586,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9172816200446199,
+      "grad_norm": 4.9544216874469615,
+      "learning_rate": 3.56746928636118e-07,
+      "loss": 1.3357,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9174532349407929,
+      "grad_norm": 4.545971443192365,
+      "learning_rate": 3.55277005520106e-07,
+      "loss": 1.4425,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9176248498369658,
+      "grad_norm": 4.722418176836132,
+      "learning_rate": 3.538100621908569e-07,
+      "loss": 1.3921,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9177964647331388,
+      "grad_norm": 4.057783399430007,
+      "learning_rate": 3.523460991015915e-07,
+      "loss": 1.6308,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9179680796293118,
+      "grad_norm": 5.273019770925544,
+      "learning_rate": 3.50885116704609e-07,
+      "loss": 1.7232,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9181396945254848,
+      "grad_norm": 4.42627069754146,
+      "learning_rate": 3.494271154512874e-07,
+      "loss": 1.5878,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9183113094216578,
+      "grad_norm": 4.342110447707852,
+      "learning_rate": 3.4797209579208067e-07,
+      "loss": 1.5794,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9184829243178307,
+      "grad_norm": 5.160650884382741,
+      "learning_rate": 3.4652005817652824e-07,
+      "loss": 1.434,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9186545392140038,
+      "grad_norm": 4.672146926715929,
+      "learning_rate": 3.4507100305324337e-07,
+      "loss": 1.447,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9188261541101768,
+      "grad_norm": 4.647768137125765,
+      "learning_rate": 3.436249308699202e-07,
+      "loss": 1.4473,
+      "step": 5354
+    },
+    {
+      "epoch": 0.9189977690063498,
+      "grad_norm": 4.3358851893594474,
+      "learning_rate": 3.4218184207332804e-07,
+      "loss": 1.4792,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9191693839025228,
+      "grad_norm": 4.0404804565843975,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 1.5571,
+      "step": 5356
+    },
+    {
+      "epoch": 0.9193409987986957,
+      "grad_norm": 3.917131038925715,
+      "learning_rate": 3.3930461642281644e-07,
+      "loss": 1.4437,
+      "step": 5357
+    },
+    {
+      "epoch": 0.9195126136948687,
+      "grad_norm": 5.9250294479410925,
+      "learning_rate": 3.3787048045783146e-07,
+      "loss": 1.6586,
+      "step": 5358
+    },
+    {
+      "epoch": 0.9196842285910417,
+      "grad_norm": 5.32193496516879,
+      "learning_rate": 3.364393296574453e-07,
+      "loss": 1.6385,
+      "step": 5359
+    },
+    {
+      "epoch": 0.9198558434872147,
+      "grad_norm": 3.6713072214805345,
+      "learning_rate": 3.3501116446381875e-07,
+      "loss": 1.4173,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9200274583833877,
+      "grad_norm": 4.325247897984546,
+      "learning_rate": 3.335859853181922e-07,
+      "loss": 1.594,
+      "step": 5361
+    },
+    {
+      "epoch": 0.9201990732795606,
+      "grad_norm": 3.9556601346905746,
+      "learning_rate": 3.321637926608823e-07,
+      "loss": 1.362,
+      "step": 5362
+    },
+    {
+      "epoch": 0.9203706881757336,
+      "grad_norm": 4.0808374972855646,
+      "learning_rate": 3.307445869312809e-07,
+      "loss": 1.4276,
+      "step": 5363
+    },
+    {
+      "epoch": 0.9205423030719067,
+      "grad_norm": 4.428345514591795,
+      "learning_rate": 3.2932836856786164e-07,
+      "loss": 1.3601,
+      "step": 5364
+    },
+    {
+      "epoch": 0.9207139179680797,
+      "grad_norm": 5.326500874434499,
+      "learning_rate": 3.279151380081691e-07,
+      "loss": 1.5781,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9208855328642526,
+      "grad_norm": 4.57421907352443,
+      "learning_rate": 3.265048956888306e-07,
+      "loss": 1.4663,
+      "step": 5366
+    },
+    {
+      "epoch": 0.9210571477604256,
+      "grad_norm": 4.613812049849862,
+      "learning_rate": 3.250976420455465e-07,
+      "loss": 1.6557,
+      "step": 5367
+    },
+    {
+      "epoch": 0.9212287626565986,
+      "grad_norm": 4.430156567713831,
+      "learning_rate": 3.236933775130968e-07,
+      "loss": 1.7358,
+      "step": 5368
+    },
+    {
+      "epoch": 0.9214003775527716,
+      "grad_norm": 5.04977908171309,
+      "learning_rate": 3.222921025253356e-07,
+      "loss": 1.4869,
+      "step": 5369
+    },
+    {
+      "epoch": 0.9215719924489446,
+      "grad_norm": 5.115593825488799,
+      "learning_rate": 3.20893817515191e-07,
+      "loss": 1.5958,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9217436073451175,
+      "grad_norm": 4.088029261173349,
+      "learning_rate": 3.194985229146741e-07,
+      "loss": 1.2138,
+      "step": 5371
+    },
+    {
+      "epoch": 0.9219152222412905,
+      "grad_norm": 4.5759428070142505,
+      "learning_rate": 3.181062191548667e-07,
+      "loss": 1.2562,
+      "step": 5372
+    },
+    {
+      "epoch": 0.9220868371374635,
+      "grad_norm": 5.254876753182029,
+      "learning_rate": 3.1671690666592925e-07,
+      "loss": 1.3768,
+      "step": 5373
+    },
+    {
+      "epoch": 0.9222584520336365,
+      "grad_norm": 4.775672443241848,
+      "learning_rate": 3.1533058587709606e-07,
+      "loss": 1.4375,
+      "step": 5374
+    },
+    {
+      "epoch": 0.9224300669298096,
+      "grad_norm": 4.162721060436665,
+      "learning_rate": 3.1394725721667794e-07,
+      "loss": 1.2617,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9226016818259825,
+      "grad_norm": 5.306961707635675,
+      "learning_rate": 3.125669211120619e-07,
+      "loss": 1.6401,
+      "step": 5376
+    },
+    {
+      "epoch": 0.9227732967221555,
+      "grad_norm": 4.104382892199951,
+      "learning_rate": 3.1118957798970895e-07,
+      "loss": 1.2813,
+      "step": 5377
+    },
+    {
+      "epoch": 0.9229449116183285,
+      "grad_norm": 3.978223757181011,
+      "learning_rate": 3.0981522827515876e-07,
+      "loss": 1.4656,
+      "step": 5378
+    },
+    {
+      "epoch": 0.9231165265145015,
+      "grad_norm": 5.180532339256917,
+      "learning_rate": 3.084438723930205e-07,
+      "loss": 1.4536,
+      "step": 5379
+    },
+    {
+      "epoch": 0.9232881414106745,
+      "grad_norm": 5.061494779717264,
+      "learning_rate": 3.070755107669843e-07,
+      "loss": 1.5193,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9234597563068474,
+      "grad_norm": 4.647210995365146,
+      "learning_rate": 3.057101438198107e-07,
+      "loss": 1.6764,
+      "step": 5381
+    },
+    {
+      "epoch": 0.9236313712030204,
+      "grad_norm": 4.596035808100414,
+      "learning_rate": 3.0434777197333785e-07,
+      "loss": 1.5186,
+      "step": 5382
+    },
+    {
+      "epoch": 0.9238029860991934,
+      "grad_norm": 4.246198579622432,
+      "learning_rate": 3.0298839564847804e-07,
+      "loss": 1.3931,
+      "step": 5383
+    },
+    {
+      "epoch": 0.9239746009953664,
+      "grad_norm": 5.472122145728372,
+      "learning_rate": 3.016320152652152e-07,
+      "loss": 1.4979,
+      "step": 5384
+    },
+    {
+      "epoch": 0.9241462158915393,
+      "grad_norm": 4.483257198830271,
+      "learning_rate": 3.0027863124261204e-07,
+      "loss": 1.6708,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9243178307877123,
+      "grad_norm": 4.283443055866963,
+      "learning_rate": 2.989282439988039e-07,
+      "loss": 1.5622,
+      "step": 5386
+    },
+    {
+      "epoch": 0.9244894456838854,
+      "grad_norm": 5.108144941988228,
+      "learning_rate": 2.975808539509983e-07,
+      "loss": 1.4247,
+      "step": 5387
+    },
+    {
+      "epoch": 0.9246610605800584,
+      "grad_norm": 4.9080051075921345,
+      "learning_rate": 2.9623646151548004e-07,
+      "loss": 1.5395,
+      "step": 5388
+    },
+    {
+      "epoch": 0.9248326754762314,
+      "grad_norm": 3.9609765253317093,
+      "learning_rate": 2.948950671076034e-07,
+      "loss": 1.292,
+      "step": 5389
+    },
+    {
+      "epoch": 0.9250042903724043,
+      "grad_norm": 4.077791367965805,
+      "learning_rate": 2.935566711418014e-07,
+      "loss": 1.46,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9251759052685773,
+      "grad_norm": 5.31914432062151,
+      "learning_rate": 2.9222127403157773e-07,
+      "loss": 1.5185,
+      "step": 5391
+    },
+    {
+      "epoch": 0.9253475201647503,
+      "grad_norm": 4.6330553464781685,
+      "learning_rate": 2.908888761895112e-07,
+      "loss": 1.518,
+      "step": 5392
+    },
+    {
+      "epoch": 0.9255191350609233,
+      "grad_norm": 4.125501160404574,
+      "learning_rate": 2.895594780272515e-07,
+      "loss": 1.4487,
+      "step": 5393
+    },
+    {
+      "epoch": 0.9256907499570963,
+      "grad_norm": 5.050516248080808,
+      "learning_rate": 2.882330799555233e-07,
+      "loss": 1.4311,
+      "step": 5394
+    },
+    {
+      "epoch": 0.9258623648532692,
+      "grad_norm": 4.276476782200002,
+      "learning_rate": 2.8690968238412444e-07,
+      "loss": 1.3596,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9260339797494422,
+      "grad_norm": 4.455226772166471,
+      "learning_rate": 2.8558928572192444e-07,
+      "loss": 1.5735,
+      "step": 5396
+    },
+    {
+      "epoch": 0.9262055946456152,
+      "grad_norm": 4.452040685222897,
+      "learning_rate": 2.8427189037686933e-07,
+      "loss": 1.4951,
+      "step": 5397
+    },
+    {
+      "epoch": 0.9263772095417883,
+      "grad_norm": 4.883370990093987,
+      "learning_rate": 2.829574967559723e-07,
+      "loss": 1.5208,
+      "step": 5398
+    },
+    {
+      "epoch": 0.9265488244379612,
+      "grad_norm": 4.10642429487041,
+      "learning_rate": 2.81646105265323e-07,
+      "loss": 1.3949,
+      "step": 5399
+    },
+    {
+      "epoch": 0.9267204393341342,
+      "grad_norm": 5.00072667775669,
+      "learning_rate": 2.803377163100818e-07,
+      "loss": 1.5265,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9268920542303072,
+      "grad_norm": 4.845865399519903,
+      "learning_rate": 2.790323302944831e-07,
+      "loss": 1.4133,
+      "step": 5401
+    },
+    {
+      "epoch": 0.9270636691264802,
+      "grad_norm": 4.690431091503764,
+      "learning_rate": 2.7772994762183426e-07,
+      "loss": 1.5244,
+      "step": 5402
+    },
+    {
+      "epoch": 0.9272352840226532,
+      "grad_norm": 5.091401768899432,
+      "learning_rate": 2.7643056869451015e-07,
+      "loss": 1.3736,
+      "step": 5403
+    },
+    {
+      "epoch": 0.9274068989188261,
+      "grad_norm": 4.591526642457146,
+      "learning_rate": 2.751341939139618e-07,
+      "loss": 1.5156,
+      "step": 5404
+    },
+    {
+      "epoch": 0.9275785138149991,
+      "grad_norm": 5.37448574464333,
