diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6150 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0997923404926186,
+  "eval_steps": 500,
+  "global_step": 872,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5686,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.6127,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5707,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.59,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5644,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5914,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5969,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.586,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5967,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.6217,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5745,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5426,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5519,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5843,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.6121,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.6276,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.6146,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5983,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.613,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5787,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5984,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5756,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.93226623535156,
+      "learning_rate": 1.1363636363636364e-07,
+      "loss": 3.5773,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.64237976074219,
+      "learning_rate": 2.2727272727272729e-07,
+      "loss": 3.5666,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.94294738769531,
+      "learning_rate": 3.409090909090909e-07,
+      "loss": 3.5691,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.58763885498047,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 3.5667,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.7476577758789,
+      "learning_rate": 5.681818181818182e-07,
+      "loss": 3.5572,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 75.63665008544922,
+      "learning_rate": 6.818181818181818e-07,
+      "loss": 3.4797,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.8890380859375,
+      "learning_rate": 7.954545454545455e-07,
+      "loss": 3.4472,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 69.56644439697266,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 3.2339,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 67.08919525146484,
+      "learning_rate": 1.0227272727272729e-06,
+      "loss": 3.1385,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 48.23773956298828,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 2.4925,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 45.45056915283203,
+      "learning_rate": 1.25e-06,
+      "loss": 2.4319,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 38.1535758972168,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 2.2507,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 35.65153503417969,
+      "learning_rate": 1.4772727272727275e-06,
+      "loss": 2.1687,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.733210563659668,
+      "learning_rate": 1.590909090909091e-06,
+      "loss": 1.7378,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.957545280456543,
+      "learning_rate": 1.7045454545454546e-06,
+      "loss": 1.6735,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.594935417175293,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.5907,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.634173393249512,
+      "learning_rate": 1.931818181818182e-06,
+      "loss": 1.5384,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.905609130859375,
+      "learning_rate": 2.0454545454545457e-06,
+      "loss": 1.5125,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0466458797454834,
+      "learning_rate": 2.1590909090909092e-06,
+      "loss": 1.4422,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.975010633468628,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.4561,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3392484188079834,
+      "learning_rate": 2.3863636363636367e-06,
+      "loss": 1.4086,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.107882261276245,
+      "learning_rate": 2.5e-06,
+      "loss": 1.4129,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7513774633407593,
+      "learning_rate": 2.6136363636363637e-06,
+      "loss": 1.3709,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5137622356414795,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.3971,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5218021869659424,
+      "learning_rate": 2.8409090909090916e-06,
+      "loss": 1.3869,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3554155826568604,
+      "learning_rate": 2.954545454545455e-06,
+      "loss": 1.3618,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2151788473129272,
+      "learning_rate": 3.0681818181818186e-06,
+      "loss": 1.3641,
+      "step": 49
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1324663162231445,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 1.3731,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0742617845535278,
+      "learning_rate": 3.2954545454545456e-06,
+      "loss": 1.3624,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1047883033752441,
+      "learning_rate": 3.409090909090909e-06,
+      "loss": 1.3796,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1034663915634155,
+      "learning_rate": 3.522727272727273e-06,
+      "loss": 1.3592,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0702307224273682,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.3698,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0497913360595703,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.368,
+      "step": 55
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0903385877609253,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 1.3601,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0922733545303345,
+      "learning_rate": 3.9772727272727275e-06,
+      "loss": 1.3738,
+      "step": 57
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0606937408447266,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.3495,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0293612480163574,
+      "learning_rate": 4.204545454545455e-06,
+      "loss": 1.3531,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.081424355506897,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 1.3758,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.160018801689148,
+      "learning_rate": 4.4318181818181824e-06,
+      "loss": 1.3745,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1710997819900513,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.391,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1680262088775635,
+      "learning_rate": 4.6590909090909095e-06,
+      "loss": 1.3737,
+      "step": 63
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.142930507659912,
+      "learning_rate": 4.772727272727273e-06,
+      "loss": 1.4017,
+      "step": 64
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.087868332862854,
+      "learning_rate": 4.8863636363636365e-06,
+      "loss": 1.3784,
+      "step": 65
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1250782012939453,
+      "learning_rate": 5e-06,
+      "loss": 1.3761,
+      "step": 66
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1870965957641602,
+      "learning_rate": 5.113636363636364e-06,
+      "loss": 1.3941,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1462435722351074,
+      "learning_rate": 5.2272727272727274e-06,
+      "loss": 1.3671,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1489235162734985,
+      "learning_rate": 5.340909090909091e-06,
+      "loss": 1.3945,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0989032983779907,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.3684,
+      "step": 70
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1220042705535889,
+      "learning_rate": 5.568181818181818e-06,
+      "loss": 1.3874,
+      "step": 71
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0458035469055176,
+      "learning_rate": 5.681818181818183e-06,
+      "loss": 1.4101,
+      "step": 72
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0890356302261353,
+      "learning_rate": 5.795454545454546e-06,
+      "loss": 1.4111,
+      "step": 73
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0306185483932495,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 1.375,
+      "step": 74
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0304069519042969,
+      "learning_rate": 6.022727272727273e-06,
+      "loss": 1.397,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0588593482971191,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 1.4001,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1896718740463257,
+      "learning_rate": 6.25e-06,
+      "loss": 1.387,
+      "step": 77
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.163329005241394,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 1.4085,
+      "step": 78
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0310306549072266,
+      "learning_rate": 6.477272727272727e-06,
+      "loss": 1.3856,
+      "step": 79
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.047789216041565,
+      "learning_rate": 6.590909090909091e-06,
+      "loss": 1.4072,
+      "step": 80
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0043468475341797,
+      "learning_rate": 6.704545454545454e-06,
+      "loss": 1.3974,
+      "step": 81
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.064086675643921,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.4015,
+      "step": 82
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1652333736419678,
+      "learning_rate": 6.931818181818183e-06,
+      "loss": 1.4042,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0301768779754639,
+      "learning_rate": 7.045454545454546e-06,
+      "loss": 1.3901,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0952802896499634,
+      "learning_rate": 7.15909090909091e-06,
+      "loss": 1.3807,
+      "step": 85
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.165692925453186,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 1.3928,
+      "step": 86
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1395169496536255,
+      "learning_rate": 7.386363636363637e-06,
+      "loss": 1.4028,
+      "step": 87
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1742874383926392,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.4118,
+      "step": 88
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3489391803741455,
+      "learning_rate": 7.613636363636364e-06,
+      "loss": 1.3887,
+      "step": 89
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2398784160614014,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 1.383,
+      "step": 90
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3594183921813965,
+      "learning_rate": 7.840909090909091e-06,
+      "loss": 1.4146,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2309668064117432,
+      "learning_rate": 7.954545454545455e-06,
+      "loss": 1.4056,
+      "step": 92
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1133555173873901,
+      "learning_rate": 8.068181818181819e-06,
+      "loss": 1.396,
+      "step": 93
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2161476612091064,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 1.4003,
+      "step": 94
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2214394807815552,
+      "learning_rate": 8.295454545454547e-06,
+      "loss": 1.3861,
+      "step": 95
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2064229249954224,
+      "learning_rate": 8.40909090909091e-06,
+      "loss": 1.4051,
+      "step": 96
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1662726402282715,
+      "learning_rate": 8.522727272727273e-06,
+      "loss": 1.3758,
+      "step": 97
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.295553207397461,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 1.4256,
+      "step": 98
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2915432453155518,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.3965,
+      "step": 99
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1948374509811401,
+      "learning_rate": 8.863636363636365e-06,
+      "loss": 1.4115,
+      "step": 100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1821595430374146,
+      "learning_rate": 8.977272727272727e-06,
+      "loss": 1.3967,
+      "step": 101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.151677131652832,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.3866,
+      "step": 102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2359035015106201,
+      "learning_rate": 9.204545454545455e-06,
+      "loss": 1.4052,
+      "step": 103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4146509170532227,
+      "learning_rate": 9.318181818181819e-06,
+      "loss": 1.3826,
+      "step": 104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2139616012573242,
+      "learning_rate": 9.431818181818183e-06,
+      "loss": 1.3798,
+      "step": 105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.205654263496399,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 1.3977,
+      "step": 106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.243720293045044,
+      "learning_rate": 9.65909090909091e-06,
+      "loss": 1.3886,
+      "step": 107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4605085849761963,
+      "learning_rate": 9.772727272727273e-06,
+      "loss": 1.372,
+      "step": 108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4088302850723267,
+      "learning_rate": 9.886363636363637e-06,
+      "loss": 1.391,
+      "step": 109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.425614595413208,
+      "learning_rate": 1e-05,
+      "loss": 1.42,
+      "step": 110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3304733037948608,
+      "learning_rate": 9.987244897959184e-06,
+      "loss": 1.3856,
+      "step": 111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3015329837799072,
+      "learning_rate": 9.974489795918369e-06,
+      "loss": 1.3929,
+      "step": 112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3881584405899048,
+      "learning_rate": 9.961734693877552e-06,
+      "loss": 1.3913,
+      "step": 113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.35237455368042,
+      "learning_rate": 9.948979591836737e-06,
+      "loss": 1.4069,
+      "step": 114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2924104928970337,
+      "learning_rate": 9.93622448979592e-06,
+      "loss": 1.3928,
+      "step": 115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3976210355758667,
+      "learning_rate": 9.923469387755103e-06,
+      "loss": 1.3904,
+      "step": 116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1047642230987549,
+      "learning_rate": 9.910714285714288e-06,
+      "loss": 1.3675,
+      "step": 117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0460392236709595,
+      "learning_rate": 9.89795918367347e-06,
+      "loss": 1.3787,
+      "step": 118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1290128231048584,
+      "learning_rate": 9.885204081632654e-06,
+      "loss": 1.3997,
+      "step": 119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1596468687057495,
+      "learning_rate": 9.872448979591838e-06,
+      "loss": 1.3838,
+      "step": 120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.18632972240448,
+      "learning_rate": 9.859693877551022e-06,
+      "loss": 1.3773,
+      "step": 121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2686792612075806,
+      "learning_rate": 9.846938775510205e-06,
+      "loss": 1.3823,
+      "step": 122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2351279258728027,
+      "learning_rate": 9.834183673469388e-06,
+      "loss": 1.3993,
+      "step": 123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3058531284332275,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 1.362,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3058429956436157,
+      "learning_rate": 9.808673469387756e-06,
+      "loss": 1.3728,
+      "step": 125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3963571786880493,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.3628,
+      "step": 126
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2370892763137817,
+      "learning_rate": 9.783163265306123e-06,
+      "loss": 1.3771,
+      "step": 127
