diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.04,
+  "eval_steps": 50,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4e-05,
+      "grad_norm": 11.292348861694336,
+      "learning_rate": 0.0,
+      "loss": 0.5879,
+      "step": 1
+    },
+    {
+      "epoch": 8e-05,
+      "grad_norm": 5.98015832901001,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.7321,
+      "step": 2
+    },
+    {
+      "epoch": 0.00012,
+      "grad_norm": 5.267852783203125,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.3428,
+      "step": 3
+    },
+    {
+      "epoch": 0.00016,
+      "grad_norm": 5.546362400054932,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.4415,
+      "step": 4
+    },
+    {
+      "epoch": 0.0002,
+      "grad_norm": 9.708696365356445,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.922,
+      "step": 5
+    },
+    {
+      "epoch": 0.00024,
+      "grad_norm": 5.526885986328125,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.3137,
+      "step": 6
+    },
+    {
+      "epoch": 0.00028,
+      "grad_norm": 5.722662448883057,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.6669,
+      "step": 7
+    },
+    {
+      "epoch": 0.00032,
+      "grad_norm": 7.009148120880127,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.3416,
+      "step": 8
+    },
+    {
+      "epoch": 0.00036,
+      "grad_norm": 10.944995880126953,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.944,
+      "step": 9
+    },
+    {
+      "epoch": 0.0004,
+      "grad_norm": 5.111751556396484,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.3114,
+      "step": 10
+    },
+    {
+      "epoch": 0.00044,
+      "grad_norm": 8.339456558227539,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.4201,
+      "step": 11
+    },
+    {
+      "epoch": 0.00048,
+      "grad_norm": 6.22341251373291,
+      "learning_rate": 2.2e-06,
+      "loss": 0.3123,
+      "step": 12
+    },
+    {
+      "epoch": 0.00052,
+      "grad_norm": 9.959796905517578,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8624,
+      "step": 13
+    },
+    {
+      "epoch": 0.00056,
+      "grad_norm": 4.146289825439453,
+      "learning_rate": 2.6e-06,
+      "loss": 0.2001,
+      "step": 14
+    },
+    {
+      "epoch": 0.0006,
+      "grad_norm": 6.458455562591553,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.402,
+      "step": 15
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 10.759032249450684,
+      "learning_rate": 3e-06,
+      "loss": 0.6514,
+      "step": 16
+    },
+    {
+      "epoch": 0.00068,
+      "grad_norm": 4.342711925506592,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.352,
+      "step": 17
+    },
+    {
+      "epoch": 0.00072,
+      "grad_norm": 7.300567626953125,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.6157,
+      "step": 18
+    },
+    {
+      "epoch": 0.00076,
+      "grad_norm": 7.142605304718018,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.1848,
+      "step": 19
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 5.487978935241699,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.4656,
+      "step": 20
+    },
+    {
+      "epoch": 0.00084,
+      "grad_norm": 9.961310386657715,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5615,
+      "step": 21
+    },
+    {
+      "epoch": 0.00088,
+      "grad_norm": 7.270428657531738,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7521,
+      "step": 22
+    },
+    {
+      "epoch": 0.00092,
+      "grad_norm": 7.73252534866333,
+      "learning_rate": 4.4e-06,
+      "loss": 0.6043,
+      "step": 23
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 9.061256408691406,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.5364,
+      "step": 24
+    },
+    {
+      "epoch": 0.001,
+      "grad_norm": 3.229898691177368,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.2068,
+      "step": 25
+    },
+    {
+      "epoch": 0.00104,
+      "grad_norm": 9.038275718688965,
+      "learning_rate": 5e-06,
+      "loss": 0.7441,
+      "step": 26
+    },
+    {
+      "epoch": 0.00108,
+      "grad_norm": 8.578261375427246,
+      "learning_rate": 5.2e-06,
+      "loss": 0.7859,
+      "step": 27
+    },
+    {
+      "epoch": 0.00112,
+      "grad_norm": 6.900351047515869,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.4085,
+      "step": 28
+    },
+    {
+      "epoch": 0.00116,
+      "grad_norm": 8.593361854553223,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.7263,
+      "step": 29
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 3.5956859588623047,
+      "learning_rate": 5.8e-06,
+      "loss": 0.4099,
+      "step": 30
+    },
+    {
+      "epoch": 0.00124,
+      "grad_norm": 8.845776557922363,
+      "learning_rate": 6e-06,
+      "loss": 0.5834,
+      "step": 31
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 10.743149757385254,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.9868,
+      "step": 32
+    },
+    {
+      "epoch": 0.00132,
+      "grad_norm": 8.158079147338867,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.8745,
+      "step": 33
+    },
+    {
+      "epoch": 0.00136,
+      "grad_norm": 3.6267876625061035,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.1236,
+      "step": 34
+    },
+    {
+      "epoch": 0.0014,
+      "grad_norm": 7.552522659301758,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.7084,
+      "step": 35
+    },
+    {
+      "epoch": 0.00144,
+      "grad_norm": 13.59090805053711,
+      "learning_rate": 7e-06,
+      "loss": 0.8113,
+      "step": 36
+    },
+    {
+      "epoch": 0.00148,
+      "grad_norm": 3.308990240097046,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.2687,
+      "step": 37
+    },
+    {
+      "epoch": 0.00152,
+      "grad_norm": 2.898580551147461,
+      "learning_rate": 7.4e-06,
+      "loss": 0.1995,
+      "step": 38
+    },
+    {
+      "epoch": 0.00156,
+      "grad_norm": 6.26355504989624,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.472,
+      "step": 39
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 4.201684474945068,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.3247,
+      "step": 40
+    },
+    {
+      "epoch": 0.00164,
+      "grad_norm": 3.5538246631622314,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.1211,
+      "step": 41
+    },
+    {
+      "epoch": 0.00168,
+      "grad_norm": 11.741997718811035,
+      "learning_rate": 8.2e-06,
+      "loss": 0.624,
+      "step": 42
+    },
+    {
+      "epoch": 0.00172,
+      "grad_norm": 6.365185737609863,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.581,
+      "step": 43
+    },
+    {
+      "epoch": 0.00176,
+      "grad_norm": 11.875202178955078,
+      "learning_rate": 8.6e-06,
+      "loss": 0.2942,
+      "step": 44
+    },
+    {
+      "epoch": 0.0018,
+      "grad_norm": 5.565279483795166,
+      "learning_rate": 8.8e-06,
+      "loss": 0.4682,
+      "step": 45
+    },
+    {
+      "epoch": 0.00184,
+      "grad_norm": 1.8691498041152954,
+      "learning_rate": 9e-06,
+      "loss": 0.1277,
+      "step": 46
+    },
+    {
+      "epoch": 0.00188,
+      "grad_norm": 10.839354515075684,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.7995,
+      "step": 47
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 3.7909772396087646,
+      "learning_rate": 9.4e-06,
+      "loss": 0.2405,
+      "step": 48
+    },
+    {
+      "epoch": 0.00196,
+      "grad_norm": 4.03000020980835,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.2248,
+      "step": 49
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 6.199543476104736,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.7268,
+      "step": 50
+    },
+    {
+      "epoch": 0.00204,
+      "grad_norm": 9.68652629852295,
+      "learning_rate": 1e-05,
+      "loss": 0.831,
+      "step": 51
+    },
+    {
+      "epoch": 0.00208,
+      "grad_norm": 7.7645721435546875,
+      "learning_rate": 1.02e-05,
+      "loss": 0.8584,
+      "step": 52
+    },
+    {
+      "epoch": 0.00212,
+      "grad_norm": 6.8255462646484375,
+      "learning_rate": 1.04e-05,
+      "loss": 0.6228,
+      "step": 53
+    },
+    {
+      "epoch": 0.00216,
+      "grad_norm": 7.695949077606201,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 0.6851,
+      "step": 54
+    },
+    {
+      "epoch": 0.0022,
+      "grad_norm": 7.131970405578613,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.4942,
+      "step": 55
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 6.7976484298706055,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.5735,
+      "step": 56
+    },
+    {
+      "epoch": 0.00228,
+      "grad_norm": 11.443472862243652,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.9056,
+      "step": 57
+    },
+    {
+      "epoch": 0.00232,
+      "grad_norm": 5.609070777893066,
+      "learning_rate": 1.14e-05,
+      "loss": 0.3772,
+      "step": 58
+    },
+    {
+      "epoch": 0.00236,
+      "grad_norm": 2.7658133506774902,
+      "learning_rate": 1.16e-05,
+      "loss": 0.2007,
+      "step": 59
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 6.5747175216674805,
+      "learning_rate": 1.18e-05,
+      "loss": 0.4634,
+      "step": 60
+    },
+    {
+      "epoch": 0.00244,
+      "grad_norm": 8.063694953918457,
+      "learning_rate": 1.2e-05,
+      "loss": 0.8941,
+      "step": 61
+    },
+    {
+      "epoch": 0.00248,
+      "grad_norm": 1.418664813041687,
+      "learning_rate": 1.22e-05,
+      "loss": 0.1862,
+      "step": 62
+    },
+    {
+      "epoch": 0.00252,
+      "grad_norm": 3.039188861846924,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.1075,
+      "step": 63
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 7.310708522796631,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.6085,
+      "step": 64
+    },
+    {
+      "epoch": 0.0026,
+      "grad_norm": 3.227430820465088,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.1181,
+      "step": 65
+    },
+    {
+      "epoch": 0.00264,
+      "grad_norm": 4.987690448760986,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.2985,
+      "step": 66
+    },
+    {
+      "epoch": 0.00268,
+      "grad_norm": 2.9371721744537354,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.2096,
+      "step": 67
+    },
+    {
+      "epoch": 0.00272,
+      "grad_norm": 10.126900672912598,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.7161,
+      "step": 68
+    },
+    {
+      "epoch": 0.00276,
+      "grad_norm": 7.239901542663574,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.5133,
+      "step": 69
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 6.861073970794678,
+      "learning_rate": 1.38e-05,
+      "loss": 0.3688,
+      "step": 70
+    },
+    {
+      "epoch": 0.00284,
+      "grad_norm": 5.571442604064941,
+      "learning_rate": 1.4e-05,
+      "loss": 0.3012,
+      "step": 71
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 3.0669121742248535,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 0.3499,
+      "step": 72
+    },
+    {
+      "epoch": 0.00292,
+      "grad_norm": 3.999744176864624,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.3531,
+      "step": 73
+    },
+    {
+      "epoch": 0.00296,
+      "grad_norm": 4.861403942108154,
+      "learning_rate": 1.46e-05,
+      "loss": 0.4143,
+      "step": 74
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 7.454000473022461,
+      "learning_rate": 1.48e-05,
+      "loss": 0.4864,
+      "step": 75
+    },
+    {
+      "epoch": 0.00304,
+      "grad_norm": 4.531309604644775,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3676,
+      "step": 76
+    },
+    {
+      "epoch": 0.00308,
+      "grad_norm": 7.034474849700928,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.5264,
+      "step": 77
+    },
+    {
+      "epoch": 0.00312,
+      "grad_norm": 9.963683128356934,
+      "learning_rate": 1.54e-05,
+      "loss": 0.3694,
+      "step": 78
+    },
+    {
+      "epoch": 0.00316,
+      "grad_norm": 1.9476526975631714,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.1869,
+      "step": 79
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 1.3104735612869263,
+      "learning_rate": 1.58e-05,
+      "loss": 0.1056,
+      "step": 80
+    },
+    {
+      "epoch": 0.00324,
+      "grad_norm": 2.2083330154418945,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2005,
+      "step": 81
+    },
+    {
+      "epoch": 0.00328,
+      "grad_norm": 5.3363165855407715,
+      "learning_rate": 1.62e-05,
+      "loss": 0.2634,
+      "step": 82
+    },
+    {
+      "epoch": 0.00332,
+      "grad_norm": 11.242843627929688,
+      "learning_rate": 1.64e-05,
+      "loss": 1.1091,
+      "step": 83
+    },
+    {
+      "epoch": 0.00336,
+      "grad_norm": 7.179255962371826,
+      "learning_rate": 1.66e-05,
+      "loss": 0.3439,
+      "step": 84
+    },
+    {
+      "epoch": 0.0034,
+      "grad_norm": 8.445199966430664,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.8232,
+      "step": 85
+    },
+    {
+      "epoch": 0.00344,
+      "grad_norm": 4.664881229400635,
+      "learning_rate": 1.7e-05,
+      "loss": 0.2455,
+      "step": 86
+    },
+    {
+      "epoch": 0.00348,
+      "grad_norm": 8.128833770751953,
+      "learning_rate": 1.72e-05,
+      "loss": 0.3129,
+      "step": 87
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 9.402512550354004,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 0.3957,
+      "step": 88
+    },
+    {
+      "epoch": 0.00356,
+      "grad_norm": 3.260019540786743,
+      "learning_rate": 1.76e-05,
+      "loss": 0.1802,
+      "step": 89
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 7.607392311096191,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 0.1656,
+      "step": 90
+    },
+    {
+      "epoch": 0.00364,
+      "grad_norm": 10.270995140075684,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5582,
+      "step": 91
+    },
+    {
+      "epoch": 0.00368,
+      "grad_norm": 4.662510395050049,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.1049,
+      "step": 92
+    },
+    {
+      "epoch": 0.00372,
+      "grad_norm": 8.008647918701172,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.6918,
+      "step": 93
+    },
+    {
+      "epoch": 0.00376,
+      "grad_norm": 13.225763320922852,
+      "learning_rate": 1.86e-05,
+      "loss": 0.7782,
+      "step": 94
+    },
+    {
+      "epoch": 0.0038,
+      "grad_norm": 5.24733829498291,
+      "learning_rate": 1.88e-05,
+      "loss": 0.2837,
+      "step": 95
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 2.362607479095459,
+      "learning_rate": 1.9e-05,
+      "loss": 0.1659,
+      "step": 96
+    },
+    {
+      "epoch": 0.00388,
+      "grad_norm": 11.296570777893066,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.5652,
+      "step": 97
+    },
+    {
+      "epoch": 0.00392,
+      "grad_norm": 12.645127296447754,
+      "learning_rate": 1.94e-05,
+      "loss": 0.3574,
+      "step": 98
+    },
+    {
+      "epoch": 0.00396,
+      "grad_norm": 5.612782001495361,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.4221,
+      "step": 99
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 8.27474594116211,
+      "learning_rate": 1.98e-05,
+      "loss": 0.3161,
+      "step": 100
+    },
+    {
+      "epoch": 0.00404,
+      "grad_norm": 3.731281280517578,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 101
+    },
+    {
+      "epoch": 0.00408,
+      "grad_norm": 5.077661991119385,
+      "learning_rate": 1.9999939076577906e-05,
+      "loss": 0.1812,
+      "step": 102
+    },
+    {
+      "epoch": 0.00412,
+      "grad_norm": 7.391285419464111,
+      "learning_rate": 1.9999756307053947e-05,
+      "loss": 0.6966,
+      "step": 103
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 6.538702011108398,
+      "learning_rate": 1.9999451693655125e-05,
+      "loss": 0.3295,
+      "step": 104
+    },
+    {
+      "epoch": 0.0042,
+      "grad_norm": 5.847230434417725,
+      "learning_rate": 1.9999025240093045e-05,
+      "loss": 0.1437,
+      "step": 105
+    },
+    {
+      "epoch": 0.00424,
+      "grad_norm": 5.70723295211792,
+      "learning_rate": 1.9998476951563914e-05,
+      "loss": 0.4592,
+      "step": 106
+    },
+    {
+      "epoch": 0.00428,
+      "grad_norm": 0.8218597173690796,
+      "learning_rate": 1.9997806834748455e-05,
+      "loss": 0.0926,
+      "step": 107
+    },
+    {
+      "epoch": 0.00432,
+      "grad_norm": 3.7447714805603027,
+      "learning_rate": 1.9997014897811834e-05,
+      "loss": 0.1708,
+      "step": 108
+    },
+    {
+      "epoch": 0.00436,
+      "grad_norm": 8.909958839416504,
+      "learning_rate": 1.9996101150403543e-05,
+      "loss": 0.5381,
+      "step": 109
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 3.5772249698638916,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.2456,
+      "step": 110
+    },
+    {
+      "epoch": 0.00444,
+      "grad_norm": 1.8850897550582886,
+      "learning_rate": 1.999390827019096e-05,
+      "loss": 0.1759,
+      "step": 111
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 7.449399948120117,
+      "learning_rate": 1.999262916410621e-05,
+      "loss": 0.3917,
+      "step": 112
+    },
+    {
+      "epoch": 0.00452,
+      "grad_norm": 4.810344219207764,
+      "learning_rate": 1.9991228300988586e-05,
+      "loss": 0.3322,
+      "step": 113
+    },
+    {
+      "epoch": 0.00456,
+      "grad_norm": 4.789427280426025,
+      "learning_rate": 1.998970569790715e-05,
+      "loss": 0.1724,
+      "step": 114
+    },
+    {
+      "epoch": 0.0046,
+      "grad_norm": 1.9709957838058472,
+      "learning_rate": 1.9988061373414342e-05,
+      "loss": 0.1774,
+      "step": 115
+    },
+    {
+      "epoch": 0.00464,
+      "grad_norm": 11.115662574768066,
+      "learning_rate": 1.9986295347545738e-05,
+      "loss": 0.6839,
+      "step": 116
+    },
+    {
+      "epoch": 0.00468,
+      "grad_norm": 1.0981431007385254,
+      "learning_rate": 1.9984407641819812e-05,
+      "loss": 0.055,
+      "step": 117
+    },
+    {
+      "epoch": 0.00472,
+      "grad_norm": 2.2565126419067383,
+      "learning_rate": 1.9982398279237657e-05,
+      "loss": 0.1603,
+      "step": 118
+    },
+    {
+      "epoch": 0.00476,
+      "grad_norm": 2.357349157333374,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.143,
+      "step": 119
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 0.5598532557487488,
+      "learning_rate": 1.9978014682920503e-05,
+      "loss": 0.0177,
+      "step": 120
+    },
+    {
+      "epoch": 0.00484,
+      "grad_norm": 6.7046284675598145,
+      "learning_rate": 1.9975640502598243e-05,
+      "loss": 0.5838,
+      "step": 121
+    },
+    {
+      "epoch": 0.00488,
+      "grad_norm": 9.128063201904297,
+      "learning_rate": 1.997314477224458e-05,
+      "loss": 0.507,
+      "step": 122
+    },
+    {
+      "epoch": 0.00492,
+      "grad_norm": 3.900993585586548,
+      "learning_rate": 1.9970527522269204e-05,
+      "loss": 0.2463,
+      "step": 123
+    },
+    {
+      "epoch": 0.00496,
+      "grad_norm": 6.569873809814453,
+      "learning_rate": 1.9967788784562474e-05,
+      "loss": 0.3339,
+      "step": 124
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 2.224302291870117,
+      "learning_rate": 1.9964928592495046e-05,
+      "loss": 0.1472,
+      "step": 125
+    },
+    {
+      "epoch": 0.00504,
+      "grad_norm": 2.9040894508361816,
+      "learning_rate": 1.9961946980917457e-05,
+      "loss": 0.215,
+      "step": 126
+    },
+    {
+      "epoch": 0.00508,
+      "grad_norm": 7.263742446899414,
+      "learning_rate": 1.9958843986159705e-05,
+      "loss": 0.3728,
+      "step": 127
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 1.1857038736343384,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.1051,
+      "step": 128
+    },
+    {
+      "epoch": 0.00516,
+      "grad_norm": 3.9250738620758057,
+      "learning_rate": 1.9952273999818312e-05,
+      "loss": 0.2019,
+      "step": 129
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 4.663411617279053,
+      "learning_rate": 1.9948807088287884e-05,
+      "loss": 0.2725,
+      "step": 130
+    },
+    {
+      "epoch": 0.00524,
+      "grad_norm": 4.305727005004883,
+      "learning_rate": 1.9945218953682736e-05,
+      "loss": 0.2478,
+      "step": 131
+    },
+    {
+      "epoch": 0.00528,
+      "grad_norm": 6.489120006561279,
