diff --git "a/checkpoint-4000/trainer_state.json" "b/checkpoint-4000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-4000/trainer_state.json"
@@ -0,0 +1,28033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.496709300881659,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00012417732522041475,
+      "grad_norm": 1.9972740411758423,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 3.1309,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002483546504408295,
+      "grad_norm": 0.4196031987667084,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.9498,
+      "step": 2
+    },
+    {
+      "epoch": 0.00037253197566124423,
+      "grad_norm": 0.9132967591285706,
+      "learning_rate": 7.5e-07,
+      "loss": 2.3547,
+      "step": 3
+    },
+    {
+      "epoch": 0.000496709300881659,
+      "grad_norm": 1.5627893209457397,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.7516,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006208866261020738,
+      "grad_norm": 1.751610279083252,
+      "learning_rate": 1.25e-06,
+      "loss": 2.8906,
+      "step": 5
+    },
+    {
+      "epoch": 0.0007450639513224885,
+      "grad_norm": 1.7836052179336548,
+      "learning_rate": 1.5e-06,
+      "loss": 2.7751,
+      "step": 6
+    },
+    {
+      "epoch": 0.0008692412765429032,
+      "grad_norm": 2.1412250995635986,
+      "learning_rate": 1.7500000000000002e-06,
+      "loss": 3.2829,
+      "step": 7
+    },
+    {
+      "epoch": 0.000993418601763318,
+      "grad_norm": 3.818275213241577,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.5971,
+      "step": 8
+    },
+    {
+      "epoch": 0.0011175959269837327,
+      "grad_norm": 2.5483787059783936,
+      "learning_rate": 2.25e-06,
+      "loss": 3.2911,
+      "step": 9
+    },
+    {
+      "epoch": 0.0012417732522041476,
+      "grad_norm": 1.694832682609558,
+      "learning_rate": 2.5e-06,
+      "loss": 2.7594,
+      "step": 10
+    },
+    {
+      "epoch": 0.0013659505774245623,
+      "grad_norm": 2.2460033893585205,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 3.2845,
+      "step": 11
+    },
+    {
+      "epoch": 0.001490127902644977,
+      "grad_norm": 0.9272519946098328,
+      "learning_rate": 3e-06,
+      "loss": 2.2749,
+      "step": 12
+    },
+    {
+      "epoch": 0.0016143052278653918,
+      "grad_norm": 1.1639480590820312,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 2.4602,
+      "step": 13
+    },
+    {
+      "epoch": 0.0017384825530858065,
+      "grad_norm": 3.024296283721924,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 3.8326,
+      "step": 14
+    },
+    {
+      "epoch": 0.0018626598783062214,
+      "grad_norm": 1.4361118078231812,
+      "learning_rate": 3.75e-06,
+      "loss": 2.6242,
+      "step": 15
+    },
+    {
+      "epoch": 0.001986837203526636,
+      "grad_norm": 1.5660126209259033,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.6636,
+      "step": 16
+    },
+    {
+      "epoch": 0.002111014528747051,
+      "grad_norm": 2.5432591438293457,
+      "learning_rate": 4.250000000000001e-06,
+      "loss": 3.3678,
+      "step": 17
+    },
+    {
+      "epoch": 0.0022351918539674654,
+      "grad_norm": 2.389925479888916,
+      "learning_rate": 4.5e-06,
+      "loss": 3.4116,
+      "step": 18
+    },
+    {
+      "epoch": 0.0023593691791878803,
+      "grad_norm": 0.580479621887207,
+      "learning_rate": 4.75e-06,
+      "loss": 2.0553,
+      "step": 19
+    },
+    {
+      "epoch": 0.002483546504408295,
+      "grad_norm": 2.27305269241333,
+      "learning_rate": 5e-06,
+      "loss": 3.7348,
+      "step": 20
+    },
+    {
+      "epoch": 0.0026077238296287096,
+      "grad_norm": 0.42597833275794983,
+      "learning_rate": 5.25e-06,
+      "loss": 1.8877,
+      "step": 21
+    },
+    {
+      "epoch": 0.0027319011548491245,
+      "grad_norm": 3.5201642513275146,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 5.1293,
+      "step": 22
+    },
+    {
+      "epoch": 0.0028560784800695394,
+      "grad_norm": 0.6278425455093384,
+      "learning_rate": 5.750000000000001e-06,
+      "loss": 2.1686,
+      "step": 23
+    },
+    {
+      "epoch": 0.002980255805289954,
+      "grad_norm": 0.47079750895500183,
+      "learning_rate": 6e-06,
+      "loss": 2.244,
+      "step": 24
+    },
+    {
+      "epoch": 0.0031044331305103688,
+      "grad_norm": 0.5824602842330933,
+      "learning_rate": 6.25e-06,
+      "loss": 2.4093,
+      "step": 25
+    },
+    {
+      "epoch": 0.0032286104557307837,
+      "grad_norm": 0.6499738097190857,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 2.4185,
+      "step": 26
+    },
+    {
+      "epoch": 0.003352787780951198,
+      "grad_norm": 0.6688817739486694,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 2.4489,
+      "step": 27
+    },
+    {
+      "epoch": 0.003476965106171613,
+      "grad_norm": 1.1021251678466797,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 2.9258,
+      "step": 28
+    },
+    {
+      "epoch": 0.003601142431392028,
+      "grad_norm": 1.060639500617981,
+      "learning_rate": 7.25e-06,
+      "loss": 3.0351,
+      "step": 29
+    },
+    {
+      "epoch": 0.0037253197566124428,
+      "grad_norm": 2.2871217727661133,
+      "learning_rate": 7.5e-06,
+      "loss": 4.6802,
+      "step": 30
+    },
+    {
+      "epoch": 0.0038494970818328572,
+      "grad_norm": 1.0577988624572754,
+      "learning_rate": 7.75e-06,
+      "loss": 2.8163,
+      "step": 31
+    },
+    {
+      "epoch": 0.003973674407053272,
+      "grad_norm": 0.3350886106491089,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.9198,
+      "step": 32
+    },
+    {
+      "epoch": 0.004097851732273687,
+      "grad_norm": 1.3288217782974243,
+      "learning_rate": 8.25e-06,
+      "loss": 3.468,
+      "step": 33
+    },
+    {
+      "epoch": 0.004222029057494102,
+      "grad_norm": 0.4030284583568573,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 2.2999,
+      "step": 34
+    },
+    {
+      "epoch": 0.004346206382714516,
+      "grad_norm": 0.42509469389915466,
+      "learning_rate": 8.75e-06,
+      "loss": 2.3681,
+      "step": 35
+    },
+    {
+      "epoch": 0.004470383707934931,
+      "grad_norm": 0.4114774465560913,
+      "learning_rate": 9e-06,
+      "loss": 2.2684,
+      "step": 36
+    },
+    {
+      "epoch": 0.004594561033155346,
+      "grad_norm": 0.480406254529953,
+      "learning_rate": 9.25e-06,
+      "loss": 2.4586,
+      "step": 37
+    },
+    {
+      "epoch": 0.004718738358375761,
+      "grad_norm": 1.4463213682174683,
+      "learning_rate": 9.5e-06,
+      "loss": 3.8188,
+      "step": 38
+    },
+    {
+      "epoch": 0.004842915683596175,
+      "grad_norm": 0.33254992961883545,
+      "learning_rate": 9.750000000000002e-06,
+      "loss": 2.11,
+      "step": 39
+    },
+    {
+      "epoch": 0.00496709300881659,
+      "grad_norm": 0.6171764135360718,
+      "learning_rate": 1e-05,
+      "loss": 2.1058,
+      "step": 40
+    },
+    {
+      "epoch": 0.005091270334037005,
+      "grad_norm": 0.7144868969917297,
+      "learning_rate": 1.025e-05,
+      "loss": 2.9112,
+      "step": 41
+    },
+    {
+      "epoch": 0.005215447659257419,
+      "grad_norm": 0.37768566608428955,
+      "learning_rate": 1.05e-05,
+      "loss": 2.4216,
+      "step": 42
+    },
+    {
+      "epoch": 0.005339624984477835,
+      "grad_norm": 0.2819634974002838,
+      "learning_rate": 1.075e-05,
+      "loss": 1.9169,
+      "step": 43
+    },
+    {
+      "epoch": 0.005463802309698249,
+      "grad_norm": 0.4746767282485962,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 2.479,
+      "step": 44
+    },
+    {
+      "epoch": 0.0055879796349186635,
+      "grad_norm": 0.6445564031600952,
+      "learning_rate": 1.125e-05,
+      "loss": 2.7089,
+      "step": 45
+    },
+    {
+      "epoch": 0.005712156960139079,
+      "grad_norm": 0.3493378162384033,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 2.0534,
+      "step": 46
+    },
+    {
+      "epoch": 0.005836334285359493,
+      "grad_norm": 0.40267983078956604,
+      "learning_rate": 1.175e-05,
+      "loss": 2.3149,
+      "step": 47
+    },
+    {
+      "epoch": 0.005960511610579908,
+      "grad_norm": 0.39035916328430176,
+      "learning_rate": 1.2e-05,
+      "loss": 2.178,
+      "step": 48
+    },
+    {
+      "epoch": 0.006084688935800323,
+      "grad_norm": 0.47994619607925415,
+      "learning_rate": 1.225e-05,
+      "loss": 2.581,
+      "step": 49
+    },
+    {
+      "epoch": 0.0062088662610207375,
+      "grad_norm": 0.8229231834411621,
+      "learning_rate": 1.25e-05,
+      "loss": 3.0865,
+      "step": 50
+    },
+    {
+      "epoch": 0.006333043586241152,
+      "grad_norm": 0.47467172145843506,
+      "learning_rate": 1.2750000000000002e-05,
+      "loss": 2.6022,
+      "step": 51
+    },
+    {
+      "epoch": 0.006457220911461567,
+      "grad_norm": 0.5355031490325928,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 2.6013,
+      "step": 52
+    },
+    {
+      "epoch": 0.006581398236681982,
+      "grad_norm": 0.27130579948425293,
+      "learning_rate": 1.3250000000000002e-05,
+      "loss": 1.8945,
+      "step": 53
+    },
+    {
+      "epoch": 0.006705575561902396,
+      "grad_norm": 0.5617486238479614,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 2.0927,
+      "step": 54
+    },
+    {
+      "epoch": 0.0068297528871228115,
+      "grad_norm": 1.4256666898727417,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 3.0528,
+      "step": 55
+    },
+    {
+      "epoch": 0.006953930212343226,
+      "grad_norm": 0.7161185145378113,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.8259,
+      "step": 56
+    },
+    {
+      "epoch": 0.0070781075375636405,
+      "grad_norm": 1.7097219228744507,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 2.6188,
+      "step": 57
+    },
+    {
+      "epoch": 0.007202284862784056,
+      "grad_norm": 0.30535292625427246,
+      "learning_rate": 1.45e-05,
+      "loss": 1.9856,
+      "step": 58
+    },
+    {
+      "epoch": 0.00732646218800447,
+      "grad_norm": 0.3676794767379761,
+      "learning_rate": 1.475e-05,
+      "loss": 2.2203,
+      "step": 59
+    },
+    {
+      "epoch": 0.0074506395132248856,
+      "grad_norm": 0.3784639239311218,
+      "learning_rate": 1.5e-05,
+      "loss": 1.9348,
+      "step": 60
+    },
+    {
+      "epoch": 0.0075748168384453,
+      "grad_norm": 0.7310217022895813,
+      "learning_rate": 1.525e-05,
+      "loss": 3.2562,
+      "step": 61
+    },
+    {
+      "epoch": 0.0076989941636657145,
+      "grad_norm": 0.46555766463279724,
+      "learning_rate": 1.55e-05,
+      "loss": 2.3449,
+      "step": 62
+    },
+    {
+      "epoch": 0.00782317148888613,
+      "grad_norm": 0.4974745810031891,
+      "learning_rate": 1.575e-05,
+      "loss": 2.764,
+      "step": 63
+    },
+    {
+      "epoch": 0.007947348814106544,
+      "grad_norm": 0.7489021420478821,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 3.1774,
+      "step": 64
+    },
+    {
+      "epoch": 0.008071526139326959,
+      "grad_norm": 0.9219920635223389,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 2.343,
+      "step": 65
+    },
+    {
+      "epoch": 0.008195703464547373,
+      "grad_norm": 0.42888444662094116,
+      "learning_rate": 1.65e-05,
+      "loss": 2.3018,
+      "step": 66
+    },
+    {
+      "epoch": 0.008319880789767788,
+      "grad_norm": 0.399170458316803,
+      "learning_rate": 1.675e-05,
+      "loss": 2.3174,
+      "step": 67
+    },
+    {
+      "epoch": 0.008444058114988204,
+      "grad_norm": 0.7459259033203125,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 3.2416,
+      "step": 68
+    },
+    {
+      "epoch": 0.008568235440208618,
+      "grad_norm": 0.7495900988578796,
+      "learning_rate": 1.725e-05,
+      "loss": 3.4573,
+      "step": 69
+    },
+    {
+      "epoch": 0.008692412765429033,
+      "grad_norm": 0.3706538677215576,
+      "learning_rate": 1.75e-05,
+      "loss": 1.9846,
+      "step": 70
+    },
+    {
+      "epoch": 0.008816590090649447,
+      "grad_norm": 0.4431997239589691,
+      "learning_rate": 1.775e-05,
+      "loss": 2.322,
+      "step": 71
+    },
+    {
+      "epoch": 0.008940767415869862,
+      "grad_norm": 0.5538950562477112,
+      "learning_rate": 1.8e-05,
+      "loss": 1.8841,
+      "step": 72
+    },
+    {
+      "epoch": 0.009064944741090276,
+      "grad_norm": 0.9639168977737427,
+      "learning_rate": 1.825e-05,
+      "loss": 3.2895,
+      "step": 73
+    },
+    {
+      "epoch": 0.009189122066310692,
+      "grad_norm": 0.5068603157997131,
+      "learning_rate": 1.85e-05,
+      "loss": 2.2791,
+      "step": 74
+    },
+    {
+      "epoch": 0.009313299391531107,
+      "grad_norm": 0.400095134973526,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 2.3856,
+      "step": 75
+    },
+    {
+      "epoch": 0.009437476716751521,
+      "grad_norm": 0.6737989187240601,
+      "learning_rate": 1.9e-05,
+      "loss": 2.5886,
+      "step": 76
+    },
+    {
+      "epoch": 0.009561654041971936,
+      "grad_norm": 0.5538077354431152,
+      "learning_rate": 1.925e-05,
+      "loss": 2.471,
+      "step": 77
+    },
+    {
+      "epoch": 0.00968583136719235,
+      "grad_norm": 1.5527266263961792,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 2.2257,
+      "step": 78
+    },
+    {
+      "epoch": 0.009810008692412766,
+      "grad_norm": 0.30248555541038513,
+      "learning_rate": 1.9750000000000002e-05,
+      "loss": 1.9698,
+      "step": 79
+    },
+    {
+      "epoch": 0.00993418601763318,
+      "grad_norm": 0.3202931582927704,
+      "learning_rate": 2e-05,
+      "loss": 2.0785,
+      "step": 80
+    },
+    {
+      "epoch": 0.010058363342853595,
+      "grad_norm": 0.9583621621131897,
+      "learning_rate": 2.025e-05,
+      "loss": 3.137,
+      "step": 81
+    },
+    {
+      "epoch": 0.01018254066807401,
+      "grad_norm": 0.33016419410705566,
+      "learning_rate": 2.05e-05,
+      "loss": 1.7716,
+      "step": 82
+    },
+    {
+      "epoch": 0.010306717993294424,
+      "grad_norm": 0.5700507760047913,
+      "learning_rate": 2.075e-05,
+      "loss": 2.3819,
+      "step": 83
+    },
+    {
+      "epoch": 0.010430895318514839,
+      "grad_norm": 0.4957675337791443,
+      "learning_rate": 2.1e-05,
+      "loss": 2.2985,
+      "step": 84
+    },
+    {
+      "epoch": 0.010555072643735255,
+      "grad_norm": 0.32864460349082947,
+      "learning_rate": 2.125e-05,
+      "loss": 2.1227,
+      "step": 85
+    },
+    {
+      "epoch": 0.01067924996895567,
+      "grad_norm": 0.47747576236724854,
+      "learning_rate": 2.15e-05,
+      "loss": 2.5522,
+      "step": 86
+    },
+    {
+      "epoch": 0.010803427294176084,
+      "grad_norm": 0.2985879182815552,
+      "learning_rate": 2.175e-05,
+      "loss": 2.0393,
+      "step": 87
+    },
+    {
+      "epoch": 0.010927604619396498,
+      "grad_norm": 0.437658429145813,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 2.3948,
+      "step": 88
+    },
+    {
+      "epoch": 0.011051781944616913,
+      "grad_norm": 0.7312447428703308,
+      "learning_rate": 2.2250000000000002e-05,
+      "loss": 2.6819,
+      "step": 89
+    },
+    {
+      "epoch": 0.011175959269837327,
+      "grad_norm": 0.43491923809051514,
+      "learning_rate": 2.25e-05,
+      "loss": 2.5222,
+      "step": 90
+    },
+    {
+      "epoch": 0.011300136595057743,
+      "grad_norm": 0.4501711428165436,
+      "learning_rate": 2.275e-05,
+      "loss": 2.2477,
+      "step": 91
+    },
+    {
+      "epoch": 0.011424313920278158,
+      "grad_norm": 0.4149808883666992,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 2.074,
+      "step": 92
+    },
+    {
+      "epoch": 0.011548491245498572,
+      "grad_norm": 0.3053185045719147,
+      "learning_rate": 2.3250000000000003e-05,
+      "loss": 1.7077,
+      "step": 93
+    },
+    {
+      "epoch": 0.011672668570718987,
+      "grad_norm": 0.787242591381073,
+      "learning_rate": 2.35e-05,
+      "loss": 3.0082,
+      "step": 94
+    },
+    {
+      "epoch": 0.011796845895939401,
+      "grad_norm": 1.549908995628357,
+      "learning_rate": 2.375e-05,
+      "loss": 3.3943,
+      "step": 95
+    },
+    {
+      "epoch": 0.011921023221159815,
+      "grad_norm": 0.41458502411842346,
+      "learning_rate": 2.4e-05,
+      "loss": 2.2152,
+      "step": 96
+    },
+    {
+      "epoch": 0.012045200546380232,
+      "grad_norm": 0.5464817881584167,
+      "learning_rate": 2.425e-05,
+      "loss": 2.4457,
+      "step": 97
+    },
+    {
+      "epoch": 0.012169377871600646,
+      "grad_norm": 0.8465514183044434,
+      "learning_rate": 2.45e-05,
+      "loss": 2.5779,
+      "step": 98
+    },
+    {
+      "epoch": 0.01229355519682106,
+      "grad_norm": 0.4009145200252533,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 2.4391,
+      "step": 99
+    },
+    {
+      "epoch": 0.012417732522041475,
+      "grad_norm": 0.3426278233528137,
+      "learning_rate": 2.5e-05,
+      "loss": 1.7594,
+      "step": 100
+    },
+    {
+      "epoch": 0.01254190984726189,
+      "grad_norm": 0.6358038783073425,
+      "learning_rate": 2.525e-05,
+      "loss": 2.7029,
+      "step": 101
+    },
+    {
+      "epoch": 0.012666087172482304,
+      "grad_norm": 0.6930362582206726,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 2.8139,
+      "step": 102
+    },
+    {
+      "epoch": 0.01279026449770272,
+      "grad_norm": 0.9735205769538879,
+      "learning_rate": 2.5750000000000002e-05,
+      "loss": 2.8593,
+      "step": 103
+    },
+    {
+      "epoch": 0.012914441822923135,
+      "grad_norm": 0.350033164024353,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.0911,
+      "step": 104
+    },
+    {
+      "epoch": 0.013038619148143549,
+      "grad_norm": 0.596856415271759,
+      "learning_rate": 2.625e-05,
+      "loss": 2.8891,
+      "step": 105
+    },
+    {
+      "epoch": 0.013162796473363964,
+      "grad_norm": 0.4276641011238098,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 2.3127,
+      "step": 106
+    },
+    {
+      "epoch": 0.013286973798584378,
+      "grad_norm": 0.7200548648834229,
+      "learning_rate": 2.6750000000000003e-05,
+      "loss": 2.6859,
+      "step": 107
+    },
+    {
+      "epoch": 0.013411151123804792,
+      "grad_norm": 0.37764063477516174,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 2.0031,
+      "step": 108
+    },
+    {
+      "epoch": 0.013535328449025209,
+      "grad_norm": 0.7167308330535889,
+      "learning_rate": 2.725e-05,
+      "loss": 3.0279,
+      "step": 109
+    },
+    {
+      "epoch": 0.013659505774245623,
+      "grad_norm": 0.5303190350532532,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 2.4144,
+      "step": 110
+    },
+    {
+      "epoch": 0.013783683099466038,
+      "grad_norm": 0.6632254123687744,
+      "learning_rate": 2.7750000000000004e-05,
+      "loss": 2.8227,
+      "step": 111
+    },
+    {
+      "epoch": 0.013907860424686452,
+      "grad_norm": 0.5389112830162048,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 2.0841,
+      "step": 112
+    },
+    {
+      "epoch": 0.014032037749906866,
+      "grad_norm": 0.46221888065338135,
+      "learning_rate": 2.825e-05,
+      "loss": 2.1262,
+      "step": 113
+    },
+    {
+      "epoch": 0.014156215075127281,
+      "grad_norm": 0.3444712162017822,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.9827,
+      "step": 114
+    },
+    {
+      "epoch": 0.014280392400347697,
+      "grad_norm": 1.7098551988601685,
+      "learning_rate": 2.8749999999999997e-05,
+      "loss": 3.4576,
+      "step": 115
+    },
+    {
+      "epoch": 0.014404569725568112,
+      "grad_norm": 0.3885643482208252,
+      "learning_rate": 2.9e-05,
+      "loss": 2.0672,
+      "step": 116
+    },
+    {
+      "epoch": 0.014528747050788526,
+      "grad_norm": 0.471099853515625,
+      "learning_rate": 2.925e-05,
+      "loss": 2.1644,
+      "step": 117
+    },
+    {
+      "epoch": 0.01465292437600894,
+      "grad_norm": 0.6742566227912903,
+      "learning_rate": 2.95e-05,
+      "loss": 2.2505,
+      "step": 118
+    },
+    {
+      "epoch": 0.014777101701229355,
+      "grad_norm": 0.7954622507095337,
+      "learning_rate": 2.975e-05,
+      "loss": 3.0283,
+      "step": 119
+    },
+    {
+      "epoch": 0.014901279026449771,
+      "grad_norm": 0.464680016040802,
+      "learning_rate": 3e-05,
+      "loss": 2.4224,
+      "step": 120
+    },
+    {
+      "epoch": 0.015025456351670186,
+      "grad_norm": 0.5577122569084167,
+      "learning_rate": 3.025e-05,
+      "loss": 2.6874,
+      "step": 121
+    },
+    {
+      "epoch": 0.0151496336768906,
+      "grad_norm": 0.4663263261318207,
+      "learning_rate": 3.05e-05,
+      "loss": 2.4357,
+      "step": 122
+    },
+    {
+      "epoch": 0.015273811002111014,
+      "grad_norm": 0.8069583177566528,
+      "learning_rate": 3.075e-05,
+      "loss": 2.7963,
+      "step": 123
+    },
+    {
+      "epoch": 0.015397988327331429,
+      "grad_norm": 0.4184455871582031,
+      "learning_rate": 3.1e-05,
+      "loss": 2.0846,
+      "step": 124
+    },
+    {
+      "epoch": 0.015522165652551843,
+      "grad_norm": 0.33683550357818604,
+      "learning_rate": 3.125e-05,
+      "loss": 2.0174,
+      "step": 125
+    },
+    {
+      "epoch": 0.01564634297777226,
+      "grad_norm": 0.4974128305912018,
+      "learning_rate": 3.15e-05,
+      "loss": 2.7199,
+      "step": 126
+    },
+    {
+      "epoch": 0.015770520302992672,
+      "grad_norm": 0.9204639196395874,
+      "learning_rate": 3.175e-05,
+      "loss": 3.4981,
+      "step": 127
+    },
+    {
+      "epoch": 0.01589469762821309,
+      "grad_norm": 0.32168489694595337,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.9273,
+      "step": 128
+    },
+    {
+      "epoch": 0.016018874953433505,
+      "grad_norm": 0.5214820504188538,
+      "learning_rate": 3.2250000000000005e-05,
+      "loss": 2.3157,
+      "step": 129
+    },
+    {
+      "epoch": 0.016143052278653917,
+      "grad_norm": 1.7353380918502808,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 3.2162,
+      "step": 130
+    },
+    {
+      "epoch": 0.016267229603874334,
+      "grad_norm": 0.6233454942703247,
+      "learning_rate": 3.275e-05,
+      "loss": 2.5917,
+      "step": 131
+    },
+    {
+      "epoch": 0.016391406929094746,
+      "grad_norm": 0.4389147460460663,
+      "learning_rate": 3.3e-05,
+      "loss": 2.281,
+      "step": 132
+    },
+    {
+      "epoch": 0.016515584254315162,
+      "grad_norm": 0.6098535060882568,
+      "learning_rate": 3.325e-05,
+      "loss": 2.8663,
+      "step": 133
+    },
+    {
+      "epoch": 0.016639761579535575,
+      "grad_norm": 0.3645295798778534,
+      "learning_rate": 3.35e-05,
+      "loss": 2.2955,
+      "step": 134
+    },
+    {
+      "epoch": 0.01676393890475599,
+      "grad_norm": 0.3560003340244293,
+      "learning_rate": 3.375000000000001e-05,
+      "loss": 1.8143,
+      "step": 135
+    },
+    {
+      "epoch": 0.016888116229976408,
+      "grad_norm": 0.3994666337966919,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 2.4423,
+      "step": 136
+    },
+    {
+      "epoch": 0.01701229355519682,
+      "grad_norm": 0.5861908197402954,
+      "learning_rate": 3.4250000000000006e-05,
+      "loss": 2.6024,
+      "step": 137
+    },
+    {
+      "epoch": 0.017136470880417237,
+      "grad_norm": 0.36611929535865784,
+      "learning_rate": 3.45e-05,
+      "loss": 2.188,
+      "step": 138
+    },
+    {
+      "epoch": 0.01726064820563765,
+      "grad_norm": 0.3659608066082001,
+      "learning_rate": 3.475e-05,
+      "loss": 2.1396,
+      "step": 139
+    },
+    {
+      "epoch": 0.017384825530858065,
+      "grad_norm": 0.33257558941841125,
+      "learning_rate": 3.5e-05,
+      "loss": 1.9805,
+      "step": 140
+    },
+    {
+      "epoch": 0.01750900285607848,
+      "grad_norm": 0.7428910732269287,
+      "learning_rate": 3.525e-05,
+      "loss": 2.8775,
+      "step": 141
+    },
+    {
+      "epoch": 0.017633180181298894,
+      "grad_norm": 0.33400580286979675,
+      "learning_rate": 3.55e-05,
+      "loss": 2.0345,
+      "step": 142
+    },
+    {
+      "epoch": 0.01775735750651931,
+      "grad_norm": 0.35283729434013367,
+      "learning_rate": 3.575e-05,
+      "loss": 1.9029,
+      "step": 143
+    },
+    {
+      "epoch": 0.017881534831739723,
+      "grad_norm": 0.3765888214111328,
+      "learning_rate": 3.6e-05,
+      "loss": 2.0839,
+      "step": 144
+    },
+    {
+      "epoch": 0.01800571215696014,
+      "grad_norm": 1.138491153717041,
+      "learning_rate": 3.625e-05,
+      "loss": 2.9589,
+      "step": 145
+    },
+    {
+      "epoch": 0.018129889482180552,
+      "grad_norm": 0.48436638712882996,
+      "learning_rate": 3.65e-05,
+      "loss": 1.9813,
+      "step": 146
+    },
+    {
+      "epoch": 0.01825406680740097,
+      "grad_norm": 0.38523510098457336,
+      "learning_rate": 3.675e-05,
+      "loss": 2.1353,
+      "step": 147
+    },
+    {
+      "epoch": 0.018378244132621385,
+      "grad_norm": 0.6861639022827148,
+      "learning_rate": 3.7e-05,
+      "loss": 2.3444,
+      "step": 148
+    },
+    {
+      "epoch": 0.018502421457841797,
+      "grad_norm": 0.4485691785812378,
+      "learning_rate": 3.7250000000000004e-05,
+      "loss": 2.3605,
+      "step": 149
+    },
+    {
+      "epoch": 0.018626598783062213,
+      "grad_norm": 0.3629736602306366,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 2.0114,
+      "step": 150
+    },
+    {
+      "epoch": 0.018750776108282626,
+      "grad_norm": 0.30963510274887085,
+      "learning_rate": 3.775e-05,
+      "loss": 1.7653,
+      "step": 151
+    },
+    {
+      "epoch": 0.018874953433503042,
+      "grad_norm": 0.2924124002456665,
+      "learning_rate": 3.8e-05,
+      "loss": 1.8532,
+      "step": 152
+    },
+    {
+      "epoch": 0.01899913075872346,
+      "grad_norm": 0.35504698753356934,
+      "learning_rate": 3.825e-05,
+      "loss": 1.9178,
+      "step": 153
+    },
+    {
+      "epoch": 0.01912330808394387,
+      "grad_norm": 0.847620964050293,
+      "learning_rate": 3.85e-05,
+      "loss": 3.1104,
+      "step": 154
+    },
+    {
+      "epoch": 0.019247485409164287,
+      "grad_norm": 0.432923287153244,
+      "learning_rate": 3.875e-05,
+      "loss": 2.0543,
+      "step": 155
+    },
+    {
+      "epoch": 0.0193716627343847,
+      "grad_norm": 0.38991379737854004,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 2.0407,
+      "step": 156
+    },
+    {
+      "epoch": 0.019495840059605116,
+      "grad_norm": 0.3152579069137573,
+      "learning_rate": 3.9250000000000005e-05,
+      "loss": 1.6679,
+      "step": 157
+    },
+    {
+      "epoch": 0.019620017384825533,
+      "grad_norm": 0.5111315846443176,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 2.3939,
+      "step": 158
+    },
+    {
+      "epoch": 0.019744194710045945,
+      "grad_norm": 0.4135592579841614,
+      "learning_rate": 3.9750000000000004e-05,
+      "loss": 1.9966,
+      "step": 159
+    },
+    {
+      "epoch": 0.01986837203526636,
+      "grad_norm": 0.38342544436454773,
+      "learning_rate": 4e-05,
+      "loss": 2.0957,
+      "step": 160
+    },
+    {
+      "epoch": 0.019992549360486774,
+      "grad_norm": 0.36585286259651184,
+      "learning_rate": 4.025e-05,
+      "loss": 2.1809,
+      "step": 161
+    },
+    {
+      "epoch": 0.02011672668570719,
+      "grad_norm": 0.3803645372390747,
+      "learning_rate": 4.05e-05,
+      "loss": 2.4555,
+      "step": 162
+    },
+    {
+      "epoch": 0.020240904010927603,
+      "grad_norm": 0.3952953517436981,
+      "learning_rate": 4.075e-05,
+      "loss": 2.0289,
+      "step": 163
+    },
+    {
+      "epoch": 0.02036508133614802,
+      "grad_norm": 0.4498346447944641,
+      "learning_rate": 4.1e-05,
+      "loss": 2.1259,
+      "step": 164
+    },
+    {
+      "epoch": 0.020489258661368435,
+      "grad_norm": 0.5039570331573486,
+      "learning_rate": 4.125e-05,
+      "loss": 2.1211,
+      "step": 165
+    },
+    {
+      "epoch": 0.020613435986588848,
+      "grad_norm": 0.552778422832489,
+      "learning_rate": 4.15e-05,
+      "loss": 2.7013,
+      "step": 166
+    },
+    {
+      "epoch": 0.020737613311809264,
+      "grad_norm": 0.48352673649787903,
+      "learning_rate": 4.175e-05,
+      "loss": 2.2531,
+      "step": 167
+    },
+    {
+      "epoch": 0.020861790637029677,
+      "grad_norm": 0.33896100521087646,
+      "learning_rate": 4.2e-05,
+      "loss": 2.0123,
+      "step": 168
+    },
+    {
+      "epoch": 0.020985967962250093,
+      "grad_norm": 0.29688796401023865,
+      "learning_rate": 4.2250000000000004e-05,
+      "loss": 1.7783,
+      "step": 169
+    },
+    {
+      "epoch": 0.02111014528747051,
+      "grad_norm": 0.3621704876422882,
+      "learning_rate": 4.25e-05,
+      "loss": 2.0276,
+      "step": 170
+    },
+    {
+      "epoch": 0.021234322612690922,
+      "grad_norm": 0.6339600682258606,
+      "learning_rate": 4.275e-05,
+      "loss": 3.0074,
+      "step": 171
+    },
+    {
+      "epoch": 0.02135849993791134,
+      "grad_norm": 0.4397406280040741,
+      "learning_rate": 4.3e-05,
+      "loss": 2.1572,
+      "step": 172
+    },
+    {
+      "epoch": 0.02148267726313175,
+      "grad_norm": 0.36329373717308044,
+      "learning_rate": 4.325e-05,
+      "loss": 1.5756,
+      "step": 173
+    },
+    {
+      "epoch": 0.021606854588352167,
+      "grad_norm": 1.2870923280715942,
+      "learning_rate": 4.35e-05,
+      "loss": 3.0002,
+      "step": 174
+    },
+    {
+      "epoch": 0.02173103191357258,
+      "grad_norm": 0.38476189970970154,
+      "learning_rate": 4.375e-05,
+      "loss": 1.9917,
+      "step": 175
+    },
+    {
+      "epoch": 0.021855209238792996,
+      "grad_norm": 0.3581407368183136,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.9033,
+      "step": 176
+    },
+    {
+      "epoch": 0.021979386564013412,
+      "grad_norm": 1.926517128944397,
+      "learning_rate": 4.4250000000000005e-05,
+      "loss": 2.8374,
+      "step": 177
+    },
+    {
+      "epoch": 0.022103563889233825,
+      "grad_norm": 0.5636987686157227,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 2.5388,
+      "step": 178
+    },
+    {
+      "epoch": 0.02222774121445424,
+      "grad_norm": 0.44197601079940796,
+      "learning_rate": 4.4750000000000004e-05,
+      "loss": 2.2843,
+      "step": 179
+    },
+    {
+      "epoch": 0.022351918539674654,
+      "grad_norm": 0.2799761891365051,
+      "learning_rate": 4.5e-05,
+      "loss": 1.7478,
+      "step": 180
+    },
+    {
+      "epoch": 0.02247609586489507,
+      "grad_norm": 0.4067891240119934,
+      "learning_rate": 4.525e-05,
+      "loss": 1.8689,
+      "step": 181
+    },
+    {
+      "epoch": 0.022600273190115486,
+      "grad_norm": 0.7503419518470764,
+      "learning_rate": 4.55e-05,
+      "loss": 2.1794,
+      "step": 182
+    },
+    {
+      "epoch": 0.0227244505153359,
+      "grad_norm": 0.4105197489261627,
+      "learning_rate": 4.575e-05,
+      "loss": 2.3061,
+      "step": 183
+    },
+    {
+      "epoch": 0.022848627840556315,
+      "grad_norm": 0.5977079272270203,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.4258,
+      "step": 184
+    },
+    {
+      "epoch": 0.022972805165776728,
+      "grad_norm": 0.34998947381973267,
+      "learning_rate": 4.6250000000000006e-05,
+      "loss": 1.9122,
+      "step": 185
+    },
+    {
+      "epoch": 0.023096982490997144,
+      "grad_norm": 0.4132118225097656,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 2.4114,
+      "step": 186
+    },
+    {
+      "epoch": 0.023221159816217557,
+      "grad_norm": 0.4211031198501587,
+      "learning_rate": 4.6750000000000005e-05,
+      "loss": 2.2286,
+      "step": 187
+    },
+    {
+      "epoch": 0.023345337141437973,
+      "grad_norm": 0.37225693464279175,
+      "learning_rate": 4.7e-05,
+      "loss": 2.2202,
+      "step": 188
+    },
+    {
+      "epoch": 0.02346951446665839,
+      "grad_norm": 0.35957542061805725,
+      "learning_rate": 4.7249999999999997e-05,
+      "loss": 1.8781,
+      "step": 189
+    },
+    {
+      "epoch": 0.023593691791878802,
+      "grad_norm": 0.6783155202865601,
+      "learning_rate": 4.75e-05,
+      "loss": 2.1989,
+      "step": 190
+    },
+    {
+      "epoch": 0.02371786911709922,
+      "grad_norm": 0.3693026900291443,
+      "learning_rate": 4.775e-05,
+      "loss": 1.975,
+      "step": 191
+    },
+    {
+      "epoch": 0.02384204644231963,
+      "grad_norm": 0.7921383380889893,
+      "learning_rate": 4.8e-05,
+      "loss": 2.7003,
+      "step": 192
+    },
+    {
+      "epoch": 0.023966223767540047,
+      "grad_norm": 0.4388650059700012,
+      "learning_rate": 4.825e-05,
+      "loss": 2.4211,
+      "step": 193
+    },
+    {
+      "epoch": 0.024090401092760463,
+      "grad_norm": 0.487396240234375,
+      "learning_rate": 4.85e-05,
+      "loss": 1.9366,
+      "step": 194
+    },
+    {
+      "epoch": 0.024214578417980876,
+      "grad_norm": 0.3785874843597412,
+      "learning_rate": 4.875e-05,
+      "loss": 2.0287,
+      "step": 195
+    },
+    {
+      "epoch": 0.024338755743201292,
+      "grad_norm": 0.4508357346057892,
+      "learning_rate": 4.9e-05,
+      "loss": 2.1802,
+      "step": 196
+    },
+    {
+      "epoch": 0.024462933068421705,
+      "grad_norm": 0.7750405669212341,
+      "learning_rate": 4.9250000000000004e-05,
+      "loss": 3.2427,
+      "step": 197
+    },
+    {
+      "epoch": 0.02458711039364212,
+      "grad_norm": 0.631817638874054,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 2.578,
+      "step": 198
+    },
+    {
+      "epoch": 0.024711287718862537,
+      "grad_norm": 0.5869089365005493,
+      "learning_rate": 4.975e-05,
+      "loss": 1.8844,
+      "step": 199
+    },
+    {
+      "epoch": 0.02483546504408295,
+      "grad_norm": 0.512671709060669,
+      "learning_rate": 5e-05,
+      "loss": 2.2825,
+      "step": 200
+    },
+    {
+      "epoch": 0.024959642369303366,
+      "grad_norm": 0.5000550150871277,
+      "learning_rate": 5e-05,
+      "loss": 2.4025,
+      "step": 201
+    },
+    {
+      "epoch": 0.02508381969452378,
+      "grad_norm": 0.40964850783348083,
+      "learning_rate": 5e-05,
+      "loss": 2.1271,
+      "step": 202
+    },
+    {
+      "epoch": 0.025207997019744195,
+      "grad_norm": 0.3266145586967468,
+      "learning_rate": 5e-05,
+      "loss": 2.0795,
+      "step": 203
+    },
+    {
+      "epoch": 0.025332174344964608,
+      "grad_norm": 0.3464200794696808,
+      "learning_rate": 5e-05,
+      "loss": 2.0552,
+      "step": 204
+    },
+    {
+      "epoch": 0.025456351670185024,
+      "grad_norm": 0.42155832052230835,
+      "learning_rate": 5e-05,
+      "loss": 2.3771,
+      "step": 205
+    },
+    {
+      "epoch": 0.02558052899540544,
+      "grad_norm": 0.3549226224422455,
+      "learning_rate": 5e-05,
+      "loss": 2.0559,
+      "step": 206
+    },
+    {
+      "epoch": 0.025704706320625853,
+      "grad_norm": 0.3902105391025543,
+      "learning_rate": 5e-05,
+      "loss": 2.06,
+      "step": 207
+    },
+    {
+      "epoch": 0.02582888364584627,
+      "grad_norm": 0.4960445761680603,
+      "learning_rate": 5e-05,
+      "loss": 2.6786,
+      "step": 208
+    },
+    {
+      "epoch": 0.025953060971066682,
+      "grad_norm": 0.5811383724212646,
+      "learning_rate": 5e-05,
+      "loss": 2.8473,
+      "step": 209
+    },
+    {
+      "epoch": 0.026077238296287098,
+      "grad_norm": 0.27700942754745483,
+      "learning_rate": 5e-05,
+      "loss": 1.6227,
+      "step": 210
+    },
+    {
+      "epoch": 0.026201415621507514,
+      "grad_norm": 0.3668477237224579,
+      "learning_rate": 5e-05,
+      "loss": 2.1087,
+      "step": 211
+    },
+    {
+      "epoch": 0.026325592946727927,
+      "grad_norm": 0.411236047744751,
+      "learning_rate": 5e-05,
+      "loss": 2.142,
+      "step": 212
+    },
+    {
+      "epoch": 0.026449770271948343,
+      "grad_norm": 0.3131048381328583,
+      "learning_rate": 5e-05,
+      "loss": 1.9008,
+      "step": 213
+    },
+    {
+      "epoch": 0.026573947597168756,
+      "grad_norm": 0.36345964670181274,
+      "learning_rate": 5e-05,
+      "loss": 2.2525,
+      "step": 214
+    },
+    {
+      "epoch": 0.026698124922389172,
+      "grad_norm": 0.2752915620803833,
+      "learning_rate": 5e-05,
+      "loss": 1.7363,
+      "step": 215
+    },
+    {
+      "epoch": 0.026822302247609585,
+      "grad_norm": 0.612686812877655,
+      "learning_rate": 5e-05,
+      "loss": 2.2539,
+      "step": 216
+    },
+    {
+      "epoch": 0.02694647957283,
+      "grad_norm": 0.5532636046409607,
+      "learning_rate": 5e-05,
+      "loss": 2.6056,
+      "step": 217
+    },
+    {
+      "epoch": 0.027070656898050417,
+      "grad_norm": 0.5926713347434998,
+      "learning_rate": 5e-05,
+      "loss": 2.9557,
+      "step": 218
+    },
+    {
+      "epoch": 0.02719483422327083,
+      "grad_norm": 0.486667275428772,
+      "learning_rate": 5e-05,
+      "loss": 2.4859,
+      "step": 219
+    },
+    {
+      "epoch": 0.027319011548491246,
+      "grad_norm": 0.35966232419013977,
+      "learning_rate": 5e-05,
+      "loss": 1.7959,
+      "step": 220
+    },
+    {
+      "epoch": 0.02744318887371166,
+      "grad_norm": 0.2921900749206543,
+      "learning_rate": 5e-05,
+      "loss": 1.8394,
+      "step": 221
+    },
+    {
+      "epoch": 0.027567366198932075,
+      "grad_norm": 0.7260276675224304,
+      "learning_rate": 5e-05,
+      "loss": 2.7059,
+      "step": 222
+    },
+    {
+      "epoch": 0.02769154352415249,
+      "grad_norm": 0.7896019816398621,
+      "learning_rate": 5e-05,
+      "loss": 2.6853,
+      "step": 223
+    },
+    {
+      "epoch": 0.027815720849372904,
+      "grad_norm": 0.31710851192474365,
+      "learning_rate": 5e-05,
+      "loss": 1.8571,
+      "step": 224
+    },
+    {
+      "epoch": 0.02793989817459332,
+      "grad_norm": 0.337398886680603,
+      "learning_rate": 5e-05,
+      "loss": 2.0157,
+      "step": 225
+    },
+    {
+      "epoch": 0.028064075499813733,
+      "grad_norm": 0.3483402132987976,
+      "learning_rate": 5e-05,
+      "loss": 1.8896,
+      "step": 226
+    },
+    {
+      "epoch": 0.02818825282503415,
+      "grad_norm": 0.29430773854255676,
+      "learning_rate": 5e-05,
+      "loss": 1.9548,
+      "step": 227
+    },
+    {
+      "epoch": 0.028312430150254562,
+      "grad_norm": 0.2870185077190399,
+      "learning_rate": 5e-05,
+      "loss": 1.8494,
+      "step": 228
+    },
+    {
+      "epoch": 0.028436607475474978,
+      "grad_norm": 0.29148703813552856,
+      "learning_rate": 5e-05,
+      "loss": 1.9176,
+      "step": 229
+    },
+    {
+      "epoch": 0.028560784800695394,
+      "grad_norm": 0.33007490634918213,
+      "learning_rate": 5e-05,
+      "loss": 2.1557,
+      "step": 230
+    },
+    {
+      "epoch": 0.028684962125915807,
+      "grad_norm": 0.4312034547328949,
+      "learning_rate": 5e-05,
+      "loss": 2.4879,
+      "step": 231
+    },
+    {
+      "epoch": 0.028809139451136223,
+      "grad_norm": 0.5464475154876709,
+      "learning_rate": 5e-05,
+      "loss": 2.306,
+      "step": 232
+    },
+    {
+      "epoch": 0.028933316776356636,
+      "grad_norm": 0.32755184173583984,
+      "learning_rate": 5e-05,
+      "loss": 2.0765,
+      "step": 233
+    },
+    {
+      "epoch": 0.029057494101577052,
+      "grad_norm": 0.32250750064849854,
+      "learning_rate": 5e-05,
+      "loss": 1.9146,
+      "step": 234
+    },
+    {
+      "epoch": 0.029181671426797468,
+      "grad_norm": 0.3579796552658081,
+      "learning_rate": 5e-05,
+      "loss": 1.8278,
+      "step": 235
+    },
+    {
+      "epoch": 0.02930584875201788,
+      "grad_norm": 0.531172513961792,
+      "learning_rate": 5e-05,
+      "loss": 2.489,
+      "step": 236
+    },
+    {
+      "epoch": 0.029430026077238297,
+      "grad_norm": 0.2668640613555908,
+      "learning_rate": 5e-05,
+      "loss": 1.8602,
+      "step": 237
+    },
+    {
+      "epoch": 0.02955420340245871,
+      "grad_norm": 0.40588808059692383,
+      "learning_rate": 5e-05,
+      "loss": 2.2541,
+      "step": 238
+    },
+    {
+      "epoch": 0.029678380727679126,
+      "grad_norm": 0.7726118564605713,
+      "learning_rate": 5e-05,
+      "loss": 2.8152,
+      "step": 239
+    },
+    {
+      "epoch": 0.029802558052899542,
+      "grad_norm": 0.4816347062587738,
+      "learning_rate": 5e-05,
+      "loss": 2.5003,
+      "step": 240
+    },
+    {
+      "epoch": 0.029926735378119955,
+      "grad_norm": 0.4347776770591736,
+      "learning_rate": 5e-05,
+      "loss": 2.4994,
+      "step": 241
+    },
+    {
+      "epoch": 0.03005091270334037,
+      "grad_norm": 0.38524556159973145,
+      "learning_rate": 5e-05,
+      "loss": 1.8947,
+      "step": 242
+    },
+    {
+      "epoch": 0.030175090028560784,
+      "grad_norm": 0.4469987452030182,
+      "learning_rate": 5e-05,
+      "loss": 2.2407,
+      "step": 243
+    },
+    {
+      "epoch": 0.0302992673537812,
+      "grad_norm": 0.38496163487434387,
+      "learning_rate": 5e-05,
+      "loss": 2.074,
+      "step": 244
+    },
+    {
+      "epoch": 0.030423444679001613,
+      "grad_norm": 0.4645942747592926,
+      "learning_rate": 5e-05,
+      "loss": 2.4636,
+      "step": 245
+    },
+    {
+      "epoch": 0.03054762200422203,
+      "grad_norm": 0.477590411901474,
+      "learning_rate": 5e-05,
+      "loss": 2.189,
+      "step": 246
+    },
+    {
+      "epoch": 0.030671799329442445,
+      "grad_norm": 0.36989307403564453,
+      "learning_rate": 5e-05,
+      "loss": 1.9196,
+      "step": 247
+    },
+    {
+      "epoch": 0.030795976654662858,
+      "grad_norm": 0.2893427312374115,
+      "learning_rate": 5e-05,
+      "loss": 1.8337,
+      "step": 248
+    },
+    {
+      "epoch": 0.030920153979883274,
+      "grad_norm": 0.5507279634475708,
+      "learning_rate": 5e-05,
+      "loss": 2.3324,
+      "step": 249
+    },
+    {
+      "epoch": 0.031044331305103687,
+      "grad_norm": 0.40870580077171326,
+      "learning_rate": 5e-05,
+      "loss": 2.0548,
+      "step": 250
+    },
+    {
+      "epoch": 0.031168508630324103,
+      "grad_norm": 0.35371220111846924,
+      "learning_rate": 5e-05,
+      "loss": 1.883,
+      "step": 251
+    },
+    {
+      "epoch": 0.03129268595554452,
+      "grad_norm": 0.44694972038269043,
+      "learning_rate": 5e-05,
+      "loss": 2.4052,
+      "step": 252
+    },
+    {
+      "epoch": 0.031416863280764935,
+      "grad_norm": 0.36996620893478394,
+      "learning_rate": 5e-05,
+      "loss": 2.0551,
+      "step": 253
+    },
+    {
+      "epoch": 0.031541040605985345,
+      "grad_norm": 0.2715965211391449,
+      "learning_rate": 5e-05,
+      "loss": 1.7275,
+      "step": 254
+    },
+    {
+      "epoch": 0.03166521793120576,
+      "grad_norm": 0.6524844169616699,
+      "learning_rate": 5e-05,
+      "loss": 2.6092,
+      "step": 255
+    },
+    {
+      "epoch": 0.03178939525642618,
+      "grad_norm": 0.6248584389686584,
+      "learning_rate": 5e-05,
+      "loss": 3.1241,
+      "step": 256
+    },
+    {
+      "epoch": 0.03191357258164659,
+      "grad_norm": 0.2754964232444763,
+      "learning_rate": 5e-05,
+      "loss": 1.7563,
+      "step": 257
+    },
+    {
+      "epoch": 0.03203774990686701,
+      "grad_norm": 0.5506348013877869,
+      "learning_rate": 5e-05,
+      "loss": 2.9112,
+      "step": 258
+    },
+    {
+      "epoch": 0.03216192723208742,
+      "grad_norm": 0.34817564487457275,
+      "learning_rate": 5e-05,
+      "loss": 2.0328,
+      "step": 259
+    },
+    {
+      "epoch": 0.032286104557307835,
+      "grad_norm": 0.46653464436531067,
+      "learning_rate": 5e-05,
+      "loss": 2.0462,
+      "step": 260
+    },
+    {
+      "epoch": 0.03241028188252825,
+      "grad_norm": 0.5370591282844543,
+      "learning_rate": 5e-05,
+      "loss": 2.3273,
+      "step": 261
+    },
+    {
+      "epoch": 0.03253445920774867,
+      "grad_norm": 0.2886908948421478,
+      "learning_rate": 5e-05,
+      "loss": 1.8567,
+      "step": 262
+    },
+    {
+      "epoch": 0.032658636532969076,
+      "grad_norm": 0.34875550866127014,
+      "learning_rate": 5e-05,
+      "loss": 2.1524,
+      "step": 263
+    },
+    {
+      "epoch": 0.03278281385818949,
+      "grad_norm": 0.3545679450035095,
+      "learning_rate": 5e-05,
+      "loss": 2.0521,
+      "step": 264
+    },
+    {
+      "epoch": 0.03290699118340991,
+      "grad_norm": 0.5829645991325378,
+      "learning_rate": 5e-05,
+      "loss": 2.5073,
+      "step": 265
+    },
+    {
+      "epoch": 0.033031168508630325,
+      "grad_norm": 0.4411691427230835,
+      "learning_rate": 5e-05,
+      "loss": 2.6498,
+      "step": 266
+    },
+    {
+      "epoch": 0.03315534583385074,
+      "grad_norm": 0.3999745845794678,
+      "learning_rate": 5e-05,
+      "loss": 2.6879,
+      "step": 267
+    },
+    {
+      "epoch": 0.03327952315907115,
+      "grad_norm": 0.5126621723175049,
+      "learning_rate": 5e-05,
+      "loss": 2.053,
+      "step": 268
+    },
+    {
+      "epoch": 0.03340370048429157,
+      "grad_norm": 0.5460847616195679,
+      "learning_rate": 5e-05,
+      "loss": 2.1279,
+      "step": 269
+    },
+    {
+      "epoch": 0.03352787780951198,
+      "grad_norm": 0.4496626853942871,
+      "learning_rate": 5e-05,
+      "loss": 2.0383,
+      "step": 270
+    },
+    {
+      "epoch": 0.0336520551347324,
+      "grad_norm": 0.5265777707099915,
+      "learning_rate": 5e-05,
+      "loss": 2.6784,
+      "step": 271
+    },
+    {
+      "epoch": 0.033776232459952815,
+      "grad_norm": 0.4580742418766022,
+      "learning_rate": 5e-05,
+      "loss": 2.5226,
+      "step": 272
+    },
+    {
+      "epoch": 0.033900409785173224,
+      "grad_norm": 0.3497520387172699,
+      "learning_rate": 5e-05,
+      "loss": 2.0537,
+      "step": 273
+    },
+    {
+      "epoch": 0.03402458711039364,
+      "grad_norm": 0.2773326635360718,
+      "learning_rate": 5e-05,
+      "loss": 1.8833,
+      "step": 274
+    },
+    {
+      "epoch": 0.03414876443561406,
+      "grad_norm": 0.3024221956729889,
+      "learning_rate": 5e-05,
+      "loss": 1.9978,
+      "step": 275
+    },
+    {
+      "epoch": 0.03427294176083447,
+      "grad_norm": 0.46817541122436523,
+      "learning_rate": 5e-05,
+      "loss": 2.033,
+      "step": 276
+    },
+    {
+      "epoch": 0.03439711908605489,
+      "grad_norm": 0.5574439764022827,
+      "learning_rate": 5e-05,
+      "loss": 2.0834,
+      "step": 277
+    },
+    {
+      "epoch": 0.0345212964112753,
+      "grad_norm": 0.44354552030563354,
+      "learning_rate": 5e-05,
+      "loss": 2.4408,
+      "step": 278
+    },
+    {
+      "epoch": 0.034645473736495715,
+      "grad_norm": 0.31415316462516785,
+      "learning_rate": 5e-05,
+      "loss": 2.0483,
+      "step": 279
+    },
+    {
+      "epoch": 0.03476965106171613,
+      "grad_norm": 0.38016122579574585,
+      "learning_rate": 5e-05,
+      "loss": 2.3632,
+      "step": 280
+    },
+    {
+      "epoch": 0.03489382838693655,
+      "grad_norm": 0.43614232540130615,
+      "learning_rate": 5e-05,
+      "loss": 1.9638,
+      "step": 281
+    },
+    {
+      "epoch": 0.03501800571215696,
+      "grad_norm": 0.4233233630657196,
+      "learning_rate": 5e-05,
+      "loss": 2.2086,
+      "step": 282
+    },
+    {
+      "epoch": 0.03514218303737737,
+      "grad_norm": 0.31526637077331543,
+      "learning_rate": 5e-05,
+      "loss": 1.9088,
+      "step": 283
+    },
+    {
+      "epoch": 0.03526636036259779,
+      "grad_norm": 0.2817949056625366,
+      "learning_rate": 5e-05,
+      "loss": 1.9809,
+      "step": 284
+    },
+    {
+      "epoch": 0.035390537687818205,
+      "grad_norm": 0.28902673721313477,
+      "learning_rate": 5e-05,
+      "loss": 1.9522,
+      "step": 285
+    },
+    {
+      "epoch": 0.03551471501303862,
+      "grad_norm": 0.31482502818107605,
+      "learning_rate": 5e-05,
+      "loss": 2.0892,
+      "step": 286
+    },
+    {
+      "epoch": 0.03563889233825904,
+      "grad_norm": 0.34513360261917114,
+      "learning_rate": 5e-05,
+      "loss": 1.8356,
+      "step": 287
+    },
+    {
+      "epoch": 0.035763069663479446,
+      "grad_norm": 0.25314414501190186,
+      "learning_rate": 5e-05,
+      "loss": 1.6583,
+      "step": 288
+    },
+    {
+      "epoch": 0.03588724698869986,
+      "grad_norm": 0.26934492588043213,
+      "learning_rate": 5e-05,
+      "loss": 1.8899,
+      "step": 289
+    },
+    {
+      "epoch": 0.03601142431392028,
+      "grad_norm": 0.5365657806396484,
+      "learning_rate": 5e-05,
+      "loss": 2.7175,
+      "step": 290
+    },
+    {
+      "epoch": 0.036135601639140695,
+      "grad_norm": 0.39509204030036926,
+      "learning_rate": 5e-05,
+      "loss": 2.1938,
+      "step": 291
+    },
+    {
+      "epoch": 0.036259778964361104,
+      "grad_norm": 0.3224128782749176,
+      "learning_rate": 5e-05,
+      "loss": 2.2362,
+      "step": 292
+    },
+    {
+      "epoch": 0.03638395628958152,
+      "grad_norm": 0.3229978382587433,
+      "learning_rate": 5e-05,
+      "loss": 2.2166,
+      "step": 293
+    },
+    {
+      "epoch": 0.03650813361480194,
+      "grad_norm": 0.306490421295166,
+      "learning_rate": 5e-05,
+      "loss": 1.9623,
+      "step": 294
+    },
+    {
+      "epoch": 0.03663231094002235,
+      "grad_norm": 0.2801074683666229,
+      "learning_rate": 5e-05,
+      "loss": 1.7569,
+      "step": 295
+    },
+    {
+      "epoch": 0.03675648826524277,
+      "grad_norm": 0.7114109992980957,
+      "learning_rate": 5e-05,
+      "loss": 2.7558,
+      "step": 296
+    },
+    {
+      "epoch": 0.03688066559046318,
+      "grad_norm": 0.40275782346725464,
+      "learning_rate": 5e-05,
+      "loss": 2.4591,
+      "step": 297
+    },
+    {
+      "epoch": 0.037004842915683595,
+      "grad_norm": 0.45890384912490845,
+      "learning_rate": 5e-05,
+      "loss": 2.5683,
+      "step": 298
+    },
+    {
+      "epoch": 0.03712902024090401,
+      "grad_norm": 0.2812986671924591,
+      "learning_rate": 5e-05,
+      "loss": 1.8466,
+      "step": 299
+    },
+    {
+      "epoch": 0.03725319756612443,
+      "grad_norm": 0.27165907621383667,
+      "learning_rate": 5e-05,
+      "loss": 1.8132,
+      "step": 300
+    },
+    {
+      "epoch": 0.03737737489134484,
+      "grad_norm": 0.7007853388786316,
+      "learning_rate": 5e-05,
+      "loss": 2.4039,
+      "step": 301
+    },
+    {
+      "epoch": 0.03750155221656525,
+      "grad_norm": 0.29555222392082214,
+      "learning_rate": 5e-05,
+      "loss": 1.8297,
+      "step": 302
+    },
+    {
+      "epoch": 0.03762572954178567,
+      "grad_norm": 0.8720472455024719,
+      "learning_rate": 5e-05,
+      "loss": 3.5648,
+      "step": 303
+    },
+    {
+      "epoch": 0.037749906867006085,
+      "grad_norm": 0.5265495777130127,
+      "learning_rate": 5e-05,
+      "loss": 2.3558,
+      "step": 304
+    },
+    {
+      "epoch": 0.0378740841922265,
+      "grad_norm": 0.30763059854507446,
+      "learning_rate": 5e-05,
+      "loss": 1.9394,
+      "step": 305
+    },
+    {
+      "epoch": 0.03799826151744692,
+      "grad_norm": 0.4869915246963501,
+      "learning_rate": 5e-05,
+      "loss": 2.4922,
+      "step": 306
+    },
+    {
+      "epoch": 0.038122438842667326,
+      "grad_norm": 0.4010864496231079,
+      "learning_rate": 5e-05,
+      "loss": 2.5176,
+      "step": 307
+    },
+    {
+      "epoch": 0.03824661616788774,
+      "grad_norm": 0.3895525634288788,
+      "learning_rate": 5e-05,
+      "loss": 2.0411,
+      "step": 308
+    },
+    {
+      "epoch": 0.03837079349310816,
+      "grad_norm": 0.33512821793556213,
+      "learning_rate": 5e-05,
+      "loss": 1.7361,
+      "step": 309
+    },
+    {
+      "epoch": 0.038494970818328575,
+      "grad_norm": 0.2946857810020447,
+      "learning_rate": 5e-05,
+      "loss": 2.1224,
+      "step": 310
+    },
+    {
+      "epoch": 0.03861914814354899,
+      "grad_norm": 0.2924429178237915,
+      "learning_rate": 5e-05,
+      "loss": 1.9855,
+      "step": 311
+    },
+    {
+      "epoch": 0.0387433254687694,
+      "grad_norm": 0.2950690686702728,
+      "learning_rate": 5e-05,
+      "loss": 2.0655,
+      "step": 312
+    },
+    {
+      "epoch": 0.03886750279398982,
+      "grad_norm": 0.3788645267486572,
+      "learning_rate": 5e-05,
+      "loss": 2.0447,
+      "step": 313
+    },
+    {
+      "epoch": 0.03899168011921023,
+      "grad_norm": 0.24583245813846588,
+      "learning_rate": 5e-05,
+      "loss": 1.9357,
+      "step": 314
+    },
+    {
+      "epoch": 0.03911585744443065,
+      "grad_norm": 0.25939449667930603,
+      "learning_rate": 5e-05,
+      "loss": 1.7458,
+      "step": 315
+    },
+    {
+      "epoch": 0.039240034769651065,
+      "grad_norm": 0.6960536241531372,
+      "learning_rate": 5e-05,
+      "loss": 2.9839,
+      "step": 316
+    },
+    {
+      "epoch": 0.039364212094871474,
+      "grad_norm": 0.7116987705230713,
+      "learning_rate": 5e-05,
+      "loss": 3.4002,
+      "step": 317
+    },
+    {
+      "epoch": 0.03948838942009189,
+      "grad_norm": 0.5138376355171204,
+      "learning_rate": 5e-05,
+      "loss": 2.2105,
+      "step": 318
+    },
+    {
+      "epoch": 0.03961256674531231,
+      "grad_norm": 0.35122495889663696,
+      "learning_rate": 5e-05,
+      "loss": 1.9606,
+      "step": 319
+    },
+    {
+      "epoch": 0.03973674407053272,
+      "grad_norm": 0.33248552680015564,
+      "learning_rate": 5e-05,
+      "loss": 1.8964,
+      "step": 320
+    },
+    {
+      "epoch": 0.03986092139575313,
+      "grad_norm": 0.5977162718772888,
+      "learning_rate": 5e-05,
+      "loss": 2.6293,
+      "step": 321
+    },
+    {
+      "epoch": 0.03998509872097355,
+      "grad_norm": 1.101988673210144,
+      "learning_rate": 5e-05,
+      "loss": 2.3351,
+      "step": 322
+    },
+    {
+      "epoch": 0.040109276046193965,
+      "grad_norm": 0.36033323407173157,
+      "learning_rate": 5e-05,
+      "loss": 2.0743,
+      "step": 323
+    },
+    {
+      "epoch": 0.04023345337141438,
+      "grad_norm": 0.44148921966552734,
+      "learning_rate": 5e-05,
+      "loss": 2.1855,
+      "step": 324
+    },
+    {
+      "epoch": 0.0403576306966348,
+      "grad_norm": 0.3606829345226288,
+      "learning_rate": 5e-05,
+      "loss": 1.8096,
+      "step": 325
+    },
+    {
+      "epoch": 0.040481808021855206,
+      "grad_norm": 0.25826698541641235,
+      "learning_rate": 5e-05,
+      "loss": 1.8324,
+      "step": 326
+    },
+    {
+      "epoch": 0.04060598534707562,
+      "grad_norm": 0.2834356427192688,
+      "learning_rate": 5e-05,
+      "loss": 2.0842,
+      "step": 327
+    },
+    {
+      "epoch": 0.04073016267229604,
+      "grad_norm": 0.33947786688804626,
+      "learning_rate": 5e-05,
+      "loss": 1.818,
+      "step": 328
+    },
+    {
+      "epoch": 0.040854339997516455,
+      "grad_norm": 0.31332477927207947,
+      "learning_rate": 5e-05,
+      "loss": 2.0323,
+      "step": 329
+    },
+    {
+      "epoch": 0.04097851732273687,
+      "grad_norm": 0.3210617005825043,
+      "learning_rate": 5e-05,
+      "loss": 2.1279,
+      "step": 330
+    },
+    {
+      "epoch": 0.04110269464795728,
+      "grad_norm": 0.327605664730072,
+      "learning_rate": 5e-05,
+      "loss": 2.3199,
+      "step": 331
+    },
+    {
+      "epoch": 0.041226871973177696,
+      "grad_norm": 0.4369778633117676,
+      "learning_rate": 5e-05,
+      "loss": 2.076,
+      "step": 332
+    },
+    {
+      "epoch": 0.04135104929839811,
+      "grad_norm": 1.08906888961792,
+      "learning_rate": 5e-05,
+      "loss": 2.6898,
+      "step": 333
+    },
+    {
+      "epoch": 0.04147522662361853,
+      "grad_norm": 0.27949631214141846,
+      "learning_rate": 5e-05,
+      "loss": 1.8686,
+      "step": 334
+    },
+    {
+      "epoch": 0.041599403948838945,
+      "grad_norm": 0.36476194858551025,
+      "learning_rate": 5e-05,
+      "loss": 2.1013,
+      "step": 335
+    },
+    {
+      "epoch": 0.041723581274059354,
+      "grad_norm": 0.3634919822216034,
+      "learning_rate": 5e-05,
+      "loss": 1.921,
+      "step": 336
+    },
+    {
+      "epoch": 0.04184775859927977,
+      "grad_norm": 0.8270204663276672,
+      "learning_rate": 5e-05,
+      "loss": 2.7156,
+      "step": 337
+    },
+    {
+      "epoch": 0.04197193592450019,
+      "grad_norm": 0.3705833852291107,
+      "learning_rate": 5e-05,
+      "loss": 2.3646,
+      "step": 338
+    },
+    {
+      "epoch": 0.0420961132497206,
+      "grad_norm": 0.4712885320186615,
+      "learning_rate": 5e-05,
+      "loss": 2.7343,
+      "step": 339
+    },
+    {
+      "epoch": 0.04222029057494102,
+      "grad_norm": 0.3547822833061218,
+      "learning_rate": 5e-05,
+      "loss": 2.0513,
+      "step": 340
+    },
+    {
+      "epoch": 0.04234446790016143,
+      "grad_norm": 0.344499409198761,
+      "learning_rate": 5e-05,
+      "loss": 1.8261,
+      "step": 341
+    },
+    {
+      "epoch": 0.042468645225381844,
+      "grad_norm": 0.4338517189025879,
+      "learning_rate": 5e-05,
+      "loss": 2.2454,
+      "step": 342
+    },
+    {
+      "epoch": 0.04259282255060226,
+      "grad_norm": 0.31831085681915283,
+      "learning_rate": 5e-05,
+      "loss": 2.1019,
+      "step": 343
+    },
+    {
+      "epoch": 0.04271699987582268,
+      "grad_norm": 0.985245943069458,
+      "learning_rate": 5e-05,
+      "loss": 3.4544,
+      "step": 344
+    },
+    {
+      "epoch": 0.042841177201043086,
+      "grad_norm": 0.4352297782897949,
+      "learning_rate": 5e-05,
+      "loss": 1.941,
+      "step": 345
+    },
+    {
+      "epoch": 0.0429653545262635,
+      "grad_norm": 0.42493873834609985,
+      "learning_rate": 5e-05,
+      "loss": 1.8747,
+      "step": 346
+    },
+    {
+      "epoch": 0.04308953185148392,
+      "grad_norm": 0.42205947637557983,
+      "learning_rate": 5e-05,
+      "loss": 1.6574,
+      "step": 347
+    },
+    {
+      "epoch": 0.043213709176704335,
+      "grad_norm": 0.7846937775611877,
+      "learning_rate": 5e-05,
+      "loss": 2.2254,
+      "step": 348
+    },
+    {
+      "epoch": 0.04333788650192475,
+      "grad_norm": 0.4324062764644623,
+      "learning_rate": 5e-05,
+      "loss": 2.1008,
+      "step": 349
+    },
+    {
+      "epoch": 0.04346206382714516,
+      "grad_norm": 0.35785984992980957,
+      "learning_rate": 5e-05,
+      "loss": 1.8846,
+      "step": 350
+    },
+    {
+      "epoch": 0.043586241152365576,
+      "grad_norm": 0.4233599901199341,
+      "learning_rate": 5e-05,
+      "loss": 1.858,
+      "step": 351
+    },
+    {
+      "epoch": 0.04371041847758599,
+      "grad_norm": 0.5496866106987,
+      "learning_rate": 5e-05,
+      "loss": 2.1612,
+      "step": 352
+    },
+    {
+      "epoch": 0.04383459580280641,
+      "grad_norm": 0.3902002274990082,
+      "learning_rate": 5e-05,
+      "loss": 1.9258,
+      "step": 353
+    },
+    {
+      "epoch": 0.043958773128026825,
+      "grad_norm": 0.2814697325229645,
+      "learning_rate": 5e-05,
+      "loss": 1.9894,
+      "step": 354
+    },
+    {
+      "epoch": 0.044082950453247234,
+      "grad_norm": 0.2832564115524292,
+      "learning_rate": 5e-05,
+      "loss": 2.2246,
+      "step": 355
+    },
+    {
+      "epoch": 0.04420712777846765,
+      "grad_norm": 0.35928434133529663,
+      "learning_rate": 5e-05,
+      "loss": 2.0396,
+      "step": 356
+    },
+    {
+      "epoch": 0.044331305103688066,
+      "grad_norm": 0.33969515562057495,
+      "learning_rate": 5e-05,
+      "loss": 1.7788,
+      "step": 357
+    },
+    {
+      "epoch": 0.04445548242890848,
+      "grad_norm": 0.3965449035167694,
+      "learning_rate": 5e-05,
+      "loss": 2.3086,
+      "step": 358
+    },
+    {
+      "epoch": 0.0445796597541289,
+      "grad_norm": 0.29177576303482056,
+      "learning_rate": 5e-05,
+      "loss": 2.1047,
+      "step": 359
+    },
+    {
+      "epoch": 0.04470383707934931,
+      "grad_norm": 0.32253915071487427,
+      "learning_rate": 5e-05,
+      "loss": 2.1016,
+      "step": 360
+    },
+    {
+      "epoch": 0.044828014404569724,
+      "grad_norm": 0.2820304036140442,
+      "learning_rate": 5e-05,
+      "loss": 2.0737,
+      "step": 361
+    },
+    {
+      "epoch": 0.04495219172979014,
+      "grad_norm": 0.30913272500038147,
+      "learning_rate": 5e-05,
+      "loss": 2.0834,
+      "step": 362
+    },
+    {
+      "epoch": 0.04507636905501056,
+      "grad_norm": 0.2592400014400482,
+      "learning_rate": 5e-05,
+      "loss": 1.6461,
+      "step": 363
+    },
+    {
+      "epoch": 0.04520054638023097,
+      "grad_norm": 0.3722885251045227,
+      "learning_rate": 5e-05,
+      "loss": 2.3949,
+      "step": 364
+    },
+    {
+      "epoch": 0.04532472370545138,
+      "grad_norm": 0.26727503538131714,
+      "learning_rate": 5e-05,
+      "loss": 1.9801,
+      "step": 365
+    },
+    {
+      "epoch": 0.0454489010306718,
+      "grad_norm": 0.25907137989997864,
+      "learning_rate": 5e-05,
+      "loss": 1.9168,
+      "step": 366
+    },
+    {
+      "epoch": 0.045573078355892215,
+      "grad_norm": 0.24896273016929626,
+      "learning_rate": 5e-05,
+      "loss": 1.8161,
+      "step": 367
+    },
+    {
+      "epoch": 0.04569725568111263,
+      "grad_norm": 0.45273929834365845,
+      "learning_rate": 5e-05,
+      "loss": 2.3741,
+      "step": 368
+    },
+    {
+      "epoch": 0.04582143300633305,
+      "grad_norm": 0.5102758407592773,
+      "learning_rate": 5e-05,
+      "loss": 2.1605,
+      "step": 369
+    },
+    {
+      "epoch": 0.045945610331553456,
+      "grad_norm": 0.23880641162395477,
+      "learning_rate": 5e-05,
+      "loss": 1.8322,
+      "step": 370
+    },
+    {
+      "epoch": 0.04606978765677387,
+      "grad_norm": 0.2831520736217499,
+      "learning_rate": 5e-05,
+      "loss": 1.8083,
+      "step": 371
+    },
+    {
+      "epoch": 0.04619396498199429,
+      "grad_norm": 0.27552568912506104,
+      "learning_rate": 5e-05,
+      "loss": 1.8244,
+      "step": 372
+    },
+    {
+      "epoch": 0.046318142307214705,
+      "grad_norm": 0.45094606280326843,
+      "learning_rate": 5e-05,
+      "loss": 2.4609,
+      "step": 373
+    },
+    {
+      "epoch": 0.046442319632435114,
+      "grad_norm": 0.5706016421318054,
+      "learning_rate": 5e-05,
+      "loss": 2.392,
+      "step": 374
+    },
+    {
+      "epoch": 0.04656649695765553,
+      "grad_norm": 0.35276082158088684,
+      "learning_rate": 5e-05,
+      "loss": 2.3236,
+      "step": 375
+    },
+    {
+      "epoch": 0.046690674282875946,
+      "grad_norm": 0.6116194128990173,
+      "learning_rate": 5e-05,
+      "loss": 3.0445,
+      "step": 376
+    },
+    {
+      "epoch": 0.04681485160809636,
+      "grad_norm": 0.48592257499694824,
+      "learning_rate": 5e-05,
+      "loss": 2.7122,
+      "step": 377
+    },
+    {
+      "epoch": 0.04693902893331678,
+      "grad_norm": 0.4276454746723175,
+      "learning_rate": 5e-05,
+      "loss": 1.7492,
+      "step": 378
+    },
+    {
+      "epoch": 0.04706320625853719,
+      "grad_norm": 0.3442023694515228,
+      "learning_rate": 5e-05,
+      "loss": 1.7294,
+      "step": 379
+    },
+    {
+      "epoch": 0.047187383583757604,
+      "grad_norm": 0.31899723410606384,
+      "learning_rate": 5e-05,
+      "loss": 1.847,
+      "step": 380
+    },
+    {
+      "epoch": 0.04731156090897802,
+      "grad_norm": 0.6390381455421448,
+      "learning_rate": 5e-05,
+      "loss": 2.3517,
+      "step": 381
+    },
+    {
+      "epoch": 0.04743573823419844,
+      "grad_norm": 0.2842358946800232,
+      "learning_rate": 5e-05,
+      "loss": 1.7723,
+      "step": 382
+    },
+    {
+      "epoch": 0.04755991555941885,
+      "grad_norm": 0.4072648286819458,
+      "learning_rate": 5e-05,
+      "loss": 2.07,
+      "step": 383
+    },
+    {
+      "epoch": 0.04768409288463926,
+      "grad_norm": 0.4015129506587982,
+      "learning_rate": 5e-05,
+      "loss": 2.0209,
+      "step": 384
+    },
+    {
+      "epoch": 0.04780827020985968,
+      "grad_norm": 0.3660687506198883,
+      "learning_rate": 5e-05,
+      "loss": 1.8115,
+      "step": 385
+    },
+    {
+      "epoch": 0.047932447535080094,
+      "grad_norm": 0.3126830458641052,
+      "learning_rate": 5e-05,
+      "loss": 1.9809,
+      "step": 386
+    },
+    {
+      "epoch": 0.04805662486030051,
+      "grad_norm": 0.33506837487220764,
+      "learning_rate": 5e-05,
+      "loss": 1.9708,
+      "step": 387
+    },
+    {
+      "epoch": 0.04818080218552093,
+      "grad_norm": 0.3971947431564331,
+      "learning_rate": 5e-05,
+      "loss": 2.2299,
+      "step": 388
+    },
+    {
+      "epoch": 0.048304979510741336,
+      "grad_norm": 0.5403782725334167,
+      "learning_rate": 5e-05,
+      "loss": 2.8499,
+      "step": 389
+    },
+    {
+      "epoch": 0.04842915683596175,
+      "grad_norm": 0.408307284116745,
+      "learning_rate": 5e-05,
+      "loss": 2.1051,
+      "step": 390
+    },
+    {
+      "epoch": 0.04855333416118217,
+      "grad_norm": 0.3732825219631195,
+      "learning_rate": 5e-05,
+      "loss": 2.5551,
+      "step": 391
+    },
+    {
+      "epoch": 0.048677511486402585,
+      "grad_norm": 0.2747226059436798,
+      "learning_rate": 5e-05,
+      "loss": 1.7471,
+      "step": 392
+    },
+    {
+      "epoch": 0.048801688811623,
+      "grad_norm": 0.7345656156539917,
+      "learning_rate": 5e-05,
+      "loss": 2.591,
+      "step": 393
+    },
+    {
+      "epoch": 0.04892586613684341,
+      "grad_norm": 0.33286187052726746,
+      "learning_rate": 5e-05,
+      "loss": 2.1689,
+      "step": 394
+    },
+    {
+      "epoch": 0.049050043462063826,
+      "grad_norm": 0.2857557237148285,
+      "learning_rate": 5e-05,
+      "loss": 1.8944,
+      "step": 395
+    },
+    {
+      "epoch": 0.04917422078728424,
+      "grad_norm": 0.37910234928131104,
+      "learning_rate": 5e-05,
+      "loss": 2.2506,
+      "step": 396
+    },
+    {
+      "epoch": 0.04929839811250466,
+      "grad_norm": 0.36602476239204407,
+      "learning_rate": 5e-05,
+      "loss": 2.3104,
+      "step": 397
+    },
+    {
+      "epoch": 0.049422575437725075,
+      "grad_norm": 0.28713685274124146,
+      "learning_rate": 5e-05,
+      "loss": 2.147,
+      "step": 398
+    },
+    {
+      "epoch": 0.049546752762945484,
+      "grad_norm": 0.23418322205543518,
+      "learning_rate": 5e-05,
+      "loss": 1.786,
+      "step": 399
+    },
+    {
+      "epoch": 0.0496709300881659,
+      "grad_norm": 0.31772157549858093,
+      "learning_rate": 5e-05,
+      "loss": 2.2039,
+      "step": 400
+    },
+    {
+      "epoch": 0.049795107413386316,
+      "grad_norm": 0.2687137722969055,
+      "learning_rate": 5e-05,
+      "loss": 1.8333,
+      "step": 401
+    },
+    {
+      "epoch": 0.04991928473860673,
+      "grad_norm": 0.2499142736196518,
+      "learning_rate": 5e-05,
+      "loss": 1.9173,
+      "step": 402
+    },
+    {
+      "epoch": 0.05004346206382714,
+      "grad_norm": 0.37822225689888,
+      "learning_rate": 5e-05,
+      "loss": 2.4174,
+      "step": 403
+    },
+    {
+      "epoch": 0.05016763938904756,
+      "grad_norm": 0.2855492830276489,
+      "learning_rate": 5e-05,
+      "loss": 1.7378,
+      "step": 404
+    },
+    {
+      "epoch": 0.050291816714267974,
+      "grad_norm": 0.2772412896156311,
+      "learning_rate": 5e-05,
+      "loss": 2.0178,
+      "step": 405
+    },
+    {
+      "epoch": 0.05041599403948839,
+      "grad_norm": 0.30689844489097595,
+      "learning_rate": 5e-05,
+      "loss": 1.947,
+      "step": 406
+    },
+    {
+      "epoch": 0.05054017136470881,
+      "grad_norm": 0.2882624864578247,
+      "learning_rate": 5e-05,
+      "loss": 1.844,
+      "step": 407
+    },
+    {
+      "epoch": 0.050664348689929216,
+      "grad_norm": 0.23683811724185944,
+      "learning_rate": 5e-05,
+      "loss": 1.7275,
+      "step": 408
+    },
+    {
+      "epoch": 0.05078852601514963,
+      "grad_norm": 0.5635033845901489,
+      "learning_rate": 5e-05,
+      "loss": 3.4076,
+      "step": 409
+    },
+    {
+      "epoch": 0.05091270334037005,
+      "grad_norm": 0.23519952595233917,
+      "learning_rate": 5e-05,
+      "loss": 1.6939,
+      "step": 410
+    },
+    {
+      "epoch": 0.051036880665590464,
+      "grad_norm": 0.33149218559265137,
+      "learning_rate": 5e-05,
+      "loss": 2.3945,
+      "step": 411
+    },
+    {
+      "epoch": 0.05116105799081088,
+      "grad_norm": 0.5098563432693481,
+      "learning_rate": 5e-05,
+      "loss": 2.8198,
+      "step": 412
+    },
+    {
+      "epoch": 0.05128523531603129,
+      "grad_norm": 0.23613744974136353,
+      "learning_rate": 5e-05,
+      "loss": 1.6647,
+      "step": 413
+    },
+    {
+      "epoch": 0.051409412641251706,
+      "grad_norm": 0.29657983779907227,
+      "learning_rate": 5e-05,
+      "loss": 1.7917,
+      "step": 414
+    },
+    {
+      "epoch": 0.05153358996647212,
+      "grad_norm": 0.326131671667099,
+      "learning_rate": 5e-05,
+      "loss": 1.8165,
+      "step": 415
+    },
+    {
+      "epoch": 0.05165776729169254,
+      "grad_norm": 0.39462924003601074,
+      "learning_rate": 5e-05,
+      "loss": 2.3772,
+      "step": 416
+    },
+    {
+      "epoch": 0.051781944616912955,
+      "grad_norm": 0.4868965744972229,
+      "learning_rate": 5e-05,
+      "loss": 2.624,
+      "step": 417
+    },
+    {
+      "epoch": 0.051906121942133364,
+      "grad_norm": 0.804303765296936,
+      "learning_rate": 5e-05,
+      "loss": 3.1658,
+      "step": 418
+    },
+    {
+      "epoch": 0.05203029926735378,
+      "grad_norm": 0.4080972671508789,
+      "learning_rate": 5e-05,
+      "loss": 2.3328,
+      "step": 419
+    },
+    {
+      "epoch": 0.052154476592574196,
+      "grad_norm": 0.5057448148727417,
+      "learning_rate": 5e-05,
+      "loss": 2.6605,
+      "step": 420
+    },
+    {
+      "epoch": 0.05227865391779461,
+      "grad_norm": 0.40663498640060425,
+      "learning_rate": 5e-05,
+      "loss": 2.2624,
+      "step": 421
+    },
+    {
+      "epoch": 0.05240283124301503,
+      "grad_norm": 0.27341869473457336,
+      "learning_rate": 5e-05,
+      "loss": 1.9175,
+      "step": 422
+    },
+    {
+      "epoch": 0.05252700856823544,
+      "grad_norm": 0.27328404784202576,
+      "learning_rate": 5e-05,
+      "loss": 2.1219,
+      "step": 423
+    },
+    {
+      "epoch": 0.052651185893455854,
+      "grad_norm": 0.6359747052192688,
+      "learning_rate": 5e-05,
+      "loss": 3.0013,
+      "step": 424
+    },
+    {
+      "epoch": 0.05277536321867627,
+      "grad_norm": 0.274902880191803,
+      "learning_rate": 5e-05,
+      "loss": 1.9396,
+      "step": 425
+    },
+    {
+      "epoch": 0.052899540543896686,
+      "grad_norm": 0.41854164004325867,
+      "learning_rate": 5e-05,
+      "loss": 2.5872,
+      "step": 426
+    },
+    {
+      "epoch": 0.0530237178691171,
+      "grad_norm": 0.28985920548439026,
+      "learning_rate": 5e-05,
+      "loss": 2.1378,
+      "step": 427
+    },
+    {
+      "epoch": 0.05314789519433751,
+      "grad_norm": 0.3510410785675049,
+      "learning_rate": 5e-05,
+      "loss": 2.4868,
+      "step": 428
+    },
+    {
+      "epoch": 0.05327207251955793,
+      "grad_norm": 0.4156201481819153,
+      "learning_rate": 5e-05,
+      "loss": 2.2873,
+      "step": 429
+    },
+    {
+      "epoch": 0.053396249844778344,
+      "grad_norm": 0.3104117512702942,
+      "learning_rate": 5e-05,
+      "loss": 2.0735,
+      "step": 430
+    },
+    {
+      "epoch": 0.05352042716999876,
+      "grad_norm": 0.3024173080921173,
+      "learning_rate": 5e-05,
+      "loss": 2.0158,
+      "step": 431
+    },
+    {
+      "epoch": 0.05364460449521917,
+      "grad_norm": 0.2899841070175171,
+      "learning_rate": 5e-05,
+      "loss": 1.955,
+      "step": 432
+    },
+    {
+      "epoch": 0.053768781820439586,
+      "grad_norm": 0.5953050851821899,
+      "learning_rate": 5e-05,
+      "loss": 2.7843,
+      "step": 433
+    },
+    {
+      "epoch": 0.05389295914566,
+      "grad_norm": 0.33898434042930603,
+      "learning_rate": 5e-05,
+      "loss": 2.2594,
+      "step": 434
+    },
+    {
+      "epoch": 0.05401713647088042,
+      "grad_norm": 0.2735304534435272,
+      "learning_rate": 5e-05,
+      "loss": 2.0111,
+      "step": 435
+    },
+    {
+      "epoch": 0.054141313796100834,
+      "grad_norm": 0.2993002235889435,
+      "learning_rate": 5e-05,
+      "loss": 1.8185,
+      "step": 436
+    },
+    {
+      "epoch": 0.054265491121321244,
+      "grad_norm": 0.6330844759941101,
+      "learning_rate": 5e-05,
+      "loss": 2.8258,
+      "step": 437
+    },
+    {
+      "epoch": 0.05438966844654166,
+      "grad_norm": 0.736097514629364,
+      "learning_rate": 5e-05,
+      "loss": 2.1661,
+      "step": 438
+    },
+    {
+      "epoch": 0.054513845771762076,
+      "grad_norm": 0.34374523162841797,
+      "learning_rate": 5e-05,
+      "loss": 1.9379,
+      "step": 439
+    },
+    {
+      "epoch": 0.05463802309698249,
+      "grad_norm": 0.6117690801620483,
+      "learning_rate": 5e-05,
+      "loss": 2.346,
+      "step": 440
+    },
+    {
+      "epoch": 0.05476220042220291,
+      "grad_norm": 0.35637909173965454,
+      "learning_rate": 5e-05,
+      "loss": 1.9468,
+      "step": 441
+    },
+    {
+      "epoch": 0.05488637774742332,
+      "grad_norm": 0.3755735456943512,
+      "learning_rate": 5e-05,
+      "loss": 2.0639,
+      "step": 442
+    },
+    {
+      "epoch": 0.055010555072643734,
+      "grad_norm": 0.28550875186920166,
+      "learning_rate": 5e-05,
+      "loss": 1.9211,
+      "step": 443
+    },
+    {
+      "epoch": 0.05513473239786415,
+      "grad_norm": 0.25539669394493103,
+      "learning_rate": 5e-05,
+      "loss": 1.7414,
+      "step": 444
+    },
+    {
+      "epoch": 0.055258909723084566,
+      "grad_norm": 0.27155032753944397,
+      "learning_rate": 5e-05,
+      "loss": 1.686,
+      "step": 445
+    },
+    {
+      "epoch": 0.05538308704830498,
+      "grad_norm": 0.38703253865242004,
+      "learning_rate": 5e-05,
+      "loss": 2.264,
+      "step": 446
+    },
+    {
+      "epoch": 0.05550726437352539,
+      "grad_norm": 0.48445382714271545,
+      "learning_rate": 5e-05,
+      "loss": 2.2648,
+      "step": 447
+    },
+    {
+      "epoch": 0.05563144169874581,
+      "grad_norm": 0.3057152032852173,
+      "learning_rate": 5e-05,
+      "loss": 1.7933,
+      "step": 448
+    },
+    {
+      "epoch": 0.055755619023966224,
+      "grad_norm": 0.3264656066894531,
+      "learning_rate": 5e-05,
+      "loss": 1.981,
+      "step": 449
+    },
+    {
+      "epoch": 0.05587979634918664,
+      "grad_norm": 0.280307412147522,
+      "learning_rate": 5e-05,
+      "loss": 1.9146,
+      "step": 450
+    },
+    {
+      "epoch": 0.05600397367440706,
+      "grad_norm": 0.2741982638835907,
+      "learning_rate": 5e-05,
+      "loss": 1.8801,
+      "step": 451
+    },
+    {
+      "epoch": 0.056128150999627466,
+      "grad_norm": 0.6853694319725037,
+      "learning_rate": 5e-05,
+      "loss": 2.9746,
+      "step": 452
+    },
+    {
+      "epoch": 0.05625232832484788,
+      "grad_norm": 0.2608783543109894,
+      "learning_rate": 5e-05,
+      "loss": 1.8815,
+      "step": 453
+    },
+    {
+      "epoch": 0.0563765056500683,
+      "grad_norm": 0.38988515734672546,
+      "learning_rate": 5e-05,
+      "loss": 2.1393,
+      "step": 454
+    },
+    {
+      "epoch": 0.056500682975288714,
+      "grad_norm": 0.3347717225551605,
+      "learning_rate": 5e-05,
+      "loss": 2.1852,
+      "step": 455
+    },
+    {
+      "epoch": 0.056624860300509124,
+      "grad_norm": 0.3986371159553528,
+      "learning_rate": 5e-05,
+      "loss": 2.7134,
+      "step": 456
+    },
+    {
+      "epoch": 0.05674903762572954,
+      "grad_norm": 0.3521873950958252,
+      "learning_rate": 5e-05,
+      "loss": 2.6786,
+      "step": 457
+    },
+    {
+      "epoch": 0.056873214950949956,
+      "grad_norm": 0.3290453851222992,
+      "learning_rate": 5e-05,
+      "loss": 2.029,
+      "step": 458
+    },
+    {
+      "epoch": 0.05699739227617037,
+      "grad_norm": 0.3097821772098541,
+      "learning_rate": 5e-05,
+      "loss": 1.9485,
+      "step": 459
+    },
+    {
+      "epoch": 0.05712156960139079,
+      "grad_norm": 0.28530988097190857,
+      "learning_rate": 5e-05,
+      "loss": 2.1007,
+      "step": 460
+    },
+    {
+      "epoch": 0.0572457469266112,
+      "grad_norm": 0.24027346074581146,
+      "learning_rate": 5e-05,
+      "loss": 1.7516,
+      "step": 461
+    },
+    {
+      "epoch": 0.057369924251831614,
+      "grad_norm": 0.49746012687683105,
+      "learning_rate": 5e-05,
+      "loss": 2.2849,
+      "step": 462
+    },
+    {
+      "epoch": 0.05749410157705203,
+      "grad_norm": 0.2805483043193817,
+      "learning_rate": 5e-05,
+      "loss": 2.1191,
+      "step": 463
+    },
+    {
+      "epoch": 0.057618278902272446,
+      "grad_norm": 0.7454423308372498,
+      "learning_rate": 5e-05,
+      "loss": 2.7142,
+      "step": 464
+    },
+    {
+      "epoch": 0.05774245622749286,
+      "grad_norm": 0.26804909110069275,
+      "learning_rate": 5e-05,
+      "loss": 1.9686,
+      "step": 465
+    },
+    {
+      "epoch": 0.05786663355271327,
+      "grad_norm": 0.4737478792667389,
+      "learning_rate": 5e-05,
+      "loss": 2.5153,
+      "step": 466
+    },
+    {
+      "epoch": 0.05799081087793369,
+      "grad_norm": 0.2857726514339447,
+      "learning_rate": 5e-05,
+      "loss": 2.0748,
+      "step": 467
+    },
+    {
+      "epoch": 0.058114988203154104,
+      "grad_norm": 0.29102930426597595,
+      "learning_rate": 5e-05,
+      "loss": 1.8484,
+      "step": 468
+    },
+    {
+      "epoch": 0.05823916552837452,
+      "grad_norm": 0.34573471546173096,
+      "learning_rate": 5e-05,
+      "loss": 2.5487,
+      "step": 469
+    },
+    {
+      "epoch": 0.058363342853594936,
+      "grad_norm": 0.25077787041664124,
+      "learning_rate": 5e-05,
+      "loss": 1.771,
+      "step": 470
+    },
+    {
+      "epoch": 0.058487520178815346,
+      "grad_norm": 0.38156285881996155,
+      "learning_rate": 5e-05,
+      "loss": 2.1605,
+      "step": 471
+    },
+    {
+      "epoch": 0.05861169750403576,
+      "grad_norm": 0.28113090991973877,
+      "learning_rate": 5e-05,
+      "loss": 2.1714,
+      "step": 472
+    },
+    {
+      "epoch": 0.05873587482925618,
+      "grad_norm": 0.2667486369609833,
+      "learning_rate": 5e-05,
+      "loss": 2.066,
+      "step": 473
+    },
+    {
+      "epoch": 0.058860052154476594,
+      "grad_norm": 0.25925296545028687,
+      "learning_rate": 5e-05,
+      "loss": 1.8001,
+      "step": 474
+    },
+    {
+      "epoch": 0.05898422947969701,
+      "grad_norm": 0.28931882977485657,
+      "learning_rate": 5e-05,
+      "loss": 2.1015,
+      "step": 475
+    },
+    {
+      "epoch": 0.05910840680491742,
+      "grad_norm": 0.26327359676361084,
+      "learning_rate": 5e-05,
+      "loss": 1.9302,
+      "step": 476
+    },
+    {
+      "epoch": 0.059232584130137836,
+      "grad_norm": 0.31456077098846436,
+      "learning_rate": 5e-05,
+      "loss": 1.986,
+      "step": 477
+    },
+    {
+      "epoch": 0.05935676145535825,
+      "grad_norm": 0.2656742036342621,
+      "learning_rate": 5e-05,
+      "loss": 1.9974,
+      "step": 478
+    },
+    {
+      "epoch": 0.05948093878057867,
+      "grad_norm": 0.7194716334342957,
+      "learning_rate": 5e-05,
+      "loss": 1.9918,
+      "step": 479
+    },
+    {
+      "epoch": 0.059605116105799084,
+      "grad_norm": 0.3521345853805542,
+      "learning_rate": 5e-05,
+      "loss": 2.1788,
+      "step": 480
+    },
+    {
+      "epoch": 0.059729293431019494,
+      "grad_norm": 0.3286682665348053,
+      "learning_rate": 5e-05,
+      "loss": 2.1944,
+      "step": 481
+    },
+    {
+      "epoch": 0.05985347075623991,
+      "grad_norm": 0.3449973464012146,
+      "learning_rate": 5e-05,
+      "loss": 2.0522,
+      "step": 482
+    },
+    {
+      "epoch": 0.059977648081460326,
+      "grad_norm": 0.28249526023864746,
+      "learning_rate": 5e-05,
+      "loss": 2.1656,
+      "step": 483
+    },
+    {
+      "epoch": 0.06010182540668074,
+      "grad_norm": 0.7991644740104675,
+      "learning_rate": 5e-05,
+      "loss": 2.122,
+      "step": 484
+    },
+    {
+      "epoch": 0.06022600273190115,
+      "grad_norm": 0.5220968723297119,
+      "learning_rate": 5e-05,
+      "loss": 2.3525,
+      "step": 485
+    },
+    {
+      "epoch": 0.06035018005712157,
+      "grad_norm": 0.35750848054885864,
+      "learning_rate": 5e-05,
+      "loss": 2.7801,
+      "step": 486
+    },
+    {
+      "epoch": 0.060474357382341984,
+      "grad_norm": 0.5980648398399353,
+      "learning_rate": 5e-05,
+      "loss": 3.0898,
+      "step": 487
+    },
+    {
+      "epoch": 0.0605985347075624,
+      "grad_norm": 0.39209190011024475,
+      "learning_rate": 5e-05,
+      "loss": 2.3064,
+      "step": 488
+    },
+    {
+      "epoch": 0.060722712032782816,
+      "grad_norm": 0.3195672035217285,
+      "learning_rate": 5e-05,
+      "loss": 1.9504,
+      "step": 489
+    },
+    {
+      "epoch": 0.060846889358003226,
+      "grad_norm": 0.6777394413948059,
+      "learning_rate": 5e-05,
+      "loss": 2.6206,
+      "step": 490
+    },
+    {
+      "epoch": 0.06097106668322364,
+      "grad_norm": 0.37489867210388184,
+      "learning_rate": 5e-05,
+      "loss": 2.2357,
+      "step": 491
+    },
+    {
+      "epoch": 0.06109524400844406,
+      "grad_norm": 0.3548474907875061,
+      "learning_rate": 5e-05,
+      "loss": 1.8687,
+      "step": 492
+    },
+    {
+      "epoch": 0.061219421333664474,
+      "grad_norm": 0.3291890621185303,
+      "learning_rate": 5e-05,