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1372427940368652,
+      "learning_rate": 9.770408163265307e-06,
+      "loss": 1.3715,
+      "step": 128
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.206458330154419,
+      "learning_rate": 9.75765306122449e-06,
+      "loss": 1.3791,
+      "step": 129
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.201851487159729,
+      "learning_rate": 9.744897959183674e-06,
+      "loss": 1.3547,
+      "step": 130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.182597041130066,
+      "learning_rate": 9.732142857142858e-06,
+      "loss": 1.3697,
+      "step": 131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1894217729568481,
+      "learning_rate": 9.719387755102042e-06,
+      "loss": 1.3512,
+      "step": 132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.448066234588623,
+      "learning_rate": 9.706632653061225e-06,
+      "loss": 1.3765,
+      "step": 133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1208083629608154,
+      "learning_rate": 9.693877551020408e-06,
+      "loss": 1.3749,
+      "step": 134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0563664436340332,
+      "learning_rate": 9.681122448979593e-06,
+      "loss": 1.352,
+      "step": 135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.41444993019104,
+      "learning_rate": 9.668367346938776e-06,
+      "loss": 1.3603,
+      "step": 136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.186637043952942,
+      "learning_rate": 9.65561224489796e-06,
+      "loss": 1.3805,
+      "step": 137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2297229766845703,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 1.3559,
+      "step": 138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3162171840667725,
+      "learning_rate": 9.630102040816327e-06,
+      "loss": 1.3651,
+      "step": 139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.358998417854309,
+      "learning_rate": 9.617346938775512e-06,
+      "loss": 1.3841,
+      "step": 140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2178298234939575,
+      "learning_rate": 9.604591836734695e-06,
+      "loss": 1.3206,
+      "step": 141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3197548389434814,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 1.3747,
+      "step": 142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2247390747070312,
+      "learning_rate": 9.579081632653063e-06,
+      "loss": 1.3524,
+      "step": 143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2377766370773315,
+      "learning_rate": 9.566326530612246e-06,
+      "loss": 1.376,
+      "step": 144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1804473400115967,
+      "learning_rate": 9.55357142857143e-06,
+      "loss": 1.3553,
+      "step": 145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2282185554504395,
+      "learning_rate": 9.540816326530612e-06,
+      "loss": 1.3956,
+      "step": 146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.222644567489624,
+      "learning_rate": 9.528061224489797e-06,
+      "loss": 1.3485,
+      "step": 147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0799063444137573,
+      "learning_rate": 9.51530612244898e-06,
+      "loss": 1.3462,
+      "step": 148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0759501457214355,
+      "learning_rate": 9.502551020408163e-06,
+      "loss": 1.3571,
+      "step": 149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.391052007675171,
+      "learning_rate": 9.489795918367348e-06,
+      "loss": 1.3797,
+      "step": 150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2906898260116577,
+      "learning_rate": 9.477040816326531e-06,
+      "loss": 1.3281,
+      "step": 151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2738357782363892,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 1.3648,
+      "step": 152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4676612615585327,
+      "learning_rate": 9.451530612244899e-06,
+      "loss": 1.3382,
+      "step": 153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2316924333572388,
+      "learning_rate": 9.438775510204082e-06,
+      "loss": 1.3448,
+      "step": 154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.183659315109253,
+      "learning_rate": 9.426020408163265e-06,
+      "loss": 1.3405,
+      "step": 155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2155417203903198,
+      "learning_rate": 9.41326530612245e-06,
+      "loss": 1.3744,
+      "step": 156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2606340646743774,
+      "learning_rate": 9.400510204081633e-06,
+      "loss": 1.3449,
+      "step": 157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1220340728759766,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 1.3704,
+      "step": 158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1462124586105347,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.349,
+      "step": 159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0098296403884888,
+      "learning_rate": 9.362244897959184e-06,
+      "loss": 1.3564,
+      "step": 160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.06204354763031,
+      "learning_rate": 9.349489795918369e-06,
+      "loss": 1.3233,
+      "step": 161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0313581228256226,
+      "learning_rate": 9.336734693877552e-06,
+      "loss": 1.335,
+      "step": 162
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0908300876617432,
+      "learning_rate": 9.323979591836737e-06,
+      "loss": 1.3394,
+      "step": 163
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1179544925689697,
+      "learning_rate": 9.31122448979592e-06,
+      "loss": 1.3774,
+      "step": 164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.416581630706787,
+      "learning_rate": 9.298469387755103e-06,
+      "loss": 1.3598,
+      "step": 165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.228285312652588,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 1.3448,
+      "step": 166
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2312650680541992,
+      "learning_rate": 9.27295918367347e-06,
+      "loss": 1.3408,
+      "step": 167
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2846862077713013,
+      "learning_rate": 9.260204081632654e-06,
+      "loss": 1.3746,
+      "step": 168
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4658101797103882,
+      "learning_rate": 9.247448979591837e-06,
+      "loss": 1.3501,
+      "step": 169
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.335235357284546,
+      "learning_rate": 9.234693877551022e-06,
+      "loss": 1.3394,
+      "step": 170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.367955207824707,
+      "learning_rate": 9.221938775510205e-06,
+      "loss": 1.3385,
+      "step": 171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3143895864486694,
+      "learning_rate": 9.209183673469388e-06,
+      "loss": 1.3301,
+      "step": 172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4805288314819336,
+      "learning_rate": 9.196428571428571e-06,
+      "loss": 1.3193,
+      "step": 173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3719357252120972,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 1.345,
+      "step": 174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4281933307647705,
+      "learning_rate": 9.170918367346939e-06,
+      "loss": 1.3412,
+      "step": 175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3063050508499146,
+      "learning_rate": 9.158163265306124e-06,
+      "loss": 1.3251,
+      "step": 176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3859777450561523,
+      "learning_rate": 9.145408163265307e-06,
+      "loss": 1.3515,
+      "step": 177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2141677141189575,
+      "learning_rate": 9.13265306122449e-06,
+      "loss": 1.333,
+      "step": 178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2974693775177002,
+      "learning_rate": 9.119897959183674e-06,
+      "loss": 1.3433,
+      "step": 179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1979682445526123,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 1.35,
+      "step": 180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1833395957946777,
+      "learning_rate": 9.094387755102042e-06,
+      "loss": 1.3202,
+      "step": 181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1826756000518799,
+      "learning_rate": 9.081632653061225e-06,
+      "loss": 1.3266,
+      "step": 182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.306878685951233,
+      "learning_rate": 9.068877551020409e-06,
+      "loss": 1.3459,
+      "step": 183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3741576671600342,
+      "learning_rate": 9.056122448979593e-06,
+      "loss": 1.3479,
+      "step": 184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4168845415115356,
+      "learning_rate": 9.043367346938776e-06,
+      "loss": 1.3221,
+      "step": 185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4802296161651611,
+      "learning_rate": 9.03061224489796e-06,
+      "loss": 1.3377,
+      "step": 186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3033307790756226,
+      "learning_rate": 9.017857142857144e-06,
+      "loss": 1.3219,
+      "step": 187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3650187253952026,
+      "learning_rate": 9.005102040816327e-06,
+      "loss": 1.3257,
+      "step": 188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.263314962387085,
+      "learning_rate": 8.992346938775512e-06,
+      "loss": 1.3142,
+      "step": 189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2871077060699463,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 1.3418,
+      "step": 190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2486342191696167,
+      "learning_rate": 8.966836734693878e-06,
+      "loss": 1.311,
+      "step": 191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.309653878211975,
+      "learning_rate": 8.954081632653061e-06,
+      "loss": 1.3147,
+      "step": 192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2959657907485962,
+      "learning_rate": 8.941326530612246e-06,
+      "loss": 1.3277,
+      "step": 193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2131139039993286,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 1.3251,
+      "step": 194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2249537706375122,
+      "learning_rate": 8.915816326530612e-06,
+      "loss": 1.3304,
+      "step": 195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1446560621261597,
+      "learning_rate": 8.903061224489795e-06,
+      "loss": 1.3042,
+      "step": 196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1486533880233765,
+      "learning_rate": 8.89030612244898e-06,
+      "loss": 1.3105,
+      "step": 197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.113291621208191,
+      "learning_rate": 8.877551020408163e-06,
+      "loss": 1.3226,
+      "step": 198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1082658767700195,
+      "learning_rate": 8.864795918367348e-06,
+      "loss": 1.3176,
+      "step": 199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0742919445037842,
+      "learning_rate": 8.852040816326531e-06,
+      "loss": 1.3221,
+      "step": 200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.058107614517212,
+      "learning_rate": 8.839285714285714e-06,
+      "loss": 1.3176,
+      "step": 201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.05449640750885,
+      "learning_rate": 8.826530612244899e-06,
+      "loss": 1.3188,
+      "step": 202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.019459843635559,
+      "learning_rate": 8.813775510204082e-06,
+      "loss": 1.3058,
+      "step": 203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0865976810455322,
+      "learning_rate": 8.801020408163265e-06,
+      "loss": 1.3143,
+      "step": 204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0287652015686035,
+      "learning_rate": 8.78826530612245e-06,
+      "loss": 1.3002,
+      "step": 205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0019289255142212,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 1.3229,
+      "step": 206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9947201013565063,
+      "learning_rate": 8.762755102040818e-06,
+      "loss": 1.3127,
+      "step": 207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0613619089126587,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.3231,
+      "step": 208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1687304973602295,
+      "learning_rate": 8.737244897959184e-06,
+      "loss": 1.3026,
+      "step": 209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2115646600723267,
+      "learning_rate": 8.724489795918369e-06,
+      "loss": 1.3363,
+      "step": 210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2404974699020386,
+      "learning_rate": 8.711734693877552e-06,
+      "loss": 1.3328,
+      "step": 211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2207895517349243,
+      "learning_rate": 8.698979591836737e-06,
+      "loss": 1.3003,
+      "step": 212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3173328638076782,
+      "learning_rate": 8.68622448979592e-06,
+      "loss": 1.3428,
+      "step": 213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3301403522491455,
+      "learning_rate": 8.673469387755103e-06,
+      "loss": 1.2952,
+      "step": 214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.389487385749817,
+      "learning_rate": 8.660714285714286e-06,
+      "loss": 1.311,
+      "step": 215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4389171600341797,
+      "learning_rate": 8.64795918367347e-06,
+      "loss": 1.3105,
+      "step": 216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4631147384643555,
+      "learning_rate": 8.635204081632654e-06,
+      "loss": 1.3253,
+      "step": 217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3370922803878784,
+      "learning_rate": 8.622448979591837e-06,
+      "loss": 1.2933,
+      "step": 218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4107997417449951,
+      "learning_rate": 8.609693877551022e-06,
+      "loss": 1.297,
+      "step": 219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1963098049163818,
+      "learning_rate": 8.596938775510205e-06,
+      "loss": 1.3093,
+      "step": 220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.245422124862671,
+      "learning_rate": 8.584183673469388e-06,
+      "loss": 1.2944,
+      "step": 221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1765896081924438,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.3108,
+      "step": 222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3151187896728516,
+      "learning_rate": 8.558673469387756e-06,
+      "loss": 1.3015,
+      "step": 223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1630090475082397,
+      "learning_rate": 8.545918367346939e-06,
+      "loss": 1.2917,
+      "step": 224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2731425762176514,
+      "learning_rate": 8.533163265306124e-06,
+      "loss": 1.299,
+      "step": 225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2473809719085693,
+      "learning_rate": 8.520408163265307e-06,
+      "loss": 1.2883,
+      "step": 226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2504627704620361,
+      "learning_rate": 8.50765306122449e-06,
+      "loss": 1.3029,
+      "step": 227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.124976634979248,
+      "learning_rate": 8.494897959183675e-06,
+      "loss": 1.3063,
+      "step": 228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2146427631378174,
+      "learning_rate": 8.482142857142858e-06,
+      "loss": 1.3139,
+      "step": 229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1318598985671997,
+      "learning_rate": 8.469387755102042e-06,
+      "loss": 1.3245,
+      "step": 230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.175479769706726,
+      "learning_rate": 8.456632653061225e-06,
+      "loss": 1.2904,
+      "step": 231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.06119966506958,
+      "learning_rate": 8.443877551020409e-06,
+      "loss": 1.2826,
+      "step": 232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0992507934570312,
+      "learning_rate": 8.431122448979593e-06,
+      "loss": 1.3068,
+      "step": 233
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0282840728759766,
+      "learning_rate": 8.418367346938776e-06,
+      "loss": 1.2794,
+      "step": 234
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0799790620803833,
