diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.15345661014348194,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010230440676232128,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.751,
+      "step": 1
+    },
+    {
+      "epoch": 0.00020460881352464257,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.8722,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003069132202869639,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3e-06,
+      "loss": 1.5554,
+      "step": 3
+    },
+    {
+      "epoch": 0.00040921762704928513,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.8484,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005115220338116064,
+      "grad_norm": 0.65625,
+      "learning_rate": 5e-06,
+      "loss": 1.6083,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006138264405739278,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6e-06,
+      "loss": 1.8859,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007161308473362491,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.5886,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008184352540985703,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.6189,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009207396608608916,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9e-06,
+      "loss": 1.7513,
+      "step": 9
+    },
+    {
+      "epoch": 0.0010230440676232129,
+      "grad_norm": 0.640625,
+      "learning_rate": 1e-05,
+      "loss": 1.8472,
+      "step": 10
+    },
+    {
+      "epoch": 0.001125348474385534,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.5958,
+      "step": 11
+    },
+    {
+      "epoch": 0.0012276528811478555,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2e-05,
+      "loss": 1.7285,
+      "step": 12
+    },
+    {
+      "epoch": 0.0013299572879101767,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.6828,
+      "step": 13
+    },
+    {
+      "epoch": 0.0014322616946724981,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.8157,
+      "step": 14
+    },
+    {
+      "epoch": 0.0015345661014348193,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5e-05,
+      "loss": 1.9777,
+      "step": 15
+    },
+    {
+      "epoch": 0.0016368705081971405,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7051,
+      "step": 16
+    },
+    {
+      "epoch": 0.001739174914959462,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.8864,
+      "step": 17
+    },
+    {
+      "epoch": 0.0018414793217217831,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8e-05,
+      "loss": 1.7858,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019437837284841043,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9e-05,
+      "loss": 1.7346,
+      "step": 19
+    },
+    {
+      "epoch": 0.0020460881352464258,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2e-05,
+      "loss": 1.9046,
+      "step": 20
+    },
+    {
+      "epoch": 0.002148392542008747,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.1e-05,
+      "loss": 1.8649,
+      "step": 21
+    },
+    {
+      "epoch": 0.002250696948771068,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.9158,
+      "step": 22
+    },
+    {
+      "epoch": 0.0023530013555333896,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.6108,
+      "step": 23
+    },
+    {
+      "epoch": 0.002455305762295711,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4e-05,
+      "loss": 1.7164,
+      "step": 24
+    },
+    {
+      "epoch": 0.002557610169058032,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.5e-05,
+      "loss": 1.9596,
+      "step": 25
+    },
+    {
+      "epoch": 0.0026599145758203534,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.0135,
+      "step": 26
+    },
+    {
+      "epoch": 0.002762218982582675,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.8614,
+      "step": 27
+    },
+    {
+      "epoch": 0.0028645233893449962,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.6934,
+      "step": 28
+    },
+    {
+      "epoch": 0.0029668277961073172,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.9e-05,
+      "loss": 1.8986,
+      "step": 29
+    },
+    {
+      "epoch": 0.0030691322028696386,
+      "grad_norm": 0.625,
+      "learning_rate": 3e-05,
+      "loss": 1.9362,
+      "step": 30
+    },
+    {
+      "epoch": 0.00317143660963196,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.1e-05,
+      "loss": 1.8173,
+      "step": 31
+    },
+    {
+      "epoch": 0.003273741016394281,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.9652,
+      "step": 32
+    },
+    {
+      "epoch": 0.0033760454231566025,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.3e-05,
+      "loss": 1.8063,
+      "step": 33
+    },
+    {
+      "epoch": 0.003478349829918924,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.804,
+      "step": 34
+    },
+    {
+      "epoch": 0.003580654236681245,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.5e-05,
+      "loss": 1.969,
+      "step": 35
+    },
+    {
+      "epoch": 0.0036829586434435663,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7977,
+      "step": 36
+    },
+    {
+      "epoch": 0.0037852630502058877,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.7e-05,
+      "loss": 1.8371,
+      "step": 37
+    },
+    {
+      "epoch": 0.0038875674569682087,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.8e-05,
+      "loss": 1.6842,
+      "step": 38
+    },
+    {
+      "epoch": 0.0039898718637305305,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.57,
+      "step": 39
+    },
+    {
+      "epoch": 0.0040921762704928515,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4e-05,
+      "loss": 1.8302,
+      "step": 40
+    },
+    {
+      "epoch": 0.0041944806772551725,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.1e-05,
+      "loss": 1.6632,
+      "step": 41
+    },
+    {
+      "epoch": 0.004296785084017494,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.2e-05,
+      "loss": 1.8011,
+      "step": 42
+    },
+    {
+      "epoch": 0.004399089490779815,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.3e-05,
+      "loss": 1.7171,
+      "step": 43
+    },
+    {
+      "epoch": 0.004501393897542136,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.7685,
+      "step": 44
+    },
+    {
+      "epoch": 0.004603698304304458,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.5e-05,
+      "loss": 1.6182,
+      "step": 45
+    },
+    {
+      "epoch": 0.004706002711066779,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.8105,
+      "step": 46
+    },
+    {
+      "epoch": 0.0048083071178291,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.7e-05,
+      "loss": 1.9284,
+      "step": 47
+    },
+    {
+      "epoch": 0.004910611524591422,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.8e-05,
+      "loss": 1.8883,
+      "step": 48
+    },
+    {
+      "epoch": 0.005012915931353743,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9e-05,
+      "loss": 1.6786,
+      "step": 49
+    },
+    {
+      "epoch": 0.005115220338116064,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5e-05,
+      "loss": 1.8476,
+      "step": 50
+    },
+    {
+      "epoch": 0.005217524744878386,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 1.6703,
+      "step": 51
+    },
+    {
+      "epoch": 0.005319829151640707,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.7675,
+      "step": 52
+    },
+    {
+      "epoch": 0.005422133558403028,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 1.5959,
+      "step": 53
+    },
+    {
+      "epoch": 0.00552443796516535,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.7945,
+      "step": 54
+    },
+    {
+      "epoch": 0.005626742371927671,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.7426,
+      "step": 55
+    },
+    {
+      "epoch": 0.0057290467786899925,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.8982,
+      "step": 56
+    },
+    {
+      "epoch": 0.0058313511854523135,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.9036,
+      "step": 57
+    },
+    {
+      "epoch": 0.0059336555922146345,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.8e-05,
+      "loss": 1.7437,
+      "step": 58
+    },
+    {
+      "epoch": 0.006035959998976956,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.9e-05,
+      "loss": 1.7137,
+      "step": 59
+    },
+    {
+      "epoch": 0.006138264405739277,
+      "grad_norm": 0.546875,
+      "learning_rate": 6e-05,
+      "loss": 1.8123,
+      "step": 60
+    },
+    {
+      "epoch": 0.006240568812501598,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.1e-05,
+      "loss": 1.6035,
+      "step": 61
+    },
+    {
+      "epoch": 0.00634287321926392,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.2e-05,
+      "loss": 1.7869,
+      "step": 62
+    },
+    {
+      "epoch": 0.006445177626026241,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.3e-05,
+      "loss": 1.559,
+      "step": 63
+    },
+    {
+      "epoch": 0.006547482032788562,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.7037,
+      "step": 64
+    },
+    {
+      "epoch": 0.006649786439550884,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.7837,
+      "step": 65
+    },
+    {
+      "epoch": 0.006752090846313205,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.6e-05,
+      "loss": 1.6767,
+      "step": 66
+    },
+    {
+      "epoch": 0.006854395253075526,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.7e-05,
+      "loss": 1.8754,
+      "step": 67
+    },
+    {
+      "epoch": 0.006956699659837848,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.8037,
+      "step": 68
+    },
+    {
+      "epoch": 0.007059004066600169,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.9e-05,
+      "loss": 1.9943,
+      "step": 69
+    },
+    {
+      "epoch": 0.00716130847336249,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7e-05,
+      "loss": 1.6343,
+      "step": 70
+    },
+    {
+      "epoch": 0.007263612880124812,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.1e-05,
+      "loss": 1.7749,
+      "step": 71
+    },
+    {
+      "epoch": 0.007365917286887133,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.2e-05,
+      "loss": 1.9159,
+      "step": 72
+    },
+    {
+      "epoch": 0.0074682216936494536,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.3e-05,
+      "loss": 1.663,
+      "step": 73
+    },
+    {
+      "epoch": 0.007570526100411775,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.4e-05,
+      "loss": 1.7922,
+      "step": 74
+    },
+    {
+      "epoch": 0.007672830507174096,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.8486,
+      "step": 75
+    },
+    {
+      "epoch": 0.007775134913936417,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.6e-05,
+      "loss": 1.5211,
+      "step": 76
+    },
+    {
+      "epoch": 0.007877439320698738,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.7e-05,
+      "loss": 1.7232,
+      "step": 77
+    },
+    {
+      "epoch": 0.007979743727461061,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.5951,
+      "step": 78
+    },
+    {
+      "epoch": 0.008082048134223382,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 1.7327,
+      "step": 79
+    },
+    {
+      "epoch": 0.008184352540985703,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 80
+    },
+    {
+      "epoch": 0.008286656947748024,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.1e-05,
+      "loss": 1.6925,
+      "step": 81
+    },
+    {
+      "epoch": 0.008388961354510345,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.2e-05,
+      "loss": 1.6937,
+      "step": 82
+    },
+    {
+      "epoch": 0.008491265761272666,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.3e-05,
+      "loss": 1.5959,
+      "step": 83
+    },
+    {
+      "epoch": 0.008593570168034989,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.4e-05,
+      "loss": 1.7659,
+      "step": 84
+    },
+    {
+      "epoch": 0.00869587457479731,
+      "grad_norm": 0.48828125,
+      "learning_rate": 8.5e-05,
+      "loss": 1.8178,
+      "step": 85
+    },
+    {
+      "epoch": 0.00879817898155963,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.6e-05,
+      "loss": 1.6062,
+      "step": 86
+    },
+    {
+      "epoch": 0.008900483388321952,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.7e-05,
+      "loss": 1.7386,
+      "step": 87
+    },
+    {
+      "epoch": 0.009002787795084273,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.5267,
+      "step": 88
+    },
+    {
+      "epoch": 0.009105092201846595,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 1.6326,
+      "step": 89
+    },
+    {
+      "epoch": 0.009207396608608916,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9e-05,
+      "loss": 1.4584,
+      "step": 90
+    },
+    {
+      "epoch": 0.009309701015371237,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.1e-05,
+      "loss": 1.6442,
+      "step": 91
+    },
+    {
+      "epoch": 0.009412005422133558,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.6235,
+      "step": 92
+    },
+    {
+      "epoch": 0.00951430982889588,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.5549,
+      "step": 93
+    },
+    {
+      "epoch": 0.0096166142356582,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.4e-05,
+      "loss": 1.6552,
+      "step": 94
+    },
+    {
+      "epoch": 0.009718918642420523,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.5e-05,
+      "loss": 1.582,
+      "step": 95
+    },
+    {
+      "epoch": 0.009821223049182844,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.6e-05,
+      "loss": 1.6172,
+      "step": 96
+    },
+    {
+      "epoch": 0.009923527455945165,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.7e-05,
+      "loss": 1.7113,
+      "step": 97
+    },
+    {
+      "epoch": 0.010025831862707486,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.8e-05,
+      "loss": 1.6369,
+      "step": 98
+    },
+    {
+      "epoch": 0.010128136269469807,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.5055,
+      "step": 99
+    },
+    {
+      "epoch": 0.010230440676232128,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 100
+    },
+    {
+      "epoch": 0.01033274508299445,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.999497487437187e-05,
+      "loss": 1.5179,
+      "step": 101
+    },
+    {
+      "epoch": 0.010435049489756772,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.998994974874373e-05,
+      "loss": 1.4241,
+      "step": 102
+    },
+    {
+      "epoch": 0.010537353896519093,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.998492462311558e-05,
+      "loss": 1.7416,
+      "step": 103
+    },
+    {
+      "epoch": 0.010639658303281414,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.997989949748744e-05,
+      "loss": 1.7483,
+      "step": 104
+    },
+    {
+      "epoch": 0.010741962710043735,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.99748743718593e-05,
+      "loss": 1.7977,
+      "step": 105
+    },
+    {
+      "epoch": 0.010844267116806056,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.996984924623116e-05,
+      "loss": 1.5858,
+      "step": 106
+    },
+    {
+      "epoch": 0.010946571523568378,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.996482412060301e-05,
+      "loss": 1.6459,
+      "step": 107
+    },
+    {
+      "epoch": 0.0110488759303307,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.995979899497487e-05,
+      "loss": 2.6572,
+      "step": 108
+    },
+    {
+      "epoch": 0.01115118033709302,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.995477386934674e-05,
+      "loss": 1.5429,
+      "step": 109
+    },
+    {
+      "epoch": 0.011253484743855341,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.99497487437186e-05,
+      "loss": 1.5996,
+      "step": 110
+    },
+    {
+      "epoch": 0.011355789150617662,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.994472361809045e-05,
+      "loss": 1.6824,
+      "step": 111
+    },
+    {
+      "epoch": 0.011458093557379985,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.993969849246232e-05,
+      "loss": 1.5821,
+      "step": 112
+    },
+    {
+      "epoch": 0.011560397964142306,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.993467336683417e-05,
+      "loss": 1.6859,
+      "step": 113
+    },
+    {
+      "epoch": 0.011662702370904627,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.992964824120603e-05,
+      "loss": 1.4029,
+      "step": 114
+    },
+    {
+      "epoch": 0.011765006777666948,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.99246231155779e-05,
+      "loss": 1.5493,
+      "step": 115
+    },
+    {
+      "epoch": 0.011867311184429269,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.991959798994976e-05,
+      "loss": 1.5848,
+      "step": 116
+    },
+    {
+      "epoch": 0.01196961559119159,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.99145728643216e-05,
+      "loss": 1.5165,
+      "step": 117
+    },
+    {
+      "epoch": 0.012071919997953913,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.990954773869348e-05,
+      "loss": 1.5062,
+      "step": 118
+    },
+    {
+      "epoch": 0.012174224404716234,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.990452261306533e-05,
+      "loss": 1.7182,
+      "step": 119
+    },
+    {
+      "epoch": 0.012276528811478555,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.989949748743719e-05,
+      "loss": 1.675,
+      "step": 120
+    },
+    {
+      "epoch": 0.012378833218240876,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.989447236180905e-05,
+      "loss": 1.6604,
+      "step": 121
+    },
+    {
+      "epoch": 0.012481137625003197,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.988944723618092e-05,
+      "loss": 1.6381,
+      "step": 122
+    },
+    {
+      "epoch": 0.012583442031765518,
+      "grad_norm": 0.375,
+      "learning_rate": 9.988442211055276e-05,
+      "loss": 1.6174,
+      "step": 123
+    },
+    {
+      "epoch": 0.01268574643852784,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.987939698492463e-05,
+      "loss": 1.546,
+      "step": 124
+    },
+    {
+      "epoch": 0.012788050845290161,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.987437185929649e-05,
+      "loss": 1.6795,
+      "step": 125
+    },
+    {
+      "epoch": 0.012890355252052482,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.986934673366835e-05,
+      "loss": 1.5977,
+      "step": 126
+    },
+    {
+      "epoch": 0.012992659658814803,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.98643216080402e-05,
+      "loss": 1.5735,
+      "step": 127
+    },
+    {
+      "epoch": 0.013094964065577124,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.985929648241207e-05,
+      "loss": 1.4447,
+      "step": 128
+    },
+    {
+      "epoch": 0.013197268472339445,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.985427135678392e-05,
+      "loss": 1.5235,
+      "step": 129
+    },
+    {
+      "epoch": 0.013299572879101768,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.984924623115578e-05,
+      "loss": 1.63,
+      "step": 130
+    },
+    {
+      "epoch": 0.013401877285864089,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.984422110552765e-05,
+      "loss": 1.6612,
+      "step": 131
+    },
+    {
+      "epoch": 0.01350418169262641,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.983919597989951e-05,
+      "loss": 1.6085,
+      "step": 132
+    },
+    {
+      "epoch": 0.01360648609938873,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.983417085427136e-05,
+      "loss": 1.5909,
+      "step": 133
+    },
+    {
+      "epoch": 0.013708790506151052,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.982914572864322e-05,
+      "loss": 1.5974,
+      "step": 134
+    },
+    {
+      "epoch": 0.013811094912913375,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.982412060301508e-05,
+      "loss": 1.4692,
+      "step": 135
+    },
+    {
+      "epoch": 0.013913399319675696,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.981909547738694e-05,
+      "loss": 1.4972,
+      "step": 136
+    },
+    {
+      "epoch": 0.014015703726438017,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.98140703517588e-05,
+      "loss": 1.6054,
+      "step": 137
+    },
+    {
+      "epoch": 0.014118008133200338,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.980904522613065e-05,
+      "loss": 1.5417,
+      "step": 138
+    },
+    {
+      "epoch": 0.014220312539962658,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.980402010050252e-05,
+      "loss": 1.6078,
+      "step": 139
+    },
+    {
+      "epoch": 0.01432261694672498,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.979899497487438e-05,
+      "loss": 1.4736,
+      "step": 140
+    },
+    {
+      "epoch": 0.014424921353487302,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.979396984924624e-05,
+      "loss": 1.6093,
+      "step": 141
+    },
+    {
+      "epoch": 0.014527225760249623,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.978894472361809e-05,
+      "loss": 1.5488,
+      "step": 142
+    },
+    {
+      "epoch": 0.014629530167011944,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.978391959798995e-05,
+      "loss": 1.526,
+      "step": 143
+    },
+    {
+      "epoch": 0.014731834573774265,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.977889447236181e-05,
+      "loss": 1.7257,
+      "step": 144
+    },
+    {
+      "epoch": 0.014834138980536586,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.977386934673367e-05,
+      "loss": 1.7617,
+      "step": 145
+    },
+    {
+      "epoch": 0.014936443387298907,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.976884422110552e-05,
+      "loss": 1.8159,
+      "step": 146
+    },
+    {
+      "epoch": 0.01503874779406123,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.97638190954774e-05,
+      "loss": 1.5534,
+      "step": 147
+    },
+    {
+      "epoch": 0.01514105220082355,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.975879396984925e-05,
+      "loss": 1.6115,
+      "step": 148
+    },
+    {
+      "epoch": 0.015243356607585872,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.975376884422111e-05,
+      "loss": 1.7144,
+      "step": 149
+    },
+    {
+      "epoch": 0.015345661014348193,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.974874371859297e-05,
+      "loss": 1.4684,
+      "step": 150
+    },
+    {
+      "epoch": 0.015447965421110514,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.974371859296483e-05,
+      "loss": 1.5664,
+      "step": 151
+    },
+    {
+      "epoch": 0.015550269827872835,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.973869346733668e-05,
+      "loss": 1.6773,
+      "step": 152
+    },
+    {
+      "epoch": 0.015652574234635157,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.973366834170856e-05,
+      "loss": 1.6853,
+      "step": 153
+    },
+    {
+      "epoch": 0.015754878641397477,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.97286432160804e-05,
+      "loss": 1.7032,
+      "step": 154
+    },
+    {
+      "epoch": 0.0158571830481598,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.972361809045227e-05,
+      "loss": 1.475,
+      "step": 155
+    },
+    {
+      "epoch": 0.015959487454922122,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.971859296482412e-05,
+      "loss": 1.5864,
+      "step": 156
+    },
+    {
+      "epoch": 0.01606179186168444,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.971356783919599e-05,
+      "loss": 1.5935,
+      "step": 157
+    },
+    {
+      "epoch": 0.016164096268446764,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.970854271356784e-05,
+      "loss": 1.7352,
+      "step": 158
+    },
+    {
+      "epoch": 0.016266400675209083,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.97035175879397e-05,
+      "loss": 1.7007,
+      "step": 159
+    },
+    {
+      "epoch": 0.016368705081971406,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.969849246231156e-05,
+      "loss": 1.4993,
+      "step": 160
+    },
+    {
+      "epoch": 0.01647100948873373,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.969346733668343e-05,
+      "loss": 1.6352,
+      "step": 161
+    },
+    {
+      "epoch": 0.016573313895496048,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.968844221105527e-05,
+      "loss": 1.5778,
+      "step": 162
+    },
+    {
+      "epoch": 0.01667561830225837,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.968341708542715e-05,
+      "loss": 1.4844,
+      "step": 163
+    },
+    {
+      "epoch": 0.01677792270902069,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.9678391959799e-05,
+      "loss": 1.3593,
+      "step": 164
+    },
+    {
+      "epoch": 0.016880227115783013,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.967336683417086e-05,
+      "loss": 1.578,
+      "step": 165
+    },
+    {