+      "learning_rate": 2.738408236807111e-07,
+      "loss": 1.5098,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9277501287111721,
+      "grad_norm": 4.2669076798826,
+      "learning_rate": 2.7255045839435054e-07,
+      "loss": 1.2476,
+      "step": 5406
+    },
+    {
+      "epoch": 0.9279217436073451,
+      "grad_norm": 4.102832677100814,
+      "learning_rate": 2.7126309845354694e-07,
+      "loss": 1.4285,
+      "step": 5407
+    },
+    {
+      "epoch": 0.9280933585035182,
+      "grad_norm": 5.35685891960044,
+      "learning_rate": 2.699787442560342e-07,
+      "loss": 1.6172,
+      "step": 5408
+    },
+    {
+      "epoch": 0.928264973399691,
+      "grad_norm": 5.3607508756824505,
+      "learning_rate": 2.6869739619862165e-07,
+      "loss": 1.505,
+      "step": 5409
+    },
+    {
+      "epoch": 0.9284365882958641,
+      "grad_norm": 3.655402870444697,
+      "learning_rate": 2.674190546771904e-07,
+      "loss": 1.3444,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9286082031920371,
+      "grad_norm": 4.080997386538315,
+      "learning_rate": 2.6614372008668675e-07,
+      "loss": 1.1752,
+      "step": 5411
+    },
+    {
+      "epoch": 0.9287798180882101,
+      "grad_norm": 5.681595602476748,
+      "learning_rate": 2.6487139282113437e-07,
+      "loss": 1.515,
+      "step": 5412
+    },
+    {
+      "epoch": 0.9289514329843831,
+      "grad_norm": 4.230696391350242,
+      "learning_rate": 2.636020732736233e-07,
+      "loss": 1.5766,
+      "step": 5413
+    },
+    {
+      "epoch": 0.929123047880556,
+      "grad_norm": 4.444666272055504,
+      "learning_rate": 2.623357618363176e-07,
+      "loss": 1.4772,
+      "step": 5414
+    },
+    {
+      "epoch": 0.929294662776729,
+      "grad_norm": 4.577090789386312,
+      "learning_rate": 2.6107245890045205e-07,
+      "loss": 1.627,
+      "step": 5415
+    },
+    {
+      "epoch": 0.929466277672902,
+      "grad_norm": 4.813376168898806,
+      "learning_rate": 2.59812164856329e-07,
+      "loss": 1.4489,
+      "step": 5416
+    },
+    {
+      "epoch": 0.929637892569075,
+      "grad_norm": 4.468312323542875,
+      "learning_rate": 2.585548800933235e-07,
+      "loss": 1.3745,
+      "step": 5417
+    },
+    {
+      "epoch": 0.9298095074652479,
+      "grad_norm": 4.115224502173242,
+      "learning_rate": 2.573006049998783e-07,
+      "loss": 1.4818,
+      "step": 5418
+    },
+    {
+      "epoch": 0.9299811223614209,
+      "grad_norm": 5.082592295099144,
+      "learning_rate": 2.5604933996351e-07,
+      "loss": 1.6077,
+      "step": 5419
+    },
+    {
+      "epoch": 0.930152737257594,
+      "grad_norm": 5.3530434976445544,
+      "learning_rate": 2.5480108537080275e-07,
+      "loss": 1.3564,
+      "step": 5420
+    },
+    {
+      "epoch": 0.930324352153767,
+      "grad_norm": 4.207748830053649,
+      "learning_rate": 2.535558416074113e-07,
+      "loss": 1.3965,
+      "step": 5421
+    },
+    {
+      "epoch": 0.93049596704994,
+      "grad_norm": 4.110903228790358,
+      "learning_rate": 2.5231360905806133e-07,
+      "loss": 1.3848,
+      "step": 5422
+    },
+    {
+      "epoch": 0.9306675819461129,
+      "grad_norm": 4.131267606119521,
+      "learning_rate": 2.510743881065447e-07,
+      "loss": 1.3226,
+      "step": 5423
+    },
+    {
+      "epoch": 0.9308391968422859,
+      "grad_norm": 5.018272101149,
+      "learning_rate": 2.4983817913572626e-07,
+      "loss": 1.4884,
+      "step": 5424
+    },
+    {
+      "epoch": 0.9310108117384589,
+      "grad_norm": 3.957374956760122,
+      "learning_rate": 2.4860498252753827e-07,
+      "loss": 1.5885,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9311824266346319,
+      "grad_norm": 4.209945118333955,
+      "learning_rate": 2.47374798662986e-07,
+      "loss": 1.262,
+      "step": 5426
+    },
+    {
+      "epoch": 0.9313540415308049,
+      "grad_norm": 4.851479119584087,
+      "learning_rate": 2.4614762792213754e-07,
+      "loss": 1.5034,
+      "step": 5427
+    },
+    {
+      "epoch": 0.9315256564269778,
+      "grad_norm": 4.737621251186431,
+      "learning_rate": 2.449234706841364e-07,
+      "loss": 1.7132,
+      "step": 5428
+    },
+    {
+      "epoch": 0.9316972713231508,
+      "grad_norm": 4.717165212880252,
+      "learning_rate": 2.437023273271888e-07,
+      "loss": 1.5048,
+      "step": 5429
+    },
+    {
+      "epoch": 0.9318688862193238,
+      "grad_norm": 4.793579595824607,
+      "learning_rate": 2.4248419822857636e-07,
+      "loss": 1.4585,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9320405011154969,
+      "grad_norm": 4.803817184445834,
+      "learning_rate": 2.41269083764647e-07,
+      "loss": 1.3786,
+      "step": 5431
+    },
+    {
+      "epoch": 0.9322121160116699,
+      "grad_norm": 4.582613285977406,
+      "learning_rate": 2.4005698431081356e-07,
+      "loss": 1.4374,
+      "step": 5432
+    },
+    {
+      "epoch": 0.9323837309078428,
+      "grad_norm": 5.148065210219013,
+      "learning_rate": 2.3884790024156115e-07,
+      "loss": 1.8179,
+      "step": 5433
+    },
+    {
+      "epoch": 0.9325553458040158,
+      "grad_norm": 4.4222829713928515,
+      "learning_rate": 2.376418319304441e-07,
+      "loss": 1.432,
+      "step": 5434
+    },
+    {
+      "epoch": 0.9327269607001888,
+      "grad_norm": 4.866943439553488,
+      "learning_rate": 2.3643877975008223e-07,
+      "loss": 1.551,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9328985755963618,
+      "grad_norm": 4.775044992338535,
+      "learning_rate": 2.3523874407216597e-07,
+      "loss": 1.448,
+      "step": 5436
+    },
+    {
+      "epoch": 0.9330701904925347,
+      "grad_norm": 4.14479706167544,
+      "learning_rate": 2.3404172526744984e-07,
+      "loss": 1.3905,
+      "step": 5437
+    },
+    {
+      "epoch": 0.9332418053887077,
+      "grad_norm": 5.441804555677087,
+      "learning_rate": 2.3284772370576137e-07,
+      "loss": 1.7822,
+      "step": 5438
+    },
+    {
+      "epoch": 0.9334134202848807,
+      "grad_norm": 4.423596700926595,
+      "learning_rate": 2.3165673975599213e-07,
+      "loss": 1.4915,
+      "step": 5439
+    },
+    {
+      "epoch": 0.9335850351810537,
+      "grad_norm": 4.150493018626314,
+      "learning_rate": 2.3046877378610443e-07,
+      "loss": 1.5582,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9337566500772267,
+      "grad_norm": 5.125956238059363,
+      "learning_rate": 2.2928382616312582e-07,
+      "loss": 1.5916,
+      "step": 5441
+    },
+    {
+      "epoch": 0.9339282649733996,
+      "grad_norm": 4.124233777280232,
+      "learning_rate": 2.2810189725315013e-07,
+      "loss": 1.4833,
+      "step": 5442
+    },
+    {
+      "epoch": 0.9340998798695727,
+      "grad_norm": 4.686028271545141,
+      "learning_rate": 2.2692298742134188e-07,
+      "loss": 1.5007,
+      "step": 5443
+    },
+    {
+      "epoch": 0.9342714947657457,
+      "grad_norm": 3.4058910946900793,
+      "learning_rate": 2.2574709703193086e-07,
+      "loss": 1.1171,
+      "step": 5444
+    },
+    {
+      "epoch": 0.9344431096619187,
+      "grad_norm": 4.879661854689699,
+      "learning_rate": 2.2457422644821535e-07,
+      "loss": 1.488,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9346147245580917,
+      "grad_norm": 4.140283141104058,
+      "learning_rate": 2.2340437603255994e-07,
+      "loss": 1.5075,
+      "step": 5446
+    },
+    {
+      "epoch": 0.9347863394542646,
+      "grad_norm": 4.383123977705686,
+      "learning_rate": 2.2223754614639437e-07,
+      "loss": 1.4522,
+      "step": 5447
+    },
+    {
+      "epoch": 0.9349579543504376,
+      "grad_norm": 5.885875304656546,
+      "learning_rate": 2.2107373715021696e-07,
+      "loss": 1.5742,
+      "step": 5448
+    },
+    {
+      "epoch": 0.9351295692466106,
+      "grad_norm": 4.944101652425936,
+      "learning_rate": 2.1991294940359343e-07,
+      "loss": 1.4143,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9353011841427836,
+      "grad_norm": 3.99872897697736,
+      "learning_rate": 2.1875518326515578e-07,
+      "loss": 1.274,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9354727990389565,
+      "grad_norm": 4.651661534779984,
+      "learning_rate": 2.176004390926001e-07,
+      "loss": 1.4185,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9356444139351295,
+      "grad_norm": 3.781247208678973,
+      "learning_rate": 2.1644871724269102e-07,
+      "loss": 1.4458,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9358160288313025,
+      "grad_norm": 4.1797280880035785,
+      "learning_rate": 2.1530001807125944e-07,
+      "loss": 1.5128,
+      "step": 5453
+    },
+    {
+      "epoch": 0.9359876437274756,
+      "grad_norm": 5.149269990553209,
+      "learning_rate": 2.1415434193320038e-07,
+      "loss": 1.4525,
+      "step": 5454
+    },
+    {
+      "epoch": 0.9361592586236486,
+      "grad_norm": 5.188102171780974,
+      "learning_rate": 2.130116891824796e-07,
+      "loss": 1.5899,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9363308735198215,
+      "grad_norm": 4.982838307095838,
+      "learning_rate": 2.1187206017212248e-07,
+      "loss": 1.687,
+      "step": 5456
+    },
+    {
+      "epoch": 0.9365024884159945,
+      "grad_norm": 4.694262359670236,
+      "learning_rate": 2.1073545525422402e-07,
+      "loss": 1.5309,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9366741033121675,
+      "grad_norm": 4.012266792074211,
+      "learning_rate": 2.0960187477994444e-07,
+      "loss": 1.5667,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9368457182083405,
+      "grad_norm": 4.6351374309440025,
+      "learning_rate": 2.0847131909950912e-07,
+      "loss": 1.5457,
+      "step": 5459
+    },
+    {
+      "epoch": 0.9370173331045135,
+      "grad_norm": 4.049863812346382,
+      "learning_rate": 2.0734378856221092e-07,
+      "loss": 1.1781,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9371889480006864,
+      "grad_norm": 4.120135750766048,
+      "learning_rate": 2.0621928351640231e-07,
+      "loss": 1.4823,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9373605628968594,
+      "grad_norm": 3.8214226303817167,
+      "learning_rate": 2.0509780430950754e-07,
+      "loss": 1.3756,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9375321777930324,
+      "grad_norm": 4.287005650149692,
+      "learning_rate": 2.0397935128801283e-07,
+      "loss": 1.5436,
+      "step": 5463
+    },
+    {
+      "epoch": 0.9377037926892055,
+      "grad_norm": 4.251647074771913,
+      "learning_rate": 2.0286392479747064e-07,
+      "loss": 1.4238,
+      "step": 5464
+    },
+    {
+      "epoch": 0.9378754075853785,
+      "grad_norm": 4.279691404522448,
+      "learning_rate": 2.0175152518249753e-07,
+      "loss": 1.6744,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9380470224815514,