+      "loss": 2.3134,
+      "step": 493
+    },
+    {
+      "epoch": 0.06134359865888489,
+      "grad_norm": 0.3681256175041199,
+      "learning_rate": 5e-05,
+      "loss": 2.2221,
+      "step": 494
+    },
+    {
+      "epoch": 0.0614677759841053,
+      "grad_norm": 0.5733017921447754,
+      "learning_rate": 5e-05,
+      "loss": 2.5418,
+      "step": 495
+    },
+    {
+      "epoch": 0.061591953309325716,
+      "grad_norm": 0.4781680107116699,
+      "learning_rate": 5e-05,
+      "loss": 2.7054,
+      "step": 496
+    },
+    {
+      "epoch": 0.06171613063454613,
+      "grad_norm": 0.2734091877937317,
+      "learning_rate": 5e-05,
+      "loss": 1.8827,
+      "step": 497
+    },
+    {
+      "epoch": 0.06184030795976655,
+      "grad_norm": 0.38738420605659485,
+      "learning_rate": 5e-05,
+      "loss": 2.6568,
+      "step": 498
+    },
+    {
+      "epoch": 0.061964485284986964,
+      "grad_norm": 0.46199506521224976,
+      "learning_rate": 5e-05,
+      "loss": 2.7171,
+      "step": 499
+    },
+    {
+      "epoch": 0.062088662610207374,
+      "grad_norm": 0.2888731062412262,
+      "learning_rate": 5e-05,
+      "loss": 1.8164,
+      "step": 500
+    },
+    {
+      "epoch": 0.06221283993542779,
+      "grad_norm": 0.30647486448287964,
+      "learning_rate": 5e-05,
+      "loss": 1.9239,
+      "step": 501
+    },
+    {
+      "epoch": 0.062337017260648206,
+      "grad_norm": 0.33426958322525024,
+      "learning_rate": 5e-05,
+      "loss": 2.1038,
+      "step": 502
+    },
+    {
+      "epoch": 0.06246119458586862,
+      "grad_norm": 0.26133787631988525,
+      "learning_rate": 5e-05,
+      "loss": 1.888,
+      "step": 503
+    },
+    {
+      "epoch": 0.06258537191108904,
+      "grad_norm": 0.25062304735183716,
+      "learning_rate": 5e-05,
+      "loss": 2.1271,
+      "step": 504
+    },
+    {
+      "epoch": 0.06270954923630945,
+      "grad_norm": 0.23644529283046722,
+      "learning_rate": 5e-05,
+      "loss": 1.6674,
+      "step": 505
+    },
+    {
+      "epoch": 0.06283372656152987,
+      "grad_norm": 0.24004316329956055,
+      "learning_rate": 5e-05,
+      "loss": 1.7753,
+      "step": 506
+    },
+    {
+      "epoch": 0.06295790388675028,
+      "grad_norm": 0.2889292240142822,
+      "learning_rate": 5e-05,
+      "loss": 1.8082,
+      "step": 507
+    },
+    {
+      "epoch": 0.06308208121197069,
+      "grad_norm": 0.3024662137031555,
+      "learning_rate": 5e-05,
+      "loss": 1.9739,
+      "step": 508
+    },
+    {
+      "epoch": 0.06320625853719111,
+      "grad_norm": 0.29193171858787537,
+      "learning_rate": 5e-05,
+      "loss": 2.0238,
+      "step": 509
+    },
+    {
+      "epoch": 0.06333043586241152,
+      "grad_norm": 0.2500467002391815,
+      "learning_rate": 5e-05,
+      "loss": 1.8968,
+      "step": 510
+    },
+    {
+      "epoch": 0.06345461318763194,
+      "grad_norm": 0.25340279936790466,
+      "learning_rate": 5e-05,
+      "loss": 1.9851,
+      "step": 511
+    },
+    {
+      "epoch": 0.06357879051285235,
+      "grad_norm": 0.3032784163951874,
+      "learning_rate": 5e-05,
+      "loss": 2.2362,
+      "step": 512
+    },
+    {
+      "epoch": 0.06370296783807276,
+      "grad_norm": 0.23703156411647797,
+      "learning_rate": 5e-05,
+      "loss": 1.8264,
+      "step": 513
+    },
+    {
+      "epoch": 0.06382714516329319,
+      "grad_norm": 0.24125397205352783,
+      "learning_rate": 5e-05,
+      "loss": 1.805,
+      "step": 514
+    },
+    {
+      "epoch": 0.0639513224885136,
+      "grad_norm": 0.27547687292099,
+      "learning_rate": 5e-05,
+      "loss": 1.8905,
+      "step": 515
+    },
+    {
+      "epoch": 0.06407549981373402,
+      "grad_norm": 0.3540593981742859,
+      "learning_rate": 5e-05,
+      "loss": 2.4013,
+      "step": 516
+    },
+    {
+      "epoch": 0.06419967713895443,
+      "grad_norm": 0.24197076261043549,
+      "learning_rate": 5e-05,
+      "loss": 1.7991,
+      "step": 517
+    },
+    {
+      "epoch": 0.06432385446417484,
+      "grad_norm": 0.24468238651752472,
+      "learning_rate": 5e-05,
+      "loss": 1.7627,
+      "step": 518
+    },
+    {
+      "epoch": 0.06444803178939526,
+      "grad_norm": 0.4829714298248291,
+      "learning_rate": 5e-05,
+      "loss": 2.6352,
+      "step": 519
+    },
+    {
+      "epoch": 0.06457220911461567,
+      "grad_norm": 0.26643821597099304,
+      "learning_rate": 5e-05,
+      "loss": 1.921,
+      "step": 520
+    },
+    {
+      "epoch": 0.06469638643983609,
+      "grad_norm": 0.2687952220439911,
+      "learning_rate": 5e-05,
+      "loss": 1.6274,
+      "step": 521
+    },
+    {
+      "epoch": 0.0648205637650565,
+      "grad_norm": 0.4201943874359131,
+      "learning_rate": 5e-05,
+      "loss": 2.5597,
+      "step": 522
+    },
+    {
+      "epoch": 0.06494474109027691,
+      "grad_norm": 0.406886488199234,
+      "learning_rate": 5e-05,
+      "loss": 2.2164,
+      "step": 523
+    },
+    {
+      "epoch": 0.06506891841549733,
+      "grad_norm": 0.33637088537216187,
+      "learning_rate": 5e-05,
+      "loss": 2.0433,
+      "step": 524
+    },
+    {
+      "epoch": 0.06519309574071774,
+      "grad_norm": 0.25890302658081055,
+      "learning_rate": 5e-05,
+      "loss": 1.904,
+      "step": 525
+    },
+    {
+      "epoch": 0.06531727306593815,
+      "grad_norm": 0.2726302742958069,
+      "learning_rate": 5e-05,
+      "loss": 1.9808,
+      "step": 526
+    },
+    {
+      "epoch": 0.06544145039115858,
+      "grad_norm": 0.2803145945072174,
+      "learning_rate": 5e-05,
+      "loss": 1.6989,
+      "step": 527
+    },
+    {
+      "epoch": 0.06556562771637899,
+      "grad_norm": 0.41411343216896057,
+      "learning_rate": 5e-05,
+      "loss": 2.27,
+      "step": 528
+    },
+    {
+      "epoch": 0.06568980504159941,
+      "grad_norm": 0.2928730249404907,
+      "learning_rate": 5e-05,
+      "loss": 1.9875,
+      "step": 529
+    },
+    {
+      "epoch": 0.06581398236681982,
+      "grad_norm": 0.29995816946029663,
+      "learning_rate": 5e-05,
+      "loss": 2.0455,
+      "step": 530
+    },
+    {
+      "epoch": 0.06593815969204023,
+      "grad_norm": 0.7726313471794128,
+      "learning_rate": 5e-05,
+      "loss": 3.541,
+      "step": 531
+    },
+    {
+      "epoch": 0.06606233701726065,
+      "grad_norm": 0.2529071569442749,
+      "learning_rate": 5e-05,
+      "loss": 2.0551,
+      "step": 532
+    },
+    {
+      "epoch": 0.06618651434248106,
+      "grad_norm": 0.32561784982681274,
+      "learning_rate": 5e-05,
+      "loss": 1.9279,
+      "step": 533
+    },
+    {
+      "epoch": 0.06631069166770148,
+      "grad_norm": 0.29991114139556885,
+      "learning_rate": 5e-05,
+      "loss": 1.9804,
+      "step": 534
+    },
+    {
+      "epoch": 0.06643486899292189,
+      "grad_norm": 0.2466021627187729,
+      "learning_rate": 5e-05,
+      "loss": 1.7065,
+      "step": 535
+    },
+    {
+      "epoch": 0.0665590463181423,
+      "grad_norm": 0.6933873295783997,
+      "learning_rate": 5e-05,
+      "loss": 2.6648,
+      "step": 536
+    },
+    {
+      "epoch": 0.06668322364336272,
+      "grad_norm": 0.23817643523216248,
+      "learning_rate": 5e-05,
+      "loss": 1.6139,
+      "step": 537
+    },
+    {
+      "epoch": 0.06680740096858313,
+      "grad_norm": 0.27212318778038025,
+      "learning_rate": 5e-05,
+      "loss": 2.012,
+      "step": 538
+    },
+    {
+      "epoch": 0.06693157829380356,
+      "grad_norm": 0.46541309356689453,
+      "learning_rate": 5e-05,
+      "loss": 2.6264,
+      "step": 539
+    },
+    {
+      "epoch": 0.06705575561902397,
+      "grad_norm": 0.418171763420105,
+      "learning_rate": 5e-05,
+      "loss": 2.7678,
+      "step": 540
+    },
+    {
+      "epoch": 0.06717993294424437,
+      "grad_norm": 0.4368995130062103,
+      "learning_rate": 5e-05,
+      "loss": 2.4146,
+      "step": 541
+    },
+    {
+      "epoch": 0.0673041102694648,
+      "grad_norm": 0.35557737946510315,
+      "learning_rate": 5e-05,
+      "loss": 1.967,
+      "step": 542
+    },
+    {
+      "epoch": 0.06742828759468521,
+      "grad_norm": 0.2717956006526947,
+      "learning_rate": 5e-05,
+      "loss": 2.0497,
+      "step": 543
+    },
+    {
+      "epoch": 0.06755246491990563,
+      "grad_norm": 0.29945579171180725,
+      "learning_rate": 5e-05,
+      "loss": 2.3428,
+      "step": 544
+    },
+    {
+      "epoch": 0.06767664224512604,
+      "grad_norm": 0.3052377700805664,
+      "learning_rate": 5e-05,
+      "loss": 2.214,
+      "step": 545
+    },
+    {
+      "epoch": 0.06780081957034645,
+      "grad_norm": 0.44460761547088623,
+      "learning_rate": 5e-05,
+      "loss": 3.1489,
+      "step": 546
+    },
+    {
+      "epoch": 0.06792499689556687,
+      "grad_norm": 0.26404836773872375,
+      "learning_rate": 5e-05,
+      "loss": 2.0027,
+      "step": 547
+    },
+    {
+      "epoch": 0.06804917422078728,
+      "grad_norm": 0.23541861772537231,
+      "learning_rate": 5e-05,
+      "loss": 1.8127,
+      "step": 548
+    },
+    {
+      "epoch": 0.0681733515460077,
+      "grad_norm": 0.3142141103744507,
+      "learning_rate": 5e-05,
+      "loss": 1.8443,
+      "step": 549
+    },
+    {
+      "epoch": 0.06829752887122811,
+      "grad_norm": 0.3869318664073944,
+      "learning_rate": 5e-05,
+      "loss": 2.2449,
+      "step": 550
+    },
+    {
+      "epoch": 0.06842170619644852,
+      "grad_norm": 0.3425566256046295,
+      "learning_rate": 5e-05,
+      "loss": 2.6453,
+      "step": 551
+    },
+    {
+      "epoch": 0.06854588352166895,
+      "grad_norm": 0.25234097242355347,
+      "learning_rate": 5e-05,
+      "loss": 1.986,
+      "step": 552
+    },
+    {
+      "epoch": 0.06867006084688936,
+      "grad_norm": 0.4374043941497803,
+      "learning_rate": 5e-05,
+      "loss": 2.7752,
+      "step": 553
+    },
+    {
+      "epoch": 0.06879423817210978,
+      "grad_norm": 0.3221331238746643,
+      "learning_rate": 5e-05,
+      "loss": 2.6341,
+      "step": 554
+    },
+    {
+      "epoch": 0.06891841549733019,
+      "grad_norm": 0.26285409927368164,
+      "learning_rate": 5e-05,
+      "loss": 2.0298,
+      "step": 555
+    },
+    {
+      "epoch": 0.0690425928225506,
+      "grad_norm": 0.2963787317276001,
+      "learning_rate": 5e-05,
+      "loss": 2.0899,
+      "step": 556
+    },
+    {
+      "epoch": 0.06916677014777102,
+      "grad_norm": 0.2507532835006714,
+      "learning_rate": 5e-05,
+      "loss": 1.928,
+      "step": 557
+    },
+    {
+      "epoch": 0.06929094747299143,
+      "grad_norm": 0.2756315767765045,
+      "learning_rate": 5e-05,
+      "loss": 1.9318,
+      "step": 558
+    },
+    {
+      "epoch": 0.06941512479821185,
+      "grad_norm": 0.3374318778514862,
+      "learning_rate": 5e-05,
+      "loss": 2.4675,
+      "step": 559
+    },
+    {
+      "epoch": 0.06953930212343226,
+      "grad_norm": 0.3347979187965393,
+      "learning_rate": 5e-05,
+      "loss": 2.1398,
+      "step": 560
+    },
+    {
+      "epoch": 0.06966347944865267,
+      "grad_norm": 0.31123411655426025,
+      "learning_rate": 5e-05,
+      "loss": 1.8448,
+      "step": 561
+    },
+    {
+      "epoch": 0.0697876567738731,
+      "grad_norm": 0.3514026701450348,
+      "learning_rate": 5e-05,
+      "loss": 2.3298,
+      "step": 562
+    },
+    {
+      "epoch": 0.0699118340990935,
+      "grad_norm": 0.24323804676532745,
+      "learning_rate": 5e-05,
+      "loss": 1.7833,
+      "step": 563
+    },
+    {
+      "epoch": 0.07003601142431393,
+      "grad_norm": 0.37998902797698975,
+      "learning_rate": 5e-05,
+      "loss": 2.1883,
+      "step": 564
+    },
+    {
+      "epoch": 0.07016018874953434,
+      "grad_norm": 0.27929794788360596,
+      "learning_rate": 5e-05,
+      "loss": 1.8131,
+      "step": 565
+    },
+    {
+      "epoch": 0.07028436607475474,
+      "grad_norm": 0.23255369067192078,
+      "learning_rate": 5e-05,
+      "loss": 1.6527,
+      "step": 566
+    },
+    {
+      "epoch": 0.07040854339997517,
+      "grad_norm": 0.3981661796569824,
+      "learning_rate": 5e-05,
+      "loss": 2.4643,
+      "step": 567
+    },
+    {
+      "epoch": 0.07053272072519558,
+      "grad_norm": 0.2854522466659546,
+      "learning_rate": 5e-05,
+      "loss": 2.0035,
+      "step": 568
+    },
+    {
+      "epoch": 0.070656898050416,
+      "grad_norm": 0.34047701954841614,
+      "learning_rate": 5e-05,
+      "loss": 2.3843,
+      "step": 569
+    },
+    {
+      "epoch": 0.07078107537563641,
+      "grad_norm": 0.2492959201335907,
+      "learning_rate": 5e-05,
+      "loss": 2.0595,
+      "step": 570
+    },
+    {
+      "epoch": 0.07090525270085682,
+      "grad_norm": 0.28671765327453613,
+      "learning_rate": 5e-05,
+      "loss": 2.0709,
+      "step": 571
+    },
+    {
+      "epoch": 0.07102943002607724,
+      "grad_norm": 0.29164648056030273,
+      "learning_rate": 5e-05,
+      "loss": 2.2407,
+      "step": 572
+    },
+    {
+      "epoch": 0.07115360735129765,
+      "grad_norm": 0.47018054127693176,
+      "learning_rate": 5e-05,
+      "loss": 2.669,
+      "step": 573
+    },
+    {
+      "epoch": 0.07127778467651807,
+      "grad_norm": 0.24436764419078827,
+      "learning_rate": 5e-05,
+      "loss": 1.663,
+      "step": 574
+    },
+    {
+      "epoch": 0.07140196200173848,
+      "grad_norm": 0.25208258628845215,
+      "learning_rate": 5e-05,
+      "loss": 1.8464,
+      "step": 575
+    },
+    {
+      "epoch": 0.07152613932695889,
+      "grad_norm": 0.25105947256088257,
+      "learning_rate": 5e-05,
+      "loss": 2.0068,
+      "step": 576
+    },
+    {
+      "epoch": 0.07165031665217932,
+      "grad_norm": 0.4490038752555847,
+      "learning_rate": 5e-05,
+      "loss": 2.7631,
+      "step": 577
+    },
+    {
+      "epoch": 0.07177449397739973,
+      "grad_norm": 0.27281132340431213,
+      "learning_rate": 5e-05,
+      "loss": 2.0265,
+      "step": 578
+    },
+    {
+      "epoch": 0.07189867130262015,
+      "grad_norm": 0.38831856846809387,
+      "learning_rate": 5e-05,
+      "loss": 2.8805,
+      "step": 579
+    },
+    {
+      "epoch": 0.07202284862784056,
+      "grad_norm": 0.2941652834415436,
+      "learning_rate": 5e-05,
+      "loss": 2.0559,
+      "step": 580
+    },
+    {
+      "epoch": 0.07214702595306097,
+      "grad_norm": 0.307485967874527,
+      "learning_rate": 5e-05,
+      "loss": 2.225,
+      "step": 581
+    },
+    {
+      "epoch": 0.07227120327828139,
+      "grad_norm": 0.24915628135204315,
+      "learning_rate": 5e-05,
+      "loss": 1.6885,
+      "step": 582
+    },
+    {
+      "epoch": 0.0723953806035018,
+      "grad_norm": 0.49729689955711365,
+      "learning_rate": 5e-05,
+      "loss": 2.7455,
+      "step": 583
+    },
+    {
+      "epoch": 0.07251955792872221,
+      "grad_norm": 0.3003414571285248,
+      "learning_rate": 5e-05,
+      "loss": 1.8831,
+      "step": 584
+    },
+    {
+      "epoch": 0.07264373525394263,
+      "grad_norm": 0.307110995054245,
+      "learning_rate": 5e-05,
+      "loss": 1.8277,
+      "step": 585
+    },
+    {
+      "epoch": 0.07276791257916304,
+      "grad_norm": 0.6576548218727112,
+      "learning_rate": 5e-05,
+      "loss": 3.3871,
+      "step": 586
+    },
+    {
+      "epoch": 0.07289208990438346,
+      "grad_norm": 0.29746583104133606,
+      "learning_rate": 5e-05,
+      "loss": 2.4285,
+      "step": 587
+    },
+    {
+      "epoch": 0.07301626722960387,
+      "grad_norm": 0.3882395327091217,
+      "learning_rate": 5e-05,
+      "loss": 2.1912,
+      "step": 588
+    },
+    {
+      "epoch": 0.07314044455482428,
+      "grad_norm": 0.3972069025039673,
+      "learning_rate": 5e-05,
+      "loss": 1.9297,
+      "step": 589
+    },
+    {
+      "epoch": 0.0732646218800447,
+      "grad_norm": 0.29125985503196716,
+      "learning_rate": 5e-05,
+      "loss": 1.7165,
+      "step": 590
+    },
+    {
+      "epoch": 0.07338879920526512,
+      "grad_norm": 0.35590019822120667,
+      "learning_rate": 5e-05,
+      "loss": 2.0228,
+      "step": 591
+    },
+    {
+      "epoch": 0.07351297653048554,
+      "grad_norm": 0.2818076014518738,
+      "learning_rate": 5e-05,
+      "loss": 2.1202,
+      "step": 592
+    },
+    {
+      "epoch": 0.07363715385570595,
+      "grad_norm": 0.32424747943878174,
+      "learning_rate": 5e-05,
+      "loss": 2.0989,
+      "step": 593
+    },
+    {
+      "epoch": 0.07376133118092636,
+      "grad_norm": 0.30949074029922485,
+      "learning_rate": 5e-05,
+      "loss": 1.7198,
+      "step": 594
+    },
+    {
+      "epoch": 0.07388550850614678,
+      "grad_norm": 0.3237742483615875,
+      "learning_rate": 5e-05,
+      "loss": 2.0506,
+      "step": 595
+    },
+    {
+      "epoch": 0.07400968583136719,
+      "grad_norm": 0.3861253559589386,
+      "learning_rate": 5e-05,
+      "loss": 2.5358,
+      "step": 596
+    },
+    {
+      "epoch": 0.07413386315658761,
+      "grad_norm": 0.2853670120239258,
+      "learning_rate": 5e-05,
+      "loss": 2.0617,
+      "step": 597
+    },
+    {
+      "epoch": 0.07425804048180802,
+      "grad_norm": 0.2880936563014984,
+      "learning_rate": 5e-05,
+      "loss": 2.3797,
+      "step": 598
+    },
+    {
+      "epoch": 0.07438221780702843,
+      "grad_norm": 0.48230239748954773,
+      "learning_rate": 5e-05,
+      "loss": 2.4877,
+      "step": 599
+    },
+    {
+      "epoch": 0.07450639513224885,
+      "grad_norm": 0.3733183443546295,
+      "learning_rate": 5e-05,
+      "loss": 2.1438,
+      "step": 600
+    },
+    {
+      "epoch": 0.07463057245746926,
+      "grad_norm": 0.2936583161354065,
+      "learning_rate": 5e-05,
+      "loss": 2.0757,
+      "step": 601
+    },
+    {
+      "epoch": 0.07475474978268969,
+      "grad_norm": 0.2757823169231415,
+      "learning_rate": 5e-05,
+      "loss": 1.6964,
+      "step": 602
+    },
+    {
+      "epoch": 0.0748789271079101,
+      "grad_norm": 0.26744693517684937,
+      "learning_rate": 5e-05,
+      "loss": 2.0538,
+      "step": 603
+    },
+    {
+      "epoch": 0.0750031044331305,
+      "grad_norm": 0.5093820691108704,
+      "learning_rate": 5e-05,
+      "loss": 2.814,
+      "step": 604
+    },
+    {
+      "epoch": 0.07512728175835093,
+      "grad_norm": 0.2654862701892853,
+      "learning_rate": 5e-05,
+      "loss": 2.0413,
+      "step": 605
+    },
+    {
+      "epoch": 0.07525145908357134,
+      "grad_norm": 0.25221195816993713,
+      "learning_rate": 5e-05,
+      "loss": 1.8044,
+      "step": 606
+    },
+    {
+      "epoch": 0.07537563640879176,
+      "grad_norm": 0.22025726735591888,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 607
+    },
+    {
+      "epoch": 0.07549981373401217,
+      "grad_norm": 0.2488354593515396,
+      "learning_rate": 5e-05,
+      "loss": 1.956,
+      "step": 608
+    },
+    {
+      "epoch": 0.07562399105923258,
+      "grad_norm": 0.3650732934474945,
+      "learning_rate": 5e-05,
+      "loss": 2.2333,
+      "step": 609
+    },
+    {
+      "epoch": 0.075748168384453,
+      "grad_norm": 0.2663559019565582,
+      "learning_rate": 5e-05,
+      "loss": 2.0696,
+      "step": 610
+    },
+    {
+      "epoch": 0.07587234570967341,
+      "grad_norm": 0.25640586018562317,
+      "learning_rate": 5e-05,
+      "loss": 1.8694,
+      "step": 611
+    },
+    {
+      "epoch": 0.07599652303489383,
+      "grad_norm": 0.2904110848903656,
+      "learning_rate": 5e-05,
+      "loss": 1.9323,
+      "step": 612
+    },
+    {
+      "epoch": 0.07612070036011424,
+      "grad_norm": 0.2875644862651825,
+      "learning_rate": 5e-05,
+      "loss": 2.2906,
+      "step": 613
+    },
+    {
+      "epoch": 0.07624487768533465,
+      "grad_norm": 0.3155750036239624,
+      "learning_rate": 5e-05,
+      "loss": 2.0073,
+      "step": 614
+    },
+    {
+      "epoch": 0.07636905501055508,
+      "grad_norm": 0.27238690853118896,
+      "learning_rate": 5e-05,
+      "loss": 2.1883,
+      "step": 615
+    },
+    {
+      "epoch": 0.07649323233577549,
+      "grad_norm": 0.4303113520145416,
+      "learning_rate": 5e-05,
+      "loss": 2.5293,
+      "step": 616
+    },
+    {
+      "epoch": 0.07661740966099591,
+      "grad_norm": 0.24898111820220947,
+      "learning_rate": 5e-05,
+      "loss": 1.8349,
+      "step": 617
+    },
+    {
+      "epoch": 0.07674158698621632,
+      "grad_norm": 0.278856098651886,
+      "learning_rate": 5e-05,
+      "loss": 2.0372,
+      "step": 618
+    },
+    {
+      "epoch": 0.07686576431143673,
+      "grad_norm": 0.2800319194793701,
+      "learning_rate": 5e-05,
+      "loss": 2.1427,
+      "step": 619
+    },
+    {
+      "epoch": 0.07698994163665715,
+      "grad_norm": 0.2848834991455078,
+      "learning_rate": 5e-05,
+      "loss": 2.483,
+      "step": 620
+    },
+    {
+      "epoch": 0.07711411896187756,
+      "grad_norm": 0.24069686233997345,
+      "learning_rate": 5e-05,
+      "loss": 1.8475,
+      "step": 621
+    },
+    {
+      "epoch": 0.07723829628709798,
+      "grad_norm": 0.7840754985809326,
+      "learning_rate": 5e-05,
+      "loss": 3.1674,
+      "step": 622
+    },
+    {
+      "epoch": 0.07736247361231839,
+      "grad_norm": 0.3103671371936798,
+      "learning_rate": 5e-05,
+      "loss": 2.3145,
+      "step": 623
+    },
+    {
+      "epoch": 0.0774866509375388,
+      "grad_norm": 0.35941794514656067,
+      "learning_rate": 5e-05,
+      "loss": 2.0401,
+      "step": 624
+    },
+    {
+      "epoch": 0.07761082826275922,
+      "grad_norm": 0.413422554731369,
+      "learning_rate": 5e-05,
+      "loss": 2.2512,
+      "step": 625
+    },
+    {
+      "epoch": 0.07773500558797963,
+      "grad_norm": 0.20849303901195526,
+      "learning_rate": 5e-05,
+      "loss": 1.3579,
+      "step": 626
+    },
+    {
+      "epoch": 0.07785918291320006,
+      "grad_norm": 0.2941526174545288,
+      "learning_rate": 5e-05,
+      "loss": 2.0278,
+      "step": 627
+    },
+    {
+      "epoch": 0.07798336023842047,
+      "grad_norm": 0.333158403635025,
+      "learning_rate": 5e-05,
+      "loss": 2.3466,
+      "step": 628
+    },
+    {
+      "epoch": 0.07810753756364087,
+      "grad_norm": 0.28543370962142944,
+      "learning_rate": 5e-05,
+      "loss": 1.8151,
+      "step": 629
+    },
+    {
+      "epoch": 0.0782317148888613,
+      "grad_norm": 0.27258527278900146,
+      "learning_rate": 5e-05,
+      "loss": 1.8972,
+      "step": 630
+    },
+    {
+      "epoch": 0.07835589221408171,
+      "grad_norm": 0.2952210307121277,
+      "learning_rate": 5e-05,
+      "loss": 1.9706,
+      "step": 631
+    },
+    {
+      "epoch": 0.07848006953930213,
+      "grad_norm": 0.2456735074520111,
+      "learning_rate": 5e-05,
+      "loss": 1.8652,
+      "step": 632
+    },
+    {
+      "epoch": 0.07860424686452254,
+      "grad_norm": 0.2551192343235016,
+      "learning_rate": 5e-05,
+      "loss": 1.9088,
+      "step": 633
+    },
+    {
+      "epoch": 0.07872842418974295,
+      "grad_norm": 0.42509379982948303,
+      "learning_rate": 5e-05,
+      "loss": 3.0928,
+      "step": 634
+    },
+    {
+      "epoch": 0.07885260151496337,
+      "grad_norm": 0.36801427602767944,
+      "learning_rate": 5e-05,
+      "loss": 2.3781,
+      "step": 635
+    },
+    {
+      "epoch": 0.07897677884018378,
+      "grad_norm": 0.4733458161354065,
+      "learning_rate": 5e-05,
+      "loss": 2.7991,
+      "step": 636
+    },
+    {
+      "epoch": 0.07910095616540419,
+      "grad_norm": 0.44228315353393555,
+      "learning_rate": 5e-05,
+      "loss": 2.6639,
+      "step": 637
+    },
+    {
+      "epoch": 0.07922513349062461,
+      "grad_norm": 0.3555773198604584,
+      "learning_rate": 5e-05,
+      "loss": 1.9481,
+      "step": 638
+    },
+    {
+      "epoch": 0.07934931081584502,
+      "grad_norm": 0.3030836582183838,
+      "learning_rate": 5e-05,
+      "loss": 1.9925,
+      "step": 639
+    },
+    {
+      "epoch": 0.07947348814106545,
+      "grad_norm": 0.3075978457927704,
+      "learning_rate": 5e-05,
+      "loss": 2.4985,
+      "step": 640
+    },
+    {
+      "epoch": 0.07959766546628586,
+      "grad_norm": 0.28234943747520447,
+      "learning_rate": 5e-05,
+      "loss": 1.8877,
+      "step": 641
+    },
+    {
+      "epoch": 0.07972184279150626,
+      "grad_norm": 0.2820385694503784,
+      "learning_rate": 5e-05,
+      "loss": 1.9299,
+      "step": 642
+    },
+    {
+      "epoch": 0.07984602011672669,
+      "grad_norm": 0.2545168995857239,
+      "learning_rate": 5e-05,
+      "loss": 1.7379,
+      "step": 643
+    },
+    {
+      "epoch": 0.0799701974419471,
+      "grad_norm": 0.6520564556121826,
+      "learning_rate": 5e-05,
+      "loss": 2.7253,
+      "step": 644
+    },
+    {
+      "epoch": 0.08009437476716752,
+      "grad_norm": 0.5718693733215332,
+      "learning_rate": 5e-05,
+      "loss": 2.8075,
+      "step": 645
+    },
+    {
+      "epoch": 0.08021855209238793,
+      "grad_norm": 0.2914111018180847,
+      "learning_rate": 5e-05,
+      "loss": 2.0151,
+      "step": 646
+    },
+    {
+      "epoch": 0.08034272941760834,
+      "grad_norm": 0.3657142221927643,
+      "learning_rate": 5e-05,
+      "loss": 2.2484,
+      "step": 647
+    },
+    {
+      "epoch": 0.08046690674282876,
+      "grad_norm": 0.27595165371894836,
+      "learning_rate": 5e-05,
+      "loss": 1.8637,
+      "step": 648
+    },
+    {
+      "epoch": 0.08059108406804917,
+      "grad_norm": 0.3122614324092865,
+      "learning_rate": 5e-05,
+      "loss": 2.0657,
+      "step": 649
+    },
+    {
+      "epoch": 0.0807152613932696,
+      "grad_norm": 0.3301526606082916,
+      "learning_rate": 5e-05,
+      "loss": 2.1419,
+      "step": 650
+    },
+    {
+      "epoch": 0.08083943871849,
+      "grad_norm": 0.28250741958618164,
+      "learning_rate": 5e-05,
+      "loss": 1.9938,
+      "step": 651
+    },
+    {
+      "epoch": 0.08096361604371041,
+      "grad_norm": 0.3018810451030731,
+      "learning_rate": 5e-05,
+      "loss": 1.9527,
+      "step": 652
+    },
+    {
+      "epoch": 0.08108779336893084,
+      "grad_norm": 0.27801814675331116,
+      "learning_rate": 5e-05,
+      "loss": 2.0605,
+      "step": 653
+    },
+    {
+      "epoch": 0.08121197069415124,
+      "grad_norm": 0.5379852652549744,
+      "learning_rate": 5e-05,
+      "loss": 2.3631,
+      "step": 654
+    },
+    {
+      "epoch": 0.08133614801937167,
+      "grad_norm": 0.2944119870662689,
+      "learning_rate": 5e-05,
+      "loss": 1.8594,
+      "step": 655
+    },
+    {
+      "epoch": 0.08146032534459208,
+      "grad_norm": 0.40263888239860535,
+      "learning_rate": 5e-05,
+      "loss": 2.6047,
+      "step": 656
+    },
+    {
+      "epoch": 0.08158450266981249,
+      "grad_norm": 0.3852522373199463,
+      "learning_rate": 5e-05,
+      "loss": 2.2848,
+      "step": 657
+    },
+    {
+      "epoch": 0.08170867999503291,
+      "grad_norm": 0.277389258146286,
+      "learning_rate": 5e-05,
+      "loss": 1.9752,
+      "step": 658
+    },
+    {
+      "epoch": 0.08183285732025332,
+      "grad_norm": 0.23953773081302643,
+      "learning_rate": 5e-05,
+      "loss": 1.9688,
+      "step": 659
+    },
+    {
+      "epoch": 0.08195703464547374,
+      "grad_norm": 0.26224371790885925,
+      "learning_rate": 5e-05,
+      "loss": 1.9153,
+      "step": 660
+    },
+    {
+      "epoch": 0.08208121197069415,
+      "grad_norm": 0.2478763610124588,
+      "learning_rate": 5e-05,
+      "loss": 1.6921,
+      "step": 661
+    },
+    {
+      "epoch": 0.08220538929591456,
+      "grad_norm": 0.7691401243209839,
+      "learning_rate": 5e-05,
+      "loss": 3.6811,
+      "step": 662
+    },
+    {
+      "epoch": 0.08232956662113498,
+      "grad_norm": 0.29780057072639465,
+      "learning_rate": 5e-05,
+      "loss": 2.0604,
+      "step": 663
+    },
+    {
+      "epoch": 0.08245374394635539,
+      "grad_norm": 0.5676976442337036,
+      "learning_rate": 5e-05,
+      "loss": 2.7832,
+      "step": 664
+    },
+    {
+      "epoch": 0.08257792127157582,
+      "grad_norm": 0.2499982863664627,
+      "learning_rate": 5e-05,
+      "loss": 1.9787,
+      "step": 665
+    },
+    {
+      "epoch": 0.08270209859679623,
+      "grad_norm": 0.3072485029697418,
+      "learning_rate": 5e-05,
+      "loss": 1.9209,
+      "step": 666
+    },
+    {
+      "epoch": 0.08282627592201663,
+      "grad_norm": 0.24421563744544983,
+      "learning_rate": 5e-05,
+      "loss": 1.81,
+      "step": 667
+    },
+    {
+      "epoch": 0.08295045324723706,
+      "grad_norm": 0.29384368658065796,
+      "learning_rate": 5e-05,
+      "loss": 2.2303,
+      "step": 668
+    },
+    {
+      "epoch": 0.08307463057245747,
+      "grad_norm": 0.3173346519470215,
+      "learning_rate": 5e-05,
+      "loss": 1.7589,
+      "step": 669
+    },
+    {
+      "epoch": 0.08319880789767789,
+      "grad_norm": 0.3035511076450348,
+      "learning_rate": 5e-05,
+      "loss": 2.1347,
+      "step": 670
+    },
+    {
+      "epoch": 0.0833229852228983,
+      "grad_norm": 0.2766481339931488,
+      "learning_rate": 5e-05,
+      "loss": 1.8904,
+      "step": 671
+    },
+    {
+      "epoch": 0.08344716254811871,
+      "grad_norm": 0.27517154812812805,
+      "learning_rate": 5e-05,
+      "loss": 1.9532,
+      "step": 672
+    },
+    {
+      "epoch": 0.08357133987333913,
+      "grad_norm": 0.2978760600090027,
+      "learning_rate": 5e-05,
+      "loss": 2.2585,
+      "step": 673
+    },
+    {
+      "epoch": 0.08369551719855954,
+      "grad_norm": 0.33277857303619385,
+      "learning_rate": 5e-05,
+      "loss": 2.2466,
+      "step": 674
+    },
+    {
+      "epoch": 0.08381969452377996,
+      "grad_norm": 0.31049874424934387,
+      "learning_rate": 5e-05,
+      "loss": 2.0607,
+      "step": 675
+    },
+    {
+      "epoch": 0.08394387184900037,
+      "grad_norm": 0.2526696026325226,
+      "learning_rate": 5e-05,
+      "loss": 1.832,
+      "step": 676
+    },
+    {
+      "epoch": 0.08406804917422078,
+      "grad_norm": 0.24298225343227386,
+      "learning_rate": 5e-05,
+      "loss": 1.9017,
+      "step": 677
+    },
+    {
+      "epoch": 0.0841922264994412,
+      "grad_norm": 0.24344177544116974,
+      "learning_rate": 5e-05,
+      "loss": 1.833,
+      "step": 678
+    },
+    {
+      "epoch": 0.08431640382466161,
+      "grad_norm": 0.23898719251155853,
+      "learning_rate": 5e-05,
+      "loss": 1.7999,
+      "step": 679
+    },
+    {
+      "epoch": 0.08444058114988204,
+      "grad_norm": 0.25833937525749207,
+      "learning_rate": 5e-05,
+      "loss": 1.8702,
+      "step": 680
+    },
+    {
+      "epoch": 0.08456475847510245,
+      "grad_norm": 0.22862203419208527,
+      "learning_rate": 5e-05,
+      "loss": 1.7092,
+      "step": 681
+    },
+    {
+      "epoch": 0.08468893580032286,
+      "grad_norm": 0.2428145557641983,
+      "learning_rate": 5e-05,
+      "loss": 1.7779,
+      "step": 682
+    },
+    {
+      "epoch": 0.08481311312554328,
+      "grad_norm": 0.3271627426147461,
+      "learning_rate": 5e-05,
+      "loss": 2.3873,
+      "step": 683
+    },
+    {
+      "epoch": 0.08493729045076369,
+      "grad_norm": 0.2865424156188965,
+      "learning_rate": 5e-05,
+      "loss": 2.32,
+      "step": 684
+    },
+    {
+      "epoch": 0.08506146777598411,
+      "grad_norm": 0.2696629762649536,
+      "learning_rate": 5e-05,
+      "loss": 1.9952,
+      "step": 685
+    },
+    {
+      "epoch": 0.08518564510120452,
+      "grad_norm": 0.46669477224349976,
+      "learning_rate": 5e-05,
+      "loss": 2.5196,
+      "step": 686
+    },
+    {
+      "epoch": 0.08530982242642493,
+      "grad_norm": 0.4242430627346039,
+      "learning_rate": 5e-05,
+      "loss": 2.4382,
+      "step": 687
+    },
+    {
+      "epoch": 0.08543399975164535,
+      "grad_norm": 0.360166072845459,
+      "learning_rate": 5e-05,
+      "loss": 2.2533,
+      "step": 688
+    },
+    {
+      "epoch": 0.08555817707686576,
+      "grad_norm": 0.28157690167427063,
+      "learning_rate": 5e-05,
+      "loss": 2.1017,
+      "step": 689
+    },
+    {
+      "epoch": 0.08568235440208617,
+      "grad_norm": 0.2724866271018982,
+      "learning_rate": 5e-05,
+      "loss": 2.0147,
+      "step": 690
+    },
+    {
+      "epoch": 0.0858065317273066,
+      "grad_norm": 0.2401788979768753,
+      "learning_rate": 5e-05,
+      "loss": 1.7621,
+      "step": 691
+    },
+    {
+      "epoch": 0.085930709052527,
+      "grad_norm": 0.5512127876281738,
+      "learning_rate": 5e-05,
+      "loss": 3.1957,
+      "step": 692
+    },
+    {
+      "epoch": 0.08605488637774743,
+      "grad_norm": 0.3149046301841736,
+      "learning_rate": 5e-05,
+      "loss": 2.0574,
+      "step": 693
+    },
+    {
+      "epoch": 0.08617906370296784,
+      "grad_norm": 0.24976606667041779,
+      "learning_rate": 5e-05,
+      "loss": 2.0597,
+      "step": 694
+    },
+    {
+      "epoch": 0.08630324102818825,
+      "grad_norm": 0.2638603746891022,
+      "learning_rate": 5e-05,
+      "loss": 1.8479,
+      "step": 695
+    },
+    {
+      "epoch": 0.08642741835340867,
+      "grad_norm": 0.21984033286571503,
+      "learning_rate": 5e-05,
+      "loss": 1.7259,
+      "step": 696
+    },
+    {
+      "epoch": 0.08655159567862908,
+      "grad_norm": 0.2649635970592499,
+      "learning_rate": 5e-05,
+      "loss": 2.1046,
+      "step": 697
+    },
+    {
+      "epoch": 0.0866757730038495,
+      "grad_norm": 0.7213672399520874,
+      "learning_rate": 5e-05,
+      "loss": 3.2027,
+      "step": 698
+    },
+    {
+      "epoch": 0.08679995032906991,
+      "grad_norm": 0.28230035305023193,
+      "learning_rate": 5e-05,
+      "loss": 2.0541,
+      "step": 699
+    },
+    {
+      "epoch": 0.08692412765429032,
+      "grad_norm": 0.28091666102409363,
+      "learning_rate": 5e-05,
+      "loss": 1.5943,
+      "step": 700
+    },
+    {
+      "epoch": 0.08704830497951074,
+      "grad_norm": 0.28518977761268616,
+      "learning_rate": 5e-05,
+      "loss": 1.9855,
+      "step": 701
+    },
+    {
+      "epoch": 0.08717248230473115,
+      "grad_norm": 0.3630225658416748,
+      "learning_rate": 5e-05,
+      "loss": 2.6693,
+      "step": 702
+    },
+    {
+      "epoch": 0.08729665962995158,
+      "grad_norm": 0.3029293715953827,
+      "learning_rate": 5e-05,
+      "loss": 1.9212,
+      "step": 703
+    },
+    {
+      "epoch": 0.08742083695517198,
+      "grad_norm": 0.325604647397995,
+      "learning_rate": 5e-05,
+      "loss": 2.104,
+      "step": 704
+    },
+    {
+      "epoch": 0.0875450142803924,
+      "grad_norm": 0.25055134296417236,
+      "learning_rate": 5e-05,
+      "loss": 2.0108,
+      "step": 705
+    },
+    {
+      "epoch": 0.08766919160561282,
+      "grad_norm": 0.2693207859992981,
+      "learning_rate": 5e-05,
+      "loss": 2.0158,
+      "step": 706
+    },
+    {
+      "epoch": 0.08779336893083323,
+      "grad_norm": 0.3679948151111603,
+      "learning_rate": 5e-05,
+      "loss": 2.136,
+      "step": 707
+    },
+    {
+      "epoch": 0.08791754625605365,
+      "grad_norm": 0.2586674094200134,
+      "learning_rate": 5e-05,
+      "loss": 1.7522,
+      "step": 708
+    },
+    {
+      "epoch": 0.08804172358127406,
+      "grad_norm": 0.23914387822151184,
+      "learning_rate": 5e-05,
+      "loss": 1.8264,
+      "step": 709
+    },
+    {
+      "epoch": 0.08816590090649447,
+      "grad_norm": 0.24564284086227417,
+      "learning_rate": 5e-05,
+      "loss": 2.0949,
+      "step": 710
+    },
+    {
+      "epoch": 0.08829007823171489,
+      "grad_norm": 0.3506222665309906,
+      "learning_rate": 5e-05,
+      "loss": 2.0817,
+      "step": 711
+    },
+    {
+      "epoch": 0.0884142555569353,
+      "grad_norm": 0.3688604235649109,
+      "learning_rate": 5e-05,
+      "loss": 2.2708,
+      "step": 712
+    },
+    {
+      "epoch": 0.08853843288215572,
+      "grad_norm": 0.4141845405101776,
+      "learning_rate": 5e-05,
+      "loss": 2.1846,
+      "step": 713
+    },
+    {
+      "epoch": 0.08866261020737613,
+      "grad_norm": 0.2667691111564636,
+      "learning_rate": 5e-05,
+      "loss": 1.9375,
+      "step": 714
+    },
+    {
+      "epoch": 0.08878678753259654,
+      "grad_norm": 0.34525224566459656,
+      "learning_rate": 5e-05,
+      "loss": 2.2076,
+      "step": 715
+    },
+    {
+      "epoch": 0.08891096485781697,
+      "grad_norm": 0.3912832736968994,
+      "learning_rate": 5e-05,
+      "loss": 2.3614,
+      "step": 716
+    },
+    {
+      "epoch": 0.08903514218303737,
+      "grad_norm": 0.4596518278121948,
+      "learning_rate": 5e-05,
+      "loss": 2.4385,
+      "step": 717
+    },
+    {
+      "epoch": 0.0891593195082578,
+      "grad_norm": 0.28569549322128296,
+      "learning_rate": 5e-05,
+      "loss": 2.4532,
+      "step": 718
+    },
+    {
+      "epoch": 0.08928349683347821,
+      "grad_norm": 0.3232335150241852,
+      "learning_rate": 5e-05,
+      "loss": 2.5114,
+      "step": 719
+    },
+    {
+      "epoch": 0.08940767415869862,
+      "grad_norm": 0.24512383341789246,
+      "learning_rate": 5e-05,
+      "loss": 1.6351,
+      "step": 720
+    },
+    {
+      "epoch": 0.08953185148391904,
+      "grad_norm": 0.40597257018089294,
+      "learning_rate": 5e-05,
+      "loss": 2.6137,
+      "step": 721
+    },
+    {
+      "epoch": 0.08965602880913945,
+      "grad_norm": 0.3255784213542938,
+      "learning_rate": 5e-05,
+      "loss": 2.0508,
+      "step": 722
+    },
+    {
+      "epoch": 0.08978020613435987,
+      "grad_norm": 0.40446966886520386,
+      "learning_rate": 5e-05,
+      "loss": 2.686,
+      "step": 723
+    },
+    {
+      "epoch": 0.08990438345958028,
+      "grad_norm": 0.3603253662586212,
+      "learning_rate": 5e-05,
+      "loss": 2.6255,
+      "step": 724
+    },
+    {
+      "epoch": 0.09002856078480069,
+      "grad_norm": 0.27847838401794434,
+      "learning_rate": 5e-05,
+      "loss": 1.9845,
+      "step": 725
+    },
+    {
+      "epoch": 0.09015273811002111,
+      "grad_norm": 0.2638871669769287,
+      "learning_rate": 5e-05,
+      "loss": 1.9401,
+      "step": 726
+    },
+    {
+      "epoch": 0.09027691543524152,
+      "grad_norm": 0.25534674525260925,
+      "learning_rate": 5e-05,
+      "loss": 1.8353,
+      "step": 727
+    },
+    {
+      "epoch": 0.09040109276046195,
+      "grad_norm": 0.24471010267734528,
+      "learning_rate": 5e-05,
+      "loss": 1.8426,
+      "step": 728
+    },
+    {
+      "epoch": 0.09052527008568236,
+      "grad_norm": 0.40118831396102905,
+      "learning_rate": 5e-05,
+      "loss": 2.9792,
+      "step": 729
+    },
+    {
+      "epoch": 0.09064944741090276,
+      "grad_norm": 0.6646387577056885,
+      "learning_rate": 5e-05,
+      "loss": 3.0327,
+      "step": 730
+    },
+    {
+      "epoch": 0.09077362473612319,
+      "grad_norm": 0.3027131259441376,
+      "learning_rate": 5e-05,
+      "loss": 2.0686,
+      "step": 731
+    },
+    {
+      "epoch": 0.0908978020613436,
+      "grad_norm": 0.2986295819282532,
+      "learning_rate": 5e-05,
+      "loss": 2.8916,
+      "step": 732
+    },
+    {
+      "epoch": 0.09102197938656402,
+      "grad_norm": 0.2983008325099945,
+      "learning_rate": 5e-05,
+      "loss": 2.0715,
+      "step": 733
+    },
+    {
+      "epoch": 0.09114615671178443,
+      "grad_norm": 0.234079971909523,
+      "learning_rate": 5e-05,
+      "loss": 1.7109,
+      "step": 734
+    },
+    {
+      "epoch": 0.09127033403700484,
+      "grad_norm": 0.3074970543384552,
+      "learning_rate": 5e-05,
+      "loss": 2.1104,
+      "step": 735
+    },
+    {
+      "epoch": 0.09139451136222526,
+      "grad_norm": 0.2545653283596039,
+      "learning_rate": 5e-05,
+      "loss": 1.9592,
+      "step": 736
+    },
+    {
+      "epoch": 0.09151868868744567,
+      "grad_norm": 0.2585062086582184,
+      "learning_rate": 5e-05,
+      "loss": 1.8301,
+      "step": 737
+    },
+    {
+      "epoch": 0.0916428660126661,
+      "grad_norm": 0.22999894618988037,
+      "learning_rate": 5e-05,
+      "loss": 1.6784,
+      "step": 738
+    },
+    {
+      "epoch": 0.0917670433378865,
+      "grad_norm": 0.2429080605506897,
+      "learning_rate": 5e-05,
+      "loss": 1.7051,
+      "step": 739
+    },
+    {
+      "epoch": 0.09189122066310691,
+      "grad_norm": 0.4000047743320465,
+      "learning_rate": 5e-05,
+      "loss": 2.4653,
+      "step": 740
+    },
+    {
+      "epoch": 0.09201539798832734,
+      "grad_norm": 0.26037660241127014,
+      "learning_rate": 5e-05,
+      "loss": 1.908,
+      "step": 741
+    },
+    {
+      "epoch": 0.09213957531354774,
+      "grad_norm": 0.2336800992488861,
+      "learning_rate": 5e-05,
+      "loss": 1.9641,
+      "step": 742
+    },
+    {
+      "epoch": 0.09226375263876817,
+      "grad_norm": 0.2122896909713745,
+      "learning_rate": 5e-05,
+      "loss": 1.6852,
+      "step": 743
+    },
+    {
+      "epoch": 0.09238792996398858,
+      "grad_norm": 0.44178128242492676,
+      "learning_rate": 5e-05,
+      "loss": 2.5299,
+      "step": 744
+    },
+    {
+      "epoch": 0.09251210728920899,
+      "grad_norm": 0.2383100986480713,
+      "learning_rate": 5e-05,
+      "loss": 1.6617,
+      "step": 745
+    },
+    {
+      "epoch": 0.09263628461442941,
+      "grad_norm": 0.2937283515930176,
+      "learning_rate": 5e-05,
+      "loss": 1.9293,
+      "step": 746
+    },
+    {
+      "epoch": 0.09276046193964982,
+      "grad_norm": 0.23060178756713867,
+      "learning_rate": 5e-05,
+      "loss": 1.965,
+      "step": 747
+    },
+    {
+      "epoch": 0.09288463926487023,
+      "grad_norm": 0.2461223602294922,
+      "learning_rate": 5e-05,
+      "loss": 1.86,
+      "step": 748
+    },
+    {
+      "epoch": 0.09300881659009065,
+      "grad_norm": 0.26430073380470276,
+      "learning_rate": 5e-05,
+      "loss": 1.8866,
+      "step": 749
+    },
+    {
+      "epoch": 0.09313299391531106,
+      "grad_norm": 0.2722196578979492,
+      "learning_rate": 5e-05,
+      "loss": 2.168,
+      "step": 750
+    },
+    {
+      "epoch": 0.09325717124053148,
+      "grad_norm": 0.24600358307361603,
+      "learning_rate": 5e-05,
+      "loss": 1.9163,
+      "step": 751
+    },
+    {
+      "epoch": 0.09338134856575189,
+      "grad_norm": 0.3493156433105469,
+      "learning_rate": 5e-05,
+      "loss": 2.2273,
+      "step": 752
+    },
+    {
+      "epoch": 0.0935055258909723,
+      "grad_norm": 0.3096140921115875,
+      "learning_rate": 5e-05,
+      "loss": 2.1582,
+      "step": 753
+    },
+    {
+      "epoch": 0.09362970321619273,
+      "grad_norm": 0.26107022166252136,
+      "learning_rate": 5e-05,
+      "loss": 1.9407,
+      "step": 754
+    },
+    {
+      "epoch": 0.09375388054141313,
+      "grad_norm": 0.3041000962257385,
+      "learning_rate": 5e-05,
+      "loss": 2.2585,
+      "step": 755
+    },
+    {
+      "epoch": 0.09387805786663356,
+      "grad_norm": 0.26501205563545227,
+      "learning_rate": 5e-05,
+      "loss": 2.3782,
+      "step": 756
+    },
+    {
+      "epoch": 0.09400223519185397,
+      "grad_norm": 0.27229952812194824,
+      "learning_rate": 5e-05,
+      "loss": 2.1154,
+      "step": 757
+    },
+    {
+      "epoch": 0.09412641251707438,
+      "grad_norm": 0.2528907060623169,
+      "learning_rate": 5e-05,
+      "loss": 1.7321,
+      "step": 758
+    },
+    {
+      "epoch": 0.0942505898422948,
+      "grad_norm": 0.22811026871204376,
+      "learning_rate": 5e-05,
+      "loss": 1.7883,
+      "step": 759
+    },
+    {
+      "epoch": 0.09437476716751521,
+      "grad_norm": 0.2483462244272232,
+      "learning_rate": 5e-05,
+      "loss": 1.9824,
+      "step": 760
+    },
+    {
+      "epoch": 0.09449894449273563,
+      "grad_norm": 0.2446499466896057,
+      "learning_rate": 5e-05,
+      "loss": 1.9688,
+      "step": 761
+    },
+    {
+      "epoch": 0.09462312181795604,
+      "grad_norm": 0.25365975499153137,
+      "learning_rate": 5e-05,
+      "loss": 1.955,
+      "step": 762
+    },
+    {
+      "epoch": 0.09474729914317645,
+      "grad_norm": 0.25691381096839905,
+      "learning_rate": 5e-05,
+      "loss": 2.0523,
+      "step": 763
+    },
+    {
+      "epoch": 0.09487147646839687,
+      "grad_norm": 0.30339887738227844,
+      "learning_rate": 5e-05,
+      "loss": 2.2387,
+      "step": 764
+    },
+    {
+      "epoch": 0.09499565379361728,
+      "grad_norm": 0.3994625210762024,
+      "learning_rate": 5e-05,
+      "loss": 2.7802,
+      "step": 765
+    },
+    {
+      "epoch": 0.0951198311188377,
+      "grad_norm": 0.2273540496826172,
+      "learning_rate": 5e-05,
+      "loss": 1.8601,
+      "step": 766
+    },
+    {
+      "epoch": 0.09524400844405811,
+      "grad_norm": 0.23096863925457,
+      "learning_rate": 5e-05,
+      "loss": 1.6603,
+      "step": 767
+    },
+    {
+      "epoch": 0.09536818576927852,
+      "grad_norm": 0.23737122118473053,
+      "learning_rate": 5e-05,
+      "loss": 1.803,
+      "step": 768
+    },
+    {
+      "epoch": 0.09549236309449895,
+      "grad_norm": 0.2674116790294647,
+      "learning_rate": 5e-05,
+      "loss": 1.9375,
+      "step": 769
+    },
+    {
+      "epoch": 0.09561654041971936,
+      "grad_norm": 0.3963793218135834,
+      "learning_rate": 5e-05,
+      "loss": 2.2229,
+      "step": 770
+    },
+    {
+      "epoch": 0.09574071774493978,
+      "grad_norm": 0.24932265281677246,
+      "learning_rate": 5e-05,
+      "loss": 1.9151,
+      "step": 771
+    },
+    {
+      "epoch": 0.09586489507016019,
+      "grad_norm": 0.3672339618206024,
+      "learning_rate": 5e-05,
+      "loss": 1.9886,
+      "step": 772
+    },
+    {
+      "epoch": 0.0959890723953806,
+      "grad_norm": 0.28162822127342224,
+      "learning_rate": 5e-05,
+      "loss": 1.8631,
+      "step": 773
+    },
+    {
+      "epoch": 0.09611324972060102,
+      "grad_norm": 0.28446900844573975,
+      "learning_rate": 5e-05,
+      "loss": 1.9035,
+      "step": 774
+    },
+    {
+      "epoch": 0.09623742704582143,
+      "grad_norm": 0.22119373083114624,
+      "learning_rate": 5e-05,
+      "loss": 1.6798,
+      "step": 775
+    },
+    {
+      "epoch": 0.09636160437104185,
+      "grad_norm": 0.325965940952301,
+      "learning_rate": 5e-05,
+      "loss": 1.715,
+      "step": 776
+    },
+    {
+      "epoch": 0.09648578169626226,
+      "grad_norm": 0.3543036878108978,
+      "learning_rate": 5e-05,
+      "loss": 2.0075,
+      "step": 777
+    },
+    {
+      "epoch": 0.09660995902148267,
+      "grad_norm": 0.302958607673645,
+      "learning_rate": 5e-05,
+      "loss": 1.8459,
+      "step": 778
+    },
+    {
+      "epoch": 0.0967341363467031,
+      "grad_norm": 0.2946341037750244,
+      "learning_rate": 5e-05,
+      "loss": 2.1246,
+      "step": 779
+    },
+    {
+      "epoch": 0.0968583136719235,
+      "grad_norm": 0.29457908868789673,
+      "learning_rate": 5e-05,
+      "loss": 2.3041,
+      "step": 780
+    },
+    {
+      "epoch": 0.09698249099714393,
+      "grad_norm": 0.5821194052696228,
+      "learning_rate": 5e-05,
+      "loss": 2.9414,
+      "step": 781
+    },
+    {
+      "epoch": 0.09710666832236434,
+      "grad_norm": 0.32491669058799744,
+      "learning_rate": 5e-05,
+      "loss": 2.2702,
+      "step": 782
+    },
+    {
+      "epoch": 0.09723084564758475,
+      "grad_norm": 0.2968323528766632,
+      "learning_rate": 5e-05,
+      "loss": 2.3479,
+      "step": 783
+    },
+    {
+      "epoch": 0.09735502297280517,
+      "grad_norm": 0.6173321604728699,
+      "learning_rate": 5e-05,
+      "loss": 3.1077,
+      "step": 784
+    },
+    {
+      "epoch": 0.09747920029802558,
+      "grad_norm": 0.40393519401550293,
+      "learning_rate": 5e-05,
+      "loss": 2.0843,
+      "step": 785
+    },
+    {
+      "epoch": 0.097603377623246,
+      "grad_norm": 0.3294332027435303,
+      "learning_rate": 5e-05,
+      "loss": 1.8889,
+      "step": 786
+    },
+    {
+      "epoch": 0.09772755494846641,
+      "grad_norm": 0.45853391289711,
+      "learning_rate": 5e-05,
+      "loss": 2.1967,
+      "step": 787
+    },
+    {
+      "epoch": 0.09785173227368682,
+      "grad_norm": 0.3437069058418274,
+      "learning_rate": 5e-05,
+      "loss": 1.9014,
+      "step": 788
+    },
+    {
+      "epoch": 0.09797590959890724,
+      "grad_norm": 0.30439493060112,
+      "learning_rate": 5e-05,
+      "loss": 2.0923,
+      "step": 789
+    },
+    {
+      "epoch": 0.09810008692412765,
+      "grad_norm": 0.32713377475738525,
+      "learning_rate": 5e-05,
+      "loss": 1.7945,
+      "step": 790
+    },
+    {
+      "epoch": 0.09822426424934808,
+      "grad_norm": 0.28058183193206787,
+      "learning_rate": 5e-05,
+      "loss": 1.7459,
+      "step": 791
+    },
+    {
+      "epoch": 0.09834844157456848,
+      "grad_norm": 0.2603336572647095,
+      "learning_rate": 5e-05,
+      "loss": 1.942,
+      "step": 792
+    },
+    {
+      "epoch": 0.0984726188997889,
+      "grad_norm": 0.2959796190261841,
+      "learning_rate": 5e-05,
+      "loss": 2.2251,
+      "step": 793
+    },
+    {
+      "epoch": 0.09859679622500932,
+      "grad_norm": 0.24663518369197845,
+      "learning_rate": 5e-05,
+      "loss": 1.88,
+      "step": 794
+    },
+    {
+      "epoch": 0.09872097355022973,
+      "grad_norm": 0.3397309482097626,
+      "learning_rate": 5e-05,
+      "loss": 2.0779,
+      "step": 795
+    },
+    {
+      "epoch": 0.09884515087545015,
+      "grad_norm": 0.2475767284631729,
+      "learning_rate": 5e-05,
+      "loss": 1.7452,
+      "step": 796
+    },
+    {
+      "epoch": 0.09896932820067056,
+      "grad_norm": 0.3029365539550781,
+      "learning_rate": 5e-05,
+      "loss": 2.0652,
+      "step": 797
+    },
+    {
+      "epoch": 0.09909350552589097,
+      "grad_norm": 0.2678433656692505,
+      "learning_rate": 5e-05,
+      "loss": 1.9127,
+      "step": 798
+    },
+    {
+      "epoch": 0.09921768285111139,
+      "grad_norm": 0.35326072573661804,
+      "learning_rate": 5e-05,
+      "loss": 2.0838,
+      "step": 799
+    },
+    {
+      "epoch": 0.0993418601763318,
+      "grad_norm": 0.28843721747398376,
+      "learning_rate": 5e-05,
+      "loss": 2.1572,
+      "step": 800
+    },
+    {
+      "epoch": 0.09946603750155221,
+      "grad_norm": 0.24140127003192902,
+      "learning_rate": 5e-05,
+      "loss": 1.6219,
+      "step": 801
+    },
+    {
+      "epoch": 0.09959021482677263,
+      "grad_norm": 0.29560938477516174,
+      "learning_rate": 5e-05,
+      "loss": 1.8265,
+      "step": 802
+    },
+    {
+      "epoch": 0.09971439215199304,
+      "grad_norm": 0.31561002135276794,
+      "learning_rate": 5e-05,
+      "loss": 2.0694,
+      "step": 803
+    },
+    {
+      "epoch": 0.09983856947721347,
+      "grad_norm": 0.3077153265476227,
+      "learning_rate": 5e-05,
+      "loss": 2.3872,
+      "step": 804
+    },
+    {
+      "epoch": 0.09996274680243387,
+      "grad_norm": 0.24623873829841614,
+      "learning_rate": 5e-05,
+      "loss": 1.7888,
+      "step": 805
+    },
+    {
+      "epoch": 0.10008692412765428,
+      "grad_norm": 0.27619126439094543,
+      "learning_rate": 5e-05,
+      "loss": 2.1839,
+      "step": 806
+    },
+    {
+      "epoch": 0.1002111014528747,
+      "grad_norm": 0.33211326599121094,
+      "learning_rate": 5e-05,
+      "loss": 2.2991,
+      "step": 807
+    },
+    {
+      "epoch": 0.10033527877809512,
+      "grad_norm": 0.23164010047912598,
+      "learning_rate": 5e-05,
+      "loss": 1.5602,
+      "step": 808
+    },
+    {
+      "epoch": 0.10045945610331554,
+      "grad_norm": 0.2822321951389313,
+      "learning_rate": 5e-05,
+      "loss": 1.94,
+      "step": 809
+    },
+    {
+      "epoch": 0.10058363342853595,
+      "grad_norm": 0.2520303428173065,
+      "learning_rate": 5e-05,
+      "loss": 2.0901,
+      "step": 810
+    },
+    {
+      "epoch": 0.10070781075375636,
+      "grad_norm": 0.26948827505111694,
+      "learning_rate": 5e-05,
+      "loss": 2.1365,
+      "step": 811
+    },
+    {
+      "epoch": 0.10083198807897678,
+      "grad_norm": 0.382692813873291,
+      "learning_rate": 5e-05,
+      "loss": 2.1824,
+      "step": 812
+    },
+    {
+      "epoch": 0.10095616540419719,
+      "grad_norm": 0.22778475284576416,
+      "learning_rate": 5e-05,
+      "loss": 2.0242,
+      "step": 813
+    },
+    {
+      "epoch": 0.10108034272941761,
+      "grad_norm": 0.43782931566238403,
+      "learning_rate": 5e-05,
+      "loss": 3.1766,
+      "step": 814
+    },
+    {
+      "epoch": 0.10120452005463802,
+      "grad_norm": 0.25693511962890625,
+      "learning_rate": 5e-05,
+      "loss": 2.0156,
+      "step": 815
+    },
+    {
+      "epoch": 0.10132869737985843,
+      "grad_norm": 0.2677198052406311,
+      "learning_rate": 5e-05,
+      "loss": 1.7291,
+      "step": 816
+    },
+    {
+      "epoch": 0.10145287470507885,
+      "grad_norm": 0.2706223428249359,
+      "learning_rate": 5e-05,
+      "loss": 1.9196,
+      "step": 817
+    },
+    {
+      "epoch": 0.10157705203029926,
+      "grad_norm": 0.30248013138771057,
+      "learning_rate": 5e-05,
+      "loss": 2.1137,
+      "step": 818
+    },
+    {
+      "epoch": 0.10170122935551969,
+      "grad_norm": 0.25752121210098267,
+      "learning_rate": 5e-05,
+      "loss": 1.8943,
+      "step": 819
+    },
+    {
+      "epoch": 0.1018254066807401,
+      "grad_norm": 0.2458408623933792,
+      "learning_rate": 5e-05,
+      "loss": 1.9843,
+      "step": 820
+    },
+    {
+      "epoch": 0.1019495840059605,
+      "grad_norm": 0.2543642818927765,
+      "learning_rate": 5e-05,
+      "loss": 1.8824,
+      "step": 821
+    },
+    {
+      "epoch": 0.10207376133118093,
+      "grad_norm": 0.5759515762329102,
+      "learning_rate": 5e-05,
+      "loss": 2.5496,
+      "step": 822
+    },
+    {
+      "epoch": 0.10219793865640134,
+      "grad_norm": 0.24595996737480164,
+      "learning_rate": 5e-05,
+      "loss": 1.763,
+      "step": 823
+    },
+    {
+      "epoch": 0.10232211598162176,
+      "grad_norm": 0.3513668477535248,
+      "learning_rate": 5e-05,
+      "loss": 2.1594,
+      "step": 824
+    },
+    {
+      "epoch": 0.10244629330684217,
+      "grad_norm": 0.24765171110630035,
+      "learning_rate": 5e-05,
+      "loss": 1.9873,
+      "step": 825
+    },
+    {
+      "epoch": 0.10257047063206258,
+      "grad_norm": 0.23468390107154846,
+      "learning_rate": 5e-05,
+      "loss": 1.6796,
+      "step": 826
+    },
+    {
+      "epoch": 0.102694647957283,
+      "grad_norm": 0.38637110590934753,
+      "learning_rate": 5e-05,
+      "loss": 2.0828,
+      "step": 827
+    },
+    {
+      "epoch": 0.10281882528250341,
+      "grad_norm": 0.22323277592658997,
+      "learning_rate": 5e-05,
+      "loss": 1.642,
+      "step": 828
+    },
+    {
+      "epoch": 0.10294300260772384,
+      "grad_norm": 0.2797977030277252,
+      "learning_rate": 5e-05,
+      "loss": 1.8465,
+      "step": 829
+    },
+    {
+      "epoch": 0.10306717993294424,
+      "grad_norm": 0.24816101789474487,
+      "learning_rate": 5e-05,
+      "loss": 1.9725,
+      "step": 830
+    },
+    {
+      "epoch": 0.10319135725816465,
+      "grad_norm": 0.3066013753414154,
+      "learning_rate": 5e-05,
+      "loss": 2.4379,
+      "step": 831
+    },
+    {
+      "epoch": 0.10331553458338508,
+      "grad_norm": 0.41610953211784363,
+      "learning_rate": 5e-05,
+      "loss": 2.9505,
+      "step": 832
+    },
+    {
+      "epoch": 0.10343971190860549,
+      "grad_norm": 0.24125410616397858,
+      "learning_rate": 5e-05,
+      "loss": 1.6854,
+      "step": 833
+    },
+    {
+      "epoch": 0.10356388923382591,
+      "grad_norm": 0.3574509918689728,
+      "learning_rate": 5e-05,
+      "loss": 2.0999,
+      "step": 834
+    },
+    {
+      "epoch": 0.10368806655904632,
+      "grad_norm": 0.23841390013694763,
+      "learning_rate": 5e-05,
+      "loss": 1.7081,
+      "step": 835
+    },
+    {
+      "epoch": 0.10381224388426673,
+      "grad_norm": 0.32144802808761597,
+      "learning_rate": 5e-05,
+      "loss": 2.5213,
+      "step": 836
+    },
+    {
+      "epoch": 0.10393642120948715,
+      "grad_norm": 0.28795325756073,
+      "learning_rate": 5e-05,
+      "loss": 2.4315,
+      "step": 837
+    },
+    {
+      "epoch": 0.10406059853470756,
+      "grad_norm": 0.23300907015800476,
+      "learning_rate": 5e-05,
+      "loss": 1.7849,
+      "step": 838
+    },
+    {
+      "epoch": 0.10418477585992798,
+      "grad_norm": 0.25263527035713196,
+      "learning_rate": 5e-05,
+      "loss": 1.9377,
+      "step": 839
+    },
+    {
+      "epoch": 0.10430895318514839,
+      "grad_norm": 0.37186625599861145,
+      "learning_rate": 5e-05,
+      "loss": 2.9579,
+      "step": 840
+    },
+    {
+      "epoch": 0.1044331305103688,
+      "grad_norm": 0.3031153380870819,
+      "learning_rate": 5e-05,
+      "loss": 1.8852,
+      "step": 841
+    },
+    {
+      "epoch": 0.10455730783558922,
+      "grad_norm": 0.31639477610588074,
+      "learning_rate": 5e-05,
+      "loss": 2.7184,
+      "step": 842
+    },
+    {
+      "epoch": 0.10468148516080963,
+      "grad_norm": 0.24418902397155762,
+      "learning_rate": 5e-05,
+      "loss": 1.855,
+      "step": 843
+    },
+    {
+      "epoch": 0.10480566248603006,
+      "grad_norm": 0.22167858481407166,
+      "learning_rate": 5e-05,
+      "loss": 1.6351,
+      "step": 844
+    },
+    {
+      "epoch": 0.10492983981125047,
+      "grad_norm": 0.27623358368873596,
+      "learning_rate": 5e-05,
+      "loss": 1.9165,
+      "step": 845
+    },
+    {
+      "epoch": 0.10505401713647088,
+      "grad_norm": 0.24267208576202393,
+      "learning_rate": 5e-05,
+      "loss": 1.8563,
+      "step": 846
+    },
+    {
+      "epoch": 0.1051781944616913,
+      "grad_norm": 0.24272559583187103,
+      "learning_rate": 5e-05,
+      "loss": 2.029,
+      "step": 847
+    },
+    {
+      "epoch": 0.10530237178691171,
+      "grad_norm": 0.2580758035182953,
+      "learning_rate": 5e-05,
+      "loss": 2.0239,
+      "step": 848
+    },
+    {
+      "epoch": 0.10542654911213213,
+      "grad_norm": 0.25721096992492676,
+      "learning_rate": 5e-05,
+      "loss": 2.1841,
+      "step": 849
+    },
+    {
+      "epoch": 0.10555072643735254,
+      "grad_norm": 0.27631187438964844,
+      "learning_rate": 5e-05,
+      "loss": 2.0753,
+      "step": 850
+    },
+    {
+      "epoch": 0.10567490376257295,
+      "grad_norm": 0.23685285449028015,
+      "learning_rate": 5e-05,
+      "loss": 2.1299,
+      "step": 851
+    },
+    {
+      "epoch": 0.10579908108779337,
+      "grad_norm": 0.362529993057251,
+      "learning_rate": 5e-05,
+      "loss": 2.4725,
+      "step": 852
+    },
+    {
+      "epoch": 0.10592325841301378,
+      "grad_norm": 0.24620841443538666,
+      "learning_rate": 5e-05,
+      "loss": 1.689,
+      "step": 853
+    },
+    {
+      "epoch": 0.1060474357382342,
+      "grad_norm": 0.39088621735572815,
+      "learning_rate": 5e-05,
+      "loss": 2.2395,
+      "step": 854
+    },
+    {
+      "epoch": 0.10617161306345461,
+      "grad_norm": 0.25166064500808716,
+      "learning_rate": 5e-05,
+      "loss": 1.9069,
+      "step": 855
+    },
+    {
+      "epoch": 0.10629579038867502,
+      "grad_norm": 0.3524984121322632,
+      "learning_rate": 5e-05,
+      "loss": 2.343,
+      "step": 856
+    },
+    {
+      "epoch": 0.10641996771389545,
+      "grad_norm": 0.24170979857444763,
+      "learning_rate": 5e-05,
+      "loss": 1.5245,
+      "step": 857
+    },
+    {
+      "epoch": 0.10654414503911586,
+      "grad_norm": 0.2690974473953247,
+      "learning_rate": 5e-05,
+      "loss": 1.8784,
+      "step": 858
+    },
+    {
+      "epoch": 0.10666832236433627,
+      "grad_norm": 0.4175858199596405,
+      "learning_rate": 5e-05,
+      "loss": 2.54,
+      "step": 859
+    },
+    {
+      "epoch": 0.10679249968955669,
+      "grad_norm": 0.2780891954898834,
+      "learning_rate": 5e-05,
+      "loss": 2.0745,
+      "step": 860
+    },
+    {
+      "epoch": 0.1069166770147771,
+      "grad_norm": 0.2890850007534027,
+      "learning_rate": 5e-05,
+      "loss": 2.1256,
+      "step": 861
+    },
+    {
+      "epoch": 0.10704085433999752,
+      "grad_norm": 0.3829166889190674,
+      "learning_rate": 5e-05,
+      "loss": 2.3878,
+      "step": 862
+    },
+    {
+      "epoch": 0.10716503166521793,
+      "grad_norm": 0.4972594082355499,
+      "learning_rate": 5e-05,
+      "loss": 3.0912,
+      "step": 863
+    },
+    {
+      "epoch": 0.10728920899043834,
+      "grad_norm": 0.428373783826828,
+      "learning_rate": 5e-05,
+      "loss": 3.0987,
+      "step": 864
+    },
+    {
+      "epoch": 0.10741338631565876,
+      "grad_norm": 0.2967165410518646,
+      "learning_rate": 5e-05,
+      "loss": 2.1655,
+      "step": 865
+    },
+    {
+      "epoch": 0.10753756364087917,
+      "grad_norm": 0.3932843804359436,
+      "learning_rate": 5e-05,
+      "loss": 2.4387,
+      "step": 866
+    },
+    {
+      "epoch": 0.1076617409660996,
+      "grad_norm": 0.3399292826652527,
+      "learning_rate": 5e-05,
+      "loss": 2.3679,
+      "step": 867
+    },
+    {
+      "epoch": 0.10778591829132,
+      "grad_norm": 0.3172128200531006,
+      "learning_rate": 5e-05,
+      "loss": 2.1567,
+      "step": 868
+    },
+    {
+      "epoch": 0.10791009561654041,
+      "grad_norm": 0.23311318457126617,
+      "learning_rate": 5e-05,
+      "loss": 1.5103,
+      "step": 869
+    },
+    {
+      "epoch": 0.10803427294176084,
+      "grad_norm": 0.28969600796699524,
+      "learning_rate": 5e-05,
+      "loss": 1.6748,
+      "step": 870
+    },
+    {
+      "epoch": 0.10815845026698125,
+      "grad_norm": 0.34855031967163086,
+      "learning_rate": 5e-05,
+      "loss": 2.2776,
+      "step": 871
+    },
+    {
+      "epoch": 0.10828262759220167,
+      "grad_norm": 0.32641398906707764,
+      "learning_rate": 5e-05,
+      "loss": 2.1035,
+      "step": 872
+    },
+    {
+      "epoch": 0.10840680491742208,
+      "grad_norm": 0.2550683319568634,
+      "learning_rate": 5e-05,
+      "loss": 1.8585,
+      "step": 873
+    },
+    {
+      "epoch": 0.10853098224264249,
+      "grad_norm": 0.2957420349121094,
+      "learning_rate": 5e-05,
+      "loss": 2.2075,
+      "step": 874
+    },
+    {
+      "epoch": 0.10865515956786291,
+      "grad_norm": 0.2820521295070648,
+      "learning_rate": 5e-05,
+      "loss": 1.9231,
+      "step": 875
+    },
+    {
+      "epoch": 0.10877933689308332,
+      "grad_norm": 0.3224237263202667,
+      "learning_rate": 5e-05,
+      "loss": 2.6275,
+      "step": 876
+    },
+    {
+      "epoch": 0.10890351421830374,
+      "grad_norm": 0.32133105397224426,
+      "learning_rate": 5e-05,
+      "loss": 2.0473,
+      "step": 877
+    },
+    {
+      "epoch": 0.10902769154352415,
+      "grad_norm": 0.30259227752685547,
+      "learning_rate": 5e-05,
+      "loss": 2.3739,
+      "step": 878
+    },
+    {
+      "epoch": 0.10915186886874456,
+      "grad_norm": 0.2607637941837311,
+      "learning_rate": 5e-05,
+      "loss": 1.7886,
+      "step": 879
+    },
+    {
+      "epoch": 0.10927604619396498,
+      "grad_norm": 0.23737330734729767,
+      "learning_rate": 5e-05,
+      "loss": 1.9364,
+      "step": 880
+    },
+    {
+      "epoch": 0.1094002235191854,
+      "grad_norm": 0.366264671087265,
+      "learning_rate": 5e-05,
+      "loss": 2.5266,
+      "step": 881
+    },
+    {
+      "epoch": 0.10952440084440582,
+      "grad_norm": 0.22696363925933838,
+      "learning_rate": 5e-05,
+      "loss": 1.7637,
+      "step": 882
+    },
+    {
+      "epoch": 0.10964857816962623,
+      "grad_norm": 0.34759801626205444,
+      "learning_rate": 5e-05,
+      "loss": 2.638,
+      "step": 883
+    },
+    {
+      "epoch": 0.10977275549484664,
+      "grad_norm": 0.36980926990509033,
+      "learning_rate": 5e-05,
+      "loss": 2.0731,
+      "step": 884
+    },
+    {
+      "epoch": 0.10989693282006706,
+      "grad_norm": 0.2537945508956909,
+      "learning_rate": 5e-05,
+      "loss": 1.7732,
+      "step": 885
+    },
+    {
+      "epoch": 0.11002111014528747,
+      "grad_norm": 0.32599031925201416,
+      "learning_rate": 5e-05,
+      "loss": 2.2379,
+      "step": 886
+    },
+    {
+      "epoch": 0.11014528747050789,
+      "grad_norm": 0.2296784669160843,
+      "learning_rate": 5e-05,
+      "loss": 1.9002,
+      "step": 887
+    },
+    {
+      "epoch": 0.1102694647957283,
+      "grad_norm": 0.24791423976421356,
+      "learning_rate": 5e-05,
+      "loss": 1.8496,
+      "step": 888
+    },
+    {
+      "epoch": 0.11039364212094871,
+      "grad_norm": 0.24951361119747162,
+      "learning_rate": 5e-05,
+      "loss": 1.9118,
+      "step": 889
+    },
+    {
+      "epoch": 0.11051781944616913,
+      "grad_norm": 0.2583346664905548,
+      "learning_rate": 5e-05,
+      "loss": 1.9558,
+      "step": 890
+    },
+    {
+      "epoch": 0.11064199677138954,
+      "grad_norm": 0.6806591749191284,
+      "learning_rate": 5e-05,
+      "loss": 2.7708,
+      "step": 891
+    },
+    {
+      "epoch": 0.11076617409660997,
+      "grad_norm": 0.28599515557289124,
+      "learning_rate": 5e-05,
+      "loss": 1.8723,
+      "step": 892
+    },
+    {
+      "epoch": 0.11089035142183037,
+      "grad_norm": 0.3603532910346985,
+      "learning_rate": 5e-05,
+      "loss": 2.1735,
+      "step": 893
+    },
+    {
+      "epoch": 0.11101452874705078,
+      "grad_norm": 0.28333306312561035,
+      "learning_rate": 5e-05,
+      "loss": 1.9305,
+      "step": 894
+    },
+    {
+      "epoch": 0.1111387060722712,
+      "grad_norm": 0.3318665027618408,
+      "learning_rate": 5e-05,
+      "loss": 2.1284,
+      "step": 895
+    },
+    {
+      "epoch": 0.11126288339749162,
+      "grad_norm": 0.32923001050949097,
+      "learning_rate": 5e-05,
+      "loss": 2.1328,
+      "step": 896
+    },
+    {
+      "epoch": 0.11138706072271204,
+      "grad_norm": 0.41261720657348633,
+      "learning_rate": 5e-05,
+      "loss": 2.6938,
+      "step": 897
+    },
+    {
+      "epoch": 0.11151123804793245,
+      "grad_norm": 0.3548716604709625,
+      "learning_rate": 5e-05,
+      "loss": 2.0738,
+      "step": 898
+    },
+    {
+      "epoch": 0.11163541537315286,
+      "grad_norm": 0.2646797001361847,
+      "learning_rate": 5e-05,
+      "loss": 1.8559,
+      "step": 899
+    },
+    {
+      "epoch": 0.11175959269837328,
+      "grad_norm": 0.31071344017982483,
+      "learning_rate": 5e-05,
+      "loss": 2.0802,
+      "step": 900
+    },
+    {
+      "epoch": 0.11188377002359369,
+      "grad_norm": 0.22349640727043152,
+      "learning_rate": 5e-05,
+      "loss": 1.6962,
+      "step": 901
+    },
+    {
+      "epoch": 0.11200794734881411,
+      "grad_norm": 0.3751986026763916,
+      "learning_rate": 5e-05,
+      "loss": 2.2044,
+      "step": 902
+    },
+    {
+      "epoch": 0.11213212467403452,
+      "grad_norm": 0.23499399423599243,
+      "learning_rate": 5e-05,
+      "loss": 1.7119,
+      "step": 903
+    },
+    {
+      "epoch": 0.11225630199925493,
+      "grad_norm": 0.2599724531173706,
+      "learning_rate": 5e-05,
+      "loss": 1.7784,
+      "step": 904
+    },
+    {
+      "epoch": 0.11238047932447535,
+      "grad_norm": 0.24458616971969604,
+      "learning_rate": 5e-05,
+      "loss": 1.7294,
+      "step": 905
+    },
+    {
+      "epoch": 0.11250465664969576,
+      "grad_norm": 0.2781731188297272,
+      "learning_rate": 5e-05,
+      "loss": 2.0202,
+      "step": 906
+    },
+    {
+      "epoch": 0.11262883397491619,
+      "grad_norm": 0.22873641550540924,
+      "learning_rate": 5e-05,
+      "loss": 1.7724,
+      "step": 907
+    },
+    {
+      "epoch": 0.1127530113001366,
+      "grad_norm": 0.27547910809516907,
+      "learning_rate": 5e-05,
+      "loss": 2.1615,
+      "step": 908
+    },
+    {
+      "epoch": 0.112877188625357,
+      "grad_norm": 0.35256701707839966,
+      "learning_rate": 5e-05,
+      "loss": 2.1046,
+      "step": 909
+    },
+    {
+      "epoch": 0.11300136595057743,
+      "grad_norm": 0.31568098068237305,
+      "learning_rate": 5e-05,
+      "loss": 2.2322,
+      "step": 910
+    },
+    {
+      "epoch": 0.11312554327579784,
+      "grad_norm": 0.3242977559566498,
+      "learning_rate": 5e-05,
+      "loss": 2.2185,
+      "step": 911
+    },
+    {
+      "epoch": 0.11324972060101825,
+      "grad_norm": 0.30523544549942017,
+      "learning_rate": 5e-05,
+      "loss": 2.3313,
+      "step": 912
+    },
+    {
+      "epoch": 0.11337389792623867,
+      "grad_norm": 0.22899942100048065,
+      "learning_rate": 5e-05,
+      "loss": 1.9237,
+      "step": 913
+    },
+    {
+      "epoch": 0.11349807525145908,
+      "grad_norm": 0.2607531249523163,
+      "learning_rate": 5e-05,
+      "loss": 1.7515,
+      "step": 914
+    },
+    {
+      "epoch": 0.1136222525766795,
+      "grad_norm": 0.25846919417381287,
+      "learning_rate": 5e-05,
+      "loss": 1.8115,
+      "step": 915
+    },
+    {
+      "epoch": 0.11374642990189991,
+      "grad_norm": 0.36639493703842163,
+      "learning_rate": 5e-05,
+      "loss": 2.5123,
+      "step": 916
+    },
+    {
+      "epoch": 0.11387060722712032,
+      "grad_norm": 0.24815653264522552,
+      "learning_rate": 5e-05,
+      "loss": 1.8082,
+      "step": 917
+    },
+    {
+      "epoch": 0.11399478455234074,
+      "grad_norm": 0.23351258039474487,
+      "learning_rate": 5e-05,
+      "loss": 1.87,
+      "step": 918
+    },
+    {
+      "epoch": 0.11411896187756115,
+      "grad_norm": 0.31201139092445374,
+      "learning_rate": 5e-05,
+      "loss": 2.3676,
+      "step": 919
+    },
+    {
+      "epoch": 0.11424313920278158,
+      "grad_norm": 0.27010735869407654,
+      "learning_rate": 5e-05,
+      "loss": 2.1338,
+      "step": 920
+    },
+    {
+      "epoch": 0.11436731652800199,
+      "grad_norm": 0.2572576403617859,
+      "learning_rate": 5e-05,
+      "loss": 2.1985,
+      "step": 921
+    },
+    {
+      "epoch": 0.1144914938532224,
+      "grad_norm": 0.2479707896709442,
+      "learning_rate": 5e-05,
+      "loss": 2.0943,
+      "step": 922
+    },
+    {
+      "epoch": 0.11461567117844282,
+      "grad_norm": 0.24821355938911438,
+      "learning_rate": 5e-05,
+      "loss": 2.0318,
+      "step": 923
+    },
+    {
+      "epoch": 0.11473984850366323,
+      "grad_norm": 0.3458319902420044,
+      "learning_rate": 5e-05,
+      "loss": 2.4627,
+      "step": 924
+    },
+    {
+      "epoch": 0.11486402582888365,
+      "grad_norm": 0.2778370976448059,
+      "learning_rate": 5e-05,
+      "loss": 2.0825,
+      "step": 925
+    },
+    {
+      "epoch": 0.11498820315410406,
+      "grad_norm": 0.24979355931282043,
+      "learning_rate": 5e-05,
+      "loss": 1.9581,
+      "step": 926
+    },
+    {
+      "epoch": 0.11511238047932447,
+      "grad_norm": 0.24766133725643158,
+      "learning_rate": 5e-05,
+      "loss": 1.9789,
+      "step": 927
+    },
+    {
+      "epoch": 0.11523655780454489,
+      "grad_norm": 0.27076855301856995,
+      "learning_rate": 5e-05,
+      "loss": 2.3324,
+      "step": 928
+    },
+    {
+      "epoch": 0.1153607351297653,
+      "grad_norm": 0.24753974378108978,
+      "learning_rate": 5e-05,
+      "loss": 1.8434,
+      "step": 929
+    },
+    {
+      "epoch": 0.11548491245498572,
+      "grad_norm": 0.3660670220851898,
+      "learning_rate": 5e-05,
+      "loss": 2.412,
+      "step": 930
+    },
+    {
+      "epoch": 0.11560908978020613,
+      "grad_norm": 0.25543397665023804,
+      "learning_rate": 5e-05,
+      "loss": 1.8726,
+      "step": 931
+    },
+    {
+      "epoch": 0.11573326710542654,
+      "grad_norm": 0.5722286105155945,
+      "learning_rate": 5e-05,
+      "loss": 2.9469,
+      "step": 932
+    },
+    {
+      "epoch": 0.11585744443064697,
+      "grad_norm": 0.24585482478141785,
+      "learning_rate": 5e-05,
+      "loss": 1.8847,
+      "step": 933
+    },
+    {
+      "epoch": 0.11598162175586738,
+      "grad_norm": 0.3858698904514313,
+      "learning_rate": 5e-05,
+      "loss": 2.2881,
+      "step": 934
+    },
+    {
+      "epoch": 0.1161057990810878,
+      "grad_norm": 0.679387092590332,
+      "learning_rate": 5e-05,
+      "loss": 3.9892,
+      "step": 935
+    },
+    {
+      "epoch": 0.11622997640630821,
+      "grad_norm": 0.4026963412761688,
+      "learning_rate": 5e-05,
+      "loss": 2.43,
+      "step": 936
+    },
+    {
+      "epoch": 0.11635415373152862,
+      "grad_norm": 0.3264620304107666,
+      "learning_rate": 5e-05,
+      "loss": 2.5949,
+      "step": 937
+    },
+    {
+      "epoch": 0.11647833105674904,
+      "grad_norm": 0.2738242447376251,
+      "learning_rate": 5e-05,
+      "loss": 1.8355,
+      "step": 938
+    },
+    {
+      "epoch": 0.11660250838196945,
+      "grad_norm": 0.30453240871429443,
+      "learning_rate": 5e-05,
+      "loss": 1.804,
+      "step": 939
+    },
+    {
+      "epoch": 0.11672668570718987,
+      "grad_norm": 0.30571305751800537,
+      "learning_rate": 5e-05,
+      "loss": 2.1956,
+      "step": 940
+    },
+    {
+      "epoch": 0.11685086303241028,
+      "grad_norm": 0.2804155945777893,
+      "learning_rate": 5e-05,
+      "loss": 1.7701,
+      "step": 941
+    },
+    {
+      "epoch": 0.11697504035763069,
+      "grad_norm": 0.3185891807079315,
+      "learning_rate": 5e-05,
+      "loss": 2.2808,
+      "step": 942
+    },
+    {
+      "epoch": 0.11709921768285111,
+      "grad_norm": 0.2636656165122986,
+      "learning_rate": 5e-05,
+      "loss": 2.0351,
+      "step": 943
+    },
+    {
+      "epoch": 0.11722339500807152,
+      "grad_norm": 0.2993179261684418,
+      "learning_rate": 5e-05,
+      "loss": 2.0193,
+      "step": 944
+    },
+    {
+      "epoch": 0.11734757233329195,
+      "grad_norm": 0.35081735253334045,
+      "learning_rate": 5e-05,
+      "loss": 2.2608,
+      "step": 945
+    },
+    {
+      "epoch": 0.11747174965851236,
+      "grad_norm": 0.23226995766162872,
+      "learning_rate": 5e-05,
+      "loss": 1.7364,
+      "step": 946
+    },
+    {
+      "epoch": 0.11759592698373277,
+      "grad_norm": 0.3048912286758423,
+      "learning_rate": 5e-05,
+      "loss": 2.0342,
+      "step": 947
+    },
+    {
+      "epoch": 0.11772010430895319,
+      "grad_norm": 0.25601980090141296,
+      "learning_rate": 5e-05,
+      "loss": 1.8795,
+      "step": 948
+    },
+    {
+      "epoch": 0.1178442816341736,
+      "grad_norm": 0.31035560369491577,
+      "learning_rate": 5e-05,
+      "loss": 2.201,
+      "step": 949
+    },
+    {
+      "epoch": 0.11796845895939402,
+      "grad_norm": 0.2591751515865326,
+      "learning_rate": 5e-05,
+      "loss": 1.8405,
+      "step": 950
+    },
+    {
+      "epoch": 0.11809263628461443,
+      "grad_norm": 0.3209805190563202,
+      "learning_rate": 5e-05,
+      "loss": 2.2541,
+      "step": 951
+    },
+    {
+      "epoch": 0.11821681360983484,
+      "grad_norm": 0.30173051357269287,
+      "learning_rate": 5e-05,
+      "loss": 2.313,
+      "step": 952
+    },
+    {
+      "epoch": 0.11834099093505526,
+      "grad_norm": 0.2397630214691162,
+      "learning_rate": 5e-05,
+      "loss": 1.8164,
+      "step": 953
+    },
+    {
+      "epoch": 0.11846516826027567,
+      "grad_norm": 0.24437163770198822,
+      "learning_rate": 5e-05,
+      "loss": 1.9928,
+      "step": 954
+    },
+    {
+      "epoch": 0.1185893455854961,
+      "grad_norm": 0.4401412606239319,
+      "learning_rate": 5e-05,
+      "loss": 2.5289,
+      "step": 955
+    },
+    {
+      "epoch": 0.1187135229107165,
+      "grad_norm": 0.41325151920318604,
+      "learning_rate": 5e-05,
+      "loss": 2.6033,
+      "step": 956
+    },
+    {
+      "epoch": 0.11883770023593691,
+      "grad_norm": 0.23168908059597015,
+      "learning_rate": 5e-05,
+      "loss": 1.7914,
+      "step": 957
+    },
+    {
+      "epoch": 0.11896187756115734,
+      "grad_norm": 0.3848232626914978,
+      "learning_rate": 5e-05,
+      "loss": 2.2646,
+      "step": 958
+    },
+    {
+      "epoch": 0.11908605488637775,
+      "grad_norm": 0.39556726813316345,
+      "learning_rate": 5e-05,
+      "loss": 2.1765,
+      "step": 959
+    },
+    {
+      "epoch": 0.11921023221159817,
+      "grad_norm": 0.31372812390327454,
+      "learning_rate": 5e-05,
+      "loss": 2.0435,
+      "step": 960
+    },
+    {
+      "epoch": 0.11933440953681858,
+      "grad_norm": 0.5253868103027344,
+      "learning_rate": 5e-05,
+      "loss": 2.3224,
+      "step": 961
+    },
+    {
+      "epoch": 0.11945858686203899,
+      "grad_norm": 0.3625740110874176,
+      "learning_rate": 5e-05,
+      "loss": 2.0779,
+      "step": 962
+    },
+    {
+      "epoch": 0.11958276418725941,
+      "grad_norm": 0.24666467308998108,
+      "learning_rate": 5e-05,
+      "loss": 1.7768,
+      "step": 963
+    },
+    {
+      "epoch": 0.11970694151247982,
+      "grad_norm": 1.0668760538101196,
+      "learning_rate": 5e-05,
+      "loss": 3.7203,
+      "step": 964
+    },
+    {
+      "epoch": 0.11983111883770024,
+      "grad_norm": 0.2652861773967743,
+      "learning_rate": 5e-05,
+      "loss": 1.8322,
+      "step": 965
+    },
+    {
+      "epoch": 0.11995529616292065,
+      "grad_norm": 0.5147027373313904,
+      "learning_rate": 5e-05,
+      "loss": 2.3742,
+      "step": 966
+    },
+    {
+      "epoch": 0.12007947348814106,
+      "grad_norm": 0.692245602607727,
+      "learning_rate": 5e-05,
+      "loss": 2.5479,
+      "step": 967
+    },
+    {
+      "epoch": 0.12020365081336148,
+      "grad_norm": 0.2837624251842499,
+      "learning_rate": 5e-05,
+      "loss": 2.2053,
+      "step": 968
+    },
+    {
+      "epoch": 0.1203278281385819,
+      "grad_norm": 0.34765708446502686,
+      "learning_rate": 5e-05,
+      "loss": 2.9103,
+      "step": 969
+    },
+    {
+      "epoch": 0.1204520054638023,
+      "grad_norm": 0.4444332420825958,
+      "learning_rate": 5e-05,
+      "loss": 2.5352,
+      "step": 970
+    },
+    {
+      "epoch": 0.12057618278902273,
+      "grad_norm": 0.40752092003822327,
+      "learning_rate": 5e-05,
+      "loss": 2.3765,
+      "step": 971
+    },
+    {
+      "epoch": 0.12070036011424314,
+      "grad_norm": 0.26922377943992615,
+      "learning_rate": 5e-05,
+      "loss": 2.0827,
+      "step": 972
+    },
+    {
+      "epoch": 0.12082453743946356,
+      "grad_norm": 0.26593759655952454,
+      "learning_rate": 5e-05,
+      "loss": 1.7902,
+      "step": 973
+    },
+    {
+      "epoch": 0.12094871476468397,
+      "grad_norm": 0.3797745704650879,
+      "learning_rate": 5e-05,
+      "loss": 2.4203,
+      "step": 974
+    },
+    {
+      "epoch": 0.12107289208990438,
+      "grad_norm": 0.28492528200149536,
+      "learning_rate": 5e-05,
+      "loss": 1.8963,
+      "step": 975
+    },
+    {
+      "epoch": 0.1211970694151248,
+      "grad_norm": 0.35106533765792847,
+      "learning_rate": 5e-05,
+      "loss": 2.001,
+      "step": 976
+    },
+    {
+      "epoch": 0.12132124674034521,
+      "grad_norm": 0.30248552560806274,
+      "learning_rate": 5e-05,
+      "loss": 1.8161,
+      "step": 977
+    },
+    {
+      "epoch": 0.12144542406556563,
+      "grad_norm": 0.2563718259334564,
+      "learning_rate": 5e-05,
+      "loss": 1.766,
+      "step": 978
+    },
+    {
+      "epoch": 0.12156960139078604,
+      "grad_norm": 0.32553622126579285,
+      "learning_rate": 5e-05,
+      "loss": 2.3881,
+      "step": 979
+    },
+    {
+      "epoch": 0.12169377871600645,
+      "grad_norm": 0.2741480767726898,
+      "learning_rate": 5e-05,
+      "loss": 2.12,
+      "step": 980
+    },
+    {
+      "epoch": 0.12181795604122687,
+      "grad_norm": 0.3366210162639618,
+      "learning_rate": 5e-05,
+      "loss": 1.9177,
+      "step": 981
+    },
+    {
+      "epoch": 0.12194213336644728,
+      "grad_norm": 0.31167927384376526,
+      "learning_rate": 5e-05,
+      "loss": 1.9777,
+      "step": 982
+    },
+    {
+      "epoch": 0.1220663106916677,
+      "grad_norm": 0.310498982667923,
+      "learning_rate": 5e-05,
+      "loss": 2.22,
+      "step": 983
+    },
+    {
+      "epoch": 0.12219048801688812,
+      "grad_norm": 0.3025086224079132,
+      "learning_rate": 5e-05,
+      "loss": 2.2006,
+      "step": 984
+    },
+    {
+      "epoch": 0.12231466534210853,
+      "grad_norm": 0.2378969043493271,
+      "learning_rate": 5e-05,
+      "loss": 1.9422,
+      "step": 985
+    },
+    {
+      "epoch": 0.12243884266732895,
+      "grad_norm": 0.28100481629371643,
+      "learning_rate": 5e-05,
+      "loss": 2.1352,
+      "step": 986
+    },
+    {
+      "epoch": 0.12256301999254936,
+      "grad_norm": 0.3556027114391327,
+      "learning_rate": 5e-05,
+      "loss": 2.6645,
+      "step": 987
+    },
+    {
+      "epoch": 0.12268719731776978,
+      "grad_norm": 0.2855495512485504,
+      "learning_rate": 5e-05,
+      "loss": 1.8554,
+      "step": 988
+    },
+    {
+      "epoch": 0.12281137464299019,
+      "grad_norm": 0.27533024549484253,
+      "learning_rate": 5e-05,
+      "loss": 1.9915,
+      "step": 989
+    },
+    {
+      "epoch": 0.1229355519682106,
+      "grad_norm": 0.2480265200138092,
+      "learning_rate": 5e-05,
+      "loss": 1.7773,
+      "step": 990
+    },
+    {
+      "epoch": 0.12305972929343102,
+      "grad_norm": 0.5346510410308838,
+      "learning_rate": 5e-05,
+      "loss": 2.8553,
+      "step": 991
+    },
+    {
+      "epoch": 0.12318390661865143,
+      "grad_norm": 0.26068994402885437,
+      "learning_rate": 5e-05,
+      "loss": 2.0096,
+      "step": 992
+    },
+    {
+      "epoch": 0.12330808394387185,
+      "grad_norm": 0.4033184051513672,
+      "learning_rate": 5e-05,
+      "loss": 2.6374,
+      "step": 993
+    },
+    {
+      "epoch": 0.12343226126909226,
+      "grad_norm": 0.2643042802810669,
+      "learning_rate": 5e-05,
+      "loss": 1.8803,
+      "step": 994
+    },
+    {
+      "epoch": 0.12355643859431267,
+      "grad_norm": 0.3660152852535248,
+      "learning_rate": 5e-05,