+      "epoch": 0.016982531522545332,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.966834170854272e-05,
+      "loss": 1.5978,
+      "step": 166
+    },
+    {
+      "epoch": 0.017084835929307655,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.966331658291458e-05,
+      "loss": 1.4406,
+      "step": 167
+    },
+    {
+      "epoch": 0.017187140336069977,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.965829145728643e-05,
+      "loss": 1.4486,
+      "step": 168
+    },
+    {
+      "epoch": 0.017289444742832297,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.96532663316583e-05,
+      "loss": 1.5925,
+      "step": 169
+    },
+    {
+      "epoch": 0.01739174914959462,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.964824120603016e-05,
+      "loss": 1.6354,
+      "step": 170
+    },
+    {
+      "epoch": 0.01749405355635694,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.964321608040202e-05,
+      "loss": 1.487,
+      "step": 171
+    },
+    {
+      "epoch": 0.01759635796311926,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.963819095477387e-05,
+      "loss": 1.6157,
+      "step": 172
+    },
+    {
+      "epoch": 0.017698662369881584,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.963316582914573e-05,
+      "loss": 1.5149,
+      "step": 173
+    },
+    {
+      "epoch": 0.017800966776643903,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.962814070351759e-05,
+      "loss": 1.5642,
+      "step": 174
+    },
+    {
+      "epoch": 0.017903271183406226,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.962311557788945e-05,
+      "loss": 1.6143,
+      "step": 175
+    },
+    {
+      "epoch": 0.018005575590168545,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.961809045226132e-05,
+      "loss": 1.5505,
+      "step": 176
+    },
+    {
+      "epoch": 0.018107879996930868,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.961306532663316e-05,
+      "loss": 1.4731,
+      "step": 177
+    },
+    {
+      "epoch": 0.01821018440369319,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.960804020100503e-05,
+      "loss": 1.4846,
+      "step": 178
+    },
+    {
+      "epoch": 0.01831248881045551,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.960301507537689e-05,
+      "loss": 1.3875,
+      "step": 179
+    },
+    {
+      "epoch": 0.018414793217217833,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.959798994974875e-05,
+      "loss": 1.5664,
+      "step": 180
+    },
+    {
+      "epoch": 0.018517097623980152,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.95929648241206e-05,
+      "loss": 1.5777,
+      "step": 181
+    },
+    {
+      "epoch": 0.018619402030742475,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.958793969849247e-05,
+      "loss": 1.4144,
+      "step": 182
+    },
+    {
+      "epoch": 0.018721706437504794,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.958291457286432e-05,
+      "loss": 1.5068,
+      "step": 183
+    },
+    {
+      "epoch": 0.018824010844267117,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.957788944723619e-05,
+      "loss": 1.5641,
+      "step": 184
+    },
+    {
+      "epoch": 0.01892631525102944,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.957286432160805e-05,
+      "loss": 1.4061,
+      "step": 185
+    },
+    {
+      "epoch": 0.01902861965779176,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.956783919597991e-05,
+      "loss": 1.5301,
+      "step": 186
+    },
+    {
+      "epoch": 0.01913092406455408,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.956281407035176e-05,
+      "loss": 1.583,
+      "step": 187
+    },
+    {
+      "epoch": 0.0192332284713164,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.955778894472362e-05,
+      "loss": 1.4345,
+      "step": 188
+    },
+    {
+      "epoch": 0.019335532878078723,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.955276381909548e-05,
+      "loss": 1.4498,
+      "step": 189
+    },
+    {
+      "epoch": 0.019437837284841046,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.954773869346734e-05,
+      "loss": 1.523,
+      "step": 190
+    },
+    {
+      "epoch": 0.019540141691603365,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.954271356783919e-05,
+      "loss": 1.5828,
+      "step": 191
+    },
+    {
+      "epoch": 0.019642446098365688,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.953768844221107e-05,
+      "loss": 1.4512,
+      "step": 192
+    },
+    {
+      "epoch": 0.019744750505128007,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.953266331658292e-05,
+      "loss": 1.4966,
+      "step": 193
+    },
+    {
+      "epoch": 0.01984705491189033,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.952763819095478e-05,
+      "loss": 1.6731,
+      "step": 194
+    },
+    {
+      "epoch": 0.01994935931865265,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.952261306532664e-05,
+      "loss": 1.7775,
+      "step": 195
+    },
+    {
+      "epoch": 0.020051663725414972,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.95175879396985e-05,
+      "loss": 1.4147,
+      "step": 196
+    },
+    {
+      "epoch": 0.020153968132177295,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.951256281407035e-05,
+      "loss": 1.7135,
+      "step": 197
+    },
+    {
+      "epoch": 0.020256272538939614,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.950753768844223e-05,
+      "loss": 1.5537,
+      "step": 198
+    },
+    {
+      "epoch": 0.020358576945701937,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.950251256281408e-05,
+      "loss": 1.6574,
+      "step": 199
+    },
+    {
+      "epoch": 0.020460881352464256,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.949748743718594e-05,
+      "loss": 1.3233,
+      "step": 200
+    },
+    {
+      "epoch": 0.02056318575922658,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.94924623115578e-05,
+      "loss": 1.7198,
+      "step": 201
+    },
+    {
+      "epoch": 0.0206654901659889,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.948743718592966e-05,
+      "loss": 1.435,
+      "step": 202
+    },
+    {
+      "epoch": 0.02076779457275122,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.948241206030151e-05,
+      "loss": 1.51,
+      "step": 203
+    },
+    {
+      "epoch": 0.020870098979513543,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.947738693467337e-05,
+      "loss": 1.4164,
+      "step": 204
+    },
+    {
+      "epoch": 0.020972403386275863,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.947236180904523e-05,
+      "loss": 1.5362,
+      "step": 205
+    },
+    {
+      "epoch": 0.021074707793038185,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.94673366834171e-05,
+      "loss": 1.5413,
+      "step": 206
+    },
+    {
+      "epoch": 0.021177012199800508,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.946231155778894e-05,
+      "loss": 1.5644,
+      "step": 207
+    },
+    {
+      "epoch": 0.021279316606562827,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.94572864321608e-05,
+      "loss": 1.5051,
+      "step": 208
+    },
+    {
+      "epoch": 0.02138162101332515,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.945226130653267e-05,
+      "loss": 1.564,
+      "step": 209
+    },
+    {
+      "epoch": 0.02148392542008747,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.944723618090453e-05,
+      "loss": 1.4634,
+      "step": 210
+    },
+    {
+      "epoch": 0.021586229826849792,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.944221105527639e-05,
+      "loss": 1.508,
+      "step": 211
+    },
+    {
+      "epoch": 0.02168853423361211,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.943718592964824e-05,
+      "loss": 1.377,
+      "step": 212
+    },
+    {
+      "epoch": 0.021790838640374434,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.94321608040201e-05,
+      "loss": 1.5139,
+      "step": 213
+    },
+    {
+      "epoch": 0.021893143047136757,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.942713567839197e-05,
+      "loss": 1.4629,
+      "step": 214
+    },
+    {
+      "epoch": 0.021995447453899076,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.942211055276383e-05,
+      "loss": 1.5722,
+      "step": 215
+    },
+    {
+      "epoch": 0.0220977518606614,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.941708542713568e-05,
+      "loss": 1.5695,
+      "step": 216
+    },
+    {
+      "epoch": 0.022200056267423718,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.941206030150754e-05,
+      "loss": 1.4521,
+      "step": 217
+    },
+    {
+      "epoch": 0.02230236067418604,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.94070351758794e-05,
+      "loss": 1.5651,
+      "step": 218
+    },
+    {
+      "epoch": 0.022404665080948363,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.940201005025126e-05,
+      "loss": 1.3817,
+      "step": 219
+    },
+    {
+      "epoch": 0.022506969487710683,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.939698492462311e-05,
+      "loss": 1.421,
+      "step": 220
+    },
+    {
+      "epoch": 0.022609273894473005,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.939195979899499e-05,
+      "loss": 1.5311,
+      "step": 221
+    },
+    {
+      "epoch": 0.022711578301235325,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.938693467336683e-05,
+      "loss": 1.6331,
+      "step": 222
+    },
+    {
+      "epoch": 0.022813882707997647,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.93819095477387e-05,
+      "loss": 1.5601,
+      "step": 223
+    },
+    {
+      "epoch": 0.02291618711475997,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.937688442211056e-05,
+      "loss": 1.4092,
+      "step": 224
+    },
+    {
+      "epoch": 0.02301849152152229,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.937185929648242e-05,
+      "loss": 1.5075,
+      "step": 225
+    },
+    {
+      "epoch": 0.023120795928284612,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.936683417085427e-05,
+      "loss": 1.3996,
+      "step": 226
+    },
+    {
+      "epoch": 0.02322310033504693,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.936180904522614e-05,
+      "loss": 1.315,
+      "step": 227
+    },
+    {
+      "epoch": 0.023325404741809254,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.935678391959799e-05,
+      "loss": 1.4644,
+      "step": 228
+    },
+    {
+      "epoch": 0.023427709148571573,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.935175879396985e-05,
+      "loss": 1.5531,
+      "step": 229
+    },
+    {
+      "epoch": 0.023530013555333896,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.934673366834172e-05,
+      "loss": 1.5297,
+      "step": 230
+    },
+    {
+      "epoch": 0.02363231796209622,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.934170854271358e-05,
+      "loss": 1.6029,
+      "step": 231
+    },
+    {
+      "epoch": 0.023734622368858538,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.933668341708543e-05,
+      "loss": 1.389,
+      "step": 232
+    },
+    {
+      "epoch": 0.02383692677562086,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.933165829145729e-05,
+      "loss": 1.5131,
+      "step": 233
+    },
+    {
+      "epoch": 0.02393923118238318,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.932663316582915e-05,
+      "loss": 1.5949,
+      "step": 234
+    },
+    {
+      "epoch": 0.024041535589145503,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.932160804020101e-05,
+      "loss": 1.4901,
+      "step": 235
+    },
+    {
+      "epoch": 0.024143839995907825,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.931658291457286e-05,
+      "loss": 1.52,
+      "step": 236
+    },
+    {
+      "epoch": 0.024246144402670144,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.931155778894474e-05,
+      "loss": 1.5003,
+      "step": 237
+    },
+    {
+      "epoch": 0.024348448809432467,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.930653266331659e-05,
+      "loss": 1.6123,
+      "step": 238
+    },
+    {
+      "epoch": 0.024450753216194786,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.930150753768845e-05,
+      "loss": 1.4051,
+      "step": 239
+    },
+    {
+      "epoch": 0.02455305762295711,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.929648241206031e-05,
+      "loss": 1.4039,
+      "step": 240
+    },
+    {
+      "epoch": 0.024655362029719432,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.929145728643217e-05,
+      "loss": 1.5335,
+      "step": 241
+    },
+    {
+      "epoch": 0.02475766643648175,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.928643216080402e-05,
+      "loss": 1.589,
+      "step": 242
+    },
+    {
+      "epoch": 0.024859970843244074,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.928140703517588e-05,
+      "loss": 1.4552,
+      "step": 243
+    },
+    {
+      "epoch": 0.024962275250006393,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.927638190954774e-05,
+      "loss": 1.5384,
+      "step": 244
+    },
+    {
+      "epoch": 0.025064579656768716,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.927135678391961e-05,
+      "loss": 1.4569,
+      "step": 245
+    },
+    {
+      "epoch": 0.025166884063531035,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.926633165829147e-05,
+      "loss": 1.3436,
+      "step": 246
+    },
+    {
+      "epoch": 0.025269188470293358,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.926130653266332e-05,
+      "loss": 1.4222,
+      "step": 247
+    },
+    {
+      "epoch": 0.02537149287705568,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.925628140703518e-05,
+      "loss": 1.6223,
+      "step": 248
+    },
+    {
+      "epoch": 0.025473797283818,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.925125628140703e-05,
+      "loss": 1.489,
+      "step": 249
+    },
+    {
+      "epoch": 0.025576101690580322,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.92462311557789e-05,
+      "loss": 1.3825,
+      "step": 250
+    },
+    {
+      "epoch": 0.02567840609734264,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.924120603015075e-05,
+      "loss": 1.6174,
+      "step": 251
+    },
+    {
+      "epoch": 0.025780710504104964,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.923618090452261e-05,
+      "loss": 1.6671,
+      "step": 252
+    },
+    {
+      "epoch": 0.025883014910867287,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.923115577889448e-05,
+      "loss": 1.5772,
+      "step": 253
+    },
+    {
+      "epoch": 0.025985319317629606,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.922613065326634e-05,
+      "loss": 1.3375,
+      "step": 254
+    },
+    {
+      "epoch": 0.02608762372439193,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.922110552763819e-05,
+      "loss": 1.6211,
+      "step": 255
+    },
+    {
+      "epoch": 0.02618992813115425,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.921608040201006e-05,
+      "loss": 1.4128,
+      "step": 256
+    },
+    {
+      "epoch": 0.02629223253791657,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.921105527638191e-05,
+      "loss": 1.6238,
+      "step": 257
+    },
+    {
+      "epoch": 0.02639453694467889,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.920603015075377e-05,
+      "loss": 1.4589,
+      "step": 258
+    },
+    {
+      "epoch": 0.026496841351441213,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.920100502512563e-05,
+      "loss": 1.5027,
+      "step": 259
+    },
+    {
+      "epoch": 0.026599145758203536,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.91959798994975e-05,
+      "loss": 1.5178,
+      "step": 260
+    },
+    {
+      "epoch": 0.026701450164965855,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.919095477386935e-05,
+      "loss": 1.6778,
+      "step": 261
+    },
+    {
+      "epoch": 0.026803754571728178,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.918592964824122e-05,
+      "loss": 1.4559,
+      "step": 262
+    },
+    {
+      "epoch": 0.026906058978490497,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.918090452261307e-05,
+      "loss": 1.451,
+      "step": 263
+    },
+    {
+      "epoch": 0.02700836338525282,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.917587939698493e-05,
+      "loss": 1.5377,
+      "step": 264
+    },
+    {
+      "epoch": 0.027110667792015142,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.917085427135678e-05,
+      "loss": 1.4026,
+      "step": 265
+    },
+    {
+      "epoch": 0.02721297219877746,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.916582914572866e-05,
+      "loss": 1.663,
+      "step": 266
+    },
+    {
+      "epoch": 0.027315276605539784,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.91608040201005e-05,
+      "loss": 1.452,
+      "step": 267
+    },
+    {
+      "epoch": 0.027417581012302104,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.915577889447237e-05,
+      "loss": 1.5737,
+      "step": 268
+    },
+    {
+      "epoch": 0.027519885419064426,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.915075376884423e-05,
+      "loss": 1.4146,
+      "step": 269
+    },
+    {
+      "epoch": 0.02762218982582675,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.914572864321609e-05,
+      "loss": 1.6943,
+      "step": 270
+    },
+    {
+      "epoch": 0.02772449423258907,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.914070351758794e-05,
+      "loss": 1.6901,
+      "step": 271
+    },
+    {
+      "epoch": 0.02782679863935139,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.913567839195981e-05,
+      "loss": 1.7098,
+      "step": 272
+    },
+    {
+      "epoch": 0.02792910304611371,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.913065326633166e-05,
+      "loss": 1.575,
+      "step": 273
+    },
+    {
+      "epoch": 0.028031407452876033,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.912562814070352e-05,
+      "loss": 1.5084,
+      "step": 274
+    },
+    {
+      "epoch": 0.028133711859638352,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.912060301507539e-05,
+      "loss": 1.4202,
+      "step": 275
+    },
+    {
+      "epoch": 0.028236016266400675,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.911557788944725e-05,
+      "loss": 1.3769,
+      "step": 276
+    },
+    {
+      "epoch": 0.028338320673162998,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.91105527638191e-05,
+      "loss": 1.5667,
+      "step": 277
+    },
+    {
+      "epoch": 0.028440625079925317,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.910552763819096e-05,
+      "loss": 1.4585,
+      "step": 278
+    },
+    {
+      "epoch": 0.02854292948668764,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.910050251256282e-05,
+      "loss": 1.5577,
+      "step": 279
+    },
+    {
+      "epoch": 0.02864523389344996,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.909547738693468e-05,
+      "loss": 1.6707,
+      "step": 280
+    },
+    {
+      "epoch": 0.02874753830021228,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.909045226130653e-05,
+      "loss": 1.6164,
+      "step": 281
+    },
+    {
+      "epoch": 0.028849842706974604,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.90854271356784e-05,
+      "loss": 1.5375,
+      "step": 282
+    },
+    {
+      "epoch": 0.028952147113736924,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.908040201005026e-05,
+      "loss": 1.5094,
+      "step": 283
+    },
+    {
+      "epoch": 0.029054451520499246,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.90753768844221e-05,
+      "loss": 1.5684,
+      "step": 284
+    },
+    {
+      "epoch": 0.029156755927261566,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.907035175879398e-05,
+      "loss": 1.4031,
+      "step": 285
+    },
+    {
+      "epoch": 0.02925906033402389,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.906532663316583e-05,
+      "loss": 1.401,
+      "step": 286
+    },
+    {
+      "epoch": 0.02936136474078621,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.906030150753769e-05,
+      "loss": 1.495,
+      "step": 287
+    },
+    {
+      "epoch": 0.02946366914754853,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.905527638190955e-05,
+      "loss": 1.4831,
+      "step": 288
+    },
+    {
+      "epoch": 0.029565973554310853,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.905025125628141e-05,
+      "loss": 1.6447,
+      "step": 289
+    },
+    {
+      "epoch": 0.029668277961073172,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.904522613065326e-05,
+      "loss": 1.6664,
+      "step": 290
+    },
+    {
+      "epoch": 0.029770582367835495,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.904020100502514e-05,
+      "loss": 1.4133,
+      "step": 291
+    },
+    {
+      "epoch": 0.029872886774597814,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.903517587939699e-05,
+      "loss": 1.528,
+      "step": 292
+    },
+    {
+      "epoch": 0.029975191181360137,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.903015075376885e-05,
+      "loss": 1.4882,
+      "step": 293
+    },
+    {
+      "epoch": 0.03007749558812246,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.902512562814071e-05,
+      "loss": 1.5297,
+      "step": 294
+    },
+    {
+      "epoch": 0.03017979999488478,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.902010050251257e-05,
+      "loss": 1.5328,
+      "step": 295
+    },
+    {
+      "epoch": 0.0302821044016471,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.901507537688442e-05,
+      "loss": 1.5717,
+      "step": 296
+    },
+    {
+      "epoch": 0.03038440880840942,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.901005025125628e-05,
+      "loss": 1.6148,
+      "step": 297
+    },
+    {
+      "epoch": 0.030486713215171744,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.900502512562815e-05,
+      "loss": 1.5215,
+      "step": 298
+    },
+    {
+      "epoch": 0.030589017621934066,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.4978,
+      "step": 299
+    },
+    {
+      "epoch": 0.030691322028696386,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.899497487437186e-05,
+      "loss": 1.4289,
+      "step": 300
+    },
+    {
+      "epoch": 0.03079362643545871,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.898994974874373e-05,
+      "loss": 1.6323,
+      "step": 301
+    },
+    {
+      "epoch": 0.030895930842221028,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.898492462311558e-05,
+      "loss": 1.3455,
+      "step": 302
+    },
+    {
+      "epoch": 0.03099823524898335,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.897989949748744e-05,
+      "loss": 1.3538,
+      "step": 303
+    },
+    {
+      "epoch": 0.03110053965574567,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.89748743718593e-05,
+      "loss": 1.6287,
+      "step": 304
+    },
+    {
+      "epoch": 0.031202844062507992,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.896984924623117e-05,
+      "loss": 1.5676,
+      "step": 305
+    },
+    {
+      "epoch": 0.031305148469270315,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.896482412060301e-05,
+      "loss": 1.4222,
+      "step": 306
+    },
+    {
+      "epoch": 0.03140745287603264,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.895979899497489e-05,
+      "loss": 1.6212,
+      "step": 307
+    },
+    {
+      "epoch": 0.031509757282794953,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.895477386934674e-05,
+      "loss": 1.5027,
+      "step": 308
+    },
+    {
+      "epoch": 0.031612061689557276,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.89497487437186e-05,
+      "loss": 1.4725,
+      "step": 309
+    },
+    {
+      "epoch": 0.0317143660963196,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.894472361809046e-05,
+      "loss": 1.4821,
+      "step": 310
+    },
+    {
+      "epoch": 0.03181667050308192,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.893969849246232e-05,
+      "loss": 1.4358,
+      "step": 311
+    },
+    {
+      "epoch": 0.031918974909844244,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.893467336683417e-05,
+      "loss": 1.3851,
+      "step": 312
+    },
+    {
+      "epoch": 0.03202127931660656,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.892964824120604e-05,
+      "loss": 1.6571,
+      "step": 313
+    },
+    {
+      "epoch": 0.03212358372336888,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.89246231155779e-05,
+      "loss": 1.4201,
+      "step": 314
+    },
+    {
+      "epoch": 0.032225888130131206,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.891959798994975e-05,
+      "loss": 1.5493,
+      "step": 315
+    },
+    {
+      "epoch": 0.03232819253689353,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.891457286432161e-05,
+      "loss": 1.5349,
+      "step": 316
+    },
+    {
+      "epoch": 0.03243049694365585,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.890954773869347e-05,
+      "loss": 1.429,
+      "step": 317