+      "grad_norm": 5.126031706890018,
+      "learning_rate": 2.0064215278677412e-07,
+      "loss": 1.6748,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9382186373777244,
+      "grad_norm": 4.303546507230215,
+      "learning_rate": 1.9953580795304628e-07,
+      "loss": 1.4509,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9383902522738974,
+      "grad_norm": 4.23722789041591,
+      "learning_rate": 1.9843249102312722e-07,
+      "loss": 1.289,
+      "step": 5468
+    },
+    {
+      "epoch": 0.9385618671700704,
+      "grad_norm": 4.465369397607201,
+      "learning_rate": 1.9733220233789097e-07,
+      "loss": 1.486,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9387334820662433,
+      "grad_norm": 4.067478812139229,
+      "learning_rate": 1.962349422372778e-07,
+      "loss": 1.3555,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9389050969624163,
+      "grad_norm": 3.9450981044114175,
+      "learning_rate": 1.9514071106028987e-07,
+      "loss": 1.3921,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9390767118585893,
+      "grad_norm": 4.62484541639502,
+      "learning_rate": 1.9404950914499787e-07,
+      "loss": 1.485,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9392483267547623,
+      "grad_norm": 4.970890379453228,
+      "learning_rate": 1.9296133682853436e-07,
+      "loss": 1.5941,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9394199416509353,
+      "grad_norm": 4.729506984296073,
+      "learning_rate": 1.9187619444709483e-07,
+      "loss": 1.4279,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9395915565471082,
+      "grad_norm": 4.426615000149878,
+      "learning_rate": 1.9079408233593998e-07,
+      "loss": 1.6161,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9397631714432813,
+      "grad_norm": 5.170515051255126,
+      "learning_rate": 1.8971500082939354e-07,
+      "loss": 1.5884,
+      "step": 5476
+    },
+    {
+      "epoch": 0.9399347863394543,
+      "grad_norm": 4.415585548610362,
+      "learning_rate": 1.886389502608421e-07,
+      "loss": 1.7015,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9401064012356273,
+      "grad_norm": 4.501590145025204,
+      "learning_rate": 1.8756593096273978e-07,
+      "loss": 1.7003,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9402780161318003,
+      "grad_norm": 4.49601989494037,
+      "learning_rate": 1.8649594326660024e-07,
+      "loss": 1.6124,
+      "step": 5479
+    },
+    {
+      "epoch": 0.9404496310279732,
+      "grad_norm": 5.03603978096081,
+      "learning_rate": 1.854289875030002e-07,
+      "loss": 1.4852,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9406212459241462,
+      "grad_norm": 4.046174708184559,
+      "learning_rate": 1.843650640015826e-07,
+      "loss": 1.5849,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9407928608203192,
+      "grad_norm": 4.255537810701056,
+      "learning_rate": 1.8330417309105342e-07,
+      "loss": 1.6783,
+      "step": 5482
+    },
+    {
+      "epoch": 0.9409644757164922,
+      "grad_norm": 4.075111191752301,
+      "learning_rate": 1.822463150991771e-07,
+      "loss": 1.1585,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9411360906126652,
+      "grad_norm": 5.953991043181289,
+      "learning_rate": 1.811914903527867e-07,
+      "loss": 1.5888,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9413077055088381,
+      "grad_norm": 5.279935510968091,
+      "learning_rate": 1.8013969917777484e-07,
+      "loss": 1.6853,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9414793204050111,
+      "grad_norm": 4.382044619510773,
+      "learning_rate": 1.7909094189909826e-07,
+      "loss": 1.5727,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9416509353011842,
+      "grad_norm": 6.203155445137067,
+      "learning_rate": 1.780452188407744e-07,
+      "loss": 1.7352,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9418225501973572,
+      "grad_norm": 4.7427327154515035,
+      "learning_rate": 1.770025303258882e-07,
+      "loss": 1.4214,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9419941650935301,
+      "grad_norm": 4.256842350675999,
+      "learning_rate": 1.7596287667658085e-07,
+      "loss": 1.5036,
+      "step": 5489
+    },
+    {
+      "epoch": 0.9421657799897031,
+      "grad_norm": 4.83648651138836,
+      "learning_rate": 1.7492625821405872e-07,
+      "loss": 1.6256,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9423373948858761,
+      "grad_norm": 4.53033515886486,
+      "learning_rate": 1.738926752585901e-07,
+      "loss": 1.659,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9425090097820491,
+      "grad_norm": 4.174373651973854,
+      "learning_rate": 1.7286212812950842e-07,
+      "loss": 1.3821,
+      "step": 5492
+    },
+    {
+      "epoch": 0.9426806246782221,
+      "grad_norm": 4.920264499057055,
+      "learning_rate": 1.7183461714520455e-07,
+      "loss": 1.6867,
+      "step": 5493
+    },
+    {
+      "epoch": 0.942852239574395,
+      "grad_norm": 4.242615109063735,
+      "learning_rate": 1.7081014262313344e-07,
+      "loss": 1.5198,
+      "step": 5494
+    },
+    {
+      "epoch": 0.943023854470568,
+      "grad_norm": 4.1052912146059395,
+      "learning_rate": 1.6978870487981082e-07,
+      "loss": 1.4706,
+      "step": 5495
+    },
+    {
+      "epoch": 0.943195469366741,
+      "grad_norm": 4.467767061948136,
+      "learning_rate": 1.687703042308164e-07,
+      "loss": 1.5312,
+      "step": 5496
+    },
+    {
+      "epoch": 0.943367084262914,
+      "grad_norm": 4.689116423577103,
+      "learning_rate": 1.6775494099078971e-07,
+      "loss": 1.4992,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9435386991590871,
+      "grad_norm": 4.1132320005324985,
+      "learning_rate": 1.6674261547343417e-07,
+      "loss": 1.5359,
+      "step": 5498
+    },
+    {
+      "epoch": 0.94371031405526,
+      "grad_norm": 5.082596245702644,
+      "learning_rate": 1.6573332799151076e-07,
+      "loss": 1.6278,
+      "step": 5499
+    },
+    {
+      "epoch": 0.943881928951433,
+      "grad_norm": 4.250666476463317,
+      "learning_rate": 1.6472707885684557e-07,
+      "loss": 1.322,
+      "step": 5500
+    },
+    {
+      "epoch": 0.944053543847606,
+      "grad_norm": 4.965076369426824,
+      "learning_rate": 1.6372386838032218e-07,
+      "loss": 1.7252,
+      "step": 5501
+    },
+    {
+      "epoch": 0.944225158743779,
+      "grad_norm": 4.612750041092019,
+      "learning_rate": 1.6272369687189038e-07,
+      "loss": 1.3538,
+      "step": 5502
+    },
+    {
+      "epoch": 0.944396773639952,
+      "grad_norm": 4.575129781882723,
+      "learning_rate": 1.6172656464055748e-07,
+      "loss": 1.469,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9445683885361249,
+      "grad_norm": 3.933610133607644,
+      "learning_rate": 1.6073247199439034e-07,
+      "loss": 1.4181,
+      "step": 5504
+    },
+    {
+      "epoch": 0.9447400034322979,
+      "grad_norm": 4.0560079355778855,
+      "learning_rate": 1.5974141924052222e-07,
+      "loss": 1.5681,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9449116183284709,
+      "grad_norm": 4.814534286449412,
+      "learning_rate": 1.5875340668514262e-07,
+      "loss": 1.3591,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9450832332246439,
+      "grad_norm": 4.002163378366633,
+      "learning_rate": 1.5776843463350288e-07,
+      "loss": 1.3633,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9452548481208168,
+      "grad_norm": 4.906710382709269,
+      "learning_rate": 1.567865033899163e-07,
+      "loss": 1.4755,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9454264630169898,
+      "grad_norm": 4.595955671730814,
+      "learning_rate": 1.5580761325775352e-07,
+      "loss": 1.786,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9455980779131629,
+      "grad_norm": 4.884715002591531,
+      "learning_rate": 1.5483176453944814e-07,
+      "loss": 1.5741,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9457696928093359,
+      "grad_norm": 4.660137512620219,
+      "learning_rate": 1.538589575364946e-07,
+      "loss": 1.7246,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9459413077055089,
+      "grad_norm": 4.914873791354064,
+      "learning_rate": 1.52889192549448e-07,
+      "loss": 1.5279,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9461129226016818,
+      "grad_norm": 5.12123560290714,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 1.7243,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9462845374978548,
+      "grad_norm": 4.709705897755593,
+      "learning_rate": 1.5095878982058442e-07,
+      "loss": 1.4513,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9464561523940278,
+      "grad_norm": 5.175152832964303,
+      "learning_rate": 1.4999815267517593e-07,
+      "loss": 1.4411,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9466277672902008,
+      "grad_norm": 4.409390626975295,
+      "learning_rate": 1.49040558738488e-07,
+      "loss": 1.3232,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9467993821863738,
+      "grad_norm": 4.354459610064533,
+      "learning_rate": 1.480860083063762e-07,
+      "loss": 1.4545,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9469709970825467,
+      "grad_norm": 3.986037892439437,
+      "learning_rate": 1.4713450167375022e-07,
+      "loss": 1.4302,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9471426119787197,
+      "grad_norm": 5.039051237389449,
+      "learning_rate": 1.4618603913458596e-07,
+      "loss": 1.3696,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9473142268748928,
+      "grad_norm": 4.902122930097657,
+      "learning_rate": 1.452406209819135e-07,
+      "loss": 1.3077,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9474858417710658,
+      "grad_norm": 5.904590933452497,
+      "learning_rate": 1.4429824750782583e-07,
+      "loss": 1.4765,
+      "step": 5521
+    },
+    {
+      "epoch": 0.9476574566672387,
+      "grad_norm": 4.994552490517991,
+      "learning_rate": 1.4335891900347453e-07,
+      "loss": 1.6367,
+      "step": 5522
+    },
+    {
+      "epoch": 0.9478290715634117,
+      "grad_norm": 4.248039818352286,
+      "learning_rate": 1.4242263575906966e-07,
+      "loss": 1.4552,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9480006864595847,
+      "grad_norm": 3.926366421870658,
+      "learning_rate": 1.4148939806387986e-07,
+      "loss": 1.3853,
+      "step": 5524
+    },
+    {
+      "epoch": 0.9481723013557577,
+      "grad_norm": 3.8921543269617787,
+      "learning_rate": 1.4055920620623443e-07,
+      "loss": 1.311,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9483439162519307,
+      "grad_norm": 4.511433098985666,
+      "learning_rate": 1.396320604735213e-07,