+      "loss": 2.3213,
+      "step": 995
+    },
+    {
+      "epoch": 0.1236806159195331,
+      "grad_norm": 0.23946939408779144,
+      "learning_rate": 5e-05,
+      "loss": 1.8943,
+      "step": 996
+    },
+    {
+      "epoch": 0.1238047932447535,
+      "grad_norm": 0.3469591438770294,
+      "learning_rate": 5e-05,
+      "loss": 2.2684,
+      "step": 997
+    },
+    {
+      "epoch": 0.12392897056997393,
+      "grad_norm": 0.2437586784362793,
+      "learning_rate": 5e-05,
+      "loss": 1.993,
+      "step": 998
+    },
+    {
+      "epoch": 0.12405314789519434,
+      "grad_norm": 0.2569863498210907,
+      "learning_rate": 5e-05,
+      "loss": 2.2131,
+      "step": 999
+    },
+    {
+      "epoch": 0.12417732522041475,
+      "grad_norm": 0.26156288385391235,
+      "learning_rate": 5e-05,
+      "loss": 1.9116,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12430150254563517,
+      "grad_norm": 0.24384045600891113,
+      "learning_rate": 5e-05,
+      "loss": 1.9923,
+      "step": 1001
+    },
+    {
+      "epoch": 0.12442567987085558,
+      "grad_norm": 0.32400116324424744,
+      "learning_rate": 5e-05,
+      "loss": 2.5213,
+      "step": 1002
+    },
+    {
+      "epoch": 0.124549857196076,
+      "grad_norm": 0.2604520916938782,
+      "learning_rate": 5e-05,
+      "loss": 1.9388,
+      "step": 1003
+    },
+    {
+      "epoch": 0.12467403452129641,
+      "grad_norm": 0.27235329151153564,
+      "learning_rate": 5e-05,
+      "loss": 2.0165,
+      "step": 1004
+    },
+    {
+      "epoch": 0.12479821184651682,
+      "grad_norm": 0.33321765065193176,
+      "learning_rate": 5e-05,
+      "loss": 2.4184,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12492238917173724,
+      "grad_norm": 0.22587352991104126,
+      "learning_rate": 5e-05,
+      "loss": 1.7306,
+      "step": 1006
+    },
+    {
+      "epoch": 0.12504656649695767,
+      "grad_norm": 0.25614914298057556,
+      "learning_rate": 5e-05,
+      "loss": 1.898,
+      "step": 1007
+    },
+    {
+      "epoch": 0.12517074382217808,
+      "grad_norm": 0.3783401548862457,
+      "learning_rate": 5e-05,
+      "loss": 2.7309,
+      "step": 1008
+    },
+    {
+      "epoch": 0.12529492114739849,
+      "grad_norm": 0.281352698802948,
+      "learning_rate": 5e-05,
+      "loss": 1.736,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1254190984726189,
+      "grad_norm": 0.26021233201026917,
+      "learning_rate": 5e-05,
+      "loss": 1.8002,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1255432757978393,
+      "grad_norm": 0.2724604308605194,
+      "learning_rate": 5e-05,
+      "loss": 2.1816,
+      "step": 1011
+    },
+    {
+      "epoch": 0.12566745312305974,
+      "grad_norm": 0.2835272550582886,
+      "learning_rate": 5e-05,
+      "loss": 2.614,
+      "step": 1012
+    },
+    {
+      "epoch": 0.12579163044828015,
+      "grad_norm": 0.25480642914772034,
+      "learning_rate": 5e-05,
+      "loss": 2.086,
+      "step": 1013
+    },
+    {
+      "epoch": 0.12591580777350056,
+      "grad_norm": 0.3954196274280548,
+      "learning_rate": 5e-05,
+      "loss": 2.0803,
+      "step": 1014
+    },
+    {
+      "epoch": 0.12603998509872097,
+      "grad_norm": 0.5192334651947021,
+      "learning_rate": 5e-05,
+      "loss": 3.2472,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12616416242394138,
+      "grad_norm": 0.2652159631252289,
+      "learning_rate": 5e-05,
+      "loss": 2.113,
+      "step": 1016
+    },
+    {
+      "epoch": 0.12628833974916182,
+      "grad_norm": 0.27216076850891113,
+      "learning_rate": 5e-05,
+      "loss": 1.9084,
+      "step": 1017
+    },
+    {
+      "epoch": 0.12641251707438222,
+      "grad_norm": 0.28878965973854065,
+      "learning_rate": 5e-05,
+      "loss": 2.0184,
+      "step": 1018
+    },
+    {
+      "epoch": 0.12653669439960263,
+      "grad_norm": 0.2580112814903259,
+      "learning_rate": 5e-05,
+      "loss": 1.7244,
+      "step": 1019
+    },
+    {
+      "epoch": 0.12666087172482304,
+      "grad_norm": 0.26858070492744446,
+      "learning_rate": 5e-05,
+      "loss": 1.9107,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12678504905004345,
+      "grad_norm": 0.28445541858673096,
+      "learning_rate": 5e-05,
+      "loss": 2.0212,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1269092263752639,
+      "grad_norm": 0.2333497256040573,
+      "learning_rate": 5e-05,
+      "loss": 1.7624,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1270334037004843,
+      "grad_norm": 0.31717449426651,
+      "learning_rate": 5e-05,
+      "loss": 2.3561,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1271575810257047,
+      "grad_norm": 0.26815950870513916,
+      "learning_rate": 5e-05,
+      "loss": 1.8738,
+      "step": 1024
+    },
+    {
+      "epoch": 0.12728175835092512,
+      "grad_norm": 0.38293471932411194,
+      "learning_rate": 5e-05,
+      "loss": 2.283,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12740593567614553,
+      "grad_norm": 0.5300688743591309,
+      "learning_rate": 5e-05,
+      "loss": 2.8997,
+      "step": 1026
+    },
+    {
+      "epoch": 0.12753011300136596,
+      "grad_norm": 0.33331796526908875,
+      "learning_rate": 5e-05,
+      "loss": 2.5145,
+      "step": 1027
+    },
+    {
+      "epoch": 0.12765429032658637,
+      "grad_norm": 0.24202688038349152,
+      "learning_rate": 5e-05,
+      "loss": 1.7676,
+      "step": 1028
+    },
+    {
+      "epoch": 0.12777846765180678,
+      "grad_norm": 0.23438231647014618,
+      "learning_rate": 5e-05,
+      "loss": 1.7615,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1279026449770272,
+      "grad_norm": 0.2875107526779175,
+      "learning_rate": 5e-05,
+      "loss": 1.7874,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1280268223022476,
+      "grad_norm": 0.22003595530986786,
+      "learning_rate": 5e-05,
+      "loss": 1.596,
+      "step": 1031
+    },
+    {
+      "epoch": 0.12815099962746804,
+      "grad_norm": 0.26276060938835144,
+      "learning_rate": 5e-05,
+      "loss": 1.9154,
+      "step": 1032
+    },
+    {
+      "epoch": 0.12827517695268845,
+      "grad_norm": 0.36533546447753906,
+      "learning_rate": 5e-05,
+      "loss": 2.8755,
+      "step": 1033
+    },
+    {
+      "epoch": 0.12839935427790886,
+      "grad_norm": 0.38885048031806946,
+      "learning_rate": 5e-05,
+      "loss": 2.9473,
+      "step": 1034
+    },
+    {
+      "epoch": 0.12852353160312927,
+      "grad_norm": 0.2708134949207306,
+      "learning_rate": 5e-05,
+      "loss": 1.9407,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12864770892834967,
+      "grad_norm": 0.22453922033309937,
+      "learning_rate": 5e-05,
+      "loss": 1.7948,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1287718862535701,
+      "grad_norm": 0.3211478292942047,
+      "learning_rate": 5e-05,
+      "loss": 2.0837,
+      "step": 1037
+    },
+    {
+      "epoch": 0.12889606357879052,
+      "grad_norm": 0.2573651671409607,
+      "learning_rate": 5e-05,
+      "loss": 1.9984,
+      "step": 1038
+    },
+    {
+      "epoch": 0.12902024090401093,
+      "grad_norm": 0.3328755795955658,
+      "learning_rate": 5e-05,
+      "loss": 2.0742,
+      "step": 1039
+    },
+    {
+      "epoch": 0.12914441822923134,
+      "grad_norm": 0.36100730299949646,
+      "learning_rate": 5e-05,
+      "loss": 2.1482,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12926859555445175,
+      "grad_norm": 0.3397010564804077,
+      "learning_rate": 5e-05,
+      "loss": 2.6753,
+      "step": 1041
+    },
+    {
+      "epoch": 0.12939277287967219,
+      "grad_norm": 0.29403823614120483,
+      "learning_rate": 5e-05,
+      "loss": 2.1762,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1295169502048926,
+      "grad_norm": 0.3259550929069519,
+      "learning_rate": 5e-05,
+      "loss": 2.0438,
+      "step": 1043
+    },
+    {
+      "epoch": 0.129641127530113,
+      "grad_norm": 0.2814289927482605,
+      "learning_rate": 5e-05,
+      "loss": 2.1891,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1297653048553334,
+      "grad_norm": 0.2861133813858032,
+      "learning_rate": 5e-05,
+      "loss": 1.9424,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12988948218055382,
+      "grad_norm": 0.2853018641471863,
+      "learning_rate": 5e-05,
+      "loss": 2.0572,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13001365950577426,
+      "grad_norm": 0.321277379989624,
+      "learning_rate": 5e-05,
+      "loss": 2.4332,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13013783683099467,
+      "grad_norm": 0.2422984093427658,
+      "learning_rate": 5e-05,
+      "loss": 1.8778,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13026201415621508,
+      "grad_norm": 0.2669890522956848,
+      "learning_rate": 5e-05,
+      "loss": 2.0212,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1303861914814355,
+      "grad_norm": 0.24418428540229797,
+      "learning_rate": 5e-05,
+      "loss": 1.9481,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1305103688066559,
+      "grad_norm": 0.2489718794822693,
+      "learning_rate": 5e-05,
+      "loss": 2.1197,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1306345461318763,
+      "grad_norm": 0.2919555902481079,
+      "learning_rate": 5e-05,
+      "loss": 2.1435,
+      "step": 1052
+    },
+    {
+      "epoch": 0.13075872345709674,
+      "grad_norm": 0.2646684944629669,
+      "learning_rate": 5e-05,
+      "loss": 2.173,
+      "step": 1053
+    },
+    {
+      "epoch": 0.13088290078231715,
+      "grad_norm": 0.3247489333152771,
+      "learning_rate": 5e-05,
+      "loss": 2.2599,
+      "step": 1054
+    },
+    {
+      "epoch": 0.13100707810753756,
+      "grad_norm": 0.22721542418003082,
+      "learning_rate": 5e-05,
+      "loss": 1.7981,
+      "step": 1055
+    },
+    {
+      "epoch": 0.13113125543275797,
+      "grad_norm": 0.4916575253009796,
+      "learning_rate": 5e-05,
+      "loss": 2.5025,
+      "step": 1056
+    },
+    {
+      "epoch": 0.13125543275797838,
+      "grad_norm": 0.45959606766700745,
+      "learning_rate": 5e-05,
+      "loss": 2.8821,
+      "step": 1057
+    },
+    {
+      "epoch": 0.13137961008319882,
+      "grad_norm": 0.25712233781814575,
+      "learning_rate": 5e-05,
+      "loss": 1.6615,
+      "step": 1058
+    },
+    {
+      "epoch": 0.13150378740841923,
+      "grad_norm": 0.3775932788848877,
+      "learning_rate": 5e-05,
+      "loss": 1.9113,
+      "step": 1059
+    },
+    {
+      "epoch": 0.13162796473363964,
+      "grad_norm": 0.6293278932571411,
+      "learning_rate": 5e-05,
+      "loss": 2.4153,
+      "step": 1060
+    },
+    {
+      "epoch": 0.13175214205886004,
+      "grad_norm": 0.3483099043369293,
+      "learning_rate": 5e-05,
+      "loss": 2.0118,
+      "step": 1061
+    },
+    {
+      "epoch": 0.13187631938408045,
+      "grad_norm": 0.3565744161605835,
+      "learning_rate": 5e-05,
+      "loss": 2.3465,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1320004967093009,
+      "grad_norm": 0.3302151560783386,
+      "learning_rate": 5e-05,
+      "loss": 1.9891,
+      "step": 1063
+    },
+    {
+      "epoch": 0.1321246740345213,
+      "grad_norm": 0.7692415714263916,
+      "learning_rate": 5e-05,
+      "loss": 1.9401,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1322488513597417,
+      "grad_norm": 0.3344350755214691,
+      "learning_rate": 5e-05,
+      "loss": 2.1948,
+      "step": 1065
+    },
+    {
+      "epoch": 0.13237302868496212,
+      "grad_norm": 0.2954474985599518,
+      "learning_rate": 5e-05,
+      "loss": 1.9716,
+      "step": 1066
+    },
+    {
+      "epoch": 0.13249720601018253,
+      "grad_norm": 0.5076240301132202,
+      "learning_rate": 5e-05,
+      "loss": 2.9537,
+      "step": 1067
+    },
+    {
+      "epoch": 0.13262138333540296,
+      "grad_norm": 0.2375306636095047,
+      "learning_rate": 5e-05,
+      "loss": 1.7894,
+      "step": 1068
+    },
+    {
+      "epoch": 0.13274556066062337,
+      "grad_norm": 0.5773283243179321,
+      "learning_rate": 5e-05,
+      "loss": 2.7361,
+      "step": 1069
+    },
+    {
+      "epoch": 0.13286973798584378,
+      "grad_norm": 0.3737553358078003,
+      "learning_rate": 5e-05,
+      "loss": 2.1509,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1329939153110642,
+      "grad_norm": 0.2960174083709717,
+      "learning_rate": 5e-05,
+      "loss": 2.3799,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1331180926362846,
+      "grad_norm": 0.28043243288993835,
+      "learning_rate": 5e-05,
+      "loss": 2.1211,
+      "step": 1072
+    },
+    {
+      "epoch": 0.13324226996150504,
+      "grad_norm": 0.28315070271492004,
+      "learning_rate": 5e-05,
+      "loss": 2.0942,
+      "step": 1073
+    },
+    {
+      "epoch": 0.13336644728672545,
+      "grad_norm": 0.2449692040681839,
+      "learning_rate": 5e-05,
+      "loss": 1.7595,
+      "step": 1074
+    },
+    {
+      "epoch": 0.13349062461194586,
+      "grad_norm": 0.6175221800804138,
+      "learning_rate": 5e-05,
+      "loss": 2.8706,
+      "step": 1075
+    },
+    {
+      "epoch": 0.13361480193716627,
+      "grad_norm": 0.37083110213279724,
+      "learning_rate": 5e-05,
+      "loss": 2.5191,
+      "step": 1076
+    },
+    {
+      "epoch": 0.13373897926238668,
+      "grad_norm": 0.25019583106040955,
+      "learning_rate": 5e-05,
+      "loss": 2.0465,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1338631565876071,
+      "grad_norm": 0.256204754114151,
+      "learning_rate": 5e-05,
+      "loss": 1.8876,
+      "step": 1078
+    },
+    {
+      "epoch": 0.13398733391282752,
+      "grad_norm": 0.29824960231781006,
+      "learning_rate": 5e-05,
+      "loss": 2.1895,
+      "step": 1079
+    },
+    {
+      "epoch": 0.13411151123804793,
+      "grad_norm": 0.24536696076393127,
+      "learning_rate": 5e-05,
+      "loss": 1.7937,
+      "step": 1080
+    },
+    {
+      "epoch": 0.13423568856326834,
+      "grad_norm": 0.47522780299186707,
+      "learning_rate": 5e-05,
+      "loss": 2.2545,
+      "step": 1081
+    },
+    {
+      "epoch": 0.13435986588848875,
+      "grad_norm": 0.4264921545982361,
+      "learning_rate": 5e-05,
+      "loss": 2.792,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1344840432137092,
+      "grad_norm": 0.23045669496059418,
+      "learning_rate": 5e-05,
+      "loss": 1.8701,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1346082205389296,
+      "grad_norm": 0.4328882694244385,
+      "learning_rate": 5e-05,
+      "loss": 2.7582,
+      "step": 1084
+    },
+    {
+      "epoch": 0.13473239786415,
+      "grad_norm": 0.31502556800842285,
+      "learning_rate": 5e-05,
+      "loss": 2.5346,
+      "step": 1085
+    },
+    {
+      "epoch": 0.13485657518937041,
+      "grad_norm": 0.3627992868423462,
+      "learning_rate": 5e-05,
+      "loss": 2.3319,
+      "step": 1086
+    },
+    {
+      "epoch": 0.13498075251459082,
+      "grad_norm": 0.3795703649520874,
+      "learning_rate": 5e-05,
+      "loss": 2.2837,
+      "step": 1087
+    },
+    {
+      "epoch": 0.13510492983981126,
+      "grad_norm": 0.3328198492527008,
+      "learning_rate": 5e-05,
+      "loss": 2.5716,
+      "step": 1088
+    },
+    {
+      "epoch": 0.13522910716503167,
+      "grad_norm": 0.2691793739795685,
+      "learning_rate": 5e-05,
+      "loss": 1.8648,
+      "step": 1089
+    },
+    {
+      "epoch": 0.13535328449025208,
+      "grad_norm": 0.2132241129875183,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1354774618154725,
+      "grad_norm": 0.29721713066101074,
+      "learning_rate": 5e-05,
+      "loss": 2.3337,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1356016391406929,
+      "grad_norm": 0.24550172686576843,
+      "learning_rate": 5e-05,
+      "loss": 1.7892,
+      "step": 1092
+    },
+    {
+      "epoch": 0.13572581646591333,
+      "grad_norm": 0.3057241439819336,
+      "learning_rate": 5e-05,
+      "loss": 2.2008,
+      "step": 1093
+    },
+    {
+      "epoch": 0.13584999379113374,
+      "grad_norm": 0.42028123140335083,
+      "learning_rate": 5e-05,
+      "loss": 2.4991,
+      "step": 1094
+    },
+    {
+      "epoch": 0.13597417111635415,
+      "grad_norm": 0.34639716148376465,
+      "learning_rate": 5e-05,
+      "loss": 2.4221,
+      "step": 1095
+    },
+    {
+      "epoch": 0.13609834844157456,
+      "grad_norm": 0.2971496284008026,
+      "learning_rate": 5e-05,
+      "loss": 1.9329,
+      "step": 1096
+    },
+    {
+      "epoch": 0.13622252576679497,
+      "grad_norm": 0.2966279685497284,
+      "learning_rate": 5e-05,
+      "loss": 2.1465,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1363467030920154,
+      "grad_norm": 0.2975577414035797,
+      "learning_rate": 5e-05,
+      "loss": 2.3422,
+      "step": 1098
+    },
+    {
+      "epoch": 0.13647088041723582,
+      "grad_norm": 0.5118456482887268,
+      "learning_rate": 5e-05,
+      "loss": 2.5421,
+      "step": 1099
+    },
+    {
+      "epoch": 0.13659505774245623,
+      "grad_norm": 0.23964305222034454,
+      "learning_rate": 5e-05,
+      "loss": 1.9165,
+      "step": 1100
+    },
+    {
+      "epoch": 0.13671923506767664,
+      "grad_norm": 0.30365464091300964,
+      "learning_rate": 5e-05,
+      "loss": 1.9171,
+      "step": 1101
+    },
+    {
+      "epoch": 0.13684341239289705,
+      "grad_norm": 0.2908100187778473,
+      "learning_rate": 5e-05,
+      "loss": 1.7971,
+      "step": 1102
+    },
+    {
+      "epoch": 0.13696758971811748,
+      "grad_norm": 0.3548651337623596,
+      "learning_rate": 5e-05,
+      "loss": 2.6751,
+      "step": 1103
+    },
+    {
+      "epoch": 0.1370917670433379,
+      "grad_norm": 0.3089209198951721,
+      "learning_rate": 5e-05,
+      "loss": 2.2546,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1372159443685583,
+      "grad_norm": 0.29616862535476685,
+      "learning_rate": 5e-05,
+      "loss": 2.1639,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1373401216937787,
+      "grad_norm": 0.31919291615486145,
+      "learning_rate": 5e-05,
+      "loss": 2.4508,
+      "step": 1106
+    },
+    {
+      "epoch": 0.13746429901899912,
+      "grad_norm": 0.2513790428638458,
+      "learning_rate": 5e-05,
+      "loss": 2.0057,
+      "step": 1107
+    },
+    {
+      "epoch": 0.13758847634421956,
+      "grad_norm": 0.33529070019721985,
+      "learning_rate": 5e-05,
+      "loss": 2.4737,
+      "step": 1108
+    },
+    {
+      "epoch": 0.13771265366943997,
+      "grad_norm": 0.25810202956199646,
+      "learning_rate": 5e-05,
+      "loss": 2.0787,
+      "step": 1109
+    },
+    {
+      "epoch": 0.13783683099466038,
+      "grad_norm": 0.27952781319618225,
+      "learning_rate": 5e-05,
+      "loss": 2.151,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13796100831988078,
+      "grad_norm": 0.22980237007141113,
+      "learning_rate": 5e-05,
+      "loss": 1.7528,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1380851856451012,
+      "grad_norm": 0.45032209157943726,
+      "learning_rate": 5e-05,
+      "loss": 2.5959,
+      "step": 1112
+    },
+    {
+      "epoch": 0.13820936297032163,
+      "grad_norm": 0.22644434869289398,
+      "learning_rate": 5e-05,
+      "loss": 1.7315,
+      "step": 1113
+    },
+    {
+      "epoch": 0.13833354029554204,
+      "grad_norm": 0.27581727504730225,
+      "learning_rate": 5e-05,
+      "loss": 2.2056,
+      "step": 1114
+    },
+    {
+      "epoch": 0.13845771762076245,
+      "grad_norm": 0.32302621006965637,
+      "learning_rate": 5e-05,
+      "loss": 2.6313,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13858189494598286,
+      "grad_norm": 0.3314933776855469,
+      "learning_rate": 5e-05,
+      "loss": 2.3385,
+      "step": 1116
+    },
+    {
+      "epoch": 0.13870607227120327,
+      "grad_norm": 0.4746491014957428,
+      "learning_rate": 5e-05,
+      "loss": 3.0895,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1388302495964237,
+      "grad_norm": 0.2554611563682556,
+      "learning_rate": 5e-05,
+      "loss": 1.8501,
+      "step": 1118
+    },
+    {
+      "epoch": 0.13895442692164411,
+      "grad_norm": 0.3268207013607025,
+      "learning_rate": 5e-05,
+      "loss": 2.7787,
+      "step": 1119
+    },
+    {
+      "epoch": 0.13907860424686452,
+      "grad_norm": 0.40559980273246765,
+      "learning_rate": 5e-05,
+      "loss": 2.3169,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13920278157208493,
+      "grad_norm": 0.27972233295440674,
+      "learning_rate": 5e-05,
+      "loss": 1.9973,
+      "step": 1121
+    },
+    {
+      "epoch": 0.13932695889730534,
+      "grad_norm": 0.2509663701057434,
+      "learning_rate": 5e-05,
+      "loss": 1.9772,
+      "step": 1122
+    },
+    {
+      "epoch": 0.13945113622252578,
+      "grad_norm": 0.2372690737247467,
+      "learning_rate": 5e-05,
+      "loss": 1.8282,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1395753135477462,
+      "grad_norm": 0.33050277829170227,
+      "learning_rate": 5e-05,
+      "loss": 2.3587,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1396994908729666,
+      "grad_norm": 0.4594133496284485,
+      "learning_rate": 5e-05,
+      "loss": 2.6167,
+      "step": 1125
+    },
+    {
+      "epoch": 0.139823668198187,
+      "grad_norm": 0.2318199723958969,
+      "learning_rate": 5e-05,
+      "loss": 1.7646,
+      "step": 1126
+    },
+    {
+      "epoch": 0.13994784552340742,
+      "grad_norm": 0.2286657989025116,
+      "learning_rate": 5e-05,
+      "loss": 1.8169,
+      "step": 1127
+    },
+    {
+      "epoch": 0.14007202284862785,
+      "grad_norm": 0.2952748239040375,
+      "learning_rate": 5e-05,
+      "loss": 1.8384,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14019620017384826,
+      "grad_norm": 0.2604459226131439,
+      "learning_rate": 5e-05,
+      "loss": 1.9408,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14032037749906867,
+      "grad_norm": 0.33005622029304504,
+      "learning_rate": 5e-05,
+      "loss": 2.5149,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14044455482428908,
+      "grad_norm": 0.5335607528686523,
+      "learning_rate": 5e-05,
+      "loss": 2.7443,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1405687321495095,
+      "grad_norm": 0.29719600081443787,
+      "learning_rate": 5e-05,
+      "loss": 2.2798,
+      "step": 1132
+    },
+    {
+      "epoch": 0.14069290947472993,
+      "grad_norm": 0.27328968048095703,
+      "learning_rate": 5e-05,
+      "loss": 2.0366,
+      "step": 1133
+    },
+    {
+      "epoch": 0.14081708679995034,
+      "grad_norm": 0.23124265670776367,
+      "learning_rate": 5e-05,
+      "loss": 1.811,
+      "step": 1134
+    },
+    {
+      "epoch": 0.14094126412517075,
+      "grad_norm": 0.2638339400291443,
+      "learning_rate": 5e-05,
+      "loss": 1.8189,
+      "step": 1135
+    },
+    {
+      "epoch": 0.14106544145039115,
+      "grad_norm": 0.26248520612716675,
+      "learning_rate": 5e-05,
+      "loss": 2.0001,
+      "step": 1136
+    },
+    {
+      "epoch": 0.14118961877561156,
+      "grad_norm": 0.2610619068145752,
+      "learning_rate": 5e-05,
+      "loss": 2.1195,
+      "step": 1137
+    },
+    {
+      "epoch": 0.141313796100832,
+      "grad_norm": 0.23357811570167542,
+      "learning_rate": 5e-05,
+      "loss": 1.7947,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1414379734260524,
+      "grad_norm": 0.22468189895153046,
+      "learning_rate": 5e-05,
+      "loss": 1.6522,
+      "step": 1139
+    },
+    {
+      "epoch": 0.14156215075127282,
+      "grad_norm": 0.26619935035705566,
+      "learning_rate": 5e-05,
+      "loss": 2.0381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.14168632807649323,
+      "grad_norm": 0.24181407690048218,
+      "learning_rate": 5e-05,
+      "loss": 1.842,
+      "step": 1141
+    },
+    {
+      "epoch": 0.14181050540171364,
+      "grad_norm": 0.28241896629333496,
+      "learning_rate": 5e-05,
+      "loss": 2.1041,
+      "step": 1142
+    },
+    {
+      "epoch": 0.14193468272693407,
+      "grad_norm": 0.3849261999130249,
+      "learning_rate": 5e-05,
+      "loss": 2.6363,
+      "step": 1143
+    },
+    {
+      "epoch": 0.14205886005215448,
+      "grad_norm": 0.2791067361831665,
+      "learning_rate": 5e-05,
+      "loss": 2.3319,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1421830373773749,
+      "grad_norm": 0.2378561645746231,
+      "learning_rate": 5e-05,
+      "loss": 1.808,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1423072147025953,
+      "grad_norm": 0.23445846140384674,
+      "learning_rate": 5e-05,
+      "loss": 1.73,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1424313920278157,
+      "grad_norm": 0.29622235894203186,
+      "learning_rate": 5e-05,
+      "loss": 2.3817,
+      "step": 1147
+    },
+    {
+      "epoch": 0.14255556935303615,
+      "grad_norm": 0.2245474010705948,
+      "learning_rate": 5e-05,
+      "loss": 1.6031,
+      "step": 1148
+    },
+    {
+      "epoch": 0.14267974667825656,
+      "grad_norm": 0.2519143521785736,
+      "learning_rate": 5e-05,
+      "loss": 1.929,
+      "step": 1149
+    },
+    {
+      "epoch": 0.14280392400347697,
+      "grad_norm": 0.3421010971069336,
+      "learning_rate": 5e-05,
+      "loss": 1.8719,
+      "step": 1150
+    },
+    {
+      "epoch": 0.14292810132869738,
+      "grad_norm": 0.2214202582836151,
+      "learning_rate": 5e-05,
+      "loss": 1.7547,
+      "step": 1151
+    },
+    {
+      "epoch": 0.14305227865391779,
+      "grad_norm": 0.3228437900543213,
+      "learning_rate": 5e-05,
+      "loss": 2.1783,
+      "step": 1152
+    },
+    {
+      "epoch": 0.14317645597913822,
+      "grad_norm": 0.24276112020015717,
+      "learning_rate": 5e-05,
+      "loss": 1.8954,
+      "step": 1153
+    },
+    {
+      "epoch": 0.14330063330435863,
+      "grad_norm": 0.3633849024772644,
+      "learning_rate": 5e-05,
+      "loss": 2.4981,
+      "step": 1154
+    },
+    {
+      "epoch": 0.14342481062957904,
+      "grad_norm": 0.2802623212337494,
+      "learning_rate": 5e-05,
+      "loss": 2.2688,
+      "step": 1155
+    },
+    {
+      "epoch": 0.14354898795479945,
+      "grad_norm": 0.281960666179657,
+      "learning_rate": 5e-05,
+      "loss": 2.02,
+      "step": 1156
+    },
+    {
+      "epoch": 0.14367316528001986,
+      "grad_norm": 0.3305203914642334,
+      "learning_rate": 5e-05,
+      "loss": 2.0643,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1437973426052403,
+      "grad_norm": 0.3385838568210602,
+      "learning_rate": 5e-05,
+      "loss": 2.2191,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1439215199304607,
+      "grad_norm": 0.29866835474967957,
+      "learning_rate": 5e-05,
+      "loss": 2.0052,
+      "step": 1159
+    },
+    {
+      "epoch": 0.14404569725568112,
+      "grad_norm": 0.42141684889793396,
+      "learning_rate": 5e-05,
+      "loss": 2.4322,
+      "step": 1160
+    },
+    {
+      "epoch": 0.14416987458090152,
+      "grad_norm": 0.30975401401519775,
+      "learning_rate": 5e-05,
+      "loss": 1.7659,
+      "step": 1161
+    },
+    {
+      "epoch": 0.14429405190612193,
+      "grad_norm": 0.2526281774044037,
+      "learning_rate": 5e-05,
+      "loss": 1.964,
+      "step": 1162
+    },
+    {
+      "epoch": 0.14441822923134234,
+      "grad_norm": 0.3553585410118103,
+      "learning_rate": 5e-05,
+      "loss": 2.407,
+      "step": 1163
+    },
+    {
+      "epoch": 0.14454240655656278,
+      "grad_norm": 0.25870001316070557,
+      "learning_rate": 5e-05,
+      "loss": 2.1386,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1446665838817832,
+      "grad_norm": 0.3086514472961426,
+      "learning_rate": 5e-05,
+      "loss": 1.9795,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1447907612070036,
+      "grad_norm": 0.32468122243881226,
+      "learning_rate": 5e-05,
+      "loss": 2.0295,
+      "step": 1166
+    },
+    {
+      "epoch": 0.144914938532224,
+      "grad_norm": 0.3432020843029022,
+      "learning_rate": 5e-05,
+      "loss": 2.2733,
+      "step": 1167
+    },
+    {
+      "epoch": 0.14503911585744442,
+      "grad_norm": 0.4500787854194641,
+      "learning_rate": 5e-05,
+      "loss": 3.1246,
+      "step": 1168
+    },
+    {
+      "epoch": 0.14516329318266485,
+      "grad_norm": 0.2316397726535797,
+      "learning_rate": 5e-05,
+      "loss": 1.8885,
+      "step": 1169
+    },
+    {
+      "epoch": 0.14528747050788526,
+      "grad_norm": 0.2591933310031891,
+      "learning_rate": 5e-05,
+      "loss": 1.833,
+      "step": 1170
+    },
+    {
+      "epoch": 0.14541164783310567,
+      "grad_norm": 0.2530015707015991,
+      "learning_rate": 5e-05,
+      "loss": 1.9319,
+      "step": 1171
+    },
+    {
+      "epoch": 0.14553582515832608,
+      "grad_norm": 0.25178420543670654,
+      "learning_rate": 5e-05,
+      "loss": 1.9035,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1456600024835465,
+      "grad_norm": 0.38378190994262695,
+      "learning_rate": 5e-05,
+      "loss": 2.866,
+      "step": 1173
+    },
+    {
+      "epoch": 0.14578417980876693,
+      "grad_norm": 0.24720364809036255,
+      "learning_rate": 5e-05,
+      "loss": 1.9238,
+      "step": 1174
+    },
+    {
+      "epoch": 0.14590835713398734,
+      "grad_norm": 0.24143683910369873,
+      "learning_rate": 5e-05,
+      "loss": 1.9584,
+      "step": 1175
+    },
+    {
+      "epoch": 0.14603253445920775,
+      "grad_norm": 0.2445477694272995,
+      "learning_rate": 5e-05,
+      "loss": 1.6613,
+      "step": 1176
+    },
+    {
+      "epoch": 0.14615671178442816,
+      "grad_norm": 0.24600107967853546,
+      "learning_rate": 5e-05,
+      "loss": 1.9561,
+      "step": 1177
+    },
+    {
+      "epoch": 0.14628088910964857,
+      "grad_norm": 0.2492237538099289,
+      "learning_rate": 5e-05,
+      "loss": 1.8048,
+      "step": 1178
+    },
+    {
+      "epoch": 0.146405066434869,
+      "grad_norm": 0.516223132610321,
+      "learning_rate": 5e-05,
+      "loss": 2.6817,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1465292437600894,
+      "grad_norm": 0.2528268098831177,
+      "learning_rate": 5e-05,
+      "loss": 2.023,
+      "step": 1180
+    },
+    {
+      "epoch": 0.14665342108530982,
+      "grad_norm": 0.2526163160800934,
+      "learning_rate": 5e-05,
+      "loss": 2.0564,
+      "step": 1181
+    },
+    {
+      "epoch": 0.14677759841053023,
+      "grad_norm": 0.2417389154434204,
+      "learning_rate": 5e-05,
+      "loss": 1.9248,
+      "step": 1182
+    },
+    {
+      "epoch": 0.14690177573575064,
+      "grad_norm": 0.47474679350852966,
+      "learning_rate": 5e-05,
+      "loss": 2.5207,
+      "step": 1183
+    },
+    {
+      "epoch": 0.14702595306097108,
+      "grad_norm": 0.24570490419864655,
+      "learning_rate": 5e-05,
+      "loss": 1.9672,
+      "step": 1184
+    },
+    {
+      "epoch": 0.14715013038619149,
+      "grad_norm": 0.27930474281311035,
+      "learning_rate": 5e-05,
+      "loss": 2.1775,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1472743077114119,
+      "grad_norm": 0.23085196316242218,
+      "learning_rate": 5e-05,
+      "loss": 1.7616,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1473984850366323,
+      "grad_norm": 0.3221113979816437,
+      "learning_rate": 5e-05,
+      "loss": 2.2469,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1475226623618527,
+      "grad_norm": 0.25005877017974854,
+      "learning_rate": 5e-05,
+      "loss": 1.86,
+      "step": 1188
+    },
+    {
+      "epoch": 0.14764683968707315,
+      "grad_norm": 0.24570053815841675,
+      "learning_rate": 5e-05,
+      "loss": 2.119,
+      "step": 1189
+    },
+    {
+      "epoch": 0.14777101701229356,
+      "grad_norm": 0.2593410909175873,
+      "learning_rate": 5e-05,
+      "loss": 2.0018,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14789519433751397,
+      "grad_norm": 0.48453813791275024,
+      "learning_rate": 5e-05,
+      "loss": 2.6275,
+      "step": 1191
+    },
+    {
+      "epoch": 0.14801937166273438,
+      "grad_norm": 0.3979604244232178,
+      "learning_rate": 5e-05,
+      "loss": 2.5155,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1481435489879548,
+      "grad_norm": 0.3578927516937256,
+      "learning_rate": 5e-05,
+      "loss": 2.4039,
+      "step": 1193
+    },
+    {
+      "epoch": 0.14826772631317522,
+      "grad_norm": 0.3434313237667084,
+      "learning_rate": 5e-05,
+      "loss": 2.6433,
+      "step": 1194
+    },
+    {
+      "epoch": 0.14839190363839563,
+      "grad_norm": 0.2263568788766861,
+      "learning_rate": 5e-05,
+      "loss": 1.6669,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14851608096361604,
+      "grad_norm": 0.3011169135570526,
+      "learning_rate": 5e-05,
+      "loss": 2.0213,
+      "step": 1196
+    },
+    {
+      "epoch": 0.14864025828883645,
+      "grad_norm": 0.255597859621048,
+      "learning_rate": 5e-05,
+      "loss": 1.8261,
+      "step": 1197
+    },
+    {
+      "epoch": 0.14876443561405686,
+      "grad_norm": 0.3074510991573334,
+      "learning_rate": 5e-05,
+      "loss": 2.3057,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1488886129392773,
+      "grad_norm": 0.305249959230423,
+      "learning_rate": 5e-05,
+      "loss": 1.8926,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1490127902644977,
+      "grad_norm": 0.3319258689880371,
+      "learning_rate": 5e-05,
+      "loss": 2.2618,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14913696758971812,
+      "grad_norm": 0.2887781262397766,
+      "learning_rate": 5e-05,
+      "loss": 2.1429,
+      "step": 1201
+    },
+    {
+      "epoch": 0.14926114491493853,
+      "grad_norm": 0.23523738980293274,
+      "learning_rate": 5e-05,
+      "loss": 1.8104,
+      "step": 1202
+    },
+    {
+      "epoch": 0.14938532224015894,
+      "grad_norm": 0.3615783154964447,
+      "learning_rate": 5e-05,
+      "loss": 2.0174,
+      "step": 1203
+    },
+    {
+      "epoch": 0.14950949956537937,
+      "grad_norm": 0.20284977555274963,
+      "learning_rate": 5e-05,
+      "loss": 1.542,
+      "step": 1204
+    },
+    {
+      "epoch": 0.14963367689059978,
+      "grad_norm": 0.23691438138484955,
+      "learning_rate": 5e-05,
+      "loss": 1.7979,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1497578542158202,
+      "grad_norm": 0.26956695318222046,
+      "learning_rate": 5e-05,
+      "loss": 2.0376,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1498820315410406,
+      "grad_norm": 0.5684540867805481,
+      "learning_rate": 5e-05,
+      "loss": 2.6706,
+      "step": 1207
+    },
+    {
+      "epoch": 0.150006208866261,
+      "grad_norm": 0.22813959419727325,
+      "learning_rate": 5e-05,
+      "loss": 1.8555,
+      "step": 1208
+    },
+    {
+      "epoch": 0.15013038619148145,
+      "grad_norm": 0.2528534531593323,
+      "learning_rate": 5e-05,
+      "loss": 1.9938,
+      "step": 1209
+    },
+    {
+      "epoch": 0.15025456351670186,
+      "grad_norm": 0.2532576024532318,
+      "learning_rate": 5e-05,
+      "loss": 1.8197,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15037874084192226,
+      "grad_norm": 0.43566054105758667,
+      "learning_rate": 5e-05,
+      "loss": 3.0556,
+      "step": 1211
+    },
+    {
+      "epoch": 0.15050291816714267,
+      "grad_norm": 0.29913580417633057,
+      "learning_rate": 5e-05,
+      "loss": 2.3617,
+      "step": 1212
+    },
+    {
+      "epoch": 0.15062709549236308,
+      "grad_norm": 0.30753758549690247,
+      "learning_rate": 5e-05,
+      "loss": 2.2932,
+      "step": 1213
+    },
+    {
+      "epoch": 0.15075127281758352,
+      "grad_norm": 0.2163032442331314,
+      "learning_rate": 5e-05,
+      "loss": 1.6103,
+      "step": 1214
+    },
+    {
+      "epoch": 0.15087545014280393,
+      "grad_norm": 0.2995399534702301,
+      "learning_rate": 5e-05,
+      "loss": 2.4001,
+      "step": 1215
+    },
+    {
+      "epoch": 0.15099962746802434,
+      "grad_norm": 0.2516692578792572,
+      "learning_rate": 5e-05,
+      "loss": 1.9123,
+      "step": 1216
+    },
+    {
+      "epoch": 0.15112380479324475,
+      "grad_norm": 0.29868268966674805,
+      "learning_rate": 5e-05,
+      "loss": 1.8208,
+      "step": 1217
+    },
+    {
+      "epoch": 0.15124798211846516,
+      "grad_norm": 0.31595051288604736,
+      "learning_rate": 5e-05,
+      "loss": 2.0664,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1513721594436856,
+      "grad_norm": 0.2653745412826538,
+      "learning_rate": 5e-05,
+      "loss": 1.7214,
+      "step": 1219
+    },
+    {
+      "epoch": 0.151496336768906,
+      "grad_norm": 0.2843726873397827,
+      "learning_rate": 5e-05,
+      "loss": 1.8578,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1516205140941264,
+      "grad_norm": 0.3057863712310791,
+      "learning_rate": 5e-05,
+      "loss": 2.5803,
+      "step": 1221
+    },
+    {
+      "epoch": 0.15174469141934682,
+      "grad_norm": 0.41839346289634705,
+      "learning_rate": 5e-05,
+      "loss": 2.6349,
+      "step": 1222
+    },
+    {
+      "epoch": 0.15186886874456723,
+      "grad_norm": 0.27506548166275024,
+      "learning_rate": 5e-05,
+      "loss": 1.9141,
+      "step": 1223
+    },
+    {
+      "epoch": 0.15199304606978767,
+      "grad_norm": 0.30796173214912415,
+      "learning_rate": 5e-05,
+      "loss": 2.0619,
+      "step": 1224
+    },
+    {
+      "epoch": 0.15211722339500808,
+      "grad_norm": 0.2962079346179962,
+      "learning_rate": 5e-05,
+      "loss": 1.6308,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1522414007202285,
+      "grad_norm": 0.29694291949272156,
+      "learning_rate": 5e-05,
+      "loss": 2.1664,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1523655780454489,
+      "grad_norm": 0.26229187846183777,
+      "learning_rate": 5e-05,
+      "loss": 1.9155,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1524897553706693,
+      "grad_norm": 0.4687798321247101,
+      "learning_rate": 5e-05,
+      "loss": 2.2301,
+      "step": 1228
+    },
+    {
+      "epoch": 0.15261393269588974,
+      "grad_norm": 0.26803359389305115,
+      "learning_rate": 5e-05,
+      "loss": 1.7888,
+      "step": 1229
+    },
+    {
+      "epoch": 0.15273811002111015,
+      "grad_norm": 0.39951369166374207,
+      "learning_rate": 5e-05,
+      "loss": 2.5015,
+      "step": 1230
+    },
+    {
+      "epoch": 0.15286228734633056,
+      "grad_norm": 0.47469648718833923,
+      "learning_rate": 5e-05,
+      "loss": 2.8467,
+      "step": 1231
+    },
+    {
+      "epoch": 0.15298646467155097,
+      "grad_norm": 0.4635626971721649,
+      "learning_rate": 5e-05,
+      "loss": 2.3634,
+      "step": 1232
+    },
+    {
+      "epoch": 0.15311064199677138,
+      "grad_norm": 0.37162482738494873,
+      "learning_rate": 5e-05,
+      "loss": 2.5226,
+      "step": 1233
+    },
+    {
+      "epoch": 0.15323481932199182,
+      "grad_norm": 0.4097093939781189,
+      "learning_rate": 5e-05,
+      "loss": 2.5152,
+      "step": 1234
+    },
+    {
+      "epoch": 0.15335899664721223,
+      "grad_norm": 0.24587558209896088,
+      "learning_rate": 5e-05,
+      "loss": 1.9299,
+      "step": 1235
+    },
+    {
+      "epoch": 0.15348317397243263,
+      "grad_norm": 0.4131929874420166,
+      "learning_rate": 5e-05,
+      "loss": 2.7207,
+      "step": 1236
+    },
+    {
+      "epoch": 0.15360735129765304,
+      "grad_norm": 0.29544156789779663,
+      "learning_rate": 5e-05,
+      "loss": 2.0303,
+      "step": 1237
+    },
+    {
+      "epoch": 0.15373152862287345,
+      "grad_norm": 0.25577372312545776,
+      "learning_rate": 5e-05,
+      "loss": 1.7898,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1538557059480939,
+      "grad_norm": 0.271610826253891,
+      "learning_rate": 5e-05,
+      "loss": 2.0402,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1539798832733143,
+      "grad_norm": 0.24255508184432983,
+      "learning_rate": 5e-05,
+      "loss": 1.9027,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1541040605985347,
+      "grad_norm": 0.25839805603027344,
+      "learning_rate": 5e-05,
+      "loss": 1.7868,
+      "step": 1241
+    },
+    {
+      "epoch": 0.15422823792375512,
+      "grad_norm": 0.2666448950767517,
+      "learning_rate": 5e-05,
+      "loss": 1.8418,
+      "step": 1242
+    },
+    {
+      "epoch": 0.15435241524897553,
+      "grad_norm": 0.7195788025856018,
+      "learning_rate": 5e-05,
+      "loss": 2.9927,
+      "step": 1243
+    },
+    {
+      "epoch": 0.15447659257419596,
+      "grad_norm": 0.6202380061149597,
+      "learning_rate": 5e-05,
+      "loss": 2.7148,
+      "step": 1244
+    },
+    {
+      "epoch": 0.15460076989941637,
+      "grad_norm": 0.47668972611427307,
+      "learning_rate": 5e-05,
+      "loss": 2.9792,
+      "step": 1245
+    },
+    {
+      "epoch": 0.15472494722463678,
+      "grad_norm": 0.28007158637046814,
+      "learning_rate": 5e-05,
+      "loss": 2.2218,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1548491245498572,
+      "grad_norm": 0.4167512357234955,
+      "learning_rate": 5e-05,
+      "loss": 2.485,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1549733018750776,
+      "grad_norm": 0.24505804479122162,
+      "learning_rate": 5e-05,
+      "loss": 1.7562,
+      "step": 1248
+    },
+    {
+      "epoch": 0.15509747920029804,
+      "grad_norm": 0.24643422663211823,
+      "learning_rate": 5e-05,
+      "loss": 1.8853,
+      "step": 1249
+    },
+    {
+      "epoch": 0.15522165652551845,
+      "grad_norm": 0.2433510273694992,
+      "learning_rate": 5e-05,
+      "loss": 1.8705,
+      "step": 1250
+    },
+    {
+      "epoch": 0.15534583385073886,
+      "grad_norm": 0.3541926443576813,
+      "learning_rate": 5e-05,
+      "loss": 2.2321,
+      "step": 1251
+    },
+    {
+      "epoch": 0.15547001117595927,
+      "grad_norm": 0.35379981994628906,
+      "learning_rate": 5e-05,
+      "loss": 2.8314,
+      "step": 1252
+    },
+    {
+      "epoch": 0.15559418850117968,
+      "grad_norm": 0.2638271749019623,
+      "learning_rate": 5e-05,
+      "loss": 1.7285,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1557183658264001,
+      "grad_norm": 0.2888670265674591,
+      "learning_rate": 5e-05,
+      "loss": 2.1675,
+      "step": 1254
+    },
+    {
+      "epoch": 0.15584254315162052,
+      "grad_norm": 0.2516119182109833,
+      "learning_rate": 5e-05,
+      "loss": 1.8925,
+      "step": 1255
+    },
+    {
+      "epoch": 0.15596672047684093,
+      "grad_norm": 0.22935748100280762,
+      "learning_rate": 5e-05,
+      "loss": 1.656,
+      "step": 1256
+    },
+    {
+      "epoch": 0.15609089780206134,
+      "grad_norm": 0.27387183904647827,
+      "learning_rate": 5e-05,
+      "loss": 2.2068,
+      "step": 1257
+    },
+    {
+      "epoch": 0.15621507512728175,
+      "grad_norm": 0.2415919452905655,
+      "learning_rate": 5e-05,
+      "loss": 1.7069,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1563392524525022,
+      "grad_norm": 0.24453413486480713,
+      "learning_rate": 5e-05,
+      "loss": 1.8001,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1564634297777226,
+      "grad_norm": 0.24439413845539093,
+      "learning_rate": 5e-05,
+      "loss": 1.7806,
+      "step": 1260
+    },
+    {
+      "epoch": 0.156587607102943,
+      "grad_norm": 0.23395656049251556,
+      "learning_rate": 5e-05,
+      "loss": 1.7242,
+      "step": 1261
+    },
+    {
+      "epoch": 0.15671178442816341,
+      "grad_norm": 0.28672462701797485,
+      "learning_rate": 5e-05,
+      "loss": 2.5583,
+      "step": 1262
+    },
+    {
+      "epoch": 0.15683596175338382,
+      "grad_norm": 0.27625617384910583,
+      "learning_rate": 5e-05,
+      "loss": 1.9203,
+      "step": 1263
+    },
+    {
+      "epoch": 0.15696013907860426,
+      "grad_norm": 0.3245868384838104,
+      "learning_rate": 5e-05,
+      "loss": 2.9431,
+      "step": 1264
+    },
+    {
+      "epoch": 0.15708431640382467,
+      "grad_norm": 0.2666500210762024,
+      "learning_rate": 5e-05,
+      "loss": 2.1769,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15720849372904508,
+      "grad_norm": 0.31889089941978455,
+      "learning_rate": 5e-05,
+      "loss": 2.7167,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1573326710542655,
+      "grad_norm": 0.3483503758907318,
+      "learning_rate": 5e-05,
+      "loss": 2.2731,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1574568483794859,
+      "grad_norm": 0.23736940324306488,
+      "learning_rate": 5e-05,
+      "loss": 1.727,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1575810257047063,
+      "grad_norm": 0.2436118721961975,
+      "learning_rate": 5e-05,
+      "loss": 1.9461,
+      "step": 1269
+    },
+    {
+      "epoch": 0.15770520302992674,
+      "grad_norm": 0.33697667717933655,
+      "learning_rate": 5e-05,
+      "loss": 2.1842,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15782938035514715,
+      "grad_norm": 0.3701704144477844,
+      "learning_rate": 5e-05,
+      "loss": 2.4626,
+      "step": 1271
+    },
+    {
+      "epoch": 0.15795355768036756,
+      "grad_norm": 0.24360108375549316,
+      "learning_rate": 5e-05,
+      "loss": 1.887,
+      "step": 1272
+    },
+    {
+      "epoch": 0.15807773500558797,
+      "grad_norm": 0.2931790351867676,
+      "learning_rate": 5e-05,
+      "loss": 2.1977,
+      "step": 1273
+    },
+    {
+      "epoch": 0.15820191233080838,
+      "grad_norm": 0.2223939448595047,
+      "learning_rate": 5e-05,
+      "loss": 1.7017,
+      "step": 1274
+    },
+    {
+      "epoch": 0.15832608965602882,
+      "grad_norm": 0.4535906910896301,
+      "learning_rate": 5e-05,
+      "loss": 3.0571,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15845026698124923,
+      "grad_norm": 0.3376586437225342,
+      "learning_rate": 5e-05,
+      "loss": 2.5278,
+      "step": 1276
+    },
+    {
+      "epoch": 0.15857444430646964,
+      "grad_norm": 0.379981130361557,
+      "learning_rate": 5e-05,
+      "loss": 2.3979,
+      "step": 1277
+    },
+    {
+      "epoch": 0.15869862163169005,
+      "grad_norm": 0.28272438049316406,
+      "learning_rate": 5e-05,
+      "loss": 1.9535,
+      "step": 1278
+    },
+    {
+      "epoch": 0.15882279895691045,
+      "grad_norm": 0.24832209944725037,
+      "learning_rate": 5e-05,
+      "loss": 1.7181,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1589469762821309,
+      "grad_norm": 0.2532510459423065,
+      "learning_rate": 5e-05,
+      "loss": 1.6482,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1590711536073513,
+      "grad_norm": 0.24660013616085052,
+      "learning_rate": 5e-05,
+      "loss": 1.7507,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1591953309325717,
+      "grad_norm": 0.2602323591709137,
+      "learning_rate": 5e-05,
+      "loss": 2.1038,
+      "step": 1282
+    },
+    {
+      "epoch": 0.15931950825779212,
+      "grad_norm": 0.30742180347442627,
+      "learning_rate": 5e-05,
+      "loss": 2.2179,
+      "step": 1283
+    },
+    {
+      "epoch": 0.15944368558301253,
+      "grad_norm": 0.26936063170433044,
+      "learning_rate": 5e-05,
+      "loss": 2.0372,
+      "step": 1284
+    },
+    {
+      "epoch": 0.15956786290823297,
+      "grad_norm": 0.2616680860519409,
+      "learning_rate": 5e-05,
+      "loss": 2.1247,
+      "step": 1285
+    },
+    {
+      "epoch": 0.15969204023345338,
+      "grad_norm": 0.42486435174942017,
+      "learning_rate": 5e-05,
+      "loss": 2.4125,
+      "step": 1286
+    },
+    {
+      "epoch": 0.15981621755867378,
+      "grad_norm": 0.26610830426216125,
+      "learning_rate": 5e-05,
+      "loss": 1.8577,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1599403948838942,
+      "grad_norm": 0.27697575092315674,
+      "learning_rate": 5e-05,
+      "loss": 2.3448,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1600645722091146,
+      "grad_norm": 0.25561267137527466,
+      "learning_rate": 5e-05,
+      "loss": 2.0115,
+      "step": 1289
+    },
+    {
+      "epoch": 0.16018874953433504,
+      "grad_norm": 0.28269773721694946,
+      "learning_rate": 5e-05,
+      "loss": 2.1156,
+      "step": 1290
+    },
+    {
+      "epoch": 0.16031292685955545,
+      "grad_norm": 0.2693529725074768,
+      "learning_rate": 5e-05,
+      "loss": 2.0133,
+      "step": 1291
+    },
+    {
+      "epoch": 0.16043710418477586,
+      "grad_norm": 0.30885741114616394,
+      "learning_rate": 5e-05,
+      "loss": 2.3881,
+      "step": 1292
+    },
+    {
+      "epoch": 0.16056128150999627,
+      "grad_norm": 0.24259135127067566,
+      "learning_rate": 5e-05,
+      "loss": 2.125,
+      "step": 1293
+    },
+    {
+      "epoch": 0.16068545883521668,
+      "grad_norm": 0.3845694661140442,
+      "learning_rate": 5e-05,
+      "loss": 2.4752,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1608096361604371,
+      "grad_norm": 0.44367334246635437,
+      "learning_rate": 5e-05,
+      "loss": 2.4092,
+      "step": 1295
+    },
+    {
+      "epoch": 0.16093381348565752,
+      "grad_norm": 0.2634572982788086,
+      "learning_rate": 5e-05,
+      "loss": 1.752,
+      "step": 1296
+    },
+    {
+      "epoch": 0.16105799081087793,
+      "grad_norm": 0.29496175050735474,
+      "learning_rate": 5e-05,
+      "loss": 2.1139,
+      "step": 1297
+    },
+    {
+      "epoch": 0.16118216813609834,
+      "grad_norm": 0.23719622194766998,
+      "learning_rate": 5e-05,
+      "loss": 1.8492,
+      "step": 1298
+    },
+    {
+      "epoch": 0.16130634546131875,
+      "grad_norm": 0.2589401602745056,
+      "learning_rate": 5e-05,
+      "loss": 2.0617,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1614305227865392,
+      "grad_norm": 0.22831667959690094,
+      "learning_rate": 5e-05,
+      "loss": 1.8122,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1615547001117596,
+      "grad_norm": 0.3076975345611572,
+      "learning_rate": 5e-05,
+      "loss": 2.1634,
+      "step": 1301
+    },
+    {
+      "epoch": 0.16167887743698,
+      "grad_norm": 0.23006132245063782,
+      "learning_rate": 5e-05,
+      "loss": 1.914,
+      "step": 1302
+    },
+    {
+      "epoch": 0.16180305476220042,
+      "grad_norm": 0.2639243006706238,
+      "learning_rate": 5e-05,
+      "loss": 2.0658,
+      "step": 1303
+    },
+    {
+      "epoch": 0.16192723208742082,
+      "grad_norm": 0.30185243487358093,
+      "learning_rate": 5e-05,
+      "loss": 2.2321,
+      "step": 1304
+    },
+    {
+      "epoch": 0.16205140941264126,
+      "grad_norm": 0.24596966803073883,
+      "learning_rate": 5e-05,
+      "loss": 1.9653,
+      "step": 1305
+    },
+    {
+      "epoch": 0.16217558673786167,
+      "grad_norm": 0.23422880470752716,
+      "learning_rate": 5e-05,
+      "loss": 1.7255,
+      "step": 1306
+    },
+    {
+      "epoch": 0.16229976406308208,
+      "grad_norm": 0.22731982171535492,
+      "learning_rate": 5e-05,
+      "loss": 1.7056,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1624239413883025,
+      "grad_norm": 0.23401689529418945,
+      "learning_rate": 5e-05,
+      "loss": 1.7543,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1625481187135229,
+      "grad_norm": 0.2320878505706787,
+      "learning_rate": 5e-05,
+      "loss": 1.851,
+      "step": 1309
+    },
+    {
+      "epoch": 0.16267229603874334,
+      "grad_norm": 0.22573159635066986,
+      "learning_rate": 5e-05,
+      "loss": 1.6737,
+      "step": 1310
+    },
+    {
+      "epoch": 0.16279647336396375,
+      "grad_norm": 0.22687570750713348,
+      "learning_rate": 5e-05,
+      "loss": 1.6772,
+      "step": 1311
+    },
+    {
+      "epoch": 0.16292065068918415,
+      "grad_norm": 0.21604809165000916,
+      "learning_rate": 5e-05,
+      "loss": 1.7014,
+      "step": 1312
+    },
+    {
+      "epoch": 0.16304482801440456,
+      "grad_norm": 0.33596667647361755,
+      "learning_rate": 5e-05,
+      "loss": 2.554,
+      "step": 1313
+    },
+    {
+      "epoch": 0.16316900533962497,
+      "grad_norm": 0.21518127620220184,
+      "learning_rate": 5e-05,
+      "loss": 1.6064,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1632931826648454,
+      "grad_norm": 0.2203385829925537,
+      "learning_rate": 5e-05,
+      "loss": 1.558,
+      "step": 1315
+    },
+    {
+      "epoch": 0.16341735999006582,
+      "grad_norm": 0.38375264406204224,
+      "learning_rate": 5e-05,
+      "loss": 2.3561,
+      "step": 1316
+    },
+    {
+      "epoch": 0.16354153731528623,
+      "grad_norm": 0.2642592489719391,
+      "learning_rate": 5e-05,
+      "loss": 2.0025,
+      "step": 1317
+    },
+    {
+      "epoch": 0.16366571464050664,
+      "grad_norm": 0.25367793440818787,
+      "learning_rate": 5e-05,
+      "loss": 1.8964,
+      "step": 1318
+    },
+    {
+      "epoch": 0.16378989196572705,
+      "grad_norm": 0.2757321000099182,
+      "learning_rate": 5e-05,
+      "loss": 2.2166,
+      "step": 1319
+    },
+    {
+      "epoch": 0.16391406929094748,
+      "grad_norm": 0.36721518635749817,
+      "learning_rate": 5e-05,
+      "loss": 1.6403,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1640382466161679,
+      "grad_norm": 0.2706136405467987,
+      "learning_rate": 5e-05,
+      "loss": 2.04,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1641624239413883,
+      "grad_norm": 0.239241361618042,
+      "learning_rate": 5e-05,
+      "loss": 2.0376,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1642866012666087,
+      "grad_norm": 0.26604607701301575,
+      "learning_rate": 5e-05,
+      "loss": 1.9828,
+      "step": 1323
+    },
+    {
+      "epoch": 0.16441077859182912,
+      "grad_norm": 0.2374410778284073,
+      "learning_rate": 5e-05,
+      "loss": 1.7973,
+      "step": 1324
+    },
+    {
+      "epoch": 0.16453495591704956,
+      "grad_norm": 0.24728651344776154,
+      "learning_rate": 5e-05,
+      "loss": 1.7262,
+      "step": 1325
+    },
+    {
+      "epoch": 0.16465913324226997,
+      "grad_norm": 0.2484746128320694,
+      "learning_rate": 5e-05,
+      "loss": 1.7974,
+      "step": 1326
+    },
+    {
+      "epoch": 0.16478331056749038,
+      "grad_norm": 0.23304536938667297,
+      "learning_rate": 5e-05,
+      "loss": 1.6588,
+      "step": 1327
+    },
+    {
+      "epoch": 0.16490748789271079,
+      "grad_norm": 0.2906743586063385,
+      "learning_rate": 5e-05,
+      "loss": 2.2776,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1650316652179312,
+      "grad_norm": 0.3603396415710449,
+      "learning_rate": 5e-05,
+      "loss": 2.0138,
+      "step": 1329
+    },
+    {
+      "epoch": 0.16515584254315163,
+      "grad_norm": 0.24917276203632355,
+      "learning_rate": 5e-05,
+      "loss": 1.923,
+      "step": 1330
+    },
+    {
+      "epoch": 0.16528001986837204,
+      "grad_norm": 0.20571613311767578,
+      "learning_rate": 5e-05,
+      "loss": 1.596,
+      "step": 1331
+    },
+    {
+      "epoch": 0.16540419719359245,
+      "grad_norm": 0.21570037305355072,
+      "learning_rate": 5e-05,
+      "loss": 1.8356,
+      "step": 1332
+    },
+    {
+      "epoch": 0.16552837451881286,
+      "grad_norm": 0.2692769467830658,
+      "learning_rate": 5e-05,
+      "loss": 2.3511,
+      "step": 1333
+    },
+    {
+      "epoch": 0.16565255184403327,
+      "grad_norm": 0.6433316469192505,
+      "learning_rate": 5e-05,
+      "loss": 3.1341,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1657767291692537,
+      "grad_norm": 0.29831385612487793,
+      "learning_rate": 5e-05,
+      "loss": 2.1875,
+      "step": 1335
+    },
+    {
+      "epoch": 0.16590090649447412,
+      "grad_norm": 0.3096681535243988,
+      "learning_rate": 5e-05,
+      "loss": 2.0789,
+      "step": 1336
+    },
+    {
+      "epoch": 0.16602508381969452,
+      "grad_norm": 0.393687903881073,
+      "learning_rate": 5e-05,
+      "loss": 2.0953,
+      "step": 1337
+    },
+    {
+      "epoch": 0.16614926114491493,
+      "grad_norm": 0.2538616359233856,
+      "learning_rate": 5e-05,
+      "loss": 1.84,
+      "step": 1338
+    },
+    {
+      "epoch": 0.16627343847013534,
+      "grad_norm": 0.24923135340213776,
+      "learning_rate": 5e-05,
+      "loss": 2.1599,
+      "step": 1339
+    },
+    {
+      "epoch": 0.16639761579535578,
+      "grad_norm": 0.2715846002101898,
+      "learning_rate": 5e-05,
+      "loss": 1.9949,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1665217931205762,
+      "grad_norm": 0.2750992774963379,
+      "learning_rate": 5e-05,
+      "loss": 1.7898,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1666459704457966,
+      "grad_norm": 0.3703128397464752,
+      "learning_rate": 5e-05,
+      "loss": 2.143,
+      "step": 1342
+    },
+    {
+      "epoch": 0.166770147771017,
+      "grad_norm": 0.24690303206443787,
+      "learning_rate": 5e-05,
+      "loss": 1.6829,
+      "step": 1343
+    },
+    {
+      "epoch": 0.16689432509623742,
+      "grad_norm": 0.23967143893241882,
+      "learning_rate": 5e-05,
+      "loss": 1.7114,
+      "step": 1344
+    },
+    {
+      "epoch": 0.16701850242145785,
+      "grad_norm": 0.23750554025173187,
+      "learning_rate": 5e-05,
+      "loss": 1.8553,
+      "step": 1345
+    },
+    {
+      "epoch": 0.16714267974667826,
+      "grad_norm": 0.29518237709999084,
+      "learning_rate": 5e-05,
+      "loss": 2.1972,
+      "step": 1346
+    },
+    {
+      "epoch": 0.16726685707189867,
+      "grad_norm": 0.3136303126811981,
+      "learning_rate": 5e-05,
+      "loss": 2.5917,
+      "step": 1347
+    },
+    {
+      "epoch": 0.16739103439711908,
+      "grad_norm": 0.2705548405647278,
+      "learning_rate": 5e-05,
+      "loss": 1.6883,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1675152117223395,
+      "grad_norm": 0.2752920985221863,
+      "learning_rate": 5e-05,
+      "loss": 2.0115,
+      "step": 1349
+    },
+    {
+      "epoch": 0.16763938904755993,
+      "grad_norm": 0.3215140700340271,
+      "learning_rate": 5e-05,
+      "loss": 2.3198,
+      "step": 1350
+    },
+    {
+      "epoch": 0.16776356637278034,
+      "grad_norm": 0.26572516560554504,
+      "learning_rate": 5e-05,
+      "loss": 2.0305,
+      "step": 1351
+    },
+    {
+      "epoch": 0.16788774369800075,
+      "grad_norm": 0.2589241862297058,
+      "learning_rate": 5e-05,
+      "loss": 1.9757,
+      "step": 1352
+    },
+    {
+      "epoch": 0.16801192102322116,
+      "grad_norm": 0.25737783312797546,
+      "learning_rate": 5e-05,
+      "loss": 1.9797,
+      "step": 1353
+    },
+    {
+      "epoch": 0.16813609834844156,
+      "grad_norm": 0.29077279567718506,
+      "learning_rate": 5e-05,
+      "loss": 2.1914,
+      "step": 1354
+    },
+    {
+      "epoch": 0.168260275673662,
+      "grad_norm": 0.30198049545288086,
+      "learning_rate": 5e-05,
+      "loss": 2.0131,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1683844529988824,
+      "grad_norm": 0.261377215385437,
+      "learning_rate": 5e-05,
+      "loss": 1.9785,
+      "step": 1356
+    },
+    {
+      "epoch": 0.16850863032410282,
+      "grad_norm": 0.23400487005710602,
+      "learning_rate": 5e-05,
+      "loss": 1.7566,
+      "step": 1357
+    },
+    {
+      "epoch": 0.16863280764932323,
+      "grad_norm": 0.24732302129268646,
+      "learning_rate": 5e-05,
+      "loss": 2.0366,
+      "step": 1358
+    },
+    {
+      "epoch": 0.16875698497454364,
+      "grad_norm": 0.26338592171669006,
+      "learning_rate": 5e-05,
+      "loss": 1.9771,
+      "step": 1359
+    },
+    {
+      "epoch": 0.16888116229976408,
+      "grad_norm": 0.263784259557724,
+      "learning_rate": 5e-05,
+      "loss": 1.9624,
+      "step": 1360
+    },
+    {
+      "epoch": 0.16900533962498449,
+      "grad_norm": 0.2597507834434509,
+      "learning_rate": 5e-05,
+      "loss": 1.7799,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1691295169502049,
+      "grad_norm": 0.34933894872665405,
+      "learning_rate": 5e-05,
+      "loss": 2.3649,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1692536942754253,
+      "grad_norm": 0.3756512999534607,
+      "learning_rate": 5e-05,
+      "loss": 2.7594,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1693778716006457,
+      "grad_norm": 0.2576308250427246,
+      "learning_rate": 5e-05,
+      "loss": 2.0799,
+      "step": 1364
+    },
+    {
+      "epoch": 0.16950204892586615,
+      "grad_norm": 0.2317439317703247,
+      "learning_rate": 5e-05,
+      "loss": 1.7362,
+      "step": 1365
+    },
+    {
+      "epoch": 0.16962622625108656,
+      "grad_norm": 0.2574308514595032,
+      "learning_rate": 5e-05,
+      "loss": 2.2173,
+      "step": 1366
+    },
+    {
+      "epoch": 0.16975040357630697,
+      "grad_norm": 0.2534666657447815,
+      "learning_rate": 5e-05,
+      "loss": 1.9303,
+      "step": 1367
+    },
+    {
+      "epoch": 0.16987458090152738,
+      "grad_norm": 0.26502034068107605,
+      "learning_rate": 5e-05,
+      "loss": 1.9592,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1699987582267478,
+      "grad_norm": 0.3469485938549042,
+      "learning_rate": 5e-05,
+      "loss": 2.3586,
+      "step": 1369
+    },
+    {
+      "epoch": 0.17012293555196822,
+      "grad_norm": 0.34820976853370667,
+      "learning_rate": 5e-05,
+      "loss": 2.7691,
+      "step": 1370
+    },
+    {
+      "epoch": 0.17024711287718863,
+      "grad_norm": 0.2686615586280823,
+      "learning_rate": 5e-05,
+      "loss": 2.3223,
+      "step": 1371
+    },
+    {
+      "epoch": 0.17037129020240904,
+      "grad_norm": 0.3581946790218353,
+      "learning_rate": 5e-05,
+      "loss": 2.6801,
+      "step": 1372
+    },
+    {
+      "epoch": 0.17049546752762945,
+      "grad_norm": 0.29164987802505493,
+      "learning_rate": 5e-05,
+      "loss": 1.9593,
+      "step": 1373
+    },
+    {
+      "epoch": 0.17061964485284986,
+      "grad_norm": 0.32622209191322327,
+      "learning_rate": 5e-05,
+      "loss": 2.2314,
+      "step": 1374
+    },
+    {
+      "epoch": 0.1707438221780703,
+      "grad_norm": 0.2485162615776062,
+      "learning_rate": 5e-05,
+      "loss": 1.9148,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1708679995032907,
+      "grad_norm": 0.25347989797592163,
+      "learning_rate": 5e-05,
+      "loss": 2.2589,
+      "step": 1376
+    },
+    {
+      "epoch": 0.17099217682851112,
+      "grad_norm": 0.26037269830703735,
+      "learning_rate": 5e-05,
+      "loss": 1.7405,
+      "step": 1377
+    },
+    {
+      "epoch": 0.17111635415373153,
+      "grad_norm": 0.24087978899478912,
+      "learning_rate": 5e-05,
+      "loss": 1.5645,
+      "step": 1378
+    },
+    {
+      "epoch": 0.17124053147895194,
+      "grad_norm": 0.3239763677120209,
+      "learning_rate": 5e-05,
+      "loss": 2.1053,
+      "step": 1379
+    },
+    {
+      "epoch": 0.17136470880417234,
+      "grad_norm": 0.3365716338157654,
+      "learning_rate": 5e-05,
+      "loss": 2.9064,
+      "step": 1380
+    },
+    {
+      "epoch": 0.17148888612939278,
+      "grad_norm": 0.24474604427814484,
+      "learning_rate": 5e-05,
+      "loss": 1.858,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1716130634546132,
+      "grad_norm": 0.2951070964336395,
+      "learning_rate": 5e-05,
+      "loss": 2.6255,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1717372407798336,
+      "grad_norm": 0.27612584829330444,
+      "learning_rate": 5e-05,
+      "loss": 2.1143,
+      "step": 1383
+    },
+    {
+      "epoch": 0.171861418105054,
+      "grad_norm": 0.32302820682525635,
+      "learning_rate": 5e-05,
+      "loss": 2.2926,
+      "step": 1384
+    },
+    {
+      "epoch": 0.17198559543027442,
+      "grad_norm": 0.34798377752304077,
+      "learning_rate": 5e-05,
+      "loss": 2.8825,
+      "step": 1385
+    },
+    {
+      "epoch": 0.17210977275549486,
+      "grad_norm": 0.23948217928409576,
+      "learning_rate": 5e-05,
+      "loss": 1.7242,
+      "step": 1386
+    },
+    {
+      "epoch": 0.17223395008071526,
+      "grad_norm": 0.2691064476966858,
+      "learning_rate": 5e-05,
+      "loss": 1.9609,
+      "step": 1387
+    },
+    {
+      "epoch": 0.17235812740593567,
+      "grad_norm": 0.5158481001853943,
+      "learning_rate": 5e-05,
+      "loss": 4.332,
+      "step": 1388
+    },
+    {
+      "epoch": 0.17248230473115608,
+      "grad_norm": 0.2692219018936157,
+      "learning_rate": 5e-05,
+      "loss": 2.1929,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1726064820563765,
+      "grad_norm": 0.26824715733528137,
+      "learning_rate": 5e-05,
+      "loss": 2.1272,
+      "step": 1390
+    },
+    {
+      "epoch": 0.17273065938159693,
+      "grad_norm": 0.23186148703098297,
+      "learning_rate": 5e-05,
+      "loss": 1.7898,
+      "step": 1391
+    },
+    {
+      "epoch": 0.17285483670681734,
+      "grad_norm": 0.24311073124408722,
+      "learning_rate": 5e-05,
+      "loss": 1.8924,
+      "step": 1392
+    },
+    {
+      "epoch": 0.17297901403203775,
+      "grad_norm": 0.37856578826904297,
+      "learning_rate": 5e-05,
+      "loss": 2.707,
+      "step": 1393
+    },
+    {
+      "epoch": 0.17310319135725816,
+      "grad_norm": 0.23472905158996582,
+      "learning_rate": 5e-05,
+      "loss": 1.81,
+      "step": 1394
+    },
+    {
+      "epoch": 0.17322736868247857,
+      "grad_norm": 0.35837453603744507,
+      "learning_rate": 5e-05,
+      "loss": 2.4733,
+      "step": 1395
+    },
+    {
+      "epoch": 0.173351546007699,
+      "grad_norm": 0.2707933783531189,
+      "learning_rate": 5e-05,
+      "loss": 2.0588,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1734757233329194,
+      "grad_norm": 0.24753183126449585,
+      "learning_rate": 5e-05,
+      "loss": 1.9286,
+      "step": 1397
+    },
+    {
+      "epoch": 0.17359990065813982,
+      "grad_norm": 0.2418799102306366,
+      "learning_rate": 5e-05,
+      "loss": 1.7855,
+      "step": 1398
+    },
+    {
+      "epoch": 0.17372407798336023,
+      "grad_norm": 0.22159628570079803,
+      "learning_rate": 5e-05,
+      "loss": 1.6137,
+      "step": 1399
+    },
+    {
+      "epoch": 0.17384825530858064,
+      "grad_norm": 0.2428511530160904,
+      "learning_rate": 5e-05,
+      "loss": 1.8336,
+      "step": 1400
+    },
+    {
+      "epoch": 0.17397243263380108,
+      "grad_norm": 0.26165932416915894,
+      "learning_rate": 5e-05,
+      "loss": 2.0209,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1740966099590215,
+      "grad_norm": 0.4102734625339508,
+      "learning_rate": 5e-05,
+      "loss": 2.7117,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1742207872842419,
+      "grad_norm": 0.2906935513019562,
+      "learning_rate": 5e-05,
+      "loss": 2.3419,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1743449646094623,
+      "grad_norm": 0.26342058181762695,
+      "learning_rate": 5e-05,
+      "loss": 1.9405,
+      "step": 1404
+    },
+    {
+      "epoch": 0.17446914193468271,
+      "grad_norm": 0.26844656467437744,
+      "learning_rate": 5e-05,
+      "loss": 2.341,
+      "step": 1405
+    },
+    {
+      "epoch": 0.17459331925990315,
+      "grad_norm": 0.3802238702774048,
+      "learning_rate": 5e-05,
+      "loss": 2.3929,
+      "step": 1406
+    },
+    {
+      "epoch": 0.17471749658512356,
+      "grad_norm": 0.2770462930202484,
+      "learning_rate": 5e-05,
+      "loss": 2.1808,
+      "step": 1407
+    },
+    {
+      "epoch": 0.17484167391034397,
+      "grad_norm": 0.2371964305639267,
+      "learning_rate": 5e-05,
+      "loss": 1.6153,
+      "step": 1408
+    },
+    {
+      "epoch": 0.17496585123556438,
+      "grad_norm": 0.24255287647247314,
+      "learning_rate": 5e-05,
+      "loss": 1.7784,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1750900285607848,
+      "grad_norm": 0.29572808742523193,
+      "learning_rate": 5e-05,
+      "loss": 2.0706,
+      "step": 1410
+    },
+    {
+      "epoch": 0.17521420588600523,
+      "grad_norm": 0.2493947595357895,
+      "learning_rate": 5e-05,
+      "loss": 1.7381,
+      "step": 1411
+    },
+    {
+      "epoch": 0.17533838321122563,
+      "grad_norm": 0.26389703154563904,
+      "learning_rate": 5e-05,
+      "loss": 2.1818,
+      "step": 1412
+    },
+    {
+      "epoch": 0.17546256053644604,
+      "grad_norm": 0.3114052712917328,
+      "learning_rate": 5e-05,
+      "loss": 2.2937,
+      "step": 1413
+    },
+    {
+      "epoch": 0.17558673786166645,
+      "grad_norm": 0.265316367149353,
+      "learning_rate": 5e-05,
+      "loss": 2.0094,
+      "step": 1414
+    },
+    {
+      "epoch": 0.17571091518688686,
+      "grad_norm": 0.23204118013381958,
+      "learning_rate": 5e-05,
+      "loss": 1.7295,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1758350925121073,
+      "grad_norm": 0.29755398631095886,
+      "learning_rate": 5e-05,
+      "loss": 2.0758,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1759592698373277,
+      "grad_norm": 0.2418096363544464,
+      "learning_rate": 5e-05,
+      "loss": 1.7773,
+      "step": 1417
+    },
+    {
+      "epoch": 0.17608344716254812,
+      "grad_norm": 0.23157909512519836,
+      "learning_rate": 5e-05,
+      "loss": 1.7074,
+      "step": 1418
+    },
+    {
+      "epoch": 0.17620762448776853,
+      "grad_norm": 0.2520146071910858,
+      "learning_rate": 5e-05,
+      "loss": 2.0032,
+      "step": 1419
+    },
+    {
+      "epoch": 0.17633180181298894,
+      "grad_norm": 0.2604805827140808,
+      "learning_rate": 5e-05,
+      "loss": 2.1325,
+      "step": 1420
+    },
+    {
+      "epoch": 0.17645597913820937,
+      "grad_norm": 0.2591821551322937,
+      "learning_rate": 5e-05,
+      "loss": 2.0302,
+      "step": 1421
+    },
+    {
+      "epoch": 0.17658015646342978,
+      "grad_norm": 0.4048409163951874,
+      "learning_rate": 5e-05,
+      "loss": 2.5586,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1767043337886502,
+      "grad_norm": 0.2755661606788635,
+      "learning_rate": 5e-05,
+      "loss": 1.9671,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1768285111138706,
+      "grad_norm": 0.4293416142463684,
+      "learning_rate": 5e-05,
+      "loss": 2.6976,
+      "step": 1424
+    },
+    {
+      "epoch": 0.176952688439091,
+      "grad_norm": 0.4016331136226654,
+      "learning_rate": 5e-05,
+      "loss": 2.3576,
+      "step": 1425
+    },
+    {
+      "epoch": 0.17707686576431145,
+      "grad_norm": 0.3287256062030792,
+      "learning_rate": 5e-05,
+      "loss": 2.2676,
+      "step": 1426
+    },
+    {
+      "epoch": 0.17720104308953186,
+      "grad_norm": 0.4586311876773834,
+      "learning_rate": 5e-05,
+      "loss": 2.9434,
+      "step": 1427
+    },
+    {
+      "epoch": 0.17732522041475227,
+      "grad_norm": 0.2618440091609955,
+      "learning_rate": 5e-05,
+      "loss": 1.9483,
+      "step": 1428
+    },
+    {
+      "epoch": 0.17744939773997268,
+      "grad_norm": 0.2931461036205292,
+      "learning_rate": 5e-05,
+      "loss": 1.9923,
+      "step": 1429
+    },
+    {
+      "epoch": 0.17757357506519308,
+      "grad_norm": 0.26656579971313477,
+      "learning_rate": 5e-05,
+      "loss": 1.9064,
+      "step": 1430
+    },
+    {
+      "epoch": 0.17769775239041352,
+      "grad_norm": 0.2721560597419739,
+      "learning_rate": 5e-05,
+      "loss": 1.8299,
+      "step": 1431
+    },
+    {
+      "epoch": 0.17782192971563393,
+      "grad_norm": 0.2622223496437073,
+      "learning_rate": 5e-05,
+      "loss": 1.9364,
+      "step": 1432
+    },
+    {
+      "epoch": 0.17794610704085434,
+      "grad_norm": 0.251390278339386,
+      "learning_rate": 5e-05,
+      "loss": 1.7021,
+      "step": 1433
+    },
+    {
+      "epoch": 0.17807028436607475,
+      "grad_norm": 0.2672654986381531,
+      "learning_rate": 5e-05,
+      "loss": 1.9786,
+      "step": 1434
+    },
+    {
+      "epoch": 0.17819446169129516,
+      "grad_norm": 0.27238285541534424,
+      "learning_rate": 5e-05,
+      "loss": 2.0547,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1783186390165156,
+      "grad_norm": 0.27467331290245056,
+      "learning_rate": 5e-05,
+      "loss": 2.3656,
+      "step": 1436
+    },
+    {
+      "epoch": 0.178442816341736,
+      "grad_norm": 0.23343338072299957,
+      "learning_rate": 5e-05,
+      "loss": 1.6826,
+      "step": 1437
+    },
+    {
+      "epoch": 0.17856699366695641,
+      "grad_norm": 0.36576956510543823,
+      "learning_rate": 5e-05,
+      "loss": 1.8527,
+      "step": 1438
+    },
+    {
+      "epoch": 0.17869117099217682,
+      "grad_norm": 0.31418344378471375,
+      "learning_rate": 5e-05,
+      "loss": 2.2803,
+      "step": 1439
+    },
+    {
+      "epoch": 0.17881534831739723,
+      "grad_norm": 0.33752280473709106,
+      "learning_rate": 5e-05,
+      "loss": 2.0256,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17893952564261767,
+      "grad_norm": 0.3247952461242676,
+      "learning_rate": 5e-05,
+      "loss": 2.3914,
+      "step": 1441
+    },
+    {
+      "epoch": 0.17906370296783808,
+      "grad_norm": 0.22806139290332794,
+      "learning_rate": 5e-05,
+      "loss": 1.6938,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1791878802930585,
+      "grad_norm": 0.22137673199176788,
+      "learning_rate": 5e-05,
+      "loss": 1.8286,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1793120576182789,
+      "grad_norm": 0.25577589869499207,
+      "learning_rate": 5e-05,
+      "loss": 1.8923,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1794362349434993,
+      "grad_norm": 0.27369439601898193,
+      "learning_rate": 5e-05,
+      "loss": 2.1492,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17956041226871974,
+      "grad_norm": 0.49142706394195557,
+      "learning_rate": 5e-05,
+      "loss": 3.1958,
+      "step": 1446
+    },
+    {
+      "epoch": 0.17968458959394015,
+      "grad_norm": 0.24683183431625366,
+      "learning_rate": 5e-05,
+      "loss": 1.888,
+      "step": 1447
+    },
+    {
+      "epoch": 0.17980876691916056,
+      "grad_norm": 0.35656219720840454,
+      "learning_rate": 5e-05,
+      "loss": 2.5539,
+      "step": 1448
+    },
+    {
+      "epoch": 0.17993294424438097,
+      "grad_norm": 0.40404483675956726,
+      "learning_rate": 5e-05,
+      "loss": 2.876,
+      "step": 1449
+    },
+    {
+      "epoch": 0.18005712156960138,
+      "grad_norm": 0.30323413014411926,
+      "learning_rate": 5e-05,
+      "loss": 1.9658,
+      "step": 1450
+    },
+    {
+      "epoch": 0.18018129889482182,
+      "grad_norm": 0.2609386742115021,
+      "learning_rate": 5e-05,
+      "loss": 1.8952,
+      "step": 1451
+    },
+    {
+      "epoch": 0.18030547622004223,
+      "grad_norm": 0.26660341024398804,
+      "learning_rate": 5e-05,
+      "loss": 2.2509,
+      "step": 1452
+    },
+    {
+      "epoch": 0.18042965354526264,
+      "grad_norm": 0.2815009653568268,
+      "learning_rate": 5e-05,
+      "loss": 1.8909,
+      "step": 1453
+    },
+    {
+      "epoch": 0.18055383087048305,
+      "grad_norm": 0.27161529660224915,
+      "learning_rate": 5e-05,
+      "loss": 1.9646,
+      "step": 1454
+    },
+    {
+      "epoch": 0.18067800819570345,
+      "grad_norm": 0.3715043067932129,
+      "learning_rate": 5e-05,
+      "loss": 2.7612,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1808021855209239,
+      "grad_norm": 0.23395122587680817,
+      "learning_rate": 5e-05,
+      "loss": 1.6155,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1809263628461443,
+      "grad_norm": 0.2671591341495514,
+      "learning_rate": 5e-05,
+      "loss": 2.0628,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1810505401713647,
+      "grad_norm": 0.2846117317676544,
+      "learning_rate": 5e-05,
+      "loss": 2.5625,
+      "step": 1458
+    },
+    {
+      "epoch": 0.18117471749658512,
+      "grad_norm": 0.3615747094154358,
+      "learning_rate": 5e-05,
+      "loss": 2.6498,
+      "step": 1459
+    },
+    {
+      "epoch": 0.18129889482180553,
+      "grad_norm": 0.26534903049468994,
+      "learning_rate": 5e-05,
+      "loss": 1.7613,
+      "step": 1460
+    },
+    {
+      "epoch": 0.18142307214702597,
+      "grad_norm": 0.35508978366851807,
+      "learning_rate": 5e-05,
+      "loss": 2.5538,
+      "step": 1461
+    },
+    {
+      "epoch": 0.18154724947224637,
+      "grad_norm": 0.31567883491516113,
+      "learning_rate": 5e-05,
+      "loss": 2.0223,
+      "step": 1462
+    },
+    {
+      "epoch": 0.18167142679746678,
+      "grad_norm": 0.24844303727149963,
+      "learning_rate": 5e-05,
+      "loss": 2.1412,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1817956041226872,
+      "grad_norm": 0.23571166396141052,
+      "learning_rate": 5e-05,
+      "loss": 1.769,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1819197814479076,
+      "grad_norm": 0.2513312101364136,
+      "learning_rate": 5e-05,
+      "loss": 2.057,
+      "step": 1465
+    },
+    {
+      "epoch": 0.18204395877312804,
+      "grad_norm": 0.25203919410705566,
+      "learning_rate": 5e-05,
+      "loss": 1.8129,
+      "step": 1466
+    },
+    {
+      "epoch": 0.18216813609834845,
+      "grad_norm": 0.5145329236984253,
+      "learning_rate": 5e-05,
+      "loss": 2.8447,
+      "step": 1467
+    },
+    {
+      "epoch": 0.18229231342356886,
+      "grad_norm": 0.2694832384586334,
+      "learning_rate": 5e-05,
+      "loss": 1.8948,
+      "step": 1468
+    },
+    {
+      "epoch": 0.18241649074878927,
+      "grad_norm": 0.24046273529529572,
+      "learning_rate": 5e-05,
+      "loss": 1.9026,
+      "step": 1469
+    },
+    {
+      "epoch": 0.18254066807400968,
+      "grad_norm": 0.30585774779319763,
+      "learning_rate": 5e-05,
+      "loss": 2.4001,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1826648453992301,
+      "grad_norm": 0.25027376413345337,
+      "learning_rate": 5e-05,
+      "loss": 1.7056,
+      "step": 1471
+    },
+    {
+      "epoch": 0.18278902272445052,
+      "grad_norm": 0.8577630519866943,
+      "learning_rate": 5e-05,
+      "loss": 3.3302,
+      "step": 1472
+    },
+    {
+      "epoch": 0.18291320004967093,
+      "grad_norm": 0.24672439694404602,
+      "learning_rate": 5e-05,
+      "loss": 1.7994,
+      "step": 1473
+    },
+    {
+      "epoch": 0.18303737737489134,
+      "grad_norm": 0.25796759128570557,
+      "learning_rate": 5e-05,
+      "loss": 1.9372,
+      "step": 1474
+    },
+    {
+      "epoch": 0.18316155470011175,
+      "grad_norm": 0.44984665513038635,
+      "learning_rate": 5e-05,
+      "loss": 2.3565,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1832857320253322,
+      "grad_norm": 0.3122994303703308,
+      "learning_rate": 5e-05,
+      "loss": 2.153,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1834099093505526,
+      "grad_norm": 0.31633636355400085,
+      "learning_rate": 5e-05,
+      "loss": 1.9941,
+      "step": 1477
+    },
+    {
+      "epoch": 0.183534086675773,
+      "grad_norm": 0.2333393096923828,
+      "learning_rate": 5e-05,
+      "loss": 1.6817,
+      "step": 1478
+    },
+    {
+      "epoch": 0.18365826400099342,
+      "grad_norm": 0.32502904534339905,
+      "learning_rate": 5e-05,
+      "loss": 1.9969,
+      "step": 1479
+    },
+    {
+      "epoch": 0.18378244132621382,
+      "grad_norm": 0.2573794722557068,
+      "learning_rate": 5e-05,
+      "loss": 1.9469,
+      "step": 1480
+    },
+    {
+      "epoch": 0.18390661865143426,
+      "grad_norm": 0.2532219886779785,
+      "learning_rate": 5e-05,
+      "loss": 1.767,
+      "step": 1481
+    },
+    {
+      "epoch": 0.18403079597665467,
+      "grad_norm": 0.2704537510871887,
+      "learning_rate": 5e-05,
+      "loss": 2.0838,
+      "step": 1482
+    },
+    {
+      "epoch": 0.18415497330187508,
+      "grad_norm": 0.2416088879108429,
+      "learning_rate": 5e-05,
+      "loss": 1.7621,
+      "step": 1483
+    },
+    {
+      "epoch": 0.1842791506270955,
+      "grad_norm": 0.2926209270954132,
+      "learning_rate": 5e-05,
+      "loss": 2.0904,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1844033279523159,
+      "grad_norm": 0.2582109570503235,
+      "learning_rate": 5e-05,
+      "loss": 2.1826,
+      "step": 1485
+    },
+    {
+      "epoch": 0.18452750527753634,
+      "grad_norm": 0.4070335626602173,
+      "learning_rate": 5e-05,
+      "loss": 2.7879,
+      "step": 1486
+    },
+    {
+      "epoch": 0.18465168260275674,
+      "grad_norm": 0.23861896991729736,
+      "learning_rate": 5e-05,
+      "loss": 1.8033,
+      "step": 1487
+    },
+    {
+      "epoch": 0.18477585992797715,
+      "grad_norm": 0.24718508124351501,
+      "learning_rate": 5e-05,
+      "loss": 1.9015,
+      "step": 1488
+    },
+    {
+      "epoch": 0.18490003725319756,
+      "grad_norm": 0.2664046585559845,
+      "learning_rate": 5e-05,
+      "loss": 1.9978,
+      "step": 1489
+    },
+    {
+      "epoch": 0.18502421457841797,
+      "grad_norm": 0.31886690855026245,
+      "learning_rate": 5e-05,
+      "loss": 2.0193,
+      "step": 1490
+    },
+    {
+      "epoch": 0.18514839190363838,
+      "grad_norm": 0.2447976917028427,
+      "learning_rate": 5e-05,
+      "loss": 1.883,
+      "step": 1491
+    },
+    {
+      "epoch": 0.18527256922885882,
+      "grad_norm": 0.23854738473892212,
+      "learning_rate": 5e-05,
+      "loss": 2.0306,
+      "step": 1492
+    },
+    {
+      "epoch": 0.18539674655407923,
+      "grad_norm": 0.25025415420532227,
+      "learning_rate": 5e-05,
+      "loss": 1.9011,
+      "step": 1493
+    },
+    {
+      "epoch": 0.18552092387929964,
+      "grad_norm": 0.22724208235740662,
+      "learning_rate": 5e-05,
+      "loss": 1.735,
+      "step": 1494
+    },
+    {
+      "epoch": 0.18564510120452005,
+      "grad_norm": 0.5102798938751221,
+      "learning_rate": 5e-05,
+      "loss": 2.8414,
+      "step": 1495
+    },
+    {
+      "epoch": 0.18576927852974046,