+    },
+    {
+      "epoch": 0.03253280135041817,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.890452261306533e-05,
+      "loss": 1.5291,
+      "step": 318
+    },
+    {
+      "epoch": 0.03263510575718049,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.889949748743718e-05,
+      "loss": 1.3211,
+      "step": 319
+    },
+    {
+      "epoch": 0.03273741016394281,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.889447236180906e-05,
+      "loss": 1.39,
+      "step": 320
+    },
+    {
+      "epoch": 0.032839714570705135,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.88894472361809e-05,
+      "loss": 1.6569,
+      "step": 321
+    },
+    {
+      "epoch": 0.03294201897746746,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.888442211055277e-05,
+      "loss": 1.3799,
+      "step": 322
+    },
+    {
+      "epoch": 0.03304432338422977,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.887939698492463e-05,
+      "loss": 1.5544,
+      "step": 323
+    },
+    {
+      "epoch": 0.033146627790992096,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.887437185929649e-05,
+      "loss": 1.3852,
+      "step": 324
+    },
+    {
+      "epoch": 0.03324893219775442,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.886934673366834e-05,
+      "loss": 1.5531,
+      "step": 325
+    },
+    {
+      "epoch": 0.03335123660451674,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.886432160804021e-05,
+      "loss": 1.5463,
+      "step": 326
+    },
+    {
+      "epoch": 0.03345354101127906,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.885929648241206e-05,
+      "loss": 1.4092,
+      "step": 327
+    },
+    {
+      "epoch": 0.03355584541804138,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.885427135678393e-05,
+      "loss": 1.419,
+      "step": 328
+    },
+    {
+      "epoch": 0.0336581498248037,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.884924623115577e-05,
+      "loss": 1.5751,
+      "step": 329
+    },
+    {
+      "epoch": 0.033760454231566026,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.884422110552765e-05,
+      "loss": 1.4857,
+      "step": 330
+    },
+    {
+      "epoch": 0.03386275863832835,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.88391959798995e-05,
+      "loss": 1.4095,
+      "step": 331
+    },
+    {
+      "epoch": 0.033965063045090664,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.883417085427136e-05,
+      "loss": 1.4,
+      "step": 332
+    },
+    {
+      "epoch": 0.03406736745185299,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.882914572864322e-05,
+      "loss": 1.5383,
+      "step": 333
+    },
+    {
+      "epoch": 0.03416967185861531,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.882412060301508e-05,
+      "loss": 1.5472,
+      "step": 334
+    },
+    {
+      "epoch": 0.03427197626537763,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.881909547738693e-05,
+      "loss": 2.0221,
+      "step": 335
+    },
+    {
+      "epoch": 0.034374280672139955,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.881407035175881e-05,
+      "loss": 1.5059,
+      "step": 336
+    },
+    {
+      "epoch": 0.03447658507890227,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.880904522613066e-05,
+      "loss": 1.5476,
+      "step": 337
+    },
+    {
+      "epoch": 0.03457888948566459,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.880402010050252e-05,
+      "loss": 1.4791,
+      "step": 338
+    },
+    {
+      "epoch": 0.034681193892426916,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.879899497487438e-05,
+      "loss": 1.4705,
+      "step": 339
+    },
+    {
+      "epoch": 0.03478349829918924,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.879396984924624e-05,
+      "loss": 1.72,
+      "step": 340
+    },
+    {
+      "epoch": 0.03488580270595156,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.878894472361809e-05,
+      "loss": 1.3656,
+      "step": 341
+    },
+    {
+      "epoch": 0.03498810711271388,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.878391959798995e-05,
+      "loss": 1.5254,
+      "step": 342
+    },
+    {
+      "epoch": 0.0350904115194762,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.877889447236182e-05,
+      "loss": 1.6494,
+      "step": 343
+    },
+    {
+      "epoch": 0.03519271592623852,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.877386934673368e-05,
+      "loss": 1.4853,
+      "step": 344
+    },
+    {
+      "epoch": 0.035295020333000846,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.876884422110553e-05,
+      "loss": 1.4587,
+      "step": 345
+    },
+    {
+      "epoch": 0.03539732473976317,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.87638190954774e-05,
+      "loss": 1.4623,
+      "step": 346
+    },
+    {
+      "epoch": 0.035499629146525484,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.875879396984925e-05,
+      "loss": 1.5349,
+      "step": 347
+    },
+    {
+      "epoch": 0.03560193355328781,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.875376884422111e-05,
+      "loss": 1.4416,
+      "step": 348
+    },
+    {
+      "epoch": 0.03570423796005013,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.874874371859297e-05,
+      "loss": 1.5453,
+      "step": 349
+    },
+    {
+      "epoch": 0.03580654236681245,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.874371859296482e-05,
+      "loss": 1.5111,
+      "step": 350
+    },
+    {
+      "epoch": 0.035908846773574775,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.873869346733668e-05,
+      "loss": 1.4418,
+      "step": 351
+    },
+    {
+      "epoch": 0.03601115118033709,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.873366834170855e-05,
+      "loss": 1.6188,
+      "step": 352
+    },
+    {
+      "epoch": 0.03611345558709941,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.872864321608041e-05,
+      "loss": 1.4287,
+      "step": 353
+    },
+    {
+      "epoch": 0.036215759993861736,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.872361809045226e-05,
+      "loss": 1.6406,
+      "step": 354
+    },
+    {
+      "epoch": 0.03631806440062406,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.871859296482413e-05,
+      "loss": 1.3695,
+      "step": 355
+    },
+    {
+      "epoch": 0.03642036880738638,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.871356783919598e-05,
+      "loss": 1.4109,
+      "step": 356
+    },
+    {
+      "epoch": 0.0365226732141487,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.870854271356784e-05,
+      "loss": 1.4585,
+      "step": 357
+    },
+    {
+      "epoch": 0.03662497762091102,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.870351758793969e-05,
+      "loss": 1.7788,
+      "step": 358
+    },
+    {
+      "epoch": 0.03672728202767334,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.869849246231157e-05,
+      "loss": 1.5165,
+      "step": 359
+    },
+    {
+      "epoch": 0.036829586434435665,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.869346733668342e-05,
+      "loss": 1.3988,
+      "step": 360
+    },
+    {
+      "epoch": 0.03693189084119798,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.868844221105528e-05,
+      "loss": 1.5767,
+      "step": 361
+    },
+    {
+      "epoch": 0.037034195247960304,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.868341708542714e-05,
+      "loss": 1.4882,
+      "step": 362
+    },
+    {
+      "epoch": 0.03713649965472263,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.8678391959799e-05,
+      "loss": 1.5221,
+      "step": 363
+    },
+    {
+      "epoch": 0.03723880406148495,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.867336683417085e-05,
+      "loss": 1.4674,
+      "step": 364
+    },
+    {
+      "epoch": 0.03734110846824727,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.866834170854273e-05,
+      "loss": 1.5144,
+      "step": 365
+    },
+    {
+      "epoch": 0.03744341287500959,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.866331658291457e-05,
+      "loss": 1.4161,
+      "step": 366
+    },
+    {
+      "epoch": 0.03754571728177191,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.865829145728644e-05,
+      "loss": 1.4226,
+      "step": 367
+    },
+    {
+      "epoch": 0.03764802168853423,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.86532663316583e-05,
+      "loss": 1.5478,
+      "step": 368
+    },
+    {
+      "epoch": 0.037750326095296556,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.864824120603016e-05,
+      "loss": 1.539,
+      "step": 369
+    },
+    {
+      "epoch": 0.03785263050205888,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.864321608040201e-05,
+      "loss": 1.489,
+      "step": 370
+    },
+    {
+      "epoch": 0.037954934908821195,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.863819095477388e-05,
+      "loss": 1.4653,
+      "step": 371
+    },
+    {
+      "epoch": 0.03805723931558352,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.863316582914573e-05,
+      "loss": 1.4207,
+      "step": 372
+    },
+    {
+      "epoch": 0.03815954372234584,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.86281407035176e-05,
+      "loss": 1.5421,
+      "step": 373
+    },
+    {
+      "epoch": 0.03826184812910816,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.862311557788944e-05,
+      "loss": 1.4693,
+      "step": 374
+    },
+    {
+      "epoch": 0.038364152535870485,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.861809045226132e-05,
+      "loss": 1.4315,
+      "step": 375
+    },
+    {
+      "epoch": 0.0384664569426328,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.861306532663317e-05,
+      "loss": 1.519,
+      "step": 376
+    },
+    {
+      "epoch": 0.038568761349395124,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.860804020100503e-05,
+      "loss": 1.3558,
+      "step": 377
+    },
+    {
+      "epoch": 0.03867106575615745,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.860301507537689e-05,
+      "loss": 1.3301,
+      "step": 378
+    },
+    {
+      "epoch": 0.03877337016291977,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.859798994974875e-05,
+      "loss": 1.5604,
+      "step": 379
+    },
+    {
+      "epoch": 0.03887567456968209,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.85929648241206e-05,
+      "loss": 1.4254,
+      "step": 380
+    },
+    {
+      "epoch": 0.03897797897644441,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.858793969849246e-05,
+      "loss": 1.5913,
+      "step": 381
+    },
+    {
+      "epoch": 0.03908028338320673,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.858291457286433e-05,
+      "loss": 1.4161,
+      "step": 382
+    },
+    {
+      "epoch": 0.03918258778996905,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.857788944723619e-05,
+      "loss": 1.5643,
+      "step": 383
+    },
+    {
+      "epoch": 0.039284892196731376,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.857286432160805e-05,
+      "loss": 1.4412,
+      "step": 384
+    },
+    {
+      "epoch": 0.0393871966034937,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.85678391959799e-05,
+      "loss": 1.3842,
+      "step": 385
+    },
+    {
+      "epoch": 0.039489501010256015,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.856281407035176e-05,
+      "loss": 1.4981,
+      "step": 386
+    },
+    {
+      "epoch": 0.03959180541701834,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.855778894472362e-05,
+      "loss": 1.4213,
+      "step": 387
+    },
+    {
+      "epoch": 0.03969410982378066,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.855276381909548e-05,
+      "loss": 1.4924,
+      "step": 388
+    },
+    {
+      "epoch": 0.03979641423054298,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.854773869346733e-05,
+      "loss": 1.424,
+      "step": 389
+    },
+    {
+      "epoch": 0.0398987186373053,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.85427135678392e-05,
+      "loss": 1.7141,
+      "step": 390
+    },
+    {
+      "epoch": 0.04000102304406762,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.853768844221106e-05,
+      "loss": 1.6048,
+      "step": 391
+    },
+    {
+      "epoch": 0.040103327450829944,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.853266331658292e-05,
+      "loss": 1.2978,
+      "step": 392
+    },
+    {
+      "epoch": 0.04020563185759227,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.852763819095477e-05,
+      "loss": 1.4341,
+      "step": 393
+    },
+    {
+      "epoch": 0.04030793626435459,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.852261306532664e-05,
+      "loss": 1.5265,
+      "step": 394
+    },
+    {
+      "epoch": 0.040410240671116905,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.851758793969849e-05,
+      "loss": 1.5615,
+      "step": 395
+    },
+    {
+      "epoch": 0.04051254507787923,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.851256281407035e-05,
+      "loss": 1.5668,
+      "step": 396
+    },
+    {
+      "epoch": 0.04061484948464155,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.850753768844222e-05,
+      "loss": 1.4082,
+      "step": 397
+    },
+    {
+      "epoch": 0.04071715389140387,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.850251256281408e-05,
+      "loss": 1.4504,
+      "step": 398
+    },
+    {
+      "epoch": 0.040819458298166196,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.849748743718593e-05,
+      "loss": 1.5333,
+      "step": 399
+    },
+    {
+      "epoch": 0.04092176270492851,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.84924623115578e-05,
+      "loss": 1.4503,
+      "step": 400
+    },
+    {
+      "epoch": 0.041024067111690835,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.848743718592965e-05,
+      "loss": 1.4399,
+      "step": 401
+    },
+    {
+      "epoch": 0.04112637151845316,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.848241206030151e-05,
+      "loss": 1.5715,
+      "step": 402
+    },
+    {
+      "epoch": 0.04122867592521548,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.847738693467337e-05,
+      "loss": 1.5476,
+      "step": 403
+    },
+    {
+      "epoch": 0.0413309803319778,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.847236180904524e-05,
+      "loss": 1.2712,
+      "step": 404
+    },
+    {
+      "epoch": 0.04143328473874012,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.846733668341709e-05,
+      "loss": 1.5434,
+      "step": 405
+    },
+    {
+      "epoch": 0.04153558914550244,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.846231155778895e-05,
+      "loss": 1.4628,
+      "step": 406
+    },
+    {
+      "epoch": 0.041637893552264764,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.845728643216081e-05,
+      "loss": 1.5675,
+      "step": 407
+    },
+    {
+      "epoch": 0.04174019795902709,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.845226130653267e-05,
+      "loss": 1.353,
+      "step": 408
+    },
+    {
+      "epoch": 0.04184250236578941,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.844723618090452e-05,
+      "loss": 1.6526,
+      "step": 409
+    },
+    {
+      "epoch": 0.041944806772551725,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.84422110552764e-05,
+      "loss": 1.4674,
+      "step": 410
+    },
+    {
+      "epoch": 0.04204711117931405,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.843718592964824e-05,
+      "loss": 1.4208,
+      "step": 411
+    },
+    {
+      "epoch": 0.04214941558607637,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.84321608040201e-05,
+      "loss": 1.5195,
+      "step": 412
+    },
+    {
+      "epoch": 0.04225171999283869,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.842713567839197e-05,
+      "loss": 1.4643,
+      "step": 413
+    },
+    {
+      "epoch": 0.042354024399601016,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.842211055276383e-05,
+      "loss": 1.3745,
+      "step": 414
+    },
+    {
+      "epoch": 0.04245632880636333,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.841708542713568e-05,
+      "loss": 1.4112,
+      "step": 415
+    },
+    {
+      "epoch": 0.042558633213125655,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.841206030150754e-05,
+      "loss": 1.2664,
+      "step": 416
+    },
+    {
+      "epoch": 0.04266093761988798,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.84070351758794e-05,
+      "loss": 1.5595,
+      "step": 417
+    },
+    {
+      "epoch": 0.0427632420266503,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.840201005025126e-05,
+      "loss": 1.4136,
+      "step": 418
+    },
+    {
+      "epoch": 0.04286554643341262,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.839698492462313e-05,
+      "loss": 1.507,
+      "step": 419
+    },
+    {
+      "epoch": 0.04296785084017494,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.839195979899497e-05,
+      "loss": 1.4721,
+      "step": 420
+    },
+    {
+      "epoch": 0.04307015524693726,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.838693467336684e-05,
+      "loss": 1.4224,
+      "step": 421
+    },
+    {
+      "epoch": 0.043172459653699584,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.83819095477387e-05,
+      "loss": 1.4816,
+      "step": 422
+    },
+    {
+      "epoch": 0.04327476406046191,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.837688442211056e-05,
+      "loss": 1.5338,
+      "step": 423
+    },
+    {
+      "epoch": 0.04337706846722422,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.837185929648241e-05,
+      "loss": 1.4247,
+      "step": 424
+    },
+    {
+      "epoch": 0.043479372873986545,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.836683417085427e-05,
+      "loss": 1.521,
+      "step": 425
+    },
+    {
+      "epoch": 0.04358167728074887,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.836180904522613e-05,
+      "loss": 1.6056,
+      "step": 426
+    },
+    {
+      "epoch": 0.04368398168751119,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.8356783919598e-05,
+      "loss": 1.6455,
+      "step": 427
+    },
+    {
+      "epoch": 0.04378628609427351,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.835175879396984e-05,
+      "loss": 1.3836,
+      "step": 428
+    },
+    {
+      "epoch": 0.04388859050103583,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.834673366834172e-05,
+      "loss": 1.4052,
+      "step": 429
+    },
+    {
+      "epoch": 0.04399089490779815,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.834170854271357e-05,
+      "loss": 1.4185,
+      "step": 430
+    },
+    {
+      "epoch": 0.044093199314560474,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.833668341708543e-05,
+      "loss": 1.5088,
+      "step": 431
+    },
+    {
+      "epoch": 0.0441955037213228,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.833165829145729e-05,
+      "loss": 1.5585,
+      "step": 432
+    },
+    {
+      "epoch": 0.04429780812808512,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.832663316582915e-05,
+      "loss": 1.646,
+      "step": 433
+    },
+    {
+      "epoch": 0.044400112534847436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.8321608040201e-05,
+      "loss": 1.3428,
+      "step": 434
+    },
+    {
+      "epoch": 0.04450241694160976,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.831658291457288e-05,
+      "loss": 1.3819,
+      "step": 435
+    },
+    {
+      "epoch": 0.04460472134837208,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.831155778894473e-05,
+      "loss": 1.4197,
+      "step": 436
+    },
+    {
+      "epoch": 0.044707025755134404,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.830653266331659e-05,
+      "loss": 1.4816,
+      "step": 437
+    },
+    {
+      "epoch": 0.04480933016189673,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.830150753768844e-05,
+      "loss": 1.5839,
+      "step": 438
+    },
+    {
+      "epoch": 0.04491163456865904,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.829648241206031e-05,
+      "loss": 1.4556,
+      "step": 439
+    },
+    {
+      "epoch": 0.045013938975421365,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.829145728643216e-05,
+      "loss": 1.4828,
+      "step": 440
+    },
+    {
+      "epoch": 0.04511624338218369,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.828643216080402e-05,
+      "loss": 1.7692,
+      "step": 441
+    },
+    {
+      "epoch": 0.04521854778894601,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.828140703517589e-05,
+      "loss": 1.3626,
+      "step": 442
+    },
+    {
+      "epoch": 0.04532085219570833,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.827638190954775e-05,
+      "loss": 1.3112,
+      "step": 443
+    },
+    {
+      "epoch": 0.04542315660247065,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.82713567839196e-05,
+      "loss": 1.3597,
+      "step": 444
+    },
+    {
+      "epoch": 0.04552546100923297,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.826633165829147e-05,
+      "loss": 1.5107,
+      "step": 445
+    },
+    {
+      "epoch": 0.045627765415995294,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.826130653266332e-05,
+      "loss": 1.513,
+      "step": 446
+    },
+    {
+      "epoch": 0.04573006982275762,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.825628140703518e-05,
+      "loss": 1.3878,
+      "step": 447
+    },
+    {
+      "epoch": 0.04583237422951994,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.825125628140704e-05,
+      "loss": 1.5252,
+      "step": 448
+    },
+    {
+      "epoch": 0.045934678636282256,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.82462311557789e-05,
+      "loss": 1.3745,
+      "step": 449
+    },
+    {
+      "epoch": 0.04603698304304458,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.824120603015075e-05,
+      "loss": 1.4215,
+      "step": 450
+    },
+    {
+      "epoch": 0.0461392874498069,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.823618090452262e-05,
+      "loss": 1.3404,
+      "step": 451
+    },
+    {
+      "epoch": 0.046241591856569224,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.823115577889448e-05,
+      "loss": 1.5122,
+      "step": 452
+    },
+    {
+      "epoch": 0.04634389626333154,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.822613065326634e-05,
+      "loss": 1.5934,
+      "step": 453
+    },
+    {
+      "epoch": 0.04644620067009386,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.822110552763819e-05,
+      "loss": 1.4601,
+      "step": 454
+    },
+    {
+      "epoch": 0.046548505076856185,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.821608040201005e-05,
+      "loss": 1.5104,
+      "step": 455
+    },
+    {
+      "epoch": 0.04665080948361851,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.821105527638191e-05,
+      "loss": 1.6105,
+      "step": 456
+    },
+    {
+      "epoch": 0.04675311389038083,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.820603015075378e-05,
+      "loss": 1.5262,
+      "step": 457
+    },
+    {
+      "epoch": 0.046855418297143146,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.820100502512564e-05,
+      "loss": 1.3612,
+      "step": 458
+    },
+    {
+      "epoch": 0.04695772270390547,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.819597989949749e-05,
+      "loss": 1.6181,
+      "step": 459
+    },
+    {
+      "epoch": 0.04706002711066779,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.819095477386935e-05,
+      "loss": 1.3604,
+      "step": 460
+    },
+    {
+      "epoch": 0.047162331517430114,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.818592964824121e-05,
+      "loss": 1.5613,
+      "step": 461
+    },
+    {
+      "epoch": 0.04726463592419244,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.818090452261307e-05,
+      "loss": 1.5407,
+      "step": 462
+    },
+    {
+      "epoch": 0.04736694033095475,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.817587939698492e-05,
+      "loss": 1.5801,
+      "step": 463
+    },
+    {
+      "epoch": 0.047469244737717076,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.81708542713568e-05,
+      "loss": 1.2577,
+      "step": 464
+    },
+    {
+      "epoch": 0.0475715491444794,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.816582914572864e-05,
+      "loss": 1.4299,
+      "step": 465
+    },
+    {
+      "epoch": 0.04767385355124172,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.81608040201005e-05,
+      "loss": 1.4913,
+      "step": 466
+    },
+    {
+      "epoch": 0.047776157958004044,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.815577889447236e-05,
+      "loss": 1.3479,
+      "step": 467
+    },
+    {
+      "epoch": 0.04787846236476636,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.815075376884423e-05,
+      "loss": 1.5501,
+      "step": 468
+    },
+    {
+      "epoch": 0.04798076677152868,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.814572864321608e-05,