+      "learning_rate": 1.9941509639723155e-05,
+      "loss": 0.3291,
+      "step": 132
+    },
+    {
+      "epoch": 0.00532,
+      "grad_norm": 4.379711151123047,
+      "learning_rate": 1.9937679191605964e-05,
+      "loss": 0.19,
+      "step": 133
+    },
+    {
+      "epoch": 0.00536,
+      "grad_norm": 7.65509033203125,
+      "learning_rate": 1.9933727656003964e-05,
+      "loss": 0.3565,
+      "step": 134
+    },
+    {
+      "epoch": 0.0054,
+      "grad_norm": 4.269521713256836,
+      "learning_rate": 1.992965508106537e-05,
+      "loss": 0.1361,
+      "step": 135
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 8.052738189697266,
+      "learning_rate": 1.9925461516413224e-05,
+      "loss": 0.5276,
+      "step": 136
+    },
+    {
+      "epoch": 0.00548,
+      "grad_norm": 2.0842795372009277,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.1089,
+      "step": 137
+    },
+    {
+      "epoch": 0.00552,
+      "grad_norm": 8.926876068115234,
+      "learning_rate": 1.9916711623830904e-05,
+      "loss": 0.5601,
+      "step": 138
+    },
+    {
+      "epoch": 0.00556,
+      "grad_norm": 2.0838093757629395,
+      "learning_rate": 1.991215540251542e-05,
+      "loss": 0.1284,
+      "step": 139
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 3.2754461765289307,
+      "learning_rate": 1.9907478404714438e-05,
+      "loss": 0.1737,
+      "step": 140
+    },
+    {
+      "epoch": 0.00564,
+      "grad_norm": 11.024544715881348,
+      "learning_rate": 1.9902680687415704e-05,
+      "loss": 0.4857,
+      "step": 141
+    },
+    {
+      "epoch": 0.00568,
+      "grad_norm": 1.295817494392395,
+      "learning_rate": 1.989776230907789e-05,
+      "loss": 0.1086,
+      "step": 142
+    },
+    {
+      "epoch": 0.00572,
+      "grad_norm": 11.132139205932617,
+      "learning_rate": 1.9892723329629885e-05,
+      "loss": 0.5312,
+      "step": 143
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 10.619361877441406,
+      "learning_rate": 1.988756381047006e-05,
+      "loss": 0.5485,
+      "step": 144
+    },
+    {
+      "epoch": 0.0058,
+      "grad_norm": 3.785047769546509,
+      "learning_rate": 1.988228381446553e-05,
+      "loss": 0.322,
+      "step": 145
+    },
+    {
+      "epoch": 0.00584,
+      "grad_norm": 1.2806971073150635,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.1364,
+      "step": 146
+    },
+    {
+      "epoch": 0.00588,
+      "grad_norm": 7.42913293838501,
+      "learning_rate": 1.987136265072988e-05,
+      "loss": 0.2496,
+      "step": 147
+    },
+    {
+      "epoch": 0.00592,
+      "grad_norm": 1.8928861618041992,
+      "learning_rate": 1.9865721616069695e-05,
+      "loss": 0.1203,
+      "step": 148
+    },
+    {
+      "epoch": 0.00596,
+      "grad_norm": 5.805448532104492,
+      "learning_rate": 1.985996037070505e-05,
+      "loss": 0.3457,
+      "step": 149
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 1.7266368865966797,
+      "learning_rate": 1.9854078984834904e-05,
+      "loss": 0.1223,
+      "step": 150
+    },
+    {
+      "epoch": 0.00604,
+      "grad_norm": 3.0544159412384033,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.1834,
+      "step": 151
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 5.120899677276611,
+      "learning_rate": 1.984195607969242e-05,
+      "loss": 0.2567,
+      "step": 152
+    },
+    {
+      "epoch": 0.00612,
+      "grad_norm": 4.808404922485352,
+      "learning_rate": 1.983571470813386e-05,
+      "loss": 0.217,
+      "step": 153
+    },
+    {
+      "epoch": 0.00616,
+      "grad_norm": 2.2536799907684326,
+      "learning_rate": 1.9829353491495545e-05,
+      "loss": 0.1423,
+      "step": 154
+    },
+    {
+      "epoch": 0.0062,
+      "grad_norm": 5.83665132522583,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.2755,
+      "step": 155
+    },
+    {
+      "epoch": 0.00624,
+      "grad_norm": 6.675769805908203,
+      "learning_rate": 1.9816271834476642e-05,
+      "loss": 0.2292,
+      "step": 156
+    },
+    {
+      "epoch": 0.00628,
+      "grad_norm": 2.519597291946411,
+      "learning_rate": 1.9809551553491918e-05,
+      "loss": 0.1591,
+      "step": 157
+    },
+    {
+      "epoch": 0.00632,
+      "grad_norm": 13.408095359802246,
+      "learning_rate": 1.9802711746217222e-05,
+      "loss": 0.7335,
+      "step": 158
+    },
+    {
+      "epoch": 0.00636,
+      "grad_norm": 5.37779426574707,
+      "learning_rate": 1.979575249599344e-05,
+      "loss": 0.2193,
+      "step": 159
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 6.302042007446289,
+      "learning_rate": 1.9788673887616852e-05,
+      "loss": 0.4127,
+      "step": 160
+    },
+    {
+      "epoch": 0.00644,
+      "grad_norm": 11.216113090515137,
+      "learning_rate": 1.9781476007338058e-05,
+      "loss": 0.445,
+      "step": 161
+    },
+    {
+      "epoch": 0.00648,
+      "grad_norm": 4.312234401702881,
+      "learning_rate": 1.9774158942860962e-05,
+      "loss": 0.1945,
+      "step": 162
+    },
+    {
+      "epoch": 0.00652,
+      "grad_norm": 4.612367153167725,
+      "learning_rate": 1.9766722783341682e-05,
+      "loss": 0.1919,
+      "step": 163
+    },
+    {
+      "epoch": 0.00656,
+      "grad_norm": 13.866952896118164,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.5538,
+      "step": 164
+    },
+    {
+      "epoch": 0.0066,
+      "grad_norm": 13.941161155700684,
+      "learning_rate": 1.9751493543055634e-05,
+      "loss": 0.5394,
+      "step": 165
+    },
+    {
+      "epoch": 0.00664,
+      "grad_norm": 9.81427001953125,
+      "learning_rate": 1.9743700647852356e-05,
+      "loss": 0.3897,
+      "step": 166
+    },
+    {
+      "epoch": 0.00668,
+      "grad_norm": 11.661919593811035,
+      "learning_rate": 1.9735789028731603e-05,
+      "loss": 0.4503,
+      "step": 167
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 9.800228118896484,
+      "learning_rate": 1.972775878209397e-05,
+      "loss": 0.7267,
+      "step": 168
+    },
+    {
+      "epoch": 0.00676,
+      "grad_norm": 8.045507431030273,
+      "learning_rate": 1.9719610005785466e-05,
+      "loss": 0.2883,
+      "step": 169
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 8.941047668457031,
+      "learning_rate": 1.971134279909636e-05,
+      "loss": 0.4091,
+      "step": 170
+    },
+    {
+      "epoch": 0.00684,
+      "grad_norm": 3.1888413429260254,
+      "learning_rate": 1.9702957262759964e-05,
+      "loss": 0.0684,
+      "step": 171
+    },
+    {
+      "epoch": 0.00688,
+      "grad_norm": 1.5820033550262451,
+      "learning_rate": 1.9694453498951392e-05,
+      "loss": 0.1145,
+      "step": 172
+    },
+    {
+      "epoch": 0.00692,
+      "grad_norm": 0.9278202056884766,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.0977,
+      "step": 173
+    },
+    {
+      "epoch": 0.00696,
+      "grad_norm": 1.4843041896820068,
+      "learning_rate": 1.9677091704819714e-05,
+      "loss": 0.0596,
+      "step": 174
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 1.244142770767212,
+      "learning_rate": 1.9668233886044597e-05,
+      "loss": 0.1367,
+      "step": 175
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 6.203249931335449,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.2484,
+      "step": 176
+    },
+    {
+      "epoch": 0.00708,
+      "grad_norm": 13.03145694732666,
+      "learning_rate": 1.9650164944723116e-05,
+      "loss": 0.3046,
+      "step": 177
+    },
+    {
+      "epoch": 0.00712,
+      "grad_norm": 7.601573467254639,
+      "learning_rate": 1.96409540423411e-05,
+      "loss": 0.5502,
+      "step": 178
+    },
+    {
+      "epoch": 0.00716,
+      "grad_norm": 15.64136791229248,
+      "learning_rate": 1.9631625667976584e-05,
+      "loss": 0.4973,
+      "step": 179
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 19.07086181640625,
+      "learning_rate": 1.9622179935292855e-05,
+      "loss": 0.528,
+      "step": 180
+    },
+    {
+      "epoch": 0.00724,
+      "grad_norm": 1.8338375091552734,
+      "learning_rate": 1.961261695938319e-05,
+      "loss": 0.1414,
+      "step": 181
+    },
+    {
+      "epoch": 0.00728,
+      "grad_norm": 8.561354637145996,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.5726,
+      "step": 182
+    },
+    {
+      "epoch": 0.00732,
+      "grad_norm": 5.592373847961426,
+      "learning_rate": 1.9593139745400575e-05,
+      "loss": 0.2327,
+      "step": 183
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 4.015738487243652,
+      "learning_rate": 1.9583225744651334e-05,
+      "loss": 0.1825,
+      "step": 184
+    },
+    {
+      "epoch": 0.0074,
+      "grad_norm": 7.303484916687012,
+      "learning_rate": 1.9573194975320672e-05,
+      "loss": 0.4096,
+      "step": 185
+    },
+    {
+      "epoch": 0.00744,
+      "grad_norm": 5.297791957855225,
+      "learning_rate": 1.9563047559630356e-05,
+      "loss": 0.2484,
+      "step": 186
+    },
+    {
+      "epoch": 0.00748,
+      "grad_norm": 3.077599287033081,
+      "learning_rate": 1.9552783621223437e-05,
+      "loss": 0.167,
+      "step": 187
+    },
+    {
+      "epoch": 0.00752,
+      "grad_norm": 0.6694372892379761,
+      "learning_rate": 1.954240328516277e-05,
+      "loss": 0.0404,
+      "step": 188
+    },
+    {
+      "epoch": 0.00756,
+      "grad_norm": 8.105072975158691,
+      "learning_rate": 1.9531906677929472e-05,
+      "loss": 0.2844,
+      "step": 189
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 4.712032318115234,
+      "learning_rate": 1.9521293927421388e-05,
+      "loss": 0.2381,
+      "step": 190
+    },
+    {
+      "epoch": 0.00764,
+      "grad_norm": 10.463435173034668,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.2274,
+      "step": 191
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 8.566754341125488,
+      "learning_rate": 1.9499720515246524e-05,
+      "loss": 0.5679,
+      "step": 192
+    },
+    {
+      "epoch": 0.00772,
+      "grad_norm": 3.0031518936157227,
+      "learning_rate": 1.9488760116444966e-05,
+      "loss": 0.0752,
+      "step": 193
+    },
+    {
+      "epoch": 0.00776,
+      "grad_norm": 2.9552342891693115,
+      "learning_rate": 1.947768410009586e-05,
+      "loss": 0.1556,
+      "step": 194
+    },
+    {
+      "epoch": 0.0078,
+      "grad_norm": 11.878022193908691,
+      "learning_rate": 1.9466492601156964e-05,
+      "loss": 0.2476,
+      "step": 195
+    },
+    {
+      "epoch": 0.00784,
+      "grad_norm": 2.7819948196411133,
+      "learning_rate": 1.945518575599317e-05,
+      "loss": 0.193,
+      "step": 196
+    },
+    {
+      "epoch": 0.00788,
+      "grad_norm": 6.761250972747803,
+      "learning_rate": 1.944376370237481e-05,
+      "loss": 0.247,
+      "step": 197
+    },
+    {
+      "epoch": 0.00792,
+      "grad_norm": 2.704380750656128,
+      "learning_rate": 1.943222657947601e-05,
+      "loss": 0.1743,
+      "step": 198
+    },
+    {
+      "epoch": 0.00796,
+      "grad_norm": 13.92269515991211,
+      "learning_rate": 1.942057452787297e-05,
+      "loss": 0.7003,
+      "step": 199
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 7.142035961151123,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.4348,
+      "step": 200
+    },
+    {
+      "epoch": 0.00804,
+      "grad_norm": 12.909127235412598,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.374,
+      "step": 201
+    },
+    {
+      "epoch": 0.00808,
+      "grad_norm": 1.0318565368652344,
+      "learning_rate": 1.938493022759556e-05,
+      "loss": 0.0671,
+      "step": 202
+    },
+    {
+      "epoch": 0.00812,
+      "grad_norm": 2.911125421524048,
+      "learning_rate": 1.937281989491892e-05,
+      "loss": 0.1913,
+      "step": 203
+    },
+    {
+      "epoch": 0.00816,
+      "grad_norm": 3.2514820098876953,
+      "learning_rate": 1.9360595357389735e-05,
+      "loss": 0.2098,
+      "step": 204
+    },
+    {
+      "epoch": 0.0082,
+      "grad_norm": 2.5129122734069824,
+      "learning_rate": 1.9348256763960146e-05,
+      "loss": 0.1238,
+      "step": 205
+    },
+    {
+      "epoch": 0.00824,
+      "grad_norm": 4.252076148986816,
+      "learning_rate": 1.9335804264972018e-05,
+      "loss": 0.1699,
+      "step": 206
+    },
+    {
+      "epoch": 0.00828,
+      "grad_norm": 19.623329162597656,
+      "learning_rate": 1.9323238012155125e-05,
+      "loss": 0.5051,
+      "step": 207
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 4.171853065490723,
+      "learning_rate": 1.9310558158625286e-05,
+      "loss": 0.204,
+      "step": 208
+    },
+    {
+      "epoch": 0.00836,
+      "grad_norm": 2.729529619216919,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.1599,
+      "step": 209
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 2.4231808185577393,
+      "learning_rate": 1.9284858268809135e-05,
+      "loss": 0.1799,
+      "step": 210
+    },
+    {
+      "epoch": 0.00844,
+      "grad_norm": 3.384650468826294,
+      "learning_rate": 1.9271838545667876e-05,
+      "loss": 0.2002,
+      "step": 211
+    },
+    {
+      "epoch": 0.00848,
+      "grad_norm": 4.145476818084717,
+      "learning_rate": 1.925870584809995e-05,
+      "loss": 0.2088,
+      "step": 212
+    },
+    {
+      "epoch": 0.00852,
+      "grad_norm": 7.054924488067627,
+      "learning_rate": 1.9245460336123136e-05,
+      "loss": 0.6404,
+      "step": 213
+    },
+    {
+      "epoch": 0.00856,
+      "grad_norm": 4.773870944976807,
+      "learning_rate": 1.923210217112981e-05,
+      "loss": 0.2465,
+      "step": 214
+    },
+    {
+      "epoch": 0.0086,
+      "grad_norm": 2.81622576713562,
+      "learning_rate": 1.9218631515885007e-05,
+      "loss": 0.1465,
+      "step": 215
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 4.764132499694824,
+      "learning_rate": 1.9205048534524405e-05,
+      "loss": 0.1961,
+      "step": 216
+    },
+    {
+      "epoch": 0.00868,
+      "grad_norm": 11.092087745666504,
+      "learning_rate": 1.9191353392552346e-05,
+      "loss": 0.4628,
+      "step": 217
+    },
+    {
+      "epoch": 0.00872,
+      "grad_norm": 1.2834681272506714,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.0698,
+      "step": 218
+    },
+    {
+      "epoch": 0.00876,
+      "grad_norm": 0.7233999371528625,
+      "learning_rate": 1.9163627295622397e-05,
+      "loss": 0.0147,
+      "step": 219
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 8.03890323638916,
+      "learning_rate": 1.914959667849825e-05,
+      "loss": 0.3302,
+      "step": 220
+    },
+    {
+      "epoch": 0.00884,
+      "grad_norm": 7.3795061111450195,
+      "learning_rate": 1.913545457642601e-05,
+      "loss": 0.536,
+      "step": 221
+    },
+    {
+      "epoch": 0.00888,
+      "grad_norm": 1.05797278881073,
+      "learning_rate": 1.9121201161722732e-05,
+      "loss": 0.0966,
+      "step": 222
+    },
+    {
+      "epoch": 0.00892,
+      "grad_norm": 4.85446834564209,
+      "learning_rate": 1.910683660806177e-05,
+      "loss": 0.1868,
+      "step": 223
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 6.681173801422119,
+      "learning_rate": 1.9092361090470688e-05,
+      "loss": 0.3458,
+      "step": 224
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 6.193958759307861,
+      "learning_rate": 1.907777478532909e-05,
+      "loss": 0.2951,
+      "step": 225
+    },
+    {
+      "epoch": 0.00904,
+      "grad_norm": 4.845807075500488,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 0.239,
+      "step": 226
+    },
+    {
+      "epoch": 0.00908,
+      "grad_norm": 4.359218597412109,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.2995,
+      "step": 227
+    },
+    {
+      "epoch": 0.00912,
+      "grad_norm": 3.135012149810791,
+      "learning_rate": 1.903335292863301e-05,
+      "loss": 0.0641,
+      "step": 228
+    },
+    {
+      "epoch": 0.00916,
+      "grad_norm": 1.4781063795089722,
+      "learning_rate": 1.901832526405114e-05,
+      "loss": 0.105,
+      "step": 229
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 3.9598007202148438,
+      "learning_rate": 1.9003187714021936e-05,
+      "loss": 0.1861,
+      "step": 230
+    },
+    {
+      "epoch": 0.00924,
+      "grad_norm": 1.2587207555770874,
+      "learning_rate": 1.8987940462991673e-05,
+      "loss": 0.0867,
+      "step": 231
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 2.5407896041870117,
+      "learning_rate": 1.8972583696743284e-05,
+      "loss": 0.0962,
+      "step": 232
+    },
+    {
+      "epoch": 0.00932,
+      "grad_norm": 8.56326961517334,
+      "learning_rate": 1.895711760239413e-05,
+      "loss": 0.2053,
+      "step": 233
+    },
+    {
+      "epoch": 0.00936,
+      "grad_norm": 4.605000972747803,
+      "learning_rate": 1.8941542368393683e-05,
+      "loss": 0.2537,
+      "step": 234
+    },
+    {
+      "epoch": 0.0094,
+      "grad_norm": 4.095175743103027,
+      "learning_rate": 1.892585818452126e-05,
+      "loss": 0.2272,
+      "step": 235
+    },
+    {
+      "epoch": 0.00944,
+      "grad_norm": 6.641960144042969,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.1965,
+      "step": 236
+    },
+    {
+      "epoch": 0.00948,
+      "grad_norm": 6.449761867523193,
+      "learning_rate": 1.889416373291298e-05,
+      "loss": 0.2243,
+      "step": 237
+    },
+    {
+      "epoch": 0.00952,
+      "grad_norm": 7.998575687408447,
+      "learning_rate": 1.8878153851364013e-05,
+      "loss": 0.27,
+      "step": 238
+    },
+    {
+      "epoch": 0.00956,
+      "grad_norm": 1.7765926122665405,
+      "learning_rate": 1.8862035792312148e-05,
+      "loss": 0.13,
+      "step": 239
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 1.8574138879776,
+      "learning_rate": 1.884580975215084e-05,
+      "loss": 0.1386,
+      "step": 240
+    },
+    {
+      "epoch": 0.00964,
+      "grad_norm": 8.573052406311035,
+      "learning_rate": 1.8829475928589272e-05,
+      "loss": 0.2261,
+      "step": 241
+    },
+    {
+      "epoch": 0.00968,
+      "grad_norm": 2.332636833190918,
+      "learning_rate": 1.8813034520649923e-05,
+      "loss": 0.1234,
+      "step": 242
+    },
+    {
+      "epoch": 0.00972,
+      "grad_norm": 1.3433829545974731,
+      "learning_rate": 1.879648572866617e-05,
+      "loss": 0.111,
+      "step": 243
+    },
+    {
+      "epoch": 0.00976,
+      "grad_norm": 2.112905502319336,
+      "learning_rate": 1.8779829754279806e-05,
+      "loss": 0.1182,
+      "step": 244
+    },
+    {
+      "epoch": 0.0098,
+      "grad_norm": 12.797957420349121,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.2537,
+      "step": 245
+    },
+    {
+      "epoch": 0.00984,
+      "grad_norm": 4.487147808074951,
+      "learning_rate": 1.874619707139396e-05,
+      "loss": 0.1813,
+      "step": 246
+    },
+    {
+      "epoch": 0.00988,
+      "grad_norm": 2.1665732860565186,
+      "learning_rate": 1.8729220772698096e-05,
+      "loss": 0.1184,
+      "step": 247
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 0.39008113741874695,
+      "learning_rate": 1.8712138111201898e-05,
+      "loss": 0.0057,
+      "step": 248
+    },
+    {
+      "epoch": 0.00996,
+      "grad_norm": 8.801874160766602,
+      "learning_rate": 1.869494929505219e-05,
+      "loss": 0.3132,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.311043620109558,
+      "learning_rate": 1.8677654533689287e-05,
+      "loss": 0.0887,
+      "step": 250
+    },
+    {
+      "epoch": 0.01004,
+      "grad_norm": 6.260326385498047,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.2965,
+      "step": 251
+    },
+    {
+      "epoch": 0.01008,
+      "grad_norm": 2.761636257171631,
+      "learning_rate": 1.864274801953705e-05,
+      "loss": 0.1431,
+      "step": 252
+    },
+    {
+      "epoch": 0.01012,
+      "grad_norm": 4.143901348114014,
+      "learning_rate": 1.8625136692072577e-05,
+      "loss": 0.1733,
+      "step": 253
+    },
+    {
+      "epoch": 0.01016,
+      "grad_norm": 6.137015342712402,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.1469,
+      "step": 254
+    },
+    {
+      "epoch": 0.0102,
+      "grad_norm": 1.619290828704834,
+      "learning_rate": 1.8589598969306646e-05,
+      "loss": 0.0697,
+      "step": 255
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 6.0061821937561035,
+      "learning_rate": 1.8571673007021124e-05,
+      "loss": 0.3148,
+      "step": 256
+    },
+    {
+      "epoch": 0.01028,