+      "learning_rate": 8.40561224489796e-06,
+      "loss": 1.2963,
+      "step": 235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9920130968093872,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 1.2985,
+      "step": 236
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9600382447242737,
+      "learning_rate": 8.380102040816327e-06,
+      "loss": 1.3017,
+      "step": 237
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0030544996261597,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 1.2824,
+      "step": 238
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0658328533172607,
+      "learning_rate": 8.354591836734695e-06,
+      "loss": 1.2738,
+      "step": 239
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0285568237304688,
+      "learning_rate": 8.341836734693878e-06,
+      "loss": 1.2932,
+      "step": 240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1148922443389893,
+      "learning_rate": 8.329081632653061e-06,
+      "loss": 1.2727,
+      "step": 241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0424553155899048,
+      "learning_rate": 8.316326530612246e-06,
+      "loss": 1.2917,
+      "step": 242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.09824538230896,
+      "learning_rate": 8.30357142857143e-06,
+      "loss": 1.2861,
+      "step": 243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1441254615783691,
+      "learning_rate": 8.290816326530612e-06,
+      "loss": 1.2894,
+      "step": 244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2667628526687622,
+      "learning_rate": 8.278061224489795e-06,
+      "loss": 1.2738,
+      "step": 245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3604241609573364,
+      "learning_rate": 8.26530612244898e-06,
+      "loss": 1.2889,
+      "step": 246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3580222129821777,
+      "learning_rate": 8.252551020408163e-06,
+      "loss": 1.3088,
+      "step": 247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2741780281066895,
+      "learning_rate": 8.239795918367348e-06,
+      "loss": 1.2915,
+      "step": 248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3847053050994873,
+      "learning_rate": 8.227040816326531e-06,
+      "loss": 1.3189,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2596681118011475,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 1.2751,
+      "step": 250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2427873611450195,
+      "learning_rate": 8.201530612244899e-06,
+      "loss": 1.2938,
+      "step": 251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2609047889709473,
+      "learning_rate": 8.188775510204082e-06,
+      "loss": 1.2948,
+      "step": 252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2723376750946045,
+      "learning_rate": 8.176020408163265e-06,
+      "loss": 1.2949,
+      "step": 253
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1737812757492065,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 1.2606,
+      "step": 254
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.237030267715454,
+      "learning_rate": 8.150510204081633e-06,
+      "loss": 1.2747,
+      "step": 255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.162434697151184,
+      "learning_rate": 8.137755102040818e-06,
+      "loss": 1.2841,
+      "step": 256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2295739650726318,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.2922,
+      "step": 257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1049543619155884,
+      "learning_rate": 8.112244897959184e-06,
+      "loss": 1.2724,
+      "step": 258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1978216171264648,
+      "learning_rate": 8.099489795918369e-06,
+      "loss": 1.294,
+      "step": 259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0655226707458496,
+      "learning_rate": 8.086734693877552e-06,
+      "loss": 1.2465,
+      "step": 260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1613059043884277,
+      "learning_rate": 8.073979591836735e-06,
+      "loss": 1.2683,
+      "step": 261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0504621267318726,
+      "learning_rate": 8.06122448979592e-06,
+      "loss": 1.2386,
+      "step": 262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.087553858757019,
+      "learning_rate": 8.048469387755103e-06,
+      "loss": 1.2491,
+      "step": 263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.946223258972168,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 1.2514,
+      "step": 264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9961144328117371,
+      "learning_rate": 8.02295918367347e-06,
+      "loss": 1.2781,
+      "step": 265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9003639221191406,
+      "learning_rate": 8.010204081632654e-06,
+      "loss": 1.2567,
+      "step": 266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0217918157577515,
+      "learning_rate": 7.997448979591837e-06,
+      "loss": 1.2775,
+      "step": 267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.002084732055664,
+      "learning_rate": 7.98469387755102e-06,
+      "loss": 1.2564,
+      "step": 268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1564198732376099,
+      "learning_rate": 7.971938775510205e-06,
+      "loss": 1.2637,
+      "step": 269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1586065292358398,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 1.2805,
+      "step": 270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2958130836486816,
+      "learning_rate": 7.946428571428571e-06,
+      "loss": 1.2673,
+      "step": 271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2151861190795898,
+      "learning_rate": 7.933673469387756e-06,
+      "loss": 1.2632,
+      "step": 272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2508829832077026,
+      "learning_rate": 7.920918367346939e-06,
+      "loss": 1.265,
+      "step": 273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1336225271224976,
+      "learning_rate": 7.908163265306124e-06,
+      "loss": 1.257,
+      "step": 274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1592440605163574,
+      "learning_rate": 7.895408163265307e-06,
+      "loss": 1.2628,
+      "step": 275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2461146116256714,
+      "learning_rate": 7.88265306122449e-06,
+      "loss": 1.2551,
+      "step": 276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3664628267288208,
+      "learning_rate": 7.869897959183675e-06,
+      "loss": 1.2658,
+      "step": 277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.201025128364563,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 1.2746,
+      "step": 278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1980972290039062,
+      "learning_rate": 7.844387755102042e-06,
+      "loss": 1.2641,
+      "step": 279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2280415296554565,
+      "learning_rate": 7.831632653061226e-06,
+      "loss": 1.2813,
+      "step": 280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1427682638168335,
+      "learning_rate": 7.818877551020409e-06,
+      "loss": 1.2531,
+      "step": 281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.162741780281067,
+      "learning_rate": 7.806122448979593e-06,
+      "loss": 1.2598,
+      "step": 282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1492100954055786,
+      "learning_rate": 7.793367346938776e-06,
+      "loss": 1.257,
+      "step": 283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2014518976211548,
+      "learning_rate": 7.78061224489796e-06,
+      "loss": 1.2553,
+      "step": 284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1967196464538574,
+      "learning_rate": 7.767857142857144e-06,
+      "loss": 1.2496,
+      "step": 285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2154698371887207,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 1.25,
+      "step": 286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1646188497543335,
+      "learning_rate": 7.74234693877551e-06,
+      "loss": 1.2374,
+      "step": 287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0902819633483887,
+      "learning_rate": 7.729591836734695e-06,
+      "loss": 1.2626,
+      "step": 288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1101752519607544,
+      "learning_rate": 7.716836734693878e-06,
+      "loss": 1.2402,
+      "step": 289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0915471315383911,
+      "learning_rate": 7.704081632653061e-06,
+      "loss": 1.2652,
+      "step": 290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.099541425704956,
+      "learning_rate": 7.691326530612245e-06,
+      "loss": 1.2388,
+      "step": 291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0305002927780151,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 1.2598,
+      "step": 292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1155153512954712,
+      "learning_rate": 7.665816326530612e-06,
+      "loss": 1.2334,
+      "step": 293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0462146997451782,
+      "learning_rate": 7.653061224489796e-06,
+      "loss": 1.2448,
+      "step": 294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0846874713897705,
+      "learning_rate": 7.64030612244898e-06,
+      "loss": 1.248,
+      "step": 295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9629496335983276,
+      "learning_rate": 7.627551020408163e-06,
+      "loss": 1.2395,
+      "step": 296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0308563709259033,
+      "learning_rate": 7.614795918367348e-06,
+      "loss": 1.2356,
+      "step": 297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.026279091835022,
+      "learning_rate": 7.602040816326531e-06,
+      "loss": 1.2351,
+      "step": 298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0194913148880005,
+      "learning_rate": 7.589285714285714e-06,
+      "loss": 1.2461,
+      "step": 299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9419485330581665,
+      "learning_rate": 7.576530612244899e-06,
+      "loss": 1.269,
+      "step": 300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0491052865982056,
+      "learning_rate": 7.563775510204082e-06,
+      "loss": 1.2414,
+      "step": 301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0324506759643555,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 1.2419,
+      "step": 302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.083568811416626,
+      "learning_rate": 7.53826530612245e-06,
+      "loss": 1.2263,
+      "step": 303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0389920473098755,
+      "learning_rate": 7.525510204081633e-06,
+      "loss": 1.269,
+      "step": 304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1106739044189453,
+      "learning_rate": 7.512755102040817e-06,
+      "loss": 1.2248,
+      "step": 305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0161393880844116,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.2392,
+      "step": 306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0761648416519165,
+      "learning_rate": 7.487244897959184e-06,
+      "loss": 1.2351,
+      "step": 307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.024794340133667,
+      "learning_rate": 7.474489795918368e-06,
+      "loss": 1.2514,
+      "step": 308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1029651165008545,
+      "learning_rate": 7.461734693877551e-06,
+      "loss": 1.2457,
+      "step": 309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0931360721588135,
+      "learning_rate": 7.448979591836736e-06,
+      "loss": 1.236,
+      "step": 310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.081899642944336,
+      "learning_rate": 7.436224489795919e-06,
+      "loss": 1.2381,
+      "step": 311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0580326318740845,
+      "learning_rate": 7.423469387755102e-06,
+      "loss": 1.2449,
+      "step": 312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1197216510772705,
+      "learning_rate": 7.410714285714287e-06,
+      "loss": 1.2424,
+      "step": 313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0201445817947388,
+      "learning_rate": 7.39795918367347e-06,
+      "loss": 1.2405,
+      "step": 314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.017836093902588,
+      "learning_rate": 7.385204081632653e-06,
+      "loss": 1.2466,
+      "step": 315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9251910448074341,
+      "learning_rate": 7.372448979591838e-06,
+      "loss": 1.2306,
+      "step": 316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9502259492874146,
+      "learning_rate": 7.359693877551021e-06,
+      "loss": 1.2475,
+      "step": 317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9273219704627991,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 1.2385,
+      "step": 318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9324339628219604,
+      "learning_rate": 7.334183673469388e-06,
+      "loss": 1.238,
+      "step": 319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9171964526176453,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 1.2172,
+      "step": 320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9358417987823486,
+      "learning_rate": 7.308673469387756e-06,
+      "loss": 1.2216,
+      "step": 321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9297531843185425,
+      "learning_rate": 7.295918367346939e-06,
+      "loss": 1.2133,
+      "step": 322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9529639482498169,
+      "learning_rate": 7.283163265306124e-06,
+      "loss": 1.2386,
+      "step": 323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9522646069526672,
+      "learning_rate": 7.270408163265307e-06,
+      "loss": 1.2225,
+      "step": 324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9964836835861206,
+      "learning_rate": 7.25765306122449e-06,
+      "loss": 1.2449,
+      "step": 325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9964749813079834,
+      "learning_rate": 7.244897959183675e-06,
+      "loss": 1.2212,
+      "step": 326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9914069175720215,
+      "learning_rate": 7.232142857142858e-06,
+      "loss": 1.2384,
+      "step": 327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9886379837989807,
+      "learning_rate": 7.219387755102042e-06,
+      "loss": 1.2302,
+      "step": 328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0108413696289062,
+      "learning_rate": 7.206632653061226e-06,
+      "loss": 1.2077,
+      "step": 329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0429081916809082,
+      "learning_rate": 7.193877551020409e-06,
+      "loss": 1.2305,
+      "step": 330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.06741464138031,
+      "learning_rate": 7.181122448979593e-06,
+      "loss": 1.2352,
+      "step": 331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0488401651382446,
+      "learning_rate": 7.168367346938776e-06,
+      "loss": 1.2199,
+      "step": 332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1095186471939087,
+      "learning_rate": 7.15561224489796e-06,
+      "loss": 1.2296,
+      "step": 333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.033311367034912,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.2062,
+      "step": 334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0777385234832764,
+      "learning_rate": 7.130102040816327e-06,
+      "loss": 1.2072,
+      "step": 335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0967751741409302,
+      "learning_rate": 7.1173469387755114e-06,
+      "loss": 1.2263,
+      "step": 336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.134886622428894,
+      "learning_rate": 7.1045918367346945e-06,
+      "loss": 1.2275,
+      "step": 337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0573385953903198,
+      "learning_rate": 7.091836734693878e-06,
+      "loss": 1.2284,
+      "step": 338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9954231977462769,
+      "learning_rate": 7.079081632653062e-06,
+      "loss": 1.218,
+      "step": 339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9254037737846375,
+      "learning_rate": 7.0663265306122455e-06,
+      "loss": 1.2101,
+      "step": 340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9205686450004578,
+      "learning_rate": 7.053571428571429e-06,
+      "loss": 1.2323,
+      "step": 341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9159219264984131,
+      "learning_rate": 7.0408163265306125e-06,
+      "loss": 1.2403,