+      "loss": 1.5409,
+      "step": 469
+    },
+    {
+      "epoch": 0.048083071178291005,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.814070351758794e-05,
+      "loss": 1.3988,
+      "step": 470
+    },
+    {
+      "epoch": 0.04818537558505333,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.81356783919598e-05,
+      "loss": 2.4224,
+      "step": 471
+    },
+    {
+      "epoch": 0.04828767999181565,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.813065326633167e-05,
+      "loss": 1.3418,
+      "step": 472
+    },
+    {
+      "epoch": 0.048389984398577966,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.812562814070351e-05,
+      "loss": 1.3554,
+      "step": 473
+    },
+    {
+      "epoch": 0.04849228880534029,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.812060301507539e-05,
+      "loss": 1.5012,
+      "step": 474
+    },
+    {
+      "epoch": 0.04859459321210261,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.811557788944724e-05,
+      "loss": 1.693,
+      "step": 475
+    },
+    {
+      "epoch": 0.048696897618864934,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.81105527638191e-05,
+      "loss": 1.6285,
+      "step": 476
+    },
+    {
+      "epoch": 0.04879920202562726,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.810552763819096e-05,
+      "loss": 1.434,
+      "step": 477
+    },
+    {
+      "epoch": 0.04890150643238957,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.810050251256282e-05,
+      "loss": 1.5222,
+      "step": 478
+    },
+    {
+      "epoch": 0.049003810839151896,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.809547738693467e-05,
+      "loss": 1.4239,
+      "step": 479
+    },
+    {
+      "epoch": 0.04910611524591422,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.809045226130655e-05,
+      "loss": 1.4467,
+      "step": 480
+    },
+    {
+      "epoch": 0.04920841965267654,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.80854271356784e-05,
+      "loss": 1.5483,
+      "step": 481
+    },
+    {
+      "epoch": 0.049310724059438864,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.808040201005026e-05,
+      "loss": 1.6205,
+      "step": 482
+    },
+    {
+      "epoch": 0.04941302846620118,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.807537688442211e-05,
+      "loss": 1.4619,
+      "step": 483
+    },
+    {
+      "epoch": 0.0495153328729635,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.807035175879398e-05,
+      "loss": 1.5014,
+      "step": 484
+    },
+    {
+      "epoch": 0.049617637279725825,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.806532663316583e-05,
+      "loss": 1.5649,
+      "step": 485
+    },
+    {
+      "epoch": 0.04971994168648815,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.806030150753769e-05,
+      "loss": 1.4837,
+      "step": 486
+    },
+    {
+      "epoch": 0.049822246093250463,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.805527638190956e-05,
+      "loss": 1.4792,
+      "step": 487
+    },
+    {
+      "epoch": 0.049924550500012786,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.805025125628142e-05,
+      "loss": 1.5272,
+      "step": 488
+    },
+    {
+      "epoch": 0.05002685490677511,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.804522613065327e-05,
+      "loss": 1.6108,
+      "step": 489
+    },
+    {
+      "epoch": 0.05012915931353743,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.804020100502513e-05,
+      "loss": 1.4578,
+      "step": 490
+    },
+    {
+      "epoch": 0.050231463720299754,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.803517587939699e-05,
+      "loss": 1.4169,
+      "step": 491
+    },
+    {
+      "epoch": 0.05033376812706207,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.803015075376885e-05,
+      "loss": 1.6052,
+      "step": 492
+    },
+    {
+      "epoch": 0.05043607253382439,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.802512562814071e-05,
+      "loss": 1.5128,
+      "step": 493
+    },
+    {
+      "epoch": 0.050538376940586716,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.802010050251256e-05,
+      "loss": 1.5498,
+      "step": 494
+    },
+    {
+      "epoch": 0.05064068134734904,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.801507537688442e-05,
+      "loss": 1.5615,
+      "step": 495
+    },
+    {
+      "epoch": 0.05074298575411136,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.801005025125629e-05,
+      "loss": 1.4943,
+      "step": 496
+    },
+    {
+      "epoch": 0.05084529016087368,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.800502512562815e-05,
+      "loss": 1.5393,
+      "step": 497
+    },
+    {
+      "epoch": 0.050947594567636,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.8e-05,
+      "loss": 1.5214,
+      "step": 498
+    },
+    {
+      "epoch": 0.05104989897439832,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.799497487437186e-05,
+      "loss": 1.3655,
+      "step": 499
+    },
+    {
+      "epoch": 0.051152203381160645,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.798994974874372e-05,
+      "loss": 1.4507,
+      "step": 500
+    },
+    {
+      "epoch": 0.05125450778792297,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.798492462311558e-05,
+      "loss": 1.4332,
+      "step": 501
+    },
+    {
+      "epoch": 0.05135681219468528,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.797989949748743e-05,
+      "loss": 1.4173,
+      "step": 502
+    },
+    {
+      "epoch": 0.051459116601447606,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.797487437185931e-05,
+      "loss": 1.5444,
+      "step": 503
+    },
+    {
+      "epoch": 0.05156142100820993,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.796984924623116e-05,
+      "loss": 1.5688,
+      "step": 504
+    },
+    {
+      "epoch": 0.05166372541497225,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.796482412060302e-05,
+      "loss": 1.5753,
+      "step": 505
+    },
+    {
+      "epoch": 0.051766029821734574,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.795979899497488e-05,
+      "loss": 1.4905,
+      "step": 506
+    },
+    {
+      "epoch": 0.05186833422849689,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.795477386934674e-05,
+      "loss": 1.34,
+      "step": 507
+    },
+    {
+      "epoch": 0.05197063863525921,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.794974874371859e-05,
+      "loss": 1.2313,
+      "step": 508
+    },
+    {
+      "epoch": 0.052072943042021536,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.794472361809047e-05,
+      "loss": 1.4807,
+      "step": 509
+    },
+    {
+      "epoch": 0.05217524744878386,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.793969849246231e-05,
+      "loss": 1.4914,
+      "step": 510
+    },
+    {
+      "epoch": 0.05227755185554618,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.793467336683418e-05,
+      "loss": 1.4883,
+      "step": 511
+    },
+    {
+      "epoch": 0.0523798562623085,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.792964824120604e-05,
+      "loss": 1.4305,
+      "step": 512
+    },
+    {
+      "epoch": 0.05248216066907082,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.79246231155779e-05,
+      "loss": 1.4724,
+      "step": 513
+    },
+    {
+      "epoch": 0.05258446507583314,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.791959798994975e-05,
+      "loss": 1.3858,
+      "step": 514
+    },
+    {
+      "epoch": 0.052686769482595465,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.791457286432161e-05,
+      "loss": 1.4575,
+      "step": 515
+    },
+    {
+      "epoch": 0.05278907388935778,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.790954773869347e-05,
+      "loss": 1.4729,
+      "step": 516
+    },
+    {
+      "epoch": 0.0528913782961201,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.790452261306533e-05,
+      "loss": 1.426,
+      "step": 517
+    },
+    {
+      "epoch": 0.052993682702882426,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.789949748743718e-05,
+      "loss": 1.4024,
+      "step": 518
+    },
+    {
+      "epoch": 0.05309598710964475,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.789447236180906e-05,
+      "loss": 1.5242,
+      "step": 519
+    },
+    {
+      "epoch": 0.05319829151640707,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.788944723618091e-05,
+      "loss": 1.3635,
+      "step": 520
+    },
+    {
+      "epoch": 0.05330059592316939,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.788442211055277e-05,
+      "loss": 1.339,
+      "step": 521
+    },
+    {
+      "epoch": 0.05340290032993171,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.787939698492463e-05,
+      "loss": 1.5414,
+      "step": 522
+    },
+    {
+      "epoch": 0.05350520473669403,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.78743718592965e-05,
+      "loss": 1.5013,
+      "step": 523
+    },
+    {
+      "epoch": 0.053607509143456356,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.786934673366834e-05,
+      "loss": 1.3263,
+      "step": 524
+    },
+    {
+      "epoch": 0.05370981355021868,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.78643216080402e-05,
+      "loss": 1.4383,
+      "step": 525
+    },
+    {
+      "epoch": 0.053812117956980994,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.785929648241207e-05,
+      "loss": 1.2826,
+      "step": 526
+    },
+    {
+      "epoch": 0.05391442236374332,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.785427135678393e-05,
+      "loss": 1.4956,
+      "step": 527
+    },
+    {
+      "epoch": 0.05401672677050564,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.784924623115579e-05,
+      "loss": 1.459,
+      "step": 528
+    },
+    {
+      "epoch": 0.05411903117726796,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.784422110552764e-05,
+      "loss": 1.3538,
+      "step": 529
+    },
+    {
+      "epoch": 0.054221335584030285,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.78391959798995e-05,
+      "loss": 1.375,
+      "step": 530
+    },
+    {
+      "epoch": 0.0543236399907926,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.783417085427136e-05,
+      "loss": 1.2744,
+      "step": 531
+    },
+    {
+      "epoch": 0.05442594439755492,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.782914572864322e-05,
+      "loss": 1.3516,
+      "step": 532
+    },
+    {
+      "epoch": 0.054528248804317246,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.782412060301507e-05,
+      "loss": 1.5406,
+      "step": 533
+    },
+    {
+      "epoch": 0.05463055321107957,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.781909547738694e-05,
+      "loss": 1.1639,
+      "step": 534
+    },
+    {
+      "epoch": 0.05473285761784189,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.78140703517588e-05,
+      "loss": 1.4223,
+      "step": 535
+    },
+    {
+      "epoch": 0.05483516202460421,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.780904522613066e-05,
+      "loss": 1.4039,
+      "step": 536
+    },
+    {
+      "epoch": 0.05493746643136653,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.780402010050251e-05,
+      "loss": 1.4364,
+      "step": 537
+    },
+    {
+      "epoch": 0.05503977083812885,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.779899497487438e-05,
+      "loss": 1.6896,
+      "step": 538
+    },
+    {
+      "epoch": 0.055142075244891176,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.779396984924623e-05,
+      "loss": 1.4943,
+      "step": 539
+    },
+    {
+      "epoch": 0.0552443796516535,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.77889447236181e-05,
+      "loss": 1.6979,
+      "step": 540
+    },
+    {
+      "epoch": 0.055346684058415814,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.778391959798996e-05,
+      "loss": 1.6321,
+      "step": 541
+    },
+    {
+      "epoch": 0.05544898846517814,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.777889447236182e-05,
+      "loss": 1.4745,
+      "step": 542
+    },
+    {
+      "epoch": 0.05555129287194046,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.777386934673367e-05,
+      "loss": 1.3399,
+      "step": 543
+    },
+    {
+      "epoch": 0.05565359727870278,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.776884422110554e-05,
+      "loss": 1.4463,
+      "step": 544
+    },
+    {
+      "epoch": 0.0557559016854651,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.776381909547739e-05,
+      "loss": 1.4417,
+      "step": 545
+    },
+    {
+      "epoch": 0.05585820609222742,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.775879396984925e-05,
+      "loss": 1.3438,
+      "step": 546
+    },
+    {
+      "epoch": 0.05596051049898974,
+      "grad_norm": 0.25,
+      "learning_rate": 9.77537688442211e-05,
+      "loss": 1.3517,
+      "step": 547
+    },
+    {
+      "epoch": 0.056062814905752066,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.774874371859298e-05,
+      "loss": 1.5476,
+      "step": 548
+    },
+    {
+      "epoch": 0.05616511931251439,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.774371859296483e-05,
+      "loss": 1.5421,
+      "step": 549
+    },
+    {
+      "epoch": 0.056267423719276705,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.773869346733669e-05,
+      "loss": 1.4897,
+      "step": 550
+    },
+    {
+      "epoch": 0.05636972812603903,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.773366834170855e-05,
+      "loss": 1.5852,
+      "step": 551
+    },
+    {
+      "epoch": 0.05647203253280135,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.772864321608041e-05,
+      "loss": 1.354,
+      "step": 552
+    },
+    {
+      "epoch": 0.05657433693956367,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.772361809045226e-05,
+      "loss": 1.3066,
+      "step": 553
+    },
+    {
+      "epoch": 0.056676641346325995,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.771859296482414e-05,
+      "loss": 1.4975,
+      "step": 554
+    },
+    {
+      "epoch": 0.05677894575308831,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.771356783919598e-05,
+      "loss": 1.4107,
+      "step": 555
+    },
+    {
+      "epoch": 0.056881250159850634,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.770854271356785e-05,
+      "loss": 1.4553,
+      "step": 556
+    },
+    {
+      "epoch": 0.05698355456661296,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.770351758793971e-05,
+      "loss": 1.4405,
+      "step": 557
+    },
+    {
+      "epoch": 0.05708585897337528,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.769849246231157e-05,
+      "loss": 1.3935,
+      "step": 558
+    },
+    {
+      "epoch": 0.0571881633801376,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.769346733668342e-05,
+      "loss": 1.3355,
+      "step": 559
+    },
+    {
+      "epoch": 0.05729046778689992,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.768844221105528e-05,
+      "loss": 2.7132,
+      "step": 560
+    },
+    {
+      "epoch": 0.05739277219366224,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.768341708542714e-05,
+      "loss": 2.1546,
+      "step": 561
+    },
+    {
+      "epoch": 0.05749507660042456,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.7678391959799e-05,
+      "loss": 1.4667,
+      "step": 562
+    },
+    {
+      "epoch": 0.057597381007186886,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.767336683417085e-05,
+      "loss": 1.33,
+      "step": 563
+    },
+    {
+      "epoch": 0.05769968541394921,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.766834170854271e-05,
+      "loss": 1.36,
+      "step": 564
+    },
+    {
+      "epoch": 0.057801989820711525,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.766331658291458e-05,
+      "loss": 1.4025,
+      "step": 565
+    },
+    {
+      "epoch": 0.05790429422747385,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.765829145728644e-05,
+      "loss": 1.3763,
+      "step": 566
+    },
+    {
+      "epoch": 0.05800659863423617,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.76532663316583e-05,
+      "loss": 1.4489,
+      "step": 567
+    },
+    {
+      "epoch": 0.05810890304099849,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.764824120603015e-05,
+      "loss": 1.4309,
+      "step": 568
+    },
+    {
+      "epoch": 0.058211207447760815,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.764321608040201e-05,
+      "loss": 1.3253,
+      "step": 569
+    },
+    {
+      "epoch": 0.05831351185452313,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.763819095477387e-05,
+      "loss": 1.2969,
+      "step": 570
+    },
+    {
+      "epoch": 0.058415816261285454,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.763316582914574e-05,
+      "loss": 1.3688,
+      "step": 571
+    },
+    {
+      "epoch": 0.05851812066804778,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.762814070351758e-05,
+      "loss": 1.52,
+      "step": 572
+    },
+    {
+      "epoch": 0.0586204250748101,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.762311557788946e-05,
+      "loss": 1.4799,
+      "step": 573
+    },
+    {
+      "epoch": 0.05872272948157242,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.761809045226131e-05,
+      "loss": 1.7948,
+      "step": 574
+    },
+    {
+      "epoch": 0.05882503388833474,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.761306532663317e-05,
+      "loss": 1.3959,
+      "step": 575
+    },
+    {
+      "epoch": 0.05892733829509706,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.760804020100502e-05,
+      "loss": 1.4353,
+      "step": 576
+    },
+    {
+      "epoch": 0.05902964270185938,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.76030150753769e-05,
+      "loss": 1.4863,
+      "step": 577
+    },
+    {
+      "epoch": 0.059131947108621706,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.759798994974874e-05,
+      "loss": 1.6738,
+      "step": 578
+    },
+    {
+      "epoch": 0.05923425151538402,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.75929648241206e-05,
+      "loss": 1.4405,
+      "step": 579
+    },
+    {
+      "epoch": 0.059336555922146345,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.758793969849247e-05,
+      "loss": 1.5586,
+      "step": 580
+    },
+    {
+      "epoch": 0.05943886032890867,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.758291457286433e-05,
+      "loss": 1.3402,
+      "step": 581
+    },
+    {
+      "epoch": 0.05954116473567099,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.757788944723618e-05,
+      "loss": 1.5638,
+      "step": 582
+    },
+    {
+      "epoch": 0.05964346914243331,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.757286432160805e-05,
+      "loss": 1.4336,
+      "step": 583
+    },
+    {
+      "epoch": 0.05974577354919563,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.75678391959799e-05,
+      "loss": 1.5737,
+      "step": 584
+    },
+    {
+      "epoch": 0.05984807795595795,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.756281407035176e-05,
+      "loss": 1.6251,
+      "step": 585
+    },
+    {
+      "epoch": 0.059950382362720274,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.755778894472363e-05,
+      "loss": 1.6241,
+      "step": 586
+    },
+    {
+      "epoch": 0.0600526867694826,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.755276381909549e-05,
+      "loss": 1.219,
+      "step": 587
+    },
+    {
+      "epoch": 0.06015499117624492,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.754773869346734e-05,
+      "loss": 1.4274,
+      "step": 588
+    },
+    {
+      "epoch": 0.060257295583007235,
+      "grad_norm": 0.25,
+      "learning_rate": 9.754271356783921e-05,
+      "loss": 1.3274,
+      "step": 589
+    },
+    {
+      "epoch": 0.06035959998976956,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.753768844221106e-05,
+      "loss": 1.3772,
+      "step": 590
+    },
+    {
+      "epoch": 0.06046190439653188,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.753266331658292e-05,
+      "loss": 1.4983,
+      "step": 591
+    },
+    {
+      "epoch": 0.0605642088032942,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.752763819095477e-05,
+      "loss": 1.4325,
+      "step": 592
+    },
+    {
+      "epoch": 0.060666513210056526,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.752261306532665e-05,
+      "loss": 1.3373,
+      "step": 593
+    },
+    {
+      "epoch": 0.06076881761681884,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.75175879396985e-05,
+      "loss": 1.5423,
+      "step": 594
+    },
+    {
+      "epoch": 0.060871122023581165,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.751256281407036e-05,
+      "loss": 1.5407,
+      "step": 595
+    },
+    {
+      "epoch": 0.06097342643034349,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.750753768844222e-05,
+      "loss": 1.3965,
+      "step": 596
+    },
+    {
+      "epoch": 0.06107573083710581,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.750251256281408e-05,
+      "loss": 1.4446,
+      "step": 597
+    },
+    {
+      "epoch": 0.06117803524386813,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.749748743718593e-05,
+      "loss": 1.4405,
+      "step": 598
+    },
+    {
+      "epoch": 0.06128033965063045,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.749246231155779e-05,
+      "loss": 1.4991,
+      "step": 599
+    },
+    {
+      "epoch": 0.06138264405739277,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.748743718592965e-05,
+      "loss": 1.4312,
+      "step": 600
+    },
+    {
+      "epoch": 0.061484948464155094,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.748241206030152e-05,
+      "loss": 1.5209,
+      "step": 601
+    },
+    {
+      "epoch": 0.06158725287091742,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.747738693467338e-05,
+      "loss": 1.2227,
+      "step": 602
+    },
+    {
+      "epoch": 0.06168955727767974,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.747236180904523e-05,
+      "loss": 1.3212,
+      "step": 603
+    },
+    {
+      "epoch": 0.061791861684442055,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.746733668341709e-05,
+      "loss": 1.3274,
+      "step": 604
+    },
+    {
+      "epoch": 0.06189416609120438,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.746231155778895e-05,
+      "loss": 1.4954,
+      "step": 605
+    },
+    {
+      "epoch": 0.0619964704979667,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.745728643216081e-05,
+      "loss": 1.5664,
+      "step": 606
+    },
+    {
+      "epoch": 0.06209877490472902,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.745226130653266e-05,
+      "loss": 1.4641,
+      "step": 607
+    },
+    {
+      "epoch": 0.06220107931149134,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.744723618090452e-05,
+      "loss": 1.7401,
+      "step": 608
+    },
+    {
+      "epoch": 0.06230338371825366,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.744221105527638e-05,
+      "loss": 1.4018,
+      "step": 609
+    },
+    {
+      "epoch": 0.062405688125015985,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.743718592964825e-05,
+      "loss": 1.4122,
+      "step": 610
+    },
+    {
+      "epoch": 0.0625079925317783,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.74321608040201e-05,
+      "loss": 1.4021,
+      "step": 611
+    },
+    {
+      "epoch": 0.06261029693854063,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.742713567839197e-05,
+      "loss": 1.3637,
+      "step": 612
+    },
+    {
+      "epoch": 0.06271260134530295,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.742211055276382e-05,
+      "loss": 1.5613,
+      "step": 613
+    },
+    {
+      "epoch": 0.06281490575206528,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.741708542713568e-05,
+      "loss": 1.3794,
+      "step": 614
+    },
+    {
+      "epoch": 0.06291721015882759,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.741206030150754e-05,
+      "loss": 1.3821,
+      "step": 615
+    },
+    {
+      "epoch": 0.06301951456558991,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.74070351758794e-05,
+      "loss": 1.4783,
+      "step": 616
+    },
+    {
+      "epoch": 0.06312181897235224,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.740201005025125e-05,
+      "loss": 1.4275,
+      "step": 617
+    },
+    {
+      "epoch": 0.06322412337911455,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.739698492462313e-05,
+      "loss": 1.2913,
+      "step": 618
+    },
+    {
+      "epoch": 0.06332642778587688,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.739195979899498e-05,
+      "loss": 1.4445,
+      "step": 619
+    },
+    {
+      "epoch": 0.0634287321926392,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.738693467336684e-05,
+      "loss": 1.3856,
+      "step": 620
+    },
+    {
+      "epoch": 0.06353103659940151,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.73819095477387e-05,