+      "grad_norm": 3.734489679336548,
+      "learning_rate": 1.855364260160507e-05,
+      "loss": 0.1311,
+      "step": 257
+    },
+    {
+      "epoch": 0.01032,
+      "grad_norm": 0.5674763321876526,
+      "learning_rate": 1.8535507972753275e-05,
+      "loss": 0.0082,
+      "step": 258
+    },
+    {
+      "epoch": 0.01036,
+      "grad_norm": 3.277397871017456,
+      "learning_rate": 1.851726934143048e-05,
+      "loss": 0.1692,
+      "step": 259
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 11.149739265441895,
+      "learning_rate": 1.849892692986864e-05,
+      "loss": 0.5392,
+      "step": 260
+    },
+    {
+      "epoch": 0.01044,
+      "grad_norm": 3.4369068145751953,
+      "learning_rate": 1.848048096156426e-05,
+      "loss": 0.1365,
+      "step": 261
+    },
+    {
+      "epoch": 0.01048,
+      "grad_norm": 18.73288345336914,
+      "learning_rate": 1.8461931661275642e-05,
+      "loss": 0.4826,
+      "step": 262
+    },
+    {
+      "epoch": 0.01052,
+      "grad_norm": 1.0672292709350586,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.0266,
+      "step": 263
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 8.23385238647461,
+      "learning_rate": 1.842452397007148e-05,
+      "loss": 0.3126,
+      "step": 264
+    },
+    {
+      "epoch": 0.0106,
+      "grad_norm": 26.22130012512207,
+      "learning_rate": 1.8405666034956842e-05,
+      "loss": 0.6388,
+      "step": 265
+    },
+    {
+      "epoch": 0.01064,
+      "grad_norm": 8.078812599182129,
+      "learning_rate": 1.8386705679454243e-05,
+      "loss": 0.878,
+      "step": 266
+    },
+    {
+      "epoch": 0.01068,
+      "grad_norm": 1.3683041334152222,
+      "learning_rate": 1.836764313458962e-05,
+      "loss": 0.0849,
+      "step": 267
+    },
+    {
+      "epoch": 0.01072,
+      "grad_norm": 20.437116622924805,
+      "learning_rate": 1.8348478632634067e-05,
+      "loss": 0.1825,
+      "step": 268
+    },
+    {
+      "epoch": 0.01076,
+      "grad_norm": 3.5117948055267334,
+      "learning_rate": 1.8329212407100996e-05,
+      "loss": 0.1735,
+      "step": 269
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 4.878694534301758,
+      "learning_rate": 1.8309844692743283e-05,
+      "loss": 0.162,
+      "step": 270
+    },
+    {
+      "epoch": 0.01084,
+      "grad_norm": 15.349081993103027,
+      "learning_rate": 1.8290375725550417e-05,
+      "loss": 0.7004,
+      "step": 271
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 6.009783744812012,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.2451,
+      "step": 272
+    },
+    {
+      "epoch": 0.01092,
+      "grad_norm": 0.46118679642677307,
+      "learning_rate": 1.8251134982782952e-05,
+      "loss": 0.0679,
+      "step": 273
+    },
+    {
+      "epoch": 0.01096,
+      "grad_norm": 5.6090826988220215,
+      "learning_rate": 1.8231363685344422e-05,
+      "loss": 0.1978,
+      "step": 274
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 7.970758438110352,
+      "learning_rate": 1.821149209133704e-05,
+      "loss": 0.3133,
+      "step": 275
+    },
+    {
+      "epoch": 0.01104,
+      "grad_norm": 7.678709030151367,
+      "learning_rate": 1.819152044288992e-05,
+      "loss": 0.2256,
+      "step": 276
+    },
+    {
+      "epoch": 0.01108,
+      "grad_norm": 11.467633247375488,
+      "learning_rate": 1.8171448983351284e-05,
+      "loss": 0.3771,
+      "step": 277
+    },
+    {
+      "epoch": 0.01112,
+      "grad_norm": 7.438530921936035,
+      "learning_rate": 1.815127795728554e-05,
+      "loss": 0.2678,
+      "step": 278
+    },
+    {
+      "epoch": 0.01116,
+      "grad_norm": 5.009653091430664,
+      "learning_rate": 1.8131007610470278e-05,
+      "loss": 0.1911,
+      "step": 279
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 2.6366703510284424,
+      "learning_rate": 1.8110638189893267e-05,
+      "loss": 0.1541,
+      "step": 280
+    },
+    {
+      "epoch": 0.01124,
+      "grad_norm": 0.9270055294036865,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.1006,
+      "step": 281
+    },
+    {
+      "epoch": 0.01128,
+      "grad_norm": 6.129909515380859,
+      "learning_rate": 1.806960312143802e-05,
+      "loss": 0.3082,
+      "step": 282
+    },
+    {
+      "epoch": 0.01132,
+      "grad_norm": 15.900835037231445,
+      "learning_rate": 1.804893797355914e-05,
+      "loss": 0.3574,
+      "step": 283
+    },
+    {
+      "epoch": 0.01136,
+      "grad_norm": 6.833189487457275,
+      "learning_rate": 1.8028174751911147e-05,
+      "loss": 0.2222,
+      "step": 284
+    },
+    {
+      "epoch": 0.0114,
+      "grad_norm": 1.9276386499404907,
+      "learning_rate": 1.8007313709487334e-05,
+      "loss": 0.1127,
+      "step": 285
+    },
+    {
+      "epoch": 0.01144,
+      "grad_norm": 2.2078306674957275,
+      "learning_rate": 1.798635510047293e-05,
+      "loss": 0.0555,
+      "step": 286
+    },
+    {
+      "epoch": 0.01148,
+      "grad_norm": 0.7225556969642639,
+      "learning_rate": 1.7965299180241963e-05,
+      "loss": 0.0675,
+      "step": 287
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 12.13077449798584,
+      "learning_rate": 1.7944146205354182e-05,
+      "loss": 0.328,
+      "step": 288
+    },
+    {
+      "epoch": 0.01156,
+      "grad_norm": 0.6283919811248779,
+      "learning_rate": 1.792289643355191e-05,
+      "loss": 0.0411,
+      "step": 289
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 7.6308465003967285,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.217,
+      "step": 290
+    },
+    {
+      "epoch": 0.01164,
+      "grad_norm": 9.79549503326416,
+      "learning_rate": 1.788010753606722e-05,
+      "loss": 0.2951,
+      "step": 291
+    },
+    {
+      "epoch": 0.01168,
+      "grad_norm": 4.455563068389893,
+      "learning_rate": 1.785856893175402e-05,
+      "loss": 0.0836,
+      "step": 292
+    },
+    {
+      "epoch": 0.01172,
+      "grad_norm": 6.815781593322754,
+      "learning_rate": 1.78369345732584e-05,
+      "loss": 0.7854,
+      "step": 293
+    },
+    {
+      "epoch": 0.01176,
+      "grad_norm": 1.091526985168457,
+      "learning_rate": 1.781520472418819e-05,
+      "loss": 0.1367,
+      "step": 294
+    },
+    {
+      "epoch": 0.0118,
+      "grad_norm": 0.8816052079200745,
+      "learning_rate": 1.7793379649314743e-05,
+      "loss": 0.1076,
+      "step": 295
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 2.441382884979248,
+      "learning_rate": 1.777145961456971e-05,
+      "loss": 0.1332,
+      "step": 296
+    },
+    {
+      "epoch": 0.01188,
+      "grad_norm": 4.5978169441223145,
+      "learning_rate": 1.7749444887041797e-05,
+      "loss": 0.1715,
+      "step": 297
+    },
+    {
+      "epoch": 0.01192,
+      "grad_norm": 11.053872108459473,
+      "learning_rate": 1.7727335734973512e-05,
+      "loss": 0.3845,
+      "step": 298
+    },
+    {
+      "epoch": 0.01196,
+      "grad_norm": 1.0838901996612549,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.0815,
+      "step": 299
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 3.6972904205322266,
+      "learning_rate": 1.7682835235935236e-05,
+      "loss": 0.1373,
+      "step": 300
+    },
+    {
+      "epoch": 0.01204,
+      "grad_norm": 1.2566019296646118,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.1075,
+      "step": 301
+    },
+    {
+      "epoch": 0.01208,
+      "grad_norm": 16.18206024169922,
+      "learning_rate": 1.7637960286346423e-05,
+      "loss": 0.3086,
+      "step": 302
+    },
+    {
+      "epoch": 0.01212,
+      "grad_norm": 12.326071739196777,
+      "learning_rate": 1.761538307536737e-05,
+      "loss": 0.6085,
+      "step": 303
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 10.141122817993164,
+      "learning_rate": 1.759271307334881e-05,
+      "loss": 0.4891,
+      "step": 304
+    },
+    {
+      "epoch": 0.0122,
+      "grad_norm": 2.3961305618286133,
+      "learning_rate": 1.7569950556517566e-05,
+      "loss": 0.1271,
+      "step": 305
+    },
+    {
+      "epoch": 0.01224,
+      "grad_norm": 3.0086827278137207,
+      "learning_rate": 1.7547095802227723e-05,
+      "loss": 0.1524,
+      "step": 306
+    },
+    {
+      "epoch": 0.01228,
+      "grad_norm": 1.8383742570877075,
+      "learning_rate": 1.7524149088957244e-05,
+      "loss": 0.1216,
+      "step": 307
+    },
+    {
+      "epoch": 0.01232,
+      "grad_norm": 1.6366280317306519,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.1151,
+      "step": 308
+    },
+    {
+      "epoch": 0.01236,
+      "grad_norm": 1.5046234130859375,
+      "learning_rate": 1.747798090498532e-05,
+      "loss": 0.1401,
+      "step": 309
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 2.2363080978393555,
+      "learning_rate": 1.7454759996828622e-05,
+      "loss": 0.1242,
+      "step": 310
+    },
+    {
+      "epoch": 0.01244,
+      "grad_norm": 5.961032390594482,
+      "learning_rate": 1.7431448254773943e-05,
+      "loss": 0.2322,
+      "step": 311
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 10.631216049194336,
+      "learning_rate": 1.74080459628675e-05,
+      "loss": 0.3674,
+      "step": 312
+    },
+    {
+      "epoch": 0.01252,
+      "grad_norm": 9.324240684509277,
+      "learning_rate": 1.7384553406258842e-05,
+      "loss": 0.376,
+      "step": 313
+    },
+    {
+      "epoch": 0.01256,
+      "grad_norm": 1.2649765014648438,
+      "learning_rate": 1.7360970871197347e-05,
+      "loss": 0.1096,
+      "step": 314
+    },
+    {
+      "epoch": 0.0126,
+      "grad_norm": 3.9201526641845703,
+      "learning_rate": 1.7337298645028764e-05,
+      "loss": 0.1382,
+      "step": 315
+    },
+    {
+      "epoch": 0.01264,
+      "grad_norm": 6.2504401206970215,
+      "learning_rate": 1.7313537016191706e-05,
+      "loss": 0.2084,
+      "step": 316
+    },
+    {
+      "epoch": 0.01268,
+      "grad_norm": 4.236605167388916,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.0875,
+      "step": 317
+    },
+    {
+      "epoch": 0.01272,
+      "grad_norm": 0.5248899459838867,
+      "learning_rate": 1.7265746709709762e-05,
+      "loss": 0.0881,
+      "step": 318
+    },
+    {
+      "epoch": 0.01276,
+      "grad_norm": 0.6480566263198853,
+      "learning_rate": 1.7241718614374678e-05,
+      "loss": 0.0644,
+      "step": 319
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 5.103775501251221,
+      "learning_rate": 1.7217602280983622e-05,
+      "loss": 0.2212,
+      "step": 320
+    },
+    {
+      "epoch": 0.01284,
+      "grad_norm": 1.8865947723388672,
+      "learning_rate": 1.7193398003386514e-05,
+      "loss": 0.0471,
+      "step": 321
+    },
+    {
+      "epoch": 0.01288,
+      "grad_norm": 2.2529821395874023,
+      "learning_rate": 1.716910607650483e-05,
+      "loss": 0.1253,
+      "step": 322
+    },
+    {
+      "epoch": 0.01292,
+      "grad_norm": 7.786603927612305,
+      "learning_rate": 1.7144726796328034e-05,
+      "loss": 0.2494,
+      "step": 323
+    },
+    {
+      "epoch": 0.01296,
+      "grad_norm": 10.334938049316406,
+      "learning_rate": 1.712026045990997e-05,
+      "loss": 0.2968,
+      "step": 324
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 8.239156723022461,
+      "learning_rate": 1.709570736536521e-05,
+      "loss": 0.2887,
+      "step": 325
+    },
+    {
+      "epoch": 0.01304,
+      "grad_norm": 2.021263599395752,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.1213,
+      "step": 326
+    },
+    {
+      "epoch": 0.01308,
+      "grad_norm": 16.105384826660156,
+      "learning_rate": 1.7046342099635948e-05,
+      "loss": 0.3676,
+      "step": 327
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 4.270312309265137,
+      "learning_rate": 1.7021530529951627e-05,
+      "loss": 0.1804,
+      "step": 328
+    },
+    {
+      "epoch": 0.01316,
+      "grad_norm": 1.5432668924331665,
+      "learning_rate": 1.6996633405133656e-05,
+      "loss": 0.13,
+      "step": 329
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 7.5215253829956055,
+      "learning_rate": 1.697165102854565e-05,
+      "loss": 0.2706,
+      "step": 330
+    },
+    {
+      "epoch": 0.01324,
+      "grad_norm": 13.240067481994629,
+      "learning_rate": 1.6946583704589973e-05,
+      "loss": 0.3287,
+      "step": 331
+    },
+    {
+      "epoch": 0.01328,
+      "grad_norm": 4.955234050750732,
+      "learning_rate": 1.692143173870407e-05,
+      "loss": 0.2224,
+      "step": 332
+    },
+    {
+      "epoch": 0.01332,
+      "grad_norm": 16.788005828857422,
+      "learning_rate": 1.68961954373567e-05,
+      "loss": 0.2455,
+      "step": 333
+    },
+    {
+      "epoch": 0.01336,
+      "grad_norm": 1.0123594999313354,
+      "learning_rate": 1.6870875108044233e-05,
+      "loss": 0.1055,
+      "step": 334
+    },
+    {
+      "epoch": 0.0134,
+      "grad_norm": 1.1115672588348389,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.0187,
+      "step": 335
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 1.612168788909912,
+      "learning_rate": 1.6819983600624986e-05,
+      "loss": 0.081,
+      "step": 336
+    },
+    {
+      "epoch": 0.01348,
+      "grad_norm": 7.835407257080078,
+      "learning_rate": 1.6794413042615168e-05,
+      "loss": 0.2851,
+      "step": 337
+    },
+    {
+      "epoch": 0.01352,
+      "grad_norm": 2.5436980724334717,
+      "learning_rate": 1.6768759696826608e-05,
+      "loss": 0.1653,
+      "step": 338
+    },
+    {
+      "epoch": 0.01356,
+      "grad_norm": 5.707899570465088,
+      "learning_rate": 1.6743023875837233e-05,
+      "loss": 0.1823,
+      "step": 339
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 0.696250319480896,
+      "learning_rate": 1.6717205893229904e-05,
+      "loss": 0.0145,
+      "step": 340
+    },
+    {
+      "epoch": 0.01364,
+      "grad_norm": 4.433291912078857,
+      "learning_rate": 1.6691306063588583e-05,
+      "loss": 0.1817,
+      "step": 341
+    },
+    {
+      "epoch": 0.01368,
+      "grad_norm": 0.9749025106430054,
+      "learning_rate": 1.6665324702494524e-05,
+      "loss": 0.1085,
+      "step": 342
+    },
+    {
+      "epoch": 0.01372,
+      "grad_norm": 1.4184889793395996,
+      "learning_rate": 1.6639262126522417e-05,
+      "loss": 0.0931,
+      "step": 343
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 2.1656088829040527,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.1122,
+      "step": 344
+    },
+    {
+      "epoch": 0.0138,
+      "grad_norm": 1.4760959148406982,
+      "learning_rate": 1.6586894601186804e-05,
+      "loss": 0.1105,
+      "step": 345
+    },
+    {
+      "epoch": 0.01384,
+      "grad_norm": 2.5393295288085938,
+      "learning_rate": 1.6560590289905074e-05,
+      "loss": 0.1367,
+      "step": 346
+    },
+    {
+      "epoch": 0.01388,
+      "grad_norm": 1.2137281894683838,
+      "learning_rate": 1.6534206039901057e-05,
+      "loss": 0.0995,
+      "step": 347
+    },
+    {
+      "epoch": 0.01392,
+      "grad_norm": 1.2161641120910645,
+      "learning_rate": 1.650774217265851e-05,
+      "loss": 0.1006,
+      "step": 348
+    },
+    {
+      "epoch": 0.01396,
+      "grad_norm": 3.101379632949829,
+      "learning_rate": 1.6481199010631312e-05,
+      "loss": 0.0612,
+      "step": 349
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 10.090313911437988,
+      "learning_rate": 1.645457687723951e-05,
+      "loss": 0.391,
+      "step": 350
+    },
+    {
+      "epoch": 0.01404,
+      "grad_norm": 2.4455316066741943,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.1184,
+      "step": 351
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 2.614978075027466,
+      "learning_rate": 1.6401096994849558e-05,
+      "loss": 0.0363,
+      "step": 352
+    },
+    {
+      "epoch": 0.01412,
+      "grad_norm": 4.515564441680908,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.2126,
+      "step": 353
+    },
+    {
+      "epoch": 0.01416,
+      "grad_norm": 6.503922462463379,
+      "learning_rate": 1.6347305132022677e-05,
+      "loss": 0.1739,
+      "step": 354
+    },
+    {
+      "epoch": 0.0142,
+      "grad_norm": 8.025157928466797,
+      "learning_rate": 1.632029302664851e-05,
+      "loss": 0.2612,
+      "step": 355
+    },
+    {
+      "epoch": 0.01424,
+      "grad_norm": 7.73246955871582,
+      "learning_rate": 1.6293203910498375e-05,
+      "loss": 0.2134,
+      "step": 356
+    },
+    {
+      "epoch": 0.01428,
+      "grad_norm": 6.67537784576416,
+      "learning_rate": 1.6266038113644605e-05,
+      "loss": 0.2887,
+      "step": 357
+    },
+    {
+      "epoch": 0.01432,
+      "grad_norm": 4.257614612579346,
+      "learning_rate": 1.6238795967093865e-05,
+      "loss": 0.1448,
+      "step": 358
+    },
+    {
+      "epoch": 0.01436,
+      "grad_norm": 10.802968978881836,
+      "learning_rate": 1.6211477802783105e-05,
+      "loss": 0.4567,
+      "step": 359
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 2.800992965698242,
+      "learning_rate": 1.6184083953575543e-05,
+      "loss": 0.1205,
+      "step": 360
+    },
+    {
+      "epoch": 0.01444,
+      "grad_norm": 3.144157886505127,
+      "learning_rate": 1.6156614753256583e-05,
+      "loss": 0.0907,
+      "step": 361
+    },
+    {
+      "epoch": 0.01448,
+      "grad_norm": 1.405945897102356,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.0699,
+      "step": 362
+    },
+    {
+      "epoch": 0.01452,
+      "grad_norm": 9.864731788635254,
+      "learning_rate": 1.610145163901268e-05,
+      "loss": 0.126,
+      "step": 363
+    },
+    {
+      "epoch": 0.01456,
+      "grad_norm": 4.276960372924805,
+      "learning_rate": 1.607375839723287e-05,
+      "loss": 0.1457,
+      "step": 364
+    },
+    {
+      "epoch": 0.0146,
+      "grad_norm": 1.2770179510116577,
+      "learning_rate": 1.6045991148623752e-05,
+      "loss": 0.0905,
+      "step": 365
+    },
+    {
+      "epoch": 0.01464,
+      "grad_norm": 11.94617748260498,
+      "learning_rate": 1.6018150231520486e-05,
+      "loss": 0.3556,
+      "step": 366
+    },
+    {
+      "epoch": 0.01468,
+      "grad_norm": 0.9788459539413452,
+      "learning_rate": 1.599023598515586e-05,
+      "loss": 0.0974,
+      "step": 367
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 7.744279384613037,
+      "learning_rate": 1.5962248749656158e-05,
+      "loss": 0.509,
+      "step": 368
+    },
+    {
+      "epoch": 0.01476,
+      "grad_norm": 2.847423553466797,
+      "learning_rate": 1.5934188866037017e-05,
+      "loss": 0.1363,
+      "step": 369
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 2.7117257118225098,
+      "learning_rate": 1.5906056676199256e-05,
+      "loss": 0.1359,
+      "step": 370
+    },
+    {
+      "epoch": 0.01484,
+      "grad_norm": 3.2945730686187744,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.1607,
+      "step": 371
+    },
+    {
+      "epoch": 0.01488,
+      "grad_norm": 7.09609317779541,
+      "learning_rate": 1.584957674987216e-05,
+      "loss": 0.1365,
+      "step": 372
+    },
+    {
+      "epoch": 0.01492,
+      "grad_norm": 1.2029359340667725,
+      "learning_rate": 1.5821229701572897e-05,
+      "loss": 0.1317,
+      "step": 373
+    },
+    {
+      "epoch": 0.01496,
+      "grad_norm": 3.021914482116699,
+      "learning_rate": 1.5792811723426787e-05,
+      "loss": 0.1172,
+      "step": 374
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 8.222455978393555,
+      "learning_rate": 1.5764323161697933e-05,
+      "loss": 0.2918,
+      "step": 375
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 2.163900852203369,
+      "learning_rate": 1.573576436351046e-05,
+      "loss": 0.0959,
+      "step": 376
+    },
+    {
+      "epoch": 0.01508,
+      "grad_norm": 4.453228950500488,
+      "learning_rate": 1.570713567684432e-05,
+      "loss": 0.1367,
+      "step": 377
+    },
+    {
+      "epoch": 0.01512,
+      "grad_norm": 2.86348032951355,
+      "learning_rate": 1.5678437450531014e-05,
+      "loss": 0.127,
+      "step": 378
+    },
+    {
+      "epoch": 0.01516,
+      "grad_norm": 0.6201307773590088,
+      "learning_rate": 1.564967003424938e-05,
+      "loss": 0.0603,
+      "step": 379
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 2.7370893955230713,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.1117,
+      "step": 380
+    },
+    {
+      "epoch": 0.01524,
+      "grad_norm": 15.885140419006348,