+      "loss": 1.5943,
+      "step": 5526
+    },
+    {
+      "epoch": 0.9485155311481036,
+      "grad_norm": 4.532233086564658,
+      "learning_rate": 1.3870796115218688e-07,
+      "loss": 1.6102,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9486871460442766,
+      "grad_norm": 5.046622565248214,
+      "learning_rate": 1.37786908527735e-07,
+      "loss": 1.3627,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9488587609404496,
+      "grad_norm": 4.5000477370654055,
+      "learning_rate": 1.368689028847303e-07,
+      "loss": 1.4448,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9490303758366226,
+      "grad_norm": 4.221493859098195,
+      "learning_rate": 1.35953944506797e-07,
+      "loss": 1.4925,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9492019907327957,
+      "grad_norm": 4.451180414540635,
+      "learning_rate": 1.350420336766134e-07,
+      "loss": 1.5551,
+      "step": 5531
+    },
+    {
+      "epoch": 0.9493736056289686,
+      "grad_norm": 4.108782222822078,
+      "learning_rate": 1.341331706759219e-07,
+      "loss": 1.5612,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9495452205251416,
+      "grad_norm": 4.725777889577813,
+      "learning_rate": 1.3322735578551794e-07,
+      "loss": 1.4624,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9497168354213146,
+      "grad_norm": 4.398872234583014,
+      "learning_rate": 1.323245892852587e-07,
+      "loss": 1.6568,
+      "step": 5534
+    },
+    {
+      "epoch": 0.9498884503174876,
+      "grad_norm": 4.415741481311043,
+      "learning_rate": 1.314248714540589e-07,
+      "loss": 1.6578,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9500600652136606,
+      "grad_norm": 4.281254580853789,
+      "learning_rate": 1.3052820256989174e-07,
+      "loss": 1.4232,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9502316801098335,
+      "grad_norm": 5.0435488226525225,
+      "learning_rate": 1.296345829097856e-07,
+      "loss": 1.5888,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9504032950060065,
+      "grad_norm": 5.46652166659214,
+      "learning_rate": 1.287440127498296e-07,
+      "loss": 1.5877,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9505749099021795,
+      "grad_norm": 4.513994269781327,
+      "learning_rate": 1.278564923651704e-07,
+      "loss": 1.5309,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9507465247983525,
+      "grad_norm": 4.309112779653492,
+      "learning_rate": 1.2697202203001192e-07,
+      "loss": 1.4586,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9509181396945254,
+      "grad_norm": 5.343333296569606,
+      "learning_rate": 1.2609060201761558e-07,
+      "loss": 1.4457,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9510897545906984,
+      "grad_norm": 4.118652672759147,
+      "learning_rate": 1.2521223260030135e-07,
+      "loss": 1.4706,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9512613694868715,
+      "grad_norm": 5.6021818516241515,
+      "learning_rate": 1.2433691404944548e-07,
+      "loss": 1.5353,
+      "step": 5543
+    },
+    {
+      "epoch": 0.9514329843830445,
+      "grad_norm": 4.450252790102887,
+      "learning_rate": 1.2346464663548053e-07,
+      "loss": 1.132,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9516045992792175,
+      "grad_norm": 4.1302892229467005,
+      "learning_rate": 1.225954306279009e-07,
+      "loss": 1.3012,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9517762141753904,
+      "grad_norm": 4.73965807784184,
+      "learning_rate": 1.2172926629525406e-07,
+      "loss": 1.5539,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9519478290715634,
+      "grad_norm": 4.197637706686702,
+      "learning_rate": 1.2086615390514477e-07,
+      "loss": 1.4117,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9521194439677364,
+      "grad_norm": 5.52844818643221,
+      "learning_rate": 1.2000609372423865e-07,
+      "loss": 1.5033,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9522910588639094,
+      "grad_norm": 4.259515764839677,
+      "learning_rate": 1.1914908601825204e-07,
+      "loss": 1.4461,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9524626737600824,
+      "grad_norm": 4.379129780854141,
+      "learning_rate": 1.1829513105196533e-07,
+      "loss": 1.4342,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9526342886562553,
+      "grad_norm": 4.061496755627444,
+      "learning_rate": 1.1744422908921193e-07,
+      "loss": 1.3807,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9528059035524283,
+      "grad_norm": 4.107654935189843,
+      "learning_rate": 1.1659638039287935e-07,
+      "loss": 1.483,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9529775184486013,
+      "grad_norm": 4.20693366150938,
+      "learning_rate": 1.1575158522491803e-07,
+      "loss": 1.2848,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9531491333447744,
+      "grad_norm": 4.39066981228261,
+      "learning_rate": 1.1490984384633142e-07,
+      "loss": 1.4661,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9533207482409474,
+      "grad_norm": 4.774239545298379,
+      "learning_rate": 1.1407115651717704e-07,
+      "loss": 1.4656,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9534923631371203,
+      "grad_norm": 4.513545789220338,
+      "learning_rate": 1.1323552349657541e-07,
+      "loss": 1.4191,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9536639780332933,
+      "grad_norm": 4.401696750785899,
+      "learning_rate": 1.1240294504269777e-07,
+      "loss": 1.4805,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9538355929294663,
+      "grad_norm": 4.419973212132363,
+      "learning_rate": 1.1157342141277283e-07,
+      "loss": 1.3783,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9540072078256393,
+      "grad_norm": 4.046858009648834,
+      "learning_rate": 1.1074695286308667e-07,
+      "loss": 1.4337,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9541788227218122,
+      "grad_norm": 4.88405537694404,
+      "learning_rate": 1.0992353964898283e-07,
+      "loss": 1.6405,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9543504376179852,
+      "grad_norm": 4.513552996548548,
+      "learning_rate": 1.0910318202485782e-07,
+      "loss": 1.3132,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9545220525141582,
+      "grad_norm": 4.689991520897372,
+      "learning_rate": 1.082858802441633e-07,
+      "loss": 1.5871,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9546936674103312,
+      "grad_norm": 3.711974722691003,
+      "learning_rate": 1.0747163455941179e-07,
+      "loss": 1.444,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9548652823065042,
+      "grad_norm": 5.398649179804981,
+      "learning_rate": 1.0666044522216645e-07,
+      "loss": 1.7823,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9550368972026771,
+      "grad_norm": 4.637181411813782,
+      "learning_rate": 1.0585231248305017e-07,
+      "loss": 1.4848,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9552085120988502,
+      "grad_norm": 4.921065162458169,
+      "learning_rate": 1.0504723659173987e-07,
+      "loss": 1.5463,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9553801269950232,
+      "grad_norm": 4.675508501946252,
+      "learning_rate": 1.0424521779696661e-07,
+      "loss": 1.6674,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9555517418911962,
+      "grad_norm": 4.166551009202157,
+      "learning_rate": 1.0344625634651773e-07,
+      "loss": 1.4283,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9557233567873692,
+      "grad_norm": 4.762150315402959,
+      "learning_rate": 1.0265035248723798e-07,
+      "loss": 1.3742,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9558949716835421,
+      "grad_norm": 4.368923435623675,
+      "learning_rate": 1.0185750646502623e-07,
+      "loss": 1.5163,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9560665865797151,
+      "grad_norm": 4.254634715659754,
+      "learning_rate": 1.0106771852483432e-07,
+      "loss": 1.409,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9562382014758881,
+      "grad_norm": 5.6605593576952415,
+      "learning_rate": 1.0028098891067262e-07,
+      "loss": 1.5272,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9564098163720611,
+      "grad_norm": 4.271190562672141,
+      "learning_rate": 9.949731786560557e-08,
+      "loss": 1.2537,
+      "step": 5573
+    },
+    {
+      "epoch": 0.956581431268234,
+      "grad_norm": 4.388912580337789,
+      "learning_rate": 9.871670563175173e-08,
+      "loss": 1.4486,
+      "step": 5574
+    },
+    {
+      "epoch": 0.956753046164407,
+      "grad_norm": 6.610827677646732,
+      "learning_rate": 9.793915245028595e-08,
+      "loss": 1.6385,
+      "step": 5575
+    },
+    {
+      "epoch": 0.95692466106058,
+      "grad_norm": 3.908251639364815,
+      "learning_rate": 9.716465856143608e-08,
+      "loss": 1.4899,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9570962759567531,
+      "grad_norm": 4.1041587282985885,
+      "learning_rate": 9.639322420448627e-08,
+      "loss": 1.4772,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9572678908529261,
+      "grad_norm": 4.103523520494298,
+      "learning_rate": 9.562484961777696e-08,
+      "loss": 1.3929,
+      "step": 5578
+    },
+    {
+      "epoch": 0.957439505749099,
+      "grad_norm": 5.03495639522032,
+      "learning_rate": 9.485953503869827e-08,
+      "loss": 1.4996,
+      "step": 5579
+    },
+    {
+      "epoch": 0.957611120645272,
+      "grad_norm": 4.3567043272124195,
+      "learning_rate": 9.409728070370217e-08,
+      "loss": 1.648,
+      "step": 5580
+    },
+    {
+      "epoch": 0.957782735541445,
+      "grad_norm": 4.989616679241597,
+      "learning_rate": 9.333808684828693e-08,
+      "loss": 1.4442,
+      "step": 5581
+    },
+    {
+      "epoch": 0.957954350437618,
+      "grad_norm": 4.236021356788596,
+      "learning_rate": 9.258195370701161e-08,
+      "loss": 1.5483,
+      "step": 5582
+    },
+    {
+      "epoch": 0.958125965333791,
+      "grad_norm": 4.267173572456094,
+      "learning_rate": 9.182888151348712e-08,
+      "loss": 1.6046,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9582975802299639,
+      "grad_norm": 4.245959539165511,
+      "learning_rate": 9.107887050037845e-08,
+      "loss": 1.448,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9584691951261369,
+      "grad_norm": 4.770992351179536,
+      "learning_rate": 9.03319208994069e-08,
+      "loss": 1.3618,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9586408100223099,
+      "grad_norm": 7.281904780279344,
+      "learning_rate": 8.958803294134455e-08,
+      "loss": 1.611,
+      "step": 5586
+    },
+    {
+      "epoch": 0.958812424918483,
+      "grad_norm": 3.938487005137157,