+      "loss": 1.4996,
+      "step": 621
+    },
+    {
+      "epoch": 0.06363334100616384,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.737688442211056e-05,
+      "loss": 1.3639,
+      "step": 622
+    },
+    {
+      "epoch": 0.06373564541292616,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.737185929648241e-05,
+      "loss": 1.6286,
+      "step": 623
+    },
+    {
+      "epoch": 0.06383794981968849,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.736683417085427e-05,
+      "loss": 1.503,
+      "step": 624
+    },
+    {
+      "epoch": 0.0639402542264508,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.736180904522614e-05,
+      "loss": 1.4468,
+      "step": 625
+    },
+    {
+      "epoch": 0.06404255863321312,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.7356783919598e-05,
+      "loss": 1.3198,
+      "step": 626
+    },
+    {
+      "epoch": 0.06414486303997545,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.735175879396985e-05,
+      "loss": 1.444,
+      "step": 627
+    },
+    {
+      "epoch": 0.06424716744673777,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.734673366834172e-05,
+      "loss": 1.4339,
+      "step": 628
+    },
+    {
+      "epoch": 0.0643494718535001,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.734170854271357e-05,
+      "loss": 1.4981,
+      "step": 629
+    },
+    {
+      "epoch": 0.06445177626026241,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.733668341708543e-05,
+      "loss": 1.3081,
+      "step": 630
+    },
+    {
+      "epoch": 0.06455408066702473,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.73316582914573e-05,
+      "loss": 1.5892,
+      "step": 631
+    },
+    {
+      "epoch": 0.06465638507378706,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.732663316582916e-05,
+      "loss": 1.4917,
+      "step": 632
+    },
+    {
+      "epoch": 0.06475868948054937,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.7321608040201e-05,
+      "loss": 1.5483,
+      "step": 633
+    },
+    {
+      "epoch": 0.0648609938873117,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.731658291457287e-05,
+      "loss": 1.4257,
+      "step": 634
+    },
+    {
+      "epoch": 0.06496329829407402,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.731155778894473e-05,
+      "loss": 1.4084,
+      "step": 635
+    },
+    {
+      "epoch": 0.06506560270083633,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.730653266331659e-05,
+      "loss": 1.2921,
+      "step": 636
+    },
+    {
+      "epoch": 0.06516790710759866,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.730150753768845e-05,
+      "loss": 1.5779,
+      "step": 637
+    },
+    {
+      "epoch": 0.06527021151436098,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.72964824120603e-05,
+      "loss": 1.4631,
+      "step": 638
+    },
+    {
+      "epoch": 0.06537251592112331,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.729145728643216e-05,
+      "loss": 1.5277,
+      "step": 639
+    },
+    {
+      "epoch": 0.06547482032788562,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.728643216080403e-05,
+      "loss": 1.3518,
+      "step": 640
+    },
+    {
+      "epoch": 0.06557712473464794,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.728140703517589e-05,
+      "loss": 1.4282,
+      "step": 641
+    },
+    {
+      "epoch": 0.06567942914141027,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.727638190954774e-05,
+      "loss": 1.3913,
+      "step": 642
+    },
+    {
+      "epoch": 0.06578173354817259,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.72713567839196e-05,
+      "loss": 1.3288,
+      "step": 643
+    },
+    {
+      "epoch": 0.06588403795493492,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.726633165829146e-05,
+      "loss": 1.4148,
+      "step": 644
+    },
+    {
+      "epoch": 0.06598634236169723,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.726130653266332e-05,
+      "loss": 1.7091,
+      "step": 645
+    },
+    {
+      "epoch": 0.06608864676845955,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.725628140703517e-05,
+      "loss": 1.4155,
+      "step": 646
+    },
+    {
+      "epoch": 0.06619095117522188,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.725125628140705e-05,
+      "loss": 1.3545,
+      "step": 647
+    },
+    {
+      "epoch": 0.06629325558198419,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.72462311557789e-05,
+      "loss": 1.4757,
+      "step": 648
+    },
+    {
+      "epoch": 0.06639555998874652,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.724120603015076e-05,
+      "loss": 1.5601,
+      "step": 649
+    },
+    {
+      "epoch": 0.06649786439550884,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.723618090452262e-05,
+      "loss": 1.5804,
+      "step": 650
+    },
+    {
+      "epoch": 0.06660016880227115,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.723115577889448e-05,
+      "loss": 1.4527,
+      "step": 651
+    },
+    {
+      "epoch": 0.06670247320903348,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.722613065326633e-05,
+      "loss": 1.3373,
+      "step": 652
+    },
+    {
+      "epoch": 0.0668047776157958,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.72211055276382e-05,
+      "loss": 1.5092,
+      "step": 653
+    },
+    {
+      "epoch": 0.06690708202255811,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.721608040201005e-05,
+      "loss": 1.5344,
+      "step": 654
+    },
+    {
+      "epoch": 0.06700938642932044,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.721105527638192e-05,
+      "loss": 1.4727,
+      "step": 655
+    },
+    {
+      "epoch": 0.06711169083608276,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.720603015075376e-05,
+      "loss": 1.5189,
+      "step": 656
+    },
+    {
+      "epoch": 0.06721399524284509,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.720100502512564e-05,
+      "loss": 1.5639,
+      "step": 657
+    },
+    {
+      "epoch": 0.0673162996496074,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.719597989949749e-05,
+      "loss": 1.374,
+      "step": 658
+    },
+    {
+      "epoch": 0.06741860405636972,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.719095477386935e-05,
+      "loss": 1.4811,
+      "step": 659
+    },
+    {
+      "epoch": 0.06752090846313205,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.718592964824121e-05,
+      "loss": 1.4028,
+      "step": 660
+    },
+    {
+      "epoch": 0.06762321286989437,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.718090452261307e-05,
+      "loss": 1.4393,
+      "step": 661
+    },
+    {
+      "epoch": 0.0677255172766567,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.717587939698492e-05,
+      "loss": 1.6355,
+      "step": 662
+    },
+    {
+      "epoch": 0.06782782168341901,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.71708542713568e-05,
+      "loss": 1.3344,
+      "step": 663
+    },
+    {
+      "epoch": 0.06793012609018133,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.716582914572865e-05,
+      "loss": 1.4102,
+      "step": 664
+    },
+    {
+      "epoch": 0.06803243049694366,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.716080402010051e-05,
+      "loss": 1.5379,
+      "step": 665
+    },
+    {
+      "epoch": 0.06813473490370597,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.715577889447237e-05,
+      "loss": 1.4142,
+      "step": 666
+    },
+    {
+      "epoch": 0.0682370393104683,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.715075376884423e-05,
+      "loss": 1.463,
+      "step": 667
+    },
+    {
+      "epoch": 0.06833934371723062,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.714572864321608e-05,
+      "loss": 1.5498,
+      "step": 668
+    },
+    {
+      "epoch": 0.06844164812399293,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.714070351758794e-05,
+      "loss": 1.3842,
+      "step": 669
+    },
+    {
+      "epoch": 0.06854395253075526,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.71356783919598e-05,
+      "loss": 1.5071,
+      "step": 670
+    },
+    {
+      "epoch": 0.06864625693751758,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.713065326633167e-05,
+      "loss": 1.5204,
+      "step": 671
+    },
+    {
+      "epoch": 0.06874856134427991,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.712562814070352e-05,
+      "loss": 1.6183,
+      "step": 672
+    },
+    {
+      "epoch": 0.06885086575104223,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.712060301507538e-05,
+      "loss": 1.2785,
+      "step": 673
+    },
+    {
+      "epoch": 0.06895317015780454,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.711557788944724e-05,
+      "loss": 1.3523,
+      "step": 674
+    },
+    {
+      "epoch": 0.06905547456456687,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.71105527638191e-05,
+      "loss": 1.4686,
+      "step": 675
+    },
+    {
+      "epoch": 0.06915777897132919,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.710552763819096e-05,
+      "loss": 1.4108,
+      "step": 676
+    },
+    {
+      "epoch": 0.06926008337809152,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.710050251256281e-05,
+      "loss": 1.2943,
+      "step": 677
+    },
+    {
+      "epoch": 0.06936238778485383,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.709547738693468e-05,
+      "loss": 1.5735,
+      "step": 678
+    },
+    {
+      "epoch": 0.06946469219161615,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.709045226130654e-05,
+      "loss": 1.4061,
+      "step": 679
+    },
+    {
+      "epoch": 0.06956699659837848,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.70854271356784e-05,
+      "loss": 1.5295,
+      "step": 680
+    },
+    {
+      "epoch": 0.0696693010051408,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.708040201005025e-05,
+      "loss": 1.5872,
+      "step": 681
+    },
+    {
+      "epoch": 0.06977160541190312,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.707537688442212e-05,
+      "loss": 1.4311,
+      "step": 682
+    },
+    {
+      "epoch": 0.06987390981866544,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.707035175879397e-05,
+      "loss": 1.4867,
+      "step": 683
+    },
+    {
+      "epoch": 0.06997621422542775,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.706532663316583e-05,
+      "loss": 1.34,
+      "step": 684
+    },
+    {
+      "epoch": 0.07007851863219008,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.706030150753768e-05,
+      "loss": 1.4995,
+      "step": 685
+    },
+    {
+      "epoch": 0.0701808230389524,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.705527638190956e-05,
+      "loss": 1.363,
+      "step": 686
+    },
+    {
+      "epoch": 0.07028312744571473,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.70502512562814e-05,
+      "loss": 1.4563,
+      "step": 687
+    },
+    {
+      "epoch": 0.07038543185247705,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.704522613065327e-05,
+      "loss": 1.381,
+      "step": 688
+    },
+    {
+      "epoch": 0.07048773625923936,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.704020100502513e-05,
+      "loss": 1.6578,
+      "step": 689
+    },
+    {
+      "epoch": 0.07059004066600169,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.703517587939699e-05,
+      "loss": 1.3945,
+      "step": 690
+    },
+    {
+      "epoch": 0.070692345072764,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.703015075376884e-05,
+      "loss": 1.4265,
+      "step": 691
+    },
+    {
+      "epoch": 0.07079464947952634,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.702512562814072e-05,
+      "loss": 1.3114,
+      "step": 692
+    },
+    {
+      "epoch": 0.07089695388628865,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.702010050251257e-05,
+      "loss": 1.6659,
+      "step": 693
+    },
+    {
+      "epoch": 0.07099925829305097,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.701507537688443e-05,
+      "loss": 1.3119,
+      "step": 694
+    },
+    {
+      "epoch": 0.0711015626998133,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.701005025125629e-05,
+      "loss": 1.5351,
+      "step": 695
+    },
+    {
+      "epoch": 0.07120386710657561,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.700502512562815e-05,
+      "loss": 1.3638,
+      "step": 696
+    },
+    {
+      "epoch": 0.07130617151333794,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.7e-05,
+      "loss": 1.6732,
+      "step": 697
+    },
+    {
+      "epoch": 0.07140847592010026,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.699497487437188e-05,
+      "loss": 1.2725,
+      "step": 698
+    },
+    {
+      "epoch": 0.07151078032686257,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.698994974874372e-05,
+      "loss": 1.4797,
+      "step": 699
+    },
+    {
+      "epoch": 0.0716130847336249,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.698492462311559e-05,
+      "loss": 1.3554,
+      "step": 700
+    },
+    {
+      "epoch": 0.07171538914038722,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.697989949748743e-05,
+      "loss": 1.6185,
+      "step": 701
+    },
+    {
+      "epoch": 0.07181769354714955,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.697487437185931e-05,
+      "loss": 1.572,
+      "step": 702
+    },
+    {
+      "epoch": 0.07191999795391187,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.696984924623116e-05,
+      "loss": 1.5729,
+      "step": 703
+    },
+    {
+      "epoch": 0.07202230236067418,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.696482412060302e-05,
+      "loss": 1.2853,
+      "step": 704
+    },
+    {
+      "epoch": 0.07212460676743651,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.695979899497488e-05,
+      "loss": 1.4226,
+      "step": 705
+    },
+    {
+      "epoch": 0.07222691117419883,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.695477386934674e-05,
+      "loss": 1.4438,
+      "step": 706
+    },
+    {
+      "epoch": 0.07232921558096116,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.694974874371859e-05,
+      "loss": 1.442,
+      "step": 707
+    },
+    {
+      "epoch": 0.07243151998772347,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.694472361809045e-05,
+      "loss": 1.408,
+      "step": 708
+    },
+    {
+      "epoch": 0.07253382439448579,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.693969849246232e-05,
+      "loss": 1.2642,
+      "step": 709
+    },
+    {
+      "epoch": 0.07263612880124812,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.693467336683418e-05,
+      "loss": 1.6749,
+      "step": 710
+    },
+    {
+      "epoch": 0.07273843320801043,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.692964824120604e-05,
+      "loss": 1.4077,
+      "step": 711
+    },
+    {
+      "epoch": 0.07284073761477276,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.692462311557789e-05,
+      "loss": 1.4429,
+      "step": 712
+    },
+    {
+      "epoch": 0.07294304202153508,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.691959798994975e-05,
+      "loss": 1.5588,
+      "step": 713
+    },
+    {
+      "epoch": 0.0730453464282974,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.691457286432161e-05,
+      "loss": 1.2885,
+      "step": 714
+    },
+    {
+      "epoch": 0.07314765083505972,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.690954773869348e-05,
+      "loss": 1.4156,
+      "step": 715
+    },
+    {
+      "epoch": 0.07324995524182204,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.690452261306532e-05,
+      "loss": 1.4163,
+      "step": 716
+    },
+    {
+      "epoch": 0.07335225964858436,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.689949748743719e-05,
+      "loss": 1.4334,
+      "step": 717
+    },
+    {
+      "epoch": 0.07345456405534669,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.689447236180905e-05,
+      "loss": 1.4348,
+      "step": 718
+    },
+    {
+      "epoch": 0.073556868462109,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.688944723618091e-05,
+      "loss": 1.5014,
+      "step": 719
+    },
+    {
+      "epoch": 0.07365917286887133,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.688442211055276e-05,
+      "loss": 1.3844,
+      "step": 720
+    },
+    {
+      "epoch": 0.07376147727563365,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.687939698492463e-05,
+      "loss": 1.3659,
+      "step": 721
+    },
+    {
+      "epoch": 0.07386378168239596,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.687437185929648e-05,
+      "loss": 1.3946,
+      "step": 722
+    },
+    {
+      "epoch": 0.07396608608915829,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.686934673366834e-05,
+      "loss": 1.4004,
+      "step": 723
+    },
+    {
+      "epoch": 0.07406839049592061,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.686432160804021e-05,
+      "loss": 1.4616,
+      "step": 724
+    },
+    {
+      "epoch": 0.07417069490268294,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.685929648241207e-05,
+      "loss": 1.2591,
+      "step": 725
+    },
+    {
+      "epoch": 0.07427299930944525,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.685427135678392e-05,
+      "loss": 1.5843,
+      "step": 726
+    },
+    {
+      "epoch": 0.07437530371620757,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.684924623115579e-05,
+      "loss": 1.4079,
+      "step": 727
+    },
+    {
+      "epoch": 0.0744776081229699,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.684422110552764e-05,
+      "loss": 1.3456,
+      "step": 728
+    },
+    {
+      "epoch": 0.07457991252973221,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.68391959798995e-05,
+      "loss": 1.2591,
+      "step": 729
+    },
+    {
+      "epoch": 0.07468221693649454,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.683417085427137e-05,
+      "loss": 1.5003,
+      "step": 730
+    },
+    {
+      "epoch": 0.07478452134325686,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.682914572864323e-05,
+      "loss": 1.2439,
+      "step": 731
+    },
+    {
+      "epoch": 0.07488682575001918,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.682412060301508e-05,
+      "loss": 1.4217,
+      "step": 732
+    },
+    {
+      "epoch": 0.0749891301567815,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.681909547738694e-05,
+      "loss": 1.6303,
+      "step": 733
+    },
+    {
+      "epoch": 0.07509143456354382,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.68140703517588e-05,
+      "loss": 1.6364,
+      "step": 734
+    },
+    {
+      "epoch": 0.07519373897030615,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.680904522613066e-05,
+      "loss": 1.51,
+      "step": 735
+    },
+    {
+      "epoch": 0.07529604337706847,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.680402010050251e-05,
+      "loss": 1.4796,
+      "step": 736
+    },
+    {
+      "epoch": 0.07539834778383078,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.679899497487439e-05,
+      "loss": 1.576,
+      "step": 737
+    },
+    {
+      "epoch": 0.07550065219059311,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.679396984924623e-05,
+      "loss": 1.6164,
+      "step": 738
+    },
+    {
+      "epoch": 0.07560295659735543,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.67889447236181e-05,
+      "loss": 1.4615,
+      "step": 739
+    },
+    {
+      "epoch": 0.07570526100411776,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.678391959798996e-05,
+      "loss": 1.204,
+      "step": 740
+    },
+    {
+      "epoch": 0.07580756541088007,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.677889447236182e-05,
+      "loss": 1.2792,
+      "step": 741
+    },
+    {
+      "epoch": 0.07590986981764239,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.677386934673367e-05,
+      "loss": 1.5512,
+      "step": 742
+    },
+    {
+      "epoch": 0.07601217422440472,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.676884422110553e-05,
+      "loss": 1.6731,
+      "step": 743
+    },
+    {
+      "epoch": 0.07611447863116703,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.67638190954774e-05,
+      "loss": 1.589,
+      "step": 744
+    },
+    {
+      "epoch": 0.07621678303792936,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.675879396984924e-05,
+      "loss": 1.3559,
+      "step": 745
+    },
+    {
+      "epoch": 0.07631908744469168,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.675376884422112e-05,
+      "loss": 1.4473,
+      "step": 746
+    },
+    {
+      "epoch": 0.076421391851454,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.674874371859297e-05,
+      "loss": 1.4798,
+      "step": 747
+    },
+    {
+      "epoch": 0.07652369625821633,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.674371859296483e-05,
+      "loss": 1.4264,
+      "step": 748
+    },
+    {
+      "epoch": 0.07662600066497864,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.673869346733668e-05,
+      "loss": 1.5053,
+      "step": 749
+    },
+    {
+      "epoch": 0.07672830507174097,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.673366834170855e-05,
+      "loss": 1.5566,
+      "step": 750
+    },
+    {
+      "epoch": 0.07683060947850329,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.67286432160804e-05,
+      "loss": 1.296,
+      "step": 751
+    },
+    {
+      "epoch": 0.0769329138852656,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.672361809045226e-05,
+      "loss": 1.4704,
+      "step": 752
+    },
+    {
+      "epoch": 0.07703521829202793,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.671859296482412e-05,
+      "loss": 1.4495,
+      "step": 753
+    },
+    {
+      "epoch": 0.07713752269879025,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.671356783919599e-05,
+      "loss": 1.5227,
+      "step": 754
+    },
+    {
+      "epoch": 0.07723982710555258,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.670854271356784e-05,
+      "loss": 1.4587,
+      "step": 755
+    },
+    {
+      "epoch": 0.0773421315123149,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.670351758793971e-05,
+      "loss": 1.6027,
+      "step": 756
+    },
+    {
+      "epoch": 0.07744443591907721,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.669849246231156e-05,
+      "loss": 1.4733,
+      "step": 757
+    },
+    {
+      "epoch": 0.07754674032583954,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.669346733668342e-05,
+      "loss": 1.3853,
+      "step": 758
+    },
+    {
+      "epoch": 0.07764904473260185,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.668844221105528e-05,
+      "loss": 1.4968,
+      "step": 759
+    },
+    {
+      "epoch": 0.07775134913936418,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.668341708542715e-05,
+      "loss": 1.5595,
+      "step": 760
+    },
+    {
+      "epoch": 0.0778536535461265,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.6678391959799e-05,
+      "loss": 1.5053,
+      "step": 761
+    },
+    {
+      "epoch": 0.07795595795288882,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.667336683417087e-05,
+      "loss": 1.4917,
+      "step": 762
+    },
+    {
+      "epoch": 0.07805826235965115,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.666834170854272e-05,
+      "loss": 1.5216,
+      "step": 763
+    },
+    {
+      "epoch": 0.07816056676641346,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.666331658291458e-05,
+      "loss": 1.3775,
+      "step": 764
+    },
+    {
+      "epoch": 0.07826287117317579,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.665829145728643e-05,
+      "loss": 1.6246,
+      "step": 765
+    },
+    {
+      "epoch": 0.0783651755799381,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.66532663316583e-05,
+      "loss": 1.2643,
+      "step": 766
+    },
+    {
+      "epoch": 0.07846747998670042,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.664824120603015e-05,
+      "loss": 1.4596,
+      "step": 767
+    },
+    {
+      "epoch": 0.07856978439346275,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.664321608040201e-05,
+      "loss": 1.4717,
+      "step": 768
+    },
+    {
+      "epoch": 0.07867208880022507,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.663819095477388e-05,
+      "loss": 1.4434,
+      "step": 769
+    },
+    {
+      "epoch": 0.0787743932069874,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.663316582914574e-05,
+      "loss": 1.353,
+      "step": 770
+    },
+    {
+      "epoch": 0.07887669761374971,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.662814070351759e-05,
+      "loss": 1.4666,
+      "step": 771
+    },
+    {
+      "epoch": 0.07897900202051203,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.662311557788946e-05,
+      "loss": 1.5806,
+      "step": 772
+    },
+    {
+      "epoch": 0.07908130642727436,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.661809045226131e-05,