+      "grad_norm": 0.7088466286659241,
+      "learning_rate": 5e-05,
+      "loss": 3.0139,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1858934558549609,
+      "grad_norm": 0.3303782641887665,
+      "learning_rate": 5e-05,
+      "loss": 2.1261,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1860176331801813,
+      "grad_norm": 0.39420267939567566,
+      "learning_rate": 5e-05,
+      "loss": 2.7424,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1861418105054017,
+      "grad_norm": 0.3486965298652649,
+      "learning_rate": 5e-05,
+      "loss": 2.6139,
+      "step": 1499
+    },
+    {
+      "epoch": 0.18626598783062212,
+      "grad_norm": 0.35490652918815613,
+      "learning_rate": 5e-05,
+      "loss": 2.8292,
+      "step": 1500
+    },
+    {
+      "epoch": 0.18639016515584253,
+      "grad_norm": 0.25819462537765503,
+      "learning_rate": 5e-05,
+      "loss": 1.8759,
+      "step": 1501
+    },
+    {
+      "epoch": 0.18651434248106297,
+      "grad_norm": 0.2749720513820648,
+      "learning_rate": 5e-05,
+      "loss": 1.9739,
+      "step": 1502
+    },
+    {
+      "epoch": 0.18663851980628338,
+      "grad_norm": 0.26363909244537354,
+      "learning_rate": 5e-05,
+      "loss": 1.9509,
+      "step": 1503
+    },
+    {
+      "epoch": 0.18676269713150379,
+      "grad_norm": 0.2797585129737854,
+      "learning_rate": 5e-05,
+      "loss": 1.6516,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1868868744567242,
+      "grad_norm": 0.2494063377380371,
+      "learning_rate": 5e-05,
+      "loss": 1.7693,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1870110517819446,
+      "grad_norm": 0.2863432466983795,
+      "learning_rate": 5e-05,
+      "loss": 1.9949,
+      "step": 1506
+    },
+    {
+      "epoch": 0.18713522910716504,
+      "grad_norm": 0.3042944371700287,
+      "learning_rate": 5e-05,
+      "loss": 2.3896,
+      "step": 1507
+    },
+    {
+      "epoch": 0.18725940643238545,
+      "grad_norm": 0.273322731256485,
+      "learning_rate": 5e-05,
+      "loss": 1.8971,
+      "step": 1508
+    },
+    {
+      "epoch": 0.18738358375760586,
+      "grad_norm": 0.5020868182182312,
+      "learning_rate": 5e-05,
+      "loss": 3.0533,
+      "step": 1509
+    },
+    {
+      "epoch": 0.18750776108282627,
+      "grad_norm": 0.31023505330085754,
+      "learning_rate": 5e-05,
+      "loss": 2.0806,
+      "step": 1510
+    },
+    {
+      "epoch": 0.18763193840804668,
+      "grad_norm": 0.25057369470596313,
+      "learning_rate": 5e-05,
+      "loss": 1.739,
+      "step": 1511
+    },
+    {
+      "epoch": 0.18775611573326711,
+      "grad_norm": 0.27967381477355957,
+      "learning_rate": 5e-05,
+      "loss": 2.0262,
+      "step": 1512
+    },
+    {
+      "epoch": 0.18788029305848752,
+      "grad_norm": 0.30165356397628784,
+      "learning_rate": 5e-05,
+      "loss": 2.2071,
+      "step": 1513
+    },
+    {
+      "epoch": 0.18800447038370793,
+      "grad_norm": 0.3059106469154358,
+      "learning_rate": 5e-05,
+      "loss": 2.167,
+      "step": 1514
+    },
+    {
+      "epoch": 0.18812864770892834,
+      "grad_norm": 0.31898683309555054,
+      "learning_rate": 5e-05,
+      "loss": 1.8459,
+      "step": 1515
+    },
+    {
+      "epoch": 0.18825282503414875,
+      "grad_norm": 0.28176411986351013,
+      "learning_rate": 5e-05,
+      "loss": 2.2245,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1883770023593692,
+      "grad_norm": 0.2570262849330902,
+      "learning_rate": 5e-05,
+      "loss": 1.8478,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1885011796845896,
+      "grad_norm": 0.24826116859912872,
+      "learning_rate": 5e-05,
+      "loss": 2.0992,
+      "step": 1518
+    },
+    {
+      "epoch": 0.18862535700981,
+      "grad_norm": 0.2679319381713867,
+      "learning_rate": 5e-05,
+      "loss": 2.107,
+      "step": 1519
+    },
+    {
+      "epoch": 0.18874953433503042,
+      "grad_norm": 0.4362642765045166,
+      "learning_rate": 5e-05,
+      "loss": 2.9655,
+      "step": 1520
+    },
+    {
+      "epoch": 0.18887371166025083,
+      "grad_norm": 0.25429272651672363,
+      "learning_rate": 5e-05,
+      "loss": 1.8646,
+      "step": 1521
+    },
+    {
+      "epoch": 0.18899788898547126,
+      "grad_norm": 0.24654313921928406,
+      "learning_rate": 5e-05,
+      "loss": 1.9151,
+      "step": 1522
+    },
+    {
+      "epoch": 0.18912206631069167,
+      "grad_norm": 0.2695249617099762,
+      "learning_rate": 5e-05,
+      "loss": 1.9894,
+      "step": 1523
+    },
+    {
+      "epoch": 0.18924624363591208,
+      "grad_norm": 0.24986152350902557,
+      "learning_rate": 5e-05,
+      "loss": 2.0994,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1893704209611325,
+      "grad_norm": 0.3021607995033264,
+      "learning_rate": 5e-05,
+      "loss": 2.1058,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1894945982863529,
+      "grad_norm": 0.25152772665023804,
+      "learning_rate": 5e-05,
+      "loss": 1.9933,
+      "step": 1526
+    },
+    {
+      "epoch": 0.18961877561157334,
+      "grad_norm": 0.22843512892723083,
+      "learning_rate": 5e-05,
+      "loss": 1.8159,
+      "step": 1527
+    },
+    {
+      "epoch": 0.18974295293679375,
+      "grad_norm": 0.6565892696380615,
+      "learning_rate": 5e-05,
+      "loss": 2.8925,
+      "step": 1528
+    },
+    {
+      "epoch": 0.18986713026201416,
+      "grad_norm": 0.3402864336967468,
+      "learning_rate": 5e-05,
+      "loss": 2.5189,
+      "step": 1529
+    },
+    {
+      "epoch": 0.18999130758723456,
+      "grad_norm": 0.25821635127067566,
+      "learning_rate": 5e-05,
+      "loss": 2.1182,
+      "step": 1530
+    },
+    {
+      "epoch": 0.19011548491245497,
+      "grad_norm": 0.37273097038269043,
+      "learning_rate": 5e-05,
+      "loss": 2.5305,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1902396622376754,
+      "grad_norm": 0.25413644313812256,
+      "learning_rate": 5e-05,
+      "loss": 1.8807,
+      "step": 1532
+    },
+    {
+      "epoch": 0.19036383956289582,
+      "grad_norm": 0.36151304841041565,
+      "learning_rate": 5e-05,
+      "loss": 2.6395,
+      "step": 1533
+    },
+    {
+      "epoch": 0.19048801688811623,
+      "grad_norm": 0.23387369513511658,
+      "learning_rate": 5e-05,
+      "loss": 1.7179,
+      "step": 1534
+    },
+    {
+      "epoch": 0.19061219421333664,
+      "grad_norm": 0.2963470220565796,
+      "learning_rate": 5e-05,
+      "loss": 2.3148,
+      "step": 1535
+    },
+    {
+      "epoch": 0.19073637153855705,
+      "grad_norm": 0.2130885273218155,
+      "learning_rate": 5e-05,
+      "loss": 1.6016,
+      "step": 1536
+    },
+    {
+      "epoch": 0.19086054886377748,
+      "grad_norm": 0.30370479822158813,
+      "learning_rate": 5e-05,
+      "loss": 2.3556,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1909847261889979,
+      "grad_norm": 0.4362613558769226,
+      "learning_rate": 5e-05,
+      "loss": 2.7549,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1911089035142183,
+      "grad_norm": 0.26847660541534424,
+      "learning_rate": 5e-05,
+      "loss": 2.0242,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1912330808394387,
+      "grad_norm": 0.2961687445640564,
+      "learning_rate": 5e-05,
+      "loss": 2.2042,
+      "step": 1540
+    },
+    {
+      "epoch": 0.19135725816465912,
+      "grad_norm": 0.5650090575218201,
+      "learning_rate": 5e-05,
+      "loss": 2.1309,
+      "step": 1541
+    },
+    {
+      "epoch": 0.19148143548987956,
+      "grad_norm": 0.2385435253381729,
+      "learning_rate": 5e-05,
+      "loss": 1.806,
+      "step": 1542
+    },
+    {
+      "epoch": 0.19160561281509997,
+      "grad_norm": 0.312453955411911,
+      "learning_rate": 5e-05,
+      "loss": 2.1856,
+      "step": 1543
+    },
+    {
+      "epoch": 0.19172979014032038,
+      "grad_norm": 0.24249126017093658,
+      "learning_rate": 5e-05,
+      "loss": 1.9009,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1918539674655408,
+      "grad_norm": 0.2648986876010895,
+      "learning_rate": 5e-05,
+      "loss": 2.1276,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1919781447907612,
+      "grad_norm": 0.28648820519447327,
+      "learning_rate": 5e-05,
+      "loss": 2.0801,
+      "step": 1546
+    },
+    {
+      "epoch": 0.19210232211598163,
+      "grad_norm": 0.263574481010437,
+      "learning_rate": 5e-05,
+      "loss": 1.9834,
+      "step": 1547
+    },
+    {
+      "epoch": 0.19222649944120204,
+      "grad_norm": 0.24371059238910675,
+      "learning_rate": 5e-05,
+      "loss": 1.9919,
+      "step": 1548
+    },
+    {
+      "epoch": 0.19235067676642245,
+      "grad_norm": 0.5045280456542969,
+      "learning_rate": 5e-05,
+      "loss": 2.663,
+      "step": 1549
+    },
+    {
+      "epoch": 0.19247485409164286,
+      "grad_norm": 0.25062283873558044,
+      "learning_rate": 5e-05,
+      "loss": 1.8231,
+      "step": 1550
+    },
+    {
+      "epoch": 0.19259903141686327,
+      "grad_norm": 0.3823087513446808,
+      "learning_rate": 5e-05,
+      "loss": 2.8503,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1927232087420837,
+      "grad_norm": 0.27880576252937317,
+      "learning_rate": 5e-05,
+      "loss": 1.6444,
+      "step": 1552
+    },
+    {
+      "epoch": 0.19284738606730412,
+      "grad_norm": 0.2568613290786743,
+      "learning_rate": 5e-05,
+      "loss": 1.8033,
+      "step": 1553
+    },
+    {
+      "epoch": 0.19297156339252453,
+      "grad_norm": 0.3594647943973541,
+      "learning_rate": 5e-05,
+      "loss": 2.2217,
+      "step": 1554
+    },
+    {
+      "epoch": 0.19309574071774493,
+      "grad_norm": 0.31526440382003784,
+      "learning_rate": 5e-05,
+      "loss": 1.865,
+      "step": 1555
+    },
+    {
+      "epoch": 0.19321991804296534,
+      "grad_norm": 0.2625117599964142,
+      "learning_rate": 5e-05,
+      "loss": 2.0194,
+      "step": 1556
+    },
+    {
+      "epoch": 0.19334409536818578,
+      "grad_norm": 0.30689170956611633,
+      "learning_rate": 5e-05,
+      "loss": 2.3,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1934682726934062,
+      "grad_norm": 0.3984525203704834,
+      "learning_rate": 5e-05,
+      "loss": 3.0515,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1935924500186266,
+      "grad_norm": 0.2914687395095825,
+      "learning_rate": 5e-05,
+      "loss": 2.1995,
+      "step": 1559
+    },
+    {
+      "epoch": 0.193716627343847,
+      "grad_norm": 0.29401513934135437,
+      "learning_rate": 5e-05,
+      "loss": 2.3355,
+      "step": 1560
+    },
+    {
+      "epoch": 0.19384080466906742,
+      "grad_norm": 0.313436895608902,
+      "learning_rate": 5e-05,
+      "loss": 2.6096,
+      "step": 1561
+    },
+    {
+      "epoch": 0.19396498199428786,
+      "grad_norm": 0.24266557395458221,
+      "learning_rate": 5e-05,
+      "loss": 1.8797,
+      "step": 1562
+    },
+    {
+      "epoch": 0.19408915931950826,
+      "grad_norm": 0.5439087748527527,
+      "learning_rate": 5e-05,
+      "loss": 3.3252,
+      "step": 1563
+    },
+    {
+      "epoch": 0.19421333664472867,
+      "grad_norm": 0.2875012159347534,
+      "learning_rate": 5e-05,
+      "loss": 1.8611,
+      "step": 1564
+    },
+    {
+      "epoch": 0.19433751396994908,
+      "grad_norm": 0.42187750339508057,
+      "learning_rate": 5e-05,
+      "loss": 2.5633,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1944616912951695,
+      "grad_norm": 0.2605687975883484,
+      "learning_rate": 5e-05,
+      "loss": 1.7025,
+      "step": 1566
+    },
+    {
+      "epoch": 0.19458586862038993,
+      "grad_norm": 0.241758793592453,
+      "learning_rate": 5e-05,
+      "loss": 1.681,
+      "step": 1567
+    },
+    {
+      "epoch": 0.19471004594561034,
+      "grad_norm": 0.2815418243408203,
+      "learning_rate": 5e-05,
+      "loss": 1.8833,
+      "step": 1568
+    },
+    {
+      "epoch": 0.19483422327083075,
+      "grad_norm": 0.23800040781497955,
+      "learning_rate": 5e-05,
+      "loss": 1.746,
+      "step": 1569
+    },
+    {
+      "epoch": 0.19495840059605116,
+      "grad_norm": 0.2602657377719879,
+      "learning_rate": 5e-05,
+      "loss": 2.0523,
+      "step": 1570
+    },
+    {
+      "epoch": 0.19508257792127157,
+      "grad_norm": 0.2506142556667328,
+      "learning_rate": 5e-05,
+      "loss": 1.9698,
+      "step": 1571
+    },
+    {
+      "epoch": 0.195206755246492,
+      "grad_norm": 0.274129718542099,
+      "learning_rate": 5e-05,
+      "loss": 2.0293,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1953309325717124,
+      "grad_norm": 0.21423161029815674,
+      "learning_rate": 5e-05,
+      "loss": 1.74,
+      "step": 1573
+    },
+    {
+      "epoch": 0.19545510989693282,
+      "grad_norm": 0.2414734810590744,
+      "learning_rate": 5e-05,
+      "loss": 1.918,
+      "step": 1574
+    },
+    {
+      "epoch": 0.19557928722215323,
+      "grad_norm": 0.303185373544693,
+      "learning_rate": 5e-05,
+      "loss": 2.1095,
+      "step": 1575
+    },
+    {
+      "epoch": 0.19570346454737364,
+      "grad_norm": 0.35101062059402466,
+      "learning_rate": 5e-05,
+      "loss": 2.2749,
+      "step": 1576
+    },
+    {
+      "epoch": 0.19582764187259408,
+      "grad_norm": 0.23602382838726044,
+      "learning_rate": 5e-05,
+      "loss": 1.8409,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1959518191978145,
+      "grad_norm": 0.23884277045726776,
+      "learning_rate": 5e-05,
+      "loss": 1.665,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1960759965230349,
+      "grad_norm": 0.2590658664703369,
+      "learning_rate": 5e-05,
+      "loss": 1.8435,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1962001738482553,
+      "grad_norm": 0.26221731305122375,
+      "learning_rate": 5e-05,
+      "loss": 2.0502,
+      "step": 1580
+    },
+    {
+      "epoch": 0.19632435117347571,
+      "grad_norm": 0.22786976397037506,
+      "learning_rate": 5e-05,
+      "loss": 1.6887,
+      "step": 1581
+    },
+    {
+      "epoch": 0.19644852849869615,
+      "grad_norm": 0.2415599822998047,
+      "learning_rate": 5e-05,
+      "loss": 1.9872,
+      "step": 1582
+    },
+    {
+      "epoch": 0.19657270582391656,
+      "grad_norm": 0.5078092217445374,
+      "learning_rate": 5e-05,
+      "loss": 3.5464,
+      "step": 1583
+    },
+    {
+      "epoch": 0.19669688314913697,
+      "grad_norm": 0.24448655545711517,
+      "learning_rate": 5e-05,
+      "loss": 1.8582,
+      "step": 1584
+    },
+    {
+      "epoch": 0.19682106047435738,
+      "grad_norm": 0.44471386075019836,
+      "learning_rate": 5e-05,
+      "loss": 2.8817,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1969452377995778,
+      "grad_norm": 0.2776273787021637,
+      "learning_rate": 5e-05,
+      "loss": 1.9268,
+      "step": 1586
+    },
+    {
+      "epoch": 0.19706941512479823,
+      "grad_norm": 0.2519216239452362,
+      "learning_rate": 5e-05,
+      "loss": 1.7644,
+      "step": 1587
+    },
+    {
+      "epoch": 0.19719359245001863,
+      "grad_norm": 0.2582288682460785,
+      "learning_rate": 5e-05,
+      "loss": 1.8588,
+      "step": 1588
+    },
+    {
+      "epoch": 0.19731776977523904,
+      "grad_norm": 0.25764769315719604,
+      "learning_rate": 5e-05,
+      "loss": 2.1778,
+      "step": 1589
+    },
+    {
+      "epoch": 0.19744194710045945,
+      "grad_norm": 0.24313528835773468,
+      "learning_rate": 5e-05,
+      "loss": 1.9725,
+      "step": 1590
+    },
+    {
+      "epoch": 0.19756612442567986,
+      "grad_norm": 0.23823384940624237,
+      "learning_rate": 5e-05,
+      "loss": 1.8506,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1976903017509003,
+      "grad_norm": 0.2740623950958252,
+      "learning_rate": 5e-05,
+      "loss": 2.1847,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1978144790761207,
+      "grad_norm": 0.5463706254959106,
+      "learning_rate": 5e-05,
+      "loss": 3.0436,
+      "step": 1593
+    },
+    {
+      "epoch": 0.19793865640134112,
+      "grad_norm": 0.25261610746383667,
+      "learning_rate": 5e-05,
+      "loss": 1.9796,
+      "step": 1594
+    },
+    {
+      "epoch": 0.19806283372656153,
+      "grad_norm": 0.23600277304649353,
+      "learning_rate": 5e-05,
+      "loss": 1.6148,
+      "step": 1595
+    },
+    {
+      "epoch": 0.19818701105178194,
+      "grad_norm": 0.31440797448158264,
+      "learning_rate": 5e-05,
+      "loss": 1.9735,
+      "step": 1596
+    },
+    {
+      "epoch": 0.19831118837700237,
+      "grad_norm": 0.26265764236450195,
+      "learning_rate": 5e-05,
+      "loss": 1.6048,
+      "step": 1597
+    },
+    {
+      "epoch": 0.19843536570222278,
+      "grad_norm": 0.29170727729797363,
+      "learning_rate": 5e-05,
+      "loss": 1.791,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1985595430274432,
+      "grad_norm": 0.2299099862575531,
+      "learning_rate": 5e-05,
+      "loss": 1.7237,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1986837203526636,
+      "grad_norm": 0.24569188058376312,
+      "learning_rate": 5e-05,
+      "loss": 1.7551,
+      "step": 1600
+    },
+    {
+      "epoch": 0.198807897677884,
+      "grad_norm": 0.31103986501693726,
+      "learning_rate": 5e-05,
+      "loss": 2.1028,
+      "step": 1601
+    },
+    {
+      "epoch": 0.19893207500310442,
+      "grad_norm": 0.5164408683776855,
+      "learning_rate": 5e-05,
+      "loss": 3.2438,
+      "step": 1602
+    },
+    {
+      "epoch": 0.19905625232832486,
+      "grad_norm": 0.2590464949607849,
+      "learning_rate": 5e-05,
+      "loss": 1.6186,
+      "step": 1603
+    },
+    {
+      "epoch": 0.19918042965354527,
+      "grad_norm": 0.26119983196258545,
+      "learning_rate": 5e-05,
+      "loss": 1.8619,
+      "step": 1604
+    },
+    {
+      "epoch": 0.19930460697876567,
+      "grad_norm": 0.2564757466316223,
+      "learning_rate": 5e-05,
+      "loss": 1.8927,
+      "step": 1605
+    },
+    {
+      "epoch": 0.19942878430398608,
+      "grad_norm": 0.23796196281909943,
+      "learning_rate": 5e-05,
+      "loss": 1.7918,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1995529616292065,
+      "grad_norm": 0.29652640223503113,
+      "learning_rate": 5e-05,
+      "loss": 1.9427,
+      "step": 1607
+    },
+    {
+      "epoch": 0.19967713895442693,
+      "grad_norm": 0.5349125862121582,
+      "learning_rate": 5e-05,
+      "loss": 2.5095,
+      "step": 1608
+    },
+    {
+      "epoch": 0.19980131627964734,
+      "grad_norm": 0.2431587427854538,
+      "learning_rate": 5e-05,
+      "loss": 1.7721,
+      "step": 1609
+    },
+    {
+      "epoch": 0.19992549360486775,
+      "grad_norm": 0.3856450617313385,
+      "learning_rate": 5e-05,
+      "loss": 3.0868,
+      "step": 1610
+    },
+    {
+      "epoch": 0.20004967093008816,
+      "grad_norm": 0.2545795738697052,
+      "learning_rate": 5e-05,
+      "loss": 1.9842,
+      "step": 1611
+    },
+    {
+      "epoch": 0.20017384825530857,
+      "grad_norm": 0.2770799696445465,
+      "learning_rate": 5e-05,
+      "loss": 2.2248,
+      "step": 1612
+    },
+    {
+      "epoch": 0.200298025580529,
+      "grad_norm": 0.25927576422691345,
+      "learning_rate": 5e-05,
+      "loss": 1.8765,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2004222029057494,
+      "grad_norm": 0.3126044273376465,
+      "learning_rate": 5e-05,
+      "loss": 2.1229,
+      "step": 1614
+    },
+    {
+      "epoch": 0.20054638023096982,
+      "grad_norm": 0.23928864300251007,
+      "learning_rate": 5e-05,
+      "loss": 1.6578,
+      "step": 1615
+    },
+    {
+      "epoch": 0.20067055755619023,
+      "grad_norm": 0.24184995889663696,
+      "learning_rate": 5e-05,
+      "loss": 1.6824,
+      "step": 1616
+    },
+    {
+      "epoch": 0.20079473488141064,
+      "grad_norm": 0.4060536324977875,
+      "learning_rate": 5e-05,
+      "loss": 2.6177,
+      "step": 1617
+    },
+    {
+      "epoch": 0.20091891220663108,
+      "grad_norm": 0.26484787464141846,
+      "learning_rate": 5e-05,
+      "loss": 1.8741,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2010430895318515,
+      "grad_norm": 0.4881327450275421,
+      "learning_rate": 5e-05,
+      "loss": 2.8875,
+      "step": 1619
+    },
+    {
+      "epoch": 0.2011672668570719,
+      "grad_norm": 0.4290279448032379,
+      "learning_rate": 5e-05,
+      "loss": 3.3519,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2012914441822923,
+      "grad_norm": 0.3032493591308594,
+      "learning_rate": 5e-05,
+      "loss": 2.0804,
+      "step": 1621
+    },
+    {
+      "epoch": 0.20141562150751272,
+      "grad_norm": 0.2680701017379761,
+      "learning_rate": 5e-05,
+      "loss": 1.9729,
+      "step": 1622
+    },
+    {
+      "epoch": 0.20153979883273315,
+      "grad_norm": 0.2793412208557129,
+      "learning_rate": 5e-05,
+      "loss": 1.9686,
+      "step": 1623
+    },
+    {
+      "epoch": 0.20166397615795356,
+      "grad_norm": 0.24099324643611908,
+      "learning_rate": 5e-05,
+      "loss": 1.9883,
+      "step": 1624
+    },
+    {
+      "epoch": 0.20178815348317397,
+      "grad_norm": 0.24999384582042694,
+      "learning_rate": 5e-05,
+      "loss": 1.8092,
+      "step": 1625
+    },
+    {
+      "epoch": 0.20191233080839438,
+      "grad_norm": 0.24993352591991425,
+      "learning_rate": 5e-05,
+      "loss": 1.9533,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2020365081336148,
+      "grad_norm": 0.24249236285686493,
+      "learning_rate": 5e-05,
+      "loss": 1.7275,
+      "step": 1627
+    },
+    {
+      "epoch": 0.20216068545883523,
+      "grad_norm": 0.22344590723514557,
+      "learning_rate": 5e-05,
+      "loss": 1.8207,
+      "step": 1628
+    },
+    {
+      "epoch": 0.20228486278405564,
+      "grad_norm": 0.32327112555503845,
+      "learning_rate": 5e-05,
+      "loss": 2.1362,
+      "step": 1629
+    },
+    {
+      "epoch": 0.20240904010927604,
+      "grad_norm": 0.2718002498149872,
+      "learning_rate": 5e-05,
+      "loss": 1.8973,
+      "step": 1630
+    },
+    {
+      "epoch": 0.20253321743449645,
+      "grad_norm": 0.3098478615283966,
+      "learning_rate": 5e-05,
+      "loss": 2.0483,
+      "step": 1631
+    },
+    {
+      "epoch": 0.20265739475971686,
+      "grad_norm": 0.2578212022781372,
+      "learning_rate": 5e-05,
+      "loss": 1.9976,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2027815720849373,
+      "grad_norm": 0.24807208776474,
+      "learning_rate": 5e-05,
+      "loss": 1.9823,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2029057494101577,
+      "grad_norm": 0.23889771103858948,
+      "learning_rate": 5e-05,
+      "loss": 2.0429,
+      "step": 1634
+    },
+    {
+      "epoch": 0.20302992673537812,
+      "grad_norm": 0.29187577962875366,
+      "learning_rate": 5e-05,
+      "loss": 2.3723,
+      "step": 1635
+    },
+    {
+      "epoch": 0.20315410406059853,
+      "grad_norm": 0.24887806177139282,
+      "learning_rate": 5e-05,
+      "loss": 1.981,
+      "step": 1636
+    },
+    {
+      "epoch": 0.20327828138581894,
+      "grad_norm": 0.2588616609573364,
+      "learning_rate": 5e-05,
+      "loss": 1.8584,
+      "step": 1637
+    },
+    {
+      "epoch": 0.20340245871103937,
+      "grad_norm": 0.37118417024612427,
+      "learning_rate": 5e-05,
+      "loss": 2.517,
+      "step": 1638
+    },
+    {
+      "epoch": 0.20352663603625978,
+      "grad_norm": 0.24805858731269836,
+      "learning_rate": 5e-05,
+      "loss": 1.9279,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2036508133614802,
+      "grad_norm": 0.2555654048919678,
+      "learning_rate": 5e-05,
+      "loss": 1.9295,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2037749906867006,
+      "grad_norm": 0.2625608742237091,
+      "learning_rate": 5e-05,
+      "loss": 1.8236,
+      "step": 1641
+    },
+    {
+      "epoch": 0.203899168011921,
+      "grad_norm": 0.25353118777275085,
+      "learning_rate": 5e-05,
+      "loss": 2.0502,
+      "step": 1642
+    },
+    {
+      "epoch": 0.20402334533714145,
+      "grad_norm": 0.295852929353714,
+      "learning_rate": 5e-05,
+      "loss": 2.1663,
+      "step": 1643
+    },
+    {
+      "epoch": 0.20414752266236186,
+      "grad_norm": 0.33449020981788635,
+      "learning_rate": 5e-05,
+      "loss": 2.5646,
+      "step": 1644
+    },
+    {
+      "epoch": 0.20427169998758227,
+      "grad_norm": 0.26184529066085815,
+      "learning_rate": 5e-05,
+      "loss": 2.1138,
+      "step": 1645
+    },
+    {
+      "epoch": 0.20439587731280268,
+      "grad_norm": 0.24221271276474,
+      "learning_rate": 5e-05,
+      "loss": 1.7526,
+      "step": 1646
+    },
+    {
+      "epoch": 0.20452005463802309,
+      "grad_norm": 0.28155601024627686,
+      "learning_rate": 5e-05,
+      "loss": 2.1057,
+      "step": 1647
+    },
+    {
+      "epoch": 0.20464423196324352,
+      "grad_norm": 0.2601582705974579,
+      "learning_rate": 5e-05,
+      "loss": 1.9401,
+      "step": 1648
+    },
+    {
+      "epoch": 0.20476840928846393,
+      "grad_norm": 0.28566107153892517,
+      "learning_rate": 5e-05,
+      "loss": 2.0327,
+      "step": 1649
+    },
+    {
+      "epoch": 0.20489258661368434,
+      "grad_norm": 0.25720515847206116,
+      "learning_rate": 5e-05,
+      "loss": 1.8847,
+      "step": 1650
+    },
+    {
+      "epoch": 0.20501676393890475,
+      "grad_norm": 0.2847684323787689,
+      "learning_rate": 5e-05,
+      "loss": 2.3069,
+      "step": 1651
+    },
+    {
+      "epoch": 0.20514094126412516,
+      "grad_norm": 0.25259941816329956,
+      "learning_rate": 5e-05,
+      "loss": 2.0781,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2052651185893456,
+      "grad_norm": 0.23993083834648132,
+      "learning_rate": 5e-05,
+      "loss": 1.7618,
+      "step": 1653
+    },
+    {
+      "epoch": 0.205389295914566,
+      "grad_norm": 0.2299104630947113,
+      "learning_rate": 5e-05,
+      "loss": 1.8144,
+      "step": 1654
+    },
+    {
+      "epoch": 0.20551347323978642,
+      "grad_norm": 0.255740225315094,
+      "learning_rate": 5e-05,
+      "loss": 2.101,
+      "step": 1655
+    },
+    {
+      "epoch": 0.20563765056500682,
+      "grad_norm": 0.24104425311088562,
+      "learning_rate": 5e-05,
+      "loss": 1.7776,
+      "step": 1656
+    },
+    {
+      "epoch": 0.20576182789022723,
+      "grad_norm": 0.3152559697628021,
+      "learning_rate": 5e-05,
+      "loss": 2.4188,
+      "step": 1657
+    },
+    {
+      "epoch": 0.20588600521544767,
+      "grad_norm": 0.2445751428604126,
+      "learning_rate": 5e-05,
+      "loss": 2.1092,
+      "step": 1658
+    },
+    {
+      "epoch": 0.20601018254066808,
+      "grad_norm": 0.25449883937835693,
+      "learning_rate": 5e-05,
+      "loss": 1.9749,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2061343598658885,
+      "grad_norm": 0.25314974784851074,
+      "learning_rate": 5e-05,
+      "loss": 2.0,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2062585371911089,
+      "grad_norm": 0.2331535667181015,
+      "learning_rate": 5e-05,
+      "loss": 1.7284,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2063827145163293,
+      "grad_norm": 0.2503470182418823,
+      "learning_rate": 5e-05,
+      "loss": 2.1081,
+      "step": 1662
+    },
+    {
+      "epoch": 0.20650689184154974,
+      "grad_norm": 0.2514774799346924,
+      "learning_rate": 5e-05,
+      "loss": 1.9872,
+      "step": 1663
+    },
+    {
+      "epoch": 0.20663106916677015,
+      "grad_norm": 0.442969411611557,
+      "learning_rate": 5e-05,
+      "loss": 2.6126,
+      "step": 1664
+    },
+    {
+      "epoch": 0.20675524649199056,
+      "grad_norm": 0.4162762463092804,
+      "learning_rate": 5e-05,
+      "loss": 2.8093,
+      "step": 1665
+    },
+    {
+      "epoch": 0.20687942381721097,
+      "grad_norm": 0.23191851377487183,
+      "learning_rate": 5e-05,
+      "loss": 1.69,
+      "step": 1666
+    },
+    {
+      "epoch": 0.20700360114243138,
+      "grad_norm": 0.27533841133117676,
+      "learning_rate": 5e-05,
+      "loss": 2.5071,
+      "step": 1667
+    },
+    {
+      "epoch": 0.20712777846765182,
+      "grad_norm": 0.23384816944599152,
+      "learning_rate": 5e-05,
+      "loss": 1.7697,
+      "step": 1668
+    },
+    {
+      "epoch": 0.20725195579287223,
+      "grad_norm": 0.2592547535896301,
+      "learning_rate": 5e-05,
+      "loss": 1.9669,
+      "step": 1669
+    },
+    {
+      "epoch": 0.20737613311809264,
+      "grad_norm": 0.222014382481575,
+      "learning_rate": 5e-05,
+      "loss": 1.6401,
+      "step": 1670
+    },
+    {
+      "epoch": 0.20750031044331305,
+      "grad_norm": 0.2831239402294159,
+      "learning_rate": 5e-05,
+      "loss": 2.1964,
+      "step": 1671
+    },
+    {
+      "epoch": 0.20762448776853346,
+      "grad_norm": 0.277029424905777,
+      "learning_rate": 5e-05,
+      "loss": 2.087,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2077486650937539,
+      "grad_norm": 0.39079350233078003,
+      "learning_rate": 5e-05,
+      "loss": 2.7042,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2078728424189743,
+      "grad_norm": 0.30425843596458435,
+      "learning_rate": 5e-05,
+      "loss": 2.11,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2079970197441947,
+      "grad_norm": 0.3006201386451721,
+      "learning_rate": 5e-05,
+      "loss": 2.0215,
+      "step": 1675
+    },
+    {
+      "epoch": 0.20812119706941512,
+      "grad_norm": 0.3124234974384308,
+      "learning_rate": 5e-05,
+      "loss": 2.3347,
+      "step": 1676
+    },
+    {
+      "epoch": 0.20824537439463553,
+      "grad_norm": 0.23930403590202332,
+      "learning_rate": 5e-05,
+      "loss": 1.8397,
+      "step": 1677
+    },
+    {
+      "epoch": 0.20836955171985597,
+      "grad_norm": 0.31715354323387146,
+      "learning_rate": 5e-05,
+      "loss": 2.6101,
+      "step": 1678
+    },
+    {
+      "epoch": 0.20849372904507638,
+      "grad_norm": 0.37763673067092896,
+      "learning_rate": 5e-05,
+      "loss": 2.2139,
+      "step": 1679
+    },
+    {
+      "epoch": 0.20861790637029679,
+      "grad_norm": 0.2503034174442291,
+      "learning_rate": 5e-05,
+      "loss": 1.8083,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2087420836955172,
+      "grad_norm": 0.23396505415439606,
+      "learning_rate": 5e-05,
+      "loss": 1.865,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2088662610207376,
+      "grad_norm": 0.4701700806617737,
+      "learning_rate": 5e-05,
+      "loss": 3.0018,
+      "step": 1682
+    },
+    {
+      "epoch": 0.20899043834595804,
+      "grad_norm": 0.2465221881866455,
+      "learning_rate": 5e-05,
+      "loss": 1.933,
+      "step": 1683
+    },
+    {
+      "epoch": 0.20911461567117845,
+      "grad_norm": 0.2720159590244293,
+      "learning_rate": 5e-05,
+      "loss": 1.7912,
+      "step": 1684
+    },
+    {
+      "epoch": 0.20923879299639886,
+      "grad_norm": 0.27453941106796265,
+      "learning_rate": 5e-05,
+      "loss": 2.0795,
+      "step": 1685
+    },
+    {
+      "epoch": 0.20936297032161927,
+      "grad_norm": 0.2305481731891632,
+      "learning_rate": 5e-05,
+      "loss": 1.869,
+      "step": 1686
+    },
+    {
+      "epoch": 0.20948714764683968,
+      "grad_norm": 0.4319351613521576,
+      "learning_rate": 5e-05,
+      "loss": 2.6607,
+      "step": 1687
+    },
+    {
+      "epoch": 0.20961132497206011,
+      "grad_norm": 0.2448023408651352,
+      "learning_rate": 5e-05,
+      "loss": 1.873,
+      "step": 1688
+    },
+    {
+      "epoch": 0.20973550229728052,
+      "grad_norm": 0.33915212750434875,
+      "learning_rate": 5e-05,
+      "loss": 2.6374,
+      "step": 1689
+    },
+    {
+      "epoch": 0.20985967962250093,
+      "grad_norm": 0.2421485185623169,
+      "learning_rate": 5e-05,
+      "loss": 1.8568,
+      "step": 1690
+    },
+    {
+      "epoch": 0.20998385694772134,
+      "grad_norm": 0.3487449884414673,
+      "learning_rate": 5e-05,
+      "loss": 2.9961,
+      "step": 1691
+    },
+    {
+      "epoch": 0.21010803427294175,
+      "grad_norm": 0.25106364488601685,
+      "learning_rate": 5e-05,
+      "loss": 1.8972,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2102322115981622,
+      "grad_norm": 0.2260889858007431,
+      "learning_rate": 5e-05,
+      "loss": 1.7283,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2103563889233826,
+      "grad_norm": 0.250914603471756,
+      "learning_rate": 5e-05,
+      "loss": 2.1342,
+      "step": 1694
+    },
+    {
+      "epoch": 0.210480566248603,
+      "grad_norm": 0.26212602853775024,
+      "learning_rate": 5e-05,
+      "loss": 2.2653,
+      "step": 1695
+    },
+    {
+      "epoch": 0.21060474357382342,
+      "grad_norm": 0.32511529326438904,
+      "learning_rate": 5e-05,
+      "loss": 2.4195,
+      "step": 1696
+    },
+    {
+      "epoch": 0.21072892089904383,
+      "grad_norm": 0.25976255536079407,
+      "learning_rate": 5e-05,
+      "loss": 1.9907,
+      "step": 1697
+    },
+    {
+      "epoch": 0.21085309822426426,
+      "grad_norm": 0.26089444756507874,
+      "learning_rate": 5e-05,
+      "loss": 1.7081,
+      "step": 1698
+    },
+    {
+      "epoch": 0.21097727554948467,
+      "grad_norm": 0.2486356496810913,
+      "learning_rate": 5e-05,
+      "loss": 1.8228,
+      "step": 1699
+    },
+    {
+      "epoch": 0.21110145287470508,
+      "grad_norm": 0.2910366952419281,
+      "learning_rate": 5e-05,
+      "loss": 2.2744,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2112256301999255,
+      "grad_norm": 0.2450498342514038,
+      "learning_rate": 5e-05,
+      "loss": 1.9658,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2113498075251459,
+      "grad_norm": 0.2864016592502594,
+      "learning_rate": 5e-05,
+      "loss": 2.2889,
+      "step": 1702
+    },
+    {
+      "epoch": 0.21147398485036634,
+      "grad_norm": 0.35383304953575134,
+      "learning_rate": 5e-05,
+      "loss": 3.0267,
+      "step": 1703
+    },
+    {
+      "epoch": 0.21159816217558675,
+      "grad_norm": 0.3253929018974304,
+      "learning_rate": 5e-05,
+      "loss": 3.065,
+      "step": 1704
+    },
+    {
+      "epoch": 0.21172233950080716,
+      "grad_norm": 0.3689143657684326,
+      "learning_rate": 5e-05,
+      "loss": 3.0832,
+      "step": 1705
+    },
+    {
+      "epoch": 0.21184651682602756,
+      "grad_norm": 0.2674231231212616,
+      "learning_rate": 5e-05,
+      "loss": 1.9696,
+      "step": 1706
+    },
+    {
+      "epoch": 0.21197069415124797,
+      "grad_norm": 0.2332857996225357,
+      "learning_rate": 5e-05,
+      "loss": 1.7146,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2120948714764684,
+      "grad_norm": 0.2352966070175171,
+      "learning_rate": 5e-05,
+      "loss": 1.7532,
+      "step": 1708
+    },
+    {
+      "epoch": 0.21221904880168882,
+      "grad_norm": 0.24981732666492462,
+      "learning_rate": 5e-05,
+      "loss": 1.8056,
+      "step": 1709
+    },
+    {
+      "epoch": 0.21234322612690923,
+      "grad_norm": 0.4079973101615906,
+      "learning_rate": 5e-05,
+      "loss": 2.4134,
+      "step": 1710
+    },
+    {
+      "epoch": 0.21246740345212964,
+      "grad_norm": 0.24902944266796112,
+      "learning_rate": 5e-05,
+      "loss": 1.9135,
+      "step": 1711
+    },
+    {
+      "epoch": 0.21259158077735005,
+      "grad_norm": 0.269527405500412,
+      "learning_rate": 5e-05,
+      "loss": 1.8756,
+      "step": 1712
+    },
+    {
+      "epoch": 0.21271575810257046,
+      "grad_norm": 0.24911288917064667,
+      "learning_rate": 5e-05,
+      "loss": 1.8201,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2128399354277909,
+      "grad_norm": 0.2403162568807602,
+      "learning_rate": 5e-05,
+      "loss": 1.7368,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2129641127530113,
+      "grad_norm": 0.37792307138442993,
+      "learning_rate": 5e-05,
+      "loss": 2.6803,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2130882900782317,
+      "grad_norm": 0.24422739446163177,
+      "learning_rate": 5e-05,
+      "loss": 1.9826,
+      "step": 1716
+    },
+    {
+      "epoch": 0.21321246740345212,
+      "grad_norm": 0.3007884919643402,
+      "learning_rate": 5e-05,
+      "loss": 2.3471,
+      "step": 1717
+    },
+    {
+      "epoch": 0.21333664472867253,
+      "grad_norm": 0.3203009068965912,
+      "learning_rate": 5e-05,
+      "loss": 1.6646,
+      "step": 1718
+    },
+    {
+      "epoch": 0.21346082205389297,
+      "grad_norm": 0.402255117893219,
+      "learning_rate": 5e-05,
+      "loss": 2.9297,
+      "step": 1719
+    },
+    {
+      "epoch": 0.21358499937911338,
+      "grad_norm": 0.3000585436820984,
+      "learning_rate": 5e-05,
+      "loss": 2.0304,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2137091767043338,
+      "grad_norm": 0.24241819977760315,
+      "learning_rate": 5e-05,
+      "loss": 1.7784,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2138333540295542,
+      "grad_norm": 0.24447214603424072,
+      "learning_rate": 5e-05,
+      "loss": 1.917,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2139575313547746,
+      "grad_norm": 0.41224199533462524,
+      "learning_rate": 5e-05,
+      "loss": 2.283,
+      "step": 1723
+    },
+    {
+      "epoch": 0.21408170867999504,
+      "grad_norm": 0.23594479262828827,
+      "learning_rate": 5e-05,
+      "loss": 1.8945,
+      "step": 1724
+    },
+    {
+      "epoch": 0.21420588600521545,
+      "grad_norm": 0.2759271562099457,
+      "learning_rate": 5e-05,
+      "loss": 2.1209,
+      "step": 1725
+    },
+    {
+      "epoch": 0.21433006333043586,
+      "grad_norm": 0.37910234928131104,
+      "learning_rate": 5e-05,
+      "loss": 2.4051,
+      "step": 1726
+    },
+    {
+      "epoch": 0.21445424065565627,
+      "grad_norm": 0.239668607711792,
+      "learning_rate": 5e-05,
+      "loss": 1.8435,
+      "step": 1727
+    },
+    {
+      "epoch": 0.21457841798087668,
+      "grad_norm": 0.3439406454563141,
+      "learning_rate": 5e-05,
+      "loss": 2.3448,
+      "step": 1728
+    },
+    {
+      "epoch": 0.21470259530609712,
+      "grad_norm": 0.24044089019298553,
+      "learning_rate": 5e-05,
+      "loss": 1.8574,
+      "step": 1729
+    },
+    {
+      "epoch": 0.21482677263131753,
+      "grad_norm": 0.3365280330181122,
+      "learning_rate": 5e-05,
+      "loss": 2.2237,
+      "step": 1730
+    },
+    {
+      "epoch": 0.21495094995653793,
+      "grad_norm": 0.2748318016529083,
+      "learning_rate": 5e-05,
+      "loss": 1.876,
+      "step": 1731
+    },
+    {
+      "epoch": 0.21507512728175834,
+      "grad_norm": 0.22987841069698334,
+      "learning_rate": 5e-05,
+      "loss": 1.6272,
+      "step": 1732
+    },
+    {
+      "epoch": 0.21519930460697875,
+      "grad_norm": 0.27578434348106384,
+      "learning_rate": 5e-05,
+      "loss": 2.2288,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2153234819321992,
+      "grad_norm": 0.2659235894680023,
+      "learning_rate": 5e-05,
+      "loss": 2.1437,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2154476592574196,
+      "grad_norm": 0.23005495965480804,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 1735
+    },
+    {
+      "epoch": 0.21557183658264,
+      "grad_norm": 0.24237008392810822,
+      "learning_rate": 5e-05,
+      "loss": 1.7442,
+      "step": 1736
+    },
+    {
+      "epoch": 0.21569601390786042,
+      "grad_norm": 0.25887253880500793,
+      "learning_rate": 5e-05,
+      "loss": 1.8883,
+      "step": 1737
+    },
+    {
+      "epoch": 0.21582019123308083,
+      "grad_norm": 0.24189727008342743,
+      "learning_rate": 5e-05,
+      "loss": 1.8856,
+      "step": 1738
+    },
+    {
+      "epoch": 0.21594436855830126,
+      "grad_norm": 0.22908759117126465,
+      "learning_rate": 5e-05,
+      "loss": 1.992,
+      "step": 1739
+    },
+    {
+      "epoch": 0.21606854588352167,
+      "grad_norm": 0.2191503942012787,
+      "learning_rate": 5e-05,
+      "loss": 1.7492,
+      "step": 1740
+    },
+    {
+      "epoch": 0.21619272320874208,
+      "grad_norm": 0.2622680068016052,
+      "learning_rate": 5e-05,
+      "loss": 1.8653,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2163169005339625,
+      "grad_norm": 0.41318443417549133,
+      "learning_rate": 5e-05,
+      "loss": 2.3992,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2164410778591829,
+      "grad_norm": 0.2093416154384613,
+      "learning_rate": 5e-05,
+      "loss": 1.668,
+      "step": 1743
+    },
+    {
+      "epoch": 0.21656525518440334,
+      "grad_norm": 0.2728491425514221,
+      "learning_rate": 5e-05,
+      "loss": 2.1623,
+      "step": 1744
+    },
+    {
+      "epoch": 0.21668943250962375,
+      "grad_norm": 0.23795777559280396,
+      "learning_rate": 5e-05,
+      "loss": 1.9294,
+      "step": 1745
+    },
+    {
+      "epoch": 0.21681360983484416,
+      "grad_norm": 0.2644427716732025,
+      "learning_rate": 5e-05,
+      "loss": 2.0939,
+      "step": 1746
+    },
+    {
+      "epoch": 0.21693778716006457,
+      "grad_norm": 0.2713290750980377,
+      "learning_rate": 5e-05,
+      "loss": 1.8957,
+      "step": 1747
+    },
+    {
+      "epoch": 0.21706196448528498,
+      "grad_norm": 0.33181139826774597,
+      "learning_rate": 5e-05,
+      "loss": 2.4719,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2171861418105054,
+      "grad_norm": 0.4779377281665802,
+      "learning_rate": 5e-05,
+      "loss": 3.0995,
+      "step": 1749
+    },
+    {
+      "epoch": 0.21731031913572582,
+      "grad_norm": 0.23001569509506226,
+      "learning_rate": 5e-05,
+      "loss": 1.8776,
+      "step": 1750
+    },
+    {
+      "epoch": 0.21743449646094623,
+      "grad_norm": 0.237697571516037,
+      "learning_rate": 5e-05,
+      "loss": 1.7986,
+      "step": 1751
+    },
+    {
+      "epoch": 0.21755867378616664,
+      "grad_norm": 0.35161593556404114,
+      "learning_rate": 5e-05,
+      "loss": 2.0178,
+      "step": 1752
+    },
+    {
+      "epoch": 0.21768285111138705,
+      "grad_norm": 0.33135032653808594,
+      "learning_rate": 5e-05,
+      "loss": 2.0187,
+      "step": 1753
+    },
+    {
+      "epoch": 0.21780702843660749,
+      "grad_norm": 0.238532155752182,
+      "learning_rate": 5e-05,
+      "loss": 1.773,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2179312057618279,
+      "grad_norm": 0.259236216545105,
+      "learning_rate": 5e-05,
+      "loss": 1.9786,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2180553830870483,
+      "grad_norm": 0.7529013752937317,
+      "learning_rate": 5e-05,
+      "loss": 3.5159,
+      "step": 1756
+    },
+    {
+      "epoch": 0.2181795604122687,
+      "grad_norm": 0.26810574531555176,
+      "learning_rate": 5e-05,
+      "loss": 1.7452,
+      "step": 1757
+    },
+    {
+      "epoch": 0.21830373773748912,
+      "grad_norm": 0.29408180713653564,
+      "learning_rate": 5e-05,
+      "loss": 1.9413,
+      "step": 1758
+    },
+    {
+      "epoch": 0.21842791506270956,
+      "grad_norm": 0.7544334530830383,
+      "learning_rate": 5e-05,
+      "loss": 2.8931,
+      "step": 1759
+    },
+    {
+      "epoch": 0.21855209238792997,
+      "grad_norm": 0.2329382449388504,
+      "learning_rate": 5e-05,
+      "loss": 1.7157,
+      "step": 1760
+    },
+    {
+      "epoch": 0.21867626971315038,
+      "grad_norm": 0.29731473326683044,
+      "learning_rate": 5e-05,
+      "loss": 2.4081,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2188004470383708,
+      "grad_norm": 0.28266817331314087,
+      "learning_rate": 5e-05,
+      "loss": 1.7907,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2189246243635912,
+      "grad_norm": 0.37541618943214417,
+      "learning_rate": 5e-05,
+      "loss": 2.9788,
+      "step": 1763
+    },
+    {
+      "epoch": 0.21904880168881163,
+      "grad_norm": 0.36282044649124146,
+      "learning_rate": 5e-05,
+      "loss": 2.604,
+      "step": 1764
+    },
+    {
+      "epoch": 0.21917297901403204,
+      "grad_norm": 0.3002357482910156,
+      "learning_rate": 5e-05,
+      "loss": 1.927,
+      "step": 1765
+    },
+    {
+      "epoch": 0.21929715633925245,
+      "grad_norm": 0.3582640290260315,
+      "learning_rate": 5e-05,
+      "loss": 2.763,
+      "step": 1766
+    },
+    {
+      "epoch": 0.21942133366447286,
+      "grad_norm": 0.27056246995925903,
+      "learning_rate": 5e-05,
+      "loss": 1.6919,
+      "step": 1767
+    },
+    {
+      "epoch": 0.21954551098969327,
+      "grad_norm": 0.2792450487613678,
+      "learning_rate": 5e-05,
+      "loss": 1.8795,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2196696883149137,
+      "grad_norm": 0.31989672780036926,
+      "learning_rate": 5e-05,
+      "loss": 1.9076,
+      "step": 1769
+    },
+    {
+      "epoch": 0.21979386564013412,
+      "grad_norm": 0.2750644087791443,
+      "learning_rate": 5e-05,
+      "loss": 1.9688,
+      "step": 1770
+    },
+    {
+      "epoch": 0.21991804296535453,
+      "grad_norm": 0.2729717791080475,
+      "learning_rate": 5e-05,
+      "loss": 2.6162,
+      "step": 1771
+    },
+    {
+      "epoch": 0.22004222029057494,
+      "grad_norm": 0.3536197543144226,
+      "learning_rate": 5e-05,
+      "loss": 2.5758,
+      "step": 1772
+    },
+    {
+      "epoch": 0.22016639761579535,
+      "grad_norm": 0.2544414699077606,
+      "learning_rate": 5e-05,
+      "loss": 1.8532,
+      "step": 1773
+    },
+    {
+      "epoch": 0.22029057494101578,
+      "grad_norm": 0.29141852259635925,
+      "learning_rate": 5e-05,
+      "loss": 1.8317,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2204147522662362,
+      "grad_norm": 0.3074047267436981,
+      "learning_rate": 5e-05,
+      "loss": 2.4807,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2205389295914566,
+      "grad_norm": 0.28151294589042664,
+      "learning_rate": 5e-05,
+      "loss": 1.777,
+      "step": 1776
+    },
+    {
+      "epoch": 0.220663106916677,
+      "grad_norm": 0.38786178827285767,
+      "learning_rate": 5e-05,
+      "loss": 2.2301,
+      "step": 1777
+    },
+    {
+      "epoch": 0.22078728424189742,
+      "grad_norm": 0.3930966556072235,
+      "learning_rate": 5e-05,
+      "loss": 2.2059,
+      "step": 1778
+    },
+    {
+      "epoch": 0.22091146156711786,
+      "grad_norm": 0.40529048442840576,
+      "learning_rate": 5e-05,
+      "loss": 2.9492,
+      "step": 1779
+    },
+    {
+      "epoch": 0.22103563889233827,
+      "grad_norm": 0.24425935745239258,
+      "learning_rate": 5e-05,
+      "loss": 1.8041,
+      "step": 1780
+    },
+    {
+      "epoch": 0.22115981621755867,
+      "grad_norm": 0.24750569462776184,
+      "learning_rate": 5e-05,
+      "loss": 1.9951,
+      "step": 1781
+    },
+    {
+      "epoch": 0.22128399354277908,
+      "grad_norm": 0.23385198414325714,
+      "learning_rate": 5e-05,
+      "loss": 1.7176,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2214081708679995,
+      "grad_norm": 0.2430528998374939,
+      "learning_rate": 5e-05,
+      "loss": 1.6822,
+      "step": 1783
+    },
+    {
+      "epoch": 0.22153234819321993,
+      "grad_norm": 0.2879598140716553,
+      "learning_rate": 5e-05,
+      "loss": 2.1233,
+      "step": 1784
+    },
+    {
+      "epoch": 0.22165652551844034,
+      "grad_norm": 0.2560439109802246,
+      "learning_rate": 5e-05,
+      "loss": 1.789,
+      "step": 1785
+    },
+    {
+      "epoch": 0.22178070284366075,
+      "grad_norm": 0.22753004729747772,
+      "learning_rate": 5e-05,
+      "loss": 1.8389,
+      "step": 1786
+    },
+    {
+      "epoch": 0.22190488016888116,
+      "grad_norm": 0.2329389899969101,
+      "learning_rate": 5e-05,
+      "loss": 1.8575,
+      "step": 1787
+    },
+    {
+      "epoch": 0.22202905749410157,
+      "grad_norm": 0.3101327121257782,
+      "learning_rate": 5e-05,
+      "loss": 2.8564,
+      "step": 1788
+    },
+    {
+      "epoch": 0.222153234819322,
+      "grad_norm": 0.28837624192237854,
+      "learning_rate": 5e-05,
+      "loss": 2.6174,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2222774121445424,
+      "grad_norm": 0.2643578350543976,
+      "learning_rate": 5e-05,
+      "loss": 1.8305,
+      "step": 1790
+    },
+    {
+      "epoch": 0.22240158946976282,
+      "grad_norm": 0.23591133952140808,
+      "learning_rate": 5e-05,
+      "loss": 1.8651,
+      "step": 1791
+    },
+    {
+      "epoch": 0.22252576679498323,
+      "grad_norm": 0.23883233964443207,
+      "learning_rate": 5e-05,
+      "loss": 1.8016,
+      "step": 1792
+    },
+    {
+      "epoch": 0.22264994412020364,
+      "grad_norm": 0.25554805994033813,
+      "learning_rate": 5e-05,
+      "loss": 1.7941,
+      "step": 1793
+    },
+    {
+      "epoch": 0.22277412144542408,
+      "grad_norm": 0.2344476580619812,
+      "learning_rate": 5e-05,
+      "loss": 1.8561,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2228982987706445,
+      "grad_norm": 0.2391137033700943,
+      "learning_rate": 5e-05,
+      "loss": 1.6769,
+      "step": 1795
+    },
+    {
+      "epoch": 0.2230224760958649,
+      "grad_norm": 0.29635146260261536,
+      "learning_rate": 5e-05,
+      "loss": 2.2268,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2231466534210853,
+      "grad_norm": 0.25983938574790955,
+      "learning_rate": 5e-05,
+      "loss": 2.1469,
+      "step": 1797
+    },
+    {
+      "epoch": 0.22327083074630572,
+      "grad_norm": 0.28056690096855164,
+      "learning_rate": 5e-05,
+      "loss": 2.4135,
+      "step": 1798
+    },
+    {
+      "epoch": 0.22339500807152615,
+      "grad_norm": 0.22447097301483154,
+      "learning_rate": 5e-05,
+      "loss": 1.6283,
+      "step": 1799
+    },
+    {
+      "epoch": 0.22351918539674656,
+      "grad_norm": 0.4292510449886322,
+      "learning_rate": 5e-05,
+      "loss": 3.0222,
+      "step": 1800
+    },
+    {
+      "epoch": 0.22364336272196697,
+      "grad_norm": 0.27744293212890625,
+      "learning_rate": 5e-05,
+      "loss": 2.0844,
+      "step": 1801
+    },
+    {
+      "epoch": 0.22376754004718738,
+      "grad_norm": 0.24944917857646942,
+      "learning_rate": 5e-05,
+      "loss": 2.1492,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2238917173724078,
+      "grad_norm": 0.2881680428981781,
+      "learning_rate": 5e-05,
+      "loss": 1.9108,
+      "step": 1803
+    },
+    {
+      "epoch": 0.22401589469762823,
+      "grad_norm": 0.22899511456489563,
+      "learning_rate": 5e-05,
+      "loss": 1.8498,
+      "step": 1804
+    },
+    {
+      "epoch": 0.22414007202284864,
+      "grad_norm": 0.2350693792104721,
+      "learning_rate": 5e-05,
+      "loss": 1.8751,
+      "step": 1805
+    },
+    {
+      "epoch": 0.22426424934806904,
+      "grad_norm": 0.23344671726226807,
+      "learning_rate": 5e-05,
+      "loss": 1.8933,
+      "step": 1806
+    },
+    {
+      "epoch": 0.22438842667328945,
+      "grad_norm": 0.2569604814052582,
+      "learning_rate": 5e-05,
+      "loss": 1.8855,
+      "step": 1807
+    },
+    {
+      "epoch": 0.22451260399850986,
+      "grad_norm": 0.2852650284767151,
+      "learning_rate": 5e-05,
+      "loss": 2.2437,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2246367813237303,
+      "grad_norm": 0.27842971682548523,
+      "learning_rate": 5e-05,
+      "loss": 2.1467,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2247609586489507,
+      "grad_norm": 0.2501443028450012,
+      "learning_rate": 5e-05,
+      "loss": 1.8606,
+      "step": 1810
+    },
+    {
+      "epoch": 0.22488513597417112,
+      "grad_norm": 0.23555615544319153,
+      "learning_rate": 5e-05,
+      "loss": 1.7129,
+      "step": 1811
+    },
+    {
+      "epoch": 0.22500931329939153,
+      "grad_norm": 0.23604919016361237,
+      "learning_rate": 5e-05,
+      "loss": 1.9276,
+      "step": 1812
+    },
+    {
+      "epoch": 0.22513349062461194,
+      "grad_norm": 0.266314834356308,
+      "learning_rate": 5e-05,
+      "loss": 2.0413,
+      "step": 1813
+    },
+    {
+      "epoch": 0.22525766794983237,
+      "grad_norm": 0.23061490058898926,
+      "learning_rate": 5e-05,
+      "loss": 1.7077,
+      "step": 1814
+    },
+    {
+      "epoch": 0.22538184527505278,
+      "grad_norm": 0.3107684552669525,
+      "learning_rate": 5e-05,
+      "loss": 2.4371,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2255060226002732,
+      "grad_norm": 0.2338070273399353,
+      "learning_rate": 5e-05,
+      "loss": 1.868,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2256301999254936,
+      "grad_norm": 0.22487351298332214,
+      "learning_rate": 5e-05,
+      "loss": 1.7228,
+      "step": 1817
+    },
+    {
+      "epoch": 0.225754377250714,
+      "grad_norm": 0.22506463527679443,
+      "learning_rate": 5e-05,
+      "loss": 1.8726,
+      "step": 1818
+    },
+    {
+      "epoch": 0.22587855457593445,
+      "grad_norm": 0.22189858555793762,
+      "learning_rate": 5e-05,
+      "loss": 1.8849,
+      "step": 1819
+    },
+    {
+      "epoch": 0.22600273190115486,
+      "grad_norm": 0.35950690507888794,
+      "learning_rate": 5e-05,
+      "loss": 3.0113,
+      "step": 1820
+    },
+    {
+      "epoch": 0.22612690922637527,
+      "grad_norm": 0.3117333650588989,
+      "learning_rate": 5e-05,
+      "loss": 2.5189,
+      "step": 1821
+    },
+    {
+      "epoch": 0.22625108655159568,
+      "grad_norm": 0.2612452507019043,
+      "learning_rate": 5e-05,
+      "loss": 2.3857,
+      "step": 1822
+    },
+    {
+      "epoch": 0.22637526387681609,
+      "grad_norm": 0.24799507856369019,
+      "learning_rate": 5e-05,
+      "loss": 1.8806,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2264994412020365,
+      "grad_norm": 0.23160864412784576,
+      "learning_rate": 5e-05,
+      "loss": 1.7434,
+      "step": 1824
+    },
+    {
+      "epoch": 0.22662361852725693,
+      "grad_norm": 0.24015772342681885,
+      "learning_rate": 5e-05,
+      "loss": 1.6574,
+      "step": 1825
+    },
+    {
+      "epoch": 0.22674779585247734,
+      "grad_norm": 0.3072112500667572,
+      "learning_rate": 5e-05,
+      "loss": 1.8564,
+      "step": 1826
+    },
+    {
+      "epoch": 0.22687197317769775,
+      "grad_norm": 0.40338465571403503,
+      "learning_rate": 5e-05,
+      "loss": 2.382,
+      "step": 1827
+    },
+    {
+      "epoch": 0.22699615050291816,
+      "grad_norm": 0.2783093750476837,
+      "learning_rate": 5e-05,
+      "loss": 2.0268,
+      "step": 1828
+    },
+    {
+      "epoch": 0.22712032782813857,
+      "grad_norm": 0.23229126632213593,
+      "learning_rate": 5e-05,
+      "loss": 1.657,
+      "step": 1829
+    },
+    {
+      "epoch": 0.227244505153359,
+      "grad_norm": 0.29547664523124695,
+      "learning_rate": 5e-05,
+      "loss": 2.0992,
+      "step": 1830
+    },
+    {
+      "epoch": 0.22736868247857941,
+      "grad_norm": 0.3030807375907898,
+      "learning_rate": 5e-05,
+      "loss": 2.1506,
+      "step": 1831
+    },
+    {
+      "epoch": 0.22749285980379982,
+      "grad_norm": 0.44966810941696167,
+      "learning_rate": 5e-05,
+      "loss": 2.7974,
+      "step": 1832
+    },
+    {
+      "epoch": 0.22761703712902023,
+      "grad_norm": 0.24478162825107574,
+      "learning_rate": 5e-05,
+      "loss": 1.805,
+      "step": 1833
+    },
+    {
+      "epoch": 0.22774121445424064,
+      "grad_norm": 0.25039052963256836,
+      "learning_rate": 5e-05,
+      "loss": 1.8524,
+      "step": 1834
+    },
+    {
+      "epoch": 0.22786539177946108,
+      "grad_norm": 0.2906454801559448,
+      "learning_rate": 5e-05,
+      "loss": 2.3401,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2279895691046815,
+      "grad_norm": 0.2517564594745636,
+      "learning_rate": 5e-05,
+      "loss": 1.8607,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2281137464299019,
+      "grad_norm": 0.3775496482849121,
+      "learning_rate": 5e-05,
+      "loss": 2.5024,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2282379237551223,
+      "grad_norm": 0.49359723925590515,
+      "learning_rate": 5e-05,
+      "loss": 2.1549,
+      "step": 1838
+    },
+    {
+      "epoch": 0.22836210108034272,
+      "grad_norm": 0.2934791147708893,
+      "learning_rate": 5e-05,
+      "loss": 1.9626,
+      "step": 1839
+    },
+    {
+      "epoch": 0.22848627840556315,
+      "grad_norm": 0.34492501616477966,
+      "learning_rate": 5e-05,
+      "loss": 2.692,
+      "step": 1840
+    },
+    {
+      "epoch": 0.22861045573078356,
+      "grad_norm": 0.23236510157585144,
+      "learning_rate": 5e-05,
+      "loss": 1.742,
+      "step": 1841
+    },
+    {
+      "epoch": 0.22873463305600397,
+      "grad_norm": 0.25195273756980896,
+      "learning_rate": 5e-05,
+      "loss": 1.787,
+      "step": 1842
+    },
+    {
+      "epoch": 0.22885881038122438,
+      "grad_norm": 0.23780708014965057,
+      "learning_rate": 5e-05,
+      "loss": 1.9003,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2289829877064448,
+      "grad_norm": 0.24621467292308807,
+      "learning_rate": 5e-05,
+      "loss": 1.9658,
+      "step": 1844
+    },
+    {
+      "epoch": 0.22910716503166523,
+      "grad_norm": 0.22110295295715332,
+      "learning_rate": 5e-05,
+      "loss": 1.6332,
+      "step": 1845
+    },
+    {
+      "epoch": 0.22923134235688564,
+      "grad_norm": 0.2576463520526886,
+      "learning_rate": 5e-05,
+      "loss": 1.8704,
+      "step": 1846
+    },
+    {
+      "epoch": 0.22935551968210605,
+      "grad_norm": 0.29511019587516785,
+      "learning_rate": 5e-05,
+      "loss": 2.3014,
+      "step": 1847
+    },
+    {
+      "epoch": 0.22947969700732646,
+      "grad_norm": 0.22384779155254364,
+      "learning_rate": 5e-05,
+      "loss": 1.7937,
+      "step": 1848
+    },
+    {
+      "epoch": 0.22960387433254686,
+      "grad_norm": 0.23958070576190948,
+      "learning_rate": 5e-05,
+      "loss": 2.0166,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2297280516577673,
+      "grad_norm": 0.22880825400352478,
+      "learning_rate": 5e-05,
+      "loss": 1.8631,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2298522289829877,
+      "grad_norm": 0.23036791384220123,
+      "learning_rate": 5e-05,
+      "loss": 1.7562,
+      "step": 1851
+    },
+    {
+      "epoch": 0.22997640630820812,
+      "grad_norm": 0.22369766235351562,
+      "learning_rate": 5e-05,
+      "loss": 1.8027,
+      "step": 1852
+    },
+    {
+      "epoch": 0.23010058363342853,
+      "grad_norm": 0.21710707247257233,
+      "learning_rate": 5e-05,
+      "loss": 1.9178,
+      "step": 1853
+    },
+    {
+      "epoch": 0.23022476095864894,
+      "grad_norm": 0.23242232203483582,
+      "learning_rate": 5e-05,
+      "loss": 1.8386,
+      "step": 1854
+    },
+    {
+      "epoch": 0.23034893828386938,
+      "grad_norm": 0.25150173902511597,
+      "learning_rate": 5e-05,
+      "loss": 2.1614,
+      "step": 1855
+    },
+    {
+      "epoch": 0.23047311560908978,
+      "grad_norm": 0.2314678281545639,
+      "learning_rate": 5e-05,
+      "loss": 1.6671,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2305972929343102,
+      "grad_norm": 0.24287739396095276,
+      "learning_rate": 5e-05,
+      "loss": 1.8269,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2307214702595306,
+      "grad_norm": 0.28835412859916687,
+      "learning_rate": 5e-05,
+      "loss": 2.0266,
+      "step": 1858
+    },
+    {
+      "epoch": 0.230845647584751,
+      "grad_norm": 0.24327930808067322,
+      "learning_rate": 5e-05,
+      "loss": 1.9676,
+      "step": 1859
+    },
+    {
+      "epoch": 0.23096982490997145,
+      "grad_norm": 0.22893552482128143,
+      "learning_rate": 5e-05,
+      "loss": 1.697,
+      "step": 1860
+    },
+    {
+      "epoch": 0.23109400223519186,
+      "grad_norm": 0.2457294762134552,
+      "learning_rate": 5e-05,
+      "loss": 2.1678,
+      "step": 1861
+    },
+    {
+      "epoch": 0.23121817956041227,
+      "grad_norm": 0.2598188519477844,
+      "learning_rate": 5e-05,
+      "loss": 2.1675,
+      "step": 1862
+    },
+    {
+      "epoch": 0.23134235688563268,
+      "grad_norm": 0.25997206568717957,
+      "learning_rate": 5e-05,
+      "loss": 2.4369,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2314665342108531,
+      "grad_norm": 0.23384526371955872,
+      "learning_rate": 5e-05,
+      "loss": 1.9082,
+      "step": 1864
+    },
+    {
+      "epoch": 0.23159071153607352,
+      "grad_norm": 0.2367602437734604,
+      "learning_rate": 5e-05,
+      "loss": 2.0341,
+      "step": 1865
+    },
+    {
+      "epoch": 0.23171488886129393,
+      "grad_norm": 0.23032382130622864,
+      "learning_rate": 5e-05,
+      "loss": 1.8408,
+      "step": 1866
+    },
+    {
+      "epoch": 0.23183906618651434,
+      "grad_norm": 0.24111077189445496,
+      "learning_rate": 5e-05,
+      "loss": 1.9525,
+      "step": 1867
+    },
+    {
+      "epoch": 0.23196324351173475,
+      "grad_norm": 0.25044921040534973,
+      "learning_rate": 5e-05,
+      "loss": 1.9545,
+      "step": 1868
+    },
+    {
+      "epoch": 0.23208742083695516,
+      "grad_norm": 0.24643723666667938,
+      "learning_rate": 5e-05,
+      "loss": 1.9355,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2322115981621756,
+      "grad_norm": 0.35006821155548096,
+      "learning_rate": 5e-05,
+      "loss": 2.2422,
+      "step": 1870
+    },
+    {
+      "epoch": 0.232335775487396,
+      "grad_norm": 0.34157219529151917,
+      "learning_rate": 5e-05,
+      "loss": 2.5443,
+      "step": 1871
+    },
+    {
+      "epoch": 0.23245995281261642,
+      "grad_norm": 0.24474504590034485,
+      "learning_rate": 5e-05,
+      "loss": 1.8225,
+      "step": 1872
+    },
+    {
+      "epoch": 0.23258413013783683,
+      "grad_norm": 0.3194948434829712,
+      "learning_rate": 5e-05,
+      "loss": 2.1807,
+      "step": 1873
+    },
+    {
+      "epoch": 0.23270830746305723,
+      "grad_norm": 0.22786596417427063,
+      "learning_rate": 5e-05,
+      "loss": 1.87,
+      "step": 1874
+    },
+    {
+      "epoch": 0.23283248478827767,
+      "grad_norm": 0.28753167390823364,
+      "learning_rate": 5e-05,
+      "loss": 1.8951,
+      "step": 1875
+    },
+    {
+      "epoch": 0.23295666211349808,
+      "grad_norm": 0.2794932723045349,
+      "learning_rate": 5e-05,
+      "loss": 1.7887,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2330808394387185,
+      "grad_norm": 0.23419396579265594,
+      "learning_rate": 5e-05,
+      "loss": 1.8828,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2332050167639389,
+      "grad_norm": 0.2391512244939804,
+      "learning_rate": 5e-05,
+      "loss": 1.7248,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2333291940891593,
+      "grad_norm": 0.22935175895690918,
+      "learning_rate": 5e-05,
+      "loss": 2.0159,
+      "step": 1879
+    },
+    {
+      "epoch": 0.23345337141437975,
+      "grad_norm": 0.2907181680202484,
+      "learning_rate": 5e-05,
+      "loss": 2.6502,
+      "step": 1880
+    },
+    {
+      "epoch": 0.23357754873960015,
+      "grad_norm": 0.22393736243247986,
+      "learning_rate": 5e-05,
+      "loss": 1.8167,
+      "step": 1881
+    },
+    {
+      "epoch": 0.23370172606482056,
+      "grad_norm": 0.23955309391021729,
+      "learning_rate": 5e-05,
+      "loss": 1.8751,
+      "step": 1882
+    },
+    {
+      "epoch": 0.23382590339004097,
+      "grad_norm": 0.29608166217803955,
+      "learning_rate": 5e-05,
+      "loss": 2.3452,
+      "step": 1883
+    },
+    {
+      "epoch": 0.23395008071526138,
+      "grad_norm": 0.2643882632255554,
+      "learning_rate": 5e-05,
+      "loss": 1.9328,
+      "step": 1884
+    },
+    {
+      "epoch": 0.23407425804048182,
+      "grad_norm": 0.3525054454803467,
+      "learning_rate": 5e-05,
+      "loss": 3.1187,
+      "step": 1885
+    },
+    {
+      "epoch": 0.23419843536570223,
+      "grad_norm": 0.23440736532211304,
+      "learning_rate": 5e-05,
+      "loss": 1.9781,
+      "step": 1886
+    },
+    {
+      "epoch": 0.23432261269092264,
+      "grad_norm": 0.2511214315891266,
+      "learning_rate": 5e-05,
+      "loss": 1.7289,
+      "step": 1887
+    },
+    {
+      "epoch": 0.23444679001614305,
+      "grad_norm": 0.25337353348731995,
+      "learning_rate": 5e-05,
+      "loss": 1.7276,
+      "step": 1888
+    },
+    {
+      "epoch": 0.23457096734136346,
+      "grad_norm": 0.3251419961452484,
+      "learning_rate": 5e-05,
+      "loss": 2.1247,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2346951446665839,
+      "grad_norm": 0.2831568717956543,
+      "learning_rate": 5e-05,
+      "loss": 1.8562,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2348193219918043,
+      "grad_norm": 0.34750649333000183,
+      "learning_rate": 5e-05,
+      "loss": 2.4733,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2349434993170247,
+      "grad_norm": 0.277805894613266,
+      "learning_rate": 5e-05,
+      "loss": 2.0188,
+      "step": 1892
+    },
+    {
+      "epoch": 0.23506767664224512,
+      "grad_norm": 0.36019158363342285,
+      "learning_rate": 5e-05,
+      "loss": 2.6607,
+      "step": 1893
+    },
+    {
+      "epoch": 0.23519185396746553,
+      "grad_norm": 0.2482622265815735,
+      "learning_rate": 5e-05,
+      "loss": 1.8025,
+      "step": 1894
+    },
+    {
+      "epoch": 0.23531603129268597,
+      "grad_norm": 0.3870713412761688,
+      "learning_rate": 5e-05,
+      "loss": 2.5754,
+      "step": 1895
+    },
+    {
+      "epoch": 0.23544020861790638,
+      "grad_norm": 0.4541047513484955,
+      "learning_rate": 5e-05,
+      "loss": 2.4462,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2355643859431268,
+      "grad_norm": 0.2338835746049881,
+      "learning_rate": 5e-05,
+      "loss": 1.9402,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2356885632683472,
+      "grad_norm": 0.32460445165634155,
+      "learning_rate": 5e-05,
+      "loss": 2.6475,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2358127405935676,
+      "grad_norm": 0.26159971952438354,
+      "learning_rate": 5e-05,
+      "loss": 2.274,
+      "step": 1899
+    },
+    {
+      "epoch": 0.23593691791878804,
+      "grad_norm": 0.3683694899082184,
+      "learning_rate": 5e-05,
+      "loss": 2.4997,
+      "step": 1900
+    },
+    {
+      "epoch": 0.23606109524400845,
+      "grad_norm": 0.257699191570282,
+      "learning_rate": 5e-05,
+      "loss": 1.7806,
+      "step": 1901
+    },
+    {
+      "epoch": 0.23618527256922886,
+      "grad_norm": 0.24818864464759827,
+      "learning_rate": 5e-05,
+      "loss": 1.7533,
+      "step": 1902
+    },
+    {
+      "epoch": 0.23630944989444927,
+      "grad_norm": 0.257609486579895,
+      "learning_rate": 5e-05,
+      "loss": 1.8761,
+      "step": 1903
+    },
+    {
+      "epoch": 0.23643362721966968,
+      "grad_norm": 0.2353687584400177,
+      "learning_rate": 5e-05,
+      "loss": 1.683,
+      "step": 1904
+    },
+    {
+      "epoch": 0.23655780454489012,
+      "grad_norm": 0.286986768245697,
+      "learning_rate": 5e-05,
+      "loss": 1.9561,
+      "step": 1905
+    },
+    {
+      "epoch": 0.23668198187011052,
+      "grad_norm": 0.31905364990234375,
+      "learning_rate": 5e-05,
+      "loss": 2.3166,
+      "step": 1906
+    },
+    {
+      "epoch": 0.23680615919533093,
+      "grad_norm": 0.2782406806945801,
+      "learning_rate": 5e-05,
+      "loss": 1.9338,
+      "step": 1907
+    },
+    {
+      "epoch": 0.23693033652055134,
+      "grad_norm": 0.2489187866449356,
+      "learning_rate": 5e-05,
+      "loss": 1.997,
+      "step": 1908
+    },
+    {
+      "epoch": 0.23705451384577175,
+      "grad_norm": 0.26325809955596924,
+      "learning_rate": 5e-05,
+      "loss": 2.0334,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2371786911709922,
+      "grad_norm": 0.25017687678337097,
+      "learning_rate": 5e-05,
+      "loss": 2.1734,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2373028684962126,
+      "grad_norm": 0.23639056086540222,
+      "learning_rate": 5e-05,
+      "loss": 1.7767,
+      "step": 1911
+    },
+    {
+      "epoch": 0.237427045821433,
+      "grad_norm": 0.3084493577480316,
+      "learning_rate": 5e-05,
+      "loss": 2.5107,
+      "step": 1912
+    },
+    {
+      "epoch": 0.23755122314665342,
+      "grad_norm": 0.5163066983222961,
+      "learning_rate": 5e-05,
+      "loss": 3.047,
+      "step": 1913
+    },
+    {
+      "epoch": 0.23767540047187383,
+      "grad_norm": 0.23216481506824493,
+      "learning_rate": 5e-05,
+      "loss": 1.8174,
+      "step": 1914
+    },
+    {
+      "epoch": 0.23779957779709426,
+      "grad_norm": 0.31820985674858093,
+      "learning_rate": 5e-05,
+      "loss": 2.2972,
+      "step": 1915
+    },
+    {
+      "epoch": 0.23792375512231467,
+      "grad_norm": 0.33713197708129883,
+      "learning_rate": 5e-05,
+      "loss": 1.9433,
+      "step": 1916
+    },
+    {
+      "epoch": 0.23804793244753508,
+      "grad_norm": 0.2533275783061981,
+      "learning_rate": 5e-05,
+      "loss": 1.7807,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2381721097727555,
+      "grad_norm": 0.3215118944644928,
+      "learning_rate": 5e-05,
+      "loss": 2.7627,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2382962870979759,
+      "grad_norm": 0.24559693038463593,
+      "learning_rate": 5e-05,
+      "loss": 1.7629,
+      "step": 1919
+    },
+    {
+      "epoch": 0.23842046442319634,
+      "grad_norm": 0.40229567885398865,
+      "learning_rate": 5e-05,
+      "loss": 3.0825,
+      "step": 1920
+    },
+    {
+      "epoch": 0.23854464174841675,
+      "grad_norm": 0.262498140335083,
+      "learning_rate": 5e-05,
+      "loss": 1.8635,
+      "step": 1921
+    },
+    {
+      "epoch": 0.23866881907363716,
+      "grad_norm": 0.2896023988723755,
+      "learning_rate": 5e-05,
+      "loss": 2.0284,
+      "step": 1922
+    },
+    {
+      "epoch": 0.23879299639885757,
+      "grad_norm": 0.3890548348426819,
+      "learning_rate": 5e-05,
+      "loss": 2.4002,
+      "step": 1923
+    },
+    {
+      "epoch": 0.23891717372407797,
+      "grad_norm": 0.2438245713710785,
+      "learning_rate": 5e-05,
+      "loss": 1.8078,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2390413510492984,
+      "grad_norm": 0.23711682856082916,
+      "learning_rate": 5e-05,
+      "loss": 1.7585,
+      "step": 1925
+    },
+    {
+      "epoch": 0.23916552837451882,
+      "grad_norm": 0.2526569366455078,
+      "learning_rate": 5e-05,
+      "loss": 1.8272,
+      "step": 1926
+    },
+    {
+      "epoch": 0.23928970569973923,
+      "grad_norm": 0.24494430422782898,
+      "learning_rate": 5e-05,
+      "loss": 1.5542,
+      "step": 1927
+    },
+    {
+      "epoch": 0.23941388302495964,
+      "grad_norm": 0.3084580898284912,
+      "learning_rate": 5e-05,
+      "loss": 2.2646,
+      "step": 1928
+    },
+    {
+      "epoch": 0.23953806035018005,
+      "grad_norm": 0.269167959690094,
+      "learning_rate": 5e-05,
+      "loss": 1.8957,
+      "step": 1929
+    },
+    {
+      "epoch": 0.23966223767540049,
+      "grad_norm": 0.31668105721473694,
+      "learning_rate": 5e-05,
+      "loss": 1.7582,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2397864150006209,
+      "grad_norm": 0.30181264877319336,
+      "learning_rate": 5e-05,
+      "loss": 2.2467,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2399105923258413,
+      "grad_norm": 0.36058130860328674,
+      "learning_rate": 5e-05,
+      "loss": 1.7317,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2400347696510617,
+      "grad_norm": 0.31209760904312134,
+      "learning_rate": 5e-05,
+      "loss": 2.2926,
+      "step": 1933
+    },
+    {
+      "epoch": 0.24015894697628212,
+      "grad_norm": 0.21819227933883667,
+      "learning_rate": 5e-05,
+      "loss": 1.5999,
+      "step": 1934
+    },
+    {
+      "epoch": 0.24028312430150253,
+      "grad_norm": 0.2392604500055313,
+      "learning_rate": 5e-05,
+      "loss": 1.8408,
+      "step": 1935
+    },
+    {
+      "epoch": 0.24040730162672297,
+      "grad_norm": 0.24267391860485077,
+      "learning_rate": 5e-05,
+      "loss": 1.642,
+      "step": 1936
+    },
+    {
+      "epoch": 0.24053147895194338,
+      "grad_norm": 0.24436531960964203,
+      "learning_rate": 5e-05,
+      "loss": 1.8484,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2406556562771638,
+      "grad_norm": 0.4588351845741272,
+      "learning_rate": 5e-05,
+      "loss": 2.8791,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2407798336023842,
+      "grad_norm": 0.2764546871185303,
+      "learning_rate": 5e-05,
+      "loss": 2.0667,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2409040109276046,
+      "grad_norm": 0.2699776887893677,
+      "learning_rate": 5e-05,
+      "loss": 2.1373,
+      "step": 1940
+    },
+    {
+      "epoch": 0.24102818825282504,
+      "grad_norm": 0.2740055322647095,
+      "learning_rate": 5e-05,
+      "loss": 1.812,
+      "step": 1941
+    },
+    {
+      "epoch": 0.24115236557804545,
+      "grad_norm": 0.31641992926597595,
+      "learning_rate": 5e-05,
+      "loss": 2.29,
+      "step": 1942
+    },
+    {
+      "epoch": 0.24127654290326586,
+      "grad_norm": 0.24028736352920532,
+      "learning_rate": 5e-05,
+      "loss": 1.7542,
+      "step": 1943
+    },
+    {
+      "epoch": 0.24140072022848627,
+      "grad_norm": 0.23450621962547302,
+      "learning_rate": 5e-05,
+      "loss": 1.7702,
+      "step": 1944
+    },
+    {
+      "epoch": 0.24152489755370668,
+      "grad_norm": 0.26582643389701843,
+      "learning_rate": 5e-05,
+      "loss": 1.9003,
+      "step": 1945
+    },
+    {
+      "epoch": 0.24164907487892712,
+      "grad_norm": 0.24697017669677734,
+      "learning_rate": 5e-05,
+      "loss": 1.859,
+      "step": 1946
+    },
+    {
+      "epoch": 0.24177325220414753,
+      "grad_norm": 0.2788417637348175,
+      "learning_rate": 5e-05,
+      "loss": 1.9654,
+      "step": 1947
+    },
+    {
+      "epoch": 0.24189742952936794,
+      "grad_norm": 0.2509775757789612,
+      "learning_rate": 5e-05,
+      "loss": 1.941,
+      "step": 1948
+    },
+    {
+      "epoch": 0.24202160685458834,
+      "grad_norm": 0.2366712987422943,
+      "learning_rate": 5e-05,
+      "loss": 1.8971,
+      "step": 1949
+    },
+    {
+      "epoch": 0.24214578417980875,
+      "grad_norm": 0.28300464153289795,
+      "learning_rate": 5e-05,
+      "loss": 1.9168,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2422699615050292,
+      "grad_norm": 0.374286025762558,
+      "learning_rate": 5e-05,
+      "loss": 2.3375,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2423941388302496,
+      "grad_norm": 0.27787190675735474,
+      "learning_rate": 5e-05,
+      "loss": 2.2976,
+      "step": 1952
+    },
+    {
+      "epoch": 0.24251831615547,
+      "grad_norm": 0.2881941795349121,
+      "learning_rate": 5e-05,
+      "loss": 1.9152,
+      "step": 1953
+    },
+    {
+      "epoch": 0.24264249348069042,
+      "grad_norm": 0.25375548005104065,
+      "learning_rate": 5e-05,
+      "loss": 2.069,
+      "step": 1954
+    },
+    {
+      "epoch": 0.24276667080591083,
+      "grad_norm": 0.2413804978132248,
+      "learning_rate": 5e-05,
+      "loss": 1.8972,
+      "step": 1955
+    },
+    {
+      "epoch": 0.24289084813113127,
+      "grad_norm": 0.2694176137447357,
+      "learning_rate": 5e-05,
+      "loss": 2.207,
+      "step": 1956
+    },
+    {
+      "epoch": 0.24301502545635167,
+      "grad_norm": 0.28330808877944946,
+      "learning_rate": 5e-05,
+      "loss": 2.2635,
+      "step": 1957
+    },
+    {
+      "epoch": 0.24313920278157208,
+      "grad_norm": 0.23573309183120728,
+      "learning_rate": 5e-05,
+      "loss": 1.9244,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2432633801067925,
+      "grad_norm": 0.22043515741825104,
+      "learning_rate": 5e-05,
+      "loss": 1.6197,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2433875574320129,
+      "grad_norm": 0.2412165403366089,
+      "learning_rate": 5e-05,
+      "loss": 1.9004,
+      "step": 1960
+    },
+    {
+      "epoch": 0.24351173475723334,
+      "grad_norm": 0.22970552742481232,
+      "learning_rate": 5e-05,
+      "loss": 1.8262,
+      "step": 1961
+    },
+    {
+      "epoch": 0.24363591208245375,
+      "grad_norm": 0.316812127828598,
+      "learning_rate": 5e-05,
+      "loss": 2.3856,
+      "step": 1962
+    },
+    {
+      "epoch": 0.24376008940767416,
+      "grad_norm": 0.3819718360900879,
+      "learning_rate": 5e-05,
+      "loss": 2.6623,
+      "step": 1963
+    },
+    {
+      "epoch": 0.24388426673289457,
+      "grad_norm": 0.2724602222442627,
+      "learning_rate": 5e-05,
+      "loss": 2.1579,
+      "step": 1964
+    },
+    {
+      "epoch": 0.24400844405811498,
+      "grad_norm": 0.2912684381008148,
+      "learning_rate": 5e-05,
+      "loss": 2.0478,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2441326213833354,
+      "grad_norm": 0.3085207939147949,
+      "learning_rate": 5e-05,
+      "loss": 2.2613,
+      "step": 1966
+    },
+    {
+      "epoch": 0.24425679870855582,
+      "grad_norm": 0.27749061584472656,
+      "learning_rate": 5e-05,
+      "loss": 2.8292,
+      "step": 1967
+    },
+    {
+      "epoch": 0.24438097603377623,
+      "grad_norm": 0.27682381868362427,
+      "learning_rate": 5e-05,
+      "loss": 1.9787,
+      "step": 1968
+    },
+    {
+      "epoch": 0.24450515335899664,
+      "grad_norm": 0.23448756337165833,
+      "learning_rate": 5e-05,
+      "loss": 1.8452,
+      "step": 1969
+    },
+    {
+      "epoch": 0.24462933068421705,
+      "grad_norm": 0.36228427290916443,
+      "learning_rate": 5e-05,
+      "loss": 2.1993,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2447535080094375,
+      "grad_norm": 0.2361755222082138,
+      "learning_rate": 5e-05,
+      "loss": 1.6303,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2448776853346579,
+      "grad_norm": 0.24990299344062805,
+      "learning_rate": 5e-05,
+      "loss": 1.9902,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2450018626598783,
+      "grad_norm": 0.2876966595649719,
+      "learning_rate": 5e-05,
+      "loss": 2.3467,
+      "step": 1973
+    },
+    {
+      "epoch": 0.24512603998509871,
+      "grad_norm": 0.2440340369939804,
+      "learning_rate": 5e-05,
+      "loss": 2.0496,
+      "step": 1974
+    },
+    {
+      "epoch": 0.24525021731031912,
+      "grad_norm": 0.2509459853172302,
+      "learning_rate": 5e-05,
+      "loss": 1.7841,
+      "step": 1975
+    },
+    {
+      "epoch": 0.24537439463553956,
+      "grad_norm": 0.24116043746471405,
+      "learning_rate": 5e-05,
+      "loss": 1.7636,
+      "step": 1976
+    },
+    {
+      "epoch": 0.24549857196075997,
+      "grad_norm": 0.3570502698421478,
+      "learning_rate": 5e-05,
+      "loss": 2.6231,
+      "step": 1977
+    },
+    {
+      "epoch": 0.24562274928598038,
+      "grad_norm": 0.4303056597709656,
+      "learning_rate": 5e-05,
+      "loss": 3.7122,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2457469266112008,
+      "grad_norm": 0.289449542760849,
+      "learning_rate": 5e-05,
+      "loss": 2.1098,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2458711039364212,
+      "grad_norm": 0.277208149433136,
+      "learning_rate": 5e-05,
+      "loss": 2.1407,
+      "step": 1980
+    },
+    {
+      "epoch": 0.24599528126164164,
+      "grad_norm": 0.31902116537094116,
+      "learning_rate": 5e-05,
+      "loss": 2.1891,
+      "step": 1981
+    },
+    {
+      "epoch": 0.24611945858686204,
+      "grad_norm": 0.2405809611082077,
+      "learning_rate": 5e-05,
+      "loss": 1.6028,
+      "step": 1982
+    },
+    {
+      "epoch": 0.24624363591208245,
+      "grad_norm": 0.34350183606147766,
+      "learning_rate": 5e-05,
+      "loss": 2.2372,
+      "step": 1983
+    },
+    {
+      "epoch": 0.24636781323730286,
+      "grad_norm": 0.25604358315467834,
+      "learning_rate": 5e-05,
+      "loss": 1.9088,
+      "step": 1984
+    },
+    {
+      "epoch": 0.24649199056252327,
+      "grad_norm": 0.2582472860813141,
+      "learning_rate": 5e-05,
+      "loss": 1.9506,
+      "step": 1985
+    },
+    {
+      "epoch": 0.2466161678877437,
+      "grad_norm": 0.42406120896339417,
+      "learning_rate": 5e-05,
+      "loss": 2.293,
+      "step": 1986
+    },
+    {
+      "epoch": 0.24674034521296412,
+      "grad_norm": 0.3463503122329712,
+      "learning_rate": 5e-05,
+      "loss": 2.605,
+      "step": 1987
+    },
+    {
+      "epoch": 0.24686452253818453,
+      "grad_norm": 0.31971120834350586,
+      "learning_rate": 5e-05,
+      "loss": 2.3052,
+      "step": 1988
+    },
+    {
+      "epoch": 0.24698869986340494,
+      "grad_norm": 0.34543928503990173,
+      "learning_rate": 5e-05,
+      "loss": 2.2367,
+      "step": 1989
+    },
+    {
+      "epoch": 0.24711287718862535,
+      "grad_norm": 0.2506909668445587,
+      "learning_rate": 5e-05,
+      "loss": 1.8307,
+      "step": 1990
+    },
+    {
+      "epoch": 0.24723705451384578,
+      "grad_norm": 0.2659740149974823,
+      "learning_rate": 5e-05,
+      "loss": 2.0522,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2473612318390662,
+      "grad_norm": 0.22190289199352264,
+      "learning_rate": 5e-05,
+      "loss": 1.653,
+      "step": 1992
+    },
+    {
+      "epoch": 0.2474854091642866,
+      "grad_norm": 0.24099688231945038,
+      "learning_rate": 5e-05,
+      "loss": 1.9008,
+      "step": 1993
+    },
+    {
+      "epoch": 0.247609586489507,
+      "grad_norm": 0.26779383420944214,
+      "learning_rate": 5e-05,
+      "loss": 2.0616,
+      "step": 1994
+    },
+    {
+      "epoch": 0.24773376381472742,
+      "grad_norm": 0.2511522173881531,
+      "learning_rate": 5e-05,
+      "loss": 1.9541,
+      "step": 1995
+    },
+    {
+      "epoch": 0.24785794113994786,
+      "grad_norm": 0.4537486433982849,