+      "step": 342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9511287212371826,
+      "learning_rate": 7.0280612244897964e-06,
+      "loss": 1.2316,
+      "step": 343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0230053663253784,
+      "learning_rate": 7.01530612244898e-06,
+      "loss": 1.2331,
+      "step": 344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.996154248714447,
+      "learning_rate": 7.0025510204081634e-06,
+      "loss": 1.2115,
+      "step": 345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9797936677932739,
+      "learning_rate": 6.989795918367348e-06,
+      "loss": 1.2189,
+      "step": 346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0028291940689087,
+      "learning_rate": 6.977040816326531e-06,
+      "loss": 1.2397,
+      "step": 347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0954959392547607,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 1.218,
+      "step": 348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0602993965148926,
+      "learning_rate": 6.951530612244899e-06,
+      "loss": 1.2181,
+      "step": 349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0718706846237183,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.2171,
+      "step": 350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0047320127487183,
+      "learning_rate": 6.926020408163265e-06,
+      "loss": 1.2092,
+      "step": 351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9684332609176636,
+      "learning_rate": 6.91326530612245e-06,
+      "loss": 1.2095,
+      "step": 352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8635539412498474,
+      "learning_rate": 6.900510204081633e-06,
+      "loss": 1.2299,
+      "step": 353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9231078028678894,
+      "learning_rate": 6.887755102040817e-06,
+      "loss": 1.2219,
+      "step": 354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8896716833114624,
+      "learning_rate": 6.875e-06,
+      "loss": 1.2223,
+      "step": 355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.922714114189148,
+      "learning_rate": 6.862244897959184e-06,
+      "loss": 1.1918,
+      "step": 356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9075490832328796,
+      "learning_rate": 6.849489795918368e-06,
+      "loss": 1.2137,
+      "step": 357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9079575538635254,
+      "learning_rate": 6.836734693877551e-06,
+      "loss": 1.1861,
+      "step": 358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8482446670532227,
+      "learning_rate": 6.823979591836736e-06,
+      "loss": 1.2168,
+      "step": 359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.856937050819397,
+      "learning_rate": 6.811224489795919e-06,
+      "loss": 1.1923,
+      "step": 360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8377067446708679,
+      "learning_rate": 6.798469387755102e-06,
+      "loss": 1.2248,
+      "step": 361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8343848586082458,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 1.1971,
+      "step": 362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8501905798912048,
+      "learning_rate": 6.77295918367347e-06,
+      "loss": 1.2159,
+      "step": 363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9495564103126526,
+      "learning_rate": 6.760204081632653e-06,
+      "loss": 1.21,
+      "step": 364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9224797487258911,
+      "learning_rate": 6.747448979591837e-06,
+      "loss": 1.219,
+      "step": 365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9395394325256348,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 1.1939,
+      "step": 366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9465500712394714,
+      "learning_rate": 6.721938775510205e-06,
+      "loss": 1.2176,
+      "step": 367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9452735185623169,
+      "learning_rate": 6.709183673469388e-06,
+      "loss": 1.2025,
+      "step": 368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9032958745956421,
+      "learning_rate": 6.696428571428571e-06,
+      "loss": 1.2151,
+      "step": 369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9356412291526794,
+      "learning_rate": 6.683673469387756e-06,
+      "loss": 1.1981,
+      "step": 370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9610426425933838,
+      "learning_rate": 6.670918367346939e-06,
+      "loss": 1.2212,
+      "step": 371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0291876792907715,
+      "learning_rate": 6.658163265306124e-06,
+      "loss": 1.2206,
+      "step": 372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0189391374588013,
+      "learning_rate": 6.645408163265307e-06,
+      "loss": 1.1959,
+      "step": 373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9816426038742065,
+      "learning_rate": 6.63265306122449e-06,
+      "loss": 1.2019,
+      "step": 374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9211484789848328,
+      "learning_rate": 6.619897959183675e-06,
+      "loss": 1.1775,
+      "step": 375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.944790244102478,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 1.1766,
+      "step": 376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9401800036430359,
+      "learning_rate": 6.594387755102042e-06,
+      "loss": 1.2087,
+      "step": 377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9268150925636292,
+      "learning_rate": 6.581632653061225e-06,
+      "loss": 1.196,
+      "step": 378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9066458940505981,
+      "learning_rate": 6.568877551020409e-06,
+      "loss": 1.1832,
+      "step": 379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9003369212150574,
+      "learning_rate": 6.556122448979593e-06,
+      "loss": 1.2205,
+      "step": 380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.842934250831604,
+      "learning_rate": 6.543367346938776e-06,
+      "loss": 1.1967,
+      "step": 381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8342365622520447,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 1.1894,
+      "step": 382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9125496745109558,
+      "learning_rate": 6.517857142857144e-06,
+      "loss": 1.2019,
+      "step": 383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9107596278190613,
+      "learning_rate": 6.505102040816327e-06,
+      "loss": 1.2032,
+      "step": 384
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.951339066028595,
+      "learning_rate": 6.4923469387755115e-06,
+      "loss": 1.1849,
+      "step": 385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9225314259529114,
+      "learning_rate": 6.4795918367346946e-06,
+      "loss": 1.2101,
+      "step": 386
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9681543707847595,
+      "learning_rate": 6.466836734693878e-06,
+      "loss": 1.2233,
+      "step": 387
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9564622640609741,
+      "learning_rate": 6.454081632653062e-06,
+      "loss": 1.2006,
+      "step": 388
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9844996333122253,
+      "learning_rate": 6.4413265306122455e-06,
+      "loss": 1.2073,
+      "step": 389
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9823477268218994,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 1.2221,
+      "step": 390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9903162121772766,
+      "learning_rate": 6.4158163265306125e-06,
+      "loss": 1.1994,
+      "step": 391
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8821483850479126,
+      "learning_rate": 6.403061224489796e-06,
+      "loss": 1.1885,
+      "step": 392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9244970679283142,
+      "learning_rate": 6.39030612244898e-06,
+      "loss": 1.2173,
+      "step": 393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8394110798835754,
+      "learning_rate": 6.3775510204081635e-06,
+      "loss": 1.1996,
+      "step": 394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9020444750785828,
+      "learning_rate": 6.364795918367348e-06,
+      "loss": 1.1937,
+      "step": 395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8169025778770447,
+      "learning_rate": 6.352040816326531e-06,
+      "loss": 1.1994,
+      "step": 396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8738958239555359,
+      "learning_rate": 6.3392857142857145e-06,
+      "loss": 1.1929,
+      "step": 397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8744160532951355,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 1.183,
+      "step": 398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.819991946220398,
+      "learning_rate": 6.313775510204082e-06,
+      "loss": 1.1602,
+      "step": 399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.778607964515686,
+      "learning_rate": 6.301020408163265e-06,
+      "loss": 1.1865,
+      "step": 400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8436089158058167,
+      "learning_rate": 6.288265306122449e-06,
+      "loss": 1.194,
+      "step": 401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8169540166854858,
+      "learning_rate": 6.275510204081633e-06,
+      "loss": 1.193,
+      "step": 402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8512654304504395,
+      "learning_rate": 6.262755102040817e-06,
+      "loss": 1.2183,
+      "step": 403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8186187148094177,
+      "learning_rate": 6.25e-06,
+      "loss": 1.1844,
+      "step": 404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.801342248916626,
+      "learning_rate": 6.237244897959183e-06,
+      "loss": 1.1791,
+      "step": 405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8259813785552979,
+      "learning_rate": 6.224489795918368e-06,
+      "loss": 1.2031,
+      "step": 406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8373045325279236,
+      "learning_rate": 6.211734693877551e-06,
+      "loss": 1.1881,
+      "step": 407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.819210410118103,
+      "learning_rate": 6.198979591836736e-06,
+      "loss": 1.1898,
+      "step": 408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.824668824672699,
+      "learning_rate": 6.186224489795919e-06,
+      "loss": 1.1845,
+      "step": 409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8842679858207703,
+      "learning_rate": 6.173469387755102e-06,
+      "loss": 1.2051,
+      "step": 410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.798000156879425,
+      "learning_rate": 6.160714285714286e-06,
+      "loss": 1.1926,
+      "step": 411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7511589527130127,
+      "learning_rate": 6.14795918367347e-06,
+      "loss": 1.1949,
+      "step": 412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7895238399505615,
+      "learning_rate": 6.135204081632653e-06,
+      "loss": 1.1733,
+      "step": 413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8007684946060181,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 1.1655,
+      "step": 414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7976680994033813,
+      "learning_rate": 6.109693877551021e-06,
+      "loss": 1.198,
+      "step": 415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7595086097717285,
+      "learning_rate": 6.096938775510205e-06,
+      "loss": 1.1914,
+      "step": 416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8117203116416931,
+      "learning_rate": 6.084183673469388e-06,
+      "loss": 1.183,
+      "step": 417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7925319075584412,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 1.1591,
+      "step": 418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9027591347694397,
+      "learning_rate": 6.058673469387756e-06,
+      "loss": 1.1888,
+      "step": 419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7940754294395447,
+      "learning_rate": 6.045918367346939e-06,
+      "loss": 1.1827,
+      "step": 420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7836580872535706,
+      "learning_rate": 6.033163265306124e-06,
+      "loss": 1.1835,
+      "step": 421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7472612261772156,
+      "learning_rate": 6.020408163265307e-06,
+      "loss": 1.1847,
+      "step": 422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7687056660652161,
+      "learning_rate": 6.00765306122449e-06,
+      "loss": 1.1722,
+      "step": 423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7652135491371155,
+      "learning_rate": 5.994897959183674e-06,
+      "loss": 1.1831,
+      "step": 424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7814399600028992,
+      "learning_rate": 5.982142857142858e-06,
+      "loss": 1.1723,
+      "step": 425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7041035890579224,
+      "learning_rate": 5.969387755102042e-06,
+      "loss": 1.1923,
+      "step": 426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7865569591522217,
+      "learning_rate": 5.956632653061225e-06,
+      "loss": 1.1857,
+      "step": 427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8004493117332458,
+      "learning_rate": 5.943877551020408e-06,
+      "loss": 1.1754,
+      "step": 428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8099337220191956,
+      "learning_rate": 5.931122448979593e-06,
+      "loss": 1.1568,
+      "step": 429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7906381487846375,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 1.1888,
+      "step": 430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8648595213890076,
+      "learning_rate": 5.905612244897959e-06,
+      "loss": 1.1801,
+      "step": 431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8521154522895813,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 1.1745,
+      "step": 432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8795692324638367,
+      "learning_rate": 5.880102040816327e-06,
+      "loss": 1.1898,
+      "step": 433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8522396087646484,
+      "learning_rate": 5.867346938775511e-06,
+      "loss": 1.2022,
+      "step": 434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8316823244094849,
+      "learning_rate": 5.854591836734695e-06,
+      "loss": 1.1899,
+      "step": 435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8392115235328674,
+      "learning_rate": 5.841836734693878e-06,
+      "loss": 1.1811,
+      "step": 436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8328803181648254,
+      "learning_rate": 5.829081632653062e-06,
+      "loss": 1.1769,
+      "step": 437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7763929963111877,
+      "learning_rate": 5.816326530612246e-06,
+      "loss": 1.1663,
+      "step": 438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7515134215354919,
+      "learning_rate": 5.8035714285714295e-06,
+      "loss": 1.1749,
+      "step": 439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7351372241973877,
+      "learning_rate": 5.790816326530613e-06,
+      "loss": 1.1696,
+      "step": 440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7608625292778015,
+      "learning_rate": 5.778061224489796e-06,
+      "loss": 1.1893,
+      "step": 441
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7726980447769165,
+      "learning_rate": 5.7653061224489805e-06,
+      "loss": 1.1992,
+      "step": 442
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.783970057964325,
+      "learning_rate": 5.7525510204081636e-06,
+      "loss": 1.1791,
+      "step": 443
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7531319856643677,
+      "learning_rate": 5.739795918367348e-06,
+      "loss": 1.183,
+      "step": 444
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7932224869728088,
+      "learning_rate": 5.7270408163265314e-06,
+      "loss": 1.196,
+      "step": 445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7444440126419067,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.1502,
+      "step": 446
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7612382769584656,
+      "learning_rate": 5.7015306122448984e-06,
+      "loss": 1.1523,
+      "step": 447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7724157571792603,
+      "learning_rate": 5.688775510204082e-06,
+      "loss": 1.1717,
+      "step": 448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7933560013771057,
+      "learning_rate": 5.6760204081632655e-06,
+      "loss": 1.1802,
+      "step": 449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.799335777759552,