+      "loss": 1.5159,
+      "step": 773
+    },
+    {
+      "epoch": 0.07918361083403667,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.661306532663317e-05,
+      "loss": 1.3781,
+      "step": 774
+    },
+    {
+      "epoch": 0.079285915240799,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.660804020100504e-05,
+      "loss": 1.4987,
+      "step": 775
+    },
+    {
+      "epoch": 0.07938821964756132,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.66030150753769e-05,
+      "loss": 1.5012,
+      "step": 776
+    },
+    {
+      "epoch": 0.07949052405432364,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.659798994974875e-05,
+      "loss": 1.4434,
+      "step": 777
+    },
+    {
+      "epoch": 0.07959282846108597,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.659296482412061e-05,
+      "loss": 1.6261,
+      "step": 778
+    },
+    {
+      "epoch": 0.07969513286784828,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.658793969849247e-05,
+      "loss": 1.4979,
+      "step": 779
+    },
+    {
+      "epoch": 0.0797974372746106,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.658291457286432e-05,
+      "loss": 1.5506,
+      "step": 780
+    },
+    {
+      "epoch": 0.07989974168137293,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.657788944723618e-05,
+      "loss": 1.5824,
+      "step": 781
+    },
+    {
+      "epoch": 0.08000204608813524,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.657286432160804e-05,
+      "loss": 1.3584,
+      "step": 782
+    },
+    {
+      "epoch": 0.08010435049489757,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.65678391959799e-05,
+      "loss": 1.3233,
+      "step": 783
+    },
+    {
+      "epoch": 0.08020665490165989,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.656281407035175e-05,
+      "loss": 1.454,
+      "step": 784
+    },
+    {
+      "epoch": 0.0803089593084222,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.655778894472363e-05,
+      "loss": 1.3839,
+      "step": 785
+    },
+    {
+      "epoch": 0.08041126371518453,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.655276381909548e-05,
+      "loss": 1.3733,
+      "step": 786
+    },
+    {
+      "epoch": 0.08051356812194685,
+      "grad_norm": 0.25,
+      "learning_rate": 9.654773869346734e-05,
+      "loss": 1.3631,
+      "step": 787
+    },
+    {
+      "epoch": 0.08061587252870918,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.65427135678392e-05,
+      "loss": 1.6933,
+      "step": 788
+    },
+    {
+      "epoch": 0.0807181769354715,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.653768844221106e-05,
+      "loss": 1.4814,
+      "step": 789
+    },
+    {
+      "epoch": 0.08082048134223381,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.653266331658291e-05,
+      "loss": 1.4724,
+      "step": 790
+    },
+    {
+      "epoch": 0.08092278574899614,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.652763819095479e-05,
+      "loss": 1.4719,
+      "step": 791
+    },
+    {
+      "epoch": 0.08102509015575846,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.652261306532664e-05,
+      "loss": 1.3178,
+      "step": 792
+    },
+    {
+      "epoch": 0.08112739456252079,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.65175879396985e-05,
+      "loss": 1.43,
+      "step": 793
+    },
+    {
+      "epoch": 0.0812296989692831,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.651256281407036e-05,
+      "loss": 1.4455,
+      "step": 794
+    },
+    {
+      "epoch": 0.08133200337604542,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.650753768844222e-05,
+      "loss": 1.2836,
+      "step": 795
+    },
+    {
+      "epoch": 0.08143430778280775,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.650251256281407e-05,
+      "loss": 1.3716,
+      "step": 796
+    },
+    {
+      "epoch": 0.08153661218957006,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.649748743718593e-05,
+      "loss": 1.4214,
+      "step": 797
+    },
+    {
+      "epoch": 0.08163891659633239,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.64924623115578e-05,
+      "loss": 1.3869,
+      "step": 798
+    },
+    {
+      "epoch": 0.08174122100309471,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.648743718592966e-05,
+      "loss": 1.5394,
+      "step": 799
+    },
+    {
+      "epoch": 0.08184352540985702,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.64824120603015e-05,
+      "loss": 1.4325,
+      "step": 800
+    },
+    {
+      "epoch": 0.08194582981661935,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.647738693467338e-05,
+      "loss": 1.5178,
+      "step": 801
+    },
+    {
+      "epoch": 0.08204813422338167,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.647236180904523e-05,
+      "loss": 1.4576,
+      "step": 802
+    },
+    {
+      "epoch": 0.082150438630144,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.646733668341709e-05,
+      "loss": 1.5449,
+      "step": 803
+    },
+    {
+      "epoch": 0.08225274303690631,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.646231155778895e-05,
+      "loss": 1.3722,
+      "step": 804
+    },
+    {
+      "epoch": 0.08235504744366863,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.645728643216081e-05,
+      "loss": 1.4105,
+      "step": 805
+    },
+    {
+      "epoch": 0.08245735185043096,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.645226130653266e-05,
+      "loss": 1.4878,
+      "step": 806
+    },
+    {
+      "epoch": 0.08255965625719328,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.644723618090454e-05,
+      "loss": 1.303,
+      "step": 807
+    },
+    {
+      "epoch": 0.0826619606639556,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.644221105527639e-05,
+      "loss": 1.4592,
+      "step": 808
+    },
+    {
+      "epoch": 0.08276426507071792,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.643718592964825e-05,
+      "loss": 1.3606,
+      "step": 809
+    },
+    {
+      "epoch": 0.08286656947748024,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.64321608040201e-05,
+      "loss": 1.3988,
+      "step": 810
+    },
+    {
+      "epoch": 0.08296887388424257,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.642713567839196e-05,
+      "loss": 1.434,
+      "step": 811
+    },
+    {
+      "epoch": 0.08307117829100488,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.642211055276382e-05,
+      "loss": 1.5135,
+      "step": 812
+    },
+    {
+      "epoch": 0.08317348269776721,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.641708542713568e-05,
+      "loss": 1.4898,
+      "step": 813
+    },
+    {
+      "epoch": 0.08327578710452953,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.641206030150755e-05,
+      "loss": 1.4777,
+      "step": 814
+    },
+    {
+      "epoch": 0.08337809151129184,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.64070351758794e-05,
+      "loss": 1.4209,
+      "step": 815
+    },
+    {
+      "epoch": 0.08348039591805417,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.640201005025126e-05,
+      "loss": 1.5343,
+      "step": 816
+    },
+    {
+      "epoch": 0.08358270032481649,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.639698492462312e-05,
+      "loss": 1.326,
+      "step": 817
+    },
+    {
+      "epoch": 0.08368500473157882,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.639195979899498e-05,
+      "loss": 1.265,
+      "step": 818
+    },
+    {
+      "epoch": 0.08378730913834113,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.638693467336683e-05,
+      "loss": 1.5861,
+      "step": 819
+    },
+    {
+      "epoch": 0.08388961354510345,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.63819095477387e-05,
+      "loss": 1.3839,
+      "step": 820
+    },
+    {
+      "epoch": 0.08399191795186578,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.637688442211055e-05,
+      "loss": 1.488,
+      "step": 821
+    },
+    {
+      "epoch": 0.0840942223586281,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.637185929648242e-05,
+      "loss": 1.4429,
+      "step": 822
+    },
+    {
+      "epoch": 0.08419652676539043,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.636683417085428e-05,
+      "loss": 1.5707,
+      "step": 823
+    },
+    {
+      "epoch": 0.08429883117215274,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.636180904522614e-05,
+      "loss": 1.5355,
+      "step": 824
+    },
+    {
+      "epoch": 0.08440113557891506,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.635678391959799e-05,
+      "loss": 1.3084,
+      "step": 825
+    },
+    {
+      "epoch": 0.08450343998567739,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.635175879396985e-05,
+      "loss": 1.5519,
+      "step": 826
+    },
+    {
+      "epoch": 0.0846057443924397,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.634673366834171e-05,
+      "loss": 1.4366,
+      "step": 827
+    },
+    {
+      "epoch": 0.08470804879920203,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.634170854271357e-05,
+      "loss": 1.3173,
+      "step": 828
+    },
+    {
+      "epoch": 0.08481035320596435,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.633668341708542e-05,
+      "loss": 1.3873,
+      "step": 829
+    },
+    {
+      "epoch": 0.08491265761272666,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.63316582914573e-05,
+      "loss": 1.4825,
+      "step": 830
+    },
+    {
+      "epoch": 0.085014962019489,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.632663316582915e-05,
+      "loss": 1.4827,
+      "step": 831
+    },
+    {
+      "epoch": 0.08511726642625131,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.632160804020101e-05,
+      "loss": 1.5157,
+      "step": 832
+    },
+    {
+      "epoch": 0.08521957083301364,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.631658291457287e-05,
+      "loss": 1.3386,
+      "step": 833
+    },
+    {
+      "epoch": 0.08532187523977595,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.631155778894473e-05,
+      "loss": 1.4663,
+      "step": 834
+    },
+    {
+      "epoch": 0.08542417964653827,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.630653266331658e-05,
+      "loss": 1.6365,
+      "step": 835
+    },
+    {
+      "epoch": 0.0855264840533006,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.630150753768846e-05,
+      "loss": 1.5475,
+      "step": 836
+    },
+    {
+      "epoch": 0.08562878846006292,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.62964824120603e-05,
+      "loss": 1.435,
+      "step": 837
+    },
+    {
+      "epoch": 0.08573109286682525,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.629145728643217e-05,
+      "loss": 1.488,
+      "step": 838
+    },
+    {
+      "epoch": 0.08583339727358756,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.628643216080403e-05,
+      "loss": 1.5689,
+      "step": 839
+    },
+    {
+      "epoch": 0.08593570168034988,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.628140703517589e-05,
+      "loss": 1.3256,
+      "step": 840
+    },
+    {
+      "epoch": 0.0860380060871122,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.627638190954774e-05,
+      "loss": 1.4685,
+      "step": 841
+    },
+    {
+      "epoch": 0.08614031049387452,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.62713567839196e-05,
+      "loss": 1.3367,
+      "step": 842
+    },
+    {
+      "epoch": 0.08624261490063684,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.626633165829146e-05,
+      "loss": 1.2637,
+      "step": 843
+    },
+    {
+      "epoch": 0.08634491930739917,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.626130653266333e-05,
+      "loss": 1.5606,
+      "step": 844
+    },
+    {
+      "epoch": 0.08644722371416148,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.625628140703517e-05,
+      "loss": 1.376,
+      "step": 845
+    },
+    {
+      "epoch": 0.08654952812092381,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.625125628140704e-05,
+      "loss": 1.3469,
+      "step": 846
+    },
+    {
+      "epoch": 0.08665183252768613,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.62462311557789e-05,
+      "loss": 1.501,
+      "step": 847
+    },
+    {
+      "epoch": 0.08675413693444844,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.624120603015076e-05,
+      "loss": 1.4197,
+      "step": 848
+    },
+    {
+      "epoch": 0.08685644134121077,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.623618090452262e-05,
+      "loss": 1.3986,
+      "step": 849
+    },
+    {
+      "epoch": 0.08695874574797309,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.623115577889447e-05,
+      "loss": 1.4805,
+      "step": 850
+    },
+    {
+      "epoch": 0.08706105015473542,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.622613065326633e-05,
+      "loss": 1.431,
+      "step": 851
+    },
+    {
+      "epoch": 0.08716335456149774,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.62211055276382e-05,
+      "loss": 1.4459,
+      "step": 852
+    },
+    {
+      "epoch": 0.08726565896826005,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.621608040201006e-05,
+      "loss": 1.5319,
+      "step": 853
+    },
+    {
+      "epoch": 0.08736796337502238,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.62110552763819e-05,
+      "loss": 1.3824,
+      "step": 854
+    },
+    {
+      "epoch": 0.0874702677817847,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.620603015075378e-05,
+      "loss": 1.3656,
+      "step": 855
+    },
+    {
+      "epoch": 0.08757257218854703,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.620100502512563e-05,
+      "loss": 1.3199,
+      "step": 856
+    },
+    {
+      "epoch": 0.08767487659530934,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.619597989949749e-05,
+      "loss": 1.3755,
+      "step": 857
+    },
+    {
+      "epoch": 0.08777718100207166,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.619095477386934e-05,
+      "loss": 1.3676,
+      "step": 858
+    },
+    {
+      "epoch": 0.08787948540883399,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.618592964824122e-05,
+      "loss": 1.3713,
+      "step": 859
+    },
+    {
+      "epoch": 0.0879817898155963,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.618090452261306e-05,
+      "loss": 1.2352,
+      "step": 860
+    },
+    {
+      "epoch": 0.08808409422235863,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.617587939698493e-05,
+      "loss": 1.337,
+      "step": 861
+    },
+    {
+      "epoch": 0.08818639862912095,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.617085427135679e-05,
+      "loss": 1.4312,
+      "step": 862
+    },
+    {
+      "epoch": 0.08828870303588326,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.616582914572865e-05,
+      "loss": 1.4133,
+      "step": 863
+    },
+    {
+      "epoch": 0.0883910074426456,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.61608040201005e-05,
+      "loss": 1.4106,
+      "step": 864
+    },
+    {
+      "epoch": 0.08849331184940791,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.615577889447237e-05,
+      "loss": 1.5351,
+      "step": 865
+    },
+    {
+      "epoch": 0.08859561625617024,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.615075376884422e-05,
+      "loss": 1.4089,
+      "step": 866
+    },
+    {
+      "epoch": 0.08869792066293256,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.614572864321608e-05,
+      "loss": 1.3923,
+      "step": 867
+    },
+    {
+      "epoch": 0.08880022506969487,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.614070351758795e-05,
+      "loss": 1.339,
+      "step": 868
+    },
+    {
+      "epoch": 0.0889025294764572,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.613567839195981e-05,
+      "loss": 1.4209,
+      "step": 869
+    },
+    {
+      "epoch": 0.08900483388321952,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.613065326633166e-05,
+      "loss": 1.338,
+      "step": 870
+    },
+    {
+      "epoch": 0.08910713828998185,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.612562814070353e-05,
+      "loss": 1.3902,
+      "step": 871
+    },
+    {
+      "epoch": 0.08920944269674416,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.612060301507538e-05,
+      "loss": 1.2726,
+      "step": 872
+    },
+    {
+      "epoch": 0.08931174710350648,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.611557788944724e-05,
+      "loss": 1.4451,
+      "step": 873
+    },
+    {
+      "epoch": 0.08941405151026881,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.611055276381909e-05,
+      "loss": 1.4751,
+      "step": 874
+    },
+    {
+      "epoch": 0.08951635591703112,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.610552763819097e-05,
+      "loss": 1.4962,
+      "step": 875
+    },
+    {
+      "epoch": 0.08961866032379345,
+      "grad_norm": 0.25,
+      "learning_rate": 9.610050251256282e-05,
+      "loss": 1.2166,
+      "step": 876
+    },
+    {
+      "epoch": 0.08972096473055577,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.609547738693468e-05,
+      "loss": 1.2974,
+      "step": 877
+    },
+    {
+      "epoch": 0.08982326913731808,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.609045226130654e-05,
+      "loss": 1.4087,
+      "step": 878
+    },
+    {
+      "epoch": 0.08992557354408041,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.60854271356784e-05,
+      "loss": 1.4147,
+      "step": 879
+    },
+    {
+      "epoch": 0.09002787795084273,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.608040201005025e-05,
+      "loss": 1.5533,
+      "step": 880
+    },
+    {
+      "epoch": 0.09013018235760506,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.607537688442211e-05,
+      "loss": 1.32,
+      "step": 881
+    },
+    {
+      "epoch": 0.09023248676436738,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.607035175879397e-05,
+      "loss": 1.4089,
+      "step": 882
+    },
+    {
+      "epoch": 0.09033479117112969,
+      "grad_norm": 0.25,
+      "learning_rate": 9.606532663316584e-05,
+      "loss": 1.4082,
+      "step": 883
+    },
+    {
+      "epoch": 0.09043709557789202,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.60603015075377e-05,
+      "loss": 1.5125,
+      "step": 884
+    },
+    {
+      "epoch": 0.09053939998465434,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.605527638190955e-05,
+      "loss": 1.3798,
+      "step": 885
+    },
+    {
+      "epoch": 0.09064170439141667,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.605025125628141e-05,
+      "loss": 1.4247,
+      "step": 886
+    },
+    {
+      "epoch": 0.09074400879817898,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.604522613065327e-05,
+      "loss": 1.427,
+      "step": 887
+    },
+    {
+      "epoch": 0.0908463132049413,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.604020100502513e-05,
+      "loss": 1.393,
+      "step": 888
+    },
+    {
+      "epoch": 0.09094861761170363,
+      "grad_norm": 0.25,
+      "learning_rate": 9.603517587939698e-05,
+      "loss": 1.3815,
+      "step": 889
+    },
+    {
+      "epoch": 0.09105092201846594,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.603015075376884e-05,
+      "loss": 1.5265,
+      "step": 890
+    },
+    {
+      "epoch": 0.09115322642522827,
+      "grad_norm": 0.25,
+      "learning_rate": 9.60251256281407e-05,
+      "loss": 1.399,
+      "step": 891
+    },
+    {
+      "epoch": 0.09125553083199059,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.602010050251257e-05,
+      "loss": 1.4182,
+      "step": 892
+    },
+    {
+      "epoch": 0.0913578352387529,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.601507537688442e-05,
+      "loss": 1.4047,
+      "step": 893
+    },
+    {
+      "epoch": 0.09146013964551523,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.601005025125629e-05,
+      "loss": 1.6254,
+      "step": 894
+    },
+    {
+      "epoch": 0.09156244405227755,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.600502512562814e-05,
+      "loss": 1.406,
+      "step": 895
+    },
+    {
+      "epoch": 0.09166474845903988,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.6e-05,
+      "loss": 1.3618,
+      "step": 896
+    },
+    {
+      "epoch": 0.0917670528658022,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.599497487437186e-05,
+      "loss": 1.4824,
+      "step": 897
+    },
+    {
+      "epoch": 0.09186935727256451,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.598994974874373e-05,
+      "loss": 1.4235,
+      "step": 898
+    },
+    {
+      "epoch": 0.09197166167932684,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.598492462311557e-05,
+      "loss": 1.4603,
+      "step": 899
+    },
+    {
+      "epoch": 0.09207396608608916,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.597989949748745e-05,
+      "loss": 1.5502,
+      "step": 900
+    },
+    {
+      "epoch": 0.09217627049285149,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.59748743718593e-05,
+      "loss": 1.4405,
+      "step": 901
+    },
+    {
+      "epoch": 0.0922785748996138,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.596984924623116e-05,
+      "loss": 1.4673,
+      "step": 902
+    },
+    {
+      "epoch": 0.09238087930637612,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.596482412060302e-05,
+      "loss": 1.4401,
+      "step": 903
+    },
+    {
+      "epoch": 0.09248318371313845,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.595979899497489e-05,
+      "loss": 1.4533,
+      "step": 904
+    },
+    {
+      "epoch": 0.09258548811990076,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.595477386934673e-05,
+      "loss": 1.5384,
+      "step": 905
+    },
+    {
+      "epoch": 0.09268779252666308,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.59497487437186e-05,
+      "loss": 1.5737,
+      "step": 906
+    },
+    {
+      "epoch": 0.09279009693342541,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.594472361809046e-05,
+      "loss": 1.3807,
+      "step": 907
+    },
+    {
+      "epoch": 0.09289240134018772,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.593969849246232e-05,
+      "loss": 1.5406,
+      "step": 908
+    },
+    {
+      "epoch": 0.09299470574695005,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.593467336683417e-05,
+      "loss": 1.4258,
+      "step": 909
+    },
+    {
+      "epoch": 0.09309701015371237,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.592964824120604e-05,
+      "loss": 1.2685,
+      "step": 910
+    },
+    {
+      "epoch": 0.09319931456047469,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.592462311557789e-05,
+      "loss": 1.5115,
+      "step": 911
+    },
+    {
+      "epoch": 0.09330161896723702,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.591959798994975e-05,
+      "loss": 1.3548,
+      "step": 912
+    },
+    {
+      "epoch": 0.09340392337399933,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.591457286432162e-05,
+      "loss": 1.5981,
+      "step": 913
+    },
+    {
+      "epoch": 0.09350622778076166,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.590954773869348e-05,
+      "loss": 1.4968,
+      "step": 914
+    },
+    {
+      "epoch": 0.09360853218752398,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.590452261306533e-05,
+      "loss": 1.5362,
+      "step": 915
+    },
+    {
+      "epoch": 0.09371083659428629,
+      "grad_norm": 0.25,
+      "learning_rate": 9.589949748743719e-05,
+      "loss": 1.2726,
+      "step": 916
+    },
+    {
+      "epoch": 0.09381314100104862,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.589447236180905e-05,
+      "loss": 1.3958,
+      "step": 917
+    },
+    {
+      "epoch": 0.09391544540781094,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.588944723618091e-05,
+      "loss": 1.4526,
+      "step": 918
+    },
+    {
+      "epoch": 0.09401774981457327,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.588442211055276e-05,
+      "loss": 1.349,
+      "step": 919
+    },
+    {
+      "epoch": 0.09412005422133558,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.587939698492462e-05,
+      "loss": 1.4615,
+      "step": 920
+    },
+    {
+      "epoch": 0.0942223586280979,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.587437185929649e-05,
+      "loss": 1.3326,
+      "step": 921
+    },
+    {
+      "epoch": 0.09432466303486023,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.586934673366835e-05,
+      "loss": 1.2981,
+      "step": 922
+    },
+    {
+      "epoch": 0.09442696744162254,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.586432160804021e-05,
+      "loss": 1.4485,
+      "step": 923
+    },
+    {
+      "epoch": 0.09452927184838487,
+      "grad_norm": 0.25,
+      "learning_rate": 9.585929648241206e-05,
+      "loss": 1.3519,
+      "step": 924
+    },
+    {
+      "epoch": 0.09463157625514719,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.585427135678392e-05,
+      "loss": 1.5551,