+      "learning_rate": 5e-05,
+      "loss": 2.7368,
+      "step": 1996
+    },
+    {
+      "epoch": 0.24798211846516827,
+      "grad_norm": 0.2445574849843979,
+      "learning_rate": 5e-05,
+      "loss": 1.7308,
+      "step": 1997
+    },
+    {
+      "epoch": 0.24810629579038868,
+      "grad_norm": 0.2323279082775116,
+      "learning_rate": 5e-05,
+      "loss": 1.9362,
+      "step": 1998
+    },
+    {
+      "epoch": 0.24823047311560908,
+      "grad_norm": 0.24628306925296783,
+      "learning_rate": 5e-05,
+      "loss": 2.1236,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2483546504408295,
+      "grad_norm": 0.27042755484580994,
+      "learning_rate": 5e-05,
+      "loss": 2.166,
+      "step": 2000
+    },
+    {
+      "epoch": 0.24847882776604993,
+      "grad_norm": 0.27119892835617065,
+      "learning_rate": 5e-05,
+      "loss": 2.0992,
+      "step": 2001
+    },
+    {
+      "epoch": 0.24860300509127034,
+      "grad_norm": 0.2787729799747467,
+      "learning_rate": 5e-05,
+      "loss": 2.1958,
+      "step": 2002
+    },
+    {
+      "epoch": 0.24872718241649075,
+      "grad_norm": 0.24131381511688232,
+      "learning_rate": 5e-05,
+      "loss": 1.9194,
+      "step": 2003
+    },
+    {
+      "epoch": 0.24885135974171116,
+      "grad_norm": 0.2580811083316803,
+      "learning_rate": 5e-05,
+      "loss": 2.0621,
+      "step": 2004
+    },
+    {
+      "epoch": 0.24897553706693157,
+      "grad_norm": 0.2595440745353699,
+      "learning_rate": 5e-05,
+      "loss": 2.125,
+      "step": 2005
+    },
+    {
+      "epoch": 0.249099714392152,
+      "grad_norm": 0.2212187796831131,
+      "learning_rate": 5e-05,
+      "loss": 1.7434,
+      "step": 2006
+    },
+    {
+      "epoch": 0.24922389171737241,
+      "grad_norm": 0.25188785791397095,
+      "learning_rate": 5e-05,
+      "loss": 1.5035,
+      "step": 2007
+    },
+    {
+      "epoch": 0.24934806904259282,
+      "grad_norm": 0.40886369347572327,
+      "learning_rate": 5e-05,
+      "loss": 2.681,
+      "step": 2008
+    },
+    {
+      "epoch": 0.24947224636781323,
+      "grad_norm": 0.25674906373023987,
+      "learning_rate": 5e-05,
+      "loss": 1.9956,
+      "step": 2009
+    },
+    {
+      "epoch": 0.24959642369303364,
+      "grad_norm": 0.22841785848140717,
+      "learning_rate": 5e-05,
+      "loss": 1.7359,
+      "step": 2010
+    },
+    {
+      "epoch": 0.24972060101825408,
+      "grad_norm": 0.22568254172801971,
+      "learning_rate": 5e-05,
+      "loss": 1.7335,
+      "step": 2011
+    },
+    {
+      "epoch": 0.2498447783434745,
+      "grad_norm": 0.3849019408226013,
+      "learning_rate": 5e-05,
+      "loss": 2.8411,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2499689556686949,
+      "grad_norm": 0.3539924919605255,
+      "learning_rate": 5e-05,
+      "loss": 1.9334,
+      "step": 2013
+    },
+    {
+      "epoch": 0.25009313299391533,
+      "grad_norm": 0.31630808115005493,
+      "learning_rate": 5e-05,
+      "loss": 2.3065,
+      "step": 2014
+    },
+    {
+      "epoch": 0.25021731031913574,
+      "grad_norm": 0.23726125061511993,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 2015
+    },
+    {
+      "epoch": 0.25034148764435615,
+      "grad_norm": 0.24230054020881653,
+      "learning_rate": 5e-05,
+      "loss": 1.5639,
+      "step": 2016
+    },
+    {
+      "epoch": 0.25046566496957656,
+      "grad_norm": 0.24119439721107483,
+      "learning_rate": 5e-05,
+      "loss": 1.7292,
+      "step": 2017
+    },
+    {
+      "epoch": 0.25058984229479697,
+      "grad_norm": 0.31925609707832336,
+      "learning_rate": 5e-05,
+      "loss": 2.2247,
+      "step": 2018
+    },
+    {
+      "epoch": 0.2507140196200174,
+      "grad_norm": 0.2717881202697754,
+      "learning_rate": 5e-05,
+      "loss": 2.1405,
+      "step": 2019
+    },
+    {
+      "epoch": 0.2508381969452378,
+      "grad_norm": 0.23317092657089233,
+      "learning_rate": 5e-05,
+      "loss": 1.8834,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2509623742704582,
+      "grad_norm": 0.2441297322511673,
+      "learning_rate": 5e-05,
+      "loss": 2.0862,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2510865515956786,
+      "grad_norm": 0.23900364339351654,
+      "learning_rate": 5e-05,
+      "loss": 1.8199,
+      "step": 2022
+    },
+    {
+      "epoch": 0.251210728920899,
+      "grad_norm": 0.2367074191570282,
+      "learning_rate": 5e-05,
+      "loss": 1.9068,
+      "step": 2023
+    },
+    {
+      "epoch": 0.2513349062461195,
+      "grad_norm": 0.24488438665866852,
+      "learning_rate": 5e-05,
+      "loss": 1.926,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2514590835713399,
+      "grad_norm": 0.22677388787269592,
+      "learning_rate": 5e-05,
+      "loss": 1.7542,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2515832608965603,
+      "grad_norm": 0.5012796521186829,
+      "learning_rate": 5e-05,
+      "loss": 3.3974,
+      "step": 2026
+    },
+    {
+      "epoch": 0.2517074382217807,
+      "grad_norm": 0.2840482294559479,
+      "learning_rate": 5e-05,
+      "loss": 2.5031,
+      "step": 2027
+    },
+    {
+      "epoch": 0.2518316155470011,
+      "grad_norm": 0.25313377380371094,
+      "learning_rate": 5e-05,
+      "loss": 1.9694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.25195579287222153,
+      "grad_norm": 0.28684139251708984,
+      "learning_rate": 5e-05,
+      "loss": 1.8056,
+      "step": 2029
+    },
+    {
+      "epoch": 0.25207997019744194,
+      "grad_norm": 0.33107808232307434,
+      "learning_rate": 5e-05,
+      "loss": 2.1542,
+      "step": 2030
+    },
+    {
+      "epoch": 0.25220414752266235,
+      "grad_norm": 0.2357482761144638,
+      "learning_rate": 5e-05,
+      "loss": 1.7588,
+      "step": 2031
+    },
+    {
+      "epoch": 0.25232832484788276,
+      "grad_norm": 0.2528238892555237,
+      "learning_rate": 5e-05,
+      "loss": 2.1084,
+      "step": 2032
+    },
+    {
+      "epoch": 0.25245250217310317,
+      "grad_norm": 0.4784279763698578,
+      "learning_rate": 5e-05,
+      "loss": 2.6304,
+      "step": 2033
+    },
+    {
+      "epoch": 0.25257667949832363,
+      "grad_norm": 0.2978458106517792,
+      "learning_rate": 5e-05,
+      "loss": 2.3928,
+      "step": 2034
+    },
+    {
+      "epoch": 0.25270085682354404,
+      "grad_norm": 0.3612816333770752,
+      "learning_rate": 5e-05,
+      "loss": 2.7444,
+      "step": 2035
+    },
+    {
+      "epoch": 0.25282503414876445,
+      "grad_norm": 0.2545587420463562,
+      "learning_rate": 5e-05,
+      "loss": 1.9201,
+      "step": 2036
+    },
+    {
+      "epoch": 0.25294921147398486,
+      "grad_norm": 0.306565523147583,
+      "learning_rate": 5e-05,
+      "loss": 2.3076,
+      "step": 2037
+    },
+    {
+      "epoch": 0.25307338879920527,
+      "grad_norm": 0.2631712555885315,
+      "learning_rate": 5e-05,
+      "loss": 1.6876,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2531975661244257,
+      "grad_norm": 0.2829129695892334,
+      "learning_rate": 5e-05,
+      "loss": 2.2984,
+      "step": 2039
+    },
+    {
+      "epoch": 0.2533217434496461,
+      "grad_norm": 0.240200012922287,
+      "learning_rate": 5e-05,
+      "loss": 1.6346,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2534459207748665,
+      "grad_norm": 0.25522780418395996,
+      "learning_rate": 5e-05,
+      "loss": 1.794,
+      "step": 2041
+    },
+    {
+      "epoch": 0.2535700981000869,
+      "grad_norm": 0.24453885853290558,
+      "learning_rate": 5e-05,
+      "loss": 1.8446,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2536942754253073,
+      "grad_norm": 0.24798159301280975,
+      "learning_rate": 5e-05,
+      "loss": 1.9012,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2538184527505278,
+      "grad_norm": 0.24530170857906342,
+      "learning_rate": 5e-05,
+      "loss": 1.823,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2539426300757482,
+      "grad_norm": 0.23647554218769073,
+      "learning_rate": 5e-05,
+      "loss": 1.7393,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2540668074009686,
+      "grad_norm": 0.24002909660339355,
+      "learning_rate": 5e-05,
+      "loss": 2.0722,
+      "step": 2046
+    },
+    {
+      "epoch": 0.254190984726189,
+      "grad_norm": 0.24300985038280487,
+      "learning_rate": 5e-05,
+      "loss": 1.6658,
+      "step": 2047
+    },
+    {
+      "epoch": 0.2543151620514094,
+      "grad_norm": 0.24109484255313873,
+      "learning_rate": 5e-05,
+      "loss": 2.0062,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2544393393766298,
+      "grad_norm": 0.5002878904342651,
+      "learning_rate": 5e-05,
+      "loss": 2.9496,
+      "step": 2049
+    },
+    {
+      "epoch": 0.25456351670185023,
+      "grad_norm": 0.2309587597846985,
+      "learning_rate": 5e-05,
+      "loss": 1.8569,
+      "step": 2050
+    },
+    {
+      "epoch": 0.25468769402707064,
+      "grad_norm": 0.2572423219680786,
+      "learning_rate": 5e-05,
+      "loss": 1.8463,
+      "step": 2051
+    },
+    {
+      "epoch": 0.25481187135229105,
+      "grad_norm": 0.2645682692527771,
+      "learning_rate": 5e-05,
+      "loss": 1.9417,
+      "step": 2052
+    },
+    {
+      "epoch": 0.25493604867751146,
+      "grad_norm": 0.25839585065841675,
+      "learning_rate": 5e-05,
+      "loss": 1.7537,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2550602260027319,
+      "grad_norm": 0.33676034212112427,
+      "learning_rate": 5e-05,
+      "loss": 1.8469,
+      "step": 2054
+    },
+    {
+      "epoch": 0.25518440332795234,
+      "grad_norm": 0.3969810903072357,
+      "learning_rate": 5e-05,
+      "loss": 2.8551,
+      "step": 2055
+    },
+    {
+      "epoch": 0.25530858065317275,
+      "grad_norm": 0.216770738363266,
+      "learning_rate": 5e-05,
+      "loss": 1.6182,
+      "step": 2056
+    },
+    {
+      "epoch": 0.25543275797839315,
+      "grad_norm": 0.2686668038368225,
+      "learning_rate": 5e-05,
+      "loss": 2.061,
+      "step": 2057
+    },
+    {
+      "epoch": 0.25555693530361356,
+      "grad_norm": 0.26434871554374695,
+      "learning_rate": 5e-05,
+      "loss": 1.9185,
+      "step": 2058
+    },
+    {
+      "epoch": 0.255681112628834,
+      "grad_norm": 0.4182075262069702,
+      "learning_rate": 5e-05,
+      "loss": 2.5353,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2558052899540544,
+      "grad_norm": 0.3168865144252777,
+      "learning_rate": 5e-05,
+      "loss": 2.292,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2559294672792748,
+      "grad_norm": 0.22851121425628662,
+      "learning_rate": 5e-05,
+      "loss": 1.8511,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2560536446044952,
+      "grad_norm": 0.28795769810676575,
+      "learning_rate": 5e-05,
+      "loss": 2.5227,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2561778219297156,
+      "grad_norm": 0.26309627294540405,
+      "learning_rate": 5e-05,
+      "loss": 1.9425,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2563019992549361,
+      "grad_norm": 0.2665860056877136,
+      "learning_rate": 5e-05,
+      "loss": 1.8391,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2564261765801565,
+      "grad_norm": 0.25377437472343445,
+      "learning_rate": 5e-05,
+      "loss": 1.8914,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2565503539053769,
+      "grad_norm": 0.31546226143836975,
+      "learning_rate": 5e-05,
+      "loss": 2.2622,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2566745312305973,
+      "grad_norm": 0.33930104970932007,
+      "learning_rate": 5e-05,
+      "loss": 2.477,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2567987085558177,
+      "grad_norm": 0.252470463514328,
+      "learning_rate": 5e-05,
+      "loss": 1.9096,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2569228858810381,
+      "grad_norm": 0.3332885503768921,
+      "learning_rate": 5e-05,
+      "loss": 2.9871,
+      "step": 2069
+    },
+    {
+      "epoch": 0.25704706320625853,
+      "grad_norm": 0.24781787395477295,
+      "learning_rate": 5e-05,
+      "loss": 2.0441,
+      "step": 2070
+    },
+    {
+      "epoch": 0.25717124053147894,
+      "grad_norm": 0.35084837675094604,
+      "learning_rate": 5e-05,
+      "loss": 2.3459,
+      "step": 2071
+    },
+    {
+      "epoch": 0.25729541785669935,
+      "grad_norm": 0.2222728729248047,
+      "learning_rate": 5e-05,
+      "loss": 1.7691,
+      "step": 2072
+    },
+    {
+      "epoch": 0.25741959518191976,
+      "grad_norm": 0.24346236884593964,
+      "learning_rate": 5e-05,
+      "loss": 1.7336,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2575437725071402,
+      "grad_norm": 0.3183536231517792,
+      "learning_rate": 5e-05,
+      "loss": 2.4286,
+      "step": 2074
+    },
+    {
+      "epoch": 0.25766794983236063,
+      "grad_norm": 0.2763809561729431,
+      "learning_rate": 5e-05,
+      "loss": 2.4786,
+      "step": 2075
+    },
+    {
+      "epoch": 0.25779212715758104,
+      "grad_norm": 0.2623912990093231,
+      "learning_rate": 5e-05,
+      "loss": 2.1623,
+      "step": 2076
+    },
+    {
+      "epoch": 0.25791630448280145,
+      "grad_norm": 0.256130188703537,
+      "learning_rate": 5e-05,
+      "loss": 2.0593,
+      "step": 2077
+    },
+    {
+      "epoch": 0.25804048180802186,
+      "grad_norm": 0.28890419006347656,
+      "learning_rate": 5e-05,
+      "loss": 2.4306,
+      "step": 2078
+    },
+    {
+      "epoch": 0.25816465913324227,
+      "grad_norm": 0.2251637578010559,
+      "learning_rate": 5e-05,
+      "loss": 1.9284,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2582888364584627,
+      "grad_norm": 0.28757646679878235,
+      "learning_rate": 5e-05,
+      "loss": 2.3646,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2584130137836831,
+      "grad_norm": 0.23034705221652985,
+      "learning_rate": 5e-05,
+      "loss": 1.6923,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2585371911089035,
+      "grad_norm": 0.2980548143386841,
+      "learning_rate": 5e-05,
+      "loss": 2.1013,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2586613684341239,
+      "grad_norm": 0.246156707406044,
+      "learning_rate": 5e-05,
+      "loss": 1.9474,
+      "step": 2083
+    },
+    {
+      "epoch": 0.25878554575934437,
+      "grad_norm": 0.2270069569349289,
+      "learning_rate": 5e-05,
+      "loss": 1.739,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2589097230845648,
+      "grad_norm": 0.39138659834861755,
+      "learning_rate": 5e-05,
+      "loss": 2.8917,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2590339004097852,
+      "grad_norm": 0.3135612905025482,
+      "learning_rate": 5e-05,
+      "loss": 2.0143,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2591580777350056,
+      "grad_norm": 0.2569997012615204,
+      "learning_rate": 5e-05,
+      "loss": 1.7597,
+      "step": 2087
+    },
+    {
+      "epoch": 0.259282255060226,
+      "grad_norm": 0.304267555475235,
+      "learning_rate": 5e-05,
+      "loss": 2.5893,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2594064323854464,
+      "grad_norm": 0.2569121718406677,
+      "learning_rate": 5e-05,
+      "loss": 2.1813,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2595306097106668,
+      "grad_norm": 0.28129658102989197,
+      "learning_rate": 5e-05,
+      "loss": 2.2705,
+      "step": 2090
+    },
+    {
+      "epoch": 0.25965478703588724,
+      "grad_norm": 0.258292555809021,
+      "learning_rate": 5e-05,
+      "loss": 2.032,
+      "step": 2091
+    },
+    {
+      "epoch": 0.25977896436110764,
+      "grad_norm": 0.26310989260673523,
+      "learning_rate": 5e-05,
+      "loss": 2.0423,
+      "step": 2092
+    },
+    {
+      "epoch": 0.25990314168632805,
+      "grad_norm": 0.23579686880111694,
+      "learning_rate": 5e-05,
+      "loss": 1.8162,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2600273190115485,
+      "grad_norm": 0.232418492436409,
+      "learning_rate": 5e-05,
+      "loss": 1.7722,
+      "step": 2094
+    },
+    {
+      "epoch": 0.26015149633676893,
+      "grad_norm": 0.28086546063423157,
+      "learning_rate": 5e-05,
+      "loss": 2.4549,
+      "step": 2095
+    },
+    {
+      "epoch": 0.26027567366198934,
+      "grad_norm": 0.25034159421920776,
+      "learning_rate": 5e-05,
+      "loss": 2.0505,
+      "step": 2096
+    },
+    {
+      "epoch": 0.26039985098720975,
+      "grad_norm": 0.228069469332695,
+      "learning_rate": 5e-05,
+      "loss": 1.6012,
+      "step": 2097
+    },
+    {
+      "epoch": 0.26052402831243016,
+      "grad_norm": 0.22270333766937256,
+      "learning_rate": 5e-05,
+      "loss": 1.8579,
+      "step": 2098
+    },
+    {
+      "epoch": 0.26064820563765057,
+      "grad_norm": 0.23570704460144043,
+      "learning_rate": 5e-05,
+      "loss": 1.9557,
+      "step": 2099
+    },
+    {
+      "epoch": 0.260772382962871,
+      "grad_norm": 0.295588880777359,
+      "learning_rate": 5e-05,
+      "loss": 2.0888,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2608965602880914,
+      "grad_norm": 0.22561363875865936,
+      "learning_rate": 5e-05,
+      "loss": 1.9189,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2610207376133118,
+      "grad_norm": 0.4077657163143158,
+      "learning_rate": 5e-05,
+      "loss": 2.7744,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2611449149385322,
+      "grad_norm": 0.29966095089912415,
+      "learning_rate": 5e-05,
+      "loss": 2.1181,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2612690922637526,
+      "grad_norm": 0.27679991722106934,
+      "learning_rate": 5e-05,
+      "loss": 2.1859,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2613932695889731,
+      "grad_norm": 0.23642289638519287,
+      "learning_rate": 5e-05,
+      "loss": 1.9934,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2615174469141935,
+      "grad_norm": 0.2690558135509491,
+      "learning_rate": 5e-05,
+      "loss": 2.2696,
+      "step": 2106
+    },
+    {
+      "epoch": 0.2616416242394139,
+      "grad_norm": 0.27591317892074585,
+      "learning_rate": 5e-05,
+      "loss": 2.0716,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2617658015646343,
+      "grad_norm": 0.32620933651924133,
+      "learning_rate": 5e-05,
+      "loss": 2.7693,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2618899788898547,
+      "grad_norm": 0.3226400911808014,
+      "learning_rate": 5e-05,
+      "loss": 2.2325,
+      "step": 2109
+    },
+    {
+      "epoch": 0.2620141562150751,
+      "grad_norm": 0.22471946477890015,
+      "learning_rate": 5e-05,
+      "loss": 1.5835,
+      "step": 2110
+    },
+    {
+      "epoch": 0.26213833354029553,
+      "grad_norm": 0.21536286175251007,
+      "learning_rate": 5e-05,
+      "loss": 1.756,
+      "step": 2111
+    },
+    {
+      "epoch": 0.26226251086551594,
+      "grad_norm": 0.2856312692165375,
+      "learning_rate": 5e-05,
+      "loss": 2.6498,
+      "step": 2112
+    },
+    {
+      "epoch": 0.26238668819073635,
+      "grad_norm": 0.29251551628112793,
+      "learning_rate": 5e-05,
+      "loss": 2.4083,
+      "step": 2113
+    },
+    {
+      "epoch": 0.26251086551595676,
+      "grad_norm": 0.25755417346954346,
+      "learning_rate": 5e-05,
+      "loss": 1.999,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2626350428411772,
+      "grad_norm": 0.24057379364967346,
+      "learning_rate": 5e-05,
+      "loss": 2.0975,
+      "step": 2115
+    },
+    {
+      "epoch": 0.26275922016639763,
+      "grad_norm": 0.25699663162231445,
+      "learning_rate": 5e-05,
+      "loss": 1.7909,
+      "step": 2116
+    },
+    {
+      "epoch": 0.26288339749161804,
+      "grad_norm": 0.2828010022640228,
+      "learning_rate": 5e-05,
+      "loss": 2.0109,
+      "step": 2117
+    },
+    {
+      "epoch": 0.26300757481683845,
+      "grad_norm": 0.2292487770318985,
+      "learning_rate": 5e-05,
+      "loss": 1.9333,
+      "step": 2118
+    },
+    {
+      "epoch": 0.26313175214205886,
+      "grad_norm": 0.34071868658065796,
+      "learning_rate": 5e-05,
+      "loss": 2.4698,
+      "step": 2119
+    },
+    {
+      "epoch": 0.26325592946727927,
+      "grad_norm": 0.22953563928604126,
+      "learning_rate": 5e-05,
+      "loss": 2.0231,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2633801067924997,
+      "grad_norm": 0.677270233631134,
+      "learning_rate": 5e-05,
+      "loss": 3.1947,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2635042841177201,
+      "grad_norm": 0.22566376626491547,
+      "learning_rate": 5e-05,
+      "loss": 1.5957,
+      "step": 2122
+    },
+    {
+      "epoch": 0.2636284614429405,
+      "grad_norm": 0.28488659858703613,
+      "learning_rate": 5e-05,
+      "loss": 2.0535,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2637526387681609,
+      "grad_norm": 0.25570377707481384,
+      "learning_rate": 5e-05,
+      "loss": 1.8684,
+      "step": 2124
+    },
+    {
+      "epoch": 0.26387681609338137,
+      "grad_norm": 0.23350538313388824,
+      "learning_rate": 5e-05,
+      "loss": 1.6609,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2640009934186018,
+      "grad_norm": 0.29248863458633423,
+      "learning_rate": 5e-05,
+      "loss": 2.0525,
+      "step": 2126
+    },
+    {
+      "epoch": 0.2641251707438222,
+      "grad_norm": 0.2242303192615509,
+      "learning_rate": 5e-05,
+      "loss": 1.6695,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2642493480690426,
+      "grad_norm": 0.2562909722328186,
+      "learning_rate": 5e-05,
+      "loss": 2.037,
+      "step": 2128
+    },
+    {
+      "epoch": 0.264373525394263,
+      "grad_norm": 0.26936212182044983,
+      "learning_rate": 5e-05,
+      "loss": 1.7756,
+      "step": 2129
+    },
+    {
+      "epoch": 0.2644977027194834,
+      "grad_norm": 0.25242146849632263,
+      "learning_rate": 5e-05,
+      "loss": 1.8722,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2646218800447038,
+      "grad_norm": 0.3085550367832184,
+      "learning_rate": 5e-05,
+      "loss": 1.7277,
+      "step": 2131
+    },
+    {
+      "epoch": 0.26474605736992424,
+      "grad_norm": 0.24068817496299744,
+      "learning_rate": 5e-05,
+      "loss": 1.9189,
+      "step": 2132
+    },
+    {
+      "epoch": 0.26487023469514465,
+      "grad_norm": 0.25240883231163025,
+      "learning_rate": 5e-05,
+      "loss": 2.131,
+      "step": 2133
+    },
+    {
+      "epoch": 0.26499441202036506,
+      "grad_norm": 0.251799613237381,
+      "learning_rate": 5e-05,
+      "loss": 1.8389,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2651185893455855,
+      "grad_norm": 0.24193397164344788,
+      "learning_rate": 5e-05,
+      "loss": 2.0534,
+      "step": 2135
+    },
+    {
+      "epoch": 0.26524276667080593,
+      "grad_norm": 0.3370469808578491,
+      "learning_rate": 5e-05,
+      "loss": 2.8748,
+      "step": 2136
+    },
+    {
+      "epoch": 0.26536694399602634,
+      "grad_norm": 0.2501845061779022,
+      "learning_rate": 5e-05,
+      "loss": 1.7923,
+      "step": 2137
+    },
+    {
+      "epoch": 0.26549112132124675,
+      "grad_norm": 0.2880485951900482,
+      "learning_rate": 5e-05,
+      "loss": 2.0938,
+      "step": 2138
+    },
+    {
+      "epoch": 0.26561529864646716,
+      "grad_norm": 0.23581233620643616,
+      "learning_rate": 5e-05,
+      "loss": 1.6483,
+      "step": 2139
+    },
+    {
+      "epoch": 0.26573947597168757,
+      "grad_norm": 0.23065868020057678,
+      "learning_rate": 5e-05,
+      "loss": 1.8004,
+      "step": 2140
+    },
+    {
+      "epoch": 0.265863653296908,
+      "grad_norm": 0.23730231821537018,
+      "learning_rate": 5e-05,
+      "loss": 1.8783,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2659878306221284,
+      "grad_norm": 0.2595699429512024,
+      "learning_rate": 5e-05,
+      "loss": 1.9,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2661120079473488,
+      "grad_norm": 0.2780103385448456,
+      "learning_rate": 5e-05,
+      "loss": 2.164,
+      "step": 2143
+    },
+    {
+      "epoch": 0.2662361852725692,
+      "grad_norm": 0.2823382616043091,
+      "learning_rate": 5e-05,
+      "loss": 2.0296,
+      "step": 2144
+    },
+    {
+      "epoch": 0.26636036259778967,
+      "grad_norm": 0.2872154414653778,
+      "learning_rate": 5e-05,
+      "loss": 2.3025,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2664845399230101,
+      "grad_norm": 0.24791646003723145,
+      "learning_rate": 5e-05,
+      "loss": 2.031,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2666087172482305,
+      "grad_norm": 0.24226826429367065,
+      "learning_rate": 5e-05,
+      "loss": 1.921,
+      "step": 2147
+    },
+    {
+      "epoch": 0.2667328945734509,
+      "grad_norm": 0.3547041416168213,
+      "learning_rate": 5e-05,
+      "loss": 2.3231,
+      "step": 2148
+    },
+    {
+      "epoch": 0.2668570718986713,
+      "grad_norm": 0.25479254126548767,
+      "learning_rate": 5e-05,
+      "loss": 2.0199,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2669812492238917,
+      "grad_norm": 0.29904577136039734,
+      "learning_rate": 5e-05,
+      "loss": 2.6155,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2671054265491121,
+      "grad_norm": 0.4940663278102875,
+      "learning_rate": 5e-05,
+      "loss": 3.405,
+      "step": 2151
+    },
+    {
+      "epoch": 0.26722960387433253,
+      "grad_norm": 0.3255084455013275,
+      "learning_rate": 5e-05,
+      "loss": 2.8758,
+      "step": 2152
+    },
+    {
+      "epoch": 0.26735378119955294,
+      "grad_norm": 0.27092376351356506,
+      "learning_rate": 5e-05,
+      "loss": 2.2788,
+      "step": 2153
+    },
+    {
+      "epoch": 0.26747795852477335,
+      "grad_norm": 0.26672711968421936,
+      "learning_rate": 5e-05,
+      "loss": 2.2986,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2676021358499938,
+      "grad_norm": 0.24492858350276947,
+      "learning_rate": 5e-05,
+      "loss": 2.0853,
+      "step": 2155
+    },
+    {
+      "epoch": 0.2677263131752142,
+      "grad_norm": 0.24383944272994995,
+      "learning_rate": 5e-05,
+      "loss": 1.8262,
+      "step": 2156
+    },
+    {
+      "epoch": 0.26785049050043463,
+      "grad_norm": 0.2774048447608948,
+      "learning_rate": 5e-05,
+      "loss": 2.0749,
+      "step": 2157
+    },
+    {
+      "epoch": 0.26797466782565504,
+      "grad_norm": 0.2561037242412567,
+      "learning_rate": 5e-05,
+      "loss": 1.9176,
+      "step": 2158
+    },
+    {
+      "epoch": 0.26809884515087545,
+      "grad_norm": 0.24214661121368408,
+      "learning_rate": 5e-05,
+      "loss": 1.8382,
+      "step": 2159
+    },
+    {
+      "epoch": 0.26822302247609586,
+      "grad_norm": 0.24293182790279388,
+      "learning_rate": 5e-05,
+      "loss": 1.9024,
+      "step": 2160
+    },
+    {
+      "epoch": 0.26834719980131627,
+      "grad_norm": 0.31856676936149597,
+      "learning_rate": 5e-05,
+      "loss": 2.7667,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2684713771265367,
+      "grad_norm": 0.24666784703731537,
+      "learning_rate": 5e-05,
+      "loss": 1.6426,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2685955544517571,
+      "grad_norm": 0.2479521483182907,
+      "learning_rate": 5e-05,
+      "loss": 1.9749,
+      "step": 2163
+    },
+    {
+      "epoch": 0.2687197317769775,
+      "grad_norm": 0.2446976900100708,
+      "learning_rate": 5e-05,
+      "loss": 1.8633,
+      "step": 2164
+    },
+    {
+      "epoch": 0.26884390910219796,
+      "grad_norm": 0.22923824191093445,
+      "learning_rate": 5e-05,
+      "loss": 1.9313,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2689680864274184,
+      "grad_norm": 0.24432528018951416,
+      "learning_rate": 5e-05,
+      "loss": 2.0022,
+      "step": 2166
+    },
+    {
+      "epoch": 0.2690922637526388,
+      "grad_norm": 0.23449097573757172,
+      "learning_rate": 5e-05,
+      "loss": 1.9619,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2692164410778592,
+      "grad_norm": 0.32519522309303284,
+      "learning_rate": 5e-05,
+      "loss": 2.0974,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2693406184030796,
+      "grad_norm": 0.27445507049560547,
+      "learning_rate": 5e-05,
+      "loss": 2.3887,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2694647957283,
+      "grad_norm": 0.24460965394973755,
+      "learning_rate": 5e-05,
+      "loss": 1.8538,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2695889730535204,
+      "grad_norm": 0.22980999946594238,
+      "learning_rate": 5e-05,
+      "loss": 1.7126,
+      "step": 2171
+    },
+    {
+      "epoch": 0.26971315037874083,
+      "grad_norm": 0.3191908001899719,
+      "learning_rate": 5e-05,
+      "loss": 2.2159,
+      "step": 2172
+    },
+    {
+      "epoch": 0.26983732770396124,
+      "grad_norm": 0.22930824756622314,
+      "learning_rate": 5e-05,
+      "loss": 1.8426,
+      "step": 2173
+    },
+    {
+      "epoch": 0.26996150502918165,
+      "grad_norm": 0.23960471153259277,
+      "learning_rate": 5e-05,
+      "loss": 1.8925,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2700856823544021,
+      "grad_norm": 0.22068379819393158,
+      "learning_rate": 5e-05,
+      "loss": 1.7616,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2702098596796225,
+      "grad_norm": 0.3168589770793915,
+      "learning_rate": 5e-05,
+      "loss": 2.4608,
+      "step": 2176
+    },
+    {
+      "epoch": 0.27033403700484293,
+      "grad_norm": 0.24010834097862244,
+      "learning_rate": 5e-05,
+      "loss": 2.0371,
+      "step": 2177
+    },
+    {
+      "epoch": 0.27045821433006334,
+      "grad_norm": 0.22494138777256012,
+      "learning_rate": 5e-05,
+      "loss": 1.7463,
+      "step": 2178
+    },
+    {
+      "epoch": 0.27058239165528375,
+      "grad_norm": 0.3759515881538391,
+      "learning_rate": 5e-05,
+      "loss": 2.3344,
+      "step": 2179
+    },
+    {
+      "epoch": 0.27070656898050416,
+      "grad_norm": 0.25227877497673035,
+      "learning_rate": 5e-05,
+      "loss": 1.8971,
+      "step": 2180
+    },
+    {
+      "epoch": 0.27083074630572457,
+      "grad_norm": 0.264453262090683,
+      "learning_rate": 5e-05,
+      "loss": 2.1299,
+      "step": 2181
+    },
+    {
+      "epoch": 0.270954923630945,
+      "grad_norm": 0.29659131169319153,
+      "learning_rate": 5e-05,
+      "loss": 1.9368,
+      "step": 2182
+    },
+    {
+      "epoch": 0.2710791009561654,
+      "grad_norm": 0.23890623450279236,
+      "learning_rate": 5e-05,
+      "loss": 2.016,
+      "step": 2183
+    },
+    {
+      "epoch": 0.2712032782813858,
+      "grad_norm": 0.35334932804107666,
+      "learning_rate": 5e-05,
+      "loss": 2.1166,
+      "step": 2184
+    },
+    {
+      "epoch": 0.27132745560660626,
+      "grad_norm": 0.23995842039585114,
+      "learning_rate": 5e-05,
+      "loss": 1.9291,
+      "step": 2185
+    },
+    {
+      "epoch": 0.27145163293182667,
+      "grad_norm": 0.3405332565307617,
+      "learning_rate": 5e-05,
+      "loss": 2.3928,
+      "step": 2186
+    },
+    {
+      "epoch": 0.2715758102570471,
+      "grad_norm": 0.29112547636032104,
+      "learning_rate": 5e-05,
+      "loss": 1.9831,
+      "step": 2187
+    },
+    {
+      "epoch": 0.2716999875822675,
+      "grad_norm": 0.3502449095249176,
+      "learning_rate": 5e-05,
+      "loss": 2.6041,
+      "step": 2188
+    },
+    {
+      "epoch": 0.2718241649074879,
+      "grad_norm": 0.2677818536758423,
+      "learning_rate": 5e-05,
+      "loss": 1.601,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2719483422327083,
+      "grad_norm": 0.24154359102249146,
+      "learning_rate": 5e-05,
+      "loss": 1.8655,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2720725195579287,
+      "grad_norm": 0.2517216205596924,
+      "learning_rate": 5e-05,
+      "loss": 1.7531,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2721966968831491,
+      "grad_norm": 0.2473239004611969,
+      "learning_rate": 5e-05,
+      "loss": 1.9925,
+      "step": 2192
+    },
+    {
+      "epoch": 0.27232087420836953,
+      "grad_norm": 0.24694471061229706,
+      "learning_rate": 5e-05,
+      "loss": 1.9079,
+      "step": 2193
+    },
+    {
+      "epoch": 0.27244505153358994,
+      "grad_norm": 0.2570899426937103,
+      "learning_rate": 5e-05,
+      "loss": 1.9159,
+      "step": 2194
+    },
+    {
+      "epoch": 0.2725692288588104,
+      "grad_norm": 0.24684719741344452,
+      "learning_rate": 5e-05,
+      "loss": 1.8469,
+      "step": 2195
+    },
+    {
+      "epoch": 0.2726934061840308,
+      "grad_norm": 0.23923180997371674,
+      "learning_rate": 5e-05,
+      "loss": 2.0765,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2728175835092512,
+      "grad_norm": 0.2378152459859848,
+      "learning_rate": 5e-05,
+      "loss": 2.0114,
+      "step": 2197
+    },
+    {
+      "epoch": 0.27294176083447164,
+      "grad_norm": 0.2399219423532486,
+      "learning_rate": 5e-05,
+      "loss": 1.7826,
+      "step": 2198
+    },
+    {
+      "epoch": 0.27306593815969205,
+      "grad_norm": 0.24805906414985657,
+      "learning_rate": 5e-05,
+      "loss": 1.9945,
+      "step": 2199
+    },
+    {
+      "epoch": 0.27319011548491245,
+      "grad_norm": 0.2733684778213501,
+      "learning_rate": 5e-05,
+      "loss": 2.1438,
+      "step": 2200
+    },
+    {
+      "epoch": 0.27331429281013286,
+      "grad_norm": 0.23346713185310364,
+      "learning_rate": 5e-05,
+      "loss": 1.6574,
+      "step": 2201
+    },
+    {
+      "epoch": 0.2734384701353533,
+      "grad_norm": 0.24403202533721924,
+      "learning_rate": 5e-05,
+      "loss": 2.1564,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2735626474605737,
+      "grad_norm": 0.2380913347005844,
+      "learning_rate": 5e-05,
+      "loss": 1.9491,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2736868247857941,
+      "grad_norm": 0.3888758718967438,
+      "learning_rate": 5e-05,
+      "loss": 2.9879,
+      "step": 2204
+    },
+    {
+      "epoch": 0.27381100211101456,
+      "grad_norm": 0.6516531705856323,
+      "learning_rate": 5e-05,
+      "loss": 3.3973,
+      "step": 2205
+    },
+    {
+      "epoch": 0.27393517943623497,
+      "grad_norm": 0.3094015121459961,
+      "learning_rate": 5e-05,
+      "loss": 2.1015,
+      "step": 2206
+    },
+    {
+      "epoch": 0.2740593567614554,
+      "grad_norm": 0.46443596482276917,
+      "learning_rate": 5e-05,
+      "loss": 2.1267,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2741835340866758,
+      "grad_norm": 0.2511177062988281,
+      "learning_rate": 5e-05,
+      "loss": 1.689,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2743077114118962,
+      "grad_norm": 0.30328696966171265,
+      "learning_rate": 5e-05,
+      "loss": 2.4356,
+      "step": 2209
+    },
+    {
+      "epoch": 0.2744318887371166,
+      "grad_norm": 0.31477105617523193,
+      "learning_rate": 5e-05,
+      "loss": 2.7878,
+      "step": 2210
+    },
+    {
+      "epoch": 0.274556066062337,
+      "grad_norm": 0.7352878451347351,
+      "learning_rate": 5e-05,
+      "loss": 3.478,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2746802433875574,
+      "grad_norm": 0.33083683252334595,
+      "learning_rate": 5e-05,
+      "loss": 2.4199,
+      "step": 2212
+    },
+    {
+      "epoch": 0.27480442071277783,
+      "grad_norm": 0.3189690113067627,
+      "learning_rate": 5e-05,
+      "loss": 1.8509,
+      "step": 2213
+    },
+    {
+      "epoch": 0.27492859803799824,
+      "grad_norm": 0.3604205846786499,
+      "learning_rate": 5e-05,
+      "loss": 1.7891,
+      "step": 2214
+    },
+    {
+      "epoch": 0.27505277536321865,
+      "grad_norm": 0.27476003766059875,
+      "learning_rate": 5e-05,
+      "loss": 1.7433,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2751769526884391,
+      "grad_norm": 0.297737717628479,
+      "learning_rate": 5e-05,
+      "loss": 1.9115,
+      "step": 2216
+    },
+    {
+      "epoch": 0.2753011300136595,
+      "grad_norm": 0.22951345145702362,
+      "learning_rate": 5e-05,
+      "loss": 1.7226,
+      "step": 2217
+    },
+    {
+      "epoch": 0.27542530733887993,
+      "grad_norm": 0.2588718831539154,
+      "learning_rate": 5e-05,
+      "loss": 2.1228,
+      "step": 2218
+    },
+    {
+      "epoch": 0.27554948466410034,
+      "grad_norm": 0.5698748230934143,
+      "learning_rate": 5e-05,
+      "loss": 3.1327,
+      "step": 2219
+    },
+    {
+      "epoch": 0.27567366198932075,
+      "grad_norm": 0.3699195683002472,
+      "learning_rate": 5e-05,
+      "loss": 2.8148,
+      "step": 2220
+    },
+    {
+      "epoch": 0.27579783931454116,
+      "grad_norm": 0.36215469241142273,
+      "learning_rate": 5e-05,
+      "loss": 1.8118,
+      "step": 2221
+    },
+    {
+      "epoch": 0.27592201663976157,
+      "grad_norm": 0.36940595507621765,
+      "learning_rate": 5e-05,
+      "loss": 2.2595,
+      "step": 2222
+    },
+    {
+      "epoch": 0.276046193964982,
+      "grad_norm": 0.2779625952243805,
+      "learning_rate": 5e-05,
+      "loss": 1.7874,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2761703712902024,
+      "grad_norm": 0.2580413520336151,
+      "learning_rate": 5e-05,
+      "loss": 1.9087,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2762945486154228,
+      "grad_norm": 0.2595154643058777,
+      "learning_rate": 5e-05,
+      "loss": 2.1569,
+      "step": 2225
+    },
+    {
+      "epoch": 0.27641872594064326,
+      "grad_norm": 0.3171026110649109,
+      "learning_rate": 5e-05,
+      "loss": 2.4358,
+      "step": 2226
+    },
+    {
+      "epoch": 0.27654290326586367,
+      "grad_norm": 0.4386424124240875,
+      "learning_rate": 5e-05,
+      "loss": 3.1282,
+      "step": 2227
+    },
+    {
+      "epoch": 0.2766670805910841,
+      "grad_norm": 0.3229656219482422,
+      "learning_rate": 5e-05,
+      "loss": 2.2222,
+      "step": 2228
+    },
+    {
+      "epoch": 0.2767912579163045,
+      "grad_norm": 0.26302337646484375,
+      "learning_rate": 5e-05,
+      "loss": 2.0297,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2769154352415249,
+      "grad_norm": 0.23068632185459137,
+      "learning_rate": 5e-05,
+      "loss": 1.609,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2770396125667453,
+      "grad_norm": 0.24477170407772064,
+      "learning_rate": 5e-05,
+      "loss": 1.8443,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2771637898919657,
+      "grad_norm": 0.281086266040802,
+      "learning_rate": 5e-05,
+      "loss": 1.8743,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2772879672171861,
+      "grad_norm": 0.36004120111465454,
+      "learning_rate": 5e-05,
+      "loss": 2.2178,
+      "step": 2233
+    },
+    {
+      "epoch": 0.27741214454240654,
+      "grad_norm": 0.3039983808994293,
+      "learning_rate": 5e-05,
+      "loss": 2.2681,
+      "step": 2234
+    },
+    {
+      "epoch": 0.27753632186762695,
+      "grad_norm": 0.24104949831962585,
+      "learning_rate": 5e-05,
+      "loss": 1.791,
+      "step": 2235
+    },
+    {
+      "epoch": 0.2776604991928474,
+      "grad_norm": 0.2971158027648926,
+      "learning_rate": 5e-05,
+      "loss": 2.2856,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2777846765180678,
+      "grad_norm": 0.6112428307533264,
+      "learning_rate": 5e-05,
+      "loss": 3.3518,
+      "step": 2237
+    },
+    {
+      "epoch": 0.27790885384328823,
+      "grad_norm": 0.24120260775089264,
+      "learning_rate": 5e-05,
+      "loss": 1.7549,
+      "step": 2238
+    },
+    {
+      "epoch": 0.27803303116850864,
+      "grad_norm": 0.28159934282302856,
+      "learning_rate": 5e-05,
+      "loss": 2.0611,
+      "step": 2239
+    },
+    {
+      "epoch": 0.27815720849372905,
+      "grad_norm": 0.23474760353565216,
+      "learning_rate": 5e-05,
+      "loss": 1.9129,
+      "step": 2240
+    },
+    {
+      "epoch": 0.27828138581894946,
+      "grad_norm": 0.2380552887916565,
+      "learning_rate": 5e-05,
+      "loss": 1.8316,
+      "step": 2241
+    },
+    {
+      "epoch": 0.27840556314416987,
+      "grad_norm": 0.2604740262031555,
+      "learning_rate": 5e-05,
+      "loss": 2.0654,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2785297404693903,
+      "grad_norm": 0.2551412880420685,
+      "learning_rate": 5e-05,
+      "loss": 1.7274,
+      "step": 2243
+    },
+    {
+      "epoch": 0.2786539177946107,
+      "grad_norm": 0.2548082172870636,
+      "learning_rate": 5e-05,
+      "loss": 1.9978,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2787780951198311,
+      "grad_norm": 0.3627321720123291,
+      "learning_rate": 5e-05,
+      "loss": 2.8858,
+      "step": 2245
+    },
+    {
+      "epoch": 0.27890227244505156,
+      "grad_norm": 0.24462072551250458,
+      "learning_rate": 5e-05,
+      "loss": 2.0826,
+      "step": 2246
+    },
+    {
+      "epoch": 0.27902644977027197,
+      "grad_norm": 0.5113473534584045,
+      "learning_rate": 5e-05,
+      "loss": 2.8601,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2791506270954924,
+      "grad_norm": 0.2635073959827423,
+      "learning_rate": 5e-05,
+      "loss": 2.3143,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2792748044207128,
+      "grad_norm": 0.3092377781867981,
+      "learning_rate": 5e-05,
+      "loss": 2.1067,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2793989817459332,
+      "grad_norm": 0.22348101437091827,
+      "learning_rate": 5e-05,
+      "loss": 1.6217,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2795231590711536,
+      "grad_norm": 0.4503592252731323,
+      "learning_rate": 5e-05,
+      "loss": 2.6548,
+      "step": 2251
+    },
+    {
+      "epoch": 0.279647336396374,
+      "grad_norm": 0.288337379693985,
+      "learning_rate": 5e-05,
+      "loss": 2.1731,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2797715137215944,
+      "grad_norm": 0.25461119413375854,
+      "learning_rate": 5e-05,
+      "loss": 2.1026,
+      "step": 2253
+    },
+    {
+      "epoch": 0.27989569104681483,
+      "grad_norm": 0.3134608566761017,
+      "learning_rate": 5e-05,
+      "loss": 2.7694,
+      "step": 2254
+    },
+    {
+      "epoch": 0.28001986837203524,
+      "grad_norm": 0.2979837954044342,
+      "learning_rate": 5e-05,
+      "loss": 2.2254,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2801440456972557,
+      "grad_norm": 0.36209309101104736,
+      "learning_rate": 5e-05,
+      "loss": 2.7028,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2802682230224761,
+      "grad_norm": 0.3788640797138214,
+      "learning_rate": 5e-05,
+      "loss": 2.7447,
+      "step": 2257
+    },
+    {
+      "epoch": 0.2803924003476965,
+      "grad_norm": 0.25468653440475464,
+      "learning_rate": 5e-05,
+      "loss": 1.6277,
+      "step": 2258
+    },
+    {
+      "epoch": 0.28051657767291693,
+      "grad_norm": 0.2354038953781128,
+      "learning_rate": 5e-05,
+      "loss": 1.7599,
+      "step": 2259
+    },
+    {
+      "epoch": 0.28064075499813734,
+      "grad_norm": 0.30122730135917664,
+      "learning_rate": 5e-05,
+      "loss": 1.8833,
+      "step": 2260
+    },
+    {
+      "epoch": 0.28076493232335775,
+      "grad_norm": 0.32940787076950073,
+      "learning_rate": 5e-05,
+      "loss": 2.7236,
+      "step": 2261
+    },
+    {
+      "epoch": 0.28088910964857816,
+      "grad_norm": 0.25223779678344727,
+      "learning_rate": 5e-05,
+      "loss": 1.7249,
+      "step": 2262
+    },
+    {
+      "epoch": 0.28101328697379857,
+      "grad_norm": 0.3634922504425049,
+      "learning_rate": 5e-05,
+      "loss": 2.8195,
+      "step": 2263
+    },
+    {
+      "epoch": 0.281137464299019,
+      "grad_norm": 0.25755050778388977,
+      "learning_rate": 5e-05,
+      "loss": 2.021,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2812616416242394,
+      "grad_norm": 0.24746693670749664,
+      "learning_rate": 5e-05,
+      "loss": 1.8628,
+      "step": 2265
+    },
+    {
+      "epoch": 0.28138581894945985,
+      "grad_norm": 0.34618109464645386,
+      "learning_rate": 5e-05,
+      "loss": 2.8502,
+      "step": 2266
+    },
+    {
+      "epoch": 0.28150999627468026,
+      "grad_norm": 0.2772164046764374,
+      "learning_rate": 5e-05,
+      "loss": 2.1182,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2816341735999007,
+      "grad_norm": 0.24964843690395355,
+      "learning_rate": 5e-05,
+      "loss": 1.9586,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2817583509251211,
+      "grad_norm": 0.2371501624584198,
+      "learning_rate": 5e-05,
+      "loss": 1.987,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2818825282503415,
+      "grad_norm": 0.3191934823989868,
+      "learning_rate": 5e-05,
+      "loss": 2.9567,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2820067055755619,
+      "grad_norm": 0.2739006280899048,
+      "learning_rate": 5e-05,
+      "loss": 2.2691,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2821308829007823,
+      "grad_norm": 0.27598080039024353,
+      "learning_rate": 5e-05,
+      "loss": 2.2476,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2822550602260027,
+      "grad_norm": 0.23905223608016968,
+      "learning_rate": 5e-05,
+      "loss": 1.7903,
+      "step": 2273
+    },
+    {
+      "epoch": 0.28237923755122313,
+      "grad_norm": 0.30678167939186096,
+      "learning_rate": 5e-05,
+      "loss": 2.5744,
+      "step": 2274
+    },
+    {
+      "epoch": 0.28250341487644354,
+      "grad_norm": 0.23026947677135468,
+      "learning_rate": 5e-05,
+      "loss": 1.703,
+      "step": 2275
+    },
+    {
+      "epoch": 0.282627592201664,
+      "grad_norm": 0.4967389404773712,
+      "learning_rate": 5e-05,
+      "loss": 3.088,
+      "step": 2276
+    },
+    {
+      "epoch": 0.2827517695268844,
+      "grad_norm": 0.2934126853942871,
+      "learning_rate": 5e-05,
+      "loss": 2.1456,
+      "step": 2277
+    },
+    {
+      "epoch": 0.2828759468521048,
+      "grad_norm": 0.3016539216041565,
+      "learning_rate": 5e-05,
+      "loss": 2.1929,
+      "step": 2278
+    },
+    {
+      "epoch": 0.28300012417732523,
+      "grad_norm": 0.31496742367744446,
+      "learning_rate": 5e-05,
+      "loss": 2.5584,
+      "step": 2279
+    },
+    {
+      "epoch": 0.28312430150254564,
+      "grad_norm": 0.30930665135383606,
+      "learning_rate": 5e-05,
+      "loss": 2.4248,
+      "step": 2280
+    },
+    {
+      "epoch": 0.28324847882776605,
+      "grad_norm": 0.2566080093383789,
+      "learning_rate": 5e-05,
+      "loss": 1.9646,
+      "step": 2281
+    },
+    {
+      "epoch": 0.28337265615298646,
+      "grad_norm": 0.23041681945323944,
+      "learning_rate": 5e-05,
+      "loss": 1.836,
+      "step": 2282
+    },
+    {
+      "epoch": 0.28349683347820687,
+      "grad_norm": 0.2395760864019394,
+      "learning_rate": 5e-05,
+      "loss": 1.8731,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2836210108034273,
+      "grad_norm": 0.308156818151474,
+      "learning_rate": 5e-05,
+      "loss": 2.3229,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2837451881286477,
+      "grad_norm": 0.37290820479393005,
+      "learning_rate": 5e-05,
+      "loss": 3.3451,
+      "step": 2285
+    },
+    {
+      "epoch": 0.28386936545386815,
+      "grad_norm": 0.24113960564136505,
+      "learning_rate": 5e-05,
+      "loss": 1.7158,
+      "step": 2286
+    },
+    {
+      "epoch": 0.28399354277908856,
+      "grad_norm": 0.3444121479988098,
+      "learning_rate": 5e-05,
+      "loss": 2.6584,
+      "step": 2287
+    },
+    {
+      "epoch": 0.28411772010430897,
+      "grad_norm": 0.23463676869869232,
+      "learning_rate": 5e-05,
+      "loss": 1.6836,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2842418974295294,
+      "grad_norm": 0.3306749761104584,
+      "learning_rate": 5e-05,
+      "loss": 2.6244,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2843660747547498,
+      "grad_norm": 0.26500290632247925,
+      "learning_rate": 5e-05,
+      "loss": 1.932,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2844902520799702,
+      "grad_norm": 0.23685447871685028,
+      "learning_rate": 5e-05,
+      "loss": 1.8187,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2846144294051906,
+      "grad_norm": 0.425360769033432,
+      "learning_rate": 5e-05,
+      "loss": 3.4631,
+      "step": 2292
+    },
+    {
+      "epoch": 0.284738606730411,
+      "grad_norm": 0.2755737602710724,
+      "learning_rate": 5e-05,
+      "loss": 2.1109,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2848627840556314,
+      "grad_norm": 0.33775594830513,
+      "learning_rate": 5e-05,
+      "loss": 2.2524,
+      "step": 2294
+    },
+    {
+      "epoch": 0.28498696138085183,
+      "grad_norm": 0.2746567726135254,
+      "learning_rate": 5e-05,
+      "loss": 2.0766,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2851111387060723,
+      "grad_norm": 0.23744775354862213,
+      "learning_rate": 5e-05,
+      "loss": 1.7551,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2852353160312927,
+      "grad_norm": 0.2792019844055176,
+      "learning_rate": 5e-05,
+      "loss": 2.132,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2853594933565131,
+      "grad_norm": 0.23549817502498627,
+      "learning_rate": 5e-05,
+      "loss": 1.6724,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2854836706817335,
+      "grad_norm": 0.44914984703063965,
+      "learning_rate": 5e-05,
+      "loss": 1.869,
+      "step": 2299
+    },
+    {
+      "epoch": 0.28560784800695393,
+      "grad_norm": 0.25877654552459717,
+      "learning_rate": 5e-05,
+      "loss": 1.8464,
+      "step": 2300
+    },
+    {
+      "epoch": 0.28573202533217434,
+      "grad_norm": 0.4616888165473938,
+      "learning_rate": 5e-05,
+      "loss": 3.1038,
+      "step": 2301
+    },
+    {
+      "epoch": 0.28585620265739475,
+      "grad_norm": 0.23121219873428345,
+      "learning_rate": 5e-05,
+      "loss": 1.569,
+      "step": 2302
+    },
+    {
+      "epoch": 0.28598037998261516,
+      "grad_norm": 0.24577587842941284,
+      "learning_rate": 5e-05,
+      "loss": 1.946,
+      "step": 2303
+    },
+    {
+      "epoch": 0.28610455730783557,
+      "grad_norm": 0.23375777900218964,
+      "learning_rate": 5e-05,
+      "loss": 1.6012,
+      "step": 2304
+    },
+    {
+      "epoch": 0.286228734633056,
+      "grad_norm": 0.3729497492313385,
+      "learning_rate": 5e-05,
+      "loss": 2.2228,
+      "step": 2305
+    },
+    {
+      "epoch": 0.28635291195827645,
+      "grad_norm": 0.26911404728889465,
+      "learning_rate": 5e-05,
+      "loss": 1.9809,
+      "step": 2306
+    },
+    {
+      "epoch": 0.28647708928349686,
+      "grad_norm": 0.22182457149028778,
+      "learning_rate": 5e-05,
+      "loss": 1.6623,
+      "step": 2307
+    },
+    {
+      "epoch": 0.28660126660871726,
+      "grad_norm": 0.25709474086761475,
+      "learning_rate": 5e-05,
+      "loss": 2.0334,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2867254439339377,
+      "grad_norm": 0.3356473445892334,
+      "learning_rate": 5e-05,
+      "loss": 2.1495,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2868496212591581,
+      "grad_norm": 0.39738699793815613,
+      "learning_rate": 5e-05,
+      "loss": 2.4122,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2869737985843785,
+      "grad_norm": 0.4199839234352112,
+      "learning_rate": 5e-05,
+      "loss": 2.985,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2870979759095989,
+      "grad_norm": 0.30248400568962097,
+      "learning_rate": 5e-05,
+      "loss": 1.7053,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2872221532348193,
+      "grad_norm": 0.2530093491077423,
+      "learning_rate": 5e-05,
+      "loss": 2.0484,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2873463305600397,
+      "grad_norm": 0.30588701367378235,
+      "learning_rate": 5e-05,
+      "loss": 2.1618,
+      "step": 2314
+    },
+    {
+      "epoch": 0.28747050788526013,
+      "grad_norm": 0.29306522011756897,
+      "learning_rate": 5e-05,
+      "loss": 2.1627,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2875946852104806,
+      "grad_norm": 0.26126742362976074,
+      "learning_rate": 5e-05,
+      "loss": 2.5206,
+      "step": 2316
+    },
+    {
+      "epoch": 0.287718862535701,
+      "grad_norm": 0.2552174925804138,
+      "learning_rate": 5e-05,
+      "loss": 1.7767,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2878430398609214,
+      "grad_norm": 0.27038562297821045,
+      "learning_rate": 5e-05,
+      "loss": 2.0549,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2879672171861418,
+      "grad_norm": 0.25663262605667114,
+      "learning_rate": 5e-05,
+      "loss": 2.0929,
+      "step": 2319
+    },
+    {
+      "epoch": 0.28809139451136223,
+      "grad_norm": 0.23965750634670258,
+      "learning_rate": 5e-05,
+      "loss": 1.8899,
+      "step": 2320
+    },
+    {
+      "epoch": 0.28821557183658264,
+      "grad_norm": 0.26159292459487915,
+      "learning_rate": 5e-05,
+      "loss": 1.8408,
+      "step": 2321
+    },
+    {
+      "epoch": 0.28833974916180305,
+      "grad_norm": 0.29132992029190063,
+      "learning_rate": 5e-05,
+      "loss": 2.3826,
+      "step": 2322
+    },
+    {
+      "epoch": 0.28846392648702346,
+      "grad_norm": 0.2629832625389099,
+      "learning_rate": 5e-05,
+      "loss": 2.0558,
+      "step": 2323
+    },
+    {
+      "epoch": 0.28858810381224387,
+      "grad_norm": 0.24224704504013062,
+      "learning_rate": 5e-05,
+      "loss": 1.9818,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2887122811374643,
+      "grad_norm": 0.4076092541217804,
+      "learning_rate": 5e-05,
+      "loss": 2.5366,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2888364584626847,
+      "grad_norm": 0.24215994775295258,
+      "learning_rate": 5e-05,
+      "loss": 1.8944,
+      "step": 2326
+    },
+    {
+      "epoch": 0.28896063578790515,
+      "grad_norm": 0.23523783683776855,
+      "learning_rate": 5e-05,
+      "loss": 1.7512,
+      "step": 2327
+    },
+    {
+      "epoch": 0.28908481311312556,
+      "grad_norm": 0.2690104842185974,
+      "learning_rate": 5e-05,
+      "loss": 1.7761,
+      "step": 2328
+    },
+    {
+      "epoch": 0.28920899043834597,
+      "grad_norm": 0.2651064097881317,
+      "learning_rate": 5e-05,
+      "loss": 1.8072,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2893331677635664,
+      "grad_norm": 0.24765053391456604,
+      "learning_rate": 5e-05,
+      "loss": 1.9649,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2894573450887868,
+      "grad_norm": 0.4669734239578247,
+      "learning_rate": 5e-05,
+      "loss": 2.5411,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2895815224140072,
+      "grad_norm": 0.24417367577552795,
+      "learning_rate": 5e-05,
+      "loss": 1.6564,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2897056997392276,
+      "grad_norm": 0.2411481738090515,
+      "learning_rate": 5e-05,
+      "loss": 1.7386,
+      "step": 2333
+    },
+    {
+      "epoch": 0.289829877064448,
+      "grad_norm": 0.23474077880382538,
+      "learning_rate": 5e-05,
+      "loss": 1.7429,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2899540543896684,
+      "grad_norm": 0.2852562963962555,
+      "learning_rate": 5e-05,
+      "loss": 1.9652,
+      "step": 2335
+    },
+    {
+      "epoch": 0.29007823171488883,
+      "grad_norm": 0.2371487021446228,
+      "learning_rate": 5e-05,
+      "loss": 1.8221,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2902024090401093,
+      "grad_norm": 0.2400091588497162,
+      "learning_rate": 5e-05,
+      "loss": 1.9796,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2903265863653297,
+      "grad_norm": 0.23056505620479584,
+      "learning_rate": 5e-05,
+      "loss": 2.0365,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2904507636905501,
+      "grad_norm": 0.2592349648475647,
+      "learning_rate": 5e-05,
+      "loss": 2.2178,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2905749410157705,
+      "grad_norm": 0.24092377722263336,
+      "learning_rate": 5e-05,
+      "loss": 1.9342,
+      "step": 2340
+    },
+    {
+      "epoch": 0.29069911834099094,
+      "grad_norm": 0.28325721621513367,
+      "learning_rate": 5e-05,
+      "loss": 2.1013,
+      "step": 2341
+    },
+    {
+      "epoch": 0.29082329566621135,
+      "grad_norm": 0.23552647233009338,
+      "learning_rate": 5e-05,
+      "loss": 1.8412,
+      "step": 2342
+    },
+    {
+      "epoch": 0.29094747299143175,
+      "grad_norm": 0.331377238035202,
+      "learning_rate": 5e-05,
+      "loss": 2.1674,
+      "step": 2343
+    },
+    {
+      "epoch": 0.29107165031665216,
+      "grad_norm": 0.23094972968101501,
+      "learning_rate": 5e-05,
+      "loss": 1.774,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2911958276418726,
+      "grad_norm": 0.2975688576698303,
+      "learning_rate": 5e-05,
+      "loss": 2.5615,
+      "step": 2345
+    },
+    {
+      "epoch": 0.291320004967093,
+      "grad_norm": 0.23471659421920776,
+      "learning_rate": 5e-05,
+      "loss": 1.6161,
+      "step": 2346
+    },
+    {
+      "epoch": 0.29144418229231345,
+      "grad_norm": 0.26555222272872925,
+      "learning_rate": 5e-05,
+      "loss": 2.0333,
+      "step": 2347
+    },
+    {
+      "epoch": 0.29156835961753386,
+      "grad_norm": 0.2195957899093628,
+      "learning_rate": 5e-05,
+      "loss": 1.7668,
+      "step": 2348
+    },
+    {
+      "epoch": 0.29169253694275427,
+      "grad_norm": 0.2603464424610138,
+      "learning_rate": 5e-05,
+      "loss": 2.3325,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2918167142679747,
+      "grad_norm": 0.23768115043640137,
+      "learning_rate": 5e-05,
+      "loss": 2.135,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2919408915931951,
+      "grad_norm": 0.2528925836086273,
+      "learning_rate": 5e-05,
+      "loss": 2.0161,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2920650689184155,
+      "grad_norm": 0.24497003853321075,
+      "learning_rate": 5e-05,
+      "loss": 2.0842,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2921892462436359,
+      "grad_norm": 0.22991429269313812,
+      "learning_rate": 5e-05,
+      "loss": 1.734,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2923134235688563,
+      "grad_norm": 0.22641606628894806,
+      "learning_rate": 5e-05,
+      "loss": 1.8019,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2924376008940767,
+      "grad_norm": 0.3363061845302582,
+      "learning_rate": 5e-05,
+      "loss": 3.125,
+      "step": 2355
+    },
+    {
+      "epoch": 0.29256177821929713,
+      "grad_norm": 0.23606544733047485,
+      "learning_rate": 5e-05,
+      "loss": 1.703,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2926859555445176,
+      "grad_norm": 0.24433138966560364,
+      "learning_rate": 5e-05,
+      "loss": 1.7999,
+      "step": 2357
+    },
+    {
+      "epoch": 0.292810132869738,
+      "grad_norm": 0.2222929149866104,
+      "learning_rate": 5e-05,
+      "loss": 1.8295,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2929343101949584,
+      "grad_norm": 0.24050135910511017,
+      "learning_rate": 5e-05,
+      "loss": 2.1841,
+      "step": 2359
+    },
+    {
+      "epoch": 0.2930584875201788,
+      "grad_norm": 0.30043894052505493,
+      "learning_rate": 5e-05,
+      "loss": 2.293,
+      "step": 2360
+    },
+    {
+      "epoch": 0.29318266484539923,
+      "grad_norm": 0.22937791049480438,
+      "learning_rate": 5e-05,
+      "loss": 1.8276,
+      "step": 2361
+    },
+    {
+      "epoch": 0.29330684217061964,
+      "grad_norm": 0.24394236505031586,
+      "learning_rate": 5e-05,
+      "loss": 1.5651,
+      "step": 2362
+    },
+    {
+      "epoch": 0.29343101949584005,
+      "grad_norm": 0.22702492773532867,
+      "learning_rate": 5e-05,
+      "loss": 1.7366,
+      "step": 2363
+    },
+    {
+      "epoch": 0.29355519682106046,
+      "grad_norm": 0.24780035018920898,
+      "learning_rate": 5e-05,
+      "loss": 1.9111,
+      "step": 2364
+    },
+    {
+      "epoch": 0.29367937414628087,
+      "grad_norm": 0.23855777084827423,
+      "learning_rate": 5e-05,
+      "loss": 1.7977,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2938035514715013,
+      "grad_norm": 0.2370656579732895,
+      "learning_rate": 5e-05,
+      "loss": 1.6724,
+      "step": 2366
+    },
+    {
+      "epoch": 0.29392772879672174,
+      "grad_norm": 0.23753634095191956,
+      "learning_rate": 5e-05,
+      "loss": 1.8294,
+      "step": 2367
+    },
+    {
+      "epoch": 0.29405190612194215,
+      "grad_norm": 0.2513243854045868,
+      "learning_rate": 5e-05,
+      "loss": 2.1125,
+      "step": 2368
+    },
+    {
+      "epoch": 0.29417608344716256,
+      "grad_norm": 0.2436743527650833,
+      "learning_rate": 5e-05,
+      "loss": 1.9807,
+      "step": 2369
+    },
+    {
+      "epoch": 0.29430026077238297,
+      "grad_norm": 0.2230265587568283,
+      "learning_rate": 5e-05,
+      "loss": 1.8254,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2944244380976034,
+      "grad_norm": 0.24146917462348938,
+      "learning_rate": 5e-05,
+      "loss": 1.9186,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2945486154228238,
+      "grad_norm": 0.2423929125070572,
+      "learning_rate": 5e-05,
+      "loss": 1.8563,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2946727927480442,
+      "grad_norm": 0.5293620228767395,
+      "learning_rate": 5e-05,
+      "loss": 3.2232,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2947969700732646,
+      "grad_norm": 0.47385889291763306,
+      "learning_rate": 5e-05,
+      "loss": 3.2621,
+      "step": 2374
+    },
+    {
+      "epoch": 0.294921147398485,
+      "grad_norm": 0.2932327091693878,
+      "learning_rate": 5e-05,
+      "loss": 2.1003,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2950453247237054,
+      "grad_norm": 0.3463650941848755,
+      "learning_rate": 5e-05,
+      "loss": 2.9055,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2951695020489259,
+      "grad_norm": 0.21354363858699799,
+      "learning_rate": 5e-05,
+      "loss": 1.7144,
+      "step": 2377
+    },
+    {
+      "epoch": 0.2952936793741463,
+      "grad_norm": 0.36624735593795776,
+      "learning_rate": 5e-05,
+      "loss": 2.5151,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2954178566993667,
+      "grad_norm": 0.26084357500076294,
+      "learning_rate": 5e-05,
+      "loss": 1.8736,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2955420340245871,
+      "grad_norm": 0.23558272421360016,
+      "learning_rate": 5e-05,
+      "loss": 1.7314,
+      "step": 2380
+    },
+    {
+      "epoch": 0.29566621134980753,
+      "grad_norm": 0.23063595592975616,
+      "learning_rate": 5e-05,
+      "loss": 1.5839,
+      "step": 2381
+    },
+    {
+      "epoch": 0.29579038867502794,
+      "grad_norm": 0.2673996090888977,
+      "learning_rate": 5e-05,
+      "loss": 2.167,
+      "step": 2382
+    },
+    {
+      "epoch": 0.29591456600024835,
+      "grad_norm": 0.2304811179637909,
+      "learning_rate": 5e-05,
+      "loss": 1.6322,
+      "step": 2383
+    },
+    {
+      "epoch": 0.29603874332546876,
+      "grad_norm": 0.23837745189666748,
+      "learning_rate": 5e-05,
+      "loss": 1.9747,
+      "step": 2384
+    },
+    {
+      "epoch": 0.29616292065068917,
+      "grad_norm": 0.23210571706295013,
+      "learning_rate": 5e-05,
+      "loss": 1.6524,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2962870979759096,
+      "grad_norm": 0.23646484315395355,
+      "learning_rate": 5e-05,
+      "loss": 1.9154,
+      "step": 2386
+    },
+    {
+      "epoch": 0.29641127530113004,
+      "grad_norm": 0.23370255529880524,
+      "learning_rate": 5e-05,
+      "loss": 1.7822,
+      "step": 2387
+    },
+    {
+      "epoch": 0.29653545262635045,
+      "grad_norm": 0.24929741024971008,
+      "learning_rate": 5e-05,
+      "loss": 1.8317,
+      "step": 2388
+    },
+    {
+      "epoch": 0.29665962995157086,
+      "grad_norm": 0.29031649231910706,
+      "learning_rate": 5e-05,
+      "loss": 2.4248,
+      "step": 2389
+    },
+    {
+      "epoch": 0.29678380727679127,
+      "grad_norm": 0.3389052748680115,
+      "learning_rate": 5e-05,
+      "loss": 2.7452,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2969079846020117,
+      "grad_norm": 0.25351041555404663,
+      "learning_rate": 5e-05,
+      "loss": 1.8065,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2970321619272321,
+      "grad_norm": 0.2346285581588745,
+      "learning_rate": 5e-05,
+      "loss": 1.8006,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2971563392524525,
+      "grad_norm": 0.2813652753829956,
+      "learning_rate": 5e-05,
+      "loss": 2.1146,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2972805165776729,
+      "grad_norm": 0.43128257989883423,
+      "learning_rate": 5e-05,
+      "loss": 2.9285,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2974046939028933,
+      "grad_norm": 0.25530242919921875,
+      "learning_rate": 5e-05,
+      "loss": 1.9098,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2975288712281137,
+      "grad_norm": 0.24373075366020203,
+      "learning_rate": 5e-05,
+      "loss": 1.841,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2976530485533342,
+      "grad_norm": 0.24467530846595764,
+      "learning_rate": 5e-05,
+      "loss": 1.8234,
+      "step": 2397
+    },
+    {
+      "epoch": 0.2977772258785546,
+      "grad_norm": 0.2736798822879791,
+      "learning_rate": 5e-05,
+      "loss": 2.1031,
+      "step": 2398
+    },
+    {
+      "epoch": 0.297901403203775,
+      "grad_norm": 0.3020813763141632,
+      "learning_rate": 5e-05,
+      "loss": 2.7447,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2980255805289954,
+      "grad_norm": 0.3021032512187958,
+      "learning_rate": 5e-05,
+      "loss": 2.6588,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2981497578542158,
+      "grad_norm": 0.23449717462062836,
+      "learning_rate": 5e-05,
+      "loss": 1.9597,
+      "step": 2401
+    },
+    {
+      "epoch": 0.29827393517943623,
+      "grad_norm": 0.2941332757472992,
+      "learning_rate": 5e-05,
+      "loss": 1.8493,
+      "step": 2402
+    },
+    {
+      "epoch": 0.29839811250465664,
+      "grad_norm": 0.27627015113830566,
+      "learning_rate": 5e-05,
+      "loss": 1.7745,
+      "step": 2403
+    },
+    {
+      "epoch": 0.29852228982987705,
+      "grad_norm": 0.24577543139457703,
+      "learning_rate": 5e-05,
+      "loss": 1.6863,
+      "step": 2404
+    },
+    {
+      "epoch": 0.29864646715509746,
+      "grad_norm": 0.46722403168678284,
+      "learning_rate": 5e-05,
+      "loss": 3.2409,
+      "step": 2405
+    },
+    {
+      "epoch": 0.29877064448031787,
+      "grad_norm": 0.26553261280059814,
+      "learning_rate": 5e-05,
+      "loss": 1.9566,
+      "step": 2406
+    },
+    {
+      "epoch": 0.29889482180553834,
+      "grad_norm": 0.22868528962135315,
+      "learning_rate": 5e-05,
+      "loss": 1.7161,
+      "step": 2407
+    },
+    {
+      "epoch": 0.29901899913075874,
+      "grad_norm": 0.2631630599498749,
+      "learning_rate": 5e-05,
+      "loss": 1.9372,
+      "step": 2408
+    },
+    {
+      "epoch": 0.29914317645597915,
+      "grad_norm": 0.2706388831138611,
+      "learning_rate": 5e-05,
+      "loss": 1.9286,
+      "step": 2409
+    },
+    {
+      "epoch": 0.29926735378119956,
+      "grad_norm": 0.24649964272975922,
+      "learning_rate": 5e-05,
+      "loss": 1.849,
+      "step": 2410
+    },
+    {
+      "epoch": 0.29939153110642,
+      "grad_norm": 0.3356122672557831,
+      "learning_rate": 5e-05,
+      "loss": 1.9797,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2995157084316404,
+      "grad_norm": 0.25064951181411743,
+      "learning_rate": 5e-05,
+      "loss": 1.7351,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2996398857568608,
+      "grad_norm": 0.33403944969177246,
+      "learning_rate": 5e-05,
+      "loss": 2.1686,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2997640630820812,
+      "grad_norm": 0.2619416415691376,
+      "learning_rate": 5e-05,
+      "loss": 1.9199,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2998882404073016,
+      "grad_norm": 0.24863988161087036,
+      "learning_rate": 5e-05,
+      "loss": 1.9177,
+      "step": 2415
+    },
+    {
+      "epoch": 0.300012417732522,
+      "grad_norm": 0.2544436454772949,
+      "learning_rate": 5e-05,
+      "loss": 1.8118,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3001365950577425,
+      "grad_norm": 0.25741347670555115,
+      "learning_rate": 5e-05,
+      "loss": 1.9319,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3002607723829629,
+      "grad_norm": 0.3917241096496582,
+      "learning_rate": 5e-05,
+      "loss": 3.2297,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3003849497081833,
+      "grad_norm": 0.25496190786361694,
+      "learning_rate": 5e-05,
+      "loss": 1.6107,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3005091270334037,
+      "grad_norm": 0.23908790946006775,
+      "learning_rate": 5e-05,
+      "loss": 1.8478,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3006333043586241,
+      "grad_norm": 0.2733415365219116,
+      "learning_rate": 5e-05,
+      "loss": 2.0505,
+      "step": 2421
+    },
+    {
+      "epoch": 0.30075748168384453,
+      "grad_norm": 0.23934225738048553,
+      "learning_rate": 5e-05,
+      "loss": 2.0059,
+      "step": 2422
+    },
+    {
+      "epoch": 0.30088165900906494,
+      "grad_norm": 0.23490947484970093,
+      "learning_rate": 5e-05,
+      "loss": 1.8443,
+      "step": 2423
+    },
+    {
+      "epoch": 0.30100583633428535,
+      "grad_norm": 0.23969963192939758,
+      "learning_rate": 5e-05,
+      "loss": 1.9726,
+      "step": 2424
+    },
+    {
+      "epoch": 0.30113001365950576,
+      "grad_norm": 0.24457332491874695,
+      "learning_rate": 5e-05,
+      "loss": 1.8928,
+      "step": 2425
+    },
+    {
+      "epoch": 0.30125419098472617,
+      "grad_norm": 0.632684051990509,
+      "learning_rate": 5e-05,
+      "loss": 3.0825,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3013783683099466,
+      "grad_norm": 0.23030449450016022,
+      "learning_rate": 5e-05,
+      "loss": 1.9712,
+      "step": 2427
+    },
+    {
+      "epoch": 0.30150254563516704,
+      "grad_norm": 0.27462518215179443,
+      "learning_rate": 5e-05,
+      "loss": 2.1805,
+      "step": 2428
+    },
+    {
+      "epoch": 0.30162672296038745,
+      "grad_norm": 0.2525634169578552,
+      "learning_rate": 5e-05,
+      "loss": 1.8778,
+      "step": 2429
+    },
+    {
+      "epoch": 0.30175090028560786,
+      "grad_norm": 0.3097149431705475,
+      "learning_rate": 5e-05,
+      "loss": 2.1009,
+      "step": 2430
+    },
+    {
+      "epoch": 0.30187507761082827,
+      "grad_norm": 0.27387306094169617,
+      "learning_rate": 5e-05,
+      "loss": 2.0801,
+      "step": 2431
+    },
+    {
+      "epoch": 0.3019992549360487,
+      "grad_norm": 0.23255454003810883,
+      "learning_rate": 5e-05,
+      "loss": 1.7646,
+      "step": 2432
+    },
+    {
+      "epoch": 0.3021234322612691,
+      "grad_norm": 0.26013875007629395,
+      "learning_rate": 5e-05,
+      "loss": 1.9605,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3022476095864895,
+      "grad_norm": 0.242452934384346,
+      "learning_rate": 5e-05,
+      "loss": 1.9056,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3023717869117099,
+      "grad_norm": 0.2807671129703522,
+      "learning_rate": 5e-05,
+      "loss": 2.0907,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3024959642369303,
+      "grad_norm": 0.2769166827201843,
+      "learning_rate": 5e-05,
+      "loss": 2.196,
+      "step": 2436
+    },
+    {
+      "epoch": 0.3026201415621507,
+      "grad_norm": 0.2445501834154129,
+      "learning_rate": 5e-05,
+      "loss": 2.1329,
+      "step": 2437
+    },
+    {
+      "epoch": 0.3027443188873712,
+      "grad_norm": 0.23443907499313354,
+      "learning_rate": 5e-05,
+      "loss": 1.8675,
+      "step": 2438
+    },
+    {
+      "epoch": 0.3028684962125916,
+      "grad_norm": 0.23849956691265106,
+      "learning_rate": 5e-05,
+      "loss": 1.7111,
+      "step": 2439
+    },
+    {
+      "epoch": 0.302992673537812,
+      "grad_norm": 0.23851902782917023,
+      "learning_rate": 5e-05,
+      "loss": 1.9384,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3031168508630324,
+      "grad_norm": 0.34767386317253113,
+      "learning_rate": 5e-05,
+      "loss": 3.1742,
+      "step": 2441
+    },
+    {
+      "epoch": 0.3032410281882528,
+      "grad_norm": 0.2418692260980606,
+      "learning_rate": 5e-05,
+      "loss": 1.916,
+      "step": 2442
+    },
+    {
+      "epoch": 0.30336520551347324,
+      "grad_norm": 0.27822381258010864,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 2443
+    },
+    {
+      "epoch": 0.30348938283869364,
+      "grad_norm": 0.262796014547348,
+      "learning_rate": 5e-05,
+      "loss": 1.9772,
+      "step": 2444
+    },
+    {
+      "epoch": 0.30361356016391405,
+      "grad_norm": 0.24244970083236694,
+      "learning_rate": 5e-05,
+      "loss": 1.6814,
+      "step": 2445
+    },
+    {
+      "epoch": 0.30373773748913446,
+      "grad_norm": 0.4442673325538635,
+      "learning_rate": 5e-05,
+      "loss": 2.0069,
+      "step": 2446
+    },
+    {
+      "epoch": 0.30386191481435487,
+      "grad_norm": 0.23442882299423218,
+      "learning_rate": 5e-05,
+      "loss": 1.8279,
+      "step": 2447
+    },
+    {
+      "epoch": 0.30398609213957534,
+      "grad_norm": 0.2544330358505249,
+      "learning_rate": 5e-05,
+      "loss": 2.1364,
+      "step": 2448
+    },
+    {
+      "epoch": 0.30411026946479575,
+      "grad_norm": 0.3845314383506775,
+      "learning_rate": 5e-05,
+      "loss": 3.0167,
+      "step": 2449
+    },
+    {
+      "epoch": 0.30423444679001616,
+      "grad_norm": 0.2309444397687912,
+      "learning_rate": 5e-05,
+      "loss": 2.0023,
+      "step": 2450
+    },
+    {
+      "epoch": 0.30435862411523656,
+      "grad_norm": 0.22534552216529846,
+      "learning_rate": 5e-05,
+      "loss": 1.7266,
+      "step": 2451
+    },
+    {
+      "epoch": 0.304482801440457,
+      "grad_norm": 0.23042988777160645,
+      "learning_rate": 5e-05,
+      "loss": 1.7444,
+      "step": 2452
+    },
+    {
+      "epoch": 0.3046069787656774,
+      "grad_norm": 0.29422980546951294,
+      "learning_rate": 5e-05,
+      "loss": 2.3088,
+      "step": 2453
+    },
+    {
+      "epoch": 0.3047311560908978,
+      "grad_norm": 0.24359145760536194,
+      "learning_rate": 5e-05,
+      "loss": 1.939,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3048553334161182,
+      "grad_norm": 0.23669804632663727,
+      "learning_rate": 5e-05,
+      "loss": 1.6311,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3049795107413386,
+      "grad_norm": 0.22479701042175293,
+      "learning_rate": 5e-05,
+      "loss": 1.4663,
+      "step": 2456
+    },
+    {
+      "epoch": 0.305103688066559,
+      "grad_norm": 0.29498857259750366,
+      "learning_rate": 5e-05,
+      "loss": 2.3553,
+      "step": 2457
+    },
+    {
+      "epoch": 0.3052278653917795,
+      "grad_norm": 0.2492772787809372,
+      "learning_rate": 5e-05,
+      "loss": 1.9945,
+      "step": 2458
+    },
+    {
+      "epoch": 0.3053520427169999,
+      "grad_norm": 0.2304142266511917,
+      "learning_rate": 5e-05,
+      "loss": 1.923,
+      "step": 2459
+    },
+    {
+      "epoch": 0.3054762200422203,
+      "grad_norm": 0.23694701492786407,
+      "learning_rate": 5e-05,
+      "loss": 1.7034,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3056003973674407,
+      "grad_norm": 0.260919988155365,
+      "learning_rate": 5e-05,
+      "loss": 2.0516,
+      "step": 2461
+    },
+    {
+      "epoch": 0.3057245746926611,
+      "grad_norm": 0.22292177379131317,
+      "learning_rate": 5e-05,
+      "loss": 1.7469,
+      "step": 2462
+    },
+    {
+      "epoch": 0.30584875201788153,
+      "grad_norm": 0.3954099416732788,
+      "learning_rate": 5e-05,
+      "loss": 2.5796,
+      "step": 2463
+    },
+    {
+      "epoch": 0.30597292934310194,
+      "grad_norm": 0.39405757188796997,
+      "learning_rate": 5e-05,
+      "loss": 3.0105,
+      "step": 2464
+    },
+    {
+      "epoch": 0.30609710666832235,
+      "grad_norm": 0.26535579562187195,
+      "learning_rate": 5e-05,
+      "loss": 1.7596,
+      "step": 2465
+    },
+    {
+      "epoch": 0.30622128399354276,
+      "grad_norm": 0.316542387008667,
+      "learning_rate": 5e-05,
+      "loss": 2.0847,
+      "step": 2466
+    },
+    {
+      "epoch": 0.30634546131876317,
+      "grad_norm": 0.2840765416622162,
+      "learning_rate": 5e-05,
+      "loss": 1.8332,
+      "step": 2467
+    },
+    {
+      "epoch": 0.30646963864398363,
+      "grad_norm": 0.2294183373451233,
+      "learning_rate": 5e-05,
+      "loss": 1.7689,
+      "step": 2468
+    },
+    {
+      "epoch": 0.30659381596920404,
+      "grad_norm": 0.31016889214515686,
+      "learning_rate": 5e-05,
+      "loss": 2.3342,
+      "step": 2469
+    },
+    {
+      "epoch": 0.30671799329442445,
+      "grad_norm": 0.5072765350341797,
+      "learning_rate": 5e-05,
+      "loss": 2.7014,
+      "step": 2470
+    },
+    {
+      "epoch": 0.30684217061964486,
+      "grad_norm": 0.26482972502708435,
+      "learning_rate": 5e-05,
+      "loss": 1.6224,
+      "step": 2471
+    },
+    {
+      "epoch": 0.30696634794486527,
+      "grad_norm": 0.23487739264965057,
+      "learning_rate": 5e-05,
+      "loss": 1.7675,
+      "step": 2472
+    },
+    {
+      "epoch": 0.3070905252700857,
+      "grad_norm": 0.24181108176708221,
+      "learning_rate": 5e-05,
+      "loss": 1.6487,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3072147025953061,
+      "grad_norm": 0.2593574821949005,
+      "learning_rate": 5e-05,
+      "loss": 2.0773,
+      "step": 2474
+    },
+    {
+      "epoch": 0.3073388799205265,
+      "grad_norm": 0.25183579325675964,
+      "learning_rate": 5e-05,
+      "loss": 1.9005,
+      "step": 2475
+    },
+    {
+      "epoch": 0.3074630572457469,
+      "grad_norm": 0.30021190643310547,
+      "learning_rate": 5e-05,
+      "loss": 1.8803,
+      "step": 2476
+    },
+    {
+      "epoch": 0.3075872345709673,
+      "grad_norm": 0.3121073842048645,
+      "learning_rate": 5e-05,
+      "loss": 2.5081,
+      "step": 2477
+    },
+    {
+      "epoch": 0.3077114118961878,
+      "grad_norm": 0.293508917093277,
+      "learning_rate": 5e-05,
+      "loss": 2.1715,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3078355892214082,
+      "grad_norm": 0.264616996049881,
+      "learning_rate": 5e-05,
+      "loss": 2.1911,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3079597665466286,
+      "grad_norm": 0.23609627783298492,
+      "learning_rate": 5e-05,
+      "loss": 1.858,
+      "step": 2480
+    },
+    {
+      "epoch": 0.308083943871849,
+      "grad_norm": 0.24882352352142334,
+      "learning_rate": 5e-05,
+      "loss": 2.0246,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3082081211970694,
+      "grad_norm": 0.2793102264404297,
+      "learning_rate": 5e-05,
+      "loss": 2.5636,
+      "step": 2482
+    },
+    {
+      "epoch": 0.3083322985222898,
+      "grad_norm": 0.34422752261161804,
+      "learning_rate": 5e-05,
+      "loss": 2.8102,
+      "step": 2483
+    },
+    {
+      "epoch": 0.30845647584751024,
+      "grad_norm": 0.512165367603302,
+      "learning_rate": 5e-05,
+      "loss": 2.8471,
+      "step": 2484
+    },
+    {
+      "epoch": 0.30858065317273065,
+      "grad_norm": 0.27156296372413635,
+      "learning_rate": 5e-05,
+      "loss": 2.0946,
+      "step": 2485
+    },
+    {
+      "epoch": 0.30870483049795105,
+      "grad_norm": 0.5863181352615356,
+      "learning_rate": 5e-05,
+      "loss": 3.4516,
+      "step": 2486
+    },
+    {
+      "epoch": 0.30882900782317146,
+      "grad_norm": 0.263992577791214,
+      "learning_rate": 5e-05,
+      "loss": 2.0513,
+      "step": 2487
+    },
+    {
+      "epoch": 0.30895318514839193,
+      "grad_norm": 0.23446765542030334,
+      "learning_rate": 5e-05,
+      "loss": 1.9179,
+      "step": 2488
+    },
+    {
+      "epoch": 0.30907736247361234,
+      "grad_norm": 0.22363054752349854,
+      "learning_rate": 5e-05,
+      "loss": 1.6051,
+      "step": 2489
+    },
+    {
+      "epoch": 0.30920153979883275,
+      "grad_norm": 0.27542364597320557,
+      "learning_rate": 5e-05,
+      "loss": 2.1036,
+      "step": 2490
+    },
+    {
+      "epoch": 0.30932571712405316,
+      "grad_norm": 0.2588249444961548,
+      "learning_rate": 5e-05,
+      "loss": 2.2362,
+      "step": 2491
+    },
+    {
+      "epoch": 0.30944989444927357,
+      "grad_norm": 0.2412181794643402,
+      "learning_rate": 5e-05,
+      "loss": 1.7954,
+      "step": 2492
+    },
+    {
+      "epoch": 0.309574071774494,
+      "grad_norm": 0.2897854745388031,
+      "learning_rate": 5e-05,
+      "loss": 2.0443,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3096982490997144,
+      "grad_norm": 0.28221622109413147,
+      "learning_rate": 5e-05,
+      "loss": 2.2036,
+      "step": 2494
+    },
+    {
+      "epoch": 0.3098224264249348,
+      "grad_norm": 0.48167553544044495,
+      "learning_rate": 5e-05,
+      "loss": 3.063,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3099466037501552,
+      "grad_norm": 0.3016217052936554,
+      "learning_rate": 5e-05,
+      "loss": 2.156,
+      "step": 2496
+    },
+    {
+      "epoch": 0.3100707810753756,
+      "grad_norm": 0.24846698343753815,
+      "learning_rate": 5e-05,