+      "learning_rate": 1.5591929034707468e-05,
+      "loss": 0.3353,
+      "step": 381
+    },
+    {
+      "epoch": 0.01528,
+      "grad_norm": 6.125364780426025,
+      "learning_rate": 1.556295615500305e-05,
+      "loss": 0.1895,
+      "step": 382
+    },
+    {
+      "epoch": 0.01532,
+      "grad_norm": 15.375330924987793,
+      "learning_rate": 1.553391549243344e-05,
+      "loss": 1.2622,
+      "step": 383
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 1.6310198307037354,
+      "learning_rate": 1.5504807400849957e-05,
+      "loss": 0.1133,
+      "step": 384
+    },
+    {
+      "epoch": 0.0154,
+      "grad_norm": 1.1787599325180054,
+      "learning_rate": 1.5475632234925505e-05,
+      "loss": 0.1216,
+      "step": 385
+    },
+    {
+      "epoch": 0.01544,
+      "grad_norm": 7.484168529510498,
+      "learning_rate": 1.5446390350150272e-05,
+      "loss": 0.1681,
+      "step": 386
+    },
+    {
+      "epoch": 0.01548,
+      "grad_norm": 1.0399556159973145,
+      "learning_rate": 1.54170821028274e-05,
+      "loss": 0.059,
+      "step": 387
+    },
+    {
+      "epoch": 0.01552,
+      "grad_norm": 16.719999313354492,
+      "learning_rate": 1.5387707850068633e-05,
+      "loss": 0.5407,
+      "step": 388
+    },
+    {
+      "epoch": 0.01556,
+      "grad_norm": 2.338648796081543,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.2332,
+      "step": 389
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 0.7824026942253113,
+      "learning_rate": 1.53287627607073e-05,
+      "loss": 0.0787,
+      "step": 390
+    },
+    {
+      "epoch": 0.01564,
+      "grad_norm": 2.81449556350708,
+      "learning_rate": 1.529919264233205e-05,
+      "loss": 0.0492,
+      "step": 391
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 9.222359657287598,
+      "learning_rate": 1.5269557954966777e-05,
+      "loss": 0.2793,
+      "step": 392
+    },
+    {
+      "epoch": 0.01572,
+      "grad_norm": 9.849822998046875,
+      "learning_rate": 1.5239859059700794e-05,
+      "loss": 0.4214,
+      "step": 393
+    },
+    {
+      "epoch": 0.01576,
+      "grad_norm": 1.1515494585037231,
+      "learning_rate": 1.5210096318405768e-05,
+      "loss": 0.137,
+      "step": 394
+    },
+    {
+      "epoch": 0.0158,
+      "grad_norm": 4.948048114776611,
+      "learning_rate": 1.5180270093731305e-05,
+      "loss": 0.189,
+      "step": 395
+    },
+    {
+      "epoch": 0.01584,
+      "grad_norm": 18.73415756225586,
+      "learning_rate": 1.5150380749100545e-05,
+      "loss": 0.2962,
+      "step": 396
+    },
+    {
+      "epoch": 0.01588,
+      "grad_norm": 3.8207757472991943,
+      "learning_rate": 1.5120428648705716e-05,
+      "loss": 0.1395,
+      "step": 397
+    },
+    {
+      "epoch": 0.01592,
+      "grad_norm": 7.444143772125244,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.1513,
+      "step": 398
+    },
+    {
+      "epoch": 0.01596,
+      "grad_norm": 0.9845011830329895,
+      "learning_rate": 1.5060337641211637e-05,
+      "loss": 0.1072,
+      "step": 399
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 14.107025146484375,
+      "learning_rate": 1.5030199466302354e-05,
+      "loss": 0.8209,
+      "step": 400
+    },
+    {
+      "epoch": 0.01604,
+      "grad_norm": 10.212411880493164,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.207,
+      "step": 401
+    },
+    {
+      "epoch": 0.01608,
+      "grad_norm": 2.0816144943237305,
+      "learning_rate": 1.4969739610275556e-05,
+      "loss": 0.1234,
+      "step": 402
+    },
+    {
+      "epoch": 0.01612,
+      "grad_norm": 3.7055130004882812,
+      "learning_rate": 1.493941866584231e-05,
+      "loss": 0.0658,
+      "step": 403
+    },
+    {
+      "epoch": 0.01616,
+      "grad_norm": 1.1933560371398926,
+      "learning_rate": 1.490903753615141e-05,
+      "loss": 0.1129,
+      "step": 404
+    },
+    {
+      "epoch": 0.0162,
+      "grad_norm": 1.2127540111541748,
+      "learning_rate": 1.4878596591387329e-05,
+      "loss": 0.1205,
+      "step": 405
+    },
+    {
+      "epoch": 0.01624,
+      "grad_norm": 1.8149640560150146,
+      "learning_rate": 1.4848096202463373e-05,
+      "loss": 0.1175,
+      "step": 406
+    },
+    {
+      "epoch": 0.01628,
+      "grad_norm": 9.266016006469727,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.2018,
+      "step": 407
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 1.833652377128601,
+      "learning_rate": 1.478691857940607e-05,
+      "loss": 0.1219,
+      "step": 408
+    },
+    {
+      "epoch": 0.01636,
+      "grad_norm": 2.9415693283081055,
+      "learning_rate": 1.4756242090702756e-05,
+      "loss": 0.1044,
+      "step": 409
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 9.741125106811523,
+      "learning_rate": 1.4725507648690542e-05,
+      "loss": 0.2795,
+      "step": 410
+    },
+    {
+      "epoch": 0.01644,
+      "grad_norm": 20.006994247436523,
+      "learning_rate": 1.469471562785891e-05,
+      "loss": 0.4019,
+      "step": 411
+    },
+    {
+      "epoch": 0.01648,
+      "grad_norm": 8.991267204284668,
+      "learning_rate": 1.4663866403398915e-05,
+      "loss": 0.1919,
+      "step": 412
+    },
+    {
+      "epoch": 0.01652,
+      "grad_norm": 1.2791274785995483,
+      "learning_rate": 1.463296035119862e-05,
+      "loss": 0.095,
+      "step": 413
+    },
+    {
+      "epoch": 0.01656,
+      "grad_norm": 0.39519622921943665,
+      "learning_rate": 1.4601997847838518e-05,
+      "loss": 0.0077,
+      "step": 414
+    },
+    {
+      "epoch": 0.0166,
+      "grad_norm": 0.6712736487388611,
+      "learning_rate": 1.4570979270586944e-05,
+      "loss": 0.0748,
+      "step": 415
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 3.1821417808532715,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.1423,
+      "step": 416
+    },
+    {
+      "epoch": 0.01668,
+      "grad_norm": 0.980666995048523,
+      "learning_rate": 1.4508775406894308e-05,
+      "loss": 0.0969,
+      "step": 417
+    },
+    {
+      "epoch": 0.01672,
+      "grad_norm": 3.8596527576446533,
+      "learning_rate": 1.4477590878387697e-05,
+      "loss": 0.2292,
+      "step": 418
+    },
+    {
+      "epoch": 0.01676,
+      "grad_norm": 15.665862083435059,
+      "learning_rate": 1.4446351791849276e-05,
+      "loss": 0.5364,
+      "step": 419
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 0.7705150842666626,
+      "learning_rate": 1.4415058527917454e-05,
+      "loss": 0.0703,
+      "step": 420
+    },
+    {
+      "epoch": 0.01684,
+      "grad_norm": 1.7829031944274902,
+      "learning_rate": 1.4383711467890776e-05,
+      "loss": 0.1003,
+      "step": 421
+    },
+    {
+      "epoch": 0.01688,
+      "grad_norm": 7.487597942352295,
+      "learning_rate": 1.4352310993723277e-05,
+      "loss": 0.1571,
+      "step": 422
+    },
+    {
+      "epoch": 0.01692,
+      "grad_norm": 1.041498064994812,
+      "learning_rate": 1.4320857488019826e-05,
+      "loss": 0.0745,
+      "step": 423
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 0.9480231404304504,
+      "learning_rate": 1.4289351334031461e-05,
+      "loss": 0.1106,
+      "step": 424
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 1.3635735511779785,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.1136,
+      "step": 425
+    },
+    {
+      "epoch": 0.01704,
+      "grad_norm": 2.6413416862487793,
+      "learning_rate": 1.4226182617406996e-05,
+      "loss": 0.1346,
+      "step": 426
+    },
+    {
+      "epoch": 0.01708,
+      "grad_norm": 2.8069653511047363,
+      "learning_rate": 1.4194520824461773e-05,
+      "loss": 0.1236,
+      "step": 427
+    },
+    {
+      "epoch": 0.01712,
+      "grad_norm": 2.4564380645751953,
+      "learning_rate": 1.4162807922604014e-05,
+      "loss": 0.1155,
+      "step": 428
+    },
+    {
+      "epoch": 0.01716,
+      "grad_norm": 9.50247859954834,
+      "learning_rate": 1.413104429824542e-05,
+      "loss": 0.1943,
+      "step": 429
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 2.2159554958343506,
+      "learning_rate": 1.4099230338415728e-05,
+      "loss": 0.1143,
+      "step": 430
+    },
+    {
+      "epoch": 0.01724,
+      "grad_norm": 10.001408576965332,
+      "learning_rate": 1.4067366430758004e-05,
+      "loss": 0.1928,
+      "step": 431
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 1.7870951890945435,
+      "learning_rate": 1.4035452963523903e-05,
+      "loss": 0.1054,
+      "step": 432
+    },
+    {
+      "epoch": 0.01732,
+      "grad_norm": 2.54266095161438,
+      "learning_rate": 1.4003490325568953e-05,
+      "loss": 0.1075,
+      "step": 433
+    },
+    {
+      "epoch": 0.01736,
+      "grad_norm": 1.184970498085022,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.0486,
+      "step": 434
+    },
+    {
+      "epoch": 0.0174,
+      "grad_norm": 1.8117538690567017,
+      "learning_rate": 1.3939419095909513e-05,
+      "loss": 0.0974,
+      "step": 435
+    },
+    {
+      "epoch": 0.01744,
+      "grad_norm": 11.567070007324219,
+      "learning_rate": 1.3907311284892737e-05,
+      "loss": 0.4168,
+      "step": 436
+    },
+    {
+      "epoch": 0.01748,
+      "grad_norm": 3.473884344100952,
+      "learning_rate": 1.3875155864521031e-05,
+      "loss": 0.1063,
+      "step": 437
+    },
+    {
+      "epoch": 0.01752,
+      "grad_norm": 0.1846083551645279,
+      "learning_rate": 1.3842953226598036e-05,
+      "loss": 0.0042,
+      "step": 438
+    },
+    {
+      "epoch": 0.01756,
+      "grad_norm": 1.5330449342727661,
+      "learning_rate": 1.3810703763502744e-05,
+      "loss": 0.1073,
+      "step": 439
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 17.389793395996094,
+      "learning_rate": 1.3778407868184674e-05,
+      "loss": 0.3572,
+      "step": 440
+    },
+    {
+      "epoch": 0.01764,
+      "grad_norm": 25.607913970947266,
+      "learning_rate": 1.3746065934159123e-05,
+      "loss": 0.6083,
+      "step": 441
+    },
+    {
+      "epoch": 0.01768,
+      "grad_norm": 13.53175163269043,
+      "learning_rate": 1.371367835550235e-05,
+      "loss": 0.3103,
+      "step": 442
+    },
+    {
+      "epoch": 0.01772,
+      "grad_norm": 1.2025660276412964,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.0695,
+      "step": 443
+    },
+    {
+      "epoch": 0.01776,
+      "grad_norm": 10.705772399902344,
+      "learning_rate": 1.3648767843376196e-05,
+      "loss": 0.256,
+      "step": 444
+    },
+    {
+      "epoch": 0.0178,
+      "grad_norm": 10.151220321655273,
+      "learning_rate": 1.3616245700820922e-05,
+      "loss": 0.1938,
+      "step": 445
+    },
+    {
+      "epoch": 0.01784,
+      "grad_norm": 3.368771553039551,
+      "learning_rate": 1.3583679495453e-05,
+      "loss": 0.1439,
+      "step": 446
+    },
+    {
+      "epoch": 0.01788,
+      "grad_norm": 1.5903120040893555,
+      "learning_rate": 1.3551069624081372e-05,
+      "loss": 0.1046,
+      "step": 447
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 12.853551864624023,
+      "learning_rate": 1.3518416484047018e-05,
+      "loss": 0.2705,
+      "step": 448
+    },
+    {
+      "epoch": 0.01796,
+      "grad_norm": 2.696713924407959,
+      "learning_rate": 1.3485720473218153e-05,
+      "loss": 0.1508,
+      "step": 449
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 4.579776287078857,
+      "learning_rate": 1.3452981989985347e-05,
+      "loss": 0.0928,
+      "step": 450
+    },
+    {
+      "epoch": 0.01804,
+      "grad_norm": 7.6984148025512695,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.2432,
+      "step": 451
+    },
+    {
+      "epoch": 0.01808,
+      "grad_norm": 2.45931339263916,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.0976,
+      "step": 452
+    },
+    {
+      "epoch": 0.01812,
+      "grad_norm": 1.6402121782302856,
+      "learning_rate": 1.3354515697502552e-05,
+      "loss": 0.1266,
+      "step": 453
+    },
+    {
+      "epoch": 0.01816,
+      "grad_norm": 13.451138496398926,
+      "learning_rate": 1.3321611318837033e-05,
+      "loss": 0.4544,
+      "step": 454
+    },
+    {
+      "epoch": 0.0182,
+      "grad_norm": 5.332578659057617,
+      "learning_rate": 1.3288666467385834e-05,
+      "loss": 0.1549,
+      "step": 455
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 2.333383560180664,
+      "learning_rate": 1.3255681544571568e-05,
+      "loss": 0.038,
+      "step": 456
+    },
+    {
+      "epoch": 0.01828,
+      "grad_norm": 3.0501410961151123,
+      "learning_rate": 1.3222656952305113e-05,
+      "loss": 0.1489,
+      "step": 457
+    },
+    {
+      "epoch": 0.01832,
+      "grad_norm": 6.710490703582764,
+      "learning_rate": 1.3189593092980701e-05,
+      "loss": 0.1758,
+      "step": 458
+    },
+    {
+      "epoch": 0.01836,
+      "grad_norm": 7.851617813110352,
+      "learning_rate": 1.3156490369471026e-05,
+      "loss": 0.178,
+      "step": 459
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 9.321209907531738,
+      "learning_rate": 1.3123349185122328e-05,
+      "loss": 0.2154,
+      "step": 460
+    },
+    {
+      "epoch": 0.01844,
+      "grad_norm": 0.979800283908844,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.0897,
+      "step": 461
+    },
+    {
+      "epoch": 0.01848,
+      "grad_norm": 12.141875267028809,
+      "learning_rate": 1.3056953049631059e-05,
+      "loss": 0.2597,
+      "step": 462
+    },
+    {
+      "epoch": 0.01852,
+      "grad_norm": 2.0872366428375244,
+      "learning_rate": 1.3023698907504447e-05,
+      "loss": 0.1225,
+      "step": 463
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 1.9920201301574707,
+      "learning_rate": 1.2990407922560869e-05,
+      "loss": 0.0584,
+      "step": 464
+    },
+    {
+      "epoch": 0.0186,
+      "grad_norm": 4.644482612609863,
+      "learning_rate": 1.2957080500440469e-05,
+      "loss": 0.227,
+      "step": 465
+    },
+    {
+      "epoch": 0.01864,
+      "grad_norm": 1.1247519254684448,
+      "learning_rate": 1.2923717047227368e-05,
+      "loss": 0.0653,
+      "step": 466
+    },
+    {
+      "epoch": 0.01868,
+      "grad_norm": 0.6912455558776855,
+      "learning_rate": 1.2890317969444716e-05,
+      "loss": 0.0936,
+      "step": 467
+    },
+    {
+      "epoch": 0.01872,
+      "grad_norm": 7.379387855529785,
+      "learning_rate": 1.2856883674049736e-05,
+      "loss": 0.2444,
+      "step": 468
+    },
+    {
+      "epoch": 0.01876,
+      "grad_norm": 9.36325740814209,
+      "learning_rate": 1.2823414568428767e-05,
+      "loss": 0.2212,
+      "step": 469
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 3.7019126415252686,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.1415,
+      "step": 470
+    },
+    {
+      "epoch": 0.01884,
+      "grad_norm": 8.02016830444336,
+      "learning_rate": 1.2756373558169992e-05,
+      "loss": 0.3409,
+      "step": 471
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 0.9227790236473083,
+      "learning_rate": 1.2722802470405744e-05,
+      "loss": 0.0719,
+      "step": 472
+    },
+    {
+      "epoch": 0.01892,
+      "grad_norm": 5.344834804534912,
+      "learning_rate": 1.2689198206152657e-05,
+      "loss": 0.3277,
+      "step": 473
+    },
+    {
+      "epoch": 0.01896,
+      "grad_norm": 2.111518144607544,
+      "learning_rate": 1.265556117486809e-05,
+      "loss": 0.0882,
+      "step": 474
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 5.066408634185791,
+      "learning_rate": 1.2621891786408648e-05,
+      "loss": 0.1565,
+      "step": 475
+    },
+    {
+      "epoch": 0.01904,
+      "grad_norm": 8.867663383483887,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.221,
+      "step": 476
+    },
+    {
+      "epoch": 0.01908,
+      "grad_norm": 5.5946221351623535,
+      "learning_rate": 1.2554457579357906e-05,
+      "loss": 0.1938,
+      "step": 477
+    },
+    {
+      "epoch": 0.01912,
+      "grad_norm": 0.7726828455924988,
+      "learning_rate": 1.252069358243114e-05,
+      "loss": 0.1011,
+      "step": 478
+    },
+    {
+      "epoch": 0.01916,
+      "grad_norm": 6.071194171905518,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.147,
+      "step": 479
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 9.249407768249512,
+      "learning_rate": 1.2453073858788027e-05,
+      "loss": 0.2743,
+      "step": 480
+    },
+    {
+      "epoch": 0.01924,
+      "grad_norm": 4.825432300567627,
+      "learning_rate": 1.2419218955996677e-05,
+      "loss": 0.1294,
+      "step": 481
+    },
+    {
+      "epoch": 0.01928,
+      "grad_norm": 4.636416912078857,
+      "learning_rate": 1.238533457578581e-05,
+      "loss": 0.1032,
+      "step": 482
+    },
+    {
+      "epoch": 0.01932,
+      "grad_norm": 17.063302993774414,
+      "learning_rate": 1.23514211310259e-05,
+      "loss": 0.3295,
+      "step": 483
+    },
+    {
+      "epoch": 0.01936,
+      "grad_norm": 2.871009349822998,
+      "learning_rate": 1.2317479034941572e-05,
+      "loss": 0.1939,
+      "step": 484
+    },
+    {
+      "epoch": 0.0194,
+      "grad_norm": 5.303359508514404,
+      "learning_rate": 1.2283508701106559e-05,
+      "loss": 0.1797,
+      "step": 485
+    },
+    {
+      "epoch": 0.01944,
+      "grad_norm": 3.2623231410980225,
+      "learning_rate": 1.2249510543438652e-05,
+      "loss": 0.1816,
+      "step": 486
+    },
+    {
+      "epoch": 0.01948,
+      "grad_norm": 4.530101776123047,
+      "learning_rate": 1.2215484976194675e-05,
+      "loss": 0.1831,
+      "step": 487
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 9.881510734558105,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.3282,
+      "step": 488
+    },
+    {
+      "epoch": 0.01956,
+      "grad_norm": 7.159479141235352,
+      "learning_rate": 1.2147353271670634e-05,
+      "loss": 0.2609,
+      "step": 489
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 21.535446166992188,
+      "learning_rate": 1.211324796455389e-05,
+      "loss": 0.8823,
+      "step": 490
+    },
+    {
+      "epoch": 0.01964,
+      "grad_norm": 0.29867467284202576,
+      "learning_rate": 1.2079116908177592e-05,
+      "loss": 0.0245,
+      "step": 491
+    },
+    {
+      "epoch": 0.01968,
+      "grad_norm": 1.0982681512832642,
+      "learning_rate": 1.2044960518417902e-05,
+      "loss": 0.101,
+      "step": 492
+    },
+    {
+      "epoch": 0.01972,
+      "grad_norm": 1.9848767518997192,
+      "learning_rate": 1.2010779211459649e-05,
+      "loss": 0.117,
+      "step": 493
+    },
+    {
+      "epoch": 0.01976,
+      "grad_norm": 0.7052813172340393,
+      "learning_rate": 1.1976573403791263e-05,
+      "loss": 0.1035,
+      "step": 494
+    },
+    {
+      "epoch": 0.0198,
+      "grad_norm": 2.783144235610962,
+      "learning_rate": 1.194234351219972e-05,
+      "loss": 0.131,
+      "step": 495
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 6.000821113586426,
+      "learning_rate": 1.190808995376545e-05,
+      "loss": 0.2743,
+      "step": 496
+    },
+    {
+      "epoch": 0.01988,
+      "grad_norm": 8.017651557922363,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.1648,
+      "step": 497
+    },
+    {
+      "epoch": 0.01992,
+      "grad_norm": 1.5148141384124756,
+      "learning_rate": 1.1839513506127202e-05,
+      "loss": 0.086,
+      "step": 498
+    },
+    {
+      "epoch": 0.01996,
+      "grad_norm": 1.072095513343811,
+      "learning_rate": 1.1805191452505602e-05,
+      "loss": 0.1358,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0247082710266113,
+      "learning_rate": 1.1770847403195836e-05,
+      "loss": 0.0278,
+      "step": 500
+    },
+    {
+      "epoch": 0.02004,
+      "grad_norm": 7.3399577140808105,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.7392,
+      "step": 501
+    },
+    {
+      "epoch": 0.02008,
+      "grad_norm": 0.9901246428489685,
+      "learning_rate": 1.1702094991660326e-05,
+      "loss": 0.0652,
+      "step": 502
+    },
+    {
+      "epoch": 0.02012,
+      "grad_norm": 3.5104010105133057,
+      "learning_rate": 1.1667687467161025e-05,
+      "loss": 0.0931,
+      "step": 503
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 0.9414449334144592,
+      "learning_rate": 1.1633259622416224e-05,
+      "loss": 0.0911,
+      "step": 504
+    },
+    {
+      "epoch": 0.0202,
+      "grad_norm": 2.5189425945281982,
+      "learning_rate": 1.159881187691835e-05,
+      "loss": 0.0971,