+      "learning_rate": 5.663265306122449e-06,
+      "loss": 1.1609,
+      "step": 450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.785335123538971,
+      "learning_rate": 5.6505102040816325e-06,
+      "loss": 1.1473,
+      "step": 451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.816709041595459,
+      "learning_rate": 5.637755102040817e-06,
+      "loss": 1.1681,
+      "step": 452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7700944542884827,
+      "learning_rate": 5.625e-06,
+      "loss": 1.1723,
+      "step": 453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8849509954452515,
+      "learning_rate": 5.6122448979591834e-06,
+      "loss": 1.1664,
+      "step": 454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8110070824623108,
+      "learning_rate": 5.599489795918368e-06,
+      "loss": 1.1668,
+      "step": 455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7989910244941711,
+      "learning_rate": 5.586734693877551e-06,
+      "loss": 1.1539,
+      "step": 456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7733786702156067,
+      "learning_rate": 5.573979591836735e-06,
+      "loss": 1.1631,
+      "step": 457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7904419302940369,
+      "learning_rate": 5.561224489795919e-06,
+      "loss": 1.1795,
+      "step": 458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7549111843109131,
+      "learning_rate": 5.548469387755102e-06,
+      "loss": 1.1718,
+      "step": 459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7582657337188721,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 1.1506,
+      "step": 460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7553848028182983,
+      "learning_rate": 5.52295918367347e-06,
+      "loss": 1.1445,
+      "step": 461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7763128280639648,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 1.1704,
+      "step": 462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7822700142860413,
+      "learning_rate": 5.497448979591837e-06,
+      "loss": 1.1687,
+      "step": 463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.809799313545227,
+      "learning_rate": 5.48469387755102e-06,
+      "loss": 1.144,
+      "step": 464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8360884189605713,
+      "learning_rate": 5.471938775510205e-06,
+      "loss": 1.1737,
+      "step": 465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8203660249710083,
+      "learning_rate": 5.459183673469388e-06,
+      "loss": 1.1602,
+      "step": 466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.841998279094696,
+      "learning_rate": 5.446428571428571e-06,
+      "loss": 1.1648,
+      "step": 467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7999197244644165,
+      "learning_rate": 5.433673469387756e-06,
+      "loss": 1.1757,
+      "step": 468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7686060070991516,
+      "learning_rate": 5.420918367346939e-06,
+      "loss": 1.1717,
+      "step": 469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7240757346153259,
+      "learning_rate": 5.408163265306123e-06,
+      "loss": 1.1621,
+      "step": 470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7930902242660522,
+      "learning_rate": 5.395408163265307e-06,
+      "loss": 1.1895,
+      "step": 471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7647054195404053,
+      "learning_rate": 5.38265306122449e-06,
+      "loss": 1.177,
+      "step": 472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7429608106613159,
+      "learning_rate": 5.369897959183674e-06,
+      "loss": 1.154,
+      "step": 473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7082961201667786,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 1.155,
+      "step": 474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7647345066070557,
+      "learning_rate": 5.344387755102042e-06,
+      "loss": 1.1478,
+      "step": 475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7455657124519348,
+      "learning_rate": 5.331632653061225e-06,
+      "loss": 1.1308,
+      "step": 476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7820383906364441,
+      "learning_rate": 5.318877551020408e-06,
+      "loss": 1.1657,
+      "step": 477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7322938442230225,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 1.1629,
+      "step": 478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7732429504394531,
+      "learning_rate": 5.293367346938776e-06,
+      "loss": 1.1714,
+      "step": 479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7132137417793274,
+      "learning_rate": 5.280612244897959e-06,
+      "loss": 1.1896,
+      "step": 480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7276777625083923,
+      "learning_rate": 5.267857142857144e-06,
+      "loss": 1.1426,
+      "step": 481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7646544575691223,
+      "learning_rate": 5.255102040816327e-06,
+      "loss": 1.1508,
+      "step": 482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7866250276565552,
+      "learning_rate": 5.242346938775511e-06,
+      "loss": 1.1376,
+      "step": 483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8272651433944702,
+      "learning_rate": 5.229591836734695e-06,
+      "loss": 1.1431,
+      "step": 484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7927102446556091,
+      "learning_rate": 5.216836734693878e-06,
+      "loss": 1.1481,
+      "step": 485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7916460037231445,
+      "learning_rate": 5.204081632653062e-06,
+      "loss": 1.1256,
+      "step": 486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7587056159973145,
+      "learning_rate": 5.191326530612245e-06,
+      "loss": 1.1555,
+      "step": 487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7142450213432312,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 1.1613,
+      "step": 488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7216023802757263,
+      "learning_rate": 5.165816326530613e-06,
+      "loss": 1.1603,
+      "step": 489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7468694448471069,
+      "learning_rate": 5.153061224489796e-06,
+      "loss": 1.1386,
+      "step": 490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7479641437530518,
+      "learning_rate": 5.1403061224489805e-06,
+      "loss": 1.1295,
+      "step": 491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7640284299850464,
+      "learning_rate": 5.127551020408164e-06,
+      "loss": 1.152,
+      "step": 492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7198285460472107,
+      "learning_rate": 5.1147959183673475e-06,
+      "loss": 1.137,
+      "step": 493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7455816864967346,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 1.144,
+      "step": 494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7048183679580688,
+      "learning_rate": 5.0892857142857146e-06,
+      "loss": 1.1534,
+      "step": 495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7082626223564148,
+      "learning_rate": 5.0765306122448985e-06,
+      "loss": 1.1539,
+      "step": 496
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7038515210151672,
+      "learning_rate": 5.063775510204082e-06,
+      "loss": 1.1679,
+      "step": 497
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7405194640159607,
+      "learning_rate": 5.0510204081632655e-06,
+      "loss": 1.1375,
+      "step": 498
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.67337566614151,
+      "learning_rate": 5.0382653061224495e-06,
+      "loss": 1.1403,
+      "step": 499
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6937934756278992,
+      "learning_rate": 5.0255102040816325e-06,
+      "loss": 1.1401,
+      "step": 500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6861513257026672,
+      "learning_rate": 5.012755102040817e-06,
+      "loss": 1.174,
+      "step": 501
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7102193236351013,
+      "learning_rate": 5e-06,
+      "loss": 1.1388,
+      "step": 502
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7060781121253967,
+      "learning_rate": 4.987244897959184e-06,
+      "loss": 1.1485,
+      "step": 503
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7078152894973755,
+      "learning_rate": 4.974489795918368e-06,
+      "loss": 1.1599,
+      "step": 504
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6945005655288696,
+      "learning_rate": 4.961734693877551e-06,
+      "loss": 1.1372,
+      "step": 505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6945128440856934,
+      "learning_rate": 4.948979591836735e-06,
+      "loss": 1.1511,
+      "step": 506
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6711966395378113,
+      "learning_rate": 4.936224489795919e-06,
+      "loss": 1.1296,
+      "step": 507
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6911993622779846,
+      "learning_rate": 4.923469387755102e-06,
+      "loss": 1.1263,
+      "step": 508
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6859057545661926,
+      "learning_rate": 4.910714285714286e-06,
+      "loss": 1.1611,
+      "step": 509
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6902860403060913,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 1.1328,
+      "step": 510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6645172834396362,
+      "learning_rate": 4.885204081632653e-06,
+      "loss": 1.1235,
+      "step": 511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6861415505409241,
+      "learning_rate": 4.872448979591837e-06,
+      "loss": 1.1472,
+      "step": 512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6925361156463623,
+      "learning_rate": 4.859693877551021e-06,
+      "loss": 1.1594,
+      "step": 513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7075345516204834,
+      "learning_rate": 4.846938775510204e-06,
+      "loss": 1.1422,
+      "step": 514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6742618680000305,
+      "learning_rate": 4.834183673469388e-06,
+      "loss": 1.1483,
+      "step": 515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6796462535858154,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 1.1222,
+      "step": 516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6587576866149902,
+      "learning_rate": 4.808673469387756e-06,
+      "loss": 1.1366,
+      "step": 517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6835523843765259,
+      "learning_rate": 4.795918367346939e-06,
+      "loss": 1.1485,
+      "step": 518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7211734056472778,
+      "learning_rate": 4.783163265306123e-06,
+      "loss": 1.1432,
+      "step": 519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.745185136795044,
+      "learning_rate": 4.770408163265306e-06,
+      "loss": 1.1461,
+      "step": 520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7311437129974365,
+      "learning_rate": 4.75765306122449e-06,
+      "loss": 1.1465,
+      "step": 521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6965100169181824,
+      "learning_rate": 4.744897959183674e-06,
+      "loss": 1.1564,
+      "step": 522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6883571147918701,
+      "learning_rate": 4.732142857142857e-06,
+      "loss": 1.1296,
+      "step": 523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6782147884368896,
+      "learning_rate": 4.719387755102041e-06,
+      "loss": 1.1427,
+      "step": 524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7144444584846497,
+      "learning_rate": 4.706632653061225e-06,
+      "loss": 1.1283,
+      "step": 525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6718636155128479,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 1.1198,
+      "step": 526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6480969786643982,
+      "learning_rate": 4.681122448979592e-06,
+      "loss": 1.1441,
+      "step": 527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6612988114356995,
+      "learning_rate": 4.668367346938776e-06,
+      "loss": 1.1525,
+      "step": 528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6701130270957947,
+      "learning_rate": 4.65561224489796e-06,
+      "loss": 1.122,
+      "step": 529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6586853265762329,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 1.1153,
+      "step": 530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6665288209915161,
+      "learning_rate": 4.630102040816327e-06,
+      "loss": 1.1443,
+      "step": 531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7110608816146851,
+      "learning_rate": 4.617346938775511e-06,
+      "loss": 1.1449,
+      "step": 532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6855413913726807,
+      "learning_rate": 4.604591836734694e-06,
+      "loss": 1.1427,
+      "step": 533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6243303418159485,
+      "learning_rate": 4.591836734693878e-06,
+      "loss": 1.146,
+      "step": 534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6160845756530762,
+      "learning_rate": 4.579081632653062e-06,
+      "loss": 1.1185,
+      "step": 535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.649599552154541,
+      "learning_rate": 4.566326530612245e-06,
+      "loss": 1.1402,
+      "step": 536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6832114458084106,
+      "learning_rate": 4.553571428571429e-06,
+      "loss": 1.1431,
+      "step": 537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6842077970504761,
+      "learning_rate": 4.540816326530613e-06,
+      "loss": 1.1403,
+      "step": 538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6485678553581238,
+      "learning_rate": 4.528061224489797e-06,
+      "loss": 1.1349,
+      "step": 539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6563675403594971,
+      "learning_rate": 4.51530612244898e-06,
+      "loss": 1.1445,
+      "step": 540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6893595457077026,
+      "learning_rate": 4.502551020408164e-06,
+      "loss": 1.1303,
+      "step": 541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6853702068328857,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.1266,
+      "step": 542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6498103141784668,
+      "learning_rate": 4.477040816326531e-06,
+      "loss": 1.1461,
+      "step": 543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6687408089637756,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 1.1364,
+      "step": 544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6929579377174377,
+      "learning_rate": 4.451530612244898e-06,
+      "loss": 1.1244,
+      "step": 545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6577277183532715,
+      "learning_rate": 4.438775510204082e-06,
+      "loss": 1.1177,
+      "step": 546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6540212035179138,
+      "learning_rate": 4.4260204081632656e-06,
+      "loss": 1.1296,
+      "step": 547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6687043309211731,
+      "learning_rate": 4.4132653061224495e-06,
+      "loss": 1.1158,
+      "step": 548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6880722641944885,
+      "learning_rate": 4.400510204081633e-06,
+      "loss": 1.1335,
+      "step": 549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6755310893058777,
+      "learning_rate": 4.3877551020408165e-06,
+      "loss": 1.1346,
+      "step": 550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6623722910881042,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 1.1401,
+      "step": 551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6454164981842041,
+      "learning_rate": 4.362244897959184e-06,
+      "loss": 1.1204,
+      "step": 552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6594773530960083,
+      "learning_rate": 4.349489795918368e-06,
+      "loss": 1.126,
+      "step": 553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6405773162841797,
+      "learning_rate": 4.336734693877551e-06,
+      "loss": 1.1312,
+      "step": 554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6566120982170105,
+      "learning_rate": 4.323979591836735e-06,
+      "loss": 1.1282,
+      "step": 555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6293421387672424,
+      "learning_rate": 4.3112244897959184e-06,
+      "loss": 1.1294,
+      "step": 556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6095276474952698,
+      "learning_rate": 4.298469387755102e-06,
+      "loss": 1.1392,