+      "learning_rate": 8.884720685601978e-08,
+      "loss": 1.4523,
+      "step": 5587
+    },
+    {
+      "epoch": 0.958984039814656,
+      "grad_norm": 4.3884354989483185,
+      "learning_rate": 8.810944287231504e-08,
+      "loss": 1.4561,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9591556547108289,
+      "grad_norm": 4.033949352532372,
+      "learning_rate": 8.737474121816691e-08,
+      "loss": 1.5197,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9593272696070019,
+      "grad_norm": 3.9625866466914257,
+      "learning_rate": 8.664310212056493e-08,
+      "loss": 1.2878,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9594988845031749,
+      "grad_norm": 3.845998593217139,
+      "learning_rate": 8.59145258055516e-08,
+      "loss": 1.2817,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9596704993993479,
+      "grad_norm": 4.619156431167044,
+      "learning_rate": 8.518901249822576e-08,
+      "loss": 1.5144,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9598421142955208,
+      "grad_norm": 4.182687252249632,
+      "learning_rate": 8.446656242273699e-08,
+      "loss": 1.4119,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9600137291916938,
+      "grad_norm": 4.933676338033103,
+      "learning_rate": 8.374717580229119e-08,
+      "loss": 1.4332,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9601853440878668,
+      "grad_norm": 4.323055042475596,
+      "learning_rate": 8.303085285914614e-08,
+      "loss": 1.6665,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9603569589840398,
+      "grad_norm": 6.3717991565874295,
+      "learning_rate": 8.231759381461368e-08,
+      "loss": 1.5652,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9605285738802128,
+      "grad_norm": 4.441330403516568,
+      "learning_rate": 8.160739888905867e-08,
+      "loss": 1.7358,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9607001887763857,
+      "grad_norm": 4.640337077955984,
+      "learning_rate": 8.090026830190001e-08,
+      "loss": 1.4833,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9608718036725588,
+      "grad_norm": 3.9971018482858973,
+      "learning_rate": 8.019620227160852e-08,
+      "loss": 1.2332,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9610434185687318,
+      "grad_norm": 4.572150917235713,
+      "learning_rate": 7.949520101570907e-08,
+      "loss": 1.6356,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9612150334649048,
+      "grad_norm": 4.053242958053849,
+      "learning_rate": 7.879726475077953e-08,
+      "loss": 1.2053,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9613866483610778,
+      "grad_norm": 4.269274464377621,
+      "learning_rate": 7.810239369245187e-08,
+      "loss": 1.5149,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9615582632572507,
+      "grad_norm": 5.222369099471507,
+      "learning_rate": 7.741058805540991e-08,
+      "loss": 1.5011,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9617298781534237,
+      "grad_norm": 4.14591109332786,
+      "learning_rate": 7.672184805338934e-08,
+      "loss": 1.4887,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9619014930495967,
+      "grad_norm": 5.256019734077203,
+      "learning_rate": 7.603617389918106e-08,
+      "loss": 1.3953,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9620731079457697,
+      "grad_norm": 4.798890066271091,
+      "learning_rate": 7.535356580462672e-08,
+      "loss": 1.699,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9622447228419427,
+      "grad_norm": 4.562986119179497,
+      "learning_rate": 7.467402398062206e-08,
+      "loss": 1.5486,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9624163377381156,
+      "grad_norm": 4.378363309374198,
+      "learning_rate": 7.399754863711583e-08,
+      "loss": 1.4836,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9625879526342886,
+      "grad_norm": 4.2700541625191635,
+      "learning_rate": 7.332413998310861e-08,
+      "loss": 1.4473,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9627595675304617,
+      "grad_norm": 4.870812737206464,
+      "learning_rate": 7.265379822665064e-08,
+      "loss": 1.5347,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9629311824266347,
+      "grad_norm": 4.932673098818039,
+      "learning_rate": 7.19865235748507e-08,
+      "loss": 1.5467,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9631027973228076,
+      "grad_norm": 4.32960987656462,
+      "learning_rate": 7.132231623386498e-08,
+      "loss": 1.5214,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9632744122189806,
+      "grad_norm": 4.5411362193102045,
+      "learning_rate": 7.066117640890491e-08,
+      "loss": 1.423,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9634460271151536,
+      "grad_norm": 4.390075950257135,
+      "learning_rate": 7.000310430423262e-08,
+      "loss": 1.4445,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9636176420113266,
+      "grad_norm": 3.8029473734423775,
+      "learning_rate": 6.934810012316218e-08,
+      "loss": 1.2234,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9637892569074996,
+      "grad_norm": 4.437628943541048,
+      "learning_rate": 6.869616406806056e-08,
+      "loss": 1.4683,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9639608718036725,
+      "grad_norm": 4.691688383601371,
+      "learning_rate": 6.804729634034779e-08,
+      "loss": 1.616,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9641324866998455,
+      "grad_norm": 4.27240001711328,
+      "learning_rate": 6.74014971404946e-08,
+      "loss": 1.5508,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9643041015960185,
+      "grad_norm": 5.11907953126176,
+      "learning_rate": 6.67587666680236e-08,
+      "loss": 1.7056,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9644757164921915,
+      "grad_norm": 4.160933818593476,
+      "learning_rate": 6.611910512150932e-08,
+      "loss": 1.4335,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9646473313883646,
+      "grad_norm": 5.029390805302452,
+      "learning_rate": 6.54825126985792e-08,
+      "loss": 1.4893,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9648189462845375,
+      "grad_norm": 4.078951692921687,
+      "learning_rate": 6.484898959591146e-08,
+      "loss": 1.489,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9649905611807105,
+      "grad_norm": 4.712814792856105,
+      "learning_rate": 6.421853600923622e-08,
+      "loss": 1.3516,
+      "step": 5623
+    },
+    {
+      "epoch": 0.9651621760768835,
+      "grad_norm": 4.677202471234876,
+      "learning_rate": 6.359115213333544e-08,
+      "loss": 1.4982,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9653337909730565,
+      "grad_norm": 4.557253335292012,
+      "learning_rate": 6.296683816204185e-08,
+      "loss": 1.5158,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9655054058692294,
+      "grad_norm": 4.61112946644263,
+      "learning_rate": 6.234559428824227e-08,
+      "loss": 1.4966,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9656770207654024,
+      "grad_norm": 4.765200495378624,
+      "learning_rate": 6.172742070387205e-08,
+      "loss": 1.5975,
+      "step": 5627
+    },
+    {
+      "epoch": 0.9658486356615754,
+      "grad_norm": 6.824739231345612,
+      "learning_rate": 6.111231759991954e-08,
+      "loss": 1.6472,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9660202505577484,
+      "grad_norm": 5.212286910268575,
+      "learning_rate": 6.05002851664227e-08,
+      "loss": 1.6497,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9661918654539214,
+      "grad_norm": 4.36262715050695,
+      "learning_rate": 5.989132359247363e-08,
+      "loss": 1.4278,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9663634803500943,
+      "grad_norm": 4.789479105511576,
+      "learning_rate": 5.928543306621293e-08,
+      "loss": 1.3255,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9665350952462674,
+      "grad_norm": 5.011786989281166,
+      "learning_rate": 5.868261377483531e-08,
+      "loss": 1.5388,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9667067101424404,
+      "grad_norm": 4.1783168379108035,
+      "learning_rate": 5.80828659045829e-08,
+      "loss": 1.3907,
+      "step": 5633
+    },
+    {
+      "epoch": 0.9668783250386134,
+      "grad_norm": 4.580360306910773,
+      "learning_rate": 5.7486189640751923e-08,
+      "loss": 1.6093,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9670499399347864,
+      "grad_norm": 4.518115070999871,
+      "learning_rate": 5.689258516768825e-08,
+      "loss": 1.4196,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9672215548309593,
+      "grad_norm": 4.6891687237239665,
+      "learning_rate": 5.6302052668789634e-08,
+      "loss": 1.6617,
+      "step": 5636
+    },
+    {
+      "epoch": 0.9673931697271323,
+      "grad_norm": 4.098383764712945,
+      "learning_rate": 5.5714592326504555e-08,
+      "loss": 1.4884,
+      "step": 5637
+    },
+    {
+      "epoch": 0.9675647846233053,
+      "grad_norm": 4.37431239787366,
+      "learning_rate": 5.5130204322330073e-08,
+      "loss": 1.4873,
+      "step": 5638
+    },
+    {
+      "epoch": 0.9677363995194783,
+      "grad_norm": 4.31552391862251,
+      "learning_rate": 5.4548888836817305e-08,
+      "loss": 1.4955,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9679080144156513,
+      "grad_norm": 4.460151511260132,
+      "learning_rate": 5.3970646049564815e-08,
+      "loss": 1.4606,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9680796293118242,
+      "grad_norm": 4.183143841970551,
+      "learning_rate": 5.339547613922635e-08,
+      "loss": 1.4728,
+      "step": 5641
+    },
+    {
+      "epoch": 0.9682512442079972,
+      "grad_norm": 3.4693265311756774,
+      "learning_rate": 5.282337928350312e-08,
+      "loss": 1.3582,
+      "step": 5642
+    },
+    {
+      "epoch": 0.9684228591041703,
+      "grad_norm": 4.38287982213237,
+      "learning_rate": 5.225435565914483e-08,
+      "loss": 1.4741,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9685944740003433,
+      "grad_norm": 4.57070422918699,
+      "learning_rate": 5.1688405441957525e-08,
+      "loss": 1.7358,
+      "step": 5644
+    },
+    {
+      "epoch": 0.9687660888965162,
+      "grad_norm": 6.233663364477786,
+      "learning_rate": 5.112552880679245e-08,
+      "loss": 1.5605,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9689377037926892,
+      "grad_norm": 3.951208875316004,
+      "learning_rate": 5.0565725927553824e-08,
+      "loss": 1.4682,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9691093186888622,
+      "grad_norm": 4.883379189603963,
+      "learning_rate": 5.000899697719552e-08,
+      "loss": 1.5666,
+      "step": 5647
+    },
+    {
+      "epoch": 0.9692809335850352,