+      "step": 505
+    },
+    {
+      "epoch": 0.02024,
+      "grad_norm": 1.3258261680603027,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.1026,
+      "step": 506
+    },
+    {
+      "epoch": 0.02028,
+      "grad_norm": 1.0318282842636108,
+      "learning_rate": 1.1529858362840383e-05,
+      "loss": 0.1053,
+      "step": 507
+    },
+    {
+      "epoch": 0.02032,
+      "grad_norm": 10.315982818603516,
+      "learning_rate": 1.1495353434437098e-05,
+      "loss": 0.3423,
+      "step": 508
+    },
+    {
+      "epoch": 0.02036,
+      "grad_norm": 6.957205772399902,
+      "learning_rate": 1.1460830285624119e-05,
+      "loss": 0.156,
+      "step": 509
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 7.96408224105835,
+      "learning_rate": 1.1426289337055119e-05,
+      "loss": 0.1663,
+      "step": 510
+    },
+    {
+      "epoch": 0.02044,
+      "grad_norm": 9.338637351989746,
+      "learning_rate": 1.1391731009600655e-05,
+      "loss": 0.2816,
+      "step": 511
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 1.9438921213150024,
+      "learning_rate": 1.1357155724343046e-05,
+      "loss": 0.1093,
+      "step": 512
+    },
+    {
+      "epoch": 0.02052,
+      "grad_norm": 4.294856071472168,
+      "learning_rate": 1.1322563902571227e-05,
+      "loss": 0.167,
+      "step": 513
+    },
+    {
+      "epoch": 0.02056,
+      "grad_norm": 20.368297576904297,
+      "learning_rate": 1.128795596577563e-05,
+      "loss": 0.5439,
+      "step": 514
+    },
+    {
+      "epoch": 0.0206,
+      "grad_norm": 12.038905143737793,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.5405,
+      "step": 515
+    },
+    {
+      "epoch": 0.02064,
+      "grad_norm": 1.97911536693573,
+      "learning_rate": 1.1218693434051475e-05,
+      "loss": 0.1148,
+      "step": 516
+    },
+    {
+      "epoch": 0.02068,
+      "grad_norm": 2.0836338996887207,
+      "learning_rate": 1.1184039683065014e-05,
+      "loss": 0.1164,
+      "step": 517
+    },
+    {
+      "epoch": 0.02072,
+      "grad_norm": 0.7486438751220703,
+      "learning_rate": 1.1149371504928667e-05,
+      "loss": 0.0099,
+      "step": 518
+    },
+    {
+      "epoch": 0.02076,
+      "grad_norm": 2.7294554710388184,
+      "learning_rate": 1.1114689322063255e-05,
+      "loss": 0.1469,
+      "step": 519
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 3.5850844383239746,
+      "learning_rate": 1.1079993557060228e-05,
+      "loss": 0.1371,
+      "step": 520
+    },
+    {
+      "epoch": 0.02084,
+      "grad_norm": 4.646721839904785,
+      "learning_rate": 1.1045284632676535e-05,
+      "loss": 0.1608,
+      "step": 521
+    },
+    {
+      "epoch": 0.02088,
+      "grad_norm": 1.6841893196105957,
+      "learning_rate": 1.1010562971829464e-05,
+      "loss": 0.0891,
+      "step": 522
+    },
+    {
+      "epoch": 0.02092,
+      "grad_norm": 1.646068811416626,
+      "learning_rate": 1.0975828997591496e-05,
+      "loss": 0.1298,
+      "step": 523
+    },
+    {
+      "epoch": 0.02096,
+      "grad_norm": 2.7768328189849854,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.114,
+      "step": 524
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 5.9844536781311035,
+      "learning_rate": 1.0906325801977804e-05,
+      "loss": 0.2891,
+      "step": 525
+    },
+    {
+      "epoch": 0.02104,
+      "grad_norm": 1.6219875812530518,
+      "learning_rate": 1.0871557427476585e-05,
+      "loss": 0.0919,
+      "step": 526
+    },
+    {
+      "epoch": 0.02108,
+      "grad_norm": 0.9565569758415222,
+      "learning_rate": 1.083677843332316e-05,
+      "loss": 0.0793,
+      "step": 527
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 9.714432716369629,
+      "learning_rate": 1.0801989243288588e-05,
+      "loss": 0.2615,
+      "step": 528
+    },
+    {
+      "epoch": 0.02116,
+      "grad_norm": 0.9550947546958923,
+      "learning_rate": 1.0767190281268187e-05,
+      "loss": 0.1092,
+      "step": 529
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 7.786430358886719,
+      "learning_rate": 1.0732381971276318e-05,
+      "loss": 0.2462,
+      "step": 530
+    },
+    {
+      "epoch": 0.02124,
+      "grad_norm": 12.679776191711426,
+      "learning_rate": 1.0697564737441254e-05,
+      "loss": 0.357,
+      "step": 531
+    },
+    {
+      "epoch": 0.02128,
+      "grad_norm": 2.85044264793396,
+      "learning_rate": 1.0662739004000005e-05,
+      "loss": 0.1231,
+      "step": 532
+    },
+    {
+      "epoch": 0.02132,
+      "grad_norm": 4.615429878234863,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.206,
+      "step": 533
+    },
+    {
+      "epoch": 0.02136,
+      "grad_norm": 1.797175407409668,
+      "learning_rate": 1.0593063735759619e-05,
+      "loss": 0.1069,
+      "step": 534
+    },
+    {
+      "epoch": 0.0214,
+      "grad_norm": 0.8161718845367432,
+      "learning_rate": 1.055821504993164e-05,
+      "loss": 0.0704,
+      "step": 535
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 2.111077308654785,
+      "learning_rate": 1.0523359562429441e-05,
+      "loss": 0.1317,
+      "step": 536
+    },
+    {
+      "epoch": 0.02148,
+      "grad_norm": 0.35279664397239685,
+      "learning_rate": 1.0488497697956134e-05,
+      "loss": 0.0048,
+      "step": 537
+    },
+    {
+      "epoch": 0.02152,
+      "grad_norm": 19.513723373413086,
+      "learning_rate": 1.0453629881292537e-05,
+      "loss": 0.3902,
+      "step": 538
+    },
+    {
+      "epoch": 0.02156,
+      "grad_norm": 1.5410431623458862,
+      "learning_rate": 1.0418756537291996e-05,
+      "loss": 0.0792,
+      "step": 539
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 1.479876160621643,
+      "learning_rate": 1.03838780908752e-05,
+      "loss": 0.0912,
+      "step": 540
+    },
+    {
+      "epoch": 0.02164,
+      "grad_norm": 9.28445053100586,
+      "learning_rate": 1.0348994967025012e-05,
+      "loss": 0.3597,
+      "step": 541
+    },
+    {
+      "epoch": 0.02168,
+      "grad_norm": 0.9638659358024597,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.0996,
+      "step": 542
+    },
+    {
+      "epoch": 0.02172,
+      "grad_norm": 4.64767599105835,
+      "learning_rate": 1.0279216387235691e-05,
+      "loss": 0.1641,
+      "step": 543
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 1.1675864458084106,
+      "learning_rate": 1.0244321781526533e-05,
+      "loss": 0.093,
+      "step": 544
+    },
+    {
+      "epoch": 0.0218,
+      "grad_norm": 0.7410765886306763,
+      "learning_rate": 1.0209424198833571e-05,
+      "loss": 0.0817,
+      "step": 545
+    },
+    {
+      "epoch": 0.02184,
+      "grad_norm": 1.4848532676696777,
+      "learning_rate": 1.0174524064372837e-05,
+      "loss": 0.1527,
+      "step": 546
+    },
+    {
+      "epoch": 0.02188,
+      "grad_norm": 7.588745594024658,
+      "learning_rate": 1.0139621803391454e-05,
+      "loss": 0.1814,
+      "step": 547
+    },
+    {
+      "epoch": 0.02192,
+      "grad_norm": 2.674877166748047,
+      "learning_rate": 1.010471784116246e-05,
+      "loss": 0.1183,
+      "step": 548
+    },
+    {
+      "epoch": 0.02196,
+      "grad_norm": 1.3598262071609497,
+      "learning_rate": 1.0069812602979617e-05,
+      "loss": 0.1129,
+      "step": 549
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 12.357645988464355,
+      "learning_rate": 1.0034906514152239e-05,
+      "loss": 0.3047,
+      "step": 550
+    },
+    {
+      "epoch": 0.02204,
+      "grad_norm": 2.2860231399536133,
+      "learning_rate": 1e-05,
+      "loss": 0.0891,
+      "step": 551
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 5.943870544433594,
+      "learning_rate": 9.965093485847766e-06,
+      "loss": 0.139,
+      "step": 552
+    },
+    {
+      "epoch": 0.02212,
+      "grad_norm": 0.9945041537284851,
+      "learning_rate": 9.930187397020385e-06,
+      "loss": 0.0761,
+      "step": 553
+    },
+    {
+      "epoch": 0.02216,
+      "grad_norm": 2.6000266075134277,
+      "learning_rate": 9.895282158837545e-06,
+      "loss": 0.1614,
+      "step": 554
+    },
+    {
+      "epoch": 0.0222,
+      "grad_norm": 1.6059635877609253,
+      "learning_rate": 9.860378196608549e-06,
+      "loss": 0.1272,
+      "step": 555
+    },
+    {
+      "epoch": 0.02224,
+      "grad_norm": 3.306328296661377,
+      "learning_rate": 9.825475935627165e-06,
+      "loss": 0.1576,
+      "step": 556
+    },
+    {
+      "epoch": 0.02228,
+      "grad_norm": 3.9381494522094727,
+      "learning_rate": 9.790575801166432e-06,
+      "loss": 0.1474,
+      "step": 557
+    },
+    {
+      "epoch": 0.02232,
+      "grad_norm": 0.29684770107269287,
+      "learning_rate": 9.75567821847347e-06,
+      "loss": 0.0055,
+      "step": 558
+    },
+    {
+      "epoch": 0.02236,
+      "grad_norm": 0.35691460967063904,
+      "learning_rate": 9.720783612764314e-06,
+      "loss": 0.0372,
+      "step": 559
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 1.2079758644104004,
+      "learning_rate": 9.685892409218718e-06,
+      "loss": 0.1032,
+      "step": 560
+    },
+    {
+      "epoch": 0.02244,
+      "grad_norm": 0.7328388094902039,
+      "learning_rate": 9.651005032974994e-06,
+      "loss": 0.0831,
+      "step": 561
+    },
+    {
+      "epoch": 0.02248,
+      "grad_norm": 3.2318665981292725,
+      "learning_rate": 9.616121909124801e-06,
+      "loss": 0.108,
+      "step": 562
+    },
+    {
+      "epoch": 0.02252,
+      "grad_norm": 0.8513982892036438,
+      "learning_rate": 9.581243462708007e-06,
+      "loss": 0.0662,
+      "step": 563
+    },
+    {
+      "epoch": 0.02256,
+      "grad_norm": 1.6477575302124023,
+      "learning_rate": 9.546370118707463e-06,
+      "loss": 0.1071,
+      "step": 564
+    },
+    {
+      "epoch": 0.0226,
+      "grad_norm": 0.9175444841384888,
+      "learning_rate": 9.511502302043867e-06,
+      "loss": 0.0801,
+      "step": 565
+    },
+    {
+      "epoch": 0.02264,
+      "grad_norm": 2.454674005508423,
+      "learning_rate": 9.476640437570562e-06,
+      "loss": 0.1216,
+      "step": 566
+    },
+    {
+      "epoch": 0.02268,
+      "grad_norm": 1.8718379735946655,
+      "learning_rate": 9.441784950068362e-06,
+      "loss": 0.0839,
+      "step": 567
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 0.1486193984746933,
+      "learning_rate": 9.406936264240386e-06,
+      "loss": 0.0035,
+      "step": 568
+    },
+    {
+      "epoch": 0.02276,
+      "grad_norm": 0.9708048105239868,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.0841,
+      "step": 569
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 7.86513614654541,
+      "learning_rate": 9.337260996000002e-06,
+      "loss": 0.1566,
+      "step": 570
+    },
+    {
+      "epoch": 0.02284,
+      "grad_norm": 0.23414605855941772,
+      "learning_rate": 9.302435262558748e-06,
+      "loss": 0.0048,
+      "step": 571
+    },
+    {
+      "epoch": 0.02288,
+      "grad_norm": 21.308835983276367,
+      "learning_rate": 9.267618028723687e-06,
+      "loss": 0.2941,
+      "step": 572
+    },
+    {
+      "epoch": 0.02292,
+      "grad_norm": 8.869210243225098,
+      "learning_rate": 9.232809718731815e-06,
+      "loss": 0.5089,
+      "step": 573
+    },
+    {
+      "epoch": 0.02296,
+      "grad_norm": 8.507433891296387,
+      "learning_rate": 9.198010756711413e-06,
+      "loss": 0.3288,
+      "step": 574
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 11.346800804138184,
+      "learning_rate": 9.163221566676847e-06,
+      "loss": 0.2901,
+      "step": 575
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.159308522939682,
+      "learning_rate": 9.128442572523418e-06,
+      "loss": 0.0019,
+      "step": 576
+    },
+    {
+      "epoch": 0.02308,
+      "grad_norm": 1.3349303007125854,
+      "learning_rate": 9.093674198022201e-06,
+      "loss": 0.1053,
+      "step": 577
+    },
+    {
+      "epoch": 0.02312,
+      "grad_norm": 1.441536545753479,
+      "learning_rate": 9.058916866814857e-06,
+      "loss": 0.1309,
+      "step": 578
+    },
+    {
+      "epoch": 0.02316,
+      "grad_norm": 4.523565292358398,
+      "learning_rate": 9.024171002408507e-06,
+      "loss": 0.078,
+      "step": 579
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 0.36291375756263733,
+      "learning_rate": 8.989437028170537e-06,
+      "loss": 0.0051,
+      "step": 580
+    },
+    {
+      "epoch": 0.02324,
+      "grad_norm": 1.2753270864486694,
+      "learning_rate": 8.954715367323468e-06,
+      "loss": 0.0985,
+      "step": 581
+    },
+    {
+      "epoch": 0.02328,
+      "grad_norm": 2.048231840133667,
+      "learning_rate": 8.920006442939772e-06,
+      "loss": 0.0299,
+      "step": 582
+    },
+    {
+      "epoch": 0.02332,
+      "grad_norm": 1.875231385231018,
+      "learning_rate": 8.885310677936746e-06,
+      "loss": 0.1038,
+      "step": 583
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 4.417097568511963,
+      "learning_rate": 8.850628495071336e-06,
+      "loss": 0.1113,
+      "step": 584
+    },
+    {
+      "epoch": 0.0234,
+      "grad_norm": 2.2510056495666504,
+      "learning_rate": 8.815960316934991e-06,
+      "loss": 0.154,
+      "step": 585
+    },
+    {
+      "epoch": 0.02344,
+      "grad_norm": 7.883542537689209,
+      "learning_rate": 8.781306565948528e-06,
+      "loss": 0.3247,
+      "step": 586
+    },
+    {
+      "epoch": 0.02348,
+      "grad_norm": 0.9818063974380493,
+      "learning_rate": 8.746667664356957e-06,
+      "loss": 0.0721,
+      "step": 587
+    },
+    {
+      "epoch": 0.02352,
+      "grad_norm": 1.361670970916748,
+      "learning_rate": 8.712044034224374e-06,
+      "loss": 0.0981,
+      "step": 588
+    },
+    {
+      "epoch": 0.02356,
+      "grad_norm": 0.7936698198318481,
+      "learning_rate": 8.677436097428775e-06,
+      "loss": 0.0742,
+      "step": 589
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 1.3088947534561157,
+      "learning_rate": 8.642844275656957e-06,
+      "loss": 0.1073,
+      "step": 590
+    },
+    {
+      "epoch": 0.02364,
+      "grad_norm": 6.919236183166504,
+      "learning_rate": 8.60826899039935e-06,
+      "loss": 0.2257,
+      "step": 591
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 1.6464182138442993,
+      "learning_rate": 8.573710662944884e-06,
+      "loss": 0.0871,
+      "step": 592
+    },
+    {
+      "epoch": 0.02372,
+      "grad_norm": 2.823197364807129,
+      "learning_rate": 8.539169714375885e-06,
+      "loss": 0.1271,
+      "step": 593
+    },
+    {
+      "epoch": 0.02376,
+      "grad_norm": 4.90531063079834,
+      "learning_rate": 8.504646565562907e-06,
+      "loss": 0.0644,
+      "step": 594
+    },
+    {
+      "epoch": 0.0238,
+      "grad_norm": 13.575413703918457,
+      "learning_rate": 8.47014163715962e-06,
+      "loss": 0.293,
+      "step": 595
+    },
+    {
+      "epoch": 0.02384,
+      "grad_norm": 7.615997314453125,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.1722,
+      "step": 596
+    },
+    {
+      "epoch": 0.02388,
+      "grad_norm": 5.033782482147217,
+      "learning_rate": 8.401188123081653e-06,
+      "loss": 0.1331,
+      "step": 597
+    },
+    {
+      "epoch": 0.02392,
+      "grad_norm": 0.8236892819404602,
+      "learning_rate": 8.366740377583781e-06,
+      "loss": 0.0988,
+      "step": 598
+    },
+    {
+      "epoch": 0.02396,
+      "grad_norm": 1.3046694993972778,
+      "learning_rate": 8.332312532838978e-06,
+      "loss": 0.0971,
+      "step": 599
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 2.7814369201660156,
+      "learning_rate": 8.297905008339677e-06,
+      "loss": 0.0838,
+      "step": 600
+    },
+    {
+      "epoch": 0.02404,
+      "grad_norm": 1.5740402936935425,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.1332,
+      "step": 601
+    },
+    {
+      "epoch": 0.02408,
+      "grad_norm": 12.72924518585205,
+      "learning_rate": 8.22915259680417e-06,
+      "loss": 0.491,
+      "step": 602
+    },
+    {
+      "epoch": 0.02412,
+      "grad_norm": 0.7020381689071655,
+      "learning_rate": 8.194808547494401e-06,
+      "loss": 0.0872,
+      "step": 603
+    },
+    {
+      "epoch": 0.02416,
+      "grad_norm": 13.676801681518555,
+      "learning_rate": 8.1604864938728e-06,
+      "loss": 0.607,
+      "step": 604
+    },
+    {
+      "epoch": 0.0242,
+      "grad_norm": 2.378636598587036,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.0269,
+      "step": 605
+    },
+    {
+      "epoch": 0.02424,
+      "grad_norm": 1.9276937246322632,
+      "learning_rate": 8.091910046234552e-06,
+      "loss": 0.0325,
+      "step": 606
+    },
+    {
+      "epoch": 0.02428,
+      "grad_norm": 1.8479126691818237,
+      "learning_rate": 8.057656487800283e-06,
+      "loss": 0.0191,
+      "step": 607
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 6.009891033172607,
+      "learning_rate": 8.023426596208739e-06,
+      "loss": 0.0646,
+      "step": 608
+    },
+    {
+      "epoch": 0.02436,
+      "grad_norm": 3.089521884918213,
+      "learning_rate": 7.989220788540356e-06,
+      "loss": 0.0337,
+      "step": 609
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 5.805949687957764,
+      "learning_rate": 7.955039481582098e-06,
+      "loss": 0.1741,
+      "step": 610
+    },
+    {
+      "epoch": 0.02444,
+      "grad_norm": 2.9937453269958496,
+      "learning_rate": 7.92088309182241e-06,
+      "loss": 0.1221,
+      "step": 611
+    },
+    {
+      "epoch": 0.02448,
+      "grad_norm": 5.99016809463501,
+      "learning_rate": 7.886752035446116e-06,
+      "loss": 0.163,
+      "step": 612
+    },
+    {
+      "epoch": 0.02452,
+      "grad_norm": 5.059927940368652,
+      "learning_rate": 7.852646728329368e-06,
+      "loss": 0.1345,
+      "step": 613
+    },
+    {
+      "epoch": 0.02456,
+      "grad_norm": 1.012635588645935,
+      "learning_rate": 7.818567586034578e-06,
+      "loss": 0.0862,
+      "step": 614
+    },
+    {
+      "epoch": 0.0246,
+      "grad_norm": 3.2993364334106445,
+      "learning_rate": 7.784515023805328e-06,
+      "loss": 0.1519,
+      "step": 615
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 7.147036075592041,
+      "learning_rate": 7.750489456561351e-06,
+      "loss": 0.2091,
+      "step": 616
+    },
+    {
+      "epoch": 0.02468,
+      "grad_norm": 1.7351012229919434,
+      "learning_rate": 7.716491298893443e-06,
+      "loss": 0.1221,
+      "step": 617
+    },
+    {
+      "epoch": 0.02472,
+      "grad_norm": 1.5768803358078003,
+      "learning_rate": 7.68252096505843e-06,
+      "loss": 0.086,
+      "step": 618
+    },
+    {
+      "epoch": 0.02476,
+      "grad_norm": 1.7902190685272217,
+      "learning_rate": 7.6485788689741e-06,
+      "loss": 0.0229,
+      "step": 619
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 1.1550565958023071,
+      "learning_rate": 7.6146654242141935e-06,
+      "loss": 0.0826,
+      "step": 620
+    },
+    {
+      "epoch": 0.02484,
+      "grad_norm": 3.248624086380005,
+      "learning_rate": 7.580781044003324e-06,
+      "loss": 0.1789,
+      "step": 621
+    },
+    {
+      "epoch": 0.02488,
+      "grad_norm": 0.29860222339630127,
+      "learning_rate": 7.546926141211975e-06,
+      "loss": 0.0323,
+      "step": 622
+    },
+    {
+      "epoch": 0.02492,
+      "grad_norm": 0.6060212850570679,
+      "learning_rate": 7.513101128351454e-06,
+      "loss": 0.007,
+      "step": 623
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 1.0408265590667725,
+      "learning_rate": 7.4793064175688635e-06,
+      "loss": 0.0655,
+      "step": 624
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 1.2011998891830444,
+      "learning_rate": 7.445542420642097e-06,
+      "loss": 0.087,
+      "step": 625
+    },
+    {
+      "epoch": 0.02504,
+      "grad_norm": 4.768725872039795,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 0.1968,
+      "step": 626
+    },
+    {
+      "epoch": 0.02508,
+      "grad_norm": 1.363713026046753,
+      "learning_rate": 7.378108213591355e-06,
+      "loss": 0.1133,
+      "step": 627
+    },
+    {
+      "epoch": 0.02512,
+      "grad_norm": 1.164609670639038,
+      "learning_rate": 7.344438825131912e-06,
+      "loss": 0.0837,
+      "step": 628
+    },
+    {
+      "epoch": 0.02516,
+      "grad_norm": 0.09954403340816498,
+      "learning_rate": 7.310801793847344e-06,
+      "loss": 0.0011,
+      "step": 629