+      "loss": 2.0878,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3101949584005961,
+      "grad_norm": 0.3203471899032593,
+      "learning_rate": 5e-05,
+      "loss": 2.0927,
+      "step": 2498
+    },
+    {
+      "epoch": 0.3103191357258165,
+      "grad_norm": 0.22517484426498413,
+      "learning_rate": 5e-05,
+      "loss": 1.7774,
+      "step": 2499
+    },
+    {
+      "epoch": 0.3104433130510369,
+      "grad_norm": 0.24136750400066376,
+      "learning_rate": 5e-05,
+      "loss": 1.9474,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3105674903762573,
+      "grad_norm": 0.265752911567688,
+      "learning_rate": 5e-05,
+      "loss": 1.9241,
+      "step": 2501
+    },
+    {
+      "epoch": 0.3106916677014777,
+      "grad_norm": 0.2521585524082184,
+      "learning_rate": 5e-05,
+      "loss": 2.0642,
+      "step": 2502
+    },
+    {
+      "epoch": 0.3108158450266981,
+      "grad_norm": 0.26665449142456055,
+      "learning_rate": 5e-05,
+      "loss": 1.9027,
+      "step": 2503
+    },
+    {
+      "epoch": 0.31094002235191853,
+      "grad_norm": 0.2607811987400055,
+      "learning_rate": 5e-05,
+      "loss": 2.3683,
+      "step": 2504
+    },
+    {
+      "epoch": 0.31106419967713894,
+      "grad_norm": 0.2730054557323456,
+      "learning_rate": 5e-05,
+      "loss": 2.1716,
+      "step": 2505
+    },
+    {
+      "epoch": 0.31118837700235935,
+      "grad_norm": 0.22732530534267426,
+      "learning_rate": 5e-05,
+      "loss": 1.503,
+      "step": 2506
+    },
+    {
+      "epoch": 0.31131255432757976,
+      "grad_norm": 0.23393070697784424,
+      "learning_rate": 5e-05,
+      "loss": 1.7821,
+      "step": 2507
+    },
+    {
+      "epoch": 0.3114367316528002,
+      "grad_norm": 0.293963760137558,
+      "learning_rate": 5e-05,
+      "loss": 2.6656,
+      "step": 2508
+    },
+    {
+      "epoch": 0.31156090897802063,
+      "grad_norm": 0.254947304725647,
+      "learning_rate": 5e-05,
+      "loss": 1.9871,
+      "step": 2509
+    },
+    {
+      "epoch": 0.31168508630324104,
+      "grad_norm": 0.24850718677043915,
+      "learning_rate": 5e-05,
+      "loss": 1.9521,
+      "step": 2510
+    },
+    {
+      "epoch": 0.31180926362846145,
+      "grad_norm": 0.2208259552717209,
+      "learning_rate": 5e-05,
+      "loss": 1.6537,
+      "step": 2511
+    },
+    {
+      "epoch": 0.31193344095368186,
+      "grad_norm": 0.2427312582731247,
+      "learning_rate": 5e-05,
+      "loss": 1.8059,
+      "step": 2512
+    },
+    {
+      "epoch": 0.31205761827890227,
+      "grad_norm": 0.4045015871524811,
+      "learning_rate": 5e-05,
+      "loss": 2.8572,
+      "step": 2513
+    },
+    {
+      "epoch": 0.3121817956041227,
+      "grad_norm": 0.23642271757125854,
+      "learning_rate": 5e-05,
+      "loss": 1.9659,
+      "step": 2514
+    },
+    {
+      "epoch": 0.3123059729293431,
+      "grad_norm": 0.2804414927959442,
+      "learning_rate": 5e-05,
+      "loss": 2.2127,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3124301502545635,
+      "grad_norm": 0.2459835261106491,
+      "learning_rate": 5e-05,
+      "loss": 2.0427,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3125543275797839,
+      "grad_norm": 0.24142514169216156,
+      "learning_rate": 5e-05,
+      "loss": 2.047,
+      "step": 2517
+    },
+    {
+      "epoch": 0.3126785049050044,
+      "grad_norm": 0.2312665432691574,
+      "learning_rate": 5e-05,
+      "loss": 2.2251,
+      "step": 2518
+    },
+    {
+      "epoch": 0.3128026822302248,
+      "grad_norm": 0.2877069115638733,
+      "learning_rate": 5e-05,
+      "loss": 2.527,
+      "step": 2519
+    },
+    {
+      "epoch": 0.3129268595554452,
+      "grad_norm": 0.2483450472354889,
+      "learning_rate": 5e-05,
+      "loss": 1.9798,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3130510368806656,
+      "grad_norm": 0.3296887278556824,
+      "learning_rate": 5e-05,
+      "loss": 3.1845,
+      "step": 2521
+    },
+    {
+      "epoch": 0.313175214205886,
+      "grad_norm": 0.3899775743484497,
+      "learning_rate": 5e-05,
+      "loss": 3.6965,
+      "step": 2522
+    },
+    {
+      "epoch": 0.3132993915311064,
+      "grad_norm": 0.28904592990875244,
+      "learning_rate": 5e-05,
+      "loss": 2.2741,
+      "step": 2523
+    },
+    {
+      "epoch": 0.31342356885632683,
+      "grad_norm": 0.23408769071102142,
+      "learning_rate": 5e-05,
+      "loss": 1.9654,
+      "step": 2524
+    },
+    {
+      "epoch": 0.31354774618154724,
+      "grad_norm": 0.23560366034507751,
+      "learning_rate": 5e-05,
+      "loss": 1.8064,
+      "step": 2525
+    },
+    {
+      "epoch": 0.31367192350676765,
+      "grad_norm": 0.35884061455726624,
+      "learning_rate": 5e-05,
+      "loss": 2.8486,
+      "step": 2526
+    },
+    {
+      "epoch": 0.31379610083198806,
+      "grad_norm": 0.37347790598869324,
+      "learning_rate": 5e-05,
+      "loss": 2.8253,
+      "step": 2527
+    },
+    {
+      "epoch": 0.3139202781572085,
+      "grad_norm": 0.2519088387489319,
+      "learning_rate": 5e-05,
+      "loss": 1.6374,
+      "step": 2528
+    },
+    {
+      "epoch": 0.31404445548242893,
+      "grad_norm": 0.28811997175216675,
+      "learning_rate": 5e-05,
+      "loss": 2.1381,
+      "step": 2529
+    },
+    {
+      "epoch": 0.31416863280764934,
+      "grad_norm": 0.3027819097042084,
+      "learning_rate": 5e-05,
+      "loss": 1.7923,
+      "step": 2530
+    },
+    {
+      "epoch": 0.31429281013286975,
+      "grad_norm": 0.2588360607624054,
+      "learning_rate": 5e-05,
+      "loss": 1.7273,
+      "step": 2531
+    },
+    {
+      "epoch": 0.31441698745809016,
+      "grad_norm": 0.29783663153648376,
+      "learning_rate": 5e-05,
+      "loss": 1.9726,
+      "step": 2532
+    },
+    {
+      "epoch": 0.31454116478331057,
+      "grad_norm": 0.31342625617980957,
+      "learning_rate": 5e-05,
+      "loss": 2.3345,
+      "step": 2533
+    },
+    {
+      "epoch": 0.314665342108531,
+      "grad_norm": 0.24071022868156433,
+      "learning_rate": 5e-05,
+      "loss": 1.8475,
+      "step": 2534
+    },
+    {
+      "epoch": 0.3147895194337514,
+      "grad_norm": 0.22726815938949585,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3149136967589718,
+      "grad_norm": 0.3996993899345398,
+      "learning_rate": 5e-05,
+      "loss": 2.6677,
+      "step": 2536
+    },
+    {
+      "epoch": 0.3150378740841922,
+      "grad_norm": 0.23101429641246796,
+      "learning_rate": 5e-05,
+      "loss": 1.7464,
+      "step": 2537
+    },
+    {
+      "epoch": 0.3151620514094126,
+      "grad_norm": 0.25499263405799866,
+      "learning_rate": 5e-05,
+      "loss": 2.2116,
+      "step": 2538
+    },
+    {
+      "epoch": 0.3152862287346331,
+      "grad_norm": 0.24672424793243408,
+      "learning_rate": 5e-05,
+      "loss": 1.7512,
+      "step": 2539
+    },
+    {
+      "epoch": 0.3154104060598535,
+      "grad_norm": 0.5668124556541443,
+      "learning_rate": 5e-05,
+      "loss": 2.8346,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3155345833850739,
+      "grad_norm": 0.23340995609760284,
+      "learning_rate": 5e-05,
+      "loss": 1.76,
+      "step": 2541
+    },
+    {
+      "epoch": 0.3156587607102943,
+      "grad_norm": 0.24544575810432434,
+      "learning_rate": 5e-05,
+      "loss": 1.8566,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3157829380355147,
+      "grad_norm": 0.33343833684921265,
+      "learning_rate": 5e-05,
+      "loss": 2.766,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3159071153607351,
+      "grad_norm": 0.3853832483291626,
+      "learning_rate": 5e-05,
+      "loss": 2.4766,
+      "step": 2544
+    },
+    {
+      "epoch": 0.31603129268595553,
+      "grad_norm": 0.24076224863529205,
+      "learning_rate": 5e-05,
+      "loss": 1.8126,
+      "step": 2545
+    },
+    {
+      "epoch": 0.31615547001117594,
+      "grad_norm": 0.3051636815071106,
+      "learning_rate": 5e-05,
+      "loss": 2.1975,
+      "step": 2546
+    },
+    {
+      "epoch": 0.31627964733639635,
+      "grad_norm": 0.25120866298675537,
+      "learning_rate": 5e-05,
+      "loss": 1.8756,
+      "step": 2547
+    },
+    {
+      "epoch": 0.31640382466161676,
+      "grad_norm": 0.2584150433540344,
+      "learning_rate": 5e-05,
+      "loss": 1.6929,
+      "step": 2548
+    },
+    {
+      "epoch": 0.3165280019868372,
+      "grad_norm": 0.2948315143585205,
+      "learning_rate": 5e-05,
+      "loss": 2.0621,
+      "step": 2549
+    },
+    {
+      "epoch": 0.31665217931205764,
+      "grad_norm": 0.2855660617351532,
+      "learning_rate": 5e-05,
+      "loss": 1.9359,
+      "step": 2550
+    },
+    {
+      "epoch": 0.31677635663727804,
+      "grad_norm": 0.2411571592092514,
+      "learning_rate": 5e-05,
+      "loss": 1.8733,
+      "step": 2551
+    },
+    {
+      "epoch": 0.31690053396249845,
+      "grad_norm": 0.26222503185272217,
+      "learning_rate": 5e-05,
+      "loss": 1.6517,
+      "step": 2552
+    },
+    {
+      "epoch": 0.31702471128771886,
+      "grad_norm": 0.3573346436023712,
+      "learning_rate": 5e-05,
+      "loss": 2.6762,
+      "step": 2553
+    },
+    {
+      "epoch": 0.3171488886129393,
+      "grad_norm": 0.26404184103012085,
+      "learning_rate": 5e-05,
+      "loss": 2.0124,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3172730659381597,
+      "grad_norm": 0.2882382869720459,
+      "learning_rate": 5e-05,
+      "loss": 2.1145,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3173972432633801,
+      "grad_norm": 0.26033490896224976,
+      "learning_rate": 5e-05,
+      "loss": 2.0684,
+      "step": 2556
+    },
+    {
+      "epoch": 0.3175214205886005,
+      "grad_norm": 0.24561190605163574,
+      "learning_rate": 5e-05,
+      "loss": 1.9723,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3176455979138209,
+      "grad_norm": 0.25734731554985046,
+      "learning_rate": 5e-05,
+      "loss": 1.9308,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3177697752390414,
+      "grad_norm": 0.23246902227401733,
+      "learning_rate": 5e-05,
+      "loss": 1.9535,
+      "step": 2559
+    },
+    {
+      "epoch": 0.3178939525642618,
+      "grad_norm": 0.2905048131942749,
+      "learning_rate": 5e-05,
+      "loss": 2.1131,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3180181298894822,
+      "grad_norm": 0.4069557785987854,
+      "learning_rate": 5e-05,
+      "loss": 3.1098,
+      "step": 2561
+    },
+    {
+      "epoch": 0.3181423072147026,
+      "grad_norm": 0.24848723411560059,
+      "learning_rate": 5e-05,
+      "loss": 1.6208,
+      "step": 2562
+    },
+    {
+      "epoch": 0.318266484539923,
+      "grad_norm": 0.23309265077114105,
+      "learning_rate": 5e-05,
+      "loss": 1.9138,
+      "step": 2563
+    },
+    {
+      "epoch": 0.3183906618651434,
+      "grad_norm": 0.23934386670589447,
+      "learning_rate": 5e-05,
+      "loss": 1.9338,
+      "step": 2564
+    },
+    {
+      "epoch": 0.31851483919036383,
+      "grad_norm": 0.24818089604377747,
+      "learning_rate": 5e-05,
+      "loss": 1.828,
+      "step": 2565
+    },
+    {
+      "epoch": 0.31863901651558424,
+      "grad_norm": 0.2427911013364792,
+      "learning_rate": 5e-05,
+      "loss": 2.1759,
+      "step": 2566
+    },
+    {
+      "epoch": 0.31876319384080465,
+      "grad_norm": 0.45614150166511536,
+      "learning_rate": 5e-05,
+      "loss": 2.8011,
+      "step": 2567
+    },
+    {
+      "epoch": 0.31888737116602506,
+      "grad_norm": 0.27476051449775696,
+      "learning_rate": 5e-05,
+      "loss": 1.8987,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3190115484912455,
+      "grad_norm": 0.29608142375946045,
+      "learning_rate": 5e-05,
+      "loss": 2.0781,
+      "step": 2569
+    },
+    {
+      "epoch": 0.31913572581646593,
+      "grad_norm": 0.25067877769470215,
+      "learning_rate": 5e-05,
+      "loss": 1.8279,
+      "step": 2570
+    },
+    {
+      "epoch": 0.31925990314168634,
+      "grad_norm": 0.2373666763305664,
+      "learning_rate": 5e-05,
+      "loss": 1.7941,
+      "step": 2571
+    },
+    {
+      "epoch": 0.31938408046690675,
+      "grad_norm": 0.31739741563796997,
+      "learning_rate": 5e-05,
+      "loss": 3.0416,
+      "step": 2572
+    },
+    {
+      "epoch": 0.31950825779212716,
+      "grad_norm": 0.2558632791042328,
+      "learning_rate": 5e-05,
+      "loss": 2.1331,
+      "step": 2573
+    },
+    {
+      "epoch": 0.31963243511734757,
+      "grad_norm": 0.3861458897590637,
+      "learning_rate": 5e-05,
+      "loss": 2.8171,
+      "step": 2574
+    },
+    {
+      "epoch": 0.319756612442568,
+      "grad_norm": 0.25404876470565796,
+      "learning_rate": 5e-05,
+      "loss": 1.8765,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3198807897677884,
+      "grad_norm": 0.26449182629585266,
+      "learning_rate": 5e-05,
+      "loss": 2.7219,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3200049670930088,
+      "grad_norm": 0.28693827986717224,
+      "learning_rate": 5e-05,
+      "loss": 1.9925,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3201291444182292,
+      "grad_norm": 0.3140489459037781,
+      "learning_rate": 5e-05,
+      "loss": 2.4472,
+      "step": 2578
+    },
+    {
+      "epoch": 0.32025332174344967,
+      "grad_norm": 0.3093940317630768,
+      "learning_rate": 5e-05,
+      "loss": 2.3313,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3203774990686701,
+      "grad_norm": 0.3359769582748413,
+      "learning_rate": 5e-05,
+      "loss": 2.0752,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3205016763938905,
+      "grad_norm": 0.22974927723407745,
+      "learning_rate": 5e-05,
+      "loss": 1.7855,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3206258537191109,
+      "grad_norm": 0.22663480043411255,
+      "learning_rate": 5e-05,
+      "loss": 1.7345,
+      "step": 2582
+    },
+    {
+      "epoch": 0.3207500310443313,
+      "grad_norm": 0.25881427526474,
+      "learning_rate": 5e-05,
+      "loss": 2.1007,
+      "step": 2583
+    },
+    {
+      "epoch": 0.3208742083695517,
+      "grad_norm": 0.26866796612739563,
+      "learning_rate": 5e-05,
+      "loss": 1.8233,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3209983856947721,
+      "grad_norm": 0.25161150097846985,
+      "learning_rate": 5e-05,
+      "loss": 1.6518,
+      "step": 2585
+    },
+    {
+      "epoch": 0.32112256301999254,
+      "grad_norm": 0.2753816545009613,
+      "learning_rate": 5e-05,
+      "loss": 2.4342,
+      "step": 2586
+    },
+    {
+      "epoch": 0.32124674034521294,
+      "grad_norm": 0.23251129686832428,
+      "learning_rate": 5e-05,
+      "loss": 1.693,
+      "step": 2587
+    },
+    {
+      "epoch": 0.32137091767043335,
+      "grad_norm": 0.2461782991886139,
+      "learning_rate": 5e-05,
+      "loss": 1.8945,
+      "step": 2588
+    },
+    {
+      "epoch": 0.3214950949956538,
+      "grad_norm": 0.2650192975997925,
+      "learning_rate": 5e-05,
+      "loss": 1.9366,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3216192723208742,
+      "grad_norm": 0.26024186611175537,
+      "learning_rate": 5e-05,
+      "loss": 1.8595,
+      "step": 2590
+    },
+    {
+      "epoch": 0.32174344964609464,
+      "grad_norm": 0.27012261748313904,
+      "learning_rate": 5e-05,
+      "loss": 2.2075,
+      "step": 2591
+    },
+    {
+      "epoch": 0.32186762697131505,
+      "grad_norm": 0.2246539145708084,
+      "learning_rate": 5e-05,
+      "loss": 1.688,
+      "step": 2592
+    },
+    {
+      "epoch": 0.32199180429653546,
+      "grad_norm": 0.23415213823318481,
+      "learning_rate": 5e-05,
+      "loss": 1.8578,
+      "step": 2593
+    },
+    {
+      "epoch": 0.32211598162175586,
+      "grad_norm": 0.2437078058719635,
+      "learning_rate": 5e-05,
+      "loss": 2.0806,
+      "step": 2594
+    },
+    {
+      "epoch": 0.3222401589469763,
+      "grad_norm": 0.24832111597061157,
+      "learning_rate": 5e-05,
+      "loss": 1.8768,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3223643362721967,
+      "grad_norm": 0.23762859404087067,
+      "learning_rate": 5e-05,
+      "loss": 1.8972,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3224885135974171,
+      "grad_norm": 0.2609170377254486,
+      "learning_rate": 5e-05,
+      "loss": 1.9046,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3226126909226375,
+      "grad_norm": 0.24320930242538452,
+      "learning_rate": 5e-05,
+      "loss": 1.8634,
+      "step": 2598
+    },
+    {
+      "epoch": 0.32273686824785797,
+      "grad_norm": 0.23969106376171112,
+      "learning_rate": 5e-05,
+      "loss": 2.0519,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3228610455730784,
+      "grad_norm": 0.23233890533447266,
+      "learning_rate": 5e-05,
+      "loss": 2.0007,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3229852228982988,
+      "grad_norm": 0.2541627287864685,
+      "learning_rate": 5e-05,
+      "loss": 1.9781,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3231094002235192,
+      "grad_norm": 0.3138279914855957,
+      "learning_rate": 5e-05,
+      "loss": 2.7624,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3232335775487396,
+      "grad_norm": 0.26152801513671875,
+      "learning_rate": 5e-05,
+      "loss": 1.887,
+      "step": 2603
+    },
+    {
+      "epoch": 0.32335775487396,
+      "grad_norm": 0.2979021370410919,
+      "learning_rate": 5e-05,
+      "loss": 2.2104,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3234819321991804,
+      "grad_norm": 0.25503554940223694,
+      "learning_rate": 5e-05,
+      "loss": 1.8952,
+      "step": 2605
+    },
+    {
+      "epoch": 0.32360610952440083,
+      "grad_norm": 0.26938286423683167,
+      "learning_rate": 5e-05,
+      "loss": 1.9496,
+      "step": 2606
+    },
+    {
+      "epoch": 0.32373028684962124,
+      "grad_norm": 0.24282211065292358,
+      "learning_rate": 5e-05,
+      "loss": 2.2124,
+      "step": 2607
+    },
+    {
+      "epoch": 0.32385446417484165,
+      "grad_norm": 0.2514600455760956,
+      "learning_rate": 5e-05,
+      "loss": 1.9283,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3239786415000621,
+      "grad_norm": 0.3718923032283783,
+      "learning_rate": 5e-05,
+      "loss": 2.8925,
+      "step": 2609
+    },
+    {
+      "epoch": 0.3241028188252825,
+      "grad_norm": 0.251128226518631,
+      "learning_rate": 5e-05,
+      "loss": 2.2661,
+      "step": 2610
+    },
+    {
+      "epoch": 0.32422699615050293,
+      "grad_norm": 0.29401934146881104,
+      "learning_rate": 5e-05,
+      "loss": 2.8054,
+      "step": 2611
+    },
+    {
+      "epoch": 0.32435117347572334,
+      "grad_norm": 0.2613525092601776,
+      "learning_rate": 5e-05,
+      "loss": 1.985,
+      "step": 2612
+    },
+    {
+      "epoch": 0.32447535080094375,
+      "grad_norm": 0.2795410752296448,
+      "learning_rate": 5e-05,
+      "loss": 1.9676,
+      "step": 2613
+    },
+    {
+      "epoch": 0.32459952812616416,
+      "grad_norm": 0.33179956674575806,
+      "learning_rate": 5e-05,
+      "loss": 1.8922,
+      "step": 2614
+    },
+    {
+      "epoch": 0.32472370545138457,
+      "grad_norm": 0.2416449785232544,
+      "learning_rate": 5e-05,
+      "loss": 1.9962,
+      "step": 2615
+    },
+    {
+      "epoch": 0.324847882776605,
+      "grad_norm": 0.23456701636314392,
+      "learning_rate": 5e-05,
+      "loss": 1.9759,
+      "step": 2616
+    },
+    {
+      "epoch": 0.3249720601018254,
+      "grad_norm": 0.2778686583042145,
+      "learning_rate": 5e-05,
+      "loss": 2.401,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3250962374270458,
+      "grad_norm": 0.3438623547554016,
+      "learning_rate": 5e-05,
+      "loss": 2.7347,
+      "step": 2618
+    },
+    {
+      "epoch": 0.32522041475226626,
+      "grad_norm": 0.3233967125415802,
+      "learning_rate": 5e-05,
+      "loss": 2.2459,
+      "step": 2619
+    },
+    {
+      "epoch": 0.32534459207748667,
+      "grad_norm": 0.2583015263080597,
+      "learning_rate": 5e-05,
+      "loss": 1.9029,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3254687694027071,
+      "grad_norm": 0.3060656487941742,
+      "learning_rate": 5e-05,
+      "loss": 1.5508,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3255929467279275,
+      "grad_norm": 0.25296443700790405,
+      "learning_rate": 5e-05,
+      "loss": 2.5846,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3257171240531479,
+      "grad_norm": 0.24775736033916473,
+      "learning_rate": 5e-05,
+      "loss": 2.3037,
+      "step": 2623
+    },
+    {
+      "epoch": 0.3258413013783683,
+      "grad_norm": 0.3886055648326874,
+      "learning_rate": 5e-05,
+      "loss": 2.3527,
+      "step": 2624
+    },
+    {
+      "epoch": 0.3259654787035887,
+      "grad_norm": 0.334266722202301,
+      "learning_rate": 5e-05,
+      "loss": 2.2881,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3260896560288091,
+      "grad_norm": 0.23398357629776,
+      "learning_rate": 5e-05,
+      "loss": 1.9165,
+      "step": 2626
+    },
+    {
+      "epoch": 0.32621383335402954,
+      "grad_norm": 0.36124876141548157,
+      "learning_rate": 5e-05,
+      "loss": 2.7855,
+      "step": 2627
+    },
+    {
+      "epoch": 0.32633801067924995,
+      "grad_norm": 0.29917779564857483,
+      "learning_rate": 5e-05,
+      "loss": 1.9819,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3264621880044704,
+      "grad_norm": 0.23874613642692566,
+      "learning_rate": 5e-05,
+      "loss": 1.8867,
+      "step": 2629
+    },
+    {
+      "epoch": 0.3265863653296908,
+      "grad_norm": 0.2507156431674957,
+      "learning_rate": 5e-05,
+      "loss": 1.9305,
+      "step": 2630
+    },
+    {
+      "epoch": 0.32671054265491123,
+      "grad_norm": 0.27563899755477905,
+      "learning_rate": 5e-05,
+      "loss": 1.9243,
+      "step": 2631
+    },
+    {
+      "epoch": 0.32683471998013164,
+      "grad_norm": 0.3420770764350891,
+      "learning_rate": 5e-05,
+      "loss": 2.1129,
+      "step": 2632
+    },
+    {
+      "epoch": 0.32695889730535205,
+      "grad_norm": 0.2556460499763489,
+      "learning_rate": 5e-05,
+      "loss": 1.9466,
+      "step": 2633
+    },
+    {
+      "epoch": 0.32708307463057246,
+      "grad_norm": 0.25012239813804626,
+      "learning_rate": 5e-05,
+      "loss": 2.0434,
+      "step": 2634
+    },
+    {
+      "epoch": 0.32720725195579287,
+      "grad_norm": 0.23320777714252472,
+      "learning_rate": 5e-05,
+      "loss": 1.7267,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3273314292810133,
+      "grad_norm": 0.2637651860713959,
+      "learning_rate": 5e-05,
+      "loss": 1.8675,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3274556066062337,
+      "grad_norm": 0.25030505657196045,
+      "learning_rate": 5e-05,
+      "loss": 1.903,
+      "step": 2637
+    },
+    {
+      "epoch": 0.3275797839314541,
+      "grad_norm": 0.29324087500572205,
+      "learning_rate": 5e-05,
+      "loss": 2.2937,
+      "step": 2638
+    },
+    {
+      "epoch": 0.32770396125667456,
+      "grad_norm": 0.26107755303382874,
+      "learning_rate": 5e-05,
+      "loss": 2.2742,
+      "step": 2639
+    },
+    {
+      "epoch": 0.32782813858189497,
+      "grad_norm": 0.22181333601474762,
+      "learning_rate": 5e-05,
+      "loss": 1.7687,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3279523159071154,
+      "grad_norm": 0.2466670721769333,
+      "learning_rate": 5e-05,
+      "loss": 1.9903,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3280764932323358,
+      "grad_norm": 0.23251676559448242,
+      "learning_rate": 5e-05,
+      "loss": 1.9068,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3282006705575562,
+      "grad_norm": 0.23802676796913147,
+      "learning_rate": 5e-05,
+      "loss": 1.9622,
+      "step": 2643
+    },
+    {
+      "epoch": 0.3283248478827766,
+      "grad_norm": 0.2624177932739258,
+      "learning_rate": 5e-05,
+      "loss": 2.1765,
+      "step": 2644
+    },
+    {
+      "epoch": 0.328449025207997,
+      "grad_norm": 0.23530355095863342,
+      "learning_rate": 5e-05,
+      "loss": 1.5327,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3285732025332174,
+      "grad_norm": 0.22901184856891632,
+      "learning_rate": 5e-05,
+      "loss": 1.7254,
+      "step": 2646
+    },
+    {
+      "epoch": 0.32869737985843783,
+      "grad_norm": 0.2742987275123596,
+      "learning_rate": 5e-05,
+      "loss": 2.1629,
+      "step": 2647
+    },
+    {
+      "epoch": 0.32882155718365824,
+      "grad_norm": 0.2945564091205597,
+      "learning_rate": 5e-05,
+      "loss": 2.4229,
+      "step": 2648
+    },
+    {
+      "epoch": 0.32894573450887865,
+      "grad_norm": 0.23239965736865997,
+      "learning_rate": 5e-05,
+      "loss": 1.9242,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3290699118340991,
+      "grad_norm": 0.3828739523887634,
+      "learning_rate": 5e-05,
+      "loss": 2.9628,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3291940891593195,
+      "grad_norm": 0.2655232846736908,
+      "learning_rate": 5e-05,
+      "loss": 1.967,
+      "step": 2651
+    },
+    {
+      "epoch": 0.32931826648453993,
+      "grad_norm": 0.24563553929328918,
+      "learning_rate": 5e-05,
+      "loss": 1.8046,
+      "step": 2652
+    },
+    {
+      "epoch": 0.32944244380976034,
+      "grad_norm": 0.24741974472999573,
+      "learning_rate": 5e-05,
+      "loss": 1.7547,
+      "step": 2653
+    },
+    {
+      "epoch": 0.32956662113498075,
+      "grad_norm": 0.25758981704711914,
+      "learning_rate": 5e-05,
+      "loss": 1.7289,
+      "step": 2654
+    },
+    {
+      "epoch": 0.32969079846020116,
+      "grad_norm": 0.2502439618110657,
+      "learning_rate": 5e-05,
+      "loss": 1.9411,
+      "step": 2655
+    },
+    {
+      "epoch": 0.32981497578542157,
+      "grad_norm": 0.273319810628891,
+      "learning_rate": 5e-05,
+      "loss": 1.8997,
+      "step": 2656
+    },
+    {
+      "epoch": 0.329939153110642,
+      "grad_norm": 0.2838249206542969,
+      "learning_rate": 5e-05,
+      "loss": 2.6379,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3300633304358624,
+      "grad_norm": 0.28816327452659607,
+      "learning_rate": 5e-05,
+      "loss": 2.399,
+      "step": 2658
+    },
+    {
+      "epoch": 0.3301875077610828,
+      "grad_norm": 0.2628127932548523,
+      "learning_rate": 5e-05,
+      "loss": 1.6436,
+      "step": 2659
+    },
+    {
+      "epoch": 0.33031168508630326,
+      "grad_norm": 0.2610509693622589,
+      "learning_rate": 5e-05,
+      "loss": 2.1179,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3304358624115237,
+      "grad_norm": 0.23995278775691986,
+      "learning_rate": 5e-05,
+      "loss": 1.9511,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3305600397367441,
+      "grad_norm": 0.244562029838562,
+      "learning_rate": 5e-05,
+      "loss": 1.8793,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3306842170619645,
+      "grad_norm": 0.5141323804855347,
+      "learning_rate": 5e-05,
+      "loss": 3.3328,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3308083943871849,
+      "grad_norm": 0.2750050127506256,
+      "learning_rate": 5e-05,
+      "loss": 1.8901,
+      "step": 2664
+    },
+    {
+      "epoch": 0.3309325717124053,
+      "grad_norm": 0.2849682867527008,
+      "learning_rate": 5e-05,
+      "loss": 2.6235,
+      "step": 2665
+    },
+    {
+      "epoch": 0.3310567490376257,
+      "grad_norm": 0.2689409852027893,
+      "learning_rate": 5e-05,
+      "loss": 1.7571,
+      "step": 2666
+    },
+    {
+      "epoch": 0.33118092636284613,
+      "grad_norm": 0.23531028628349304,
+      "learning_rate": 5e-05,
+      "loss": 1.7193,
+      "step": 2667
+    },
+    {
+      "epoch": 0.33130510368806654,
+      "grad_norm": 0.25513288378715515,
+      "learning_rate": 5e-05,
+      "loss": 1.7936,
+      "step": 2668
+    },
+    {
+      "epoch": 0.33142928101328695,
+      "grad_norm": 0.3676108121871948,
+      "learning_rate": 5e-05,
+      "loss": 2.3917,
+      "step": 2669
+    },
+    {
+      "epoch": 0.3315534583385074,
+      "grad_norm": 0.2444765716791153,
+      "learning_rate": 5e-05,
+      "loss": 1.7087,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3316776356637278,
+      "grad_norm": 0.25341418385505676,
+      "learning_rate": 5e-05,
+      "loss": 1.879,
+      "step": 2671
+    },
+    {
+      "epoch": 0.33180181298894823,
+      "grad_norm": 0.27764201164245605,
+      "learning_rate": 5e-05,
+      "loss": 1.8237,
+      "step": 2672
+    },
+    {
+      "epoch": 0.33192599031416864,
+      "grad_norm": 0.3166208267211914,
+      "learning_rate": 5e-05,
+      "loss": 2.9127,
+      "step": 2673
+    },
+    {
+      "epoch": 0.33205016763938905,
+      "grad_norm": 0.2800692319869995,
+      "learning_rate": 5e-05,
+      "loss": 1.941,
+      "step": 2674
+    },
+    {
+      "epoch": 0.33217434496460946,
+      "grad_norm": 0.22468514740467072,
+      "learning_rate": 5e-05,
+      "loss": 1.8493,
+      "step": 2675
+    },
+    {
+      "epoch": 0.33229852228982987,
+      "grad_norm": 0.27179422974586487,
+      "learning_rate": 5e-05,
+      "loss": 1.8602,
+      "step": 2676
+    },
+    {
+      "epoch": 0.3324226996150503,
+      "grad_norm": 0.2641296982765198,
+      "learning_rate": 5e-05,
+      "loss": 2.0543,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3325468769402707,
+      "grad_norm": 0.23599618673324585,
+      "learning_rate": 5e-05,
+      "loss": 1.8476,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3326710542654911,
+      "grad_norm": 0.2693142592906952,
+      "learning_rate": 5e-05,
+      "loss": 2.3955,
+      "step": 2679
+    },
+    {
+      "epoch": 0.33279523159071156,
+      "grad_norm": 0.32645127177238464,
+      "learning_rate": 5e-05,
+      "loss": 2.8382,
+      "step": 2680
+    },
+    {
+      "epoch": 0.33291940891593197,
+      "grad_norm": 0.3135913610458374,
+      "learning_rate": 5e-05,
+      "loss": 2.8269,
+      "step": 2681
+    },
+    {
+      "epoch": 0.3330435862411524,
+      "grad_norm": 0.2425813525915146,
+      "learning_rate": 5e-05,
+      "loss": 2.1296,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3331677635663728,
+      "grad_norm": 0.24746792018413544,
+      "learning_rate": 5e-05,
+      "loss": 1.9992,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3332919408915932,
+      "grad_norm": 0.2597034275531769,
+      "learning_rate": 5e-05,
+      "loss": 2.1613,
+      "step": 2684
+    },
+    {
+      "epoch": 0.3334161182168136,
+      "grad_norm": 0.25011998414993286,
+      "learning_rate": 5e-05,
+      "loss": 1.714,
+      "step": 2685
+    },
+    {
+      "epoch": 0.333540295542034,
+      "grad_norm": 0.25739601254463196,
+      "learning_rate": 5e-05,
+      "loss": 1.8186,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3336644728672544,
+      "grad_norm": 0.2503407597541809,
+      "learning_rate": 5e-05,
+      "loss": 2.0762,
+      "step": 2687
+    },
+    {
+      "epoch": 0.33378865019247483,
+      "grad_norm": 0.22531376779079437,
+      "learning_rate": 5e-05,
+      "loss": 1.7984,
+      "step": 2688
+    },
+    {
+      "epoch": 0.33391282751769524,
+      "grad_norm": 0.2579403817653656,
+      "learning_rate": 5e-05,
+      "loss": 2.1103,
+      "step": 2689
+    },
+    {
+      "epoch": 0.3340370048429157,
+      "grad_norm": 0.40785160660743713,
+      "learning_rate": 5e-05,
+      "loss": 3.0645,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3341611821681361,
+      "grad_norm": 0.2736961543560028,
+      "learning_rate": 5e-05,
+      "loss": 1.8795,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3342853594933565,
+      "grad_norm": 0.2668151259422302,
+      "learning_rate": 5e-05,
+      "loss": 1.9112,
+      "step": 2692
+    },
+    {
+      "epoch": 0.33440953681857694,
+      "grad_norm": 0.28688573837280273,
+      "learning_rate": 5e-05,
+      "loss": 2.3506,
+      "step": 2693
+    },
+    {
+      "epoch": 0.33453371414379734,
+      "grad_norm": 0.3707292675971985,
+      "learning_rate": 5e-05,
+      "loss": 2.82,
+      "step": 2694
+    },
+    {
+      "epoch": 0.33465789146901775,
+      "grad_norm": 0.31736403703689575,
+      "learning_rate": 5e-05,
+      "loss": 2.3905,
+      "step": 2695
+    },
+    {
+      "epoch": 0.33478206879423816,
+      "grad_norm": 0.277593731880188,
+      "learning_rate": 5e-05,
+      "loss": 2.0318,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3349062461194586,
+      "grad_norm": 0.2929210364818573,
+      "learning_rate": 5e-05,
+      "loss": 1.8892,
+      "step": 2697
+    },
+    {
+      "epoch": 0.335030423444679,
+      "grad_norm": 0.26633548736572266,
+      "learning_rate": 5e-05,
+      "loss": 1.9027,
+      "step": 2698
+    },
+    {
+      "epoch": 0.3351546007698994,
+      "grad_norm": 0.4682208001613617,
+      "learning_rate": 5e-05,
+      "loss": 2.5928,
+      "step": 2699
+    },
+    {
+      "epoch": 0.33527877809511986,
+      "grad_norm": 0.25876596570014954,
+      "learning_rate": 5e-05,
+      "loss": 1.7666,
+      "step": 2700
+    },
+    {
+      "epoch": 0.33540295542034027,
+      "grad_norm": 0.22832977771759033,
+      "learning_rate": 5e-05,
+      "loss": 1.7001,
+      "step": 2701
+    },
+    {
+      "epoch": 0.3355271327455607,
+      "grad_norm": 0.24146679043769836,
+      "learning_rate": 5e-05,
+      "loss": 1.8001,
+      "step": 2702
+    },
+    {
+      "epoch": 0.3356513100707811,
+      "grad_norm": 0.3086354732513428,
+      "learning_rate": 5e-05,
+      "loss": 2.4505,
+      "step": 2703
+    },
+    {
+      "epoch": 0.3357754873960015,
+      "grad_norm": 0.25541186332702637,
+      "learning_rate": 5e-05,
+      "loss": 1.9138,
+      "step": 2704
+    },
+    {
+      "epoch": 0.3358996647212219,
+      "grad_norm": 0.23455943167209625,
+      "learning_rate": 5e-05,
+      "loss": 1.6462,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3360238420464423,
+      "grad_norm": 0.2655121386051178,
+      "learning_rate": 5e-05,
+      "loss": 2.4973,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3361480193716627,
+      "grad_norm": 0.2663942873477936,
+      "learning_rate": 5e-05,
+      "loss": 1.8188,
+      "step": 2707
+    },
+    {
+      "epoch": 0.33627219669688313,
+      "grad_norm": 0.24645750224590302,
+      "learning_rate": 5e-05,
+      "loss": 1.8151,
+      "step": 2708
+    },
+    {
+      "epoch": 0.33639637402210354,
+      "grad_norm": 0.23590873181819916,
+      "learning_rate": 5e-05,
+      "loss": 1.847,
+      "step": 2709
+    },
+    {
+      "epoch": 0.336520551347324,
+      "grad_norm": 0.2445584535598755,
+      "learning_rate": 5e-05,
+      "loss": 1.9029,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3366447286725444,
+      "grad_norm": 0.23135963082313538,
+      "learning_rate": 5e-05,
+      "loss": 1.7899,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3367689059977648,
+      "grad_norm": 0.2910292446613312,
+      "learning_rate": 5e-05,
+      "loss": 2.4689,
+      "step": 2712
+    },
+    {
+      "epoch": 0.33689308332298523,
+      "grad_norm": 0.2866653800010681,
+      "learning_rate": 5e-05,
+      "loss": 2.1988,
+      "step": 2713
+    },
+    {
+      "epoch": 0.33701726064820564,
+      "grad_norm": 0.246674582362175,
+      "learning_rate": 5e-05,
+      "loss": 1.806,
+      "step": 2714
+    },
+    {
+      "epoch": 0.33714143797342605,
+      "grad_norm": 0.24526937305927277,
+      "learning_rate": 5e-05,
+      "loss": 1.6654,
+      "step": 2715
+    },
+    {
+      "epoch": 0.33726561529864646,
+      "grad_norm": 0.2305716723203659,
+      "learning_rate": 5e-05,
+      "loss": 1.806,
+      "step": 2716
+    },
+    {
+      "epoch": 0.33738979262386687,
+      "grad_norm": 0.36503860354423523,
+      "learning_rate": 5e-05,
+      "loss": 2.9692,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3375139699490873,
+      "grad_norm": 0.2545728385448456,
+      "learning_rate": 5e-05,
+      "loss": 1.7236,
+      "step": 2718
+    },
+    {
+      "epoch": 0.3376381472743077,
+      "grad_norm": 0.24277035892009735,
+      "learning_rate": 5e-05,
+      "loss": 1.6737,
+      "step": 2719
+    },
+    {
+      "epoch": 0.33776232459952815,
+      "grad_norm": 0.2667277753353119,
+      "learning_rate": 5e-05,
+      "loss": 1.9704,
+      "step": 2720
+    },
+    {
+      "epoch": 0.33788650192474856,
+      "grad_norm": 0.32830843329429626,
+      "learning_rate": 5e-05,
+      "loss": 2.5226,
+      "step": 2721
+    },
+    {
+      "epoch": 0.33801067924996897,
+      "grad_norm": 0.22645482420921326,
+      "learning_rate": 5e-05,
+      "loss": 1.6984,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3381348565751894,
+      "grad_norm": 0.2628045380115509,
+      "learning_rate": 5e-05,
+      "loss": 2.0692,
+      "step": 2723
+    },
+    {
+      "epoch": 0.3382590339004098,
+      "grad_norm": 0.2586096227169037,
+      "learning_rate": 5e-05,
+      "loss": 2.0922,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3383832112256302,
+      "grad_norm": 0.25160637497901917,
+      "learning_rate": 5e-05,
+      "loss": 1.8023,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3385073885508506,
+      "grad_norm": 0.2570488154888153,
+      "learning_rate": 5e-05,
+      "loss": 2.1305,
+      "step": 2726
+    },
+    {
+      "epoch": 0.338631565876071,
+      "grad_norm": 0.22564050555229187,
+      "learning_rate": 5e-05,
+      "loss": 1.6308,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3387557432012914,
+      "grad_norm": 0.27493584156036377,
+      "learning_rate": 5e-05,
+      "loss": 1.729,
+      "step": 2728
+    },
+    {
+      "epoch": 0.33887992052651184,
+      "grad_norm": 0.24422194063663483,
+      "learning_rate": 5e-05,
+      "loss": 1.9328,
+      "step": 2729
+    },
+    {
+      "epoch": 0.3390040978517323,
+      "grad_norm": 0.24400582909584045,
+      "learning_rate": 5e-05,
+      "loss": 2.0376,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3391282751769527,
+      "grad_norm": 0.25130966305732727,
+      "learning_rate": 5e-05,
+      "loss": 2.3058,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3392524525021731,
+      "grad_norm": 0.28701239824295044,
+      "learning_rate": 5e-05,
+      "loss": 2.3345,
+      "step": 2732
+    },
+    {
+      "epoch": 0.33937662982739353,
+      "grad_norm": 0.24303431808948517,
+      "learning_rate": 5e-05,
+      "loss": 2.074,
+      "step": 2733
+    },
+    {
+      "epoch": 0.33950080715261394,
+      "grad_norm": 0.26334697008132935,
+      "learning_rate": 5e-05,
+      "loss": 2.1504,
+      "step": 2734
+    },
+    {
+      "epoch": 0.33962498447783435,
+      "grad_norm": 0.2602726221084595,
+      "learning_rate": 5e-05,
+      "loss": 1.8967,
+      "step": 2735
+    },
+    {
+      "epoch": 0.33974916180305476,
+      "grad_norm": 0.26299959421157837,
+      "learning_rate": 5e-05,
+      "loss": 1.8993,
+      "step": 2736
+    },
+    {
+      "epoch": 0.33987333912827516,
+      "grad_norm": 0.23371700942516327,
+      "learning_rate": 5e-05,
+      "loss": 1.9863,
+      "step": 2737
+    },
+    {
+      "epoch": 0.3399975164534956,
+      "grad_norm": 0.2522105574607849,
+      "learning_rate": 5e-05,
+      "loss": 1.9815,
+      "step": 2738
+    },
+    {
+      "epoch": 0.340121693778716,
+      "grad_norm": 0.25059840083122253,
+      "learning_rate": 5e-05,
+      "loss": 1.7434,
+      "step": 2739
+    },
+    {
+      "epoch": 0.34024587110393645,
+      "grad_norm": 0.34855976700782776,
+      "learning_rate": 5e-05,
+      "loss": 2.2668,
+      "step": 2740
+    },
+    {
+      "epoch": 0.34037004842915686,
+      "grad_norm": 0.3817572593688965,
+      "learning_rate": 5e-05,
+      "loss": 2.8648,
+      "step": 2741
+    },
+    {
+      "epoch": 0.34049422575437727,
+      "grad_norm": 0.271109402179718,
+      "learning_rate": 5e-05,
+      "loss": 1.9262,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3406184030795977,
+      "grad_norm": 0.3439183235168457,
+      "learning_rate": 5e-05,
+      "loss": 2.9919,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3407425804048181,
+      "grad_norm": 0.3174643814563751,
+      "learning_rate": 5e-05,
+      "loss": 2.19,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3408667577300385,
+      "grad_norm": 0.4047298729419708,
+      "learning_rate": 5e-05,
+      "loss": 3.497,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3409909350552589,
+      "grad_norm": 0.2773181200027466,
+      "learning_rate": 5e-05,
+      "loss": 2.4118,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3411151123804793,
+      "grad_norm": 0.38656723499298096,
+      "learning_rate": 5e-05,
+      "loss": 2.938,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3412392897056997,
+      "grad_norm": 0.255911260843277,
+      "learning_rate": 5e-05,
+      "loss": 2.184,
+      "step": 2748
+    },
+    {
+      "epoch": 0.34136346703092013,
+      "grad_norm": 0.29141995310783386,
+      "learning_rate": 5e-05,
+      "loss": 2.1438,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3414876443561406,
+      "grad_norm": 0.2653254270553589,
+      "learning_rate": 5e-05,
+      "loss": 2.0359,
+      "step": 2750
+    },
+    {
+      "epoch": 0.341611821681361,
+      "grad_norm": 0.2267696112394333,
+      "learning_rate": 5e-05,
+      "loss": 1.839,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3417359990065814,
+      "grad_norm": 0.2552401125431061,
+      "learning_rate": 5e-05,
+      "loss": 2.0243,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3418601763318018,
+      "grad_norm": 0.24397550523281097,
+      "learning_rate": 5e-05,
+      "loss": 1.5411,
+      "step": 2753
+    },
+    {
+      "epoch": 0.34198435365702223,
+      "grad_norm": 0.2944405674934387,
+      "learning_rate": 5e-05,
+      "loss": 1.8792,
+      "step": 2754
+    },
+    {
+      "epoch": 0.34210853098224264,
+      "grad_norm": 0.2713506519794464,
+      "learning_rate": 5e-05,
+      "loss": 1.8701,
+      "step": 2755
+    },
+    {
+      "epoch": 0.34223270830746305,
+      "grad_norm": 0.2588428258895874,
+      "learning_rate": 5e-05,
+      "loss": 1.7006,
+      "step": 2756
+    },
+    {
+      "epoch": 0.34235688563268346,
+      "grad_norm": 0.32673966884613037,
+      "learning_rate": 5e-05,
+      "loss": 2.9641,
+      "step": 2757
+    },
+    {
+      "epoch": 0.34248106295790387,
+      "grad_norm": 0.29470935463905334,
+      "learning_rate": 5e-05,
+      "loss": 2.1214,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3426052402831243,
+      "grad_norm": 0.2547260820865631,
+      "learning_rate": 5e-05,
+      "loss": 1.7745,
+      "step": 2759
+    },
+    {
+      "epoch": 0.3427294176083447,
+      "grad_norm": 0.25234735012054443,
+      "learning_rate": 5e-05,
+      "loss": 1.9462,
+      "step": 2760
+    },
+    {
+      "epoch": 0.34285359493356515,
+      "grad_norm": 0.3391054570674896,
+      "learning_rate": 5e-05,
+      "loss": 2.7421,
+      "step": 2761
+    },
+    {
+      "epoch": 0.34297777225878556,
+      "grad_norm": 0.24231423437595367,
+      "learning_rate": 5e-05,
+      "loss": 1.7373,
+      "step": 2762
+    },
+    {
+      "epoch": 0.34310194958400597,
+      "grad_norm": 0.2596467435359955,
+      "learning_rate": 5e-05,
+      "loss": 2.0618,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3432261269092264,
+      "grad_norm": 0.24184861779212952,
+      "learning_rate": 5e-05,
+      "loss": 1.8985,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3433503042344468,
+      "grad_norm": 0.25749269127845764,
+      "learning_rate": 5e-05,
+      "loss": 2.2415,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3434744815596672,
+      "grad_norm": 0.23320280015468597,
+      "learning_rate": 5e-05,
+      "loss": 1.8246,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3435986588848876,
+      "grad_norm": 0.2523958384990692,
+      "learning_rate": 5e-05,
+      "loss": 1.854,
+      "step": 2767
+    },
+    {
+      "epoch": 0.343722836210108,
+      "grad_norm": 0.27736178040504456,
+      "learning_rate": 5e-05,
+      "loss": 2.3073,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3438470135353284,
+      "grad_norm": 0.22502301633358002,
+      "learning_rate": 5e-05,
+      "loss": 1.8195,
+      "step": 2769
+    },
+    {
+      "epoch": 0.34397119086054884,
+      "grad_norm": 0.2587378919124603,
+      "learning_rate": 5e-05,
+      "loss": 1.9351,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3440953681857693,
+      "grad_norm": 0.4828055500984192,
+      "learning_rate": 5e-05,
+      "loss": 2.6267,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3442195455109897,
+      "grad_norm": 0.41918686032295227,
+      "learning_rate": 5e-05,
+      "loss": 2.611,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3443437228362101,
+      "grad_norm": 0.23209670186042786,
+      "learning_rate": 5e-05,
+      "loss": 1.9922,
+      "step": 2773
+    },
+    {
+      "epoch": 0.34446790016143053,
+      "grad_norm": 0.22860324382781982,
+      "learning_rate": 5e-05,
+      "loss": 1.6626,
+      "step": 2774
+    },
+    {
+      "epoch": 0.34459207748665094,
+      "grad_norm": 0.2887635827064514,
+      "learning_rate": 5e-05,
+      "loss": 1.8943,
+      "step": 2775
+    },
+    {
+      "epoch": 0.34471625481187135,
+      "grad_norm": 0.29355379939079285,
+      "learning_rate": 5e-05,
+      "loss": 2.1005,
+      "step": 2776
+    },
+    {
+      "epoch": 0.34484043213709176,
+      "grad_norm": 0.38025107979774475,
+      "learning_rate": 5e-05,
+      "loss": 2.6319,
+      "step": 2777
+    },
+    {
+      "epoch": 0.34496460946231217,
+      "grad_norm": 0.23533344268798828,
+      "learning_rate": 5e-05,
+      "loss": 1.7306,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3450887867875326,
+      "grad_norm": 0.25997745990753174,
+      "learning_rate": 5e-05,
+      "loss": 1.9689,
+      "step": 2779
+    },
+    {
+      "epoch": 0.345212964112753,
+      "grad_norm": 0.2836317718029022,
+      "learning_rate": 5e-05,
+      "loss": 1.8641,
+      "step": 2780
+    },
+    {
+      "epoch": 0.34533714143797345,
+      "grad_norm": 0.28725767135620117,
+      "learning_rate": 5e-05,
+      "loss": 2.373,
+      "step": 2781
+    },
+    {
+      "epoch": 0.34546131876319386,
+      "grad_norm": 0.2814556956291199,
+      "learning_rate": 5e-05,
+      "loss": 2.1512,
+      "step": 2782
+    },
+    {
+      "epoch": 0.34558549608841427,
+      "grad_norm": 0.6949014067649841,
+      "learning_rate": 5e-05,
+      "loss": 2.4138,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3457096734136347,
+      "grad_norm": 0.24526333808898926,
+      "learning_rate": 5e-05,
+      "loss": 1.7468,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3458338507388551,
+      "grad_norm": 0.2844972312450409,
+      "learning_rate": 5e-05,
+      "loss": 1.88,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3459580280640755,
+      "grad_norm": 0.2742244005203247,
+      "learning_rate": 5e-05,
+      "loss": 1.7497,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3460822053892959,
+      "grad_norm": 0.34010806679725647,
+      "learning_rate": 5e-05,
+      "loss": 2.2237,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3462063827145163,
+      "grad_norm": 0.22840379178524017,
+      "learning_rate": 5e-05,
+      "loss": 1.4986,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3463305600397367,
+      "grad_norm": 0.3029632568359375,
+      "learning_rate": 5e-05,
+      "loss": 2.168,
+      "step": 2789
+    },
+    {
+      "epoch": 0.34645473736495713,
+      "grad_norm": 0.23877695202827454,
+      "learning_rate": 5e-05,
+      "loss": 1.7529,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3465789146901776,
+      "grad_norm": 0.2599693834781647,
+      "learning_rate": 5e-05,
+      "loss": 1.9903,
+      "step": 2791
+    },
+    {
+      "epoch": 0.346703092015398,
+      "grad_norm": 0.2574094533920288,
+      "learning_rate": 5e-05,
+      "loss": 1.8108,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3468272693406184,
+      "grad_norm": 0.24192070960998535,
+      "learning_rate": 5e-05,
+      "loss": 1.7852,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3469514466658388,
+      "grad_norm": 0.3200379014015198,
+      "learning_rate": 5e-05,
+      "loss": 2.4943,
+      "step": 2794
+    },
+    {
+      "epoch": 0.34707562399105923,
+      "grad_norm": 0.24337546527385712,
+      "learning_rate": 5e-05,
+      "loss": 1.7751,
+      "step": 2795
+    },
+    {
+      "epoch": 0.34719980131627964,
+      "grad_norm": 0.2670900225639343,
+      "learning_rate": 5e-05,
+      "loss": 1.9847,
+      "step": 2796
+    },
+    {
+      "epoch": 0.34732397864150005,
+      "grad_norm": 0.2614905536174774,
+      "learning_rate": 5e-05,
+      "loss": 1.8303,
+      "step": 2797
+    },
+    {
+      "epoch": 0.34744815596672046,
+      "grad_norm": 0.2653239667415619,
+      "learning_rate": 5e-05,
+      "loss": 1.7893,
+      "step": 2798
+    },
+    {
+      "epoch": 0.34757233329194087,
+      "grad_norm": 0.23396077752113342,
+      "learning_rate": 5e-05,
+      "loss": 1.8971,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3476965106171613,
+      "grad_norm": 0.24407877027988434,
+      "learning_rate": 5e-05,
+      "loss": 1.8667,
+      "step": 2800
+    },
+    {
+      "epoch": 0.34782068794238175,
+      "grad_norm": 0.26523271203041077,
+      "learning_rate": 5e-05,
+      "loss": 2.0719,
+      "step": 2801
+    },
+    {
+      "epoch": 0.34794486526760215,
+      "grad_norm": 0.279845654964447,
+      "learning_rate": 5e-05,
+      "loss": 2.0785,
+      "step": 2802
+    },
+    {
+      "epoch": 0.34806904259282256,
+      "grad_norm": 0.3161846101284027,
+      "learning_rate": 5e-05,
+      "loss": 2.1301,
+      "step": 2803
+    },
+    {
+      "epoch": 0.348193219918043,
+      "grad_norm": 0.2765732407569885,
+      "learning_rate": 5e-05,
+      "loss": 2.0976,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3483173972432634,
+      "grad_norm": 0.33240991830825806,
+      "learning_rate": 5e-05,
+      "loss": 2.2428,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3484415745684838,
+      "grad_norm": 0.24005700647830963,
+      "learning_rate": 5e-05,
+      "loss": 1.7809,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3485657518937042,
+      "grad_norm": 0.2592582404613495,
+      "learning_rate": 5e-05,
+      "loss": 1.9745,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3486899292189246,
+      "grad_norm": 0.2999490797519684,
+      "learning_rate": 5e-05,
+      "loss": 2.6955,
+      "step": 2808
+    },
+    {
+      "epoch": 0.348814106544145,
+      "grad_norm": 0.259401798248291,
+      "learning_rate": 5e-05,
+      "loss": 1.8179,
+      "step": 2809
+    },
+    {
+      "epoch": 0.34893828386936543,
+      "grad_norm": 0.24388451874256134,
+      "learning_rate": 5e-05,
+      "loss": 1.8819,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3490624611945859,
+      "grad_norm": 0.27560216188430786,
+      "learning_rate": 5e-05,
+      "loss": 1.8509,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3491866385198063,
+      "grad_norm": 0.373429536819458,
+      "learning_rate": 5e-05,
+      "loss": 3.1491,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3493108158450267,
+      "grad_norm": 0.2791976034641266,
+      "learning_rate": 5e-05,
+      "loss": 2.2035,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3494349931702471,
+      "grad_norm": 0.2509307563304901,
+      "learning_rate": 5e-05,
+      "loss": 1.9645,
+      "step": 2814
+    },
+    {
+      "epoch": 0.34955917049546753,
+      "grad_norm": 0.2831667959690094,
+      "learning_rate": 5e-05,
+      "loss": 2.1522,
+      "step": 2815
+    },
+    {
+      "epoch": 0.34968334782068794,
+      "grad_norm": 0.28687188029289246,
+      "learning_rate": 5e-05,
+      "loss": 2.1085,
+      "step": 2816
+    },
+    {
+      "epoch": 0.34980752514590835,
+      "grad_norm": 0.3539823591709137,
+      "learning_rate": 5e-05,
+      "loss": 2.9421,
+      "step": 2817
+    },
+    {
+      "epoch": 0.34993170247112876,
+      "grad_norm": 0.2686798572540283,
+      "learning_rate": 5e-05,
+      "loss": 1.8668,
+      "step": 2818
+    },
+    {
+      "epoch": 0.35005587979634917,
+      "grad_norm": 0.24746398627758026,
+      "learning_rate": 5e-05,
+      "loss": 1.6538,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3501800571215696,
+      "grad_norm": 0.24276474118232727,
+      "learning_rate": 5e-05,
+      "loss": 1.9963,
+      "step": 2820
+    },
+    {
+      "epoch": 0.35030423444679004,
+      "grad_norm": 0.2582096755504608,
+      "learning_rate": 5e-05,
+      "loss": 2.192,
+      "step": 2821
+    },
+    {
+      "epoch": 0.35042841177201045,
+      "grad_norm": 0.2585619390010834,
+      "learning_rate": 5e-05,
+      "loss": 2.5368,
+      "step": 2822
+    },
+    {
+      "epoch": 0.35055258909723086,
+      "grad_norm": 0.27060192823410034,
+      "learning_rate": 5e-05,
+      "loss": 2.0067,
+      "step": 2823
+    },
+    {
+      "epoch": 0.35067676642245127,
+      "grad_norm": 0.3271196782588959,
+      "learning_rate": 5e-05,
+      "loss": 3.0698,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3508009437476717,
+      "grad_norm": 0.27114972472190857,
+      "learning_rate": 5e-05,
+      "loss": 1.9461,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3509251210728921,
+      "grad_norm": 0.2507452666759491,
+      "learning_rate": 5e-05,
+      "loss": 1.784,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3510492983981125,
+      "grad_norm": 0.25736528635025024,
+      "learning_rate": 5e-05,
+      "loss": 1.927,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3511734757233329,
+      "grad_norm": 0.25758883357048035,
+      "learning_rate": 5e-05,
+      "loss": 1.9181,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3512976530485533,
+      "grad_norm": 0.2678426206111908,
+      "learning_rate": 5e-05,
+      "loss": 1.8063,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3514218303737737,
+      "grad_norm": 0.220537930727005,
+      "learning_rate": 5e-05,
+      "loss": 1.8122,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3515460076989942,
+      "grad_norm": 0.23005208373069763,
+      "learning_rate": 5e-05,
+      "loss": 1.892,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3516701850242146,
+      "grad_norm": 0.3395572304725647,
+      "learning_rate": 5e-05,
+      "loss": 2.6256,
+      "step": 2832
+    },
+    {
+      "epoch": 0.351794362349435,
+      "grad_norm": 0.26800087094306946,
+      "learning_rate": 5e-05,
+      "loss": 2.1189,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3519185396746554,
+      "grad_norm": 0.23922783136367798,
+      "learning_rate": 5e-05,
+      "loss": 1.9573,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3520427169998758,
+      "grad_norm": 0.45247089862823486,
+      "learning_rate": 5e-05,
+      "loss": 1.9869,
+      "step": 2835
+    },
+    {
+      "epoch": 0.35216689432509624,
+      "grad_norm": 0.23645295202732086,
+      "learning_rate": 5e-05,
+      "loss": 1.5156,
+      "step": 2836
+    },
+    {
+      "epoch": 0.35229107165031665,
+      "grad_norm": 0.5739084482192993,
+      "learning_rate": 5e-05,
+      "loss": 3.5006,
+      "step": 2837
+    },
+    {
+      "epoch": 0.35241524897553705,
+      "grad_norm": 0.3439748287200928,
+      "learning_rate": 5e-05,
+      "loss": 2.6406,
+      "step": 2838
+    },
+    {
+      "epoch": 0.35253942630075746,
+      "grad_norm": 0.4188997149467468,
+      "learning_rate": 5e-05,
+      "loss": 2.4354,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3526636036259779,
+      "grad_norm": 0.2544313967227936,
+      "learning_rate": 5e-05,
+      "loss": 1.738,
+      "step": 2840
+    },
+    {
+      "epoch": 0.35278778095119834,
+      "grad_norm": 0.49748292565345764,
+      "learning_rate": 5e-05,
+      "loss": 3.0784,
+      "step": 2841
+    },
+    {
+      "epoch": 0.35291195827641875,
+      "grad_norm": 0.26805004477500916,
+      "learning_rate": 5e-05,
+      "loss": 2.0459,
+      "step": 2842
+    },
+    {
+      "epoch": 0.35303613560163916,
+      "grad_norm": 0.24519136548042297,
+      "learning_rate": 5e-05,
+      "loss": 1.941,
+      "step": 2843
+    },
+    {
+      "epoch": 0.35316031292685957,
+      "grad_norm": 0.2823317050933838,
+      "learning_rate": 5e-05,
+      "loss": 1.7439,
+      "step": 2844
+    },
+    {
+      "epoch": 0.35328449025208,
+      "grad_norm": 0.29722580313682556,
+      "learning_rate": 5e-05,
+      "loss": 2.2139,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3534086675773004,
+      "grad_norm": 0.2921779751777649,
+      "learning_rate": 5e-05,
+      "loss": 1.9884,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3535328449025208,
+      "grad_norm": 0.2357889711856842,
+      "learning_rate": 5e-05,
+      "loss": 1.7237,
+      "step": 2847
+    },
+    {
+      "epoch": 0.3536570222277412,
+      "grad_norm": 0.23678836226463318,
+      "learning_rate": 5e-05,
+      "loss": 1.9301,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3537811995529616,
+      "grad_norm": 0.36033767461776733,
+      "learning_rate": 5e-05,
+      "loss": 2.1311,
+      "step": 2849
+    },
+    {
+      "epoch": 0.353905376878182,
+      "grad_norm": 0.2586762011051178,
+      "learning_rate": 5e-05,
+      "loss": 1.7927,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3540295542034025,
+      "grad_norm": 0.3213244676589966,
+      "learning_rate": 5e-05,
+      "loss": 2.9524,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3541537315286229,
+      "grad_norm": 0.28174933791160583,
+      "learning_rate": 5e-05,
+      "loss": 2.695,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3542779088538433,
+      "grad_norm": 0.2555718421936035,
+      "learning_rate": 5e-05,
+      "loss": 1.774,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3544020861790637,
+      "grad_norm": 0.25043463706970215,
+      "learning_rate": 5e-05,
+      "loss": 1.9787,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3545262635042841,
+      "grad_norm": 0.31411176919937134,
+      "learning_rate": 5e-05,
+      "loss": 2.3861,
+      "step": 2855
+    },
+    {
+      "epoch": 0.35465044082950453,
+      "grad_norm": 0.2472156286239624,
+      "learning_rate": 5e-05,
+      "loss": 1.9671,
+      "step": 2856
+    },
+    {
+      "epoch": 0.35477461815472494,
+      "grad_norm": 0.2894991934299469,
+      "learning_rate": 5e-05,
+      "loss": 1.6472,
+      "step": 2857
+    },
+    {
+      "epoch": 0.35489879547994535,
+      "grad_norm": 0.2334747165441513,
+      "learning_rate": 5e-05,
+      "loss": 1.7737,
+      "step": 2858
+    },
+    {
+      "epoch": 0.35502297280516576,
+      "grad_norm": 0.39418351650238037,
+      "learning_rate": 5e-05,
+      "loss": 2.671,
+      "step": 2859
+    },
+    {
+      "epoch": 0.35514715013038617,
+      "grad_norm": 0.25583016872406006,
+      "learning_rate": 5e-05,
+      "loss": 2.1107,
+      "step": 2860
+    },
+    {
+      "epoch": 0.35527132745560663,
+      "grad_norm": 0.38431593775749207,
+      "learning_rate": 5e-05,
+      "loss": 3.1659,
+      "step": 2861
+    },
+    {
+      "epoch": 0.35539550478082704,
+      "grad_norm": 0.2511475086212158,
+      "learning_rate": 5e-05,
+      "loss": 1.7431,
+      "step": 2862
+    },
+    {
+      "epoch": 0.35551968210604745,
+      "grad_norm": 0.31024259328842163,
+      "learning_rate": 5e-05,
+      "loss": 1.8739,
+      "step": 2863
+    },
+    {
+      "epoch": 0.35564385943126786,
+      "grad_norm": 0.29655084013938904,
+      "learning_rate": 5e-05,
+      "loss": 1.8697,
+      "step": 2864
+    },
+    {
+      "epoch": 0.35576803675648827,
+      "grad_norm": 0.2990925908088684,
+      "learning_rate": 5e-05,
+      "loss": 1.8586,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3558922140817087,
+      "grad_norm": 0.23770806193351746,
+      "learning_rate": 5e-05,
+      "loss": 1.843,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3560163914069291,
+      "grad_norm": 0.27660104632377625,
+      "learning_rate": 5e-05,
+      "loss": 2.0953,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3561405687321495,
+      "grad_norm": 0.2534283399581909,
+      "learning_rate": 5e-05,
+      "loss": 1.5889,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3562647460573699,
+      "grad_norm": 0.26595988869667053,
+      "learning_rate": 5e-05,
+      "loss": 2.1211,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3563889233825903,
+      "grad_norm": 0.2988552153110504,
+      "learning_rate": 5e-05,
+      "loss": 1.921,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3565131007078107,
+      "grad_norm": 0.2611449062824249,
+      "learning_rate": 5e-05,
+      "loss": 1.8725,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3566372780330312,
+      "grad_norm": 0.22860753536224365,
+      "learning_rate": 5e-05,
+      "loss": 1.806,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3567614553582516,
+      "grad_norm": 0.2686806321144104,
+      "learning_rate": 5e-05,
+      "loss": 1.8902,
+      "step": 2873
+    },
+    {
+      "epoch": 0.356885632683472,
+      "grad_norm": 0.23576751351356506,
+      "learning_rate": 5e-05,
+      "loss": 1.8558,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3570098100086924,
+      "grad_norm": 0.29896944761276245,
+      "learning_rate": 5e-05,
+      "loss": 2.1818,
+      "step": 2875
+    },
+    {
+      "epoch": 0.35713398733391283,
+      "grad_norm": 0.2467663735151291,
+      "learning_rate": 5e-05,
+      "loss": 1.6375,
+      "step": 2876
+    },
+    {
+      "epoch": 0.35725816465913324,
+      "grad_norm": 0.26486945152282715,
+      "learning_rate": 5e-05,
+      "loss": 2.0852,
+      "step": 2877
+    },
+    {
+      "epoch": 0.35738234198435365,
+      "grad_norm": 0.3457930088043213,
+      "learning_rate": 5e-05,
+      "loss": 2.5569,
+      "step": 2878
+    },
+    {
+      "epoch": 0.35750651930957406,
+      "grad_norm": 0.30438005924224854,
+      "learning_rate": 5e-05,
+      "loss": 2.4509,
+      "step": 2879
+    },
+    {
+      "epoch": 0.35763069663479446,
+      "grad_norm": 0.2636890411376953,
+      "learning_rate": 5e-05,
+      "loss": 2.0939,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3577548739600149,
+      "grad_norm": 0.260224312543869,
+      "learning_rate": 5e-05,
+      "loss": 2.107,
+      "step": 2881
+    },
+    {
+      "epoch": 0.35787905128523534,
+      "grad_norm": 0.47993791103363037,
+      "learning_rate": 5e-05,
+      "loss": 2.8968,
+      "step": 2882
+    },
+    {
+      "epoch": 0.35800322861045575,
+      "grad_norm": 0.3107258975505829,
+      "learning_rate": 5e-05,
+      "loss": 2.4332,
+      "step": 2883
+    },
+    {
+      "epoch": 0.35812740593567616,
+      "grad_norm": 0.24944131076335907,
+      "learning_rate": 5e-05,
+      "loss": 1.9468,
+      "step": 2884
+    },
+    {
+      "epoch": 0.35825158326089657,
+      "grad_norm": 0.30719128251075745,
+      "learning_rate": 5e-05,
+      "loss": 2.6837,
+      "step": 2885
+    },
+    {
+      "epoch": 0.358375760586117,
+      "grad_norm": 0.3071680963039398,
+      "learning_rate": 5e-05,
+      "loss": 2.5383,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3584999379113374,
+      "grad_norm": 0.33120197057724,
+      "learning_rate": 5e-05,
+      "loss": 2.2859,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3586241152365578,
+      "grad_norm": 0.22975130379199982,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3587482925617782,
+      "grad_norm": 0.29023388028144836,
+      "learning_rate": 5e-05,
+      "loss": 1.9831,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3588724698869986,
+      "grad_norm": 0.28600960969924927,
+      "learning_rate": 5e-05,
+      "loss": 2.3528,
+      "step": 2890
+    },
+    {
+      "epoch": 0.358996647212219,
+      "grad_norm": 0.2461761087179184,
+      "learning_rate": 5e-05,
+      "loss": 1.7646,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3591208245374395,
+      "grad_norm": 0.3595605492591858,
+      "learning_rate": 5e-05,
+      "loss": 2.4973,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3592450018626599,
+      "grad_norm": 0.25128260254859924,
+      "learning_rate": 5e-05,
+      "loss": 2.0308,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3593691791878803,
+      "grad_norm": 0.3829328119754791,
+      "learning_rate": 5e-05,
+      "loss": 3.1022,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3594933565131007,
+      "grad_norm": 0.2549037039279938,
+      "learning_rate": 5e-05,
+      "loss": 2.2065,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3596175338383211,
+      "grad_norm": 0.2445104420185089,
+      "learning_rate": 5e-05,
+      "loss": 1.8553,
+      "step": 2896
+    },
+    {
+      "epoch": 0.35974171116354153,
+      "grad_norm": 0.25458046793937683,
+      "learning_rate": 5e-05,
+      "loss": 1.8411,
+      "step": 2897
+    },
+    {
+      "epoch": 0.35986588848876194,
+      "grad_norm": 0.34245961904525757,
+      "learning_rate": 5e-05,
+      "loss": 2.0043,
+      "step": 2898
+    },
+    {
+      "epoch": 0.35999006581398235,
+      "grad_norm": 0.2767278552055359,
+      "learning_rate": 5e-05,
+      "loss": 2.3841,
+      "step": 2899
+    },
+    {
+      "epoch": 0.36011424313920276,
+      "grad_norm": 0.27056533098220825,
+      "learning_rate": 5e-05,
+      "loss": 1.9964,
+      "step": 2900
+    },
+    {
+      "epoch": 0.36023842046442317,
+      "grad_norm": 0.24606953561306,
+      "learning_rate": 5e-05,
+      "loss": 1.6673,
+      "step": 2901
+    },
+    {
+      "epoch": 0.36036259778964363,
+      "grad_norm": 0.25140783190727234,
+      "learning_rate": 5e-05,
+      "loss": 1.7711,
+      "step": 2902
+    },
+    {
+      "epoch": 0.36048677511486404,
+      "grad_norm": 0.25976094603538513,
+      "learning_rate": 5e-05,
+      "loss": 1.9651,
+      "step": 2903
+    },
+    {
+      "epoch": 0.36061095244008445,
+      "grad_norm": 0.30861905217170715,
+      "learning_rate": 5e-05,
+      "loss": 2.4284,
+      "step": 2904
+    },
+    {
+      "epoch": 0.36073512976530486,
+      "grad_norm": 0.2554759681224823,
+      "learning_rate": 5e-05,
+      "loss": 1.6431,
+      "step": 2905
+    },
+    {
+      "epoch": 0.36085930709052527,
+      "grad_norm": 0.45448505878448486,
+      "learning_rate": 5e-05,
+      "loss": 2.5967,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3609834844157457,
+      "grad_norm": 0.24251489341259003,
+      "learning_rate": 5e-05,
+      "loss": 1.9179,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3611076617409661,
+      "grad_norm": 0.28351014852523804,
+      "learning_rate": 5e-05,
+      "loss": 2.3383,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3612318390661865,
+      "grad_norm": 0.3094342052936554,
+      "learning_rate": 5e-05,
+      "loss": 2.4563,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3613560163914069,
+      "grad_norm": 0.25680920481681824,
+      "learning_rate": 5e-05,
+      "loss": 1.774,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3614801937166273,
+      "grad_norm": 0.25995075702667236,
+      "learning_rate": 5e-05,
+      "loss": 1.9433,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3616043710418478,
+      "grad_norm": 0.2613772749900818,
+      "learning_rate": 5e-05,
+      "loss": 2.2555,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3617285483670682,
+      "grad_norm": 0.2489282786846161,
+      "learning_rate": 5e-05,
+      "loss": 1.937,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3618527256922886,
+      "grad_norm": 0.23898090422153473,
+      "learning_rate": 5e-05,
+      "loss": 1.5986,
+      "step": 2914
+    },
+    {
+      "epoch": 0.361976903017509,
+      "grad_norm": 0.23500391840934753,
+      "learning_rate": 5e-05,
+      "loss": 1.8752,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3621010803427294,
+      "grad_norm": 0.24279162287712097,
+      "learning_rate": 5e-05,
+      "loss": 1.7808,
+      "step": 2916
+    },
+    {
+      "epoch": 0.36222525766794983,
+      "grad_norm": 0.24957755208015442,
+      "learning_rate": 5e-05,
+      "loss": 2.1986,
+      "step": 2917
+    },
+    {
+      "epoch": 0.36234943499317024,
+      "grad_norm": 0.40013664960861206,
+      "learning_rate": 5e-05,
+      "loss": 2.8486,
+      "step": 2918
+    },
+    {
+      "epoch": 0.36247361231839065,
+      "grad_norm": 0.27326327562332153,
+      "learning_rate": 5e-05,
+      "loss": 2.5734,
+      "step": 2919
+    },
+    {
+      "epoch": 0.36259778964361106,
+      "grad_norm": 0.2527517080307007,
+      "learning_rate": 5e-05,
+      "loss": 1.9826,
+      "step": 2920
+    },
+    {
+      "epoch": 0.36272196696883147,
+      "grad_norm": 0.25144174695014954,
+      "learning_rate": 5e-05,
+      "loss": 1.7099,
+      "step": 2921
+    },
+    {
+      "epoch": 0.36284614429405193,
+      "grad_norm": 0.26282787322998047,
+      "learning_rate": 5e-05,
+      "loss": 1.723,
+      "step": 2922
+    },
+    {
+      "epoch": 0.36297032161927234,
+      "grad_norm": 0.24631480872631073,
+      "learning_rate": 5e-05,
+      "loss": 2.0645,
+      "step": 2923
+    },
+    {
+      "epoch": 0.36309449894449275,
+      "grad_norm": 0.2382902204990387,
+      "learning_rate": 5e-05,
+      "loss": 1.8514,
+      "step": 2924
+    },
+    {
+      "epoch": 0.36321867626971316,
+      "grad_norm": 0.2269415706396103,
+      "learning_rate": 5e-05,
+      "loss": 1.8595,
+      "step": 2925
+    },
+    {
+      "epoch": 0.36334285359493357,
+      "grad_norm": 0.3079279065132141,
+      "learning_rate": 5e-05,
+      "loss": 2.2139,
+      "step": 2926
+    },
+    {
+      "epoch": 0.363467030920154,
+      "grad_norm": 0.25768083333969116,
+      "learning_rate": 5e-05,
+      "loss": 1.7746,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3635912082453744,
+      "grad_norm": 0.27373677492141724,
+      "learning_rate": 5e-05,
+      "loss": 2.4941,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3637153855705948,
+      "grad_norm": 0.23796309530735016,
+      "learning_rate": 5e-05,
+      "loss": 1.8657,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3638395628958152,
+      "grad_norm": 0.24157953262329102,
+      "learning_rate": 5e-05,
+      "loss": 1.927,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3639637402210356,
+      "grad_norm": 0.23541103303432465,
+      "learning_rate": 5e-05,
+      "loss": 1.9038,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3640879175462561,
+      "grad_norm": 0.29722172021865845,
+      "learning_rate": 5e-05,
+      "loss": 2.4113,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3642120948714765,
+      "grad_norm": 0.4047267436981201,
+      "learning_rate": 5e-05,
+      "loss": 2.958,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3643362721966969,
+      "grad_norm": 0.25163036584854126,
+      "learning_rate": 5e-05,
+      "loss": 2.0361,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3644604495219173,
+      "grad_norm": 0.22635877132415771,
+      "learning_rate": 5e-05,
+      "loss": 1.681,
+      "step": 2935
+    },
+    {
+      "epoch": 0.3645846268471377,
+      "grad_norm": 0.23741143941879272,
+      "learning_rate": 5e-05,
+      "loss": 1.7251,
+      "step": 2936
+    },
+    {
+      "epoch": 0.3647088041723581,
+      "grad_norm": 0.28617551922798157,
+      "learning_rate": 5e-05,
+      "loss": 2.1297,
+      "step": 2937
+    },
+    {
+      "epoch": 0.36483298149757853,
+      "grad_norm": 0.27699974179267883,
+      "learning_rate": 5e-05,
+      "loss": 2.1108,
+      "step": 2938
+    },
+    {
+      "epoch": 0.36495715882279894,
+      "grad_norm": 0.22934754192829132,
+      "learning_rate": 5e-05,
+      "loss": 1.5672,
+      "step": 2939
+    },
+    {
+      "epoch": 0.36508133614801935,
+      "grad_norm": 0.290359228849411,
+      "learning_rate": 5e-05,
+      "loss": 2.3294,
+      "step": 2940
+    },
+    {
+      "epoch": 0.36520551347323976,
+      "grad_norm": 0.23602132499217987,
+      "learning_rate": 5e-05,
+      "loss": 1.7783,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3653296907984602,
+      "grad_norm": 0.358762264251709,
+      "learning_rate": 5e-05,
+      "loss": 2.8368,
+      "step": 2942
+    },
+    {
+      "epoch": 0.36545386812368064,
+      "grad_norm": 0.22189496457576752,
+      "learning_rate": 5e-05,
+      "loss": 1.6123,
+      "step": 2943
+    },
+    {
+      "epoch": 0.36557804544890105,
+      "grad_norm": 0.2998792231082916,
+      "learning_rate": 5e-05,
+      "loss": 2.1474,
+      "step": 2944
+    },
+    {
+      "epoch": 0.36570222277412145,
+      "grad_norm": 0.5736348628997803,
+      "learning_rate": 5e-05,
+      "loss": 3.4069,
+      "step": 2945
+    },
+    {
+      "epoch": 0.36582640009934186,
+      "grad_norm": 0.2841154932975769,
+      "learning_rate": 5e-05,
+      "loss": 2.2049,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3659505774245623,
+      "grad_norm": 0.23865941166877747,
+      "learning_rate": 5e-05,
+      "loss": 1.7721,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3660747547497827,
+      "grad_norm": 0.22311750054359436,
+      "learning_rate": 5e-05,
+      "loss": 1.713,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3661989320750031,
+      "grad_norm": 0.25039970874786377,
+      "learning_rate": 5e-05,
+      "loss": 1.9432,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3663231094002235,
+      "grad_norm": 0.3837128281593323,
+      "learning_rate": 5e-05,
+      "loss": 2.735,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3664472867254439,
+      "grad_norm": 0.23523934185504913,
+      "learning_rate": 5e-05,
+      "loss": 1.7337,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3665714640506644,
+      "grad_norm": 0.24660135805606842,
+      "learning_rate": 5e-05,
+      "loss": 2.1519,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3666956413758848,
+      "grad_norm": 0.23488986492156982,
+      "learning_rate": 5e-05,
+      "loss": 1.8632,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3668198187011052,
+      "grad_norm": 0.24533359706401825,
+      "learning_rate": 5e-05,
+      "loss": 1.8495,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3669439960263256,
+      "grad_norm": 0.22848857939243317,
+      "learning_rate": 5e-05,
+      "loss": 1.649,
+      "step": 2955
+    },
+    {
+      "epoch": 0.367068173351546,
+      "grad_norm": 0.29741668701171875,
+      "learning_rate": 5e-05,
+      "loss": 2.4278,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3671923506767664,
+      "grad_norm": 0.2609666883945465,
+      "learning_rate": 5e-05,
+      "loss": 1.8476,
+      "step": 2957
+    },
+    {
+      "epoch": 0.36731652800198683,
+      "grad_norm": 0.23846489191055298,
+      "learning_rate": 5e-05,
+      "loss": 1.7979,
+      "step": 2958
+    },
+    {
+      "epoch": 0.36744070532720724,
+      "grad_norm": 0.2519788444042206,
+      "learning_rate": 5e-05,
+      "loss": 1.9826,
+      "step": 2959
+    },
+    {
+      "epoch": 0.36756488265242765,
+      "grad_norm": 0.2400158792734146,
+      "learning_rate": 5e-05,
+      "loss": 2.1669,
+      "step": 2960
+    },
+    {
+      "epoch": 0.36768905997764806,
+      "grad_norm": 0.2306850552558899,
+      "learning_rate": 5e-05,
+      "loss": 1.7329,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3678132373028685,
+      "grad_norm": 0.30840304493904114,
+      "learning_rate": 5e-05,
+      "loss": 2.4956,
+      "step": 2962
+    },
+    {
+      "epoch": 0.36793741462808893,
+      "grad_norm": 0.24921832978725433,
+      "learning_rate": 5e-05,
+      "loss": 1.8325,
+      "step": 2963
+    },
+    {
+      "epoch": 0.36806159195330934,
+      "grad_norm": 0.39718326926231384,
+      "learning_rate": 5e-05,
+      "loss": 2.9512,
+      "step": 2964
+    },
+    {
+      "epoch": 0.36818576927852975,
+      "grad_norm": 0.2490706890821457,
+      "learning_rate": 5e-05,
+      "loss": 1.9539,
+      "step": 2965
+    },
+    {
+      "epoch": 0.36830994660375016,
+      "grad_norm": 0.28754961490631104,
+      "learning_rate": 5e-05,
+      "loss": 2.2399,
+      "step": 2966
+    },
+    {
+      "epoch": 0.36843412392897057,
+      "grad_norm": 0.23727400600910187,
+      "learning_rate": 5e-05,
+      "loss": 1.5687,
+      "step": 2967
+    },
+    {
+      "epoch": 0.368558301254191,
+      "grad_norm": 0.305165559053421,
+      "learning_rate": 5e-05,
+      "loss": 1.9375,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3686824785794114,
+      "grad_norm": 0.28188270330429077,
+      "learning_rate": 5e-05,
+      "loss": 2.0947,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3688066559046318,
+      "grad_norm": 0.26603934168815613,
+      "learning_rate": 5e-05,
+      "loss": 2.0655,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3689308332298522,
+      "grad_norm": 0.2558739483356476,
+      "learning_rate": 5e-05,
+      "loss": 1.9905,
+      "step": 2971
+    },
+    {
+      "epoch": 0.36905501055507267,
+      "grad_norm": 0.25798073410987854,
+      "learning_rate": 5e-05,
+      "loss": 1.7421,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3691791878802931,
+      "grad_norm": 0.2610643804073334,
+      "learning_rate": 5e-05,
+      "loss": 1.8458,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3693033652055135,
+      "grad_norm": 0.2506898045539856,
+      "learning_rate": 5e-05,
+      "loss": 1.6336,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3694275425307339,
+      "grad_norm": 0.23491260409355164,
+      "learning_rate": 5e-05,
+      "loss": 1.9506,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3695517198559543,
+      "grad_norm": 0.35647445917129517,
+      "learning_rate": 5e-05,
+      "loss": 2.8191,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3696758971811747,
+      "grad_norm": 0.2702166438102722,
+      "learning_rate": 5e-05,
+      "loss": 2.1857,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3698000745063951,
+      "grad_norm": 0.33380377292633057,
+      "learning_rate": 5e-05,
+      "loss": 2.4216,
+      "step": 2978
+    },
+    {
+      "epoch": 0.36992425183161554,
+      "grad_norm": 0.24738530814647675,
+      "learning_rate": 5e-05,
+      "loss": 1.8974,
+      "step": 2979
+    },
+    {
+      "epoch": 0.37004842915683595,
+      "grad_norm": 0.2719271779060364,
+      "learning_rate": 5e-05,
+      "loss": 2.2517,
+      "step": 2980
+    },
+    {
+      "epoch": 0.37017260648205635,
+      "grad_norm": 0.29274439811706543,
+      "learning_rate": 5e-05,
+      "loss": 2.6857,
+      "step": 2981
+    },
+    {
+      "epoch": 0.37029678380727676,
+      "grad_norm": 0.23504342138767242,
+      "learning_rate": 5e-05,
+      "loss": 1.8174,
+      "step": 2982
+    },
+    {
+      "epoch": 0.37042096113249723,
+      "grad_norm": 0.26716282963752747,
+      "learning_rate": 5e-05,
+      "loss": 2.2242,
+      "step": 2983
+    },
+    {
+      "epoch": 0.37054513845771764,
+      "grad_norm": 0.2464313805103302,
+      "learning_rate": 5e-05,
+      "loss": 1.8624,
+      "step": 2984
+    },
+    {
+      "epoch": 0.37066931578293805,
+      "grad_norm": 0.2445826530456543,
+      "learning_rate": 5e-05,
+      "loss": 1.9853,
+      "step": 2985
+    },
+    {
+      "epoch": 0.37079349310815846,
+      "grad_norm": 0.22842368483543396,
+      "learning_rate": 5e-05,
+      "loss": 1.593,
+      "step": 2986
+    },
+    {
+      "epoch": 0.37091767043337887,
+      "grad_norm": 0.34224873781204224,
+      "learning_rate": 5e-05,
+      "loss": 2.9265,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3710418477585993,
+      "grad_norm": 0.24113371968269348,
+      "learning_rate": 5e-05,
+      "loss": 1.9255,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3711660250838197,
+      "grad_norm": 0.2746451497077942,
+      "learning_rate": 5e-05,
+      "loss": 2.3531,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3712902024090401,
+      "grad_norm": 0.38261857628822327,
+      "learning_rate": 5e-05,
+      "loss": 2.7068,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3714143797342605,
+      "grad_norm": 0.27590250968933105,
+      "learning_rate": 5e-05,
+      "loss": 3.0103,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3715385570594809,
+      "grad_norm": 0.2449302077293396,
+      "learning_rate": 5e-05,
+      "loss": 1.8079,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3716627343847014,
+      "grad_norm": 0.27541065216064453,
+      "learning_rate": 5e-05,
+      "loss": 2.1987,
+      "step": 2993
+    },
+    {
+      "epoch": 0.3717869117099218,
+      "grad_norm": 0.2601505517959595,
+      "learning_rate": 5e-05,
+      "loss": 2.1561,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3719110890351422,
+      "grad_norm": 0.26446446776390076,
+      "learning_rate": 5e-05,
+      "loss": 2.583,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3720352663603626,
+      "grad_norm": 0.3079470992088318,
+      "learning_rate": 5e-05,
+      "loss": 2.1146,
+      "step": 2996
+    },
+    {
+      "epoch": 0.372159443685583,
+      "grad_norm": 0.257081538438797,
+      "learning_rate": 5e-05,
+      "loss": 1.9308,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3722836210108034,
+      "grad_norm": 0.35725364089012146,
+      "learning_rate": 5e-05,
+      "loss": 2.7511,
+      "step": 2998
+    },
+    {
+      "epoch": 0.37240779833602383,
+      "grad_norm": 0.25790297985076904,
+      "learning_rate": 5e-05,