+      "step": 557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6369320750236511,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.1396,
+      "step": 558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6458733677864075,
+      "learning_rate": 4.272959183673469e-06,
+      "loss": 1.121,
+      "step": 559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6436740159988403,
+      "learning_rate": 4.260204081632653e-06,
+      "loss": 1.1105,
+      "step": 560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.65425705909729,
+      "learning_rate": 4.247448979591837e-06,
+      "loss": 1.0976,
+      "step": 561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6265342831611633,
+      "learning_rate": 4.234693877551021e-06,
+      "loss": 1.1275,
+      "step": 562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5955907702445984,
+      "learning_rate": 4.221938775510204e-06,
+      "loss": 1.119,
+      "step": 563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6000078320503235,
+      "learning_rate": 4.209183673469388e-06,
+      "loss": 1.1018,
+      "step": 564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6637157201766968,
+      "learning_rate": 4.196428571428572e-06,
+      "loss": 1.1459,
+      "step": 565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6360584497451782,
+      "learning_rate": 4.183673469387755e-06,
+      "loss": 1.1192,
+      "step": 566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7045226693153381,
+      "learning_rate": 4.170918367346939e-06,
+      "loss": 1.1258,
+      "step": 567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6479504108428955,
+      "learning_rate": 4.158163265306123e-06,
+      "loss": 1.131,
+      "step": 568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6127601861953735,
+      "learning_rate": 4.145408163265306e-06,
+      "loss": 1.1197,
+      "step": 569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6368497610092163,
+      "learning_rate": 4.13265306122449e-06,
+      "loss": 1.1236,
+      "step": 570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6386613249778748,
+      "learning_rate": 4.119897959183674e-06,
+      "loss": 1.1188,
+      "step": 571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.587577760219574,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 1.1234,
+      "step": 572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5672041177749634,
+      "learning_rate": 4.094387755102041e-06,
+      "loss": 1.1357,
+      "step": 573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5886520743370056,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 1.101,
+      "step": 574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.611720085144043,
+      "learning_rate": 4.068877551020409e-06,
+      "loss": 1.118,
+      "step": 575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6223128437995911,
+      "learning_rate": 4.056122448979592e-06,
+      "loss": 1.1232,
+      "step": 576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.620149552822113,
+      "learning_rate": 4.043367346938776e-06,
+      "loss": 1.1163,
+      "step": 577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6052107214927673,
+      "learning_rate": 4.03061224489796e-06,
+      "loss": 1.1252,
+      "step": 578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6238659024238586,
+      "learning_rate": 4.017857142857143e-06,
+      "loss": 1.1146,
+      "step": 579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6172593235969543,
+      "learning_rate": 4.005102040816327e-06,
+      "loss": 1.1265,
+      "step": 580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6135565042495728,
+      "learning_rate": 3.99234693877551e-06,
+      "loss": 1.108,
+      "step": 581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5971201062202454,
+      "learning_rate": 3.979591836734694e-06,
+      "loss": 1.1218,
+      "step": 582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6215802431106567,
+      "learning_rate": 3.966836734693878e-06,
+      "loss": 1.1328,
+      "step": 583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6160227060317993,
+      "learning_rate": 3.954081632653062e-06,
+      "loss": 1.1146,
+      "step": 584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.586821973323822,
+      "learning_rate": 3.941326530612245e-06,
+      "loss": 1.0882,
+      "step": 585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5976683497428894,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 1.1128,
+      "step": 586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6293664574623108,
+      "learning_rate": 3.915816326530613e-06,
+      "loss": 1.0857,
+      "step": 587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.598693311214447,
+      "learning_rate": 3.903061224489797e-06,
+      "loss": 1.121,
+      "step": 588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6131600141525269,
+      "learning_rate": 3.89030612244898e-06,
+      "loss": 1.1177,
+      "step": 589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5944992899894714,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 1.1065,
+      "step": 590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.618480920791626,
+      "learning_rate": 3.864795918367348e-06,
+      "loss": 1.1004,
+      "step": 591
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.62984699010849,
+      "learning_rate": 3.852040816326531e-06,
+      "loss": 1.1242,
+      "step": 592
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6211289167404175,
+      "learning_rate": 3.839285714285715e-06,
+      "loss": 1.1078,
+      "step": 593
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6085867285728455,
+      "learning_rate": 3.826530612244898e-06,
+      "loss": 1.0988,
+      "step": 594
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.606937050819397,
+      "learning_rate": 3.8137755102040817e-06,
+      "loss": 1.1233,
+      "step": 595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6092613935470581,
+      "learning_rate": 3.8010204081632656e-06,
+      "loss": 1.1143,
+      "step": 596
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5972644686698914,
+      "learning_rate": 3.7882653061224496e-06,
+      "loss": 1.1287,
+      "step": 597
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6045836210250854,
+      "learning_rate": 3.7755102040816327e-06,
+      "loss": 1.1208,
+      "step": 598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.598455548286438,
+      "learning_rate": 3.7627551020408166e-06,
+      "loss": 1.1156,
+      "step": 599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6008214950561523,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.1008,
+      "step": 600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6205429434776306,
+      "learning_rate": 3.737244897959184e-06,
+      "loss": 1.0959,
+      "step": 601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5987078547477722,
+      "learning_rate": 3.724489795918368e-06,
+      "loss": 1.1104,
+      "step": 602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6508322358131409,
+      "learning_rate": 3.711734693877551e-06,
+      "loss": 1.1084,
+      "step": 603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6050755381584167,
+      "learning_rate": 3.698979591836735e-06,
+      "loss": 1.0987,
+      "step": 604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6434103846549988,
+      "learning_rate": 3.686224489795919e-06,
+      "loss": 1.096,
+      "step": 605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5728334188461304,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 1.1214,
+      "step": 606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.588948130607605,
+      "learning_rate": 3.660714285714286e-06,
+      "loss": 1.1125,
+      "step": 607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6084712743759155,
+      "learning_rate": 3.6479591836734694e-06,
+      "loss": 1.1193,
+      "step": 608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6148542165756226,
+      "learning_rate": 3.6352040816326534e-06,
+      "loss": 1.0953,
+      "step": 609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5736531615257263,
+      "learning_rate": 3.6224489795918373e-06,
+      "loss": 1.097,
+      "step": 610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5873562693595886,
+      "learning_rate": 3.609693877551021e-06,
+      "loss": 1.1221,
+      "step": 611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6174765229225159,
+      "learning_rate": 3.5969387755102043e-06,
+      "loss": 1.1282,
+      "step": 612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6031166911125183,
+      "learning_rate": 3.584183673469388e-06,
+      "loss": 1.1025,
+      "step": 613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6206244230270386,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 1.1228,
+      "step": 614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5885174870491028,
+      "learning_rate": 3.5586734693877557e-06,
+      "loss": 1.1154,
+      "step": 615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5698478817939758,
+      "learning_rate": 3.545918367346939e-06,
+      "loss": 1.1015,
+      "step": 616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5861687064170837,
+      "learning_rate": 3.5331632653061227e-06,
+      "loss": 1.1024,
+      "step": 617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5565373301506042,
+      "learning_rate": 3.5204081632653062e-06,
+      "loss": 1.0974,
+      "step": 618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5508489012718201,
+      "learning_rate": 3.50765306122449e-06,
+      "loss": 1.1067,
+      "step": 619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5698209404945374,
+      "learning_rate": 3.494897959183674e-06,
+      "loss": 1.1008,
+      "step": 620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5689241290092468,
+      "learning_rate": 3.482142857142857e-06,
+      "loss": 1.1046,
+      "step": 621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5628039836883545,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 1.1155,
+      "step": 622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5791564583778381,
+      "learning_rate": 3.456632653061225e-06,
+      "loss": 1.1209,
+      "step": 623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5574327707290649,
+      "learning_rate": 3.4438775510204086e-06,
+      "loss": 1.1108,
+      "step": 624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5694270730018616,
+      "learning_rate": 3.431122448979592e-06,
+      "loss": 1.1218,
+      "step": 625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5663242936134338,
+      "learning_rate": 3.4183673469387756e-06,
+      "loss": 1.1257,
+      "step": 626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5686165690422058,
+      "learning_rate": 3.4056122448979595e-06,
+      "loss": 1.105,
+      "step": 627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5621039271354675,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 1.0964,
+      "step": 628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.563789427280426,
+      "learning_rate": 3.3801020408163266e-06,
+      "loss": 1.1057,
+      "step": 629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5354847311973572,
+      "learning_rate": 3.3673469387755105e-06,
+      "loss": 1.0998,
+      "step": 630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5717220306396484,
+      "learning_rate": 3.354591836734694e-06,
+      "loss": 1.1188,
+      "step": 631
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6077321171760559,
+      "learning_rate": 3.341836734693878e-06,
+      "loss": 1.0788,
+      "step": 632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5820598006248474,
+      "learning_rate": 3.329081632653062e-06,
+      "loss": 1.1001,
+      "step": 633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5642079710960388,
+      "learning_rate": 3.316326530612245e-06,
+      "loss": 1.1198,
+      "step": 634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.555977463722229,
+      "learning_rate": 3.303571428571429e-06,
+      "loss": 1.1024,
+      "step": 635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5402265787124634,
+      "learning_rate": 3.2908163265306124e-06,
+      "loss": 1.105,
+      "step": 636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.566189169883728,
+      "learning_rate": 3.2780612244897963e-06,
+      "loss": 1.1138,
+      "step": 637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5901975035667419,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 1.0954,
+      "step": 638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5786535739898682,
+      "learning_rate": 3.2525510204081634e-06,
+      "loss": 1.1033,
+      "step": 639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5403233170509338,
+      "learning_rate": 3.2397959183673473e-06,
+      "loss": 1.1109,
+      "step": 640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5486624836921692,
+      "learning_rate": 3.227040816326531e-06,
+      "loss": 1.1187,
+      "step": 641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5421629548072815,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 1.1036,
+      "step": 642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5433635115623474,
+      "learning_rate": 3.201530612244898e-06,
+      "loss": 1.1052,
+      "step": 643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5454135537147522,
+      "learning_rate": 3.1887755102040818e-06,
+      "loss": 1.1061,
+      "step": 644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5338699221611023,
+      "learning_rate": 3.1760204081632657e-06,
+      "loss": 1.1037,
+      "step": 645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5373792052268982,
+      "learning_rate": 3.1632653061224496e-06,
+      "loss": 1.119,
+      "step": 646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5319128036499023,
+      "learning_rate": 3.1505102040816327e-06,
+      "loss": 1.1083,
+      "step": 647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5329867005348206,
+      "learning_rate": 3.1377551020408166e-06,
+      "loss": 1.0911,
+      "step": 648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.539853572845459,
+      "learning_rate": 3.125e-06,
+      "loss": 1.1104,
+      "step": 649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.555448591709137,
+      "learning_rate": 3.112244897959184e-06,
+      "loss": 1.1012,
+      "step": 650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5585979223251343,
+      "learning_rate": 3.099489795918368e-06,
+      "loss": 1.1,
+      "step": 651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5565794110298157,
+      "learning_rate": 3.086734693877551e-06,
+      "loss": 1.0868,
+      "step": 652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.583026111125946,
+      "learning_rate": 3.073979591836735e-06,
+      "loss": 1.1007,
+      "step": 653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5382760167121887,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 1.0872,
+      "step": 654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5512996315956116,
+      "learning_rate": 3.0484693877551025e-06,
+      "loss": 1.0912,
+      "step": 655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5412560105323792,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 1.085,
+      "step": 656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5486542582511902,
+      "learning_rate": 3.0229591836734695e-06,
+      "loss": 1.0838,
+      "step": 657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5376004576683044,
+      "learning_rate": 3.0102040816326534e-06,
+      "loss": 1.0753,
+      "step": 658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6080600023269653,
+      "learning_rate": 2.997448979591837e-06,
+      "loss": 1.1179,
+      "step": 659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56026691198349,
+      "learning_rate": 2.984693877551021e-06,
+      "loss": 1.1046,
+      "step": 660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5503661036491394,
+      "learning_rate": 2.971938775510204e-06,
+      "loss": 1.0967,
+      "step": 661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5562008023262024,
+      "learning_rate": 2.959183673469388e-06,
+      "loss": 1.1024,
+      "step": 662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5466132760047913,
+      "learning_rate": 2.946428571428572e-06,
+      "loss": 1.0958,
+      "step": 663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5139285922050476,
+      "learning_rate": 2.9336734693877553e-06,
+      "loss": 1.0891,