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 5.015952110290527,
+      "learning_rate": 7.277197529594257e-06,
+      "loss": 0.1249,
+      "step": 630
+    },
+    {
+      "epoch": 0.02524,
+      "grad_norm": 7.982026100158691,
+      "learning_rate": 7.243626441830009e-06,
+      "loss": 0.2077,
+      "step": 631
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 0.9444122910499573,
+      "learning_rate": 7.210088939607709e-06,
+      "loss": 0.0786,
+      "step": 632
+    },
+    {
+      "epoch": 0.02532,
+      "grad_norm": 0.5538161396980286,
+      "learning_rate": 7.176585431571235e-06,
+      "loss": 0.0346,
+      "step": 633
+    },
+    {
+      "epoch": 0.02536,
+      "grad_norm": 1.0498833656311035,
+      "learning_rate": 7.143116325950266e-06,
+      "loss": 0.1088,
+      "step": 634
+    },
+    {
+      "epoch": 0.0254,
+      "grad_norm": 1.4593381881713867,
+      "learning_rate": 7.109682030555283e-06,
+      "loss": 0.0833,
+      "step": 635
+    },
+    {
+      "epoch": 0.02544,
+      "grad_norm": 3.288681745529175,
+      "learning_rate": 7.076282952772634e-06,
+      "loss": 0.1028,
+      "step": 636
+    },
+    {
+      "epoch": 0.02548,
+      "grad_norm": 2.5986289978027344,
+      "learning_rate": 7.042919499559538e-06,
+      "loss": 0.0785,
+      "step": 637
+    },
+    {
+      "epoch": 0.02552,
+      "grad_norm": 0.8563199639320374,
+      "learning_rate": 7.009592077439135e-06,
+      "loss": 0.0724,
+      "step": 638
+    },
+    {
+      "epoch": 0.02556,
+      "grad_norm": 7.134288787841797,
+      "learning_rate": 6.976301092495556e-06,
+      "loss": 0.1831,
+      "step": 639
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.8803825974464417,
+      "learning_rate": 6.943046950368944e-06,
+      "loss": 0.0693,
+      "step": 640
+    },
+    {
+      "epoch": 0.02564,
+      "grad_norm": 15.41519546508789,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.2031,
+      "step": 641
+    },
+    {
+      "epoch": 0.02568,
+      "grad_norm": 3.8874897956848145,
+      "learning_rate": 6.876650814877675e-06,
+      "loss": 0.1145,
+      "step": 642
+    },
+    {
+      "epoch": 0.02572,
+      "grad_norm": 4.336860179901123,
+      "learning_rate": 6.843509630528977e-06,
+      "loss": 0.1496,
+      "step": 643
+    },
+    {
+      "epoch": 0.02576,
+      "grad_norm": 4.7061686515808105,
+      "learning_rate": 6.8104069070193e-06,
+      "loss": 0.1333,
+      "step": 644
+    },
+    {
+      "epoch": 0.0258,
+      "grad_norm": 4.266148567199707,
+      "learning_rate": 6.777343047694891e-06,
+      "loss": 0.1025,
+      "step": 645
+    },
+    {
+      "epoch": 0.02584,
+      "grad_norm": 0.8166195750236511,
+      "learning_rate": 6.744318455428436e-06,
+      "loss": 0.0794,
+      "step": 646
+    },
+    {
+      "epoch": 0.02588,
+      "grad_norm": 1.1786655187606812,
+      "learning_rate": 6.711333532614168e-06,
+      "loss": 0.0907,
+      "step": 647
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 10.657480239868164,
+      "learning_rate": 6.67838868116297e-06,
+      "loss": 0.1798,
+      "step": 648
+    },
+    {
+      "epoch": 0.02596,
+      "grad_norm": 7.715270042419434,
+      "learning_rate": 6.645484302497452e-06,
+      "loss": 0.7763,
+      "step": 649
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 29.16922950744629,
+      "learning_rate": 6.612620797547087e-06,
+      "loss": 0.3572,
+      "step": 650
+    },
+    {
+      "epoch": 0.02604,
+      "grad_norm": 12.742008209228516,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.3084,
+      "step": 651
+    },
+    {
+      "epoch": 0.02608,
+      "grad_norm": 1.2898739576339722,
+      "learning_rate": 6.547018010014654e-06,
+      "loss": 0.0494,
+      "step": 652
+    },
+    {
+      "epoch": 0.02612,
+      "grad_norm": 1.4708659648895264,
+      "learning_rate": 6.5142795267818505e-06,
+      "loss": 0.1083,
+      "step": 653
+    },
+    {
+      "epoch": 0.02616,
+      "grad_norm": 0.5371432900428772,
+      "learning_rate": 6.481583515952983e-06,
+      "loss": 0.0535,
+      "step": 654
+    },
+    {
+      "epoch": 0.0262,
+      "grad_norm": 0.43113499879837036,
+      "learning_rate": 6.448930375918632e-06,
+      "loss": 0.0426,
+      "step": 655
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 1.2268309593200684,
+      "learning_rate": 6.4163205045469975e-06,
+      "loss": 0.0676,
+      "step": 656
+    },
+    {
+      "epoch": 0.02628,
+      "grad_norm": 0.6629493236541748,
+      "learning_rate": 6.383754299179079e-06,
+      "loss": 0.061,
+      "step": 657
+    },
+    {
+      "epoch": 0.02632,
+      "grad_norm": 0.23154549300670624,
+      "learning_rate": 6.351232156623803e-06,
+      "loss": 0.0034,
+      "step": 658
+    },
+    {
+      "epoch": 0.02636,
+      "grad_norm": 6.892587661743164,
+      "learning_rate": 6.318754473153221e-06,
+      "loss": 0.1939,
+      "step": 659
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 6.524229049682617,
+      "learning_rate": 6.286321644497655e-06,
+      "loss": 0.198,
+      "step": 660
+    },
+    {
+      "epoch": 0.02644,
+      "grad_norm": 0.6682811975479126,
+      "learning_rate": 6.25393406584088e-06,
+      "loss": 0.085,
+      "step": 661
+    },
+    {
+      "epoch": 0.02648,
+      "grad_norm": 5.873569488525391,
+      "learning_rate": 6.22159213181533e-06,
+      "loss": 0.1278,
+      "step": 662
+    },
+    {
+      "epoch": 0.02652,
+      "grad_norm": 0.07708533108234406,
+      "learning_rate": 6.18929623649726e-06,
+      "loss": 0.002,
+      "step": 663
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 2.1431870460510254,
+      "learning_rate": 6.157046773401964e-06,
+      "loss": 0.0909,
+      "step": 664
+    },
+    {
+      "epoch": 0.0266,
+      "grad_norm": 1.0127865076065063,
+      "learning_rate": 6.124844135478971e-06,
+      "loss": 0.0403,
+      "step": 665
+    },
+    {
+      "epoch": 0.02664,
+      "grad_norm": 1.1291205883026123,
+      "learning_rate": 6.092688715107265e-06,
+      "loss": 0.0653,
+      "step": 666
+    },
+    {
+      "epoch": 0.02668,
+      "grad_norm": 1.1704145669937134,
+      "learning_rate": 6.06058090409049e-06,
+      "loss": 0.0777,
+      "step": 667
+    },
+    {
+      "epoch": 0.02672,
+      "grad_norm": 1.5688973665237427,
+      "learning_rate": 6.028521093652195e-06,
+      "loss": 0.0939,
+      "step": 668
+    },
+    {
+      "epoch": 0.02676,
+      "grad_norm": 6.895656108856201,
+      "learning_rate": 5.996509674431053e-06,
+      "loss": 0.1557,
+      "step": 669
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 9.109591484069824,
+      "learning_rate": 5.9645470364761e-06,
+      "loss": 0.2866,
+      "step": 670
+    },
+    {
+      "epoch": 0.02684,
+      "grad_norm": 1.38040292263031,
+      "learning_rate": 5.932633569242e-06,
+      "loss": 0.0791,
+      "step": 671
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 1.4117307662963867,
+      "learning_rate": 5.900769661584273e-06,
+      "loss": 0.0594,
+      "step": 672
+    },
+    {
+      "epoch": 0.02692,
+      "grad_norm": 12.57364559173584,
+      "learning_rate": 5.868955701754584e-06,
+      "loss": 0.2203,
+      "step": 673
+    },
+    {
+      "epoch": 0.02696,
+      "grad_norm": 1.3525148630142212,
+      "learning_rate": 5.83719207739599e-06,
+      "loss": 0.077,
+      "step": 674
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 1.0946614742279053,
+      "learning_rate": 5.8054791755382286e-06,
+      "loss": 0.0761,
+      "step": 675
+    },
+    {
+      "epoch": 0.02704,
+      "grad_norm": 1.022179365158081,
+      "learning_rate": 5.773817382593008e-06,
+      "loss": 0.0755,
+      "step": 676
+    },
+    {
+      "epoch": 0.02708,
+      "grad_norm": 1.0013043880462646,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.0422,
+      "step": 677
+    },
+    {
+      "epoch": 0.02712,
+      "grad_norm": 31.10577392578125,
+      "learning_rate": 5.710648665968543e-06,
+      "loss": 0.3709,
+      "step": 678
+    },
+    {
+      "epoch": 0.02716,
+      "grad_norm": 3.08500599861145,
+      "learning_rate": 5.679142511980176e-06,
+      "loss": 0.1306,
+      "step": 679
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 1.874846339225769,
+      "learning_rate": 5.647689006276727e-06,
+      "loss": 0.0872,
+      "step": 680
+    },
+    {
+      "epoch": 0.02724,
+      "grad_norm": 1.731837272644043,
+      "learning_rate": 5.616288532109225e-06,
+      "loss": 0.0737,
+      "step": 681
+    },
+    {
+      "epoch": 0.02728,
+      "grad_norm": 0.6736709475517273,
+      "learning_rate": 5.584941472082549e-06,
+      "loss": 0.0753,
+      "step": 682
+    },
+    {
+      "epoch": 0.02732,
+      "grad_norm": 11.032455444335938,
+      "learning_rate": 5.553648208150728e-06,
+      "loss": 0.2352,
+      "step": 683
+    },
+    {
+      "epoch": 0.02736,
+      "grad_norm": 5.9419660568237305,
+      "learning_rate": 5.522409121612304e-06,
+      "loss": 0.1268,
+      "step": 684
+    },
+    {
+      "epoch": 0.0274,
+      "grad_norm": 11.134778022766113,
+      "learning_rate": 5.491224593105695e-06,
+      "loss": 0.2759,
+      "step": 685
+    },
+    {
+      "epoch": 0.02744,
+      "grad_norm": 1.1338247060775757,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.0691,
+      "step": 686
+    },
+    {
+      "epoch": 0.02748,
+      "grad_norm": 17.681621551513672,
+      "learning_rate": 5.429020729413062e-06,
+      "loss": 0.4142,
+      "step": 687
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 1.2438162565231323,
+      "learning_rate": 5.398002152161484e-06,
+      "loss": 0.0653,
+      "step": 688
+    },
+    {
+      "epoch": 0.02756,
+      "grad_norm": 11.666088104248047,
+      "learning_rate": 5.367039648801386e-06,
+      "loss": 0.217,
+      "step": 689
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 1.2870872020721436,
+      "learning_rate": 5.336133596601089e-06,
+      "loss": 0.0763,
+      "step": 690
+    },
+    {
+      "epoch": 0.02764,
+      "grad_norm": 5.913383483886719,
+      "learning_rate": 5.305284372141095e-06,
+      "loss": 0.1354,
+      "step": 691
+    },
+    {
+      "epoch": 0.02768,
+      "grad_norm": 1.071195363998413,
+      "learning_rate": 5.274492351309462e-06,
+      "loss": 0.0923,
+      "step": 692
+    },
+    {
+      "epoch": 0.02772,
+      "grad_norm": 9.876152038574219,
+      "learning_rate": 5.243757909297247e-06,
+      "loss": 0.0964,
+      "step": 693
+    },
+    {
+      "epoch": 0.02776,
+      "grad_norm": 1.5717387199401855,
+      "learning_rate": 5.213081420593933e-06,
+      "loss": 0.0962,
+      "step": 694
+    },
+    {
+      "epoch": 0.0278,
+      "grad_norm": 0.6891898512840271,
+      "learning_rate": 5.1824632589828465e-06,
+      "loss": 0.0741,
+      "step": 695
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 1.2878379821777344,
+      "learning_rate": 5.151903797536631e-06,
+      "loss": 0.0933,
+      "step": 696
+    },
+    {
+      "epoch": 0.02788,
+      "grad_norm": 1.9482656717300415,
+      "learning_rate": 5.121403408612672e-06,
+      "loss": 0.1362,
+      "step": 697
+    },
+    {
+      "epoch": 0.02792,
+      "grad_norm": 0.20593877136707306,
+      "learning_rate": 5.090962463848592e-06,
+      "loss": 0.0054,
+      "step": 698
+    },
+    {
+      "epoch": 0.02796,
+      "grad_norm": 0.09866832941770554,
+      "learning_rate": 5.060581334157693e-06,
+      "loss": 0.001,
+      "step": 699
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 1.5989842414855957,
+      "learning_rate": 5.030260389724447e-06,
+      "loss": 0.079,
+      "step": 700
+    },
+    {
+      "epoch": 0.02804,
+      "grad_norm": 13.688101768493652,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.3425,
+      "step": 701
+    },
+    {
+      "epoch": 0.02808,
+      "grad_norm": 13.463900566101074,
+      "learning_rate": 4.96980053369765e-06,
+      "loss": 0.5455,
+      "step": 702
+    },
+    {
+      "epoch": 0.02812,
+      "grad_norm": 6.7879767417907715,
+      "learning_rate": 4.939662358788364e-06,
+      "loss": 0.1146,
+      "step": 703
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 2.1676604747772217,
+      "learning_rate": 4.909585842496287e-06,
+      "loss": 0.0229,
+      "step": 704
+    },
+    {
+      "epoch": 0.0282,
+      "grad_norm": 28.17415428161621,
+      "learning_rate": 4.879571351294287e-06,
+      "loss": 0.5631,
+      "step": 705
+    },
+    {
+      "epoch": 0.02824,
+      "grad_norm": 0.885439932346344,
+      "learning_rate": 4.849619250899458e-06,
+      "loss": 0.035,
+      "step": 706
+    },
+    {
+      "epoch": 0.02828,
+      "grad_norm": 0.5848221182823181,
+      "learning_rate": 4.8197299062687e-06,
+      "loss": 0.0075,
+      "step": 707
+    },
+    {
+      "epoch": 0.02832,
+      "grad_norm": 0.7367845773696899,
+      "learning_rate": 4.78990368159424e-06,
+      "loss": 0.0568,
+      "step": 708
+    },
+    {
+      "epoch": 0.02836,
+      "grad_norm": 1.1661707162857056,
+      "learning_rate": 4.76014094029921e-06,
+      "loss": 0.0847,
+      "step": 709
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 1.238843321800232,
+      "learning_rate": 4.7304420450332244e-06,
+      "loss": 0.0981,
+      "step": 710
+    },
+    {
+      "epoch": 0.02844,
+      "grad_norm": 0.17394839227199554,
+      "learning_rate": 4.700807357667953e-06,
+      "loss": 0.0014,
+      "step": 711
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 1.34825599193573,
+      "learning_rate": 4.671237239292699e-06,
+      "loss": 0.0855,
+      "step": 712
+    },
+    {
+      "epoch": 0.02852,
+      "grad_norm": 0.776859700679779,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.0671,
+      "step": 713
+    },
+    {
+      "epoch": 0.02856,
+      "grad_norm": 0.7578673958778381,
+      "learning_rate": 4.612292149931369e-06,
+      "loss": 0.057,
+      "step": 714
+    },
+    {
+      "epoch": 0.0286,
+      "grad_norm": 3.891252040863037,
+      "learning_rate": 4.582917897172603e-06,
+      "loss": 0.1218,
+      "step": 715
+    },
+    {
+      "epoch": 0.02864,
+      "grad_norm": 0.37234005331993103,
+      "learning_rate": 4.5536096498497295e-06,
+      "loss": 0.0729,
+      "step": 716
+    },
+    {
+      "epoch": 0.02868,
+      "grad_norm": 6.480273246765137,
+      "learning_rate": 4.524367765074499e-06,
+      "loss": 0.1493,
+      "step": 717
+    },
+    {
+      "epoch": 0.02872,
+      "grad_norm": 4.18231201171875,
+      "learning_rate": 4.495192599150045e-06,
+      "loss": 0.1093,
+      "step": 718
+    },
+    {
+      "epoch": 0.02876,
+      "grad_norm": 6.959336757659912,
+      "learning_rate": 4.46608450756656e-06,
+      "loss": 0.1812,
+      "step": 719
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 13.548670768737793,
+      "learning_rate": 4.437043844996952e-06,
+      "loss": 0.5146,
+      "step": 720
+    },
+    {
+      "epoch": 0.02884,
+      "grad_norm": 0.648603618144989,
+      "learning_rate": 4.408070965292534e-06,
+      "loss": 0.0475,
+      "step": 721
+    },
+    {
+      "epoch": 0.02888,
+      "grad_norm": 0.5452705025672913,
+      "learning_rate": 4.379166221478697e-06,
+      "loss": 0.0463,
+      "step": 722
+    },
+    {
+      "epoch": 0.02892,
+      "grad_norm": 1.2327595949172974,
+      "learning_rate": 4.350329965750622e-06,
+      "loss": 0.0877,
+      "step": 723
+    },
+    {
+      "epoch": 0.02896,
+      "grad_norm": 2.062779426574707,
+      "learning_rate": 4.321562549468991e-06,
+      "loss": 0.0999,
+      "step": 724
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 1.144892692565918,
+      "learning_rate": 4.292864323155684e-06,
+      "loss": 0.0787,
+      "step": 725
+    },
+    {
+      "epoch": 0.02904,
+      "grad_norm": 11.709745407104492,
+      "learning_rate": 4.264235636489542e-06,
+      "loss": 0.2323,
+      "step": 726
+    },
+    {
+      "epoch": 0.02908,
+      "grad_norm": 1.1318130493164062,
+      "learning_rate": 4.235676838302069e-06,
+      "loss": 0.108,
+      "step": 727
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 0.4052428901195526,
+      "learning_rate": 4.207188276573214e-06,
+      "loss": 0.0661,
+      "step": 728
+    },
+    {
+      "epoch": 0.02916,
+      "grad_norm": 1.4315210580825806,
+      "learning_rate": 4.178770298427107e-06,
+      "loss": 0.0944,
+      "step": 729
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 0.3364134132862091,
+      "learning_rate": 4.150423250127846e-06,
+      "loss": 0.0056,
+      "step": 730
+    },
+    {
+      "epoch": 0.02924,
+      "grad_norm": 20.966676712036133,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.499,
+      "step": 731
+    },
+    {
+      "epoch": 0.02928,
+      "grad_norm": 1.064622402191162,
+      "learning_rate": 4.093943323800746e-06,
+      "loss": 0.1109,
+      "step": 732
+    },
+    {
+      "epoch": 0.02932,
+      "grad_norm": 6.967021465301514,
+      "learning_rate": 4.065811133962987e-06,
+      "loss": 0.0934,
+      "step": 733
+    },
+    {
+      "epoch": 0.02936,
+      "grad_norm": 5.16302490234375,
+      "learning_rate": 4.037751250343841e-06,
+      "loss": 0.1504,
+      "step": 734
+    },
+    {
+      "epoch": 0.0294,
+      "grad_norm": 0.8019863367080688,
+      "learning_rate": 4.009764014844143e-06,
+      "loss": 0.0772,
+      "step": 735
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 1.4126380681991577,
+      "learning_rate": 3.981849768479516e-06,
+      "loss": 0.0979,
+      "step": 736
+    },
+    {
+      "epoch": 0.02948,
+      "grad_norm": 0.9765819907188416,
+      "learning_rate": 3.954008851376252e-06,
+      "loss": 0.1274,
+      "step": 737
+    },
+    {
+      "epoch": 0.02952,
+      "grad_norm": 1.3779306411743164,
+      "learning_rate": 3.9262416027671354e-06,
+      "loss": 0.0122,
+      "step": 738
+    },
+    {
+      "epoch": 0.02956,
+      "grad_norm": 0.624764084815979,
+      "learning_rate": 3.898548360987325e-06,
+      "loss": 0.0657,
+      "step": 739
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.7635204792022705,
+      "learning_rate": 3.8709294634702374e-06,
+      "loss": 0.0481,
+      "step": 740
+    },
+    {
+      "epoch": 0.02964,
+      "grad_norm": 1.3860241174697876,
+      "learning_rate": 3.8433852467434175e-06,
+      "loss": 0.1005,
+      "step": 741
+    },
+    {
+      "epoch": 0.02968,
+      "grad_norm": 1.1818790435791016,
+      "learning_rate": 3.81591604642446e-06,
+      "loss": 0.0906,
+      "step": 742
+    },
+    {
+      "epoch": 0.02972,
+      "grad_norm": 0.9501907825469971,
+      "learning_rate": 3.7885221972168974e-06,
+      "loss": 0.0656,
+      "step": 743
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 1.6521177291870117,
+      "learning_rate": 3.7612040329061405e-06,
+      "loss": 0.0772,
+      "step": 744
+    },
+    {
+      "epoch": 0.0298,
+      "grad_norm": 0.3046473264694214,
+      "learning_rate": 3.7339618863553983e-06,
+      "loss": 0.0315,
+      "step": 745
+    },
+    {
+      "epoch": 0.02984,
+      "grad_norm": 6.267678260803223,
+      "learning_rate": 3.7067960895016277e-06,
+      "loss": 0.1364,
+      "step": 746
+    },
+    {
+      "epoch": 0.02988,
+      "grad_norm": 2.650761604309082,
+      "learning_rate": 3.679706973351491e-06,
+      "loss": 0.0768,
+      "step": 747
+    },
+    {
+      "epoch": 0.02992,
+      "grad_norm": 3.264446973800659,
+      "learning_rate": 3.6526948679773256e-06,
+      "loss": 0.1276,
+      "step": 748
+    },
+    {
+      "epoch": 0.02996,
+      "grad_norm": 2.756331443786621,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.1008,
+      "step": 749
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.337132453918457,
+      "learning_rate": 3.598903005150444e-06,
+      "loss": 0.3695,
+      "step": 750
+    },
+    {
+      "epoch": 0.03004,
+      "grad_norm": 10.05043888092041,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 0.2941,
+      "step": 751
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 3.301450729370117,
+      "learning_rate": 3.545423122760493e-06,
+      "loss": 0.104,
+      "step": 752
+    },
+    {
+      "epoch": 0.03012,
+      "grad_norm": 0.5295363664627075,
+      "learning_rate": 3.5188009893686916e-06,
+      "loss": 0.0539,
+      "step": 753
+    },
+    {
+      "epoch": 0.03016,