+      "loss": 1.8293,
+      "step": 2999
+    },
+    {
+      "epoch": 0.37253197566124424,
+      "grad_norm": 0.23215192556381226,
+      "learning_rate": 5e-05,
+      "loss": 1.8764,
+      "step": 3000
+    },
+    {
+      "epoch": 0.37265615298646465,
+      "grad_norm": 0.2362801432609558,
+      "learning_rate": 5e-05,
+      "loss": 1.7332,
+      "step": 3001
+    },
+    {
+      "epoch": 0.37278033031168506,
+      "grad_norm": 0.27961817383766174,
+      "learning_rate": 5e-05,
+      "loss": 2.1698,
+      "step": 3002
+    },
+    {
+      "epoch": 0.3729045076369055,
+      "grad_norm": 0.2894803583621979,
+      "learning_rate": 5e-05,
+      "loss": 2.3506,
+      "step": 3003
+    },
+    {
+      "epoch": 0.37302868496212593,
+      "grad_norm": 0.270277202129364,
+      "learning_rate": 5e-05,
+      "loss": 1.9794,
+      "step": 3004
+    },
+    {
+      "epoch": 0.37315286228734634,
+      "grad_norm": 0.28091710805892944,
+      "learning_rate": 5e-05,
+      "loss": 2.5896,
+      "step": 3005
+    },
+    {
+      "epoch": 0.37327703961256675,
+      "grad_norm": 0.2714832127094269,
+      "learning_rate": 5e-05,
+      "loss": 1.8102,
+      "step": 3006
+    },
+    {
+      "epoch": 0.37340121693778716,
+      "grad_norm": 0.2245844006538391,
+      "learning_rate": 5e-05,
+      "loss": 1.6682,
+      "step": 3007
+    },
+    {
+      "epoch": 0.37352539426300757,
+      "grad_norm": 0.2773212492465973,
+      "learning_rate": 5e-05,
+      "loss": 2.1679,
+      "step": 3008
+    },
+    {
+      "epoch": 0.373649571588228,
+      "grad_norm": 0.48758503794670105,
+      "learning_rate": 5e-05,
+      "loss": 3.7298,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3737737489134484,
+      "grad_norm": 0.315660297870636,
+      "learning_rate": 5e-05,
+      "loss": 2.2886,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3738979262386688,
+      "grad_norm": 0.2623748779296875,
+      "learning_rate": 5e-05,
+      "loss": 1.9723,
+      "step": 3011
+    },
+    {
+      "epoch": 0.3740221035638892,
+      "grad_norm": 0.2322576344013214,
+      "learning_rate": 5e-05,
+      "loss": 1.6699,
+      "step": 3012
+    },
+    {
+      "epoch": 0.3741462808891097,
+      "grad_norm": 0.2369822859764099,
+      "learning_rate": 5e-05,
+      "loss": 1.7811,
+      "step": 3013
+    },
+    {
+      "epoch": 0.3742704582143301,
+      "grad_norm": 0.28074389696121216,
+      "learning_rate": 5e-05,
+      "loss": 2.1291,
+      "step": 3014
+    },
+    {
+      "epoch": 0.3743946355395505,
+      "grad_norm": 0.4174252450466156,
+      "learning_rate": 5e-05,
+      "loss": 2.8298,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3745188128647709,
+      "grad_norm": 0.23224391043186188,
+      "learning_rate": 5e-05,
+      "loss": 1.8149,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3746429901899913,
+      "grad_norm": 0.24870234727859497,
+      "learning_rate": 5e-05,
+      "loss": 1.9722,
+      "step": 3017
+    },
+    {
+      "epoch": 0.3747671675152117,
+      "grad_norm": 0.2690788209438324,
+      "learning_rate": 5e-05,
+      "loss": 2.0374,
+      "step": 3018
+    },
+    {
+      "epoch": 0.37489134484043213,
+      "grad_norm": 0.25183412432670593,
+      "learning_rate": 5e-05,
+      "loss": 1.8474,
+      "step": 3019
+    },
+    {
+      "epoch": 0.37501552216565254,
+      "grad_norm": 0.2733623683452606,
+      "learning_rate": 5e-05,
+      "loss": 1.9086,
+      "step": 3020
+    },
+    {
+      "epoch": 0.37513969949087295,
+      "grad_norm": 0.27452918887138367,
+      "learning_rate": 5e-05,
+      "loss": 2.5931,
+      "step": 3021
+    },
+    {
+      "epoch": 0.37526387681609336,
+      "grad_norm": 0.25457167625427246,
+      "learning_rate": 5e-05,
+      "loss": 1.7968,
+      "step": 3022
+    },
+    {
+      "epoch": 0.3753880541413138,
+      "grad_norm": 0.2338773012161255,
+      "learning_rate": 5e-05,
+      "loss": 1.7335,
+      "step": 3023
+    },
+    {
+      "epoch": 0.37551223146653423,
+      "grad_norm": 0.3729333281517029,
+      "learning_rate": 5e-05,
+      "loss": 3.6581,
+      "step": 3024
+    },
+    {
+      "epoch": 0.37563640879175464,
+      "grad_norm": 0.23814961314201355,
+      "learning_rate": 5e-05,
+      "loss": 1.8734,
+      "step": 3025
+    },
+    {
+      "epoch": 0.37576058611697505,
+      "grad_norm": 0.34748411178588867,
+      "learning_rate": 5e-05,
+      "loss": 2.5288,
+      "step": 3026
+    },
+    {
+      "epoch": 0.37588476344219546,
+      "grad_norm": 0.2398747205734253,
+      "learning_rate": 5e-05,
+      "loss": 1.8456,
+      "step": 3027
+    },
+    {
+      "epoch": 0.37600894076741587,
+      "grad_norm": 0.2524730861186981,
+      "learning_rate": 5e-05,
+      "loss": 2.1373,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3761331180926363,
+      "grad_norm": 0.24938613176345825,
+      "learning_rate": 5e-05,
+      "loss": 1.793,
+      "step": 3029
+    },
+    {
+      "epoch": 0.3762572954178567,
+      "grad_norm": 0.28319695591926575,
+      "learning_rate": 5e-05,
+      "loss": 2.0707,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3763814727430771,
+      "grad_norm": 0.30263859033584595,
+      "learning_rate": 5e-05,
+      "loss": 2.0355,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3765056500682975,
+      "grad_norm": 0.378948837518692,
+      "learning_rate": 5e-05,
+      "loss": 2.8577,
+      "step": 3032
+    },
+    {
+      "epoch": 0.37662982739351797,
+      "grad_norm": 0.2410057783126831,
+      "learning_rate": 5e-05,
+      "loss": 1.8902,
+      "step": 3033
+    },
+    {
+      "epoch": 0.3767540047187384,
+      "grad_norm": 0.4981403946876526,
+      "learning_rate": 5e-05,
+      "loss": 3.5521,
+      "step": 3034
+    },
+    {
+      "epoch": 0.3768781820439588,
+      "grad_norm": 0.3131057322025299,
+      "learning_rate": 5e-05,
+      "loss": 1.9305,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3770023593691792,
+      "grad_norm": 0.2883146107196808,
+      "learning_rate": 5e-05,
+      "loss": 2.1817,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3771265366943996,
+      "grad_norm": 0.3231603801250458,
+      "learning_rate": 5e-05,
+      "loss": 2.2781,
+      "step": 3037
+    },
+    {
+      "epoch": 0.37725071401962,
+      "grad_norm": 0.28259745240211487,
+      "learning_rate": 5e-05,
+      "loss": 1.9511,
+      "step": 3038
+    },
+    {
+      "epoch": 0.3773748913448404,
+      "grad_norm": 0.25022047758102417,
+      "learning_rate": 5e-05,
+      "loss": 2.1814,
+      "step": 3039
+    },
+    {
+      "epoch": 0.37749906867006083,
+      "grad_norm": 0.330522745847702,
+      "learning_rate": 5e-05,
+      "loss": 2.1897,
+      "step": 3040
+    },
+    {
+      "epoch": 0.37762324599528124,
+      "grad_norm": 0.2720044255256653,
+      "learning_rate": 5e-05,
+      "loss": 1.886,
+      "step": 3041
+    },
+    {
+      "epoch": 0.37774742332050165,
+      "grad_norm": 0.24297118186950684,
+      "learning_rate": 5e-05,
+      "loss": 1.8226,
+      "step": 3042
+    },
+    {
+      "epoch": 0.3778716006457221,
+      "grad_norm": 0.2740362584590912,
+      "learning_rate": 5e-05,
+      "loss": 2.7123,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3779957779709425,
+      "grad_norm": 0.2390010505914688,
+      "learning_rate": 5e-05,
+      "loss": 2.0261,
+      "step": 3044
+    },
+    {
+      "epoch": 0.37811995529616294,
+      "grad_norm": 0.23411433398723602,
+      "learning_rate": 5e-05,
+      "loss": 1.7159,
+      "step": 3045
+    },
+    {
+      "epoch": 0.37824413262138334,
+      "grad_norm": 0.25204479694366455,
+      "learning_rate": 5e-05,
+      "loss": 1.9457,
+      "step": 3046
+    },
+    {
+      "epoch": 0.37836830994660375,
+      "grad_norm": 0.30702662467956543,
+      "learning_rate": 5e-05,
+      "loss": 2.9443,
+      "step": 3047
+    },
+    {
+      "epoch": 0.37849248727182416,
+      "grad_norm": 0.24769099056720734,
+      "learning_rate": 5e-05,
+      "loss": 1.8795,
+      "step": 3048
+    },
+    {
+      "epoch": 0.37861666459704457,
+      "grad_norm": 0.3070371747016907,
+      "learning_rate": 5e-05,
+      "loss": 2.0756,
+      "step": 3049
+    },
+    {
+      "epoch": 0.378740841922265,
+      "grad_norm": 0.2854249179363251,
+      "learning_rate": 5e-05,
+      "loss": 2.884,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3788650192474854,
+      "grad_norm": 0.2308148890733719,
+      "learning_rate": 5e-05,
+      "loss": 1.7609,
+      "step": 3051
+    },
+    {
+      "epoch": 0.3789891965727058,
+      "grad_norm": 0.26897910237312317,
+      "learning_rate": 5e-05,
+      "loss": 1.9236,
+      "step": 3052
+    },
+    {
+      "epoch": 0.37911337389792626,
+      "grad_norm": 0.247316375374794,
+      "learning_rate": 5e-05,
+      "loss": 1.8316,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3792375512231467,
+      "grad_norm": 0.2425869256258011,
+      "learning_rate": 5e-05,
+      "loss": 1.7214,
+      "step": 3054
+    },
+    {
+      "epoch": 0.3793617285483671,
+      "grad_norm": 0.25680699944496155,
+      "learning_rate": 5e-05,
+      "loss": 2.181,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3794859058735875,
+      "grad_norm": 0.24144139885902405,
+      "learning_rate": 5e-05,
+      "loss": 1.8867,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3796100831988079,
+      "grad_norm": 0.254015177488327,
+      "learning_rate": 5e-05,
+      "loss": 1.9029,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3797342605240283,
+      "grad_norm": 0.2578611671924591,
+      "learning_rate": 5e-05,
+      "loss": 2.0254,
+      "step": 3058
+    },
+    {
+      "epoch": 0.3798584378492487,
+      "grad_norm": 0.25550350546836853,
+      "learning_rate": 5e-05,
+      "loss": 1.8893,
+      "step": 3059
+    },
+    {
+      "epoch": 0.37998261517446913,
+      "grad_norm": 0.25460711121559143,
+      "learning_rate": 5e-05,
+      "loss": 2.0101,
+      "step": 3060
+    },
+    {
+      "epoch": 0.38010679249968954,
+      "grad_norm": 0.272898405790329,
+      "learning_rate": 5e-05,
+      "loss": 2.3676,
+      "step": 3061
+    },
+    {
+      "epoch": 0.38023096982490995,
+      "grad_norm": 0.25451529026031494,
+      "learning_rate": 5e-05,
+      "loss": 1.8787,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3803551471501304,
+      "grad_norm": 0.3461076021194458,
+      "learning_rate": 5e-05,
+      "loss": 3.0211,
+      "step": 3063
+    },
+    {
+      "epoch": 0.3804793244753508,
+      "grad_norm": 0.33161431550979614,
+      "learning_rate": 5e-05,
+      "loss": 2.2867,
+      "step": 3064
+    },
+    {
+      "epoch": 0.38060350180057123,
+      "grad_norm": 0.2653665840625763,
+      "learning_rate": 5e-05,
+      "loss": 1.9843,
+      "step": 3065
+    },
+    {
+      "epoch": 0.38072767912579164,
+      "grad_norm": 0.23903197050094604,
+      "learning_rate": 5e-05,
+      "loss": 1.6816,
+      "step": 3066
+    },
+    {
+      "epoch": 0.38085185645101205,
+      "grad_norm": 0.30219846963882446,
+      "learning_rate": 5e-05,
+      "loss": 2.396,
+      "step": 3067
+    },
+    {
+      "epoch": 0.38097603377623246,
+      "grad_norm": 0.31168368458747864,
+      "learning_rate": 5e-05,
+      "loss": 2.7844,
+      "step": 3068
+    },
+    {
+      "epoch": 0.38110021110145287,
+      "grad_norm": 0.2687379717826843,
+      "learning_rate": 5e-05,
+      "loss": 1.8187,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3812243884266733,
+      "grad_norm": 0.2609599232673645,
+      "learning_rate": 5e-05,
+      "loss": 1.8091,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3813485657518937,
+      "grad_norm": 0.23718082904815674,
+      "learning_rate": 5e-05,
+      "loss": 1.7568,
+      "step": 3071
+    },
+    {
+      "epoch": 0.3814727430771141,
+      "grad_norm": 0.23861859738826752,
+      "learning_rate": 5e-05,
+      "loss": 1.7172,
+      "step": 3072
+    },
+    {
+      "epoch": 0.38159692040233456,
+      "grad_norm": 0.27551525831222534,
+      "learning_rate": 5e-05,
+      "loss": 2.0051,
+      "step": 3073
+    },
+    {
+      "epoch": 0.38172109772755497,
+      "grad_norm": 0.27747848629951477,
+      "learning_rate": 5e-05,
+      "loss": 1.8958,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3818452750527754,
+      "grad_norm": 0.2449251115322113,
+      "learning_rate": 5e-05,
+      "loss": 1.7087,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3819694523779958,
+      "grad_norm": 0.24824030697345734,
+      "learning_rate": 5e-05,
+      "loss": 1.9133,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3820936297032162,
+      "grad_norm": 0.3007103204727173,
+      "learning_rate": 5e-05,
+      "loss": 2.3038,
+      "step": 3077
+    },
+    {
+      "epoch": 0.3822178070284366,
+      "grad_norm": 0.25122156739234924,
+      "learning_rate": 5e-05,
+      "loss": 1.8165,
+      "step": 3078
+    },
+    {
+      "epoch": 0.382341984353657,
+      "grad_norm": 0.2511250078678131,
+      "learning_rate": 5e-05,
+      "loss": 1.7301,
+      "step": 3079
+    },
+    {
+      "epoch": 0.3824661616788774,
+      "grad_norm": 0.2638435959815979,
+      "learning_rate": 5e-05,
+      "loss": 1.7817,
+      "step": 3080
+    },
+    {
+      "epoch": 0.38259033900409783,
+      "grad_norm": 0.25206121802330017,
+      "learning_rate": 5e-05,
+      "loss": 1.7258,
+      "step": 3081
+    },
+    {
+      "epoch": 0.38271451632931824,
+      "grad_norm": 0.2951272130012512,
+      "learning_rate": 5e-05,
+      "loss": 2.5512,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3828386936545387,
+      "grad_norm": 0.3649168908596039,
+      "learning_rate": 5e-05,
+      "loss": 3.1462,
+      "step": 3083
+    },
+    {
+      "epoch": 0.3829628709797591,
+      "grad_norm": 0.2575666308403015,
+      "learning_rate": 5e-05,
+      "loss": 1.9446,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3830870483049795,
+      "grad_norm": 0.2706921696662903,
+      "learning_rate": 5e-05,
+      "loss": 2.0286,
+      "step": 3085
+    },
+    {
+      "epoch": 0.38321122563019994,
+      "grad_norm": 0.23469194769859314,
+      "learning_rate": 5e-05,
+      "loss": 1.7326,
+      "step": 3086
+    },
+    {
+      "epoch": 0.38333540295542035,
+      "grad_norm": 0.25182586908340454,
+      "learning_rate": 5e-05,
+      "loss": 1.906,
+      "step": 3087
+    },
+    {
+      "epoch": 0.38345958028064075,
+      "grad_norm": 0.23961196839809418,
+      "learning_rate": 5e-05,
+      "loss": 1.7912,
+      "step": 3088
+    },
+    {
+      "epoch": 0.38358375760586116,
+      "grad_norm": 0.344236820936203,
+      "learning_rate": 5e-05,
+      "loss": 2.5077,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3837079349310816,
+      "grad_norm": 0.253130704164505,
+      "learning_rate": 5e-05,
+      "loss": 1.7863,
+      "step": 3090
+    },
+    {
+      "epoch": 0.383832112256302,
+      "grad_norm": 0.22801761329174042,
+      "learning_rate": 5e-05,
+      "loss": 1.8451,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3839562895815224,
+      "grad_norm": 0.2922053039073944,
+      "learning_rate": 5e-05,
+      "loss": 1.9032,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3840804669067428,
+      "grad_norm": 0.23056386411190033,
+      "learning_rate": 5e-05,
+      "loss": 1.6836,
+      "step": 3093
+    },
+    {
+      "epoch": 0.38420464423196327,
+      "grad_norm": 0.27705180644989014,
+      "learning_rate": 5e-05,
+      "loss": 2.3663,
+      "step": 3094
+    },
+    {
+      "epoch": 0.3843288215571837,
+      "grad_norm": 0.26883646845817566,
+      "learning_rate": 5e-05,
+      "loss": 2.502,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3844529988824041,
+      "grad_norm": 0.2381926029920578,
+      "learning_rate": 5e-05,
+      "loss": 1.6875,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3845771762076245,
+      "grad_norm": 0.3499280512332916,
+      "learning_rate": 5e-05,
+      "loss": 3.1822,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3847013535328449,
+      "grad_norm": 0.26643460988998413,
+      "learning_rate": 5e-05,
+      "loss": 2.0534,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3848255308580653,
+      "grad_norm": 0.24763643741607666,
+      "learning_rate": 5e-05,
+      "loss": 1.7935,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3849497081832857,
+      "grad_norm": 0.23745521903038025,
+      "learning_rate": 5e-05,
+      "loss": 1.7142,
+      "step": 3100
+    },
+    {
+      "epoch": 0.38507388550850613,
+      "grad_norm": 0.25083643198013306,
+      "learning_rate": 5e-05,
+      "loss": 1.9099,
+      "step": 3101
+    },
+    {
+      "epoch": 0.38519806283372654,
+      "grad_norm": 0.31027424335479736,
+      "learning_rate": 5e-05,
+      "loss": 2.4803,
+      "step": 3102
+    },
+    {
+      "epoch": 0.38532224015894695,
+      "grad_norm": 0.21859607100486755,
+      "learning_rate": 5e-05,
+      "loss": 1.6418,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3854464174841674,
+      "grad_norm": 0.281548410654068,
+      "learning_rate": 5e-05,
+      "loss": 2.3032,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3855705948093878,
+      "grad_norm": 0.26583942770957947,
+      "learning_rate": 5e-05,
+      "loss": 2.0453,
+      "step": 3105
+    },
+    {
+      "epoch": 0.38569477213460823,
+      "grad_norm": 0.272785484790802,
+      "learning_rate": 5e-05,
+      "loss": 2.1647,
+      "step": 3106
+    },
+    {
+      "epoch": 0.38581894945982864,
+      "grad_norm": 0.3125622272491455,
+      "learning_rate": 5e-05,
+      "loss": 2.4732,
+      "step": 3107
+    },
+    {
+      "epoch": 0.38594312678504905,
+      "grad_norm": 0.25268444418907166,
+      "learning_rate": 5e-05,
+      "loss": 2.3328,
+      "step": 3108
+    },
+    {
+      "epoch": 0.38606730411026946,
+      "grad_norm": 0.2394634634256363,
+      "learning_rate": 5e-05,
+      "loss": 1.8094,
+      "step": 3109
+    },
+    {
+      "epoch": 0.38619148143548987,
+      "grad_norm": 0.247454434633255,
+      "learning_rate": 5e-05,
+      "loss": 1.8495,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3863156587607103,
+      "grad_norm": 0.2566412389278412,
+      "learning_rate": 5e-05,
+      "loss": 1.8408,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3864398360859307,
+      "grad_norm": 0.2542460262775421,
+      "learning_rate": 5e-05,
+      "loss": 1.9761,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3865640134111511,
+      "grad_norm": 0.24918216466903687,
+      "learning_rate": 5e-05,
+      "loss": 1.9928,
+      "step": 3113
+    },
+    {
+      "epoch": 0.38668819073637156,
+      "grad_norm": 0.251039057970047,
+      "learning_rate": 5e-05,
+      "loss": 1.8508,
+      "step": 3114
+    },
+    {
+      "epoch": 0.38681236806159197,
+      "grad_norm": 0.23671133816242218,
+      "learning_rate": 5e-05,
+      "loss": 1.6716,
+      "step": 3115
+    },
+    {
+      "epoch": 0.3869365453868124,
+      "grad_norm": 0.2513390779495239,
+      "learning_rate": 5e-05,
+      "loss": 1.8331,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3870607227120328,
+      "grad_norm": 0.26725152134895325,
+      "learning_rate": 5e-05,
+      "loss": 2.3166,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3871849000372532,
+      "grad_norm": 0.34937217831611633,
+      "learning_rate": 5e-05,
+      "loss": 3.1555,
+      "step": 3118
+    },
+    {
+      "epoch": 0.3873090773624736,
+      "grad_norm": 0.2650000751018524,
+      "learning_rate": 5e-05,
+      "loss": 2.6926,
+      "step": 3119
+    },
+    {
+      "epoch": 0.387433254687694,
+      "grad_norm": 0.3080534040927887,
+      "learning_rate": 5e-05,
+      "loss": 2.4796,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3875574320129144,
+      "grad_norm": 0.2609560489654541,
+      "learning_rate": 5e-05,
+      "loss": 1.8447,
+      "step": 3121
+    },
+    {
+      "epoch": 0.38768160933813484,
+      "grad_norm": 0.30605852603912354,
+      "learning_rate": 5e-05,
+      "loss": 2.2455,
+      "step": 3122
+    },
+    {
+      "epoch": 0.38780578666335525,
+      "grad_norm": 0.24625812470912933,
+      "learning_rate": 5e-05,
+      "loss": 1.9883,
+      "step": 3123
+    },
+    {
+      "epoch": 0.3879299639885757,
+      "grad_norm": 0.3390367925167084,
+      "learning_rate": 5e-05,
+      "loss": 2.4521,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3880541413137961,
+      "grad_norm": 0.27835553884506226,
+      "learning_rate": 5e-05,
+      "loss": 2.2179,
+      "step": 3125
+    },
+    {
+      "epoch": 0.38817831863901653,
+      "grad_norm": 0.2376982569694519,
+      "learning_rate": 5e-05,
+      "loss": 1.7662,
+      "step": 3126
+    },
+    {
+      "epoch": 0.38830249596423694,
+      "grad_norm": 0.3349287509918213,
+      "learning_rate": 5e-05,
+      "loss": 2.6532,
+      "step": 3127
+    },
+    {
+      "epoch": 0.38842667328945735,
+      "grad_norm": 0.2711982727050781,
+      "learning_rate": 5e-05,
+      "loss": 1.9078,
+      "step": 3128
+    },
+    {
+      "epoch": 0.38855085061467776,
+      "grad_norm": 0.23891501128673553,
+      "learning_rate": 5e-05,
+      "loss": 2.0745,
+      "step": 3129
+    },
+    {
+      "epoch": 0.38867502793989817,
+      "grad_norm": 0.24813856184482574,
+      "learning_rate": 5e-05,
+      "loss": 2.1678,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3887992052651186,
+      "grad_norm": 0.23392683267593384,
+      "learning_rate": 5e-05,
+      "loss": 1.7592,
+      "step": 3131
+    },
+    {
+      "epoch": 0.388923382590339,
+      "grad_norm": 0.2949017882347107,
+      "learning_rate": 5e-05,
+      "loss": 2.6806,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3890475599155594,
+      "grad_norm": 0.25584590435028076,
+      "learning_rate": 5e-05,
+      "loss": 1.7985,
+      "step": 3133
+    },
+    {
+      "epoch": 0.38917173724077986,
+      "grad_norm": 0.2581401467323303,
+      "learning_rate": 5e-05,
+      "loss": 1.921,
+      "step": 3134
+    },
+    {
+      "epoch": 0.38929591456600027,
+      "grad_norm": 0.23445293307304382,
+      "learning_rate": 5e-05,
+      "loss": 1.7685,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3894200918912207,
+      "grad_norm": 0.26438209414482117,
+      "learning_rate": 5e-05,
+      "loss": 2.1228,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3895442692164411,
+      "grad_norm": 0.28357917070388794,
+      "learning_rate": 5e-05,
+      "loss": 2.276,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3896684465416615,
+      "grad_norm": 0.24636438488960266,
+      "learning_rate": 5e-05,
+      "loss": 1.7983,
+      "step": 3138
+    },
+    {
+      "epoch": 0.3897926238668819,
+      "grad_norm": 0.2388295829296112,
+      "learning_rate": 5e-05,
+      "loss": 1.8797,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3899168011921023,
+      "grad_norm": 0.26258060336112976,
+      "learning_rate": 5e-05,
+      "loss": 1.6985,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3900409785173227,
+      "grad_norm": 0.3425505757331848,
+      "learning_rate": 5e-05,
+      "loss": 2.526,
+      "step": 3141
+    },
+    {
+      "epoch": 0.39016515584254313,
+      "grad_norm": 0.24155297875404358,
+      "learning_rate": 5e-05,
+      "loss": 1.9104,
+      "step": 3142
+    },
+    {
+      "epoch": 0.39028933316776354,
+      "grad_norm": 0.2627064287662506,
+      "learning_rate": 5e-05,
+      "loss": 1.9048,
+      "step": 3143
+    },
+    {
+      "epoch": 0.390413510492984,
+      "grad_norm": 0.24964413046836853,
+      "learning_rate": 5e-05,
+      "loss": 2.2252,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3905376878182044,
+      "grad_norm": 0.24718742072582245,
+      "learning_rate": 5e-05,
+      "loss": 1.7927,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3906618651434248,
+      "grad_norm": 0.25128018856048584,
+      "learning_rate": 5e-05,
+      "loss": 2.1623,
+      "step": 3146
+    },
+    {
+      "epoch": 0.39078604246864523,
+      "grad_norm": 0.24500875174999237,
+      "learning_rate": 5e-05,
+      "loss": 1.8136,
+      "step": 3147
+    },
+    {
+      "epoch": 0.39091021979386564,
+      "grad_norm": 0.23893284797668457,
+      "learning_rate": 5e-05,
+      "loss": 2.0048,
+      "step": 3148
+    },
+    {
+      "epoch": 0.39103439711908605,
+      "grad_norm": 0.3076295852661133,
+      "learning_rate": 5e-05,
+      "loss": 2.1997,
+      "step": 3149
+    },
+    {
+      "epoch": 0.39115857444430646,
+      "grad_norm": 0.2501981854438782,
+      "learning_rate": 5e-05,
+      "loss": 1.9411,
+      "step": 3150
+    },
+    {
+      "epoch": 0.39128275176952687,
+      "grad_norm": 0.2917187511920929,
+      "learning_rate": 5e-05,
+      "loss": 1.9085,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3914069290947473,
+      "grad_norm": 0.23988878726959229,
+      "learning_rate": 5e-05,
+      "loss": 2.1462,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3915311064199677,
+      "grad_norm": 0.2696390748023987,
+      "learning_rate": 5e-05,
+      "loss": 1.999,
+      "step": 3153
+    },
+    {
+      "epoch": 0.39165528374518815,
+      "grad_norm": 0.30964720249176025,
+      "learning_rate": 5e-05,
+      "loss": 2.5373,
+      "step": 3154
+    },
+    {
+      "epoch": 0.39177946107040856,
+      "grad_norm": 0.24554970860481262,
+      "learning_rate": 5e-05,
+      "loss": 2.0891,
+      "step": 3155
+    },
+    {
+      "epoch": 0.391903638395629,
+      "grad_norm": 0.3089919090270996,
+      "learning_rate": 5e-05,
+      "loss": 2.5649,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3920278157208494,
+      "grad_norm": 0.28048399090766907,
+      "learning_rate": 5e-05,
+      "loss": 2.3251,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3921519930460698,
+      "grad_norm": 0.27881699800491333,
+      "learning_rate": 5e-05,
+      "loss": 1.9704,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3922761703712902,
+      "grad_norm": 0.3241457939147949,
+      "learning_rate": 5e-05,
+      "loss": 2.6925,
+      "step": 3159
+    },
+    {
+      "epoch": 0.3924003476965106,
+      "grad_norm": 0.28475019335746765,
+      "learning_rate": 5e-05,
+      "loss": 2.258,
+      "step": 3160
+    },
+    {
+      "epoch": 0.392524525021731,
+      "grad_norm": 0.24232326447963715,
+      "learning_rate": 5e-05,
+      "loss": 1.7258,
+      "step": 3161
+    },
+    {
+      "epoch": 0.39264870234695143,
+      "grad_norm": 0.32122868299484253,
+      "learning_rate": 5e-05,
+      "loss": 2.2273,
+      "step": 3162
+    },
+    {
+      "epoch": 0.39277287967217184,
+      "grad_norm": 0.30831006169319153,
+      "learning_rate": 5e-05,
+      "loss": 2.4585,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3928970569973923,
+      "grad_norm": 0.2527689039707184,
+      "learning_rate": 5e-05,
+      "loss": 1.7256,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3930212343226127,
+      "grad_norm": 0.27650725841522217,
+      "learning_rate": 5e-05,
+      "loss": 2.028,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3931454116478331,
+      "grad_norm": 0.290206640958786,
+      "learning_rate": 5e-05,
+      "loss": 1.9673,
+      "step": 3166
+    },
+    {
+      "epoch": 0.39326958897305353,
+      "grad_norm": 0.2424802929162979,
+      "learning_rate": 5e-05,
+      "loss": 2.0775,
+      "step": 3167
+    },
+    {
+      "epoch": 0.39339376629827394,
+      "grad_norm": 0.35881492495536804,
+      "learning_rate": 5e-05,
+      "loss": 2.3298,
+      "step": 3168
+    },
+    {
+      "epoch": 0.39351794362349435,
+      "grad_norm": 0.28162121772766113,
+      "learning_rate": 5e-05,
+      "loss": 2.1162,
+      "step": 3169
+    },
+    {
+      "epoch": 0.39364212094871476,
+      "grad_norm": 0.3731600344181061,
+      "learning_rate": 5e-05,
+      "loss": 2.7778,
+      "step": 3170
+    },
+    {
+      "epoch": 0.39376629827393517,
+      "grad_norm": 0.24918660521507263,
+      "learning_rate": 5e-05,
+      "loss": 2.3232,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3938904755991556,
+      "grad_norm": 0.3939272463321686,
+      "learning_rate": 5e-05,
+      "loss": 2.7767,
+      "step": 3172
+    },
+    {
+      "epoch": 0.394014652924376,
+      "grad_norm": 0.2590550482273102,
+      "learning_rate": 5e-05,
+      "loss": 1.9365,
+      "step": 3173
+    },
+    {
+      "epoch": 0.39413883024959645,
+      "grad_norm": 0.5520243644714355,
+      "learning_rate": 5e-05,
+      "loss": 3.5578,
+      "step": 3174
+    },
+    {
+      "epoch": 0.39426300757481686,
+      "grad_norm": 0.2481030821800232,
+      "learning_rate": 5e-05,
+      "loss": 1.9333,
+      "step": 3175
+    },
+    {
+      "epoch": 0.39438718490003727,
+      "grad_norm": 0.28480684757232666,
+      "learning_rate": 5e-05,
+      "loss": 2.1246,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3945113622252577,
+      "grad_norm": 0.3174760639667511,
+      "learning_rate": 5e-05,
+      "loss": 2.2703,
+      "step": 3177
+    },
+    {
+      "epoch": 0.3946355395504781,
+      "grad_norm": 0.23890171945095062,
+      "learning_rate": 5e-05,
+      "loss": 1.8264,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3947597168756985,
+      "grad_norm": 0.2582869827747345,
+      "learning_rate": 5e-05,
+      "loss": 2.0858,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3948838942009189,
+      "grad_norm": 0.266282856464386,
+      "learning_rate": 5e-05,
+      "loss": 2.0402,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3950080715261393,
+      "grad_norm": 0.24979424476623535,
+      "learning_rate": 5e-05,
+      "loss": 2.0849,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3951322488513597,
+      "grad_norm": 0.254221111536026,
+      "learning_rate": 5e-05,
+      "loss": 1.9893,
+      "step": 3182
+    },
+    {
+      "epoch": 0.39525642617658013,
+      "grad_norm": 0.23809747397899628,
+      "learning_rate": 5e-05,
+      "loss": 1.8721,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3953806035018006,
+      "grad_norm": 0.2598564326763153,
+      "learning_rate": 5e-05,
+      "loss": 1.9059,
+      "step": 3184
+    },
+    {
+      "epoch": 0.395504780827021,
+      "grad_norm": 0.2539629638195038,
+      "learning_rate": 5e-05,
+      "loss": 2.2068,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3956289581522414,
+      "grad_norm": 0.2955545485019684,
+      "learning_rate": 5e-05,
+      "loss": 2.1472,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3957531354774618,
+      "grad_norm": 0.24721388518810272,
+      "learning_rate": 5e-05,
+      "loss": 1.7596,
+      "step": 3187
+    },
+    {
+      "epoch": 0.39587731280268224,
+      "grad_norm": 0.2526703178882599,
+      "learning_rate": 5e-05,
+      "loss": 1.9629,
+      "step": 3188
+    },
+    {
+      "epoch": 0.39600149012790264,
+      "grad_norm": 0.2255893051624298,
+      "learning_rate": 5e-05,
+      "loss": 1.6443,
+      "step": 3189
+    },
+    {
+      "epoch": 0.39612566745312305,
+      "grad_norm": 0.23943160474300385,
+      "learning_rate": 5e-05,
+      "loss": 1.9351,
+      "step": 3190
+    },
+    {
+      "epoch": 0.39624984477834346,
+      "grad_norm": 0.435127317905426,
+      "learning_rate": 5e-05,
+      "loss": 2.8826,
+      "step": 3191
+    },
+    {
+      "epoch": 0.39637402210356387,
+      "grad_norm": 0.2501561641693115,
+      "learning_rate": 5e-05,
+      "loss": 1.5746,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3964981994287843,
+      "grad_norm": 0.28033673763275146,
+      "learning_rate": 5e-05,
+      "loss": 2.426,
+      "step": 3193
+    },
+    {
+      "epoch": 0.39662237675400475,
+      "grad_norm": 0.2348051369190216,
+      "learning_rate": 5e-05,
+      "loss": 1.7486,
+      "step": 3194
+    },
+    {
+      "epoch": 0.39674655407922516,
+      "grad_norm": 0.316873162984848,
+      "learning_rate": 5e-05,
+      "loss": 2.9575,
+      "step": 3195
+    },
+    {
+      "epoch": 0.39687073140444556,
+      "grad_norm": 0.27446675300598145,
+      "learning_rate": 5e-05,
+      "loss": 1.9949,
+      "step": 3196
+    },
+    {
+      "epoch": 0.396994908729666,
+      "grad_norm": 0.260734885931015,
+      "learning_rate": 5e-05,
+      "loss": 1.9581,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3971190860548864,
+      "grad_norm": 0.36495643854141235,
+      "learning_rate": 5e-05,
+      "loss": 2.9605,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3972432633801068,
+      "grad_norm": 0.24668940901756287,
+      "learning_rate": 5e-05,
+      "loss": 1.9434,
+      "step": 3199
+    },
+    {
+      "epoch": 0.3973674407053272,
+      "grad_norm": 0.2473028302192688,
+      "learning_rate": 5e-05,
+      "loss": 1.8453,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3974916180305476,
+      "grad_norm": 0.2859460413455963,
+      "learning_rate": 5e-05,
+      "loss": 1.9514,
+      "step": 3201
+    },
+    {
+      "epoch": 0.397615795355768,
+      "grad_norm": 0.26227250695228577,
+      "learning_rate": 5e-05,
+      "loss": 1.8888,
+      "step": 3202
+    },
+    {
+      "epoch": 0.39773997268098843,
+      "grad_norm": 0.2822369337081909,
+      "learning_rate": 5e-05,
+      "loss": 1.9258,
+      "step": 3203
+    },
+    {
+      "epoch": 0.39786415000620884,
+      "grad_norm": 0.2533319294452667,
+      "learning_rate": 5e-05,
+      "loss": 1.9191,
+      "step": 3204
+    },
+    {
+      "epoch": 0.3979883273314293,
+      "grad_norm": 0.2517368495464325,
+      "learning_rate": 5e-05,
+      "loss": 1.8751,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3981125046566497,
+      "grad_norm": 0.25612884759902954,
+      "learning_rate": 5e-05,
+      "loss": 1.9213,
+      "step": 3206
+    },
+    {
+      "epoch": 0.3982366819818701,
+      "grad_norm": 0.2861345708370209,
+      "learning_rate": 5e-05,
+      "loss": 2.5041,
+      "step": 3207
+    },
+    {
+      "epoch": 0.39836085930709053,
+      "grad_norm": 0.28101029992103577,
+      "learning_rate": 5e-05,
+      "loss": 2.3078,
+      "step": 3208
+    },
+    {
+      "epoch": 0.39848503663231094,
+      "grad_norm": 0.2397671341896057,
+      "learning_rate": 5e-05,
+      "loss": 1.4569,
+      "step": 3209
+    },
+    {
+      "epoch": 0.39860921395753135,
+      "grad_norm": 0.24958276748657227,
+      "learning_rate": 5e-05,
+      "loss": 1.8315,
+      "step": 3210
+    },
+    {
+      "epoch": 0.39873339128275176,
+      "grad_norm": 0.3403858244419098,
+      "learning_rate": 5e-05,
+      "loss": 2.5607,
+      "step": 3211
+    },
+    {
+      "epoch": 0.39885756860797217,
+      "grad_norm": 0.2546115517616272,
+      "learning_rate": 5e-05,
+      "loss": 1.9461,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3989817459331926,
+      "grad_norm": 0.2427891343832016,
+      "learning_rate": 5e-05,
+      "loss": 1.8502,
+      "step": 3213
+    },
+    {
+      "epoch": 0.399105923258413,
+      "grad_norm": 0.41200706362724304,
+      "learning_rate": 5e-05,
+      "loss": 3.0953,
+      "step": 3214
+    },
+    {
+      "epoch": 0.39923010058363345,
+      "grad_norm": 0.28418055176734924,
+      "learning_rate": 5e-05,
+      "loss": 2.0967,
+      "step": 3215
+    },
+    {
+      "epoch": 0.39935427790885386,
+      "grad_norm": 0.343633234500885,
+      "learning_rate": 5e-05,
+      "loss": 3.0882,
+      "step": 3216
+    },
+    {
+      "epoch": 0.39947845523407427,
+      "grad_norm": 0.247809499502182,
+      "learning_rate": 5e-05,
+      "loss": 1.741,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3996026325592947,
+      "grad_norm": 0.27520865201950073,
+      "learning_rate": 5e-05,
+      "loss": 1.9052,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3997268098845151,
+      "grad_norm": 0.2212163209915161,
+      "learning_rate": 5e-05,
+      "loss": 1.4233,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3998509872097355,
+      "grad_norm": 0.2825237512588501,
+      "learning_rate": 5e-05,
+      "loss": 2.4915,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3999751645349559,
+      "grad_norm": 0.23847058415412903,
+      "learning_rate": 5e-05,
+      "loss": 1.675,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4000993418601763,
+      "grad_norm": 0.25368937849998474,
+      "learning_rate": 5e-05,
+      "loss": 2.0468,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4002235191853967,
+      "grad_norm": 0.2578575909137726,
+      "learning_rate": 5e-05,
+      "loss": 1.8379,
+      "step": 3223
+    },
+    {
+      "epoch": 0.40034769651061713,
+      "grad_norm": 0.2366873025894165,
+      "learning_rate": 5e-05,
+      "loss": 1.7687,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4004718738358376,
+      "grad_norm": 0.37953028082847595,
+      "learning_rate": 5e-05,
+      "loss": 2.9504,
+      "step": 3225
+    },
+    {
+      "epoch": 0.400596051161058,
+      "grad_norm": 0.23708204925060272,
+      "learning_rate": 5e-05,
+      "loss": 1.657,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4007202284862784,
+      "grad_norm": 0.24100476503372192,
+      "learning_rate": 5e-05,
+      "loss": 1.735,
+      "step": 3227
+    },
+    {
+      "epoch": 0.4008444058114988,
+      "grad_norm": 0.24960608780384064,
+      "learning_rate": 5e-05,
+      "loss": 1.816,
+      "step": 3228
+    },
+    {
+      "epoch": 0.40096858313671924,
+      "grad_norm": 0.28130728006362915,
+      "learning_rate": 5e-05,
+      "loss": 2.1257,
+      "step": 3229
+    },
+    {
+      "epoch": 0.40109276046193965,
+      "grad_norm": 0.23559513688087463,
+      "learning_rate": 5e-05,
+      "loss": 1.8406,
+      "step": 3230
+    },
+    {
+      "epoch": 0.40121693778716006,
+      "grad_norm": 0.24949392676353455,
+      "learning_rate": 5e-05,
+      "loss": 2.1763,
+      "step": 3231
+    },
+    {
+      "epoch": 0.40134111511238046,
+      "grad_norm": 0.23313212394714355,
+      "learning_rate": 5e-05,
+      "loss": 1.8966,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4014652924376009,
+      "grad_norm": 0.26847589015960693,
+      "learning_rate": 5e-05,
+      "loss": 2.0881,
+      "step": 3233
+    },
+    {
+      "epoch": 0.4015894697628213,
+      "grad_norm": 0.266902893781662,
+      "learning_rate": 5e-05,
+      "loss": 1.8349,
+      "step": 3234
+    },
+    {
+      "epoch": 0.40171364708804175,
+      "grad_norm": 0.2542303502559662,
+      "learning_rate": 5e-05,
+      "loss": 2.0281,
+      "step": 3235
+    },
+    {
+      "epoch": 0.40183782441326216,
+      "grad_norm": 0.2559458911418915,
+      "learning_rate": 5e-05,
+      "loss": 1.8881,
+      "step": 3236
+    },
+    {
+      "epoch": 0.40196200173848257,
+      "grad_norm": 0.3651553690433502,
+      "learning_rate": 5e-05,
+      "loss": 2.8641,
+      "step": 3237
+    },
+    {
+      "epoch": 0.402086179063703,
+      "grad_norm": 0.26255086064338684,
+      "learning_rate": 5e-05,
+      "loss": 2.2252,
+      "step": 3238
+    },
+    {
+      "epoch": 0.4022103563889234,
+      "grad_norm": 0.22931768000125885,
+      "learning_rate": 5e-05,
+      "loss": 1.6029,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4023345337141438,
+      "grad_norm": 0.23497271537780762,
+      "learning_rate": 5e-05,
+      "loss": 1.7451,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4024587110393642,
+      "grad_norm": 0.24078533053398132,
+      "learning_rate": 5e-05,
+      "loss": 1.7578,
+      "step": 3241
+    },
+    {
+      "epoch": 0.4025828883645846,
+      "grad_norm": 0.2341689169406891,
+      "learning_rate": 5e-05,
+      "loss": 1.8351,
+      "step": 3242
+    },
+    {
+      "epoch": 0.402707065689805,
+      "grad_norm": 0.23458215594291687,
+      "learning_rate": 5e-05,
+      "loss": 1.8435,
+      "step": 3243
+    },
+    {
+      "epoch": 0.40283124301502543,
+      "grad_norm": 0.2809026539325714,
+      "learning_rate": 5e-05,
+      "loss": 2.086,
+      "step": 3244
+    },
+    {
+      "epoch": 0.4029554203402459,
+      "grad_norm": 0.26391589641571045,
+      "learning_rate": 5e-05,
+      "loss": 2.0628,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4030795976654663,
+      "grad_norm": 0.25057706236839294,
+      "learning_rate": 5e-05,
+      "loss": 1.8822,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4032037749906867,
+      "grad_norm": 0.33236244320869446,
+      "learning_rate": 5e-05,
+      "loss": 2.2004,
+      "step": 3247
+    },
+    {
+      "epoch": 0.4033279523159071,
+      "grad_norm": 0.22801382839679718,
+      "learning_rate": 5e-05,
+      "loss": 1.673,
+      "step": 3248
+    },
+    {
+      "epoch": 0.40345212964112753,
+      "grad_norm": 0.2859856188297272,
+      "learning_rate": 5e-05,
+      "loss": 2.0837,
+      "step": 3249
+    },
+    {
+      "epoch": 0.40357630696634794,
+      "grad_norm": 0.2515939176082611,
+      "learning_rate": 5e-05,
+      "loss": 1.8873,
+      "step": 3250
+    },
+    {
+      "epoch": 0.40370048429156835,
+      "grad_norm": 0.2529393136501312,
+      "learning_rate": 5e-05,
+      "loss": 1.9659,
+      "step": 3251
+    },
+    {
+      "epoch": 0.40382466161678876,
+      "grad_norm": 0.2500138282775879,
+      "learning_rate": 5e-05,
+      "loss": 1.8937,
+      "step": 3252
+    },
+    {
+      "epoch": 0.40394883894200917,
+      "grad_norm": 0.2902529835700989,
+      "learning_rate": 5e-05,
+      "loss": 1.8985,
+      "step": 3253
+    },
+    {
+      "epoch": 0.4040730162672296,
+      "grad_norm": 0.3275989890098572,
+      "learning_rate": 5e-05,
+      "loss": 1.9777,
+      "step": 3254
+    },
+    {
+      "epoch": 0.40419719359245004,
+      "grad_norm": 0.24015851318836212,
+      "learning_rate": 5e-05,
+      "loss": 1.8576,
+      "step": 3255
+    },
+    {
+      "epoch": 0.40432137091767045,
+      "grad_norm": 0.31689026951789856,
+      "learning_rate": 5e-05,
+      "loss": 2.3551,
+      "step": 3256
+    },
+    {
+      "epoch": 0.40444554824289086,
+      "grad_norm": 0.2728320360183716,
+      "learning_rate": 5e-05,
+      "loss": 1.9237,
+      "step": 3257
+    },
+    {
+      "epoch": 0.40456972556811127,
+      "grad_norm": 0.328977108001709,
+      "learning_rate": 5e-05,
+      "loss": 2.9301,
+      "step": 3258
+    },
+    {
+      "epoch": 0.4046939028933317,
+      "grad_norm": 0.2534151077270508,
+      "learning_rate": 5e-05,
+      "loss": 1.9114,
+      "step": 3259
+    },
+    {
+      "epoch": 0.4048180802185521,
+      "grad_norm": 0.230416938662529,
+      "learning_rate": 5e-05,
+      "loss": 1.7749,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4049422575437725,
+      "grad_norm": 0.27188584208488464,
+      "learning_rate": 5e-05,
+      "loss": 2.0906,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4050664348689929,
+      "grad_norm": 0.31253334879875183,
+      "learning_rate": 5e-05,
+      "loss": 2.76,
+      "step": 3262
+    },
+    {
+      "epoch": 0.4051906121942133,
+      "grad_norm": 0.27397364377975464,
+      "learning_rate": 5e-05,
+      "loss": 2.1069,
+      "step": 3263
+    },
+    {
+      "epoch": 0.4053147895194337,
+      "grad_norm": 0.2573283016681671,
+      "learning_rate": 5e-05,
+      "loss": 1.6221,
+      "step": 3264
+    },
+    {
+      "epoch": 0.4054389668446542,
+      "grad_norm": 0.2459263950586319,
+      "learning_rate": 5e-05,
+      "loss": 1.6137,
+      "step": 3265
+    },
+    {
+      "epoch": 0.4055631441698746,
+      "grad_norm": 0.410357266664505,
+      "learning_rate": 5e-05,
+      "loss": 2.8048,
+      "step": 3266
+    },
+    {
+      "epoch": 0.405687321495095,
+      "grad_norm": 0.2518337368965149,
+      "learning_rate": 5e-05,
+      "loss": 1.9025,
+      "step": 3267
+    },
+    {
+      "epoch": 0.4058114988203154,
+      "grad_norm": 0.25045186281204224,
+      "learning_rate": 5e-05,
+      "loss": 1.8734,
+      "step": 3268
+    },
+    {
+      "epoch": 0.40593567614553583,
+      "grad_norm": 0.3234350383281708,
+      "learning_rate": 5e-05,
+      "loss": 2.3716,
+      "step": 3269
+    },
+    {
+      "epoch": 0.40605985347075624,
+      "grad_norm": 0.25751304626464844,
+      "learning_rate": 5e-05,
+      "loss": 1.8244,
+      "step": 3270
+    },
+    {
+      "epoch": 0.40618403079597665,
+      "grad_norm": 0.23443768918514252,
+      "learning_rate": 5e-05,
+      "loss": 1.692,
+      "step": 3271
+    },
+    {
+      "epoch": 0.40630820812119706,
+      "grad_norm": 0.27722761034965515,
+      "learning_rate": 5e-05,
+      "loss": 1.849,
+      "step": 3272
+    },
+    {
+      "epoch": 0.40643238544641747,
+      "grad_norm": 0.25484806299209595,
+      "learning_rate": 5e-05,
+      "loss": 2.1674,
+      "step": 3273
+    },
+    {
+      "epoch": 0.4065565627716379,
+      "grad_norm": 0.25720641016960144,
+      "learning_rate": 5e-05,
+      "loss": 2.1471,
+      "step": 3274
+    },
+    {
+      "epoch": 0.40668074009685834,
+      "grad_norm": 0.2818875014781952,
+      "learning_rate": 5e-05,
+      "loss": 2.2734,
+      "step": 3275
+    },
+    {
+      "epoch": 0.40680491742207875,
+      "grad_norm": 0.25783732533454895,
+      "learning_rate": 5e-05,
+      "loss": 2.1214,
+      "step": 3276
+    },
+    {
+      "epoch": 0.40692909474729916,
+      "grad_norm": 0.2418941706418991,
+      "learning_rate": 5e-05,
+      "loss": 1.9046,
+      "step": 3277
+    },
+    {
+      "epoch": 0.40705327207251957,
+      "grad_norm": 0.42483940720558167,
+      "learning_rate": 5e-05,
+      "loss": 3.0115,
+      "step": 3278
+    },
+    {
+      "epoch": 0.40717744939774,
+      "grad_norm": 0.2835560142993927,
+      "learning_rate": 5e-05,
+      "loss": 2.2405,
+      "step": 3279
+    },
+    {
+      "epoch": 0.4073016267229604,
+      "grad_norm": 0.2531037926673889,
+      "learning_rate": 5e-05,
+      "loss": 1.7515,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4074258040481808,
+      "grad_norm": 0.6472195386886597,
+      "learning_rate": 5e-05,
+      "loss": 3.4052,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4075499813734012,
+      "grad_norm": 0.31920236349105835,
+      "learning_rate": 5e-05,
+      "loss": 2.6918,
+      "step": 3282
+    },
+    {
+      "epoch": 0.4076741586986216,
+      "grad_norm": 0.26303863525390625,
+      "learning_rate": 5e-05,
+      "loss": 2.403,
+      "step": 3283
+    },
+    {
+      "epoch": 0.407798336023842,
+      "grad_norm": 0.23327696323394775,
+      "learning_rate": 5e-05,
+      "loss": 1.6611,
+      "step": 3284
+    },
+    {
+      "epoch": 0.4079225133490625,
+      "grad_norm": 0.28720566630363464,
+      "learning_rate": 5e-05,
+      "loss": 2.842,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4080466906742829,
+      "grad_norm": 0.24408192932605743,
+      "learning_rate": 5e-05,
+      "loss": 1.6506,
+      "step": 3286
+    },
+    {
+      "epoch": 0.4081708679995033,
+      "grad_norm": 0.3013565242290497,
+      "learning_rate": 5e-05,
+      "loss": 2.0958,
+      "step": 3287
+    },
+    {
+      "epoch": 0.4082950453247237,
+      "grad_norm": 0.2916112542152405,
+      "learning_rate": 5e-05,
+      "loss": 2.3352,
+      "step": 3288
+    },
+    {
+      "epoch": 0.4084192226499441,
+      "grad_norm": 0.3762753903865814,
+      "learning_rate": 5e-05,
+      "loss": 2.6158,
+      "step": 3289
+    },
+    {
+      "epoch": 0.40854339997516453,
+      "grad_norm": 0.28964895009994507,
+      "learning_rate": 5e-05,
+      "loss": 2.1702,
+      "step": 3290
+    },
+    {
+      "epoch": 0.40866757730038494,
+      "grad_norm": 0.26771649718284607,
+      "learning_rate": 5e-05,
+      "loss": 1.7568,
+      "step": 3291
+    },
+    {
+      "epoch": 0.40879175462560535,
+      "grad_norm": 0.2647267282009125,
+      "learning_rate": 5e-05,
+      "loss": 1.7622,
+      "step": 3292
+    },
+    {
+      "epoch": 0.40891593195082576,
+      "grad_norm": 0.27162161469459534,
+      "learning_rate": 5e-05,
+      "loss": 1.8978,
+      "step": 3293
+    },
+    {
+      "epoch": 0.40904010927604617,
+      "grad_norm": 0.33049115538597107,
+      "learning_rate": 5e-05,
+      "loss": 2.6634,
+      "step": 3294
+    },
+    {
+      "epoch": 0.40916428660126664,
+      "grad_norm": 0.21929199993610382,
+      "learning_rate": 5e-05,
+      "loss": 1.6448,
+      "step": 3295
+    },
+    {
+      "epoch": 0.40928846392648704,
+      "grad_norm": 0.25452035665512085,
+      "learning_rate": 5e-05,
+      "loss": 1.8088,
+      "step": 3296
+    },
+    {
+      "epoch": 0.40941264125170745,
+      "grad_norm": 0.24368155002593994,
+      "learning_rate": 5e-05,
+      "loss": 1.8273,
+      "step": 3297
+    },
+    {
+      "epoch": 0.40953681857692786,
+      "grad_norm": 0.2961564064025879,
+      "learning_rate": 5e-05,
+      "loss": 2.4794,
+      "step": 3298
+    },
+    {
+      "epoch": 0.4096609959021483,
+      "grad_norm": 0.24463674426078796,
+      "learning_rate": 5e-05,
+      "loss": 1.9223,
+      "step": 3299
+    },
+    {
+      "epoch": 0.4097851732273687,
+      "grad_norm": 0.2337009757757187,
+      "learning_rate": 5e-05,
+      "loss": 1.8809,
+      "step": 3300
+    },
+    {
+      "epoch": 0.4099093505525891,
+      "grad_norm": 0.24919812381267548,
+      "learning_rate": 5e-05,
+      "loss": 2.1978,
+      "step": 3301
+    },
+    {
+      "epoch": 0.4100335278778095,
+      "grad_norm": 0.24201802909374237,
+      "learning_rate": 5e-05,
+      "loss": 1.7145,
+      "step": 3302
+    },
+    {
+      "epoch": 0.4101577052030299,
+      "grad_norm": 0.32483869791030884,
+      "learning_rate": 5e-05,
+      "loss": 2.4252,
+      "step": 3303
+    },
+    {
+      "epoch": 0.4102818825282503,
+      "grad_norm": 0.4176599979400635,
+      "learning_rate": 5e-05,
+      "loss": 2.7853,
+      "step": 3304
+    },
+    {
+      "epoch": 0.4104060598534708,
+      "grad_norm": 0.24182365834712982,
+      "learning_rate": 5e-05,
+      "loss": 1.7093,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4105302371786912,
+      "grad_norm": 0.24902278184890747,
+      "learning_rate": 5e-05,
+      "loss": 1.7321,
+      "step": 3306
+    },
+    {
+      "epoch": 0.4106544145039116,
+      "grad_norm": 0.2560373842716217,
+      "learning_rate": 5e-05,
+      "loss": 1.8098,
+      "step": 3307
+    },
+    {
+      "epoch": 0.410778591829132,
+      "grad_norm": 0.23286306858062744,
+      "learning_rate": 5e-05,
+      "loss": 1.6582,
+      "step": 3308
+    },
+    {
+      "epoch": 0.4109027691543524,
+      "grad_norm": 0.2629183530807495,
+      "learning_rate": 5e-05,
+      "loss": 2.2551,
+      "step": 3309
+    },
+    {
+      "epoch": 0.41102694647957283,
+      "grad_norm": 0.29891636967658997,
+      "learning_rate": 5e-05,
+      "loss": 2.641,
+      "step": 3310
+    },
+    {
+      "epoch": 0.41115112380479324,
+      "grad_norm": 0.3004927933216095,
+      "learning_rate": 5e-05,
+      "loss": 2.2943,
+      "step": 3311
+    },
+    {
+      "epoch": 0.41127530113001365,
+      "grad_norm": 0.2863631844520569,
+      "learning_rate": 5e-05,
+      "loss": 2.0059,
+      "step": 3312
+    },
+    {
+      "epoch": 0.41139947845523406,
+      "grad_norm": 0.2744470238685608,
+      "learning_rate": 5e-05,
+      "loss": 2.5977,
+      "step": 3313
+    },
+    {
+      "epoch": 0.41152365578045447,
+      "grad_norm": 0.23051901161670685,
+      "learning_rate": 5e-05,
+      "loss": 1.6559,
+      "step": 3314
+    },
+    {
+      "epoch": 0.4116478331056749,
+      "grad_norm": 0.2972249686717987,
+      "learning_rate": 5e-05,
+      "loss": 2.3126,
+      "step": 3315
+    },
+    {
+      "epoch": 0.41177201043089534,
+      "grad_norm": 0.31656894087791443,
+      "learning_rate": 5e-05,
+      "loss": 2.7207,
+      "step": 3316
+    },
+    {
+      "epoch": 0.41189618775611575,
+      "grad_norm": 0.23948752880096436,
+      "learning_rate": 5e-05,
+      "loss": 1.7337,
+      "step": 3317
+    },
+    {
+      "epoch": 0.41202036508133616,
+      "grad_norm": 0.22887474298477173,
+      "learning_rate": 5e-05,
+      "loss": 1.8173,
+      "step": 3318
+    },
+    {
+      "epoch": 0.41214454240655657,
+      "grad_norm": 0.2630671560764313,
+      "learning_rate": 5e-05,
+      "loss": 1.7794,
+      "step": 3319
+    },
+    {
+      "epoch": 0.412268719731777,
+      "grad_norm": 0.22800284624099731,
+      "learning_rate": 5e-05,
+      "loss": 1.4876,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4123928970569974,
+      "grad_norm": 0.2434392273426056,
+      "learning_rate": 5e-05,
+      "loss": 1.8354,
+      "step": 3321
+    },
+    {
+      "epoch": 0.4125170743822178,
+      "grad_norm": 0.2796170115470886,
+      "learning_rate": 5e-05,
+      "loss": 2.0999,
+      "step": 3322
+    },
+    {
+      "epoch": 0.4126412517074382,
+      "grad_norm": 0.24913708865642548,
+      "learning_rate": 5e-05,
+      "loss": 1.8913,
+      "step": 3323
+    },
+    {
+      "epoch": 0.4127654290326586,
+      "grad_norm": 0.23611408472061157,
+      "learning_rate": 5e-05,
+      "loss": 1.6581,
+      "step": 3324
+    },
+    {
+      "epoch": 0.412889606357879,
+      "grad_norm": 0.25136780738830566,
+      "learning_rate": 5e-05,
+      "loss": 2.2604,
+      "step": 3325
+    },
+    {
+      "epoch": 0.4130137836830995,
+      "grad_norm": 0.40644875168800354,
+      "learning_rate": 5e-05,
+      "loss": 2.4211,
+      "step": 3326
+    },
+    {
+      "epoch": 0.4131379610083199,
+      "grad_norm": 0.2564259171485901,
+      "learning_rate": 5e-05,
+      "loss": 1.5958,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4132621383335403,
+      "grad_norm": 0.292250394821167,
+      "learning_rate": 5e-05,
+      "loss": 1.8867,
+      "step": 3328
+    },
+    {
+      "epoch": 0.4133863156587607,
+      "grad_norm": 0.27501770853996277,
+      "learning_rate": 5e-05,
+      "loss": 2.0115,
+      "step": 3329
+    },
+    {
+      "epoch": 0.4135104929839811,
+      "grad_norm": 0.26127877831459045,
+      "learning_rate": 5e-05,
+      "loss": 1.7981,
+      "step": 3330
+    },
+    {
+      "epoch": 0.41363467030920154,
+      "grad_norm": 0.213522806763649,
+      "learning_rate": 5e-05,
+      "loss": 1.416,
+      "step": 3331
+    },
+    {
+      "epoch": 0.41375884763442194,
+      "grad_norm": 0.2548137605190277,
+      "learning_rate": 5e-05,
+      "loss": 1.9685,
+      "step": 3332
+    },
+    {
+      "epoch": 0.41388302495964235,
+      "grad_norm": 0.25160858035087585,
+      "learning_rate": 5e-05,
+      "loss": 2.066,
+      "step": 3333
+    },
+    {
+      "epoch": 0.41400720228486276,
+      "grad_norm": 0.2430408000946045,
+      "learning_rate": 5e-05,
+      "loss": 1.9191,
+      "step": 3334
+    },
+    {
+      "epoch": 0.4141313796100832,
+      "grad_norm": 0.2609919011592865,
+      "learning_rate": 5e-05,
+      "loss": 1.7538,
+      "step": 3335
+    },
+    {
+      "epoch": 0.41425555693530364,
+      "grad_norm": 0.2683079242706299,
+      "learning_rate": 5e-05,
+      "loss": 1.7518,
+      "step": 3336
+    },
+    {
+      "epoch": 0.41437973426052405,
+      "grad_norm": 0.28265538811683655,
+      "learning_rate": 5e-05,
+      "loss": 2.15,
+      "step": 3337
+    },
+    {
+      "epoch": 0.41450391158574446,
+      "grad_norm": 0.24519746005535126,
+      "learning_rate": 5e-05,
+      "loss": 1.6722,
+      "step": 3338
+    },
+    {
+      "epoch": 0.41462808891096486,
+      "grad_norm": 0.2990780174732208,
+      "learning_rate": 5e-05,
+      "loss": 2.0942,
+      "step": 3339
+    },
+    {
+      "epoch": 0.4147522662361853,
+      "grad_norm": 0.3654913902282715,
+      "learning_rate": 5e-05,
+      "loss": 2.9664,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4148764435614057,
+      "grad_norm": 0.24000787734985352,
+      "learning_rate": 5e-05,
+      "loss": 1.9836,
+      "step": 3341
+    },
+    {
+      "epoch": 0.4150006208866261,
+      "grad_norm": 0.24698716402053833,
+      "learning_rate": 5e-05,
+      "loss": 1.674,
+      "step": 3342
+    },
+    {
+      "epoch": 0.4151247982118465,
+      "grad_norm": 0.2373656928539276,
+      "learning_rate": 5e-05,
+      "loss": 1.728,
+      "step": 3343
+    },
+    {
+      "epoch": 0.4152489755370669,
+      "grad_norm": 0.26102742552757263,
+      "learning_rate": 5e-05,
+      "loss": 2.0658,
+      "step": 3344
+    },
+    {
+      "epoch": 0.4153731528622873,
+      "grad_norm": 0.2378949373960495,
+      "learning_rate": 5e-05,
+      "loss": 1.8418,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4154973301875078,
+      "grad_norm": 0.2490576207637787,
+      "learning_rate": 5e-05,
+      "loss": 1.821,
+      "step": 3346
+    },
+    {
+      "epoch": 0.4156215075127282,
+      "grad_norm": 0.25114601850509644,
+      "learning_rate": 5e-05,
+      "loss": 2.0123,
+      "step": 3347
+    },
+    {
+      "epoch": 0.4157456848379486,
+      "grad_norm": 0.3963528573513031,
+      "learning_rate": 5e-05,
+      "loss": 2.5569,
+      "step": 3348
+    },
+    {
+      "epoch": 0.415869862163169,
+      "grad_norm": 0.2629963457584381,
+      "learning_rate": 5e-05,
+      "loss": 1.6892,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4159940394883894,
+      "grad_norm": 0.23557481169700623,
+      "learning_rate": 5e-05,
+      "loss": 1.8269,
+      "step": 3350
+    },
+    {
+      "epoch": 0.41611821681360983,
+      "grad_norm": 0.2852698862552643,
+      "learning_rate": 5e-05,
+      "loss": 1.8967,
+      "step": 3351
+    },
+    {
+      "epoch": 0.41624239413883024,
+      "grad_norm": 0.3937217891216278,
+      "learning_rate": 5e-05,
+      "loss": 2.485,
+      "step": 3352
+    },
+    {
+      "epoch": 0.41636657146405065,
+      "grad_norm": 0.25239717960357666,
+      "learning_rate": 5e-05,
+      "loss": 1.6328,
+      "step": 3353
+    },
+    {
+      "epoch": 0.41649074878927106,
+      "grad_norm": 0.30600252747535706,
+      "learning_rate": 5e-05,
+      "loss": 1.7857,
+      "step": 3354
+    },
+    {
+      "epoch": 0.41661492611449147,
+      "grad_norm": 0.2959606945514679,
+      "learning_rate": 5e-05,
+      "loss": 1.8176,
+      "step": 3355
+    },
+    {
+      "epoch": 0.41673910343971193,
+      "grad_norm": 0.3044343590736389,
+      "learning_rate": 5e-05,
+      "loss": 1.9752,
+      "step": 3356
+    },
+    {
+      "epoch": 0.41686328076493234,
+      "grad_norm": 0.2534724771976471,
+      "learning_rate": 5e-05,
+      "loss": 2.3616,
+      "step": 3357
+    },
+    {
+      "epoch": 0.41698745809015275,
+      "grad_norm": 0.26968103647232056,
+      "learning_rate": 5e-05,
+      "loss": 1.9957,
+      "step": 3358
+    },
+    {
+      "epoch": 0.41711163541537316,
+      "grad_norm": 0.3110410273075104,
+      "learning_rate": 5e-05,
+      "loss": 2.2647,
+      "step": 3359
+    },
+    {
+      "epoch": 0.41723581274059357,
+      "grad_norm": 0.47166141867637634,
+      "learning_rate": 5e-05,
+      "loss": 2.5188,
+      "step": 3360
+    },
+    {
+      "epoch": 0.417359990065814,
+      "grad_norm": 0.331869512796402,
+      "learning_rate": 5e-05,
+      "loss": 2.0572,
+      "step": 3361
+    },
+    {
+      "epoch": 0.4174841673910344,
+      "grad_norm": 0.31614527106285095,
+      "learning_rate": 5e-05,
+      "loss": 2.462,
+      "step": 3362
+    },
+    {
+      "epoch": 0.4176083447162548,
+      "grad_norm": 0.24280445277690887,
+      "learning_rate": 5e-05,
+      "loss": 1.9101,
+      "step": 3363
+    },
+    {
+      "epoch": 0.4177325220414752,
+      "grad_norm": 0.28824475407600403,
+      "learning_rate": 5e-05,
+      "loss": 2.279,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4178566993666956,
+      "grad_norm": 0.2633785307407379,
+      "learning_rate": 5e-05,
+      "loss": 1.8472,
+      "step": 3365
+    },
+    {
+      "epoch": 0.4179808766919161,
+      "grad_norm": 0.2650469243526459,
+      "learning_rate": 5e-05,
+      "loss": 1.8278,
+      "step": 3366
+    },
+    {
+      "epoch": 0.4181050540171365,
+      "grad_norm": 0.2833096385002136,
+      "learning_rate": 5e-05,
+      "loss": 1.8323,
+      "step": 3367
+    },
+    {
+      "epoch": 0.4182292313423569,
+      "grad_norm": 0.23373261094093323,
+      "learning_rate": 5e-05,
+      "loss": 1.8023,
+      "step": 3368
+    },
+    {
+      "epoch": 0.4183534086675773,
+      "grad_norm": 0.24790047109127045,
+      "learning_rate": 5e-05,
+      "loss": 1.9374,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4184775859927977,
+      "grad_norm": 0.4289436340332031,
+      "learning_rate": 5e-05,
+      "loss": 2.9487,
+      "step": 3370
+    },
+    {
+      "epoch": 0.4186017633180181,
+      "grad_norm": 0.24313776195049286,
+      "learning_rate": 5e-05,
+      "loss": 1.7685,
+      "step": 3371
+    },
+    {
+      "epoch": 0.41872594064323854,
+      "grad_norm": 0.24773679673671722,
+      "learning_rate": 5e-05,
+      "loss": 1.8596,
+      "step": 3372
+    },
+    {
+      "epoch": 0.41885011796845895,
+      "grad_norm": 0.25414812564849854,
+      "learning_rate": 5e-05,
+      "loss": 2.0478,
+      "step": 3373
+    },
+    {
+      "epoch": 0.41897429529367936,
+      "grad_norm": 0.2624247670173645,
+      "learning_rate": 5e-05,
+      "loss": 2.2904,
+      "step": 3374
+    },
+    {
+      "epoch": 0.41909847261889976,
+      "grad_norm": 0.26548388600349426,
+      "learning_rate": 5e-05,
+      "loss": 2.1267,
+      "step": 3375
+    },
+    {
+      "epoch": 0.41922264994412023,
+      "grad_norm": 0.2785142958164215,
+      "learning_rate": 5e-05,
+      "loss": 1.8295,
+      "step": 3376
+    },
+    {
+      "epoch": 0.41934682726934064,
+      "grad_norm": 0.23659725487232208,
+      "learning_rate": 5e-05,
+      "loss": 1.7218,
+      "step": 3377
+    },
+    {
+      "epoch": 0.41947100459456105,
+      "grad_norm": 0.26557525992393494,
+      "learning_rate": 5e-05,
+      "loss": 2.4336,
+      "step": 3378
+    },
+    {
+      "epoch": 0.41959518191978146,
+      "grad_norm": 0.2445652037858963,
+      "learning_rate": 5e-05,
+      "loss": 1.7085,
+      "step": 3379
+    },
+    {
+      "epoch": 0.41971935924500187,
+      "grad_norm": 0.3549094498157501,
+      "learning_rate": 5e-05,
+      "loss": 2.2662,
+      "step": 3380
+    },
+    {
+      "epoch": 0.4198435365702223,
+      "grad_norm": 0.22882160544395447,
+      "learning_rate": 5e-05,
+      "loss": 1.7271,
+      "step": 3381
+    },
+    {
+      "epoch": 0.4199677138954427,
+      "grad_norm": 0.27897635102272034,
+      "learning_rate": 5e-05,
+      "loss": 2.3793,
+      "step": 3382
+    },
+    {
+      "epoch": 0.4200918912206631,
+      "grad_norm": 0.2471127063035965,
+      "learning_rate": 5e-05,
+      "loss": 2.0686,
+      "step": 3383
+    },
+    {
+      "epoch": 0.4202160685458835,
+      "grad_norm": 0.2623070478439331,
+      "learning_rate": 5e-05,
+      "loss": 2.0878,
+      "step": 3384
+    },
+    {
+      "epoch": 0.4203402458711039,
+      "grad_norm": 0.23333360254764557,
+      "learning_rate": 5e-05,
+      "loss": 1.9169,
+      "step": 3385
+    },
+    {
+      "epoch": 0.4204644231963244,
+      "grad_norm": 0.3164125382900238,
+      "learning_rate": 5e-05,
+      "loss": 2.664,
+      "step": 3386
+    },
+    {
+      "epoch": 0.4205886005215448,
+      "grad_norm": 0.24894979596138,
+      "learning_rate": 5e-05,
+      "loss": 2.116,
+      "step": 3387
+    },
+    {
+      "epoch": 0.4207127778467652,
+      "grad_norm": 0.23133230209350586,
+      "learning_rate": 5e-05,
+      "loss": 2.0747,
+      "step": 3388
+    },
+    {
+      "epoch": 0.4208369551719856,
+      "grad_norm": 0.44651320576667786,
+      "learning_rate": 5e-05,
+      "loss": 3.222,
+      "step": 3389
+    },
+    {
+      "epoch": 0.420961132497206,
+      "grad_norm": 0.25321146845817566,
+      "learning_rate": 5e-05,
+      "loss": 1.806,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4210853098224264,
+      "grad_norm": 0.28276702761650085,
+      "learning_rate": 5e-05,
+      "loss": 1.8903,
+      "step": 3391
+    },
+    {
+      "epoch": 0.42120948714764683,
+      "grad_norm": 0.33389943838119507,
+      "learning_rate": 5e-05,
+      "loss": 1.9715,
+      "step": 3392
+    },
+    {
+      "epoch": 0.42133366447286724,
+      "grad_norm": 0.2813355624675751,
+      "learning_rate": 5e-05,
+      "loss": 1.9086,
+      "step": 3393
+    },
+    {
+      "epoch": 0.42145784179808765,
+      "grad_norm": 0.4019930958747864,
+      "learning_rate": 5e-05,
+      "loss": 2.6213,
+      "step": 3394
+    },
+    {
+      "epoch": 0.42158201912330806,
+      "grad_norm": 0.25617924332618713,
+      "learning_rate": 5e-05,
+      "loss": 1.9279,
+      "step": 3395
+    },
+    {
+      "epoch": 0.4217061964485285,
+      "grad_norm": 0.2738133370876312,
+      "learning_rate": 5e-05,
+      "loss": 1.8295,
+      "step": 3396
+    },
+    {
+      "epoch": 0.42183037377374893,
+      "grad_norm": 0.281930536031723,
+      "learning_rate": 5e-05,
+      "loss": 2.0109,
+      "step": 3397
+    },
+    {
+      "epoch": 0.42195455109896934,
+      "grad_norm": 0.25403085350990295,
+      "learning_rate": 5e-05,
+      "loss": 1.6643,
+      "step": 3398
+    },
+    {
+      "epoch": 0.42207872842418975,
+      "grad_norm": 0.25405240058898926,
+      "learning_rate": 5e-05,
+      "loss": 2.0965,
+      "step": 3399
+    },
+    {
+      "epoch": 0.42220290574941016,
+      "grad_norm": 0.2487897425889969,
+      "learning_rate": 5e-05,
+      "loss": 2.0132,
+      "step": 3400
+    },
+    {
+      "epoch": 0.42232708307463057,
+      "grad_norm": 0.2592264711856842,
+      "learning_rate": 5e-05,
+      "loss": 2.0487,
+      "step": 3401
+    },
+    {
+      "epoch": 0.422451260399851,
+      "grad_norm": 0.2716180384159088,
+      "learning_rate": 5e-05,
+      "loss": 1.9696,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4225754377250714,
+      "grad_norm": 0.2793247699737549,
+      "learning_rate": 5e-05,
+      "loss": 2.581,
+      "step": 3403
+    },
+    {
+      "epoch": 0.4226996150502918,
+      "grad_norm": 0.22381696105003357,
+      "learning_rate": 5e-05,
+      "loss": 1.7447,
+      "step": 3404
+    },
+    {
+      "epoch": 0.4228237923755122,
+      "grad_norm": 0.5200352668762207,
+      "learning_rate": 5e-05,
+      "loss": 3.9713,
+      "step": 3405
+    },
+    {
+      "epoch": 0.4229479697007327,
+      "grad_norm": 0.2567685842514038,
+      "learning_rate": 5e-05,
+      "loss": 1.8026,
+      "step": 3406
+    },
+    {
+      "epoch": 0.4230721470259531,
+      "grad_norm": 0.23617665469646454,
+      "learning_rate": 5e-05,
+      "loss": 1.6888,
+      "step": 3407
+    },
+    {
+      "epoch": 0.4231963243511735,
+      "grad_norm": 0.2345484048128128,
+      "learning_rate": 5e-05,
+      "loss": 1.8151,
+      "step": 3408
+    },
+    {
+      "epoch": 0.4233205016763939,
+      "grad_norm": 0.2495543509721756,
+      "learning_rate": 5e-05,
+      "loss": 1.8372,
+      "step": 3409
+    },
+    {
+      "epoch": 0.4234446790016143,
+      "grad_norm": 0.38246211409568787,
+      "learning_rate": 5e-05,
+      "loss": 2.4178,
+      "step": 3410
+    },
+    {
+      "epoch": 0.4235688563268347,
+      "grad_norm": 0.30209487676620483,
+      "learning_rate": 5e-05,
+      "loss": 2.5151,
+      "step": 3411
+    },
+    {
+      "epoch": 0.42369303365205513,
+      "grad_norm": 0.25495272874832153,
+      "learning_rate": 5e-05,
+      "loss": 2.1185,
+      "step": 3412
+    },
+    {
+      "epoch": 0.42381721097727554,
+      "grad_norm": 0.2760424315929413,
+      "learning_rate": 5e-05,
+      "loss": 1.8604,
+      "step": 3413
+    },
+    {
+      "epoch": 0.42394138830249595,
+      "grad_norm": 0.27497777342796326,
+      "learning_rate": 5e-05,
+      "loss": 2.052,
+      "step": 3414
+    },
+    {
+      "epoch": 0.42406556562771636,
+      "grad_norm": 0.34112584590911865,
+      "learning_rate": 5e-05,
+      "loss": 2.0636,
+      "step": 3415
+    },
+    {
+      "epoch": 0.4241897429529368,
+      "grad_norm": 0.2603628933429718,
+      "learning_rate": 5e-05,
+      "loss": 1.8236,
+      "step": 3416
+    },
+    {
+      "epoch": 0.42431392027815723,
+      "grad_norm": 0.2946011424064636,
+      "learning_rate": 5e-05,
+      "loss": 2.8446,
+      "step": 3417
+    },
+    {
+      "epoch": 0.42443809760337764,
+      "grad_norm": 0.2663310170173645,
+      "learning_rate": 5e-05,
+      "loss": 1.8928,
+      "step": 3418
+    },
+    {
+      "epoch": 0.42456227492859805,
+      "grad_norm": 0.2847248911857605,
+      "learning_rate": 5e-05,
+      "loss": 2.1012,
+      "step": 3419
+    },
+    {
+      "epoch": 0.42468645225381846,
+      "grad_norm": 0.26694995164871216,
+      "learning_rate": 5e-05,
+      "loss": 1.9227,
+      "step": 3420
+    },
+    {
+      "epoch": 0.42481062957903887,
+      "grad_norm": 0.24277377128601074,
+      "learning_rate": 5e-05,
+      "loss": 1.5964,
+      "step": 3421
+    },
+    {
+      "epoch": 0.4249348069042593,
+      "grad_norm": 0.271227091550827,
+      "learning_rate": 5e-05,
+      "loss": 2.174,
+      "step": 3422
+    },
+    {
+      "epoch": 0.4250589842294797,
+      "grad_norm": 0.23910784721374512,
+      "learning_rate": 5e-05,
+      "loss": 1.8477,
+      "step": 3423
+    },
+    {
+      "epoch": 0.4251831615547001,
+      "grad_norm": 0.2732751965522766,
+      "learning_rate": 5e-05,
+      "loss": 2.191,
+      "step": 3424
+    },
+    {
+      "epoch": 0.4253073388799205,
+      "grad_norm": 0.23070304095745087,
+      "learning_rate": 5e-05,
+      "loss": 1.7937,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4254315162051409,
+      "grad_norm": 0.2512621581554413,
+      "learning_rate": 5e-05,
+      "loss": 1.7886,
+      "step": 3426
+    },
+    {
+      "epoch": 0.4255556935303614,
+      "grad_norm": 0.2547200620174408,
+      "learning_rate": 5e-05,
+      "loss": 2.0049,
+      "step": 3427
+    },
+    {
+      "epoch": 0.4256798708555818,
+      "grad_norm": 0.2578077018260956,
+      "learning_rate": 5e-05,
+      "loss": 1.8749,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4258040481808022,
+      "grad_norm": 0.24885417520999908,
+      "learning_rate": 5e-05,
+      "loss": 1.864,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4259282255060226,
+      "grad_norm": 0.24219416081905365,
+      "learning_rate": 5e-05,
+      "loss": 1.5677,
+      "step": 3430
+    },
+    {
+      "epoch": 0.426052402831243,
+      "grad_norm": 0.2584877014160156,
+      "learning_rate": 5e-05,
+      "loss": 2.0624,
+      "step": 3431
+    },
+    {
+      "epoch": 0.4261765801564634,
+      "grad_norm": 0.2401028722524643,
+      "learning_rate": 5e-05,
+      "loss": 1.614,
+      "step": 3432
+    },
+    {
+      "epoch": 0.42630075748168383,
+      "grad_norm": 0.23328331112861633,
+      "learning_rate": 5e-05,
+      "loss": 2.0586,
+      "step": 3433
+    },
+    {
+      "epoch": 0.42642493480690424,
+      "grad_norm": 0.24837568402290344,
+      "learning_rate": 5e-05,
+      "loss": 1.7939,
+      "step": 3434
+    },
+    {
+      "epoch": 0.42654911213212465,
+      "grad_norm": 0.23399226367473602,
+      "learning_rate": 5e-05,
+      "loss": 1.7452,
+      "step": 3435
+    },
+    {
+      "epoch": 0.42667328945734506,
+      "grad_norm": 0.235508531332016,
+      "learning_rate": 5e-05,
+      "loss": 1.6515,
+      "step": 3436
+    },
+    {
+      "epoch": 0.4267974667825655,
+      "grad_norm": 0.260160356760025,
+      "learning_rate": 5e-05,
+      "loss": 2.0762,
+      "step": 3437
+    },
+    {
+      "epoch": 0.42692164410778594,
+      "grad_norm": 0.5180080533027649,
+      "learning_rate": 5e-05,
+      "loss": 2.9584,
+      "step": 3438
+    },
+    {
+      "epoch": 0.42704582143300635,
+      "grad_norm": 0.2382991462945938,
+      "learning_rate": 5e-05,
+      "loss": 1.9452,
+      "step": 3439
+    },
+    {
+      "epoch": 0.42716999875822675,
+      "grad_norm": 0.2545469105243683,
+      "learning_rate": 5e-05,
+      "loss": 1.7956,
+      "step": 3440
+    },
+    {
+      "epoch": 0.42729417608344716,
+      "grad_norm": 0.22551454603672028,
+      "learning_rate": 5e-05,
+      "loss": 1.6773,
+      "step": 3441
+    },
+    {
+      "epoch": 0.4274183534086676,
+      "grad_norm": 0.2589209973812103,
+      "learning_rate": 5e-05,
+      "loss": 2.4389,
+      "step": 3442
+    },
+    {
+      "epoch": 0.427542530733888,
+      "grad_norm": 0.2337065041065216,
+      "learning_rate": 5e-05,
+      "loss": 1.6075,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4276667080591084,
+      "grad_norm": 0.23506352305412292,
+      "learning_rate": 5e-05,
+      "loss": 1.4543,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4277908853843288,
+      "grad_norm": 0.2294839322566986,
+      "learning_rate": 5e-05,
+      "loss": 1.6723,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4279150627095492,
+      "grad_norm": 0.2340175211429596,
+      "learning_rate": 5e-05,
+      "loss": 1.8835,
+      "step": 3446
+    },
+    {
+      "epoch": 0.4280392400347697,
+      "grad_norm": 0.24174371361732483,
+      "learning_rate": 5e-05,
+      "loss": 1.9464,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4281634173599901,
+      "grad_norm": 0.29117172956466675,
+      "learning_rate": 5e-05,
+      "loss": 2.0613,
+      "step": 3448
+    },
+    {
+      "epoch": 0.4282875946852105,
+      "grad_norm": 0.2395106703042984,
+      "learning_rate": 5e-05,
+      "loss": 1.8796,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4284117720104309,
+      "grad_norm": 0.2718517780303955,
+      "learning_rate": 5e-05,
+      "loss": 2.1484,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4285359493356513,
+      "grad_norm": 0.24703308939933777,
+      "learning_rate": 5e-05,
+      "loss": 1.6969,
+      "step": 3451
+    },
+    {
+      "epoch": 0.4286601266608717,
+      "grad_norm": 0.2534920871257782,
+      "learning_rate": 5e-05,
+      "loss": 1.9946,
+      "step": 3452
+    },
+    {
+      "epoch": 0.42878430398609213,
+      "grad_norm": 0.22707927227020264,
+      "learning_rate": 5e-05,
+      "loss": 1.8416,
+      "step": 3453
+    },
+    {
+      "epoch": 0.42890848131131254,
+      "grad_norm": 0.3400515019893646,
+      "learning_rate": 5e-05,
+      "loss": 2.4781,
+      "step": 3454
+    },
+    {
+      "epoch": 0.42903265863653295,
+      "grad_norm": 0.24067482352256775,
+      "learning_rate": 5e-05,
+      "loss": 1.686,
+      "step": 3455
+    },
+    {
+      "epoch": 0.42915683596175336,
+      "grad_norm": 0.2489548921585083,
+      "learning_rate": 5e-05,
+      "loss": 1.7761,
+      "step": 3456
+    },
+    {
+      "epoch": 0.4292810132869738,
+      "grad_norm": 0.279099702835083,
+      "learning_rate": 5e-05,
+      "loss": 2.1486,
+      "step": 3457
+    },
+    {
+      "epoch": 0.42940519061219423,
+      "grad_norm": 0.26058727502822876,
+      "learning_rate": 5e-05,
+      "loss": 1.916,
+      "step": 3458
+    },
+    {
+      "epoch": 0.42952936793741464,
+      "grad_norm": 0.24335864186286926,
+      "learning_rate": 5e-05,
+      "loss": 1.7841,
+      "step": 3459
+    },
+    {
+      "epoch": 0.42965354526263505,
+      "grad_norm": 0.23764993250370026,
+      "learning_rate": 5e-05,
+      "loss": 1.5529,
+      "step": 3460
+    },
+    {
+      "epoch": 0.42977772258785546,
+      "grad_norm": 0.24684032797813416,
+      "learning_rate": 5e-05,
+      "loss": 1.8382,
+      "step": 3461
+    },
+    {
+      "epoch": 0.42990189991307587,
+      "grad_norm": 0.2514871060848236,
+      "learning_rate": 5e-05,
+      "loss": 2.1144,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4300260772382963,
+      "grad_norm": 0.49423691630363464,
+      "learning_rate": 5e-05,
+      "loss": 3.249,
+      "step": 3463
+    },
+    {
+      "epoch": 0.4301502545635167,
+      "grad_norm": 0.23919124901294708,
+      "learning_rate": 5e-05,
+      "loss": 1.6312,
+      "step": 3464
+    },
+    {
+      "epoch": 0.4302744318887371,
+      "grad_norm": 0.2864331007003784,
+      "learning_rate": 5e-05,
+      "loss": 1.9462,
+      "step": 3465
+    },
+    {
+      "epoch": 0.4303986092139575,
+      "grad_norm": 0.3184867799282074,
+      "learning_rate": 5e-05,
+      "loss": 2.6204,
+      "step": 3466
+    },
+    {
+      "epoch": 0.43052278653917797,
+      "grad_norm": 0.3039170503616333,
+      "learning_rate": 5e-05,
+      "loss": 2.2353,
+      "step": 3467
+    },
+    {
+      "epoch": 0.4306469638643984,
+      "grad_norm": 0.22405976057052612,
+      "learning_rate": 5e-05,
+      "loss": 1.6845,
+      "step": 3468
+    },
+    {
+      "epoch": 0.4307711411896188,
+      "grad_norm": 0.2478765845298767,
+      "learning_rate": 5e-05,
+      "loss": 1.8022,
+      "step": 3469
+    },
+    {
+      "epoch": 0.4308953185148392,
+      "grad_norm": 0.2638530731201172,
+      "learning_rate": 5e-05,
+      "loss": 2.0671,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4310194958400596,
+      "grad_norm": 0.26117587089538574,
+      "learning_rate": 5e-05,
+      "loss": 2.0145,
+      "step": 3471
+    },
+    {
+      "epoch": 0.43114367316528,
+      "grad_norm": 0.25023388862609863,
+      "learning_rate": 5e-05,
+      "loss": 1.9008,
+      "step": 3472
+    },
+    {
+      "epoch": 0.4312678504905004,
+      "grad_norm": 0.25707265734672546,
+      "learning_rate": 5e-05,
+      "loss": 2.142,
+      "step": 3473
+    },
+    {
+      "epoch": 0.43139202781572084,
+      "grad_norm": 0.2396668642759323,
+      "learning_rate": 5e-05,
+      "loss": 1.8341,
+      "step": 3474
+    },
+    {
+      "epoch": 0.43151620514094124,
+      "grad_norm": 0.3423267602920532,
+      "learning_rate": 5e-05,
+      "loss": 2.9799,
+      "step": 3475
+    },
+    {
+      "epoch": 0.43164038246616165,
+      "grad_norm": 0.2501693069934845,
+      "learning_rate": 5e-05,
+      "loss": 1.7378,
+      "step": 3476
+    },
+    {
+      "epoch": 0.4317645597913821,
+      "grad_norm": 0.26930058002471924,
+      "learning_rate": 5e-05,
+      "loss": 2.0415,
+      "step": 3477
+    },
+    {
+      "epoch": 0.43188873711660253,
+      "grad_norm": 0.25953057408332825,
+      "learning_rate": 5e-05,
+      "loss": 1.777,
+      "step": 3478
+    },
+    {
+      "epoch": 0.43201291444182294,
+      "grad_norm": 0.3278453052043915,
+      "learning_rate": 5e-05,
+      "loss": 2.3855,
+      "step": 3479
+    },
+    {
+      "epoch": 0.43213709176704335,
+      "grad_norm": 0.2620750367641449,
+      "learning_rate": 5e-05,
+      "loss": 2.0836,
+      "step": 3480
+    },
+    {
+      "epoch": 0.43226126909226376,
+      "grad_norm": 0.2250550091266632,
+      "learning_rate": 5e-05,
+      "loss": 1.6408,
+      "step": 3481
+    },
+    {
+      "epoch": 0.43238544641748416,
+      "grad_norm": 0.2998541593551636,
+      "learning_rate": 5e-05,
+      "loss": 1.9747,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4325096237427046,
+      "grad_norm": 0.3165997862815857,
+      "learning_rate": 5e-05,
+      "loss": 2.2144,
+      "step": 3483
+    },
+    {
+      "epoch": 0.432633801067925,
+      "grad_norm": 0.2683013379573822,
+      "learning_rate": 5e-05,
+      "loss": 1.6844,
+      "step": 3484
+    },
+    {
+      "epoch": 0.4327579783931454,
+      "grad_norm": 0.2699425220489502,
+      "learning_rate": 5e-05,
+      "loss": 2.0721,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4328821557183658,
+      "grad_norm": 0.3582903742790222,
+      "learning_rate": 5e-05,
+      "loss": 2.5996,
+      "step": 3486
+    },
+    {
+      "epoch": 0.43300633304358627,
+      "grad_norm": 0.26281240582466125,
+      "learning_rate": 5e-05,
+      "loss": 1.79,
+      "step": 3487
+    },
+    {
+      "epoch": 0.4331305103688067,
+      "grad_norm": 0.2991257309913635,
+      "learning_rate": 5e-05,
+      "loss": 1.8693,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4332546876940271,
+      "grad_norm": 0.27996915578842163,
+      "learning_rate": 5e-05,
+      "loss": 1.9153,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4333788650192475,
+      "grad_norm": 0.25853192806243896,
+      "learning_rate": 5e-05,
+      "loss": 1.8618,
+      "step": 3490
+    },
+    {
+      "epoch": 0.4335030423444679,
+      "grad_norm": 0.2344485968351364,
+      "learning_rate": 5e-05,
+      "loss": 1.7231,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4336272196696883,
+      "grad_norm": 0.2821890115737915,
+      "learning_rate": 5e-05,
+      "loss": 1.9678,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4337513969949087,
+      "grad_norm": 0.248769611120224,
+      "learning_rate": 5e-05,
+      "loss": 1.8066,
+      "step": 3493
+    },
+    {
+      "epoch": 0.43387557432012913,
+      "grad_norm": 0.39514604210853577,
+      "learning_rate": 5e-05,
+      "loss": 2.4974,
+      "step": 3494
+    },
+    {
+      "epoch": 0.43399975164534954,
+      "grad_norm": 0.4927578270435333,
+      "learning_rate": 5e-05,
+      "loss": 3.142,
+      "step": 3495
+    },
+    {
+      "epoch": 0.43412392897056995,
+      "grad_norm": 0.23819269239902496,
+      "learning_rate": 5e-05,
+      "loss": 1.9909,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4342481062957904,
+      "grad_norm": 0.23613961040973663,
+      "learning_rate": 5e-05,
+      "loss": 1.7541,
+      "step": 3497
+    },
+    {
+      "epoch": 0.4343722836210108,
+      "grad_norm": 0.2758491337299347,
+      "learning_rate": 5e-05,
+      "loss": 1.8724,
+      "step": 3498
+    },
+    {
+      "epoch": 0.43449646094623123,
+      "grad_norm": 0.2569511830806732,
+      "learning_rate": 5e-05,
+      "loss": 1.8173,
+      "step": 3499
+    },
+    {
+      "epoch": 0.43462063827145164,
+      "grad_norm": 0.27499228715896606,
+      "learning_rate": 5e-05,