+      "grad_norm": 4.299041108648855,
+      "learning_rate": 4.945534212772107e-08,
+      "loss": 1.2871,
+      "step": 5648
+    },
+    {
+      "epoch": 0.9694525484812082,
+      "grad_norm": 4.369021210298357,
+      "learning_rate": 4.8904761550186975e-08,
+      "loss": 1.373,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9696241633773811,
+      "grad_norm": 4.637321801526171,
+      "learning_rate": 4.835725541469605e-08,
+      "loss": 1.6171,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9697957782735541,
+      "grad_norm": 4.04450574766994,
+      "learning_rate": 4.781282389040409e-08,
+      "loss": 1.7241,
+      "step": 5651
+    },
+    {
+      "epoch": 0.9699673931697271,
+      "grad_norm": 4.562322279166501,
+      "learning_rate": 4.727146714551545e-08,
+      "loss": 1.6226,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9701390080659001,
+      "grad_norm": 4.096937042355048,
+      "learning_rate": 4.673318534728522e-08,
+      "loss": 1.5468,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9703106229620732,
+      "grad_norm": 4.197094126313419,
+      "learning_rate": 4.619797866201814e-08,
+      "loss": 1.3607,
+      "step": 5654
+    },
+    {
+      "epoch": 0.970482237858246,
+      "grad_norm": 4.642033432673629,
+      "learning_rate": 4.566584725507084e-08,
+      "loss": 1.3341,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9706538527544191,
+      "grad_norm": 4.367462577968273,
+      "learning_rate": 4.513679129084514e-08,
+      "loss": 1.7135,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9708254676505921,
+      "grad_norm": 5.232433403446701,
+      "learning_rate": 4.4610810932798065e-08,
+      "loss": 1.1977,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9709970825467651,
+      "grad_norm": 5.736348041399838,
+      "learning_rate": 4.4087906343432963e-08,
+      "loss": 1.6098,
+      "step": 5658
+    },
+    {
+      "epoch": 0.9711686974429381,
+      "grad_norm": 3.9031832149356522,
+      "learning_rate": 4.356807768430282e-08,
+      "loss": 1.5015,
+      "step": 5659
+    },
+    {
+      "epoch": 0.971340312339111,
+      "grad_norm": 4.910530433607537,
+      "learning_rate": 4.3051325116013623e-08,
+      "loss": 1.569,
+      "step": 5660
+    },
+    {
+      "epoch": 0.971511927235284,
+      "grad_norm": 4.072917186403384,
+      "learning_rate": 4.253764879821653e-08,
+      "loss": 1.306,
+      "step": 5661
+    },
+    {
+      "epoch": 0.971683542131457,
+      "grad_norm": 4.408817359383308,
+      "learning_rate": 4.202704888961684e-08,
+      "loss": 1.4006,
+      "step": 5662
+    },
+    {
+      "epoch": 0.97185515702763,
+      "grad_norm": 5.17611130219878,
+      "learning_rate": 4.15195255479639e-08,
+      "loss": 1.3705,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9720267719238029,
+      "grad_norm": 4.2267671743095105,
+      "learning_rate": 4.101507893006229e-08,
+      "loss": 1.5273,
+      "step": 5664
+    },
+    {
+      "epoch": 0.972198386819976,
+      "grad_norm": 4.229414580797337,
+      "learning_rate": 4.05137091917629e-08,
+      "loss": 1.3282,
+      "step": 5665
+    },
+    {
+      "epoch": 0.972370001716149,
+      "grad_norm": 4.817358352905265,
+      "learning_rate": 4.001541648796625e-08,
+      "loss": 1.4523,
+      "step": 5666
+    },
+    {
+      "epoch": 0.972541616612322,
+      "grad_norm": 4.265060752973862,
+      "learning_rate": 3.9520200972621434e-08,
+      "loss": 1.4072,
+      "step": 5667
+    },
+    {
+      "epoch": 0.972713231508495,
+      "grad_norm": 4.628357517939475,
+      "learning_rate": 3.9028062798729395e-08,
+      "loss": 1.5124,
+      "step": 5668
+    },
+    {
+      "epoch": 0.9728848464046679,
+      "grad_norm": 3.862514053606767,
+      "learning_rate": 3.853900211833739e-08,
+      "loss": 1.3437,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9730564613008409,
+      "grad_norm": 3.8870141712613506,
+      "learning_rate": 3.805301908254455e-08,
+      "loss": 1.5777,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9732280761970139,
+      "grad_norm": 4.3758078102897064,
+      "learning_rate": 3.7570113841497446e-08,
+      "loss": 1.4638,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9733996910931869,
+      "grad_norm": 5.253787985608671,
+      "learning_rate": 3.7090286544391174e-08,
+      "loss": 1.73,
+      "step": 5672
+    },
+    {
+      "epoch": 0.9735713059893599,
+      "grad_norm": 4.817950438761102,
+      "learning_rate": 3.6613537339471596e-08,
+      "loss": 1.4443,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9737429208855328,
+      "grad_norm": 4.7533919681960635,
+      "learning_rate": 3.613986637403311e-08,
+      "loss": 1.3559,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9739145357817058,
+      "grad_norm": 4.952485798782778,
+      "learning_rate": 3.5669273794418646e-08,
+      "loss": 1.3697,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9740861506778788,
+      "grad_norm": 3.9253608493935936,
+      "learning_rate": 3.5201759746019695e-08,
+      "loss": 1.3966,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9742577655740519,
+      "grad_norm": 5.427079815018357,
+      "learning_rate": 3.473732437327737e-08,
+      "loss": 1.4305,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9744293804702249,
+      "grad_norm": 4.714282667722972,
+      "learning_rate": 3.427596781968134e-08,
+      "loss": 1.5055,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9746009953663978,
+      "grad_norm": 5.180743626011448,
+      "learning_rate": 3.381769022776982e-08,
+      "loss": 1.5293,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9747726102625708,
+      "grad_norm": 5.6085071842119625,
+      "learning_rate": 3.3362491739131755e-08,
+      "loss": 1.5025,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9749442251587438,
+      "grad_norm": 4.040738831024504,
+      "learning_rate": 3.291037249440132e-08,
+      "loss": 1.2808,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9751158400549168,
+      "grad_norm": 4.063090505117431,
+      "learning_rate": 3.246133263326234e-08,
+      "loss": 1.5642,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9752874549510897,
+      "grad_norm": 4.216981942615482,
+      "learning_rate": 3.201537229445051e-08,
+      "loss": 1.6496,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9754590698472627,
+      "grad_norm": 4.525753118471045,
+      "learning_rate": 3.1572491615745606e-08,
+      "loss": 1.389,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9756306847434357,
+      "grad_norm": 4.178939500728314,
+      "learning_rate": 3.1132690733978175e-08,
+      "loss": 1.5836,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9758022996396087,
+      "grad_norm": 4.874749820659827,
+      "learning_rate": 3.0695969785027316e-08,
+      "loss": 1.5868,
+      "step": 5686
+    },
+    {
+      "epoch": 0.9759739145357818,
+      "grad_norm": 4.961118001545191,
+      "learning_rate": 3.026232890382064e-08,
+      "loss": 1.4784,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9761455294319547,
+      "grad_norm": 4.252176699457493,
+      "learning_rate": 2.98317682243332e-08,
+      "loss": 1.4968,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9763171443281277,
+      "grad_norm": 4.829474844550253,
+      "learning_rate": 2.940428787958749e-08,
+      "loss": 1.4216,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9764887592243007,
+      "grad_norm": 4.372735145280348,
+      "learning_rate": 2.8979888001658963e-08,
+      "loss": 1.7711,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9766603741204737,
+      "grad_norm": 4.466729583742108,
+      "learning_rate": 2.8558568721666068e-08,
+      "loss": 1.519,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9768319890166467,
+      "grad_norm": 4.742783426323677,
+      "learning_rate": 2.814033016977802e-08,
+      "loss": 1.5538,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9770036039128196,
+      "grad_norm": 3.7933557580084694,
+      "learning_rate": 2.7725172475211447e-08,
+      "loss": 1.3787,
+      "step": 5693
+    },
+    {
+      "epoch": 0.9771752188089926,
+      "grad_norm": 4.717837072080277,
+      "learning_rate": 2.7313095766230423e-08,
+      "loss": 1.5531,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9773468337051656,
+      "grad_norm": 4.833001945925235,
+      "learning_rate": 2.6904100170150883e-08,
+      "loss": 1.2479,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9775184486013386,
+      "grad_norm": 5.626134413663522,
+      "learning_rate": 2.6498185813332854e-08,
+      "loss": 1.4713,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9776900634975115,
+      "grad_norm": 4.736983593289776,
+      "learning_rate": 2.6095352821184916e-08,
+      "loss": 1.5569,
+      "step": 5697
+    },
+    {
+      "epoch": 0.9778616783936845,
+      "grad_norm": 4.336991277328159,
+      "learning_rate": 2.5695601318165287e-08,
+      "loss": 1.3707,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9780332932898576,
+      "grad_norm": 4.205624960758902,
+      "learning_rate": 2.5298931427777395e-08,
+      "loss": 1.6069,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9782049081860306,
+      "grad_norm": 5.009211819509579,
+      "learning_rate": 2.4905343272577655e-08,
+      "loss": 1.7994,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9783765230822036,
+      "grad_norm": 5.352526680031248,
+      "learning_rate": 2.4514836974165458e-08,
+      "loss": 1.4851,
+      "step": 5701
+    },
+    {
+      "epoch": 0.9785481379783765,
+      "grad_norm": 3.9432798655582304,
+      "learning_rate": 2.4127412653188743e-08,
+      "loss": 1.157,
+      "step": 5702
+    },
+    {
+      "epoch": 0.9787197528745495,
+      "grad_norm": 4.0367487302013485,
+      "learning_rate": 2.3743070429345095e-08,
+      "loss": 1.3278,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9788913677707225,
+      "grad_norm": 4.554090750731708,
+      "learning_rate": 2.336181042137953e-08,
+      "loss": 1.3468,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9790629826668955,
+      "grad_norm": 4.195006522439029,
+      "learning_rate": 2.2983632747084483e-08,
+      "loss": 1.5046,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9792345975630685,
+      "grad_norm": 4.6361011590546815,
+      "learning_rate": 2.2608537523298724e-08,
+      "loss": 1.5538,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9794062124592414,
+      "grad_norm": 4.364224186014652,
+      "learning_rate": 2.2236524865910658e-08,
+      "loss": 1.5663,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9795778273554144,
+      "grad_norm": 3.962000859663549,
+      "learning_rate": 2.1867594889855016e-08,