+      "grad_norm": 2.950563430786133,
+      "learning_rate": 3.492257827341492e-06,
+      "loss": 0.1269,
+      "step": 754
+    },
+    {
+      "epoch": 0.0302,
+      "grad_norm": 13.975685119628906,
+      "learning_rate": 3.4657939600989453e-06,
+      "loss": 0.3576,
+      "step": 755
+    },
+    {
+      "epoch": 0.03024,
+      "grad_norm": 1.9617775678634644,
+      "learning_rate": 3.4394097100949286e-06,
+      "loss": 0.0924,
+      "step": 756
+    },
+    {
+      "epoch": 0.03028,
+      "grad_norm": 0.7575762867927551,
+      "learning_rate": 3.4131053988131947e-06,
+      "loss": 0.0567,
+      "step": 757
+    },
+    {
+      "epoch": 0.03032,
+      "grad_norm": 1.4084222316741943,
+      "learning_rate": 3.3868813467634833e-06,
+      "loss": 0.0828,
+      "step": 758
+    },
+    {
+      "epoch": 0.03036,
+      "grad_norm": 8.285188674926758,
+      "learning_rate": 3.360737873477584e-06,
+      "loss": 0.1885,
+      "step": 759
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 0.47943201661109924,
+      "learning_rate": 3.3346752975054763e-06,
+      "loss": 0.0555,
+      "step": 760
+    },
+    {
+      "epoch": 0.03044,
+      "grad_norm": 2.4846839904785156,
+      "learning_rate": 3.308693936411421e-06,
+      "loss": 0.1066,
+      "step": 761
+    },
+    {
+      "epoch": 0.03048,
+      "grad_norm": 0.5868697762489319,
+      "learning_rate": 3.2827941067700996e-06,
+      "loss": 0.0503,
+      "step": 762
+    },
+    {
+      "epoch": 0.03052,
+      "grad_norm": 5.280800819396973,
+      "learning_rate": 3.2569761241627694e-06,
+      "loss": 0.1604,
+      "step": 763
+    },
+    {
+      "epoch": 0.03056,
+      "grad_norm": 9.40414047241211,
+      "learning_rate": 3.2312403031733943e-06,
+      "loss": 0.3161,
+      "step": 764
+    },
+    {
+      "epoch": 0.0306,
+      "grad_norm": 1.6196728944778442,
+      "learning_rate": 3.2055869573848374e-06,
+      "loss": 0.0384,
+      "step": 765
+    },
+    {
+      "epoch": 0.03064,
+      "grad_norm": 0.7027143239974976,
+      "learning_rate": 3.1800163993750166e-06,
+      "loss": 0.0733,
+      "step": 766
+    },
+    {
+      "epoch": 0.03068,
+      "grad_norm": 1.2309304475784302,
+      "learning_rate": 3.1545289407131128e-06,
+      "loss": 0.0465,
+      "step": 767
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 1.710585594177246,
+      "learning_rate": 3.1291248919557717e-06,
+      "loss": 0.1003,
+      "step": 768
+    },
+    {
+      "epoch": 0.03076,
+      "grad_norm": 9.546642303466797,
+      "learning_rate": 3.103804562643302e-06,
+      "loss": 0.1976,
+      "step": 769
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 1.5926058292388916,
+      "learning_rate": 3.0785682612959334e-06,
+      "loss": 0.1058,
+      "step": 770
+    },
+    {
+      "epoch": 0.03084,
+      "grad_norm": 1.0405668020248413,
+      "learning_rate": 3.0534162954100264e-06,
+      "loss": 0.0931,
+      "step": 771
+    },
+    {
+      "epoch": 0.03088,
+      "grad_norm": 1.2072910070419312,
+      "learning_rate": 3.028348971454356e-06,
+      "loss": 0.0762,
+      "step": 772
+    },
+    {
+      "epoch": 0.03092,
+      "grad_norm": 1.7974003553390503,
+      "learning_rate": 3.003366594866345e-06,
+      "loss": 0.0779,
+      "step": 773
+    },
+    {
+      "epoch": 0.03096,
+      "grad_norm": 2.5940818786621094,
+      "learning_rate": 2.978469470048376e-06,
+      "loss": 0.1066,
+      "step": 774
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 1.2954779863357544,
+      "learning_rate": 2.953657900364053e-06,
+      "loss": 0.1004,
+      "step": 775
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 0.42695632576942444,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.0045,
+      "step": 776
+    },
+    {
+      "epoch": 0.03108,
+      "grad_norm": 1.0006481409072876,
+      "learning_rate": 2.9042926346347932e-06,
+      "loss": 0.0719,
+      "step": 777
+    },
+    {
+      "epoch": 0.03112,
+      "grad_norm": 0.8708807229995728,
+      "learning_rate": 2.8797395400900362e-06,
+      "loss": 0.0676,
+      "step": 778
+    },
+    {
+      "epoch": 0.03116,
+      "grad_norm": 0.4236277937889099,
+      "learning_rate": 2.855273203671969e-06,
+      "loss": 0.0611,
+      "step": 779
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.6245424747467041,
+      "learning_rate": 2.830893923495173e-06,
+      "loss": 0.0342,
+      "step": 780
+    },
+    {
+      "epoch": 0.03124,
+      "grad_norm": 2.0277907848358154,
+      "learning_rate": 2.8066019966134907e-06,
+      "loss": 0.0871,
+      "step": 781
+    },
+    {
+      "epoch": 0.03128,
+      "grad_norm": 2.040543794631958,
+      "learning_rate": 2.7823977190163788e-06,
+      "loss": 0.111,
+      "step": 782
+    },
+    {
+      "epoch": 0.03132,
+      "grad_norm": 3.2347772121429443,
+      "learning_rate": 2.7582813856253276e-06,
+      "loss": 0.1475,
+      "step": 783
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 7.035062313079834,
+      "learning_rate": 2.7342532902902418e-06,
+      "loss": 0.1662,
+      "step": 784
+    },
+    {
+      "epoch": 0.0314,
+      "grad_norm": 0.8405598402023315,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.0935,
+      "step": 785
+    },
+    {
+      "epoch": 0.03144,
+      "grad_norm": 7.133166790008545,
+      "learning_rate": 2.6864629838082957e-06,
+      "loss": 0.1322,
+      "step": 786
+    },
+    {
+      "epoch": 0.03148,
+      "grad_norm": 1.5037939548492432,
+      "learning_rate": 2.6627013549712355e-06,
+      "loss": 0.0834,
+      "step": 787
+    },
+    {
+      "epoch": 0.03152,
+      "grad_norm": 2.583517074584961,
+      "learning_rate": 2.639029128802657e-06,
+      "loss": 0.124,
+      "step": 788
+    },
+    {
+      "epoch": 0.03156,
+      "grad_norm": 10.736677169799805,
+      "learning_rate": 2.615446593741161e-06,
+      "loss": 0.2191,
+      "step": 789
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 8.237153053283691,
+      "learning_rate": 2.5919540371325005e-06,
+      "loss": 0.1808,
+      "step": 790
+    },
+    {
+      "epoch": 0.03164,
+      "grad_norm": 10.916925430297852,
+      "learning_rate": 2.5685517452260566e-06,
+      "loss": 0.3326,
+      "step": 791
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 4.187741279602051,
+      "learning_rate": 2.5452400031713786e-06,
+      "loss": 0.1341,
+      "step": 792
+    },
+    {
+      "epoch": 0.03172,
+      "grad_norm": 2.8332037925720215,
+      "learning_rate": 2.522019095014683e-06,
+      "loss": 0.0622,
+      "step": 793
+    },
+    {
+      "epoch": 0.03176,
+      "grad_norm": 1.0867068767547607,
+      "learning_rate": 2.4988893036954045e-06,
+      "loss": 0.0562,
+      "step": 794
+    },
+    {
+      "epoch": 0.0318,
+      "grad_norm": 0.9876936674118042,
+      "learning_rate": 2.4758509110427576e-06,
+      "loss": 0.0732,
+      "step": 795
+    },
+    {
+      "epoch": 0.03184,
+      "grad_norm": 7.9589667320251465,
+      "learning_rate": 2.45290419777228e-06,
+      "loss": 0.692,
+      "step": 796
+    },
+    {
+      "epoch": 0.03188,
+      "grad_norm": 1.9825248718261719,
+      "learning_rate": 2.4300494434824373e-06,
+      "loss": 0.0938,
+      "step": 797
+    },
+    {
+      "epoch": 0.03192,
+      "grad_norm": 11.507606506347656,
+      "learning_rate": 2.407286926651192e-06,
+      "loss": 0.1712,
+      "step": 798
+    },
+    {
+      "epoch": 0.03196,
+      "grad_norm": 1.5084614753723145,
+      "learning_rate": 2.3846169246326345e-06,
+      "loss": 0.1238,
+      "step": 799
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 2.1997289657592773,
+      "learning_rate": 2.362039713653581e-06,
+      "loss": 0.0638,
+      "step": 800
+    },
+    {
+      "epoch": 0.03204,
+      "grad_norm": 0.70892333984375,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.0677,
+      "step": 801
+    },
+    {
+      "epoch": 0.03208,
+      "grad_norm": 1.5294110774993896,
+      "learning_rate": 2.317164764064769e-06,
+      "loss": 0.081,
+      "step": 802
+    },
+    {
+      "epoch": 0.03212,
+      "grad_norm": 1.5677021741867065,
+      "learning_rate": 2.2948675722421086e-06,
+      "loss": 0.1059,
+      "step": 803
+    },
+    {
+      "epoch": 0.03216,
+      "grad_norm": 0.844959020614624,
+      "learning_rate": 2.27266426502649e-06,
+      "loss": 0.0355,
+      "step": 804
+    },
+    {
+      "epoch": 0.0322,
+      "grad_norm": 16.42185401916504,
+      "learning_rate": 2.2505551129582047e-06,
+      "loss": 0.2225,
+      "step": 805
+    },
+    {
+      "epoch": 0.03224,
+      "grad_norm": 1.5545107126235962,
+      "learning_rate": 2.2285403854302912e-06,
+      "loss": 0.0648,
+      "step": 806
+    },
+    {
+      "epoch": 0.03228,
+      "grad_norm": 1.0437160730361938,
+      "learning_rate": 2.206620350685257e-06,
+      "loss": 0.0829,
+      "step": 807
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 1.4130158424377441,
+      "learning_rate": 2.1847952758118118e-06,
+      "loss": 0.0913,
+      "step": 808
+    },
+    {
+      "epoch": 0.03236,
+      "grad_norm": 7.748847961425781,
+      "learning_rate": 2.163065426741603e-06,
+      "loss": 0.1787,
+      "step": 809
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 3.5241408348083496,
+      "learning_rate": 2.1414310682459805e-06,
+      "loss": 0.1238,
+      "step": 810
+    },
+    {
+      "epoch": 0.03244,
+      "grad_norm": 3.9404430389404297,
+      "learning_rate": 2.119892463932781e-06,
+      "loss": 0.0459,
+      "step": 811
+    },
+    {
+      "epoch": 0.03248,
+      "grad_norm": 1.528969168663025,
+      "learning_rate": 2.098449876243096e-06,
+      "loss": 0.0988,
+      "step": 812
+    },
+    {
+      "epoch": 0.03252,
+      "grad_norm": 0.5191096663475037,
+      "learning_rate": 2.0771035664480944e-06,
+      "loss": 0.032,
+      "step": 813
+    },
+    {
+      "epoch": 0.03256,
+      "grad_norm": 0.8545474410057068,
+      "learning_rate": 2.0558537946458177e-06,
+      "loss": 0.0793,
+      "step": 814
+    },
+    {
+      "epoch": 0.0326,
+      "grad_norm": 0.8879887461662292,
+      "learning_rate": 2.0347008197580376e-06,
+      "loss": 0.0525,
+      "step": 815
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 6.960842132568359,
+      "learning_rate": 2.013644899527074e-06,
+      "loss": 0.147,
+      "step": 816
+    },
+    {
+      "epoch": 0.03268,
+      "grad_norm": 6.885871410369873,
+      "learning_rate": 1.9926862905126663e-06,
+      "loss": 0.1587,
+      "step": 817
+    },
+    {
+      "epoch": 0.03272,
+      "grad_norm": 0.9178568124771118,
+      "learning_rate": 1.9718252480888567e-06,
+      "loss": 0.0838,
+      "step": 818
+    },
+    {
+      "epoch": 0.03276,
+      "grad_norm": 1.1158992052078247,
+      "learning_rate": 1.95106202644086e-06,
+      "loss": 0.0985,
+      "step": 819
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 5.152473449707031,
+      "learning_rate": 1.930396878561983e-06,
+      "loss": 0.129,
+      "step": 820
+    },
+    {
+      "epoch": 0.03284,
+      "grad_norm": 8.667509078979492,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.1737,
+      "step": 821
+    },
+    {
+      "epoch": 0.03288,
+      "grad_norm": 13.668463706970215,
+      "learning_rate": 1.8893618101067357e-06,
+      "loss": 0.2594,
+      "step": 822
+    },
+    {
+      "epoch": 0.03292,
+      "grad_norm": 24.340282440185547,
+      "learning_rate": 1.8689923895297247e-06,
+      "loss": 0.808,
+      "step": 823
+    },
+    {
+      "epoch": 0.03296,
+      "grad_norm": 8.698612213134766,
+      "learning_rate": 1.848722042714457e-06,
+      "loss": 0.1098,
+      "step": 824
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 1.3748273849487305,
+      "learning_rate": 1.8285510166487154e-06,
+      "loss": 0.0712,
+      "step": 825
+    },
+    {
+      "epoch": 0.03304,
+      "grad_norm": 11.213031768798828,
+      "learning_rate": 1.808479557110081e-06,
+      "loss": 0.5415,
+      "step": 826
+    },
+    {
+      "epoch": 0.03308,
+      "grad_norm": 0.40597566962242126,
+      "learning_rate": 1.7885079086629598e-06,
+      "loss": 0.0523,
+      "step": 827
+    },
+    {
+      "epoch": 0.03312,
+      "grad_norm": 13.828956604003906,
+      "learning_rate": 1.7686363146555807e-06,
+      "loss": 0.2404,
+      "step": 828
+    },
+    {
+      "epoch": 0.03316,
+      "grad_norm": 1.4844040870666504,
+      "learning_rate": 1.7488650172170496e-06,
+      "loss": 0.0503,
+      "step": 829
+    },
+    {
+      "epoch": 0.0332,
+      "grad_norm": 0.9149099588394165,
+      "learning_rate": 1.7291942572543806e-06,
+      "loss": 0.0863,
+      "step": 830
+    },
+    {
+      "epoch": 0.03324,
+      "grad_norm": 0.08156435191631317,
+      "learning_rate": 1.709624274449584e-06,
+      "loss": 0.0019,
+      "step": 831
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 1.3342599868774414,
+      "learning_rate": 1.6901553072567189e-06,
+      "loss": 0.0687,
+      "step": 832
+    },
+    {
+      "epoch": 0.03332,
+      "grad_norm": 1.5423862934112549,
+      "learning_rate": 1.6707875928990059e-06,
+      "loss": 0.0981,
+      "step": 833
+    },
+    {
+      "epoch": 0.03336,
+      "grad_norm": 7.227363586425781,
+      "learning_rate": 1.651521367365936e-06,
+      "loss": 0.1822,
+      "step": 834
+    },
+    {
+      "epoch": 0.0334,
+      "grad_norm": 6.690497875213623,
+      "learning_rate": 1.6323568654103838e-06,
+      "loss": 0.1642,
+      "step": 835
+    },
+    {
+      "epoch": 0.03344,
+      "grad_norm": 4.9608917236328125,
+      "learning_rate": 1.6132943205457607e-06,
+      "loss": 0.1618,
+      "step": 836
+    },
+    {
+      "epoch": 0.03348,
+      "grad_norm": 12.707261085510254,
+      "learning_rate": 1.5943339650431578e-06,
+      "loss": 0.2171,
+      "step": 837
+    },
+    {
+      "epoch": 0.03352,
+      "grad_norm": 1.1373966932296753,
+      "learning_rate": 1.5754760299285255e-06,
+      "loss": 0.1038,
+      "step": 838
+    },
+    {
+      "epoch": 0.03356,
+      "grad_norm": 11.289196968078613,
+      "learning_rate": 1.5567207449798517e-06,
+      "loss": 0.1857,
+      "step": 839
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 11.109713554382324,
+      "learning_rate": 1.538068338724361e-06,
+      "loss": 0.1932,
+      "step": 840
+    },
+    {
+      "epoch": 0.03364,
+      "grad_norm": 2.7628889083862305,
+      "learning_rate": 1.5195190384357405e-06,
+      "loss": 0.0588,
+      "step": 841
+    },
+    {
+      "epoch": 0.03368,
+      "grad_norm": 0.5399174094200134,
+      "learning_rate": 1.5010730701313626e-06,
+      "loss": 0.0668,
+      "step": 842
+    },
+    {
+      "epoch": 0.03372,
+      "grad_norm": 0.814849317073822,
+      "learning_rate": 1.4827306585695234e-06,
+      "loss": 0.0732,
+      "step": 843
+    },
+    {
+      "epoch": 0.03376,
+      "grad_norm": 10.353022575378418,
+      "learning_rate": 1.4644920272467245e-06,
+      "loss": 0.1981,
+      "step": 844
+    },
+    {
+      "epoch": 0.0338,
+      "grad_norm": 11.560691833496094,
+      "learning_rate": 1.446357398394934e-06,
+      "loss": 0.165,
+      "step": 845
+    },
+    {
+      "epoch": 0.03384,
+      "grad_norm": 8.962580680847168,
+      "learning_rate": 1.4283269929788779e-06,
+      "loss": 0.3251,
+      "step": 846
+    },
+    {
+      "epoch": 0.03388,
+      "grad_norm": 3.3217689990997314,
+      "learning_rate": 1.4104010306933558e-06,
+      "loss": 0.0914,
+      "step": 847
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 1.688917875289917,
+      "learning_rate": 1.3925797299605649e-06,
+      "loss": 0.0939,
+      "step": 848
+    },
+    {
+      "epoch": 0.03396,
+      "grad_norm": 4.323922157287598,
+      "learning_rate": 1.3748633079274254e-06,
+      "loss": 0.0978,
+      "step": 849
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 0.6816321611404419,
+      "learning_rate": 1.3572519804629537e-06,
+      "loss": 0.0876,
+      "step": 850
+    },
+    {
+      "epoch": 0.03404,
+      "grad_norm": 3.219444513320923,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.0493,
+      "step": 851
+    },
+    {
+      "epoch": 0.03408,
+      "grad_norm": 12.79563045501709,
+      "learning_rate": 1.322345466310717e-06,
+      "loss": 0.2188,
+      "step": 852
+    },
+    {
+      "epoch": 0.03412,
+      "grad_norm": 6.8010430335998535,
+      "learning_rate": 1.30505070494781e-06,
+      "loss": 0.1731,
+      "step": 853
+    },
+    {
+      "epoch": 0.03416,
+      "grad_norm": 2.5388362407684326,
+      "learning_rate": 1.2878618887981064e-06,
+      "loss": 0.0854,
+      "step": 854
+    },
+    {
+      "epoch": 0.0342,
+      "grad_norm": 0.30296051502227783,
+      "learning_rate": 1.2707792273019049e-06,
+      "loss": 0.0056,
+      "step": 855
+    },
+    {
+      "epoch": 0.03424,
+      "grad_norm": 25.67251205444336,
+      "learning_rate": 1.2538029286060428e-06,
+      "loss": 0.5755,
+      "step": 856
+    },
+    {
+      "epoch": 0.03428,
+      "grad_norm": 1.5421030521392822,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.0604,
+      "step": 857
+    },
+    {
+      "epoch": 0.03432,
+      "grad_norm": 0.8187874555587769,
+      "learning_rate": 1.2201702457201948e-06,
+      "loss": 0.0664,
+      "step": 858
+    },
+    {
+      "epoch": 0.03436,
+      "grad_norm": 1.1444604396820068,
+      "learning_rate": 1.2035142713338366e-06,
+      "loss": 0.0741,
+      "step": 859
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 1.2242153882980347,
+      "learning_rate": 1.1869654793500784e-06,
+      "loss": 0.1002,
+      "step": 860
+    },
+    {
+      "epoch": 0.03444,
+      "grad_norm": 18.177553176879883,
+      "learning_rate": 1.1705240714107301e-06,
+      "loss": 0.7042,
+      "step": 861
+    },
+    {
+      "epoch": 0.03448,
+      "grad_norm": 10.509775161743164,
+      "learning_rate": 1.1541902478491607e-06,
+      "loss": 0.1339,
+      "step": 862
+    },
+    {
+      "epoch": 0.03452,
+      "grad_norm": 11.771379470825195,
+      "learning_rate": 1.1379642076878528e-06,
+      "loss": 0.373,
+      "step": 863
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 1.0272094011306763,
+      "learning_rate": 1.1218461486359878e-06,
+      "loss": 0.0938,
+      "step": 864
+    },
+    {
+      "epoch": 0.0346,
+      "grad_norm": 0.9082252383232117,
+      "learning_rate": 1.1058362670870248e-06,
+      "loss": 0.0634,
+      "step": 865
+    },
+    {
+      "epoch": 0.03464,
+      "grad_norm": 0.8541867733001709,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.0828,
+      "step": 866
+    },
+    {
+      "epoch": 0.03468,
+      "grad_norm": 0.860985279083252,
+      "learning_rate": 1.0741418154787443e-06,
+      "loss": 0.0766,
+      "step": 867
+    },
+    {
+      "epoch": 0.03472,
+      "grad_norm": 1.5044641494750977,
+      "learning_rate": 1.058457631606319e-06,
+      "loss": 0.087,
+      "step": 868
+    },
+    {
+      "epoch": 0.03476,
+      "grad_norm": 1.125762939453125,
+      "learning_rate": 1.042882397605871e-06,
+      "loss": 0.1008,
+      "step": 869
+    },
+    {
+      "epoch": 0.0348,
+      "grad_norm": 2.0497515201568604,
+      "learning_rate": 1.0274163032567165e-06,
+      "loss": 0.1171,
+      "step": 870
+    },
+    {
+      "epoch": 0.03484,
+      "grad_norm": 21.08624267578125,
+      "learning_rate": 1.012059537008332e-06,
+      "loss": 1.0167,
+      "step": 871
+    },
+    {
+      "epoch": 0.03488,
+      "grad_norm": 4.431634902954102,
+      "learning_rate": 9.968122859780648e-07,
+      "loss": 0.0944,
+      "step": 872
+    },
+    {
+      "epoch": 0.03492,
+      "grad_norm": 2.0454936027526855,
+      "learning_rate": 9.816747359488632e-07,
+      "loss": 0.1134,
+      "step": 873
+    },
+    {
+      "epoch": 0.03496,
+      "grad_norm": 13.005719184875488,
+      "learning_rate": 9.666470713669918e-07,
+      "loss": 0.3284,
+      "step": 874
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 12.257450103759766,
+      "learning_rate": 9.517294753398066e-07,
+      "loss": 0.2192,
+      "step": 875
+    },
+    {
+      "epoch": 0.03504,
+      "grad_norm": 0.9450737237930298,
+      "learning_rate": 9.369221296335007e-07,
+      "loss": 0.0801,
+      "step": 876
+    },
+    {
+      "epoch": 0.03508,
+      "grad_norm": 0.7794105410575867,
+      "learning_rate": 9.222252146709143e-07,