+      "step": 664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5482622981071472,
+      "learning_rate": 2.920918367346939e-06,
+      "loss": 1.1028,
+      "step": 665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5475634336471558,
+      "learning_rate": 2.908163265306123e-06,
+      "loss": 1.0863,
+      "step": 666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5316178798675537,
+      "learning_rate": 2.8954081632653063e-06,
+      "loss": 1.0858,
+      "step": 667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5294205546379089,
+      "learning_rate": 2.8826530612244902e-06,
+      "loss": 1.099,
+      "step": 668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5268580317497253,
+      "learning_rate": 2.869897959183674e-06,
+      "loss": 1.0836,
+      "step": 669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5232135057449341,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.091,
+      "step": 670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5184506773948669,
+      "learning_rate": 2.844387755102041e-06,
+      "loss": 1.0723,
+      "step": 671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5274559855461121,
+      "learning_rate": 2.8316326530612247e-06,
+      "loss": 1.0967,
+      "step": 672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5404233932495117,
+      "learning_rate": 2.8188775510204086e-06,
+      "loss": 1.0946,
+      "step": 673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5538030862808228,
+      "learning_rate": 2.8061224489795917e-06,
+      "loss": 1.0926,
+      "step": 674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5240524411201477,
+      "learning_rate": 2.7933673469387757e-06,
+      "loss": 1.0954,
+      "step": 675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5182491540908813,
+      "learning_rate": 2.7806122448979596e-06,
+      "loss": 1.098,
+      "step": 676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5135244727134705,
+      "learning_rate": 2.767857142857143e-06,
+      "loss": 1.0896,
+      "step": 677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.508006751537323,
+      "learning_rate": 2.7551020408163266e-06,
+      "loss": 1.1097,
+      "step": 678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5241822004318237,
+      "learning_rate": 2.74234693877551e-06,
+      "loss": 1.0893,
+      "step": 679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5380449295043945,
+      "learning_rate": 2.729591836734694e-06,
+      "loss": 1.0935,
+      "step": 680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5482093095779419,
+      "learning_rate": 2.716836734693878e-06,
+      "loss": 1.08,
+      "step": 681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5596063733100891,
+      "learning_rate": 2.7040816326530615e-06,
+      "loss": 1.098,
+      "step": 682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5343565940856934,
+      "learning_rate": 2.691326530612245e-06,
+      "loss": 1.0778,
+      "step": 683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5278717279434204,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 1.0821,
+      "step": 684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.536170482635498,
+      "learning_rate": 2.6658163265306125e-06,
+      "loss": 1.1046,
+      "step": 685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5198537707328796,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 1.1022,
+      "step": 686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.507614016532898,
+      "learning_rate": 2.6403061224489795e-06,
+      "loss": 1.082,
+      "step": 687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5223891735076904,
+      "learning_rate": 2.6275510204081634e-06,
+      "loss": 1.0692,
+      "step": 688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5129234194755554,
+      "learning_rate": 2.6147959183673473e-06,
+      "loss": 1.0866,
+      "step": 689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5176243782043457,
+      "learning_rate": 2.602040816326531e-06,
+      "loss": 1.0954,
+      "step": 690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5537042617797852,
+      "learning_rate": 2.5892857142857148e-06,
+      "loss": 1.0884,
+      "step": 691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5302846431732178,
+      "learning_rate": 2.576530612244898e-06,
+      "loss": 1.0809,
+      "step": 692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5051791071891785,
+      "learning_rate": 2.563775510204082e-06,
+      "loss": 1.0884,
+      "step": 693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5340541005134583,
+      "learning_rate": 2.5510204081632657e-06,
+      "loss": 1.0967,
+      "step": 694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5301932096481323,
+      "learning_rate": 2.5382653061224492e-06,
+      "loss": 1.1186,
+      "step": 695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5282807946205139,
+      "learning_rate": 2.5255102040816328e-06,
+      "loss": 1.0987,
+      "step": 696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49649927020072937,
+      "learning_rate": 2.5127551020408163e-06,
+      "loss": 1.0759,
+      "step": 697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5125253796577454,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0888,
+      "step": 698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.502814531326294,
+      "learning_rate": 2.487244897959184e-06,
+      "loss": 1.0712,
+      "step": 699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5090625882148743,
+      "learning_rate": 2.4744897959183676e-06,
+      "loss": 1.0822,
+      "step": 700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5162419080734253,
+      "learning_rate": 2.461734693877551e-06,
+      "loss": 1.0973,
+      "step": 701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5318601131439209,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 1.0992,
+      "step": 702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5034762024879456,
+      "learning_rate": 2.4362244897959186e-06,
+      "loss": 1.0737,
+      "step": 703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5142484903335571,
+      "learning_rate": 2.423469387755102e-06,
+      "loss": 1.0993,
+      "step": 704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5127019882202148,
+      "learning_rate": 2.410714285714286e-06,
+      "loss": 1.0901,
+      "step": 705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.526448130607605,
+      "learning_rate": 2.3979591836734696e-06,
+      "loss": 1.0787,
+      "step": 706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5332403779029846,
+      "learning_rate": 2.385204081632653e-06,
+      "loss": 1.0685,
+      "step": 707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5180906057357788,
+      "learning_rate": 2.372448979591837e-06,
+      "loss": 1.0878,
+      "step": 708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5121365785598755,
+      "learning_rate": 2.3596938775510205e-06,
+      "loss": 1.0657,
+      "step": 709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5013092756271362,
+      "learning_rate": 2.3469387755102044e-06,
+      "loss": 1.0805,
+      "step": 710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5061728358268738,
+      "learning_rate": 2.334183673469388e-06,
+      "loss": 1.0877,
+      "step": 711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5036882162094116,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 1.0864,
+      "step": 712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5100290179252625,
+      "learning_rate": 2.3086734693877554e-06,
+      "loss": 1.0826,
+      "step": 713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5077418088912964,
+      "learning_rate": 2.295918367346939e-06,
+      "loss": 1.0978,
+      "step": 714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.505440890789032,
+      "learning_rate": 2.2831632653061224e-06,
+      "loss": 1.0913,
+      "step": 715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5235510468482971,
+      "learning_rate": 2.2704081632653064e-06,
+      "loss": 1.0969,
+      "step": 716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5321602821350098,
+      "learning_rate": 2.25765306122449e-06,
+      "loss": 1.0843,
+      "step": 717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5095252394676208,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 1.0817,
+      "step": 718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5119527578353882,
+      "learning_rate": 2.2321428571428573e-06,
+      "loss": 1.0967,
+      "step": 719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5339480042457581,
+      "learning_rate": 2.219387755102041e-06,
+      "loss": 1.0872,
+      "step": 720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5127794146537781,
+      "learning_rate": 2.2066326530612248e-06,
+      "loss": 1.0788,
+      "step": 721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5119395852088928,
+      "learning_rate": 2.1938775510204083e-06,
+      "loss": 1.0591,
+      "step": 722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5011995434761047,
+      "learning_rate": 2.181122448979592e-06,
+      "loss": 1.0838,
+      "step": 723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5067063570022583,
+      "learning_rate": 2.1683673469387757e-06,
+      "loss": 1.0846,
+      "step": 724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5018548369407654,
+      "learning_rate": 2.1556122448979592e-06,
+      "loss": 1.0852,
+      "step": 725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5224974751472473,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.0722,
+      "step": 726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5073409676551819,
+      "learning_rate": 2.1301020408163267e-06,
+      "loss": 1.0983,
+      "step": 727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5096011757850647,
+      "learning_rate": 2.1173469387755106e-06,
+      "loss": 1.0708,
+      "step": 728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5168206691741943,
+      "learning_rate": 2.104591836734694e-06,
+      "loss": 1.0731,
+      "step": 729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5258650779724121,
+      "learning_rate": 2.0918367346938776e-06,
+      "loss": 1.0913,
+      "step": 730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4983409643173218,
+      "learning_rate": 2.0790816326530616e-06,
+      "loss": 1.0616,
+      "step": 731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5054186582565308,
+      "learning_rate": 2.066326530612245e-06,
+      "loss": 1.0925,
+      "step": 732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5062727928161621,
+      "learning_rate": 2.0535714285714286e-06,
+      "loss": 1.0937,
+      "step": 733
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5056551098823547,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 1.0859,
+      "step": 734
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5142880082130432,
+      "learning_rate": 2.028061224489796e-06,
+      "loss": 1.0837,
+      "step": 735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5090142488479614,
+      "learning_rate": 2.01530612244898e-06,
+      "loss": 1.0669,
+      "step": 736
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5223041772842407,
+      "learning_rate": 2.0025510204081635e-06,
+      "loss": 1.0775,
+      "step": 737
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5301938056945801,
+      "learning_rate": 1.989795918367347e-06,
+      "loss": 1.1039,
+      "step": 738
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.512309193611145,
+      "learning_rate": 1.977040816326531e-06,
+      "loss": 1.0749,
+      "step": 739
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5122722387313843,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 1.0885,
+      "step": 740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5142942070960999,
+      "learning_rate": 1.9515306122448984e-06,
+      "loss": 1.0954,
+      "step": 741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5099963545799255,
+      "learning_rate": 1.938775510204082e-06,
+      "loss": 1.0924,
+      "step": 742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5308962464332581,
+      "learning_rate": 1.9260204081632654e-06,
+      "loss": 1.075,
+      "step": 743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5025672316551208,
+      "learning_rate": 1.913265306122449e-06,
+      "loss": 1.0915,
+      "step": 744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5081654191017151,
+      "learning_rate": 1.9005102040816328e-06,
+      "loss": 1.062,
+      "step": 745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5062184929847717,
+      "learning_rate": 1.8877551020408163e-06,
+      "loss": 1.0687,
+      "step": 746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48296964168548584,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 1.063,
+      "step": 747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4898391664028168,
+      "learning_rate": 1.862244897959184e-06,
+      "loss": 1.0841,
+      "step": 748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5018566250801086,
+      "learning_rate": 1.8494897959183675e-06,
+      "loss": 1.0871,
+      "step": 749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5019533634185791,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 1.0819,
+      "step": 750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4855716824531555,
+      "learning_rate": 1.8239795918367347e-06,
+      "loss": 1.0835,
+      "step": 751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4964170753955841,
+      "learning_rate": 1.8112244897959187e-06,
+      "loss": 1.0932,
+      "step": 752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4965682029724121,
+      "learning_rate": 1.7984693877551022e-06,
+      "loss": 1.0714,
+      "step": 753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4963724911212921,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 1.0714,
+      "step": 754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49885648488998413,
+      "learning_rate": 1.7729591836734694e-06,
+      "loss": 1.0824,
+      "step": 755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49728772044181824,
+      "learning_rate": 1.7602040816326531e-06,
+      "loss": 1.0873,
+      "step": 756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5012596845626831,
+      "learning_rate": 1.747448979591837e-06,
+      "loss": 1.07,
+      "step": 757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49209412932395935,
+      "learning_rate": 1.7346938775510206e-06,
+      "loss": 1.0808,
+      "step": 758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4979517459869385,
+      "learning_rate": 1.7219387755102043e-06,
+      "loss": 1.0632,
+      "step": 759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49073830246925354,
+      "learning_rate": 1.7091836734693878e-06,
+      "loss": 1.0824,
+      "step": 760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5171460509300232,
+      "learning_rate": 1.6964285714285717e-06,
+      "loss": 1.0614,
+      "step": 761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49084386229515076,
+      "learning_rate": 1.6836734693877552e-06,
+      "loss": 1.0899,
+      "step": 762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5033416152000427,
+      "learning_rate": 1.670918367346939e-06,
+      "loss": 1.1004,
+      "step": 763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49504750967025757,
+      "learning_rate": 1.6581632653061225e-06,
+      "loss": 1.0621,
+      "step": 764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4942651391029358,
+      "learning_rate": 1.6454081632653062e-06,
+      "loss": 1.0776,
+      "step": 765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4950351417064667,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 1.0887,
+      "step": 766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5089950561523438,
+      "learning_rate": 1.6198979591836736e-06,
+      "loss": 1.0815,
+      "step": 767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5063705444335938,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 1.0835,
+      "step": 768
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4940871298313141,
+      "learning_rate": 1.5943877551020409e-06,
+      "loss": 1.0913,
+      "step": 769
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4951344430446625,
+      "learning_rate": 1.5816326530612248e-06,
+      "loss": 1.0625,
+      "step": 770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5025525093078613,
+      "learning_rate": 1.5688775510204083e-06,
+      "loss": 1.097,
+      "step": 771