+      "loss": 1.8167,
+      "step": 3500
+    },
+    {
+      "epoch": 0.43474481559667205,
+      "grad_norm": 0.2717130482196808,
+      "learning_rate": 5e-05,
+      "loss": 1.997,
+      "step": 3501
+    },
+    {
+      "epoch": 0.43486899292189246,
+      "grad_norm": 0.22984832525253296,
+      "learning_rate": 5e-05,
+      "loss": 1.6562,
+      "step": 3502
+    },
+    {
+      "epoch": 0.43499317024711287,
+      "grad_norm": 0.249094620347023,
+      "learning_rate": 5e-05,
+      "loss": 2.1802,
+      "step": 3503
+    },
+    {
+      "epoch": 0.4351173475723333,
+      "grad_norm": 0.25387057662010193,
+      "learning_rate": 5e-05,
+      "loss": 1.723,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4352415248975537,
+      "grad_norm": 0.2589643895626068,
+      "learning_rate": 5e-05,
+      "loss": 1.9266,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4353657022227741,
+      "grad_norm": 0.24656008183956146,
+      "learning_rate": 5e-05,
+      "loss": 1.6418,
+      "step": 3506
+    },
+    {
+      "epoch": 0.43548987954799456,
+      "grad_norm": 0.2364129275083542,
+      "learning_rate": 5e-05,
+      "loss": 1.8593,
+      "step": 3507
+    },
+    {
+      "epoch": 0.43561405687321497,
+      "grad_norm": 0.2388906031847,
+      "learning_rate": 5e-05,
+      "loss": 1.7846,
+      "step": 3508
+    },
+    {
+      "epoch": 0.4357382341984354,
+      "grad_norm": 0.23374655842781067,
+      "learning_rate": 5e-05,
+      "loss": 1.4973,
+      "step": 3509
+    },
+    {
+      "epoch": 0.4358624115236558,
+      "grad_norm": 0.2885374128818512,
+      "learning_rate": 5e-05,
+      "loss": 2.0409,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4359865888488762,
+      "grad_norm": 0.25313517451286316,
+      "learning_rate": 5e-05,
+      "loss": 1.765,
+      "step": 3511
+    },
+    {
+      "epoch": 0.4361107661740966,
+      "grad_norm": 0.2653907835483551,
+      "learning_rate": 5e-05,
+      "loss": 2.1002,
+      "step": 3512
+    },
+    {
+      "epoch": 0.436234943499317,
+      "grad_norm": 0.29488423466682434,
+      "learning_rate": 5e-05,
+      "loss": 2.4307,
+      "step": 3513
+    },
+    {
+      "epoch": 0.4363591208245374,
+      "grad_norm": 0.3677636384963989,
+      "learning_rate": 5e-05,
+      "loss": 3.1984,
+      "step": 3514
+    },
+    {
+      "epoch": 0.43648329814975784,
+      "grad_norm": 0.22892718017101288,
+      "learning_rate": 5e-05,
+      "loss": 1.8443,
+      "step": 3515
+    },
+    {
+      "epoch": 0.43660747547497825,
+      "grad_norm": 0.26028916239738464,
+      "learning_rate": 5e-05,
+      "loss": 1.9034,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4367316528001987,
+      "grad_norm": 0.3582562506198883,
+      "learning_rate": 5e-05,
+      "loss": 2.6214,
+      "step": 3517
+    },
+    {
+      "epoch": 0.4368558301254191,
+      "grad_norm": 0.29447731375694275,
+      "learning_rate": 5e-05,
+      "loss": 2.8645,
+      "step": 3518
+    },
+    {
+      "epoch": 0.43698000745063953,
+      "grad_norm": 0.26604095101356506,
+      "learning_rate": 5e-05,
+      "loss": 2.0877,
+      "step": 3519
+    },
+    {
+      "epoch": 0.43710418477585994,
+      "grad_norm": 0.24459725618362427,
+      "learning_rate": 5e-05,
+      "loss": 1.9573,
+      "step": 3520
+    },
+    {
+      "epoch": 0.43722836210108035,
+      "grad_norm": 0.2366792857646942,
+      "learning_rate": 5e-05,
+      "loss": 1.7362,
+      "step": 3521
+    },
+    {
+      "epoch": 0.43735253942630076,
+      "grad_norm": 0.2448722869157791,
+      "learning_rate": 5e-05,
+      "loss": 1.9418,
+      "step": 3522
+    },
+    {
+      "epoch": 0.43747671675152117,
+      "grad_norm": 0.2431691735982895,
+      "learning_rate": 5e-05,
+      "loss": 1.8435,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4376008940767416,
+      "grad_norm": 0.2532593905925751,
+      "learning_rate": 5e-05,
+      "loss": 2.2679,
+      "step": 3524
+    },
+    {
+      "epoch": 0.437725071401962,
+      "grad_norm": 0.26807379722595215,
+      "learning_rate": 5e-05,
+      "loss": 2.7244,
+      "step": 3525
+    },
+    {
+      "epoch": 0.4378492487271824,
+      "grad_norm": 0.2699643671512604,
+      "learning_rate": 5e-05,
+      "loss": 2.2042,
+      "step": 3526
+    },
+    {
+      "epoch": 0.43797342605240286,
+      "grad_norm": 0.23252953588962555,
+      "learning_rate": 5e-05,
+      "loss": 1.9205,
+      "step": 3527
+    },
+    {
+      "epoch": 0.43809760337762327,
+      "grad_norm": 0.2471930980682373,
+      "learning_rate": 5e-05,
+      "loss": 2.0198,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4382217807028437,
+      "grad_norm": 0.25804635882377625,
+      "learning_rate": 5e-05,
+      "loss": 2.1311,
+      "step": 3529
+    },
+    {
+      "epoch": 0.4383459580280641,
+      "grad_norm": 0.2536420226097107,
+      "learning_rate": 5e-05,
+      "loss": 1.9975,
+      "step": 3530
+    },
+    {
+      "epoch": 0.4384701353532845,
+      "grad_norm": 0.23809343576431274,
+      "learning_rate": 5e-05,
+      "loss": 1.7388,
+      "step": 3531
+    },
+    {
+      "epoch": 0.4385943126785049,
+      "grad_norm": 0.30638399720191956,
+      "learning_rate": 5e-05,
+      "loss": 2.4784,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4387184900037253,
+      "grad_norm": 0.2595226466655731,
+      "learning_rate": 5e-05,
+      "loss": 2.2256,
+      "step": 3533
+    },
+    {
+      "epoch": 0.4388426673289457,
+      "grad_norm": 0.35969939827919006,
+      "learning_rate": 5e-05,
+      "loss": 2.8512,
+      "step": 3534
+    },
+    {
+      "epoch": 0.43896684465416613,
+      "grad_norm": 0.23786823451519012,
+      "learning_rate": 5e-05,
+      "loss": 1.8273,
+      "step": 3535
+    },
+    {
+      "epoch": 0.43909102197938654,
+      "grad_norm": 0.293653666973114,
+      "learning_rate": 5e-05,
+      "loss": 2.2023,
+      "step": 3536
+    },
+    {
+      "epoch": 0.43921519930460695,
+      "grad_norm": 0.2556871473789215,
+      "learning_rate": 5e-05,
+      "loss": 1.9264,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4393393766298274,
+      "grad_norm": 0.236263245344162,
+      "learning_rate": 5e-05,
+      "loss": 1.8319,
+      "step": 3538
+    },
+    {
+      "epoch": 0.4394635539550478,
+      "grad_norm": 0.24975807964801788,
+      "learning_rate": 5e-05,
+      "loss": 1.8151,
+      "step": 3539
+    },
+    {
+      "epoch": 0.43958773128026823,
+      "grad_norm": 0.2677541971206665,
+      "learning_rate": 5e-05,
+      "loss": 2.1277,
+      "step": 3540
+    },
+    {
+      "epoch": 0.43971190860548864,
+      "grad_norm": 0.24418696761131287,
+      "learning_rate": 5e-05,
+      "loss": 1.7876,
+      "step": 3541
+    },
+    {
+      "epoch": 0.43983608593070905,
+      "grad_norm": 0.272847443819046,
+      "learning_rate": 5e-05,
+      "loss": 2.1651,
+      "step": 3542
+    },
+    {
+      "epoch": 0.43996026325592946,
+      "grad_norm": 0.24691055715084076,
+      "learning_rate": 5e-05,
+      "loss": 2.2033,
+      "step": 3543
+    },
+    {
+      "epoch": 0.44008444058114987,
+      "grad_norm": 0.24864870309829712,
+      "learning_rate": 5e-05,
+      "loss": 2.0583,
+      "step": 3544
+    },
+    {
+      "epoch": 0.4402086179063703,
+      "grad_norm": 0.2729315161705017,
+      "learning_rate": 5e-05,
+      "loss": 2.1486,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4403327952315907,
+      "grad_norm": 0.31387004256248474,
+      "learning_rate": 5e-05,
+      "loss": 2.4992,
+      "step": 3546
+    },
+    {
+      "epoch": 0.4404569725568111,
+      "grad_norm": 0.25067535042762756,
+      "learning_rate": 5e-05,
+      "loss": 2.1168,
+      "step": 3547
+    },
+    {
+      "epoch": 0.44058114988203156,
+      "grad_norm": 0.2915966808795929,
+      "learning_rate": 5e-05,
+      "loss": 2.731,
+      "step": 3548
+    },
+    {
+      "epoch": 0.440705327207252,
+      "grad_norm": 0.4438944458961487,
+      "learning_rate": 5e-05,
+      "loss": 2.565,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4408295045324724,
+      "grad_norm": 0.34015870094299316,
+      "learning_rate": 5e-05,
+      "loss": 2.0059,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4409536818576928,
+      "grad_norm": 0.23272156715393066,
+      "learning_rate": 5e-05,
+      "loss": 1.8236,
+      "step": 3551
+    },
+    {
+      "epoch": 0.4410778591829132,
+      "grad_norm": 0.2400742471218109,
+      "learning_rate": 5e-05,
+      "loss": 1.8719,
+      "step": 3552
+    },
+    {
+      "epoch": 0.4412020365081336,
+      "grad_norm": 0.25298693776130676,
+      "learning_rate": 5e-05,
+      "loss": 2.2296,
+      "step": 3553
+    },
+    {
+      "epoch": 0.441326213833354,
+      "grad_norm": 0.25967708230018616,
+      "learning_rate": 5e-05,
+      "loss": 1.9136,
+      "step": 3554
+    },
+    {
+      "epoch": 0.44145039115857443,
+      "grad_norm": 0.42393115162849426,
+      "learning_rate": 5e-05,
+      "loss": 2.341,
+      "step": 3555
+    },
+    {
+      "epoch": 0.44157456848379484,
+      "grad_norm": 0.2446475327014923,
+      "learning_rate": 5e-05,
+      "loss": 1.7585,
+      "step": 3556
+    },
+    {
+      "epoch": 0.44169874580901525,
+      "grad_norm": 0.2373843789100647,
+      "learning_rate": 5e-05,
+      "loss": 1.6981,
+      "step": 3557
+    },
+    {
+      "epoch": 0.4418229231342357,
+      "grad_norm": 0.2612886130809784,
+      "learning_rate": 5e-05,
+      "loss": 2.4158,
+      "step": 3558
+    },
+    {
+      "epoch": 0.4419471004594561,
+      "grad_norm": 0.2711560130119324,
+      "learning_rate": 5e-05,
+      "loss": 2.4973,
+      "step": 3559
+    },
+    {
+      "epoch": 0.44207127778467653,
+      "grad_norm": 0.23028486967086792,
+      "learning_rate": 5e-05,
+      "loss": 1.7704,
+      "step": 3560
+    },
+    {
+      "epoch": 0.44219545510989694,
+      "grad_norm": 0.28811317682266235,
+      "learning_rate": 5e-05,
+      "loss": 2.1237,
+      "step": 3561
+    },
+    {
+      "epoch": 0.44231963243511735,
+      "grad_norm": 0.27257639169692993,
+      "learning_rate": 5e-05,
+      "loss": 2.1849,
+      "step": 3562
+    },
+    {
+      "epoch": 0.44244380976033776,
+      "grad_norm": 0.2721487879753113,
+      "learning_rate": 5e-05,
+      "loss": 1.9617,
+      "step": 3563
+    },
+    {
+      "epoch": 0.44256798708555817,
+      "grad_norm": 0.23857803642749786,
+      "learning_rate": 5e-05,
+      "loss": 1.8732,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4426921644107786,
+      "grad_norm": 0.26304367184638977,
+      "learning_rate": 5e-05,
+      "loss": 2.0674,
+      "step": 3565
+    },
+    {
+      "epoch": 0.442816341735999,
+      "grad_norm": 0.23700681328773499,
+      "learning_rate": 5e-05,
+      "loss": 1.5074,
+      "step": 3566
+    },
+    {
+      "epoch": 0.4429405190612194,
+      "grad_norm": 0.26313093304634094,
+      "learning_rate": 5e-05,
+      "loss": 2.1541,
+      "step": 3567
+    },
+    {
+      "epoch": 0.44306469638643986,
+      "grad_norm": 0.2672451436519623,
+      "learning_rate": 5e-05,
+      "loss": 2.2609,
+      "step": 3568
+    },
+    {
+      "epoch": 0.44318887371166027,
+      "grad_norm": 0.2410428673028946,
+      "learning_rate": 5e-05,
+      "loss": 1.7374,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4433130510368807,
+      "grad_norm": 0.24282702803611755,
+      "learning_rate": 5e-05,
+      "loss": 1.8673,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4434372283621011,
+      "grad_norm": 0.26424652338027954,
+      "learning_rate": 5e-05,
+      "loss": 1.8558,
+      "step": 3571
+    },
+    {
+      "epoch": 0.4435614056873215,
+      "grad_norm": 0.2430584728717804,
+      "learning_rate": 5e-05,
+      "loss": 2.0269,
+      "step": 3572
+    },
+    {
+      "epoch": 0.4436855830125419,
+      "grad_norm": 0.24861913919448853,
+      "learning_rate": 5e-05,
+      "loss": 2.0461,
+      "step": 3573
+    },
+    {
+      "epoch": 0.4438097603377623,
+      "grad_norm": 0.24841603636741638,
+      "learning_rate": 5e-05,
+      "loss": 1.8209,
+      "step": 3574
+    },
+    {
+      "epoch": 0.4439339376629827,
+      "grad_norm": 0.2474074512720108,
+      "learning_rate": 5e-05,
+      "loss": 1.8925,
+      "step": 3575
+    },
+    {
+      "epoch": 0.44405811498820313,
+      "grad_norm": 0.23635539412498474,
+      "learning_rate": 5e-05,
+      "loss": 1.7828,
+      "step": 3576
+    },
+    {
+      "epoch": 0.44418229231342354,
+      "grad_norm": 0.2570820748806,
+      "learning_rate": 5e-05,
+      "loss": 2.4142,
+      "step": 3577
+    },
+    {
+      "epoch": 0.444306469638644,
+      "grad_norm": 0.35572877526283264,
+      "learning_rate": 5e-05,
+      "loss": 2.547,
+      "step": 3578
+    },
+    {
+      "epoch": 0.4444306469638644,
+      "grad_norm": 0.2536236345767975,
+      "learning_rate": 5e-05,
+      "loss": 1.7313,
+      "step": 3579
+    },
+    {
+      "epoch": 0.4445548242890848,
+      "grad_norm": 0.2337023764848709,
+      "learning_rate": 5e-05,
+      "loss": 1.7236,
+      "step": 3580
+    },
+    {
+      "epoch": 0.44467900161430524,
+      "grad_norm": 0.2915627062320709,
+      "learning_rate": 5e-05,
+      "loss": 2.1736,
+      "step": 3581
+    },
+    {
+      "epoch": 0.44480317893952565,
+      "grad_norm": 0.24729955196380615,
+      "learning_rate": 5e-05,
+      "loss": 1.8573,
+      "step": 3582
+    },
+    {
+      "epoch": 0.44492735626474605,
+      "grad_norm": 0.2714468836784363,
+      "learning_rate": 5e-05,
+      "loss": 2.5826,
+      "step": 3583
+    },
+    {
+      "epoch": 0.44505153358996646,
+      "grad_norm": 0.24210494756698608,
+      "learning_rate": 5e-05,
+      "loss": 1.634,
+      "step": 3584
+    },
+    {
+      "epoch": 0.4451757109151869,
+      "grad_norm": 0.28783100843429565,
+      "learning_rate": 5e-05,
+      "loss": 2.8731,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4452998882404073,
+      "grad_norm": 0.257840096950531,
+      "learning_rate": 5e-05,
+      "loss": 2.0271,
+      "step": 3586
+    },
+    {
+      "epoch": 0.4454240655656277,
+      "grad_norm": 0.27797460556030273,
+      "learning_rate": 5e-05,
+      "loss": 2.3401,
+      "step": 3587
+    },
+    {
+      "epoch": 0.44554824289084816,
+      "grad_norm": 0.26590582728385925,
+      "learning_rate": 5e-05,
+      "loss": 2.5732,
+      "step": 3588
+    },
+    {
+      "epoch": 0.44567242021606857,
+      "grad_norm": 0.2545289993286133,
+      "learning_rate": 5e-05,
+      "loss": 2.055,
+      "step": 3589
+    },
+    {
+      "epoch": 0.445796597541289,
+      "grad_norm": 0.27943652868270874,
+      "learning_rate": 5e-05,
+      "loss": 2.6795,
+      "step": 3590
+    },
+    {
+      "epoch": 0.4459207748665094,
+      "grad_norm": 0.2627931833267212,
+      "learning_rate": 5e-05,
+      "loss": 2.0831,
+      "step": 3591
+    },
+    {
+      "epoch": 0.4460449521917298,
+      "grad_norm": 0.2529352903366089,
+      "learning_rate": 5e-05,
+      "loss": 1.7692,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4461691295169502,
+      "grad_norm": 0.2594928741455078,
+      "learning_rate": 5e-05,
+      "loss": 1.9044,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4462933068421706,
+      "grad_norm": 0.249286487698555,
+      "learning_rate": 5e-05,
+      "loss": 1.7774,
+      "step": 3594
+    },
+    {
+      "epoch": 0.446417484167391,
+      "grad_norm": 0.36518681049346924,
+      "learning_rate": 5e-05,
+      "loss": 3.0051,
+      "step": 3595
+    },
+    {
+      "epoch": 0.44654166149261143,
+      "grad_norm": 0.2684462368488312,
+      "learning_rate": 5e-05,
+      "loss": 2.1218,
+      "step": 3596
+    },
+    {
+      "epoch": 0.44666583881783184,
+      "grad_norm": 0.27154868841171265,
+      "learning_rate": 5e-05,
+      "loss": 2.225,
+      "step": 3597
+    },
+    {
+      "epoch": 0.4467900161430523,
+      "grad_norm": 0.24538029730319977,
+      "learning_rate": 5e-05,
+      "loss": 1.7131,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4469141934682727,
+      "grad_norm": 0.24630869925022125,
+      "learning_rate": 5e-05,
+      "loss": 1.709,
+      "step": 3599
+    },
+    {
+      "epoch": 0.4470383707934931,
+      "grad_norm": 0.2505440413951874,
+      "learning_rate": 5e-05,
+      "loss": 1.8827,
+      "step": 3600
+    },
+    {
+      "epoch": 0.44716254811871353,
+      "grad_norm": 0.2851535379886627,
+      "learning_rate": 5e-05,
+      "loss": 2.1343,
+      "step": 3601
+    },
+    {
+      "epoch": 0.44728672544393394,
+      "grad_norm": 0.26223599910736084,
+      "learning_rate": 5e-05,
+      "loss": 1.8654,
+      "step": 3602
+    },
+    {
+      "epoch": 0.44741090276915435,
+      "grad_norm": 0.24509179592132568,
+      "learning_rate": 5e-05,
+      "loss": 1.9708,
+      "step": 3603
+    },
+    {
+      "epoch": 0.44753508009437476,
+      "grad_norm": 0.30279234051704407,
+      "learning_rate": 5e-05,
+      "loss": 2.2681,
+      "step": 3604
+    },
+    {
+      "epoch": 0.44765925741959517,
+      "grad_norm": 0.257269948720932,
+      "learning_rate": 5e-05,
+      "loss": 1.9764,
+      "step": 3605
+    },
+    {
+      "epoch": 0.4477834347448156,
+      "grad_norm": 0.261173278093338,
+      "learning_rate": 5e-05,
+      "loss": 2.3032,
+      "step": 3606
+    },
+    {
+      "epoch": 0.447907612070036,
+      "grad_norm": 0.3636569678783417,
+      "learning_rate": 5e-05,
+      "loss": 2.8534,
+      "step": 3607
+    },
+    {
+      "epoch": 0.44803178939525645,
+      "grad_norm": 0.22757330536842346,
+      "learning_rate": 5e-05,
+      "loss": 1.7358,
+      "step": 3608
+    },
+    {
+      "epoch": 0.44815596672047686,
+      "grad_norm": 0.23848308622837067,
+      "learning_rate": 5e-05,
+      "loss": 1.7782,
+      "step": 3609
+    },
+    {
+      "epoch": 0.44828014404569727,
+      "grad_norm": 0.25233447551727295,
+      "learning_rate": 5e-05,
+      "loss": 1.7118,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4484043213709177,
+      "grad_norm": 0.270443856716156,
+      "learning_rate": 5e-05,
+      "loss": 2.2098,
+      "step": 3611
+    },
+    {
+      "epoch": 0.4485284986961381,
+      "grad_norm": 0.33944758772850037,
+      "learning_rate": 5e-05,
+      "loss": 2.7892,
+      "step": 3612
+    },
+    {
+      "epoch": 0.4486526760213585,
+      "grad_norm": 0.3116800785064697,
+      "learning_rate": 5e-05,
+      "loss": 2.8959,
+      "step": 3613
+    },
+    {
+      "epoch": 0.4487768533465789,
+      "grad_norm": 0.2481691986322403,
+      "learning_rate": 5e-05,
+      "loss": 1.9309,
+      "step": 3614
+    },
+    {
+      "epoch": 0.4489010306717993,
+      "grad_norm": 0.25150108337402344,
+      "learning_rate": 5e-05,
+      "loss": 1.9542,
+      "step": 3615
+    },
+    {
+      "epoch": 0.4490252079970197,
+      "grad_norm": 0.2553020715713501,
+      "learning_rate": 5e-05,
+      "loss": 1.8089,
+      "step": 3616
+    },
+    {
+      "epoch": 0.44914938532224014,
+      "grad_norm": 0.2526760697364807,
+      "learning_rate": 5e-05,
+      "loss": 2.1786,
+      "step": 3617
+    },
+    {
+      "epoch": 0.4492735626474606,
+      "grad_norm": 0.29077330231666565,
+      "learning_rate": 5e-05,
+      "loss": 2.1449,
+      "step": 3618
+    },
+    {
+      "epoch": 0.449397739972681,
+      "grad_norm": 0.2767769992351532,
+      "learning_rate": 5e-05,
+      "loss": 2.0351,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4495219172979014,
+      "grad_norm": 0.2682516276836395,
+      "learning_rate": 5e-05,
+      "loss": 1.7652,
+      "step": 3620
+    },
+    {
+      "epoch": 0.44964609462312183,
+      "grad_norm": 0.2579279839992523,
+      "learning_rate": 5e-05,
+      "loss": 2.1802,
+      "step": 3621
+    },
+    {
+      "epoch": 0.44977027194834224,
+      "grad_norm": 0.28205546736717224,
+      "learning_rate": 5e-05,
+      "loss": 2.1566,
+      "step": 3622
+    },
+    {
+      "epoch": 0.44989444927356265,
+      "grad_norm": 0.2978857457637787,
+      "learning_rate": 5e-05,
+      "loss": 2.1345,
+      "step": 3623
+    },
+    {
+      "epoch": 0.45001862659878306,
+      "grad_norm": 0.23637983202934265,
+      "learning_rate": 5e-05,
+      "loss": 1.8078,
+      "step": 3624
+    },
+    {
+      "epoch": 0.45014280392400347,
+      "grad_norm": 0.2526599168777466,
+      "learning_rate": 5e-05,
+      "loss": 1.7115,
+      "step": 3625
+    },
+    {
+      "epoch": 0.4502669812492239,
+      "grad_norm": 0.46913304924964905,
+      "learning_rate": 5e-05,
+      "loss": 2.0266,
+      "step": 3626
+    },
+    {
+      "epoch": 0.4503911585744443,
+      "grad_norm": 0.24125505983829498,
+      "learning_rate": 5e-05,
+      "loss": 1.8029,
+      "step": 3627
+    },
+    {
+      "epoch": 0.45051533589966475,
+      "grad_norm": 0.36426547169685364,
+      "learning_rate": 5e-05,
+      "loss": 3.1303,
+      "step": 3628
+    },
+    {
+      "epoch": 0.45063951322488516,
+      "grad_norm": 0.2742426097393036,
+      "learning_rate": 5e-05,
+      "loss": 2.3707,
+      "step": 3629
+    },
+    {
+      "epoch": 0.45076369055010557,
+      "grad_norm": 0.28186655044555664,
+      "learning_rate": 5e-05,
+      "loss": 2.044,
+      "step": 3630
+    },
+    {
+      "epoch": 0.450887867875326,
+      "grad_norm": 0.2517007291316986,
+      "learning_rate": 5e-05,
+      "loss": 2.1004,
+      "step": 3631
+    },
+    {
+      "epoch": 0.4510120452005464,
+      "grad_norm": 0.2443215698003769,
+      "learning_rate": 5e-05,
+      "loss": 1.9075,
+      "step": 3632
+    },
+    {
+      "epoch": 0.4511362225257668,
+      "grad_norm": 0.30474352836608887,
+      "learning_rate": 5e-05,
+      "loss": 2.6504,
+      "step": 3633
+    },
+    {
+      "epoch": 0.4512603998509872,
+      "grad_norm": 0.24801456928253174,
+      "learning_rate": 5e-05,
+      "loss": 1.9028,
+      "step": 3634
+    },
+    {
+      "epoch": 0.4513845771762076,
+      "grad_norm": 0.24547287821769714,
+      "learning_rate": 5e-05,
+      "loss": 1.7863,
+      "step": 3635
+    },
+    {
+      "epoch": 0.451508754501428,
+      "grad_norm": 0.2354976385831833,
+      "learning_rate": 5e-05,
+      "loss": 1.8193,
+      "step": 3636
+    },
+    {
+      "epoch": 0.45163293182664843,
+      "grad_norm": 0.22470952570438385,
+      "learning_rate": 5e-05,
+      "loss": 1.4583,
+      "step": 3637
+    },
+    {
+      "epoch": 0.4517571091518689,
+      "grad_norm": 0.24142299592494965,
+      "learning_rate": 5e-05,
+      "loss": 1.6473,
+      "step": 3638
+    },
+    {
+      "epoch": 0.4518812864770893,
+      "grad_norm": 0.2592025399208069,
+      "learning_rate": 5e-05,
+      "loss": 1.8388,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4520054638023097,
+      "grad_norm": 0.23507888615131378,
+      "learning_rate": 5e-05,
+      "loss": 1.4999,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4521296411275301,
+      "grad_norm": 0.44732698798179626,
+      "learning_rate": 5e-05,
+      "loss": 2.5855,
+      "step": 3641
+    },
+    {
+      "epoch": 0.45225381845275053,
+      "grad_norm": 0.2762181758880615,
+      "learning_rate": 5e-05,
+      "loss": 2.441,
+      "step": 3642
+    },
+    {
+      "epoch": 0.45237799577797094,
+      "grad_norm": 0.2453155517578125,
+      "learning_rate": 5e-05,
+      "loss": 1.8225,
+      "step": 3643
+    },
+    {
+      "epoch": 0.45250217310319135,
+      "grad_norm": 0.2706897556781769,
+      "learning_rate": 5e-05,
+      "loss": 2.0773,
+      "step": 3644
+    },
+    {
+      "epoch": 0.45262635042841176,
+      "grad_norm": 0.24604804813861847,
+      "learning_rate": 5e-05,
+      "loss": 1.7866,
+      "step": 3645
+    },
+    {
+      "epoch": 0.45275052775363217,
+      "grad_norm": 0.2511715292930603,
+      "learning_rate": 5e-05,
+      "loss": 1.6364,
+      "step": 3646
+    },
+    {
+      "epoch": 0.4528747050788526,
+      "grad_norm": 0.2393648475408554,
+      "learning_rate": 5e-05,
+      "loss": 1.8031,
+      "step": 3647
+    },
+    {
+      "epoch": 0.452998882404073,
+      "grad_norm": 0.32893240451812744,
+      "learning_rate": 5e-05,
+      "loss": 2.9534,
+      "step": 3648
+    },
+    {
+      "epoch": 0.45312305972929345,
+      "grad_norm": 0.2726787328720093,
+      "learning_rate": 5e-05,
+      "loss": 2.158,
+      "step": 3649
+    },
+    {
+      "epoch": 0.45324723705451386,
+      "grad_norm": 0.24405358731746674,
+      "learning_rate": 5e-05,
+      "loss": 1.8065,
+      "step": 3650
+    },
+    {
+      "epoch": 0.45337141437973427,
+      "grad_norm": 0.289327472448349,
+      "learning_rate": 5e-05,
+      "loss": 1.9226,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4534955917049547,
+      "grad_norm": 0.2605152726173401,
+      "learning_rate": 5e-05,
+      "loss": 1.9381,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4536197690301751,
+      "grad_norm": 0.5991039872169495,
+      "learning_rate": 5e-05,
+      "loss": 2.9309,
+      "step": 3653
+    },
+    {
+      "epoch": 0.4537439463553955,
+      "grad_norm": 0.2568708658218384,
+      "learning_rate": 5e-05,
+      "loss": 2.0779,
+      "step": 3654
+    },
+    {
+      "epoch": 0.4538681236806159,
+      "grad_norm": 0.3010365068912506,
+      "learning_rate": 5e-05,
+      "loss": 2.0548,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4539923010058363,
+      "grad_norm": 0.274974524974823,
+      "learning_rate": 5e-05,
+      "loss": 2.4077,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4541164783310567,
+      "grad_norm": 0.28107506036758423,
+      "learning_rate": 5e-05,
+      "loss": 2.2441,
+      "step": 3657
+    },
+    {
+      "epoch": 0.45424065565627714,
+      "grad_norm": 0.23606228828430176,
+      "learning_rate": 5e-05,
+      "loss": 1.8557,
+      "step": 3658
+    },
+    {
+      "epoch": 0.4543648329814976,
+      "grad_norm": 0.39235198497772217,
+      "learning_rate": 5e-05,
+      "loss": 2.724,
+      "step": 3659
+    },
+    {
+      "epoch": 0.454489010306718,
+      "grad_norm": 0.24546925723552704,
+      "learning_rate": 5e-05,
+      "loss": 2.1356,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4546131876319384,
+      "grad_norm": 0.2840039134025574,
+      "learning_rate": 5e-05,
+      "loss": 2.5487,
+      "step": 3661
+    },
+    {
+      "epoch": 0.45473736495715883,
+      "grad_norm": 0.2581782042980194,
+      "learning_rate": 5e-05,
+      "loss": 2.0953,
+      "step": 3662
+    },
+    {
+      "epoch": 0.45486154228237924,
+      "grad_norm": 0.28797447681427,
+      "learning_rate": 5e-05,
+      "loss": 2.7819,
+      "step": 3663
+    },
+    {
+      "epoch": 0.45498571960759965,
+      "grad_norm": 0.30103740096092224,
+      "learning_rate": 5e-05,
+      "loss": 2.3426,
+      "step": 3664
+    },
+    {
+      "epoch": 0.45510989693282006,
+      "grad_norm": 0.26766613125801086,
+      "learning_rate": 5e-05,
+      "loss": 2.08,
+      "step": 3665
+    },
+    {
+      "epoch": 0.45523407425804047,
+      "grad_norm": 0.271410197019577,
+      "learning_rate": 5e-05,
+      "loss": 1.9295,
+      "step": 3666
+    },
+    {
+      "epoch": 0.4553582515832609,
+      "grad_norm": 0.28217509388923645,
+      "learning_rate": 5e-05,
+      "loss": 1.7145,
+      "step": 3667
+    },
+    {
+      "epoch": 0.4554824289084813,
+      "grad_norm": 0.2470954954624176,
+      "learning_rate": 5e-05,
+      "loss": 1.9962,
+      "step": 3668
+    },
+    {
+      "epoch": 0.45560660623370175,
+      "grad_norm": 0.2428179234266281,
+      "learning_rate": 5e-05,
+      "loss": 1.8761,
+      "step": 3669
+    },
+    {
+      "epoch": 0.45573078355892216,
+      "grad_norm": 0.2993158996105194,
+      "learning_rate": 5e-05,
+      "loss": 2.0549,
+      "step": 3670
+    },
+    {
+      "epoch": 0.45585496088414257,
+      "grad_norm": 0.25695863366127014,
+      "learning_rate": 5e-05,
+      "loss": 1.9042,
+      "step": 3671
+    },
+    {
+      "epoch": 0.455979138209363,
+      "grad_norm": 0.255073606967926,
+      "learning_rate": 5e-05,
+      "loss": 1.9973,
+      "step": 3672
+    },
+    {
+      "epoch": 0.4561033155345834,
+      "grad_norm": 0.2622082531452179,
+      "learning_rate": 5e-05,
+      "loss": 2.1832,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4562274928598038,
+      "grad_norm": 0.2978021800518036,
+      "learning_rate": 5e-05,
+      "loss": 2.3905,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4563516701850242,
+      "grad_norm": 0.23439711332321167,
+      "learning_rate": 5e-05,
+      "loss": 1.6926,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4564758475102446,
+      "grad_norm": 0.25187382102012634,
+      "learning_rate": 5e-05,
+      "loss": 1.9492,
+      "step": 3676
+    },
+    {
+      "epoch": 0.456600024835465,
+      "grad_norm": 0.2353617548942566,
+      "learning_rate": 5e-05,
+      "loss": 1.9207,
+      "step": 3677
+    },
+    {
+      "epoch": 0.45672420216068543,
+      "grad_norm": 0.2954971194267273,
+      "learning_rate": 5e-05,
+      "loss": 2.7884,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4568483794859059,
+      "grad_norm": 0.342014342546463,
+      "learning_rate": 5e-05,
+      "loss": 2.7057,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4569725568111263,
+      "grad_norm": 0.2446756511926651,
+      "learning_rate": 5e-05,
+      "loss": 1.7201,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4570967341363467,
+      "grad_norm": 0.3007001280784607,
+      "learning_rate": 5e-05,
+      "loss": 2.8884,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4572209114615671,
+      "grad_norm": 0.3524860739707947,
+      "learning_rate": 5e-05,
+      "loss": 2.4902,
+      "step": 3682
+    },
+    {
+      "epoch": 0.45734508878678753,
+      "grad_norm": 0.2736489474773407,
+      "learning_rate": 5e-05,
+      "loss": 1.9327,
+      "step": 3683
+    },
+    {
+      "epoch": 0.45746926611200794,
+      "grad_norm": 0.24661853909492493,
+      "learning_rate": 5e-05,
+      "loss": 1.9474,
+      "step": 3684
+    },
+    {
+      "epoch": 0.45759344343722835,
+      "grad_norm": 0.2539280652999878,
+      "learning_rate": 5e-05,
+      "loss": 1.7299,
+      "step": 3685
+    },
+    {
+      "epoch": 0.45771762076244876,
+      "grad_norm": 0.3460065424442291,
+      "learning_rate": 5e-05,
+      "loss": 2.5973,
+      "step": 3686
+    },
+    {
+      "epoch": 0.45784179808766917,
+      "grad_norm": 0.24186161160469055,
+      "learning_rate": 5e-05,
+      "loss": 2.0177,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4579659754128896,
+      "grad_norm": 0.2821299135684967,
+      "learning_rate": 5e-05,
+      "loss": 1.7887,
+      "step": 3688
+    },
+    {
+      "epoch": 0.45809015273811005,
+      "grad_norm": 0.2463959902524948,
+      "learning_rate": 5e-05,
+      "loss": 1.8642,
+      "step": 3689
+    },
+    {
+      "epoch": 0.45821433006333046,
+      "grad_norm": 0.2415582686662674,
+      "learning_rate": 5e-05,
+      "loss": 1.7125,
+      "step": 3690
+    },
+    {
+      "epoch": 0.45833850738855086,
+      "grad_norm": 0.24925467371940613,
+      "learning_rate": 5e-05,
+      "loss": 2.0802,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4584626847137713,
+      "grad_norm": 0.284193754196167,
+      "learning_rate": 5e-05,
+      "loss": 2.1535,
+      "step": 3692
+    },
+    {
+      "epoch": 0.4585868620389917,
+      "grad_norm": 0.5420135855674744,
+      "learning_rate": 5e-05,
+      "loss": 3.2934,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4587110393642121,
+      "grad_norm": 0.2871832549571991,
+      "learning_rate": 5e-05,
+      "loss": 2.0187,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4588352166894325,
+      "grad_norm": 0.28680795431137085,
+      "learning_rate": 5e-05,
+      "loss": 2.315,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4589593940146529,
+      "grad_norm": 0.24108101427555084,
+      "learning_rate": 5e-05,
+      "loss": 1.6863,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4590835713398733,
+      "grad_norm": 0.22612926363945007,
+      "learning_rate": 5e-05,
+      "loss": 1.7276,
+      "step": 3697
+    },
+    {
+      "epoch": 0.45920774866509373,
+      "grad_norm": 0.39188769459724426,
+      "learning_rate": 5e-05,
+      "loss": 1.5337,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4593319259903142,
+      "grad_norm": 0.2721399664878845,
+      "learning_rate": 5e-05,
+      "loss": 2.2841,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4594561033155346,
+      "grad_norm": 0.2642127573490143,
+      "learning_rate": 5e-05,
+      "loss": 2.0429,
+      "step": 3700
+    },
+    {
+      "epoch": 0.459580280640755,
+      "grad_norm": 0.34524235129356384,
+      "learning_rate": 5e-05,
+      "loss": 2.5238,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4597044579659754,
+      "grad_norm": 0.23760610818862915,
+      "learning_rate": 5e-05,
+      "loss": 1.9163,
+      "step": 3702
+    },
+    {
+      "epoch": 0.45982863529119583,
+      "grad_norm": 0.25335967540740967,
+      "learning_rate": 5e-05,
+      "loss": 1.7642,
+      "step": 3703
+    },
+    {
+      "epoch": 0.45995281261641624,
+      "grad_norm": 0.2770371437072754,
+      "learning_rate": 5e-05,
+      "loss": 2.0273,
+      "step": 3704
+    },
+    {
+      "epoch": 0.46007698994163665,
+      "grad_norm": 0.25973621010780334,
+      "learning_rate": 5e-05,
+      "loss": 1.823,
+      "step": 3705
+    },
+    {
+      "epoch": 0.46020116726685706,
+      "grad_norm": 0.35052424669265747,
+      "learning_rate": 5e-05,
+      "loss": 3.1005,
+      "step": 3706
+    },
+    {
+      "epoch": 0.46032534459207747,
+      "grad_norm": 0.25554993748664856,
+      "learning_rate": 5e-05,
+      "loss": 2.0084,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4604495219172979,
+      "grad_norm": 0.28023436665534973,
+      "learning_rate": 5e-05,
+      "loss": 2.4759,
+      "step": 3708
+    },
+    {
+      "epoch": 0.46057369924251834,
+      "grad_norm": 0.3017418682575226,
+      "learning_rate": 5e-05,
+      "loss": 2.0529,
+      "step": 3709
+    },
+    {
+      "epoch": 0.46069787656773875,
+      "grad_norm": 0.2600777745246887,
+      "learning_rate": 5e-05,
+      "loss": 1.7135,
+      "step": 3710
+    },
+    {
+      "epoch": 0.46082205389295916,
+      "grad_norm": 0.27555832266807556,
+      "learning_rate": 5e-05,
+      "loss": 1.8721,
+      "step": 3711
+    },
+    {
+      "epoch": 0.46094623121817957,
+      "grad_norm": 0.35168084502220154,
+      "learning_rate": 5e-05,
+      "loss": 2.5732,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4610704085434,
+      "grad_norm": 0.2278173416852951,
+      "learning_rate": 5e-05,
+      "loss": 1.6557,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4611945858686204,
+      "grad_norm": 0.25641727447509766,
+      "learning_rate": 5e-05,
+      "loss": 1.8153,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4613187631938408,
+      "grad_norm": 0.2860358655452728,
+      "learning_rate": 5e-05,
+      "loss": 2.7876,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4614429405190612,
+      "grad_norm": 0.2561151385307312,
+      "learning_rate": 5e-05,
+      "loss": 1.5212,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4615671178442816,
+      "grad_norm": 0.2888529300689697,
+      "learning_rate": 5e-05,
+      "loss": 2.2057,
+      "step": 3717
+    },
+    {
+      "epoch": 0.461691295169502,
+      "grad_norm": 0.266047865152359,
+      "learning_rate": 5e-05,
+      "loss": 1.9522,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4618154724947225,
+      "grad_norm": 0.24097436666488647,
+      "learning_rate": 5e-05,
+      "loss": 1.8926,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4619396498199429,
+      "grad_norm": 0.28417128324508667,
+      "learning_rate": 5e-05,
+      "loss": 2.3777,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4620638271451633,
+      "grad_norm": 0.2749319076538086,
+      "learning_rate": 5e-05,
+      "loss": 2.5085,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4621880044703837,
+      "grad_norm": 0.38441285490989685,
+      "learning_rate": 5e-05,
+      "loss": 2.2511,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4623121817956041,
+      "grad_norm": 0.3329639434814453,
+      "learning_rate": 5e-05,
+      "loss": 2.1478,
+      "step": 3723
+    },
+    {
+      "epoch": 0.46243635912082454,
+      "grad_norm": 0.45397889614105225,
+      "learning_rate": 5e-05,
+      "loss": 3.0332,
+      "step": 3724
+    },
+    {
+      "epoch": 0.46256053644604495,
+      "grad_norm": 0.2769295871257782,
+      "learning_rate": 5e-05,
+      "loss": 2.6312,
+      "step": 3725
+    },
+    {
+      "epoch": 0.46268471377126535,
+      "grad_norm": 0.2944956421852112,
+      "learning_rate": 5e-05,
+      "loss": 2.8046,
+      "step": 3726
+    },
+    {
+      "epoch": 0.46280889109648576,
+      "grad_norm": 0.2855566740036011,
+      "learning_rate": 5e-05,
+      "loss": 2.7002,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4629330684217062,
+      "grad_norm": 0.29095032811164856,
+      "learning_rate": 5e-05,
+      "loss": 2.0279,
+      "step": 3728
+    },
+    {
+      "epoch": 0.46305724574692664,
+      "grad_norm": 0.29610753059387207,
+      "learning_rate": 5e-05,
+      "loss": 2.0261,
+      "step": 3729
+    },
+    {
+      "epoch": 0.46318142307214705,
+      "grad_norm": 0.2529889941215515,
+      "learning_rate": 5e-05,
+      "loss": 2.4328,
+      "step": 3730
+    },
+    {
+      "epoch": 0.46330560039736746,
+      "grad_norm": 0.2528243958950043,
+      "learning_rate": 5e-05,
+      "loss": 1.705,
+      "step": 3731
+    },
+    {
+      "epoch": 0.46342977772258787,
+      "grad_norm": 0.3699294924736023,
+      "learning_rate": 5e-05,
+      "loss": 1.9492,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4635539550478083,
+      "grad_norm": 0.23367568850517273,
+      "learning_rate": 5e-05,
+      "loss": 1.8059,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4636781323730287,
+      "grad_norm": 0.22545285522937775,
+      "learning_rate": 5e-05,
+      "loss": 1.5707,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4638023096982491,
+      "grad_norm": 0.3032660186290741,
+      "learning_rate": 5e-05,
+      "loss": 2.5091,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4639264870234695,
+      "grad_norm": 0.3749956786632538,
+      "learning_rate": 5e-05,
+      "loss": 2.9199,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4640506643486899,
+      "grad_norm": 0.26235827803611755,
+      "learning_rate": 5e-05,
+      "loss": 1.8312,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4641748416739103,
+      "grad_norm": 0.2523275315761566,
+      "learning_rate": 5e-05,
+      "loss": 1.8933,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4642990189991308,
+      "grad_norm": 0.24115239083766937,
+      "learning_rate": 5e-05,
+      "loss": 1.8593,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4644231963243512,
+      "grad_norm": 0.24133779108524323,
+      "learning_rate": 5e-05,
+      "loss": 1.9513,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4645473736495716,
+      "grad_norm": 0.25049519538879395,
+      "learning_rate": 5e-05,
+      "loss": 1.9626,
+      "step": 3741
+    },
+    {
+      "epoch": 0.464671550974792,
+      "grad_norm": 0.32509729266166687,
+      "learning_rate": 5e-05,
+      "loss": 2.6754,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4647957283000124,
+      "grad_norm": 0.22493858635425568,
+      "learning_rate": 5e-05,
+      "loss": 1.3969,
+      "step": 3743
+    },
+    {
+      "epoch": 0.46491990562523283,
+      "grad_norm": 0.32038432359695435,
+      "learning_rate": 5e-05,
+      "loss": 2.8724,
+      "step": 3744
+    },
+    {
+      "epoch": 0.46504408295045324,
+      "grad_norm": 0.2756808400154114,
+      "learning_rate": 5e-05,
+      "loss": 1.8665,
+      "step": 3745
+    },
+    {
+      "epoch": 0.46516826027567365,
+      "grad_norm": 0.22919133305549622,
+      "learning_rate": 5e-05,
+      "loss": 1.622,
+      "step": 3746
+    },
+    {
+      "epoch": 0.46529243760089406,
+      "grad_norm": 0.2500326335430145,
+      "learning_rate": 5e-05,
+      "loss": 1.7911,
+      "step": 3747
+    },
+    {
+      "epoch": 0.46541661492611447,
+      "grad_norm": 0.2787139415740967,
+      "learning_rate": 5e-05,
+      "loss": 2.3524,
+      "step": 3748
+    },
+    {
+      "epoch": 0.46554079225133493,
+      "grad_norm": 0.3825012445449829,
+      "learning_rate": 5e-05,
+      "loss": 2.6704,
+      "step": 3749
+    },
+    {
+      "epoch": 0.46566496957655534,
+      "grad_norm": 0.23502719402313232,
+      "learning_rate": 5e-05,
+      "loss": 1.6898,
+      "step": 3750
+    },
+    {
+      "epoch": 0.46578914690177575,
+      "grad_norm": 0.29871129989624023,
+      "learning_rate": 5e-05,
+      "loss": 2.3541,
+      "step": 3751
+    },
+    {
+      "epoch": 0.46591332422699616,
+      "grad_norm": 0.24409402906894684,
+      "learning_rate": 5e-05,
+      "loss": 1.6336,
+      "step": 3752
+    },
+    {
+      "epoch": 0.46603750155221657,
+      "grad_norm": 0.31112971901893616,
+      "learning_rate": 5e-05,
+      "loss": 2.3301,
+      "step": 3753
+    },
+    {
+      "epoch": 0.466161678877437,
+      "grad_norm": 0.46290358901023865,
+      "learning_rate": 5e-05,
+      "loss": 2.7657,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4662858562026574,
+      "grad_norm": 0.29431286454200745,
+      "learning_rate": 5e-05,
+      "loss": 2.1471,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4664100335278778,
+      "grad_norm": 0.26218414306640625,
+      "learning_rate": 5e-05,
+      "loss": 2.3939,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4665342108530982,
+      "grad_norm": 0.3483566641807556,
+      "learning_rate": 5e-05,
+      "loss": 2.2484,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4666583881783186,
+      "grad_norm": 0.24024175107479095,
+      "learning_rate": 5e-05,
+      "loss": 1.6138,
+      "step": 3758
+    },
+    {
+      "epoch": 0.466782565503539,
+      "grad_norm": 0.2772156298160553,
+      "learning_rate": 5e-05,
+      "loss": 1.9293,
+      "step": 3759
+    },
+    {
+      "epoch": 0.4669067428287595,
+      "grad_norm": 0.24673230946063995,
+      "learning_rate": 5e-05,
+      "loss": 1.6703,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4670309201539799,
+      "grad_norm": 0.24363486468791962,
+      "learning_rate": 5e-05,
+      "loss": 1.9547,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4671550974792003,
+      "grad_norm": 0.24472695589065552,
+      "learning_rate": 5e-05,
+      "loss": 1.8546,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4672792748044207,
+      "grad_norm": 0.2591949701309204,
+      "learning_rate": 5e-05,
+      "loss": 1.9849,
+      "step": 3763
+    },
+    {
+      "epoch": 0.46740345212964113,
+      "grad_norm": 0.3027430772781372,
+      "learning_rate": 5e-05,
+      "loss": 2.3229,
+      "step": 3764
+    },
+    {
+      "epoch": 0.46752762945486154,
+      "grad_norm": 0.2586989402770996,
+      "learning_rate": 5e-05,
+      "loss": 1.9121,
+      "step": 3765
+    },
+    {
+      "epoch": 0.46765180678008195,
+      "grad_norm": 0.2581965923309326,
+      "learning_rate": 5e-05,
+      "loss": 2.0683,
+      "step": 3766
+    },
+    {
+      "epoch": 0.46777598410530236,
+      "grad_norm": 0.32932889461517334,
+      "learning_rate": 5e-05,
+      "loss": 2.7389,
+      "step": 3767
+    },
+    {
+      "epoch": 0.46790016143052277,
+      "grad_norm": 0.23542195558547974,
+      "learning_rate": 5e-05,
+      "loss": 1.4886,
+      "step": 3768
+    },
+    {
+      "epoch": 0.4680243387557432,
+      "grad_norm": 0.2686495780944824,
+      "learning_rate": 5e-05,
+      "loss": 2.0493,
+      "step": 3769
+    },
+    {
+      "epoch": 0.46814851608096364,
+      "grad_norm": 0.2341269552707672,
+      "learning_rate": 5e-05,
+      "loss": 1.7391,
+      "step": 3770
+    },
+    {
+      "epoch": 0.46827269340618405,
+      "grad_norm": 0.24881036579608917,
+      "learning_rate": 5e-05,
+      "loss": 1.8926,
+      "step": 3771
+    },
+    {
+      "epoch": 0.46839687073140446,
+      "grad_norm": 0.31741994619369507,
+      "learning_rate": 5e-05,
+      "loss": 2.5013,
+      "step": 3772
+    },
+    {
+      "epoch": 0.46852104805662487,
+      "grad_norm": 0.23007549345493317,
+      "learning_rate": 5e-05,
+      "loss": 1.7307,
+      "step": 3773
+    },
+    {
+      "epoch": 0.4686452253818453,
+      "grad_norm": 0.2672252357006073,
+      "learning_rate": 5e-05,
+      "loss": 2.5262,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4687694027070657,
+      "grad_norm": 0.26503461599349976,
+      "learning_rate": 5e-05,
+      "loss": 1.9904,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4688935800322861,
+      "grad_norm": 0.24616654217243195,
+      "learning_rate": 5e-05,
+      "loss": 1.7568,
+      "step": 3776
+    },
+    {
+      "epoch": 0.4690177573575065,
+      "grad_norm": 0.24603234231472015,
+      "learning_rate": 5e-05,
+      "loss": 1.7584,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4691419346827269,
+      "grad_norm": 0.2972221076488495,
+      "learning_rate": 5e-05,
+      "loss": 3.0205,
+      "step": 3778
+    },
+    {
+      "epoch": 0.4692661120079473,
+      "grad_norm": 0.281521737575531,
+      "learning_rate": 5e-05,
+      "loss": 2.445,
+      "step": 3779
+    },
+    {
+      "epoch": 0.4693902893331678,
+      "grad_norm": 0.24010278284549713,
+      "learning_rate": 5e-05,
+      "loss": 1.7375,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4695144666583882,
+      "grad_norm": 0.2598876655101776,
+      "learning_rate": 5e-05,
+      "loss": 1.7829,
+      "step": 3781
+    },
+    {
+      "epoch": 0.4696386439836086,
+      "grad_norm": 0.32101139426231384,
+      "learning_rate": 5e-05,
+      "loss": 2.3779,
+      "step": 3782
+    },
+    {
+      "epoch": 0.469762821308829,
+      "grad_norm": 0.26489296555519104,
+      "learning_rate": 5e-05,
+      "loss": 1.9479,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4698869986340494,
+      "grad_norm": 0.24717245995998383,
+      "learning_rate": 5e-05,
+      "loss": 1.6296,
+      "step": 3784
+    },
+    {
+      "epoch": 0.47001117595926983,
+      "grad_norm": 0.24983420968055725,
+      "learning_rate": 5e-05,
+      "loss": 1.5642,
+      "step": 3785
+    },
+    {
+      "epoch": 0.47013535328449024,
+      "grad_norm": 0.25838613510131836,
+      "learning_rate": 5e-05,
+      "loss": 1.7181,
+      "step": 3786
+    },
+    {
+      "epoch": 0.47025953060971065,
+      "grad_norm": 0.307891845703125,
+      "learning_rate": 5e-05,
+      "loss": 2.2243,
+      "step": 3787
+    },
+    {
+      "epoch": 0.47038370793493106,
+      "grad_norm": 0.3163226246833801,
+      "learning_rate": 5e-05,
+      "loss": 2.7928,
+      "step": 3788
+    },
+    {
+      "epoch": 0.47050788526015147,
+      "grad_norm": 0.3786616027355194,
+      "learning_rate": 5e-05,
+      "loss": 1.9988,
+      "step": 3789
+    },
+    {
+      "epoch": 0.47063206258537194,
+      "grad_norm": 0.24822497367858887,
+      "learning_rate": 5e-05,
+      "loss": 1.8376,
+      "step": 3790
+    },
+    {
+      "epoch": 0.47075623991059234,
+      "grad_norm": 0.2994300425052643,
+      "learning_rate": 5e-05,
+      "loss": 2.0891,
+      "step": 3791
+    },
+    {
+      "epoch": 0.47088041723581275,
+      "grad_norm": 0.2565656304359436,
+      "learning_rate": 5e-05,
+      "loss": 1.932,
+      "step": 3792
+    },
+    {
+      "epoch": 0.47100459456103316,
+      "grad_norm": 0.2400045543909073,
+      "learning_rate": 5e-05,
+      "loss": 1.905,
+      "step": 3793
+    },
+    {
+      "epoch": 0.4711287718862536,
+      "grad_norm": 0.2752191424369812,
+      "learning_rate": 5e-05,
+      "loss": 2.3036,
+      "step": 3794
+    },
+    {
+      "epoch": 0.471252949211474,
+      "grad_norm": 0.2367798238992691,
+      "learning_rate": 5e-05,
+      "loss": 1.7053,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4713771265366944,
+      "grad_norm": 0.27545031905174255,
+      "learning_rate": 5e-05,
+      "loss": 2.1721,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4715013038619148,
+      "grad_norm": 0.2445094734430313,
+      "learning_rate": 5e-05,
+      "loss": 1.8001,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4716254811871352,
+      "grad_norm": 0.27212825417518616,
+      "learning_rate": 5e-05,
+      "loss": 2.0131,
+      "step": 3798
+    },
+    {
+      "epoch": 0.4717496585123556,
+      "grad_norm": 0.3546438217163086,
+      "learning_rate": 5e-05,
+      "loss": 2.7153,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4718738358375761,
+      "grad_norm": 0.26469165086746216,
+      "learning_rate": 5e-05,
+      "loss": 1.9978,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4719980131627965,
+      "grad_norm": 0.26541659235954285,
+      "learning_rate": 5e-05,
+      "loss": 1.9269,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4721221904880169,
+      "grad_norm": 0.25470834970474243,
+      "learning_rate": 5e-05,
+      "loss": 2.1003,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4722463678132373,
+      "grad_norm": 0.2495511770248413,
+      "learning_rate": 5e-05,
+      "loss": 2.165,
+      "step": 3803
+    },
+    {
+      "epoch": 0.4723705451384577,
+      "grad_norm": 0.3341478109359741,
+      "learning_rate": 5e-05,
+      "loss": 2.4154,
+      "step": 3804
+    },
+    {
+      "epoch": 0.47249472246367813,
+      "grad_norm": 0.2762279510498047,
+      "learning_rate": 5e-05,
+      "loss": 2.3299,
+      "step": 3805
+    },
+    {
+      "epoch": 0.47261889978889854,
+      "grad_norm": 0.23955392837524414,
+      "learning_rate": 5e-05,
+      "loss": 1.5116,
+      "step": 3806
+    },
+    {
+      "epoch": 0.47274307711411895,
+      "grad_norm": 0.27507612109184265,
+      "learning_rate": 5e-05,
+      "loss": 2.1639,
+      "step": 3807
+    },
+    {
+      "epoch": 0.47286725443933936,
+      "grad_norm": 0.22767682373523712,
+      "learning_rate": 5e-05,
+      "loss": 1.6815,
+      "step": 3808
+    },
+    {
+      "epoch": 0.47299143176455977,
+      "grad_norm": 0.25320327281951904,
+      "learning_rate": 5e-05,
+      "loss": 1.8617,
+      "step": 3809
+    },
+    {
+      "epoch": 0.47311560908978023,
+      "grad_norm": 0.2898845076560974,
+      "learning_rate": 5e-05,
+      "loss": 2.6267,
+      "step": 3810
+    },
+    {
+      "epoch": 0.47323978641500064,
+      "grad_norm": 0.22428782284259796,
+      "learning_rate": 5e-05,
+      "loss": 1.6021,
+      "step": 3811
+    },
+    {
+      "epoch": 0.47336396374022105,
+      "grad_norm": 0.23096810281276703,
+      "learning_rate": 5e-05,
+      "loss": 1.6784,
+      "step": 3812
+    },
+    {
+      "epoch": 0.47348814106544146,
+      "grad_norm": 0.25984397530555725,
+      "learning_rate": 5e-05,
+      "loss": 2.1357,
+      "step": 3813
+    },
+    {
+      "epoch": 0.47361231839066187,
+      "grad_norm": 0.2621871829032898,
+      "learning_rate": 5e-05,
+      "loss": 2.0393,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4737364957158823,
+      "grad_norm": 0.23887096345424652,
+      "learning_rate": 5e-05,
+      "loss": 1.9216,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4738606730411027,
+      "grad_norm": 0.2649695873260498,
+      "learning_rate": 5e-05,
+      "loss": 2.1198,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4739848503663231,
+      "grad_norm": 0.23518216609954834,
+      "learning_rate": 5e-05,
+      "loss": 1.6495,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4741090276915435,
+      "grad_norm": 0.36547526717185974,
+      "learning_rate": 5e-05,
+      "loss": 2.8033,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4742332050167639,
+      "grad_norm": 0.3875651955604553,
+      "learning_rate": 5e-05,
+      "loss": 2.9284,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4743573823419844,
+      "grad_norm": 0.2534061074256897,
+      "learning_rate": 5e-05,
+      "loss": 1.9848,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4744815596672048,
+      "grad_norm": 0.26507750153541565,
+      "learning_rate": 5e-05,
+      "loss": 2.1233,
+      "step": 3821
+    },
+    {
+      "epoch": 0.4746057369924252,
+      "grad_norm": 0.29006949067115784,
+      "learning_rate": 5e-05,
+      "loss": 2.0835,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4747299143176456,
+      "grad_norm": 0.2911592125892639,
+      "learning_rate": 5e-05,
+      "loss": 2.0187,
+      "step": 3823
+    },
+    {
+      "epoch": 0.474854091642866,
+      "grad_norm": 0.2968220114707947,
+      "learning_rate": 5e-05,
+      "loss": 2.2836,
+      "step": 3824
+    },
+    {
+      "epoch": 0.4749782689680864,
+      "grad_norm": 0.24679730832576752,
+      "learning_rate": 5e-05,
+      "loss": 1.7089,
+      "step": 3825
+    },
+    {
+      "epoch": 0.47510244629330683,
+      "grad_norm": 0.2892230451107025,
+      "learning_rate": 5e-05,
+      "loss": 2.2021,
+      "step": 3826
+    },
+    {
+      "epoch": 0.47522662361852724,
+      "grad_norm": 0.2447524517774582,
+      "learning_rate": 5e-05,
+      "loss": 1.8284,
+      "step": 3827
+    },
+    {
+      "epoch": 0.47535080094374765,
+      "grad_norm": 0.28342849016189575,
+      "learning_rate": 5e-05,
+      "loss": 1.6486,
+      "step": 3828
+    },
+    {
+      "epoch": 0.47547497826896806,
+      "grad_norm": 0.3388791084289551,
+      "learning_rate": 5e-05,
+      "loss": 2.2087,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4755991555941885,
+      "grad_norm": 0.2463005632162094,
+      "learning_rate": 5e-05,
+      "loss": 1.8514,
+      "step": 3830
+    },
+    {
+      "epoch": 0.47572333291940894,
+      "grad_norm": 0.23328736424446106,
+      "learning_rate": 5e-05,
+      "loss": 1.6566,
+      "step": 3831
+    },
+    {
+      "epoch": 0.47584751024462935,
+      "grad_norm": 0.6054593324661255,
+      "learning_rate": 5e-05,
+      "loss": 3.4001,
+      "step": 3832
+    },
+    {
+      "epoch": 0.47597168756984976,
+      "grad_norm": 0.30469247698783875,
+      "learning_rate": 5e-05,
+      "loss": 2.2101,
+      "step": 3833
+    },
+    {
+      "epoch": 0.47609586489507016,
+      "grad_norm": 0.23499421775341034,
+      "learning_rate": 5e-05,
+      "loss": 1.8245,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4762200422202906,
+      "grad_norm": 0.24627360701560974,
+      "learning_rate": 5e-05,
+      "loss": 1.746,
+      "step": 3835
+    },
+    {
+      "epoch": 0.476344219545511,
+      "grad_norm": 0.254901260137558,
+      "learning_rate": 5e-05,
+      "loss": 1.9631,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4764683968707314,
+      "grad_norm": 0.24958916008472443,
+      "learning_rate": 5e-05,
+      "loss": 1.8511,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4765925741959518,
+      "grad_norm": 0.25346264243125916,
+      "learning_rate": 5e-05,
+      "loss": 1.7831,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4767167515211722,
+      "grad_norm": 0.2501986622810364,
+      "learning_rate": 5e-05,
+      "loss": 2.2045,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4768409288463927,
+      "grad_norm": 0.23319843411445618,
+      "learning_rate": 5e-05,
+      "loss": 1.5974,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4769651061716131,
+      "grad_norm": 0.3162629306316376,
+      "learning_rate": 5e-05,
+      "loss": 2.056,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4770892834968335,
+      "grad_norm": 0.24404089152812958,
+      "learning_rate": 5e-05,
+      "loss": 1.6963,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4772134608220539,
+      "grad_norm": 0.24072685837745667,
+      "learning_rate": 5e-05,
+      "loss": 1.9766,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4773376381472743,
+      "grad_norm": 0.2808510363101959,
+      "learning_rate": 5e-05,
+      "loss": 1.989,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4774618154724947,
+      "grad_norm": 0.24510209262371063,
+      "learning_rate": 5e-05,
+      "loss": 1.8321,
+      "step": 3845
+    },
+    {
+      "epoch": 0.47758599279771513,
+      "grad_norm": 0.23880541324615479,
+      "learning_rate": 5e-05,
+      "loss": 1.5521,
+      "step": 3846
+    },
+    {
+      "epoch": 0.47771017012293554,
+      "grad_norm": 0.2707190215587616,
+      "learning_rate": 5e-05,
+      "loss": 2.2021,
+      "step": 3847
+    },
+    {
+      "epoch": 0.47783434744815595,
+      "grad_norm": 0.24554884433746338,
+      "learning_rate": 5e-05,
+      "loss": 1.7241,
+      "step": 3848
+    },
+    {
+      "epoch": 0.47795852477337636,
+      "grad_norm": 0.2338775098323822,
+      "learning_rate": 5e-05,
+      "loss": 1.8223,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4780827020985968,
+      "grad_norm": 0.27805590629577637,
+      "learning_rate": 5e-05,
+      "loss": 2.1541,
+      "step": 3850
+    },
+    {
+      "epoch": 0.47820687942381723,
+      "grad_norm": 0.24526168406009674,
+      "learning_rate": 5e-05,
+      "loss": 2.0317,
+      "step": 3851
+    },
+    {
+      "epoch": 0.47833105674903764,
+      "grad_norm": 0.23409128189086914,
+      "learning_rate": 5e-05,
+      "loss": 1.8411,
+      "step": 3852
+    },
+    {
+      "epoch": 0.47845523407425805,
+      "grad_norm": 0.2632734179496765,
+      "learning_rate": 5e-05,
+      "loss": 2.305,
+      "step": 3853
+    },
+    {
+      "epoch": 0.47857941139947846,
+      "grad_norm": 0.34554019570350647,
+      "learning_rate": 5e-05,
+      "loss": 2.7506,
+      "step": 3854
+    },
+    {
+      "epoch": 0.47870358872469887,
+      "grad_norm": 0.26898279786109924,
+      "learning_rate": 5e-05,
+      "loss": 2.0283,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4788277660499193,
+      "grad_norm": 0.4894627630710602,
+      "learning_rate": 5e-05,
+      "loss": 3.9556,
+      "step": 3856
+    },
+    {
+      "epoch": 0.4789519433751397,
+      "grad_norm": 0.25828030705451965,
+      "learning_rate": 5e-05,
+      "loss": 1.7904,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4790761207003601,
+      "grad_norm": 0.2767735719680786,
+      "learning_rate": 5e-05,
+      "loss": 2.4639,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4792002980255805,
+      "grad_norm": 0.2716367244720459,
+      "learning_rate": 5e-05,
+      "loss": 1.9747,
+      "step": 3859
+    },
+    {
+      "epoch": 0.47932447535080097,
+      "grad_norm": 0.2745836079120636,
+      "learning_rate": 5e-05,
+      "loss": 2.0554,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4794486526760214,
+      "grad_norm": 0.29028451442718506,
+      "learning_rate": 5e-05,
+      "loss": 2.1883,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4795728300012418,
+      "grad_norm": 0.3114234209060669,
+      "learning_rate": 5e-05,
+      "loss": 2.3263,
+      "step": 3862
+    },
+    {
+      "epoch": 0.4796970073264622,
+      "grad_norm": 0.2626296281814575,
+      "learning_rate": 5e-05,
+      "loss": 1.6638,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4798211846516826,
+      "grad_norm": 0.3213566839694977,
+      "learning_rate": 5e-05,
+      "loss": 2.117,
+      "step": 3864
+    },
+    {
+      "epoch": 0.479945361976903,
+      "grad_norm": 0.3131120204925537,
+      "learning_rate": 5e-05,
+      "loss": 2.0017,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4800695393021234,
+      "grad_norm": 0.30492836236953735,
+      "learning_rate": 5e-05,
+      "loss": 2.4067,
+      "step": 3866
+    },
+    {
+      "epoch": 0.48019371662734384,
+      "grad_norm": 0.26359227299690247,
+      "learning_rate": 5e-05,
+      "loss": 1.727,
+      "step": 3867
+    },
+    {
+      "epoch": 0.48031789395256425,
+      "grad_norm": 0.24697642028331757,
+      "learning_rate": 5e-05,
+      "loss": 1.8881,
+      "step": 3868
+    },
+    {
+      "epoch": 0.48044207127778465,
+      "grad_norm": 0.247593492269516,
+      "learning_rate": 5e-05,
+      "loss": 1.9391,
+      "step": 3869
+    },
+    {
+      "epoch": 0.48056624860300506,
+      "grad_norm": 0.27301526069641113,
+      "learning_rate": 5e-05,
+      "loss": 2.0986,
+      "step": 3870
+    },
+    {
+      "epoch": 0.48069042592822553,
+      "grad_norm": 0.2713072896003723,
+      "learning_rate": 5e-05,
+      "loss": 1.9867,
+      "step": 3871
+    },
+    {
+      "epoch": 0.48081460325344594,
+      "grad_norm": 0.2710086703300476,
+      "learning_rate": 5e-05,
+      "loss": 1.9851,
+      "step": 3872
+    },
+    {
+      "epoch": 0.48093878057866635,
+      "grad_norm": 0.2588740885257721,
+      "learning_rate": 5e-05,
+      "loss": 1.8256,
+      "step": 3873
+    },
+    {
+      "epoch": 0.48106295790388676,
+      "grad_norm": 0.3533439040184021,
+      "learning_rate": 5e-05,
+      "loss": 2.4111,
+      "step": 3874
+    },
+    {
+      "epoch": 0.48118713522910717,
+      "grad_norm": 0.24429315328598022,
+      "learning_rate": 5e-05,
+      "loss": 1.8134,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4813113125543276,
+      "grad_norm": 0.25022098422050476,
+      "learning_rate": 5e-05,
+      "loss": 1.891,
+      "step": 3876
+    },
+    {
+      "epoch": 0.481435489879548,
+      "grad_norm": 0.32831573486328125,
+      "learning_rate": 5e-05,
+      "loss": 2.787,
+      "step": 3877
+    },
+    {
+      "epoch": 0.4815596672047684,
+      "grad_norm": 0.2303081750869751,
+      "learning_rate": 5e-05,
+      "loss": 1.9263,
+      "step": 3878
+    },
+    {
+      "epoch": 0.4816838445299888,
+      "grad_norm": 0.27254289388656616,
+      "learning_rate": 5e-05,
+      "loss": 1.9381,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4818080218552092,
+      "grad_norm": 0.2308296263217926,
+      "learning_rate": 5e-05,
+      "loss": 1.6974,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4819321991804297,
+      "grad_norm": 0.2550886273384094,
+      "learning_rate": 5e-05,
+      "loss": 2.1294,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4820563765056501,
+      "grad_norm": 0.2510768473148346,
+      "learning_rate": 5e-05,
+      "loss": 1.8386,
+      "step": 3882
+    },
+    {
+      "epoch": 0.4821805538308705,
+      "grad_norm": 0.24736586213111877,
+      "learning_rate": 5e-05,
+      "loss": 1.7903,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4823047311560909,
+      "grad_norm": 0.25087112188339233,
+      "learning_rate": 5e-05,
+      "loss": 1.852,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4824289084813113,
+      "grad_norm": 0.258667528629303,
+      "learning_rate": 5e-05,
+      "loss": 2.1423,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4825530858065317,
+      "grad_norm": 0.24961206316947937,
+      "learning_rate": 5e-05,
+      "loss": 2.0568,
+      "step": 3886
+    },
+    {
+      "epoch": 0.48267726313175213,
+      "grad_norm": 0.24150574207305908,
+      "learning_rate": 5e-05,
+      "loss": 1.8317,
+      "step": 3887
+    },
+    {
+      "epoch": 0.48280144045697254,
+      "grad_norm": 0.2294166535139084,
+      "learning_rate": 5e-05,
+      "loss": 1.7282,
+      "step": 3888
+    },
+    {
+      "epoch": 0.48292561778219295,
+      "grad_norm": 0.33131325244903564,
+      "learning_rate": 5e-05,
+      "loss": 2.9518,
+      "step": 3889
+    },
+    {
+      "epoch": 0.48304979510741336,
+      "grad_norm": 0.31336402893066406,
+      "learning_rate": 5e-05,
+      "loss": 2.7053,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4831739724326338,
+      "grad_norm": 0.5472186803817749,
+      "learning_rate": 5e-05,
+      "loss": 3.4151,
+      "step": 3891
+    },
+    {
+      "epoch": 0.48329814975785423,
+      "grad_norm": 0.29547688364982605,
+      "learning_rate": 5e-05,
+      "loss": 2.1959,
+      "step": 3892
+    },
+    {
+      "epoch": 0.48342232708307464,
+      "grad_norm": 0.22589442133903503,
+      "learning_rate": 5e-05,
+      "loss": 1.5488,
+      "step": 3893
+    },
+    {
+      "epoch": 0.48354650440829505,
+      "grad_norm": 0.30734023451805115,
+      "learning_rate": 5e-05,
+      "loss": 2.1366,
+      "step": 3894
+    },
+    {
+      "epoch": 0.48367068173351546,
+      "grad_norm": 0.265781968832016,
+      "learning_rate": 5e-05,
+      "loss": 1.6933,
+      "step": 3895
+    },
+    {
+      "epoch": 0.48379485905873587,
+      "grad_norm": 0.2778889834880829,
+      "learning_rate": 5e-05,
+      "loss": 2.3719,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4839190363839563,
+      "grad_norm": 0.2705020606517792,
+      "learning_rate": 5e-05,
+      "loss": 2.1464,
+      "step": 3897
+    },
+    {
+      "epoch": 0.4840432137091767,
+      "grad_norm": 0.234904944896698,
+      "learning_rate": 5e-05,
+      "loss": 1.9484,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4841673910343971,
+      "grad_norm": 0.24568559229373932,
+      "learning_rate": 5e-05,
+      "loss": 1.8188,
+      "step": 3899
+    },
+    {
+      "epoch": 0.4842915683596175,
+      "grad_norm": 0.26236119866371155,
+      "learning_rate": 5e-05,
+      "loss": 1.7656,
+      "step": 3900
+    },
+    {
+      "epoch": 0.484415745684838,
+      "grad_norm": 0.25771012902259827,
+      "learning_rate": 5e-05,
+      "loss": 2.3894,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4845399230100584,
+      "grad_norm": 0.25078269839286804,
+      "learning_rate": 5e-05,
+      "loss": 1.9254,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4846641003352788,
+      "grad_norm": 0.31102854013442993,
+      "learning_rate": 5e-05,
+      "loss": 2.2158,
+      "step": 3903
+    },
+    {
+      "epoch": 0.4847882776604992,
+      "grad_norm": 0.23618362843990326,
+      "learning_rate": 5e-05,
+      "loss": 1.8634,
+      "step": 3904
+    },
+    {
+      "epoch": 0.4849124549857196,
+      "grad_norm": 0.23569625616073608,
+      "learning_rate": 5e-05,
+      "loss": 1.6641,
+      "step": 3905
+    },
+    {
+      "epoch": 0.48503663231094,
+      "grad_norm": 0.3056221604347229,
+      "learning_rate": 5e-05,
+      "loss": 2.3442,
+      "step": 3906
+    },
+    {
+      "epoch": 0.48516080963616043,
+      "grad_norm": 0.35795995593070984,
+      "learning_rate": 5e-05,
+      "loss": 2.6869,
+      "step": 3907
+    },
+    {
+      "epoch": 0.48528498696138084,
+      "grad_norm": 0.2756887674331665,
+      "learning_rate": 5e-05,
+      "loss": 1.8886,
+      "step": 3908
+    },
+    {
+      "epoch": 0.48540916428660125,
+      "grad_norm": 0.3128836750984192,
+      "learning_rate": 5e-05,
+      "loss": 2.661,
+      "step": 3909
+    },
+    {
+      "epoch": 0.48553334161182166,
+      "grad_norm": 0.3112477660179138,
+      "learning_rate": 5e-05,
+      "loss": 2.0602,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4856575189370421,
+      "grad_norm": 0.24958962202072144,
+      "learning_rate": 5e-05,
+      "loss": 1.9602,
+      "step": 3911
+    },
+    {
+      "epoch": 0.48578169626226253,
+      "grad_norm": 0.2798370122909546,
+      "learning_rate": 5e-05,
+      "loss": 2.1783,
+      "step": 3912
+    },
+    {
+      "epoch": 0.48590587358748294,
+      "grad_norm": 0.26496008038520813,
+      "learning_rate": 5e-05,
+      "loss": 2.2958,
+      "step": 3913
+    },
+    {
+      "epoch": 0.48603005091270335,
+      "grad_norm": 0.5013224482536316,
+      "learning_rate": 5e-05,
+      "loss": 3.5574,
+      "step": 3914
+    },
+    {
+      "epoch": 0.48615422823792376,
+      "grad_norm": 0.2615935206413269,
+      "learning_rate": 5e-05,
+      "loss": 2.1525,
+      "step": 3915
+    },
+    {
+      "epoch": 0.48627840556314417,
+      "grad_norm": 0.23061394691467285,
+      "learning_rate": 5e-05,
+      "loss": 1.767,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4864025828883646,
+      "grad_norm": 0.2588561177253723,
+      "learning_rate": 5e-05,
+      "loss": 1.9703,
+      "step": 3917
+    },
+    {
+      "epoch": 0.486526760213585,
+      "grad_norm": 0.2811073958873749,
+      "learning_rate": 5e-05,
+      "loss": 2.8899,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4866509375388054,
+      "grad_norm": 0.3036738634109497,
+      "learning_rate": 5e-05,
+      "loss": 2.1682,
+      "step": 3919
+    },
+    {
+      "epoch": 0.4867751148640258,
+      "grad_norm": 0.2848130464553833,
+      "learning_rate": 5e-05,
+      "loss": 2.2619,
+      "step": 3920
+    },
+    {
+      "epoch": 0.48689929218924627,
+      "grad_norm": 0.24663934111595154,
+      "learning_rate": 5e-05,
+      "loss": 2.0169,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4870234695144667,
+      "grad_norm": 0.24509768187999725,
+      "learning_rate": 5e-05,
+      "loss": 1.6234,
+      "step": 3922
+    },
+    {
+      "epoch": 0.4871476468396871,
+      "grad_norm": 0.3036537170410156,
+      "learning_rate": 5e-05,
+      "loss": 2.3312,
+      "step": 3923
+    },
+    {
+      "epoch": 0.4872718241649075,
+      "grad_norm": 0.2622587978839874,
+      "learning_rate": 5e-05,
+      "loss": 1.8345,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4873960014901279,
+      "grad_norm": 0.33658429980278015,
+      "learning_rate": 5e-05,
+      "loss": 2.5304,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4875201788153483,
+      "grad_norm": 0.2482730597257614,
+      "learning_rate": 5e-05,
+      "loss": 1.872,
+      "step": 3926
+    },
+    {
+      "epoch": 0.4876443561405687,
+      "grad_norm": 0.30108222365379333,
+      "learning_rate": 5e-05,
+      "loss": 2.4997,
+      "step": 3927
+    },
+    {
+      "epoch": 0.48776853346578913,
+      "grad_norm": 0.2725526690483093,
+      "learning_rate": 5e-05,
+      "loss": 2.6676,
+      "step": 3928
+    },
+    {
+      "epoch": 0.48789271079100954,
+      "grad_norm": 0.2819851040840149,
+      "learning_rate": 5e-05,
+      "loss": 1.7515,
+      "step": 3929
+    },
+    {
+      "epoch": 0.48801688811622995,
+      "grad_norm": 0.23388932645320892,
+      "learning_rate": 5e-05,
+      "loss": 1.644,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4881410654414504,
+      "grad_norm": 0.23615571856498718,
+      "learning_rate": 5e-05,
+      "loss": 1.9233,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4882652427666708,
+      "grad_norm": 0.28926417231559753,
+      "learning_rate": 5e-05,
+      "loss": 2.016,
+      "step": 3932
+    },
+    {
+      "epoch": 0.48838942009189124,
+      "grad_norm": 0.2590161859989166,
+      "learning_rate": 5e-05,
+      "loss": 2.1942,
+      "step": 3933
+    },
+    {
+      "epoch": 0.48851359741711164,
+      "grad_norm": 0.2569301724433899,
+      "learning_rate": 5e-05,
+      "loss": 1.9491,
+      "step": 3934
+    },
+    {
+      "epoch": 0.48863777474233205,
+      "grad_norm": 0.24590009450912476,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 3935
+    },
+    {
+      "epoch": 0.48876195206755246,
+      "grad_norm": 0.2507183849811554,
+      "learning_rate": 5e-05,
+      "loss": 1.9813,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4888861293927729,
+      "grad_norm": 0.2806329131126404,
+      "learning_rate": 5e-05,
+      "loss": 2.1556,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4890103067179933,
+      "grad_norm": 0.2593044340610504,
+      "learning_rate": 5e-05,
+      "loss": 2.1487,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4891344840432137,
+      "grad_norm": 0.2982097566127777,
+      "learning_rate": 5e-05,
+      "loss": 3.0302,
+      "step": 3939
+    },
+    {
+      "epoch": 0.4892586613684341,
+      "grad_norm": 0.29035472869873047,
+      "learning_rate": 5e-05,
+      "loss": 2.4651,
+      "step": 3940
+    },
+    {
+      "epoch": 0.48938283869365456,
+      "grad_norm": 0.39894893765449524,
+      "learning_rate": 5e-05,
+      "loss": 1.6706,
+      "step": 3941
+    },
+    {
+      "epoch": 0.489507016018875,
+      "grad_norm": 0.24433690309524536,
+      "learning_rate": 5e-05,
+      "loss": 2.0285,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4896311933440954,
+      "grad_norm": 0.2709929943084717,
+      "learning_rate": 5e-05,
+      "loss": 3.0003,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4897553706693158,
+      "grad_norm": 0.24825885891914368,
+      "learning_rate": 5e-05,
+      "loss": 1.7201,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4898795479945362,
+      "grad_norm": 0.28034350275993347,
+      "learning_rate": 5e-05,
+      "loss": 2.0749,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4900037253197566,
+      "grad_norm": 0.2617465555667877,
+      "learning_rate": 5e-05,
+      "loss": 1.8295,
+      "step": 3946
+    },
+    {
+      "epoch": 0.490127902644977,
+      "grad_norm": 0.48099711537361145,
+      "learning_rate": 5e-05,
+      "loss": 2.5138,
+      "step": 3947
+    },
+    {
+      "epoch": 0.49025207997019743,
+      "grad_norm": 0.24878370761871338,
+      "learning_rate": 5e-05,
+      "loss": 1.6884,
+      "step": 3948
+    },
+    {
+      "epoch": 0.49037625729541784,
+      "grad_norm": 0.28135889768600464,
+      "learning_rate": 5e-05,
+      "loss": 2.2166,
+      "step": 3949
+    },
+    {
+      "epoch": 0.49050043462063825,
+      "grad_norm": 0.2540428638458252,
+      "learning_rate": 5e-05,
+      "loss": 2.0954,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4906246119458587,
+      "grad_norm": 0.2989576458930969,
+      "learning_rate": 5e-05,
+      "loss": 2.9127,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4907487892710791,
+      "grad_norm": 0.35138800740242004,
+      "learning_rate": 5e-05,
+      "loss": 2.1535,
+      "step": 3952
+    },
+    {
+      "epoch": 0.49087296659629953,
+      "grad_norm": 0.2919287085533142,
+      "learning_rate": 5e-05,
+      "loss": 2.4477,
+      "step": 3953
+    },
+    {
+      "epoch": 0.49099714392151994,
+      "grad_norm": 0.28278595209121704,
+      "learning_rate": 5e-05,
+      "loss": 2.2832,
+      "step": 3954
+    },
+    {
+      "epoch": 0.49112132124674035,
+      "grad_norm": 0.24463866651058197,
+      "learning_rate": 5e-05,
+      "loss": 2.0512,
+      "step": 3955
+    },
+    {
+      "epoch": 0.49124549857196076,
+      "grad_norm": 0.25053003430366516,
+      "learning_rate": 5e-05,
+      "loss": 2.3639,
+      "step": 3956
+    },
+    {
+      "epoch": 0.49136967589718117,
+      "grad_norm": 0.24183261394500732,
+      "learning_rate": 5e-05,
+      "loss": 1.7411,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4914938532224016,
+      "grad_norm": 0.2536546587944031,
+      "learning_rate": 5e-05,
+      "loss": 2.0555,
+      "step": 3958
+    },
+    {
+      "epoch": 0.491618030547622,
+      "grad_norm": 0.2728576362133026,
+      "learning_rate": 5e-05,
+      "loss": 2.2935,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4917422078728424,
+      "grad_norm": 0.35700732469558716,
+      "learning_rate": 5e-05,
+      "loss": 2.4759,
+      "step": 3960
+    },
+    {
+      "epoch": 0.49186638519806286,
+      "grad_norm": 0.251714825630188,
+      "learning_rate": 5e-05,
+      "loss": 1.5906,
+      "step": 3961
+    },
+    {
+      "epoch": 0.49199056252328327,
+      "grad_norm": 0.27333173155784607,
+      "learning_rate": 5e-05,
+      "loss": 2.0215,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4921147398485037,
+      "grad_norm": 0.26484155654907227,
+      "learning_rate": 5e-05,
+      "loss": 2.2396,
+      "step": 3963
+    },
+    {
+      "epoch": 0.4922389171737241,
+      "grad_norm": 0.2907828688621521,
+      "learning_rate": 5e-05,
+      "loss": 1.8995,
+      "step": 3964
+    },
+    {
+      "epoch": 0.4923630944989445,
+      "grad_norm": 0.24295277893543243,
+      "learning_rate": 5e-05,
+      "loss": 1.9824,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4924872718241649,
+      "grad_norm": 0.24281099438667297,
+      "learning_rate": 5e-05,
+      "loss": 1.8336,
+      "step": 3966
+    },
+    {
+      "epoch": 0.4926114491493853,
+      "grad_norm": 0.28015756607055664,
+      "learning_rate": 5e-05,
+      "loss": 2.0993,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4927356264746057,
+      "grad_norm": 0.24537059664726257,
+      "learning_rate": 5e-05,
+      "loss": 1.9016,
+      "step": 3968
+    },
+    {
+      "epoch": 0.49285980379982613,
+      "grad_norm": 0.28075307607650757,
+      "learning_rate": 5e-05,
+      "loss": 2.1867,
+      "step": 3969
+    },
+    {
+      "epoch": 0.49298398112504654,
+      "grad_norm": 0.23990865051746368,
+      "learning_rate": 5e-05,
+      "loss": 1.8074,
+      "step": 3970
+    },
+    {
+      "epoch": 0.493108158450267,
+      "grad_norm": 0.2412898689508438,
+      "learning_rate": 5e-05,
+      "loss": 1.8406,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4932323357754874,
+      "grad_norm": 0.2641030550003052,
+      "learning_rate": 5e-05,
+      "loss": 2.1272,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4933565131007078,
+      "grad_norm": 0.29628852009773254,
+      "learning_rate": 5e-05,
+      "loss": 2.1693,
+      "step": 3973
+    },
+    {
+      "epoch": 0.49348069042592824,
+      "grad_norm": 0.23926807940006256,
+      "learning_rate": 5e-05,
+      "loss": 1.5727,
+      "step": 3974
+    },
+    {
+      "epoch": 0.49360486775114865,
+      "grad_norm": 0.25644993782043457,
+      "learning_rate": 5e-05,
+      "loss": 2.301,
+      "step": 3975
+    },
+    {
+      "epoch": 0.49372904507636906,
+      "grad_norm": 0.23636257648468018,
+      "learning_rate": 5e-05,
+      "loss": 1.8857,
+      "step": 3976
+    },
+    {
+      "epoch": 0.49385322240158946,
+      "grad_norm": 0.24126331508159637,
+      "learning_rate": 5e-05,
+      "loss": 1.8049,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4939773997268099,
+      "grad_norm": 0.2942225933074951,
+      "learning_rate": 5e-05,
+      "loss": 2.2518,
+      "step": 3978
+    },
+    {
+      "epoch": 0.4941015770520303,
+      "grad_norm": 0.23714369535446167,
+      "learning_rate": 5e-05,
+      "loss": 1.5358,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4942257543772507,
+      "grad_norm": 0.25190216302871704,
+      "learning_rate": 5e-05,
+      "loss": 1.9641,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4943499317024711,
+      "grad_norm": 0.2622401714324951,
+      "learning_rate": 5e-05,
+      "loss": 1.9031,
+      "step": 3981
+    },
+    {
+      "epoch": 0.49447410902769157,
+      "grad_norm": 0.23984573781490326,
+      "learning_rate": 5e-05,
+      "loss": 1.8997,
+      "step": 3982
+    },
+    {
+      "epoch": 0.494598286352912,
+      "grad_norm": 0.2397886961698532,
+      "learning_rate": 5e-05,
+      "loss": 1.8365,
+      "step": 3983
+    },
+    {
+      "epoch": 0.4947224636781324,
+      "grad_norm": 0.23667512834072113,
+      "learning_rate": 5e-05,
+      "loss": 1.8198,
+      "step": 3984
+    },
+    {
+      "epoch": 0.4948466410033528,
+      "grad_norm": 0.2407924085855484,
+      "learning_rate": 5e-05,
+      "loss": 1.8705,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4949708183285732,
+      "grad_norm": 0.31896474957466125,
+      "learning_rate": 5e-05,
+      "loss": 2.3651,
+      "step": 3986
+    },
+    {
+      "epoch": 0.4950949956537936,
+      "grad_norm": 0.2520856261253357,
+      "learning_rate": 5e-05,
+      "loss": 1.7471,
+      "step": 3987
+    },
+    {
+      "epoch": 0.495219172979014,
+      "grad_norm": 0.29035767912864685,
+      "learning_rate": 5e-05,
+      "loss": 2.7617,
+      "step": 3988
+    },
+    {
+      "epoch": 0.49534335030423443,
+      "grad_norm": 0.2656653821468353,
+      "learning_rate": 5e-05,
+      "loss": 2.0585,
+      "step": 3989
+    },
+    {
+      "epoch": 0.49546752762945484,
+      "grad_norm": 0.2466040849685669,
+      "learning_rate": 5e-05,
+      "loss": 1.9326,
+      "step": 3990
+    },
+    {
+      "epoch": 0.49559170495467525,
+      "grad_norm": 0.24749986827373505,
+      "learning_rate": 5e-05,
+      "loss": 1.8203,
+      "step": 3991
+    },
+    {
+      "epoch": 0.4957158822798957,
+      "grad_norm": 0.2577872574329376,
+      "learning_rate": 5e-05,
+      "loss": 1.846,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4958400596051161,
+      "grad_norm": 0.2301517277956009,
+      "learning_rate": 5e-05,
+      "loss": 1.6691,
+      "step": 3993
+    },
+    {
+      "epoch": 0.49596423693033653,
+      "grad_norm": 0.23217348754405975,
+      "learning_rate": 5e-05,
+      "loss": 1.7943,
+      "step": 3994
+    },
+    {
+      "epoch": 0.49608841425555694,
+      "grad_norm": 0.24767589569091797,
+      "learning_rate": 5e-05,
+      "loss": 2.0368,
+      "step": 3995
+    },
+    {
+      "epoch": 0.49621259158077735,
+      "grad_norm": 0.37302926182746887,
+      "learning_rate": 5e-05,
+      "loss": 2.9305,
+      "step": 3996
+    },
+    {
+      "epoch": 0.49633676890599776,
+      "grad_norm": 0.2919287085533142,
+      "learning_rate": 5e-05,
+      "loss": 2.0212,
+      "step": 3997
+    },
+    {
+      "epoch": 0.49646094623121817,
+      "grad_norm": 0.2996363341808319,
+      "learning_rate": 5e-05,
+      "loss": 2.9136,
+      "step": 3998
+    },
+    {
+      "epoch": 0.4965851235564386,
+      "grad_norm": 0.24742484092712402,
+      "learning_rate": 5e-05,
+      "loss": 1.7882,
+      "step": 3999
+    },
+    {
+      "epoch": 0.496709300881659,
+      "grad_norm": 0.24979206919670105,
+      "learning_rate": 5e-05,
+      "loss": 1.8959,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8053,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.165995710316544e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}