+      "loss": 1.4628,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9797494422515874,
+      "grad_norm": 5.393321284172536,
+      "learning_rate": 2.150174770911506e-08,
+      "loss": 1.4604,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9799210571477605,
+      "grad_norm": 6.0966067474000605,
+      "learning_rate": 2.1138983436721495e-08,
+      "loss": 1.7479,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9800926720439335,
+      "grad_norm": 4.64818726595364,
+      "learning_rate": 2.0779302184751326e-08,
+      "loss": 1.5483,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9802642869401064,
+      "grad_norm": 4.056208246221719,
+      "learning_rate": 2.0422704064330113e-08,
+      "loss": 1.6097,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9804359018362794,
+      "grad_norm": 3.7183952983824478,
+      "learning_rate": 2.0069189185630834e-08,
+      "loss": 1.5155,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9806075167324524,
+      "grad_norm": 4.2485997105560465,
+      "learning_rate": 1.9718757657873898e-08,
+      "loss": 1.2681,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9807791316286254,
+      "grad_norm": 3.9133277974927365,
+      "learning_rate": 1.937140958932604e-08,
+      "loss": 1.4072,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9809507465247983,
+      "grad_norm": 4.463848783905341,
+      "learning_rate": 1.9027145087303635e-08,
+      "loss": 1.3942,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9811223614209713,
+      "grad_norm": 5.331659535666475,
+      "learning_rate": 1.8685964258168265e-08,
+      "loss": 1.6365,
+      "step": 5717
+    },
+    {
+      "epoch": 0.9812939763171443,
+      "grad_norm": 5.0894895373248215,
+      "learning_rate": 1.834786720733006e-08,
+      "loss": 1.2009,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9814655912133173,
+      "grad_norm": 5.975640074130434,
+      "learning_rate": 1.8012854039244354e-08,
+      "loss": 1.5197,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9816372061094903,
+      "grad_norm": 4.888002329586236,
+      "learning_rate": 1.7680924857417237e-08,
+      "loss": 1.4681,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9818088210056632,
+      "grad_norm": 4.271523709414777,
+      "learning_rate": 1.735207976439779e-08,
+      "loss": 1.3427,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9819804359018363,
+      "grad_norm": 4.245891997905698,
+      "learning_rate": 1.702631886178585e-08,
+      "loss": 1.5707,
+      "step": 5722
+    },
+    {
+      "epoch": 0.9821520507980093,
+      "grad_norm": 4.5541735441126905,
+      "learning_rate": 1.6703642250227582e-08,
+      "loss": 1.4318,
+      "step": 5723
+    },
+    {
+      "epoch": 0.9823236656941823,
+      "grad_norm": 4.508941451792651,
+      "learning_rate": 1.6384050029413235e-08,
+      "loss": 1.6497,
+      "step": 5724
+    },
+    {
+      "epoch": 0.9824952805903553,
+      "grad_norm": 5.930182070393091,
+      "learning_rate": 1.6067542298083826e-08,
+      "loss": 1.3516,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9826668954865282,
+      "grad_norm": 4.431456946840624,
+      "learning_rate": 1.5754119154026693e-08,
+      "loss": 1.5274,
+      "step": 5726
+    },
+    {
+      "epoch": 0.9828385103827012,
+      "grad_norm": 4.764582813739088,
+      "learning_rate": 1.5443780694074373e-08,
+      "loss": 1.6669,
+      "step": 5727
+    },
+    {
+      "epoch": 0.9830101252788742,
+      "grad_norm": 5.3877084889889515,
+      "learning_rate": 1.513652701410795e-08,
+      "loss": 1.5626,
+      "step": 5728
+    },
+    {
+      "epoch": 0.9831817401750472,
+      "grad_norm": 4.056051069896262,
+      "learning_rate": 1.483235820905593e-08,
+      "loss": 1.4734,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9833533550712202,
+      "grad_norm": 4.569765400541128,
+      "learning_rate": 1.4531274372890925e-08,
+      "loss": 1.5172,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9835249699673931,
+      "grad_norm": 4.076733099517534,
+      "learning_rate": 1.4233275598635188e-08,
+      "loss": 1.412,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9836965848635661,
+      "grad_norm": 4.8909774703152396,
+      "learning_rate": 1.3938361978358406e-08,
+      "loss": 1.6001,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9838681997597392,
+      "grad_norm": 4.349696469923076,
+      "learning_rate": 1.3646533603173252e-08,
+      "loss": 1.3631,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9840398146559122,
+      "grad_norm": 5.019159936530047,
+      "learning_rate": 1.3357790563243156e-08,
+      "loss": 1.485,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9842114295520851,
+      "grad_norm": 5.361966139277498,
+      "learning_rate": 1.3072132947776762e-08,
+      "loss": 1.5821,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9843830444482581,
+      "grad_norm": 5.066564835395103,
+      "learning_rate": 1.2789560845029026e-08,
+      "loss": 1.5014,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9845546593444311,
+      "grad_norm": 3.947854313608584,
+      "learning_rate": 1.2510074342301225e-08,
+      "loss": 1.4026,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9847262742406041,
+      "grad_norm": 4.89044729056931,
+      "learning_rate": 1.2233673525943179e-08,
+      "loss": 1.4504,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9848978891367771,
+      "grad_norm": 4.717725721957725,
+      "learning_rate": 1.196035848135102e-08,
+      "loss": 1.4044,
+      "step": 5739
+    },
+    {
+      "epoch": 0.98506950403295,
+      "grad_norm": 4.9418648925350785,
+      "learning_rate": 1.169012929296609e-08,
+      "loss": 1.6687,
+      "step": 5740
+    },
+    {
+      "epoch": 0.985241118929123,
+      "grad_norm": 4.180740359090647,
+      "learning_rate": 1.1422986044276051e-08,
+      "loss": 1.5111,
+      "step": 5741
+    },
+    {
+      "epoch": 0.985412733825296,
+      "grad_norm": 5.175074666934108,
+      "learning_rate": 1.1158928817817104e-08,
+      "loss": 1.5831,
+      "step": 5742
+    },
+    {
+      "epoch": 0.985584348721469,
+      "grad_norm": 4.412875403121942,
+      "learning_rate": 1.0897957695171767e-08,
+      "loss": 1.451,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9857559636176421,
+      "grad_norm": 4.077973343323954,
+      "learning_rate": 1.0640072756967768e-08,
+      "loss": 1.3742,
+      "step": 5744
+    },
+    {
+      "epoch": 0.985927578513815,
+      "grad_norm": 4.4413614868434665,
+      "learning_rate": 1.038527408288026e-08,
+      "loss": 1.3252,
+      "step": 5745
+    },
+    {
+      "epoch": 0.986099193409988,
+      "grad_norm": 4.215948662693966,
+      "learning_rate": 1.0133561751629605e-08,
+      "loss": 1.4889,
+      "step": 5746
+    },
+    {
+      "epoch": 0.986270808306161,
+      "grad_norm": 5.365747097472438,
+      "learning_rate": 9.884935840984711e-09,
+      "loss": 1.6181,
+      "step": 5747
+    },
+    {
+      "epoch": 0.986442423202334,
+      "grad_norm": 3.9984124006778257,
+      "learning_rate": 9.639396427759684e-09,
+      "loss": 1.3671,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9866140380985069,
+      "grad_norm": 3.8998902758765883,
+      "learning_rate": 9.396943587813844e-09,
+      "loss": 1.4449,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9867856529946799,
+      "grad_norm": 4.888190120951397,
+      "learning_rate": 9.157577396056161e-09,
+      "loss": 1.5563,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9869572678908529,
+      "grad_norm": 4.64076007560539,
+      "learning_rate": 8.921297926439698e-09,
+      "loss": 1.4811,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9871288827870259,
+      "grad_norm": 4.37464919950965,
+      "learning_rate": 8.68810525196273e-09,
+      "loss": 1.484,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9873004976831989,
+      "grad_norm": 5.055811491546965,
+      "learning_rate": 8.457999444673181e-09,
+      "loss": 1.5054,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9874721125793718,
+      "grad_norm": 5.008744153516659,
+      "learning_rate": 8.230980575661962e-09,
+      "loss": 1.3461,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9876437274755449,
+      "grad_norm": 4.4398751615219965,
+      "learning_rate": 8.007048715068522e-09,
+      "loss": 1.5646,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9878153423717179,
+      "grad_norm": 4.034380623128741,
+      "learning_rate": 7.786203932078628e-09,
+      "loss": 1.2722,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9879869572678909,
+      "grad_norm": 5.089243824954915,
+      "learning_rate": 7.568446294921039e-09,
+      "loss": 1.6698,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9881585721640639,
+      "grad_norm": 6.214270049157232,
+      "learning_rate": 7.353775870875268e-09,
+      "loss": 1.5897,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9883301870602368,
+      "grad_norm": 4.3666436609758215,
+      "learning_rate": 7.142192726263819e-09,
+      "loss": 1.5412,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9885018019564098,
+      "grad_norm": 4.281301235759054,
+      "learning_rate": 6.933696926456624e-09,
+      "loss": 1.4412,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9886734168525828,
+      "grad_norm": 4.6665979094306005,
+      "learning_rate": 6.728288535868821e-09,
+      "loss": 1.4537,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9888450317487558,
+      "grad_norm": 4.898945659991402,
+      "learning_rate": 6.525967617964091e-09,
+      "loss": 1.3742,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9890166466449288,
+      "grad_norm": 3.9163507060149705,
+      "learning_rate": 6.326734235249099e-09,
+      "loss": 1.5333,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9891882615411017,
+      "grad_norm": 4.747806602033526,
+      "learning_rate": 6.130588449277941e-09,
+      "loss": 1.553,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9893598764372747,
+      "grad_norm": 4.461696847607399,
+      "learning_rate": 5.9375303206521406e-09,
+      "loss": 1.7084,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9895314913334478,
+      "grad_norm": 4.288740577237806,
+      "learning_rate": 5.747559909016209e-09,
+      "loss": 1.4774,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9897031062296208,
+      "grad_norm": 4.189912234005733,
+      "learning_rate": 5.5606772730643074e-09,
+      "loss": 1.484,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9898747211257937,
+      "grad_norm": 4.53852672439193,
+      "learning_rate": 5.376882470533585e-09,
+      "loss": 1.7763,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9900463360219667,
+      "grad_norm": 4.275803342711082,