+      "step": 925
+    },
+    {
+      "epoch": 0.0947338806619095,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.584924623115578e-05,
+      "loss": 1.2236,
+      "step": 926
+    },
+    {
+      "epoch": 0.09483618506867184,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.584422110552764e-05,
+      "loss": 1.3476,
+      "step": 927
+    },
+    {
+      "epoch": 0.09493848947543415,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.583919597989949e-05,
+      "loss": 1.4733,
+      "step": 928
+    },
+    {
+      "epoch": 0.09504079388219648,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.583417085427137e-05,
+      "loss": 1.4563,
+      "step": 929
+    },
+    {
+      "epoch": 0.0951430982889588,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.582914572864322e-05,
+      "loss": 1.5595,
+      "step": 930
+    },
+    {
+      "epoch": 0.09524540269572111,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.582412060301508e-05,
+      "loss": 1.3932,
+      "step": 931
+    },
+    {
+      "epoch": 0.09534770710248344,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.581909547738694e-05,
+      "loss": 1.4058,
+      "step": 932
+    },
+    {
+      "epoch": 0.09545001150924576,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.58140703517588e-05,
+      "loss": 1.4904,
+      "step": 933
+    },
+    {
+      "epoch": 0.09555231591600809,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.580904522613065e-05,
+      "loss": 1.3863,
+      "step": 934
+    },
+    {
+      "epoch": 0.0956546203227704,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.580402010050251e-05,
+      "loss": 1.4024,
+      "step": 935
+    },
+    {
+      "epoch": 0.09575692472953272,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.579899497487438e-05,
+      "loss": 1.4687,
+      "step": 936
+    },
+    {
+      "epoch": 0.09585922913629505,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.579396984924624e-05,
+      "loss": 1.3781,
+      "step": 937
+    },
+    {
+      "epoch": 0.09596153354305736,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.578894472361809e-05,
+      "loss": 1.501,
+      "step": 938
+    },
+    {
+      "epoch": 0.0960638379498197,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.578391959798996e-05,
+      "loss": 2.1388,
+      "step": 939
+    },
+    {
+      "epoch": 0.09616614235658201,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.577889447236181e-05,
+      "loss": 1.3809,
+      "step": 940
+    },
+    {
+      "epoch": 0.09626844676334433,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.577386934673367e-05,
+      "loss": 1.4965,
+      "step": 941
+    },
+    {
+      "epoch": 0.09637075117010666,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.576884422110553e-05,
+      "loss": 1.3862,
+      "step": 942
+    },
+    {
+      "epoch": 0.09647305557686897,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.57638190954774e-05,
+      "loss": 1.369,
+      "step": 943
+    },
+    {
+      "epoch": 0.0965753599836313,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.575879396984924e-05,
+      "loss": 1.5884,
+      "step": 944
+    },
+    {
+      "epoch": 0.09667766439039362,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.575376884422112e-05,
+      "loss": 1.4377,
+      "step": 945
+    },
+    {
+      "epoch": 0.09677996879715593,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.574874371859297e-05,
+      "loss": 1.3259,
+      "step": 946
+    },
+    {
+      "epoch": 0.09688227320391826,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.574371859296483e-05,
+      "loss": 1.547,
+      "step": 947
+    },
+    {
+      "epoch": 0.09698457761068058,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.573869346733669e-05,
+      "loss": 1.5853,
+      "step": 948
+    },
+    {
+      "epoch": 0.09708688201744291,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.573366834170855e-05,
+      "loss": 1.2984,
+      "step": 949
+    },
+    {
+      "epoch": 0.09718918642420522,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.57286432160804e-05,
+      "loss": 1.5164,
+      "step": 950
+    },
+    {
+      "epoch": 0.09729149083096754,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.572361809045227e-05,
+      "loss": 1.4572,
+      "step": 951
+    },
+    {
+      "epoch": 0.09739379523772987,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.571859296482413e-05,
+      "loss": 1.4137,
+      "step": 952
+    },
+    {
+      "epoch": 0.09749609964449218,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.571356783919599e-05,
+      "loss": 1.4252,
+      "step": 953
+    },
+    {
+      "epoch": 0.09759840405125451,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.570854271356784e-05,
+      "loss": 1.4936,
+      "step": 954
+    },
+    {
+      "epoch": 0.09770070845801683,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.57035175879397e-05,
+      "loss": 1.3336,
+      "step": 955
+    },
+    {
+      "epoch": 0.09780301286477915,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.569849246231156e-05,
+      "loss": 1.3072,
+      "step": 956
+    },
+    {
+      "epoch": 0.09790531727154148,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.569346733668342e-05,
+      "loss": 1.4865,
+      "step": 957
+    },
+    {
+      "epoch": 0.09800762167830379,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.568844221105529e-05,
+      "loss": 1.5836,
+      "step": 958
+    },
+    {
+      "epoch": 0.09810992608506612,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.568341708542713e-05,
+      "loss": 1.4528,
+      "step": 959
+    },
+    {
+      "epoch": 0.09821223049182844,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.5678391959799e-05,
+      "loss": 1.5253,
+      "step": 960
+    },
+    {
+      "epoch": 0.09831453489859075,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.567336683417086e-05,
+      "loss": 1.319,
+      "step": 961
+    },
+    {
+      "epoch": 0.09841683930535308,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.566834170854272e-05,
+      "loss": 1.2911,
+      "step": 962
+    },
+    {
+      "epoch": 0.0985191437121154,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.566331658291457e-05,
+      "loss": 1.5144,
+      "step": 963
+    },
+    {
+      "epoch": 0.09862144811887773,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.565829145728644e-05,
+      "loss": 1.3487,
+      "step": 964
+    },
+    {
+      "epoch": 0.09872375252564004,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.565326633165829e-05,
+      "loss": 1.4208,
+      "step": 965
+    },
+    {
+      "epoch": 0.09882605693240236,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.564824120603016e-05,
+      "loss": 1.383,
+      "step": 966
+    },
+    {
+      "epoch": 0.09892836133916469,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.5643216080402e-05,
+      "loss": 1.3051,
+      "step": 967
+    },
+    {
+      "epoch": 0.099030665745927,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.563819095477388e-05,
+      "loss": 1.3995,
+      "step": 968
+    },
+    {
+      "epoch": 0.09913297015268932,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.563316582914573e-05,
+      "loss": 1.3906,
+      "step": 969
+    },
+    {
+      "epoch": 0.09923527455945165,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.562814070351759e-05,
+      "loss": 1.4869,
+      "step": 970
+    },
+    {
+      "epoch": 0.09933757896621397,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.562311557788945e-05,
+      "loss": 1.4067,
+      "step": 971
+    },
+    {
+      "epoch": 0.0994398833729763,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.561809045226131e-05,
+      "loss": 1.3756,
+      "step": 972
+    },
+    {
+      "epoch": 0.09954218777973861,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.561306532663316e-05,
+      "loss": 1.5439,
+      "step": 973
+    },
+    {
+      "epoch": 0.09964449218650093,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.560804020100504e-05,
+      "loss": 1.4481,
+      "step": 974
+    },
+    {
+      "epoch": 0.09974679659326326,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.560301507537689e-05,
+      "loss": 1.5344,
+      "step": 975
+    },
+    {
+      "epoch": 0.09984910100002557,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.559798994974875e-05,
+      "loss": 1.3456,
+      "step": 976
+    },
+    {
+      "epoch": 0.0999514054067879,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.559296482412061e-05,
+      "loss": 1.7686,
+      "step": 977
+    },
+    {
+      "epoch": 0.10005370981355022,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.558793969849247e-05,
+      "loss": 1.3367,
+      "step": 978
+    },
+    {
+      "epoch": 0.10015601422031253,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.558291457286432e-05,
+      "loss": 1.6469,
+      "step": 979
+    },
+    {
+      "epoch": 0.10025831862707486,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.55778894472362e-05,
+      "loss": 1.469,
+      "step": 980
+    },
+    {
+      "epoch": 0.10036062303383718,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.557286432160804e-05,
+      "loss": 1.5007,
+      "step": 981
+    },
+    {
+      "epoch": 0.10046292744059951,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.556783919597991e-05,
+      "loss": 1.3586,
+      "step": 982
+    },
+    {
+      "epoch": 0.10056523184736182,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.556281407035176e-05,
+      "loss": 1.3461,
+      "step": 983
+    },
+    {
+      "epoch": 0.10066753625412414,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.555778894472363e-05,
+      "loss": 1.6054,
+      "step": 984
+    },
+    {
+      "epoch": 0.10076984066088647,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.555276381909548e-05,
+      "loss": 1.3886,
+      "step": 985
+    },
+    {
+      "epoch": 0.10087214506764879,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.554773869346734e-05,
+      "loss": 1.2513,
+      "step": 986
+    },
+    {
+      "epoch": 0.10097444947441112,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.55427135678392e-05,
+      "loss": 1.4395,
+      "step": 987
+    },
+    {
+      "epoch": 0.10107675388117343,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.553768844221107e-05,
+      "loss": 1.2448,
+      "step": 988
+    },
+    {
+      "epoch": 0.10117905828793575,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.553266331658291e-05,
+      "loss": 1.3882,
+      "step": 989
+    },
+    {
+      "epoch": 0.10128136269469808,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.552763819095478e-05,
+      "loss": 1.4483,
+      "step": 990
+    },
+    {
+      "epoch": 0.10138366710146039,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.552261306532664e-05,
+      "loss": 1.3888,
+      "step": 991
+    },
+    {
+      "epoch": 0.10148597150822272,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.55175879396985e-05,
+      "loss": 1.3309,
+      "step": 992
+    },
+    {
+      "epoch": 0.10158827591498504,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.551256281407036e-05,
+      "loss": 1.2854,
+      "step": 993
+    },
+    {
+      "epoch": 0.10169058032174735,
+      "grad_norm": 0.25,
+      "learning_rate": 9.550753768844221e-05,
+      "loss": 1.2986,
+      "step": 994
+    },
+    {
+      "epoch": 0.10179288472850968,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.550251256281407e-05,
+      "loss": 1.2908,
+      "step": 995
+    },
+    {
+      "epoch": 0.101895189135272,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.549748743718593e-05,
+      "loss": 1.4374,
+      "step": 996
+    },
+    {
+      "epoch": 0.10199749354203433,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.54924623115578e-05,
+      "loss": 1.4637,
+      "step": 997
+    },
+    {
+      "epoch": 0.10209979794879664,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.548743718592965e-05,
+      "loss": 1.2439,
+      "step": 998
+    },
+    {
+      "epoch": 0.10220210235555896,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.548241206030151e-05,
+      "loss": 1.4241,
+      "step": 999
+    },
+    {
+      "epoch": 0.10230440676232129,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.547738693467337e-05,
+      "loss": 1.465,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1024067111690836,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.547236180904523e-05,
+      "loss": 1.4816,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10250901557584594,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.546733668341708e-05,
+      "loss": 1.393,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10261131998260825,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.546231155778896e-05,
+      "loss": 1.276,
+      "step": 1003
+    },
+    {
+      "epoch": 0.10271362438937057,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.54572864321608e-05,
+      "loss": 1.5037,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1028159287961329,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.545226130653267e-05,
+      "loss": 1.3716,
+      "step": 1005
+    },
+    {
+      "epoch": 0.10291823320289521,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.544723618090453e-05,
+      "loss": 1.3784,
+      "step": 1006
+    },
+    {
+      "epoch": 0.10302053760965754,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.544221105527639e-05,
+      "loss": 1.1921,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10312284201641986,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.543718592964824e-05,
+      "loss": 1.6314,
+      "step": 1008
+    },
+    {
+      "epoch": 0.10322514642318217,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.543216080402011e-05,
+      "loss": 1.4715,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1033274508299445,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.542713567839196e-05,
+      "loss": 1.5602,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10342975523670682,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.542211055276382e-05,
+      "loss": 1.6179,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10353205964346915,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.541708542713569e-05,
+      "loss": 1.3969,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10363436405023146,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.541206030150755e-05,
+      "loss": 1.5656,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10373666845699378,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.54070351758794e-05,
+      "loss": 1.4676,
+      "step": 1014
+    },
+    {
+      "epoch": 0.10383897286375611,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.540201005025126e-05,
+      "loss": 1.411,
+      "step": 1015
+    },
+    {
+      "epoch": 0.10394127727051843,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.539698492462312e-05,
+      "loss": 1.3581,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10404358167728076,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.539195979899498e-05,
+      "loss": 1.4245,
+      "step": 1017
+    },
+    {
+      "epoch": 0.10414588608404307,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.538693467336683e-05,
+      "loss": 1.4888,
+      "step": 1018
+    },
+    {
+      "epoch": 0.10424819049080539,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.538190954773871e-05,
+      "loss": 1.5287,
+      "step": 1019
+    },
+    {
+      "epoch": 0.10435049489756772,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.537688442211056e-05,
+      "loss": 1.3537,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10445279930433003,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.537185929648242e-05,
+      "loss": 1.473,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10455510371109236,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.536683417085428e-05,
+      "loss": 1.5107,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10465740811785468,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.536180904522614e-05,
+      "loss": 1.5981,
+      "step": 1023
+    },
+    {
+      "epoch": 0.104759712524617,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.535678391959799e-05,
+      "loss": 1.477,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10486201693137932,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.535175879396985e-05,
+      "loss": 1.4047,
+      "step": 1025
+    },
+    {
+      "epoch": 0.10496432133814164,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.534673366834171e-05,
+      "loss": 1.5122,
+      "step": 1026
+    },
+    {
+      "epoch": 0.10506662574490397,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.534170854271358e-05,
+      "loss": 1.3144,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10516893015166628,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.533668341708543e-05,
+      "loss": 1.3699,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1052712345584286,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.533165829145729e-05,
+      "loss": 1.3045,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10537353896519093,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.532663316582915e-05,
+      "loss": 1.5869,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10547584337195325,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.532160804020101e-05,
+      "loss": 1.3302,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10557814777871556,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.531658291457287e-05,
+      "loss": 1.5478,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10568045218547789,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.531155778894472e-05,
+      "loss": 1.2986,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1057827565922402,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.530653266331658e-05,
+      "loss": 1.4628,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10588506099900254,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.530150753768845e-05,
+      "loss": 1.4972,
+      "step": 1035
+    },
+    {
+      "epoch": 0.10598736540576485,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.529648241206031e-05,
+      "loss": 1.2783,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10608966981252717,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.529145728643216e-05,
+      "loss": 1.5302,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1061919742192895,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.528643216080403e-05,
+      "loss": 1.45,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10629427862605181,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.528140703517588e-05,
+      "loss": 1.4843,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10639658303281414,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.527638190954774e-05,
+      "loss": 1.4724,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10649888743957646,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.52713567839196e-05,
+      "loss": 1.4652,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10660119184633877,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.526633165829147e-05,
+      "loss": 1.5171,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1067034962531011,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.526130653266331e-05,
+      "loss": 1.3151,
+      "step": 1043
+    },
+    {
+      "epoch": 0.10680580065986342,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.525628140703518e-05,
+      "loss": 1.526,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10690810506662575,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.525125628140704e-05,
+      "loss": 1.4842,
+      "step": 1045
+    },
+    {
+      "epoch": 0.10701040947338807,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.52462311557789e-05,
+      "loss": 1.4954,
+      "step": 1046
+    },
+    {
+      "epoch": 0.10711271388015038,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.524120603015075e-05,
+      "loss": 1.4233,
+      "step": 1047
+    },
+    {
+      "epoch": 0.10721501828691271,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.523618090452263e-05,
+      "loss": 1.5185,
+      "step": 1048
+    },
+    {
+      "epoch": 0.10731732269367503,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.523115577889447e-05,
+      "loss": 1.353,
+      "step": 1049
+    },
+    {
+      "epoch": 0.10741962710043736,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.522613065326634e-05,
+      "loss": 1.7096,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10752193150719967,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.52211055276382e-05,
+      "loss": 1.4951,
+      "step": 1051
+    },
+    {
+      "epoch": 0.10762423591396199,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.521608040201006e-05,
+      "loss": 1.4952,
+      "step": 1052
+    },
+    {
+      "epoch": 0.10772654032072432,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.521105527638191e-05,
+      "loss": 1.5536,
+      "step": 1053
+    },
+    {
+      "epoch": 0.10782884472748663,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.520603015075378e-05,
+      "loss": 1.4218,
+      "step": 1054
+    },
+    {
+      "epoch": 0.10793114913424896,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.520100502512563e-05,
+      "loss": 1.3663,
+      "step": 1055
+    },
+    {
+      "epoch": 0.10803345354101128,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.51959798994975e-05,
+      "loss": 1.4538,
+      "step": 1056
+    },
+    {
+      "epoch": 0.1081357579477736,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.519095477386936e-05,
+      "loss": 1.2673,
+      "step": 1057
+    },
+    {
+      "epoch": 0.10823806235453592,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.518592964824122e-05,
+      "loss": 1.4603,
+      "step": 1058
+    },
+    {
+      "epoch": 0.10834036676129824,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.518090452261307e-05,
+      "loss": 1.423,
+      "step": 1059
+    },
+    {
+      "epoch": 0.10844267116806057,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.517587939698493e-05,
+      "loss": 1.417,
+      "step": 1060
+    },
+    {
+      "epoch": 0.10854497557482289,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.517085427135679e-05,
+      "loss": 1.3901,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1086472799815852,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.516582914572865e-05,
+      "loss": 1.2679,
+      "step": 1062
+    },
+    {
+      "epoch": 0.10874958438834753,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.51608040201005e-05,
+      "loss": 1.3982,
+      "step": 1063
+    },
+    {
+      "epoch": 0.10885188879510985,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.515577889447236e-05,
+      "loss": 1.4699,
+      "step": 1064
+    },
+    {
+      "epoch": 0.10895419320187218,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.515075376884423e-05,
+      "loss": 1.5496,
+      "step": 1065
+    },
+    {
+      "epoch": 0.10905649760863449,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.514572864321609e-05,
+      "loss": 1.4047,
+      "step": 1066
+    },
+    {
+      "epoch": 0.10915880201539681,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.514070351758795e-05,
+      "loss": 1.4302,
+      "step": 1067
+    },
+    {
+      "epoch": 0.10926110642215914,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.51356783919598e-05,
+      "loss": 1.4157,
+      "step": 1068
+    },
+    {
+      "epoch": 0.10936341082892145,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.513065326633166e-05,
+      "loss": 1.4388,
+      "step": 1069
+    },
+    {
+      "epoch": 0.10946571523568378,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.512562814070352e-05,
+      "loss": 1.3216,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1095680196424461,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.512060301507538e-05,
+      "loss": 1.3952,
+      "step": 1071
+    },
+    {
+      "epoch": 0.10967032404920841,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.511557788944723e-05,
+      "loss": 1.4657,
+      "step": 1072
+    },
+    {
+      "epoch": 0.10977262845597074,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.511055276381911e-05,
+      "loss": 1.6761,
+      "step": 1073
+    },
+    {
+      "epoch": 0.10987493286273306,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.510552763819096e-05,
+      "loss": 1.4271,
+      "step": 1074
+    },
+    {
+      "epoch": 0.10997723726949539,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.510050251256282e-05,
+      "loss": 1.4589,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1100795416762577,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.509547738693467e-05,
+      "loss": 1.2355,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11018184608302002,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.509045226130654e-05,