+      "loss": 0.0385,
+      "step": 877
+    },
+    {
+      "epoch": 0.03512,
+      "grad_norm": 1.245734453201294,
+      "learning_rate": 9.076389095293148e-07,
+      "loss": 0.0696,
+      "step": 878
+    },
+    {
+      "epoch": 0.03516,
+      "grad_norm": 1.142150640487671,
+      "learning_rate": 8.931633919382299e-07,
+      "loss": 0.0789,
+      "step": 879
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 3.991765260696411,
+      "learning_rate": 8.787988382772705e-07,
+      "loss": 0.1099,
+      "step": 880
+    },
+    {
+      "epoch": 0.03524,
+      "grad_norm": 1.9521669149398804,
+      "learning_rate": 8.645454235739903e-07,
+      "loss": 0.1204,
+      "step": 881
+    },
+    {
+      "epoch": 0.03528,
+      "grad_norm": 0.5069582462310791,
+      "learning_rate": 8.504033215017527e-07,
+      "loss": 0.045,
+      "step": 882
+    },
+    {
+      "epoch": 0.03532,
+      "grad_norm": 2.455803632736206,
+      "learning_rate": 8.363727043776037e-07,
+      "loss": 0.0743,
+      "step": 883
+    },
+    {
+      "epoch": 0.03536,
+      "grad_norm": 1.673046588897705,
+      "learning_rate": 8.224537431601886e-07,
+      "loss": 0.0927,
+      "step": 884
+    },
+    {
+      "epoch": 0.0354,
+      "grad_norm": 16.8658447265625,
+      "learning_rate": 8.086466074476562e-07,
+      "loss": 0.2701,
+      "step": 885
+    },
+    {
+      "epoch": 0.03544,
+      "grad_norm": 0.9749300479888916,
+      "learning_rate": 7.949514654755963e-07,
+      "loss": 0.0953,
+      "step": 886
+    },
+    {
+      "epoch": 0.03548,
+      "grad_norm": 9.792045593261719,
+      "learning_rate": 7.81368484114996e-07,
+      "loss": 0.194,
+      "step": 887
+    },
+    {
+      "epoch": 0.03552,
+      "grad_norm": 1.3447405099868774,
+      "learning_rate": 7.678978288701911e-07,
+      "loss": 0.0851,
+      "step": 888
+    },
+    {
+      "epoch": 0.03556,
+      "grad_norm": 1.419684648513794,
+      "learning_rate": 7.545396638768698e-07,
+      "loss": 0.0787,
+      "step": 889
+    },
+    {
+      "epoch": 0.0356,
+      "grad_norm": 10.248550415039062,
+      "learning_rate": 7.412941519000527e-07,
+      "loss": 0.1078,
+      "step": 890
+    },
+    {
+      "epoch": 0.03564,
+      "grad_norm": 0.23045428097248077,
+      "learning_rate": 7.281614543321269e-07,
+      "loss": 0.0053,
+      "step": 891
+    },
+    {
+      "epoch": 0.03568,
+      "grad_norm": 1.2161086797714233,
+      "learning_rate": 7.151417311908648e-07,
+      "loss": 0.0821,
+      "step": 892
+    },
+    {
+      "epoch": 0.03572,
+      "grad_norm": 2.187063694000244,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.1181,
+      "step": 893
+    },
+    {
+      "epoch": 0.03576,
+      "grad_norm": 1.0227221250534058,
+      "learning_rate": 6.894418413747183e-07,
+      "loss": 0.0732,
+      "step": 894
+    },
+    {
+      "epoch": 0.0358,
+      "grad_norm": 1.4367542266845703,
+      "learning_rate": 6.767619878448783e-07,
+      "loss": 0.1242,
+      "step": 895
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 4.0704145431518555,
+      "learning_rate": 6.641957350279838e-07,
+      "loss": 0.0455,
+      "step": 896
+    },
+    {
+      "epoch": 0.03588,
+      "grad_norm": 5.700128078460693,
+      "learning_rate": 6.517432360398556e-07,
+      "loss": 0.1158,
+      "step": 897
+    },
+    {
+      "epoch": 0.03592,
+      "grad_norm": 4.539000988006592,
+      "learning_rate": 6.394046426102673e-07,
+      "loss": 0.1306,
+      "step": 898
+    },
+    {
+      "epoch": 0.03596,
+      "grad_norm": 0.7464107871055603,
+      "learning_rate": 6.271801050810856e-07,
+      "loss": 0.0593,
+      "step": 899
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 1.4375019073486328,
+      "learning_rate": 6.150697724044407e-07,
+      "loss": 0.0833,
+      "step": 900
+    },
+    {
+      "epoch": 0.03604,
+      "grad_norm": 1.3866355419158936,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.0846,
+      "step": 901
+    },
+    {
+      "epoch": 0.03608,
+      "grad_norm": 0.9201920628547668,
+      "learning_rate": 5.911923104577455e-07,
+      "loss": 0.0617,
+      "step": 902
+    },
+    {
+      "epoch": 0.03612,
+      "grad_norm": 0.6311770081520081,
+      "learning_rate": 5.794254721270331e-07,
+      "loss": 0.0643,
+      "step": 903
+    },
+    {
+      "epoch": 0.03616,
+      "grad_norm": 0.8222041726112366,
+      "learning_rate": 5.677734205239904e-07,
+      "loss": 0.0128,
+      "step": 904
+    },
+    {
+      "epoch": 0.0362,
+      "grad_norm": 1.4449503421783447,
+      "learning_rate": 5.562362976251901e-07,
+      "loss": 0.1284,
+      "step": 905
+    },
+    {
+      "epoch": 0.03624,
+      "grad_norm": 1.0866655111312866,
+      "learning_rate": 5.448142440068316e-07,
+      "loss": 0.0889,
+      "step": 906
+    },
+    {
+      "epoch": 0.03628,
+      "grad_norm": 10.500232696533203,
+      "learning_rate": 5.335073988430373e-07,
+      "loss": 0.0727,
+      "step": 907
+    },
+    {
+      "epoch": 0.03632,
+      "grad_norm": 1.060487985610962,
+      "learning_rate": 5.223158999041444e-07,
+      "loss": 0.087,
+      "step": 908
+    },
+    {
+      "epoch": 0.03636,
+      "grad_norm": 3.081806182861328,
+      "learning_rate": 5.112398835550348e-07,
+      "loss": 0.0241,
+      "step": 909
+    },
+    {
+      "epoch": 0.0364,
+      "grad_norm": 2.607464551925659,
+      "learning_rate": 5.002794847534765e-07,
+      "loss": 0.1021,
+      "step": 910
+    },
+    {
+      "epoch": 0.03644,
+      "grad_norm": 9.156312942504883,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.1661,
+      "step": 911
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 1.549623727798462,
+      "learning_rate": 4.787060725786141e-07,
+      "loss": 0.0163,
+      "step": 912
+    },
+    {
+      "epoch": 0.03652,
+      "grad_norm": 1.2202539443969727,
+      "learning_rate": 4.6809332207053083e-07,
+      "loss": 0.0976,
+      "step": 913
+    },
+    {
+      "epoch": 0.03656,
+      "grad_norm": 1.5141152143478394,
+      "learning_rate": 4.575967148372318e-07,
+      "loss": 0.0446,
+      "step": 914
+    },
+    {
+      "epoch": 0.0366,
+      "grad_norm": 1.1905561685562134,
+      "learning_rate": 4.4721637877656377e-07,
+      "loss": 0.0985,
+      "step": 915
+    },
+    {
+      "epoch": 0.03664,
+      "grad_norm": 2.1446890830993652,
+      "learning_rate": 4.3695244036964567e-07,
+      "loss": 0.1046,
+      "step": 916
+    },
+    {
+      "epoch": 0.03668,
+      "grad_norm": 3.9344406127929688,
+      "learning_rate": 4.268050246793276e-07,
+      "loss": 0.1419,
+      "step": 917
+    },
+    {
+      "epoch": 0.03672,
+      "grad_norm": 0.7304320931434631,
+      "learning_rate": 4.167742553486676e-07,
+      "loss": 0.0715,
+      "step": 918
+    },
+    {
+      "epoch": 0.03676,
+      "grad_norm": 1.6561803817749023,
+      "learning_rate": 4.068602545994249e-07,
+      "loss": 0.0869,
+      "step": 919
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 2.3948023319244385,
+      "learning_rate": 3.9706314323056936e-07,
+      "loss": 0.0335,
+      "step": 920
+    },
+    {
+      "epoch": 0.03684,
+      "grad_norm": 3.3992624282836914,
+      "learning_rate": 3.8738304061681107e-07,
+      "loss": 0.1426,
+      "step": 921
+    },
+    {
+      "epoch": 0.03688,
+      "grad_norm": 0.8297941088676453,
+      "learning_rate": 3.7782006470714614e-07,
+      "loss": 0.1059,
+      "step": 922
+    },
+    {
+      "epoch": 0.03692,
+      "grad_norm": 5.778967380523682,
+      "learning_rate": 3.68374332023419e-07,
+      "loss": 0.1153,
+      "step": 923
+    },
+    {
+      "epoch": 0.03696,
+      "grad_norm": 0.74250727891922,
+      "learning_rate": 3.590459576589e-07,
+      "loss": 0.0761,
+      "step": 924
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 27.099145889282227,
+      "learning_rate": 3.498350552768859e-07,
+      "loss": 0.4689,
+      "step": 925
+    },
+    {
+      "epoch": 0.03704,
+      "grad_norm": 0.7006922364234924,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 0.0714,
+      "step": 926
+    },
+    {
+      "epoch": 0.03708,
+      "grad_norm": 10.563543319702148,
+      "learning_rate": 3.3176611395540625e-07,
+      "loss": 0.2877,
+      "step": 927
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.94381183385849,
+      "learning_rate": 3.2290829518028867e-07,
+      "loss": 0.0865,
+      "step": 928
+    },
+    {
+      "epoch": 0.03716,
+      "grad_norm": 2.9429616928100586,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.0303,
+      "step": 929
+    },
+    {
+      "epoch": 0.0372,
+      "grad_norm": 1.4685101509094238,
+      "learning_rate": 3.0554650104861137e-07,
+      "loss": 0.0945,
+      "step": 930
+    },
+    {
+      "epoch": 0.03724,
+      "grad_norm": 1.4263782501220703,
+      "learning_rate": 2.970427372400353e-07,
+      "loss": 0.0838,
+      "step": 931
+    },
+    {
+      "epoch": 0.03728,
+      "grad_norm": 0.6240054368972778,
+      "learning_rate": 2.8865720090364037e-07,
+      "loss": 0.0874,
+      "step": 932
+    },
+    {
+      "epoch": 0.03732,
+      "grad_norm": 0.9149302244186401,
+      "learning_rate": 2.8038999421453827e-07,
+      "loss": 0.0597,
+      "step": 933
+    },
+    {
+      "epoch": 0.03736,
+      "grad_norm": 5.361192226409912,
+      "learning_rate": 2.7224121790603517e-07,
+      "loss": 0.1355,
+      "step": 934
+    },
+    {
+      "epoch": 0.0374,
+      "grad_norm": 1.5716146230697632,
+      "learning_rate": 2.6421097126839714e-07,
+      "loss": 0.0657,
+      "step": 935
+    },
+    {
+      "epoch": 0.03744,
+      "grad_norm": 3.323272228240967,
+      "learning_rate": 2.5629935214764866e-07,
+      "loss": 0.1591,
+      "step": 936
+    },
+    {
+      "epoch": 0.03748,
+      "grad_norm": 2.960763454437256,
+      "learning_rate": 2.4850645694436736e-07,
+      "loss": 0.0861,
+      "step": 937
+    },
+    {
+      "epoch": 0.03752,
+      "grad_norm": 0.1669314205646515,
+      "learning_rate": 2.4083238061252565e-07,
+      "loss": 0.0017,
+      "step": 938
+    },
+    {
+      "epoch": 0.03756,
+      "grad_norm": 5.447648048400879,
+      "learning_rate": 2.332772166583208e-07,
+      "loss": 0.1693,
+      "step": 939
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 0.5760691165924072,
+      "learning_rate": 2.2584105713904126e-07,
+      "loss": 0.0419,
+      "step": 940
+    },
+    {
+      "epoch": 0.03764,
+      "grad_norm": 4.979199409484863,
+      "learning_rate": 2.1852399266194312e-07,
+      "loss": 0.1206,
+      "step": 941
+    },
+    {
+      "epoch": 0.03768,
+      "grad_norm": 6.860245704650879,
+      "learning_rate": 2.1132611238315004e-07,
+      "loss": 0.1544,
+      "step": 942
+    },
+    {
+      "epoch": 0.03772,
+      "grad_norm": 1.1236414909362793,
+      "learning_rate": 2.0424750400655947e-07,
+      "loss": 0.0703,
+      "step": 943
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 1.7695764303207397,
+      "learning_rate": 1.9728825378278248e-07,
+      "loss": 0.0881,
+      "step": 944
+    },
+    {
+      "epoch": 0.0378,
+      "grad_norm": 0.4883894622325897,
+      "learning_rate": 1.9044844650808468e-07,
+      "loss": 0.0458,
+      "step": 945
+    },
+    {
+      "epoch": 0.03784,
+      "grad_norm": 2.519953727722168,
+      "learning_rate": 1.8372816552336025e-07,
+      "loss": 0.122,
+      "step": 946
+    },
+    {
+      "epoch": 0.03788,
+      "grad_norm": 5.409066200256348,
+      "learning_rate": 1.7712749271311392e-07,
+      "loss": 0.1894,
+      "step": 947
+    },
+    {
+      "epoch": 0.03792,
+      "grad_norm": 1.3381481170654297,
+      "learning_rate": 1.706465085044584e-07,
+      "loss": 0.0874,
+      "step": 948
+    },
+    {
+      "epoch": 0.03796,
+      "grad_norm": 0.5155236124992371,
+      "learning_rate": 1.6428529186614195e-07,
+      "loss": 0.0588,
+      "step": 949
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 1.525451898574829,
+      "learning_rate": 1.580439203075812e-07,
+      "loss": 0.0838,
+      "step": 950
+    },
+    {
+      "epoch": 0.03804,
+      "grad_norm": 0.7758452296257019,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 0.0752,
+      "step": 951
+    },
+    {
+      "epoch": 0.03808,
+      "grad_norm": 0.34528812766075134,
+      "learning_rate": 1.4592101516509916e-07,
+      "loss": 0.007,
+      "step": 952
+    },
+    {
+      "epoch": 0.03812,
+      "grad_norm": 7.501368999481201,
+      "learning_rate": 1.400396292949513e-07,
+      "loss": 0.1826,
+      "step": 953
+    },
+    {
+      "epoch": 0.03816,
+      "grad_norm": 0.36205366253852844,
+      "learning_rate": 1.3427838393030634e-07,
+      "loss": 0.0415,
+      "step": 954
+    },
+    {
+      "epoch": 0.0382,
+      "grad_norm": 8.520697593688965,
+      "learning_rate": 1.2863734927012094e-07,
+      "loss": 0.1506,
+      "step": 955
+    },
+    {
+      "epoch": 0.03824,
+      "grad_norm": 0.6122540831565857,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.0508,
+      "step": 956
+    },
+    {
+      "epoch": 0.03828,
+      "grad_norm": 1.9029324054718018,
+      "learning_rate": 1.1771618553447217e-07,
+      "loss": 0.1221,
+      "step": 957
+    },
+    {
+      "epoch": 0.03832,
+      "grad_norm": 0.8956604599952698,
+      "learning_rate": 1.1243618952994195e-07,
+      "loss": 0.0756,
+      "step": 958
+    },
+    {
+      "epoch": 0.03836,
+      "grad_norm": 3.979543685913086,
+      "learning_rate": 1.0727667037011668e-07,
+      "loss": 0.1272,
+      "step": 959
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 5.174049377441406,
+      "learning_rate": 1.0223769092211012e-07,
+      "loss": 0.1288,
+      "step": 960
+    },
+    {
+      "epoch": 0.03844,
+      "grad_norm": 5.619006156921387,
+      "learning_rate": 9.731931258429638e-08,
+      "loss": 0.2043,
+      "step": 961
+    },
+    {
+      "epoch": 0.03848,
+      "grad_norm": 15.900406837463379,
+      "learning_rate": 9.252159528556404e-08,
+      "loss": 0.406,
+      "step": 962
+    },
+    {
+      "epoch": 0.03852,
+      "grad_norm": 1.215280294418335,
+      "learning_rate": 8.784459748458318e-08,
+      "loss": 0.085,
+      "step": 963
+    },
+    {
+      "epoch": 0.03856,
+      "grad_norm": 1.2929669618606567,
+      "learning_rate": 8.328837616909612e-08,
+      "loss": 0.0793,
+      "step": 964
+    },
+    {
+      "epoch": 0.0386,
+      "grad_norm": 1.868042230606079,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.0764,
+      "step": 965
+    },
+    {
+      "epoch": 0.03864,
+      "grad_norm": 0.8173092603683472,
+      "learning_rate": 7.453848358678018e-08,
+      "loss": 0.0644,
+      "step": 966
+    },
+    {
+      "epoch": 0.03868,
+      "grad_norm": 0.7322748303413391,
+      "learning_rate": 7.034491893463059e-08,
+      "loss": 0.0635,
+      "step": 967
+    },
+    {
+      "epoch": 0.03872,
+      "grad_norm": 2.6058056354522705,
+      "learning_rate": 6.627234399603554e-08,
+      "loss": 0.0731,
+      "step": 968
+    },
+    {
+      "epoch": 0.03876,
+      "grad_norm": 1.2672457695007324,
+      "learning_rate": 6.232080839403631e-08,
+      "loss": 0.0829,
+      "step": 969
+    },
+    {
+      "epoch": 0.0388,
+      "grad_norm": 8.187602043151855,
+      "learning_rate": 5.849036027684607e-08,
+      "loss": 0.0919,
+      "step": 970
+    },
+    {
+      "epoch": 0.03884,
+      "grad_norm": 15.100421905517578,
+      "learning_rate": 5.4781046317267103e-08,
+      "loss": 0.3653,
+      "step": 971
+    },
+    {
+      "epoch": 0.03888,
+      "grad_norm": 0.7564333081245422,
+      "learning_rate": 5.119291171211793e-08,
+      "loss": 0.0483,
+      "step": 972
+    },
+    {
+      "epoch": 0.03892,
+      "grad_norm": 0.9435470700263977,
+      "learning_rate": 4.772600018168816e-08,
+      "loss": 0.0647,
+      "step": 973
+    },
+    {
+      "epoch": 0.03896,
+      "grad_norm": 7.892863750457764,
+      "learning_rate": 4.438035396920004e-08,
+      "loss": 0.1639,
+      "step": 974
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 1.6465458869934082,
+      "learning_rate": 4.115601384029666e-08,
+      "loss": 0.0853,
+      "step": 975
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 3.5936317443847656,
+      "learning_rate": 3.805301908254455e-08,
+      "loss": 0.1082,
+      "step": 976
+    },
+    {
+      "epoch": 0.03908,
+      "grad_norm": 1.2942278385162354,
+      "learning_rate": 3.50714075049563e-08,
+      "loss": 0.0734,
+      "step": 977
+    },
+    {
+      "epoch": 0.03912,
+      "grad_norm": 24.666431427001953,
+      "learning_rate": 3.22112154375287e-08,
+      "loss": 0.2415,
+      "step": 978
+    },
+    {
+      "epoch": 0.03916,
+      "grad_norm": 1.4653372764587402,
+      "learning_rate": 2.947247773079753e-08,
+      "loss": 0.0609,
+      "step": 979
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 3.331188440322876,
+      "learning_rate": 2.6855227755419046e-08,
+      "loss": 0.0926,
+      "step": 980
+    },
+    {
+      "epoch": 0.03924,
+      "grad_norm": 10.378106117248535,
+      "learning_rate": 2.4359497401758026e-08,
+      "loss": 0.1696,
+      "step": 981
+    },
+    {
+      "epoch": 0.03928,
+      "grad_norm": 0.7052443027496338,
+      "learning_rate": 2.1985317079500358e-08,
+      "loss": 0.0743,
+      "step": 982
+    },
+    {
+      "epoch": 0.03932,
+      "grad_norm": 4.7255401611328125,
+      "learning_rate": 1.973271571728441e-08,
+      "loss": 0.1439,
+      "step": 983
+    },
+    {
+      "epoch": 0.03936,
+      "grad_norm": 0.7315016388893127,
+      "learning_rate": 1.7601720762346895e-08,
+      "loss": 0.0818,
+      "step": 984
+    },
+    {
+      "epoch": 0.0394,
+      "grad_norm": 12.701497077941895,
+      "learning_rate": 1.5592358180189782e-08,
+      "loss": 0.2157,
+      "step": 985
+    },
+    {
+      "epoch": 0.03944,
+      "grad_norm": 1.58987295627594,
+      "learning_rate": 1.370465245426167e-08,
+      "loss": 0.1019,
+      "step": 986
+    },
+    {
+      "epoch": 0.03948,
+      "grad_norm": 3.5525593757629395,
+      "learning_rate": 1.1938626585660252e-08,
+      "loss": 0.1624,
+      "step": 987
+    },
+    {
+      "epoch": 0.03952,
+      "grad_norm": 0.7596866488456726,
+      "learning_rate": 1.0294302092853647e-08,
+      "loss": 0.0644,
+      "step": 988
+    },
+    {
+      "epoch": 0.03956,
+      "grad_norm": 2.9007773399353027,
+      "learning_rate": 8.771699011416169e-09,
+      "loss": 0.0722,
+      "step": 989
+    },
+    {
+      "epoch": 0.0396,
+      "grad_norm": 0.923038899898529,
+      "learning_rate": 7.370835893788508e-09,
+      "loss": 0.0547,
+      "step": 990
+    },
+    {
+      "epoch": 0.03964,
+      "grad_norm": 11.018105506896973,
+      "learning_rate": 6.091729809042379e-09,
+      "loss": 0.2337,
+      "step": 991
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.34633147716522217,
+      "learning_rate": 4.9343963426840006e-09,
+      "loss": 0.0327,
+      "step": 992
+    },
+    {
+      "epoch": 0.03972,
+      "grad_norm": 3.647501230239868,
+      "learning_rate": 3.898849596456477e-09,
+      "loss": 0.1359,
+      "step": 993
+    },
+    {
+      "epoch": 0.03976,
+      "grad_norm": 1.3227728605270386,
+      "learning_rate": 2.9851021881688314e-09,
+      "loss": 0.093,
+      "step": 994
+    },
+    {
+      "epoch": 0.0398,
+      "grad_norm": 1.4203863143920898,
+      "learning_rate": 2.193165251545004e-09,
+      "loss": 0.1156,
+      "step": 995
+    },
+    {
+      "epoch": 0.03984,
+      "grad_norm": 0.789050281047821,
+      "learning_rate": 1.5230484360873043e-09,
+      "loss": 0.0812,
+      "step": 996
+    },
+    {
+      "epoch": 0.03988,
+      "grad_norm": 3.081073760986328,
+      "learning_rate": 9.74759906957612e-10,
+      "loss": 0.1119,
+      "step": 997
+    },
+    {
+      "epoch": 0.03992,
+      "grad_norm": 1.709783911705017,
+      "learning_rate": 5.483063448785686e-10,
+      "loss": 0.0756,
+      "step": 998
+    },
+    {
+      "epoch": 0.03996,
+      "grad_norm": 0.7825138568878174,
+      "learning_rate": 2.436929460525317e-10,
+      "loss": 0.0582,
+      "step": 999
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7134876251220703,
+      "learning_rate": 6.092342209607083e-11,
+      "loss": 0.046,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}