+      "learning_rate": 5.196175558209727e-09,
+      "loss": 1.3581,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9902179509181397,
+      "grad_norm": 4.9091876608330605,
+      "learning_rate": 5.018556591921409e-09,
+      "loss": 1.4312,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9903895658143127,
+      "grad_norm": 4.107493584968732,
+      "learning_rate": 4.844025626544735e-09,
+      "loss": 1.3582,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9905611807104857,
+      "grad_norm": 3.7564380181909702,
+      "learning_rate": 4.672582716004348e-09,
+      "loss": 1.1795,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9907327956066586,
+      "grad_norm": 4.412765304725561,
+      "learning_rate": 4.504227913265658e-09,
+      "loss": 1.6317,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9909044105028316,
+      "grad_norm": 5.24320877700556,
+      "learning_rate": 4.338961270344832e-09,
+      "loss": 1.4063,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9910760253990046,
+      "grad_norm": 5.178813482280333,
+      "learning_rate": 4.17678283830103e-09,
+      "loss": 1.4973,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9912476402951776,
+      "grad_norm": 4.381699413398468,
+      "learning_rate": 4.017692667239726e-09,
+      "loss": 1.379,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9914192551913507,
+      "grad_norm": 4.730069800027837,
+      "learning_rate": 3.861690806312712e-09,
+      "loss": 1.4277,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9915908700875236,
+      "grad_norm": 4.374380855516868,
+      "learning_rate": 3.708777303718103e-09,
+      "loss": 1.72,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9917624849836966,
+      "grad_norm": 4.196938287157688,
+      "learning_rate": 3.5589522066992178e-09,
+      "loss": 1.5164,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9919340998798696,
+      "grad_norm": 4.305441793976212,
+      "learning_rate": 3.4122155615456954e-09,
+      "loss": 1.5696,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9921057147760426,
+      "grad_norm": 4.294337267009494,
+      "learning_rate": 3.2685674135912727e-09,
+      "loss": 1.6286,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9922773296722156,
+      "grad_norm": 4.188989705906731,
+      "learning_rate": 3.128007807217115e-09,
+      "loss": 1.4182,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9924489445683885,
+      "grad_norm": 5.3859698793037225,
+      "learning_rate": 2.990536785850706e-09,
+      "loss": 1.4849,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9926205594645615,
+      "grad_norm": 4.107292820838975,
+      "learning_rate": 2.8561543919647382e-09,
+      "loss": 1.5823,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9927921743607345,
+      "grad_norm": 3.793923421622299,
+      "learning_rate": 2.7248606670760012e-09,
+      "loss": 1.3456,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9929637892569075,
+      "grad_norm": 4.144068948793065,
+      "learning_rate": 2.596655651748714e-09,
+      "loss": 1.641,
+      "step": 5786
+    },
+    {
+      "epoch": 0.9931354041530804,
+      "grad_norm": 4.619523835567354,
+      "learning_rate": 2.471539385592303e-09,
+      "loss": 1.5316,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9933070190492534,
+      "grad_norm": 4.407734077777609,
+      "learning_rate": 2.349511907263624e-09,
+      "loss": 1.536,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9934786339454265,
+      "grad_norm": 4.236732643427535,
+      "learning_rate": 2.23057325446252e-09,
+      "loss": 1.2486,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9936502488415995,
+      "grad_norm": 4.524850324515112,
+      "learning_rate": 2.1147234639351534e-09,
+      "loss": 1.5897,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9938218637377725,
+      "grad_norm": 4.425977313268584,
+      "learning_rate": 2.0019625714740032e-09,
+      "loss": 1.4732,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9939934786339454,
+      "grad_norm": 4.938136277441804,
+      "learning_rate": 1.8922906119178686e-09,
+      "loss": 1.5872,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9941650935301184,
+      "grad_norm": 4.569109669252175,
+      "learning_rate": 1.7857076191507562e-09,
+      "loss": 1.6914,
+      "step": 5793
+    },
+    {
+      "epoch": 0.9943367084262914,
+      "grad_norm": 4.333319479768491,
+      "learning_rate": 1.6822136261018807e-09,
+      "loss": 1.6915,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9945083233224644,
+      "grad_norm": 4.840521875000851,
+      "learning_rate": 1.5818086647445551e-09,
+      "loss": 1.6062,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9946799382186374,
+      "grad_norm": 4.108093311140296,
+      "learning_rate": 1.4844927661017416e-09,
+      "loss": 1.4952,
+      "step": 5796
+    },
+    {
+      "epoch": 0.9948515531148103,
+      "grad_norm": 4.924556212858059,
+      "learning_rate": 1.3902659602382795e-09,
+      "loss": 1.4156,
+      "step": 5797
+    },
+    {
+      "epoch": 0.9950231680109833,
+      "grad_norm": 3.7023606293630182,
+      "learning_rate": 1.299128276266437e-09,
+      "loss": 1.3734,
+      "step": 5798
+    },
+    {
+      "epoch": 0.9951947829071564,
+      "grad_norm": 4.253331909254967,
+      "learning_rate": 1.2110797423436903e-09,
+      "loss": 1.3812,
+      "step": 5799
+    },
+    {
+      "epoch": 0.9953663978033294,
+      "grad_norm": 5.31903207893972,
+      "learning_rate": 1.1261203856727243e-09,
+      "loss": 1.6717,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9955380126995023,
+      "grad_norm": 4.373451600826322,
+      "learning_rate": 1.0442502325025417e-09,
+      "loss": 1.7196,
+      "step": 5801
+    },
+    {
+      "epoch": 0.9957096275956753,
+      "grad_norm": 4.10867964278865,
+      "learning_rate": 9.654693081273537e-10,
+      "loss": 1.3952,
+      "step": 5802
+    },
+    {
+      "epoch": 0.9958812424918483,
+      "grad_norm": 4.555893974185902,
+      "learning_rate": 8.897776368865796e-10,
+      "loss": 1.3633,
+      "step": 5803
+    },
+    {
+      "epoch": 0.9960528573880213,
+      "grad_norm": 3.8567048783853206,
+      "learning_rate": 8.171752421670676e-10,
+      "loss": 1.5258,
+      "step": 5804
+    },
+    {
+      "epoch": 0.9962244722841943,
+      "grad_norm": 4.438392567757818,
+      "learning_rate": 7.476621463975431e-10,
+      "loss": 1.577,
+      "step": 5805
+    },
+    {
+      "epoch": 0.9963960871803672,
+      "grad_norm": 4.957109327806115,
+      "learning_rate": 6.812383710552706e-10,
+      "loss": 1.4023,
+      "step": 5806
+    },
+    {
+      "epoch": 0.9965677020765402,
+      "grad_norm": 4.601743178471645,
+      "learning_rate": 6.179039366616125e-10,
+      "loss": 1.4835,
+      "step": 5807
+    },
+    {
+      "epoch": 0.9967393169727132,
+      "grad_norm": 4.128581428788443,
+      "learning_rate": 5.576588627853596e-10,
+      "loss": 1.3219,
+      "step": 5808
+    },
+    {
+      "epoch": 0.9969109318688862,
+      "grad_norm": 3.6985861575738586,
+      "learning_rate": 5.005031680394012e-10,
+      "loss": 1.2912,
+      "step": 5809
+    },
+    {
+      "epoch": 0.9970825467650593,
+      "grad_norm": 4.224786090285248,
+      "learning_rate": 4.4643687008183445e-10,
+      "loss": 1.275,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9972541616612322,
+      "grad_norm": 4.841573319769585,
+      "learning_rate": 3.954599856159647e-10,
+      "loss": 1.8621,
+      "step": 5811
+    },
+    {
+      "epoch": 0.9974257765574052,
+      "grad_norm": 6.116871958726782,
+      "learning_rate": 3.475725303925259e-10,
+      "loss": 1.531,
+      "step": 5812
+    },
+    {
+      "epoch": 0.9975973914535782,
+      "grad_norm": 4.581249786183579,
+      "learning_rate": 3.0277451920635024e-10,
+      "loss": 1.5031,
+      "step": 5813
+    },
+    {
+      "epoch": 0.9977690063497512,
+      "grad_norm": 4.707212027960239,
+      "learning_rate": 2.610659658974779e-10,
+      "loss": 1.6873,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9979406212459242,
+      "grad_norm": 4.3219901955773565,
+      "learning_rate": 2.2244688335226749e-10,
+      "loss": 1.4523,
+      "step": 5815
+    },
+    {
+      "epoch": 0.9981122361420971,
+      "grad_norm": 4.552469182606364,
+      "learning_rate": 1.8691728350228589e-10,
+      "loss": 1.3889,
+      "step": 5816
+    },
+    {
+      "epoch": 0.9982838510382701,
+      "grad_norm": 5.152270907638353,
+      "learning_rate": 1.5447717732541834e-10,
+      "loss": 1.344,
+      "step": 5817
+    },
+    {
+      "epoch": 0.9984554659344431,
+      "grad_norm": 3.785638594551996,
+      "learning_rate": 1.2512657484253788e-10,
+      "loss": 1.4309,
+      "step": 5818
+    },
+    {
+      "epoch": 0.9986270808306161,
+      "grad_norm": 5.645873329631196,
+      "learning_rate": 9.886548512305638e-11,
+      "loss": 1.6786,
+      "step": 5819
+    },
+    {
+      "epoch": 0.998798695726789,
+      "grad_norm": 4.406337412937685,
+      "learning_rate": 7.569391628048372e-11,
+      "loss": 1.3381,
+      "step": 5820
+    },
+    {
+      "epoch": 0.998970310622962,
+      "grad_norm": 4.677360176059494,
+      "learning_rate": 5.5611875472427746e-11,
+      "loss": 1.7457,
+      "step": 5821
+    },
+    {
+      "epoch": 0.999141925519135,
+      "grad_norm": 3.9042113518042725,
+      "learning_rate": 3.8619368903924926e-11,
+      "loss": 1.4341,
+      "step": 5822
+    },
+    {
+      "epoch": 0.9993135404153081,
+      "grad_norm": 3.6987967483322937,
+      "learning_rate": 2.471640182633017e-11,
+      "loss": 1.238,
+      "step": 5823
+    },
+    {
+      "epoch": 0.9994851553114811,
+      "grad_norm": 5.509262200567748,
+      "learning_rate": 1.390297853287592e-11,
+      "loss": 1.5012,
+      "step": 5824
+    },
+    {
+      "epoch": 0.999656770207654,
+      "grad_norm": 3.9329386319957775,
+      "learning_rate": 6.179102365333478e-12,
+      "loss": 1.386,
+      "step": 5825
+    },
+    {
+      "epoch": 0.999828385103827,
+      "grad_norm": 5.197591851477818,
+      "learning_rate": 1.5447757106823447e-12,
+      "loss": 1.6331,
+      "step": 5826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.3208345121338105,
+      "learning_rate": 0.0,
+      "loss": 1.3919,
+      "step": 5827
+    },
+    {
+      "epoch": 1.0,
+      "step": 5827,
+      "total_flos": 1.29494026024072e+18,
+      "train_loss": 1.686343349435838,
+      "train_runtime": 157705.4771,
+      "train_samples_per_second": 2.364,
+      "train_steps_per_second": 0.037
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5827,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.29494026024072e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}