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5122385025024414,
+      "learning_rate": 1.556122448979592e-06,
+      "loss": 1.0823,
+      "step": 772
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4933209717273712,
+      "learning_rate": 1.5433673469387756e-06,
+      "loss": 1.0802,
+      "step": 773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.513375461101532,
+      "learning_rate": 1.5306122448979593e-06,
+      "loss": 1.089,
+      "step": 774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5016956925392151,
+      "learning_rate": 1.5178571428571428e-06,
+      "loss": 1.0624,
+      "step": 775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.48563146591186523,
+      "learning_rate": 1.5051020408163267e-06,
+      "loss": 1.0812,
+      "step": 776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49049103260040283,
+      "learning_rate": 1.4923469387755104e-06,
+      "loss": 1.0829,
+      "step": 777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5047158002853394,
+      "learning_rate": 1.479591836734694e-06,
+      "loss": 1.0933,
+      "step": 778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4831494092941284,
+      "learning_rate": 1.4668367346938777e-06,
+      "loss": 1.0883,
+      "step": 779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4787144958972931,
+      "learning_rate": 1.4540816326530614e-06,
+      "loss": 1.0638,
+      "step": 780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4906410276889801,
+      "learning_rate": 1.4413265306122451e-06,
+      "loss": 1.0729,
+      "step": 781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48908156156539917,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.0749,
+      "step": 782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48697564005851746,
+      "learning_rate": 1.4158163265306123e-06,
+      "loss": 1.0904,
+      "step": 783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5088189840316772,
+      "learning_rate": 1.4030612244897959e-06,
+      "loss": 1.0691,
+      "step": 784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5027821063995361,
+      "learning_rate": 1.3903061224489798e-06,
+      "loss": 1.073,
+      "step": 785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4827651381492615,
+      "learning_rate": 1.3775510204081633e-06,
+      "loss": 1.0669,
+      "step": 786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5015129446983337,
+      "learning_rate": 1.364795918367347e-06,
+      "loss": 1.0739,
+      "step": 787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4779413640499115,
+      "learning_rate": 1.3520408163265307e-06,
+      "loss": 1.0801,
+      "step": 788
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4807455837726593,
+      "learning_rate": 1.3392857142857143e-06,
+      "loss": 1.0599,
+      "step": 789
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49424856901168823,
+      "learning_rate": 1.3265306122448982e-06,
+      "loss": 1.0793,
+      "step": 790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4978589415550232,
+      "learning_rate": 1.3137755102040817e-06,
+      "loss": 1.0783,
+      "step": 791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49701061844825745,
+      "learning_rate": 1.3010204081632654e-06,
+      "loss": 1.0784,
+      "step": 792
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.078125,
+      "eval_runtime": 119.7086,
+      "eval_samples_per_second": 68.516,
+      "eval_steps_per_second": 34.258,
+      "step": 792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48835909366607666,
+      "learning_rate": 1.288265306122449e-06,
+      "loss": 1.0646,
+      "step": 793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4778783321380615,
+      "learning_rate": 1.2755102040816329e-06,
+      "loss": 1.0798,
+      "step": 794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48762667179107666,
+      "learning_rate": 1.2627551020408164e-06,
+      "loss": 1.0637,
+      "step": 795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4785713255405426,
+      "learning_rate": 1.25e-06,
+      "loss": 1.0451,
+      "step": 796
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4821445345878601,
+      "learning_rate": 1.2372448979591838e-06,
+      "loss": 1.0519,
+      "step": 797
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4918241500854492,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 1.0555,
+      "step": 798
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.47542768716812134,
+      "learning_rate": 1.211734693877551e-06,
+      "loss": 1.0716,
+      "step": 799
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4874721169471741,
+      "learning_rate": 1.1989795918367348e-06,
+      "loss": 1.0691,
+      "step": 800
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4866637885570526,
+      "learning_rate": 1.1862244897959185e-06,
+      "loss": 1.0581,
+      "step": 801
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48888394236564636,
+      "learning_rate": 1.1734693877551022e-06,
+      "loss": 1.0514,
+      "step": 802
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48897311091423035,
+      "learning_rate": 1.160714285714286e-06,
+      "loss": 1.0798,
+      "step": 803
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4837857782840729,
+      "learning_rate": 1.1479591836734695e-06,
+      "loss": 1.0658,
+      "step": 804
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4774113595485687,
+      "learning_rate": 1.1352040816326532e-06,
+      "loss": 1.0721,
+      "step": 805
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4952887296676636,
+      "learning_rate": 1.122448979591837e-06,
+      "loss": 1.0591,
+      "step": 806
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4889083802700043,
+      "learning_rate": 1.1096938775510204e-06,
+      "loss": 1.0418,
+      "step": 807
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49243977665901184,
+      "learning_rate": 1.0969387755102041e-06,
+      "loss": 1.0455,
+      "step": 808
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.495235800743103,
+      "learning_rate": 1.0841836734693879e-06,
+      "loss": 1.0556,
+      "step": 809
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4856567084789276,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 1.0485,
+      "step": 810
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49020230770111084,
+      "learning_rate": 1.0586734693877553e-06,
+      "loss": 1.0468,
+      "step": 811
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.47778093814849854,
+      "learning_rate": 1.0459183673469388e-06,
+      "loss": 1.062,
+      "step": 812
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.47502633929252625,
+      "learning_rate": 1.0331632653061225e-06,
+      "loss": 1.0676,
+      "step": 813
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4991995096206665,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 1.0664,
+      "step": 814
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5038496255874634,
+      "learning_rate": 1.00765306122449e-06,
+      "loss": 1.0635,
+      "step": 815
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4831908345222473,
+      "learning_rate": 9.948979591836735e-07,
+      "loss": 1.0793,
+      "step": 816
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.48362016677856445,
+      "learning_rate": 9.821428571428572e-07,
+      "loss": 1.059,
+      "step": 817
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4806089699268341,
+      "learning_rate": 9.69387755102041e-07,
+      "loss": 1.054,
+      "step": 818
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4738354980945587,
+      "learning_rate": 9.566326530612244e-07,
+      "loss": 1.0863,
+      "step": 819
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.49784815311431885,
+      "learning_rate": 9.438775510204082e-07,
+      "loss": 1.0824,
+      "step": 820
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.48209649324417114,
+      "learning_rate": 9.31122448979592e-07,
+      "loss": 1.0669,
+      "step": 821
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4703362286090851,
+      "learning_rate": 9.183673469387756e-07,
+      "loss": 1.046,
+      "step": 822
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.48009294271469116,
+      "learning_rate": 9.056122448979593e-07,
+      "loss": 1.0486,
+      "step": 823
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4972153306007385,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 1.0625,
+      "step": 824
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4931214451789856,
+      "learning_rate": 8.801020408163266e-07,
+      "loss": 1.0627,
+      "step": 825
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5141698718070984,
+      "learning_rate": 8.673469387755103e-07,
+      "loss": 1.0644,
+      "step": 826
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4702330529689789,
+      "learning_rate": 8.545918367346939e-07,
+      "loss": 1.0533,
+      "step": 827
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4845568835735321,
+      "learning_rate": 8.418367346938776e-07,
+      "loss": 1.0402,
+      "step": 828
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4730464518070221,
+      "learning_rate": 8.290816326530612e-07,
+      "loss": 1.0373,
+      "step": 829
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.49070635437965393,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 1.0584,
+      "step": 830
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.47989973425865173,
+      "learning_rate": 8.035714285714287e-07,
+      "loss": 1.0455,
+      "step": 831
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.48409467935562134,
+      "learning_rate": 7.908163265306124e-07,
+      "loss": 1.0597,
+      "step": 832
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4711008667945862,
+      "learning_rate": 7.78061224489796e-07,
+      "loss": 1.0606,
+      "step": 833
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4811023771762848,
+      "learning_rate": 7.653061224489796e-07,
+      "loss": 1.0624,
+      "step": 834
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4816673994064331,
+      "learning_rate": 7.525510204081634e-07,
+      "loss": 1.0628,
+      "step": 835
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5013629198074341,
+      "learning_rate": 7.39795918367347e-07,
+      "loss": 1.0311,
+      "step": 836
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.49255239963531494,
+      "learning_rate": 7.270408163265307e-07,
+      "loss": 1.0812,
+      "step": 837
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.49093401432037354,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.0478,
+      "step": 838
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.46841347217559814,
+      "learning_rate": 7.015306122448979e-07,
+      "loss": 1.0526,
+      "step": 839
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.48145774006843567,
+      "learning_rate": 6.887755102040817e-07,
+      "loss": 1.0521,
+      "step": 840
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.48966702818870544,
+      "learning_rate": 6.760204081632654e-07,
+      "loss": 1.0602,
+      "step": 841
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.47580987215042114,
+      "learning_rate": 6.632653061224491e-07,
+      "loss": 1.056,
+      "step": 842
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.48572811484336853,
+      "learning_rate": 6.505102040816327e-07,
+      "loss": 1.0443,
+      "step": 843
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.4886782467365265,
+      "learning_rate": 6.377551020408164e-07,
+      "loss": 1.0551,
+      "step": 844
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4674248695373535,
+      "learning_rate": 6.25e-07,
+      "loss": 1.0462,
+      "step": 845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.48394903540611267,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 1.0645,
+      "step": 846
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4802626073360443,
+      "learning_rate": 5.994897959183674e-07,
+      "loss": 1.0481,
+      "step": 847
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.49721604585647583,
+      "learning_rate": 5.867346938775511e-07,
+      "loss": 1.0787,
+      "step": 848
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4986632764339447,
+      "learning_rate": 5.739795918367347e-07,
+      "loss": 1.0496,
+      "step": 849
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5088779330253601,
+      "learning_rate": 5.612244897959184e-07,
+      "loss": 1.0568,
+      "step": 850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4756823778152466,
+      "learning_rate": 5.484693877551021e-07,
+      "loss": 1.0526,
+      "step": 851
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.48517045378685,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 1.0592,
+      "step": 852
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.47856712341308594,
+      "learning_rate": 5.229591836734694e-07,
+      "loss": 1.0521,
+      "step": 853
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4860280454158783,
+      "learning_rate": 5.102040816326531e-07,
+      "loss": 1.0696,
+      "step": 854
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.48423293232917786,
+      "learning_rate": 4.974489795918367e-07,
+      "loss": 1.0502,
+      "step": 855
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4771590530872345,
+      "learning_rate": 4.846938775510205e-07,
+      "loss": 1.0739,
+      "step": 856
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4849052131175995,
+      "learning_rate": 4.719387755102041e-07,
+      "loss": 1.0596,
+      "step": 857
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.48700183629989624,
+      "learning_rate": 4.591836734693878e-07,
+      "loss": 1.0519,
+      "step": 858
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.48286789655685425,
+      "learning_rate": 4.4642857142857147e-07,
+      "loss": 1.0695,
+      "step": 859
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4692079722881317,
+      "learning_rate": 4.3367346938775514e-07,
+      "loss": 1.0601,
+      "step": 860
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.48513585329055786,
+      "learning_rate": 4.209183673469388e-07,
+      "loss": 1.0673,
+      "step": 861
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4999236464500427,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 1.0706,
+      "step": 862
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.48926645517349243,
+      "learning_rate": 3.954081632653062e-07,
+      "loss": 1.0432,
+      "step": 863
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4762410819530487,
+      "learning_rate": 3.826530612244898e-07,
+      "loss": 1.0508,
+      "step": 864
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.47507932782173157,
+      "learning_rate": 3.698979591836735e-07,
+      "loss": 1.0535,
+      "step": 865
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.47133708000183105,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 1.0429,
+      "step": 866
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4802726209163666,
+      "learning_rate": 3.443877551020408e-07,
+      "loss": 1.0506,
+      "step": 867
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4963398277759552,
+      "learning_rate": 3.3163265306122455e-07,
+      "loss": 1.0778,
+      "step": 868
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4718584418296814,
+      "learning_rate": 3.188775510204082e-07,
+      "loss": 1.0353,
+      "step": 869
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4984806180000305,
+      "learning_rate": 3.0612244897959183e-07,
+      "loss": 1.0669,
+      "step": 870
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4719929099082947,
+      "learning_rate": 2.9336734693877556e-07,
+      "loss": 1.0425,
+      "step": 871
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.48878350853919983,
+      "learning_rate": 2.806122448979592e-07,
+      "loss": 1.0578,
+      "step": 872
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 1.0732421875,
+      "eval_runtime": 119.8939,
+      "eval_samples_per_second": 68.41,
+      "eval_steps_per_second": 34.205,
+      "step": 872
+    },
+    {
+      "epoch": 1.1,
+      "step": 872,
+      "total_flos": 1.197452692704723e+18,
+      "train_loss": 1.2847854858691539,
+      "train_runtime": 109212.3273,
+      "train_samples_per_second": 8.178,
+      "train_steps_per_second": 0.008
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 872,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1,
+  "total_flos": 1.197452692704723e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}