+      "loss": 1.4498,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11028415048978235,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.508542713567839e-05,
+      "loss": 1.2292,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11038645489654467,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.508040201005025e-05,
+      "loss": 1.3677,
+      "step": 1079
+    },
+    {
+      "epoch": 0.110488759303307,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.507537688442212e-05,
+      "loss": 1.561,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11059106371006931,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.507035175879398e-05,
+      "loss": 1.5049,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11069336811683163,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.506532663316583e-05,
+      "loss": 1.522,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11079567252359396,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.50603015075377e-05,
+      "loss": 1.505,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11089797693035627,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.505527638190955e-05,
+      "loss": 1.6358,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1110002813371186,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.505025125628141e-05,
+      "loss": 1.4234,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11110258574388092,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.504522613065327e-05,
+      "loss": 1.4225,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11120489015064323,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.504020100502514e-05,
+      "loss": 1.3776,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11130719455740556,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.503517587939698e-05,
+      "loss": 1.5821,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11140949896416788,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.503015075376886e-05,
+      "loss": 1.6204,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1115118033709302,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.502512562814071e-05,
+      "loss": 1.454,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11161410777769253,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.502010050251257e-05,
+      "loss": 1.3912,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11171641218445484,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.501507537688442e-05,
+      "loss": 1.5202,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11181871659121717,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.50100502512563e-05,
+      "loss": 1.6027,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11192102099797949,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.500502512562814e-05,
+      "loss": 1.446,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1120233254047418,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.5e-05,
+      "loss": 1.3225,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11212562981150413,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.499497487437187e-05,
+      "loss": 1.4946,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11222793421826645,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.498994974874373e-05,
+      "loss": 1.6362,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11233023862502878,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.498492462311558e-05,
+      "loss": 1.5194,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1124325430317911,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.497989949748744e-05,
+      "loss": 1.4037,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11253484743855341,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.49748743718593e-05,
+      "loss": 1.4331,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11263715184531574,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.496984924623116e-05,
+      "loss": 1.4914,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11273945625207805,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.496482412060303e-05,
+      "loss": 1.5384,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11284176065884038,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.495979899497487e-05,
+      "loss": 1.5169,
+      "step": 1103
+    },
+    {
+      "epoch": 0.1129440650656027,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.495477386934674e-05,
+      "loss": 1.4449,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11304636947236502,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.49497487437186e-05,
+      "loss": 1.3401,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11314867387912735,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.494472361809046e-05,
+      "loss": 1.4034,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11325097828588966,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.493969849246231e-05,
+      "loss": 1.4483,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11335328269265199,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.493467336683417e-05,
+      "loss": 1.4076,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1134555870994143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.492964824120603e-05,
+      "loss": 1.5632,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11355789150617662,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.49246231155779e-05,
+      "loss": 1.4001,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11366019591293895,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.491959798994974e-05,
+      "loss": 1.4381,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11376250031970127,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.491457286432162e-05,
+      "loss": 1.5039,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1138648047264636,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.490954773869347e-05,
+      "loss": 1.4125,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11396710913322591,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.490452261306533e-05,
+      "loss": 1.3993,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11406941353998823,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.489949748743719e-05,
+      "loss": 1.2866,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11417171794675056,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.489447236180905e-05,
+      "loss": 1.4639,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11427402235351287,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.48894472361809e-05,
+      "loss": 1.5235,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1143763267602752,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.488442211055278e-05,
+      "loss": 1.3859,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11447863116703752,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.487939698492463e-05,
+      "loss": 1.6651,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11458093557379984,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.487437185929649e-05,
+      "loss": 1.37,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11468323998056217,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.486934673366835e-05,
+      "loss": 1.2772,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11478554438732448,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.486432160804021e-05,
+      "loss": 1.3493,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11488784879408681,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.485929648241206e-05,
+      "loss": 1.33,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11499015320084913,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.485427135678392e-05,
+      "loss": 1.5194,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11509245760761144,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.484924623115578e-05,
+      "loss": 1.4868,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11519476201437377,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.484422110552765e-05,
+      "loss": 1.4331,
+      "step": 1126
+    },
+    {
+      "epoch": 0.11529706642113609,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.48391959798995e-05,
+      "loss": 1.3157,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11539937082789842,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.483417085427137e-05,
+      "loss": 1.2798,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11550167523466073,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.482914572864322e-05,
+      "loss": 1.5228,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11560397964142305,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.482412060301508e-05,
+      "loss": 1.4965,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11570628404818538,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.481909547738694e-05,
+      "loss": 1.5075,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1158085884549477,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.48140703517588e-05,
+      "loss": 1.4205,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11591089286171002,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.480904522613065e-05,
+      "loss": 1.4306,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11601319726847234,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.480402010050252e-05,
+      "loss": 1.3427,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11611550167523466,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.479899497487438e-05,
+      "loss": 1.3539,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11621780608199699,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.479396984924624e-05,
+      "loss": 1.4657,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1163201104887593,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.47889447236181e-05,
+      "loss": 1.4823,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11642241489552163,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.478391959798995e-05,
+      "loss": 1.5584,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11652471930228395,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.477889447236181e-05,
+      "loss": 1.4358,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11662702370904626,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.477386934673366e-05,
+      "loss": 1.2843,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11672932811580859,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.476884422110554e-05,
+      "loss": 1.3476,
+      "step": 1141
+    },
+    {
+      "epoch": 0.11683163252257091,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.476381909547739e-05,
+      "loss": 1.5734,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11693393692933324,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.475879396984925e-05,
+      "loss": 1.4396,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11703624133609555,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.475376884422111e-05,
+      "loss": 1.3774,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11713854574285787,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.474874371859297e-05,
+      "loss": 1.3264,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1172408501496202,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.474371859296482e-05,
+      "loss": 1.4975,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11734315455638251,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.47386934673367e-05,
+      "loss": 1.2728,
+      "step": 1147
+    },
+    {
+      "epoch": 0.11744545896314484,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.473366834170854e-05,
+      "loss": 1.5111,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11754776336990716,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.47286432160804e-05,
+      "loss": 1.3132,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11765006777666948,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.472361809045227e-05,
+      "loss": 1.4369,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1177523721834318,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.471859296482413e-05,
+      "loss": 1.2832,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11785467659019412,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.471356783919598e-05,
+      "loss": 1.4,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11795698099695644,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.470854271356784e-05,
+      "loss": 1.3983,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11805928540371877,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.47035175879397e-05,
+      "loss": 1.4439,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11816158981048108,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.469849246231156e-05,
+      "loss": 1.4838,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11826389421724341,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.469346733668341e-05,
+      "loss": 1.3073,
+      "step": 1156
+    },
+    {
+      "epoch": 0.11836619862400573,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.468844221105529e-05,
+      "loss": 1.5469,
+      "step": 1157
+    },
+    {
+      "epoch": 0.11846850303076804,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.468341708542714e-05,
+      "loss": 1.3623,
+      "step": 1158
+    },
+    {
+      "epoch": 0.11857080743753037,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.4678391959799e-05,
+      "loss": 1.5889,
+      "step": 1159
+    },
+    {
+      "epoch": 0.11867311184429269,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.467336683417086e-05,
+      "loss": 1.5929,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11877541625105502,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.466834170854272e-05,
+      "loss": 1.3558,
+      "step": 1161
+    },
+    {
+      "epoch": 0.11887772065781733,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.466331658291457e-05,
+      "loss": 1.5178,
+      "step": 1162
+    },
+    {
+      "epoch": 0.11898002506457965,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.465829145728645e-05,
+      "loss": 1.3711,
+      "step": 1163
+    },
+    {
+      "epoch": 0.11908232947134198,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.46532663316583e-05,
+      "loss": 1.4921,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1191846338781043,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.464824120603016e-05,
+      "loss": 1.4712,
+      "step": 1165
+    },
+    {
+      "epoch": 0.11928693828486663,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.464321608040202e-05,
+      "loss": 1.4254,
+      "step": 1166
+    },
+    {
+      "epoch": 0.11938924269162894,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.463819095477388e-05,
+      "loss": 1.4187,
+      "step": 1167
+    },
+    {
+      "epoch": 0.11949154709839126,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.463316582914573e-05,
+      "loss": 1.4991,
+      "step": 1168
+    },
+    {
+      "epoch": 0.11959385150515359,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.462814070351759e-05,
+      "loss": 1.4256,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1196961559119159,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.462311557788945e-05,
+      "loss": 1.4597,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11979846031867823,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.461809045226132e-05,
+      "loss": 1.4307,
+      "step": 1171
+    },
+    {
+      "epoch": 0.11990076472544055,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.461306532663317e-05,
+      "loss": 1.3689,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12000306913220286,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.460804020100503e-05,
+      "loss": 1.491,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1201053735389652,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.460301507537689e-05,
+      "loss": 1.3938,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12020767794572751,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.459798994974874e-05,
+      "loss": 1.5504,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12030998235248984,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.459296482412061e-05,
+      "loss": 1.4502,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12041228675925215,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.458793969849246e-05,
+      "loss": 1.4807,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12051459116601447,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.458291457286432e-05,
+      "loss": 1.6139,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1206168955727768,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.457788944723619e-05,
+      "loss": 1.5465,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12071919997953912,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.457286432160805e-05,
+      "loss": 1.4039,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12082150438630145,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.45678391959799e-05,
+      "loss": 1.2959,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12092380879306376,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.456281407035177e-05,
+      "loss": 1.4552,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12102611319982608,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.455778894472362e-05,
+      "loss": 1.4646,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1211284176065884,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.455276381909548e-05,
+      "loss": 1.4623,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12123072201335072,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.454773869346733e-05,
+      "loss": 1.2866,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12133302642011305,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.45427135678392e-05,
+      "loss": 1.3128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12143533082687537,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.453768844221105e-05,
+      "loss": 1.4771,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12153763523363768,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.453266331658292e-05,
+      "loss": 1.3215,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12163993964040001,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.452763819095478e-05,
+      "loss": 1.4485,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12174224404716233,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.452261306532664e-05,
+      "loss": 1.4667,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12184454845392466,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.451758793969849e-05,
+      "loss": 1.5411,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12194685286068697,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.451256281407037e-05,
+      "loss": 1.619,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12204915726744929,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.450753768844221e-05,
+      "loss": 1.5097,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12215146167421162,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.450251256281408e-05,
+      "loss": 1.3796,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12225376608097394,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.449748743718594e-05,
+      "loss": 1.5279,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12235607048773627,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.44924623115578e-05,
+      "loss": 1.5335,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12245837489449858,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.448743718592965e-05,
+      "loss": 1.4219,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1225606793012609,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.448241206030152e-05,
+      "loss": 1.447,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12266298370802323,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.447738693467337e-05,
+      "loss": 1.5136,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12276528811478554,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.447236180904523e-05,
+      "loss": 1.3942,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12286759252154787,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.446733668341708e-05,
+      "loss": 1.4906,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12296989692831019,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.446231155778896e-05,
+      "loss": 1.4453,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1230722013350725,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.445728643216081e-05,
+      "loss": 1.3323,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12317450574183483,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.445226130653267e-05,
+      "loss": 1.5045,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12327681014859715,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.444723618090453e-05,
+      "loss": 1.3194,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12337911455535948,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.444221105527638e-05,
+      "loss": 1.2611,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1234814189621218,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.443718592964824e-05,
+      "loss": 1.5023,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12358372336888411,
+      "grad_norm": 0.25,
+      "learning_rate": 9.44321608040201e-05,
+      "loss": 1.2581,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12368602777564644,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.442713567839197e-05,
+      "loss": 1.4364,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12378833218240876,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.442211055276381e-05,
+      "loss": 1.3114,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12389063658917109,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.441708542713569e-05,
+      "loss": 1.3654,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1239929409959334,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.441206030150754e-05,
+      "loss": 1.4166,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12409524540269572,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.44070351758794e-05,
+      "loss": 1.3545,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12419754980945805,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.440201005025126e-05,
+      "loss": 1.5287,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12429985421622036,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.439698492462312e-05,
+      "loss": 1.4725,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12440215862298268,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.439195979899497e-05,
+      "loss": 1.4171,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12450446302974501,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.438693467336683e-05,
+      "loss": 1.3828,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12460676743650732,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.43819095477387e-05,
+      "loss": 1.1724,
+      "step": 1218
+    },
+    {
+      "epoch": 0.12470907184326965,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.437688442211056e-05,
+      "loss": 1.2788,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12481137625003197,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.437185929648241e-05,
+      "loss": 1.5043,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12491368065679428,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.436683417085428e-05,
+      "loss": 1.5568,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1250159850635566,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.436180904522613e-05,
+      "loss": 1.5162,
+      "step": 1222
+    },
+    {
+      "epoch": 0.12511828947031894,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.4356783919598e-05,
+      "loss": 1.5519,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12522059387708126,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.435175879396986e-05,
+      "loss": 1.5123,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12532289828384358,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.434673366834172e-05,
+      "loss": 1.4902,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1254252026906059,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.434170854271357e-05,
+      "loss": 1.4239,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1255275070973682,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.433668341708544e-05,
+      "loss": 1.5619,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12562981150413055,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.433165829145729e-05,
+      "loss": 1.4035,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12573211591089287,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.432663316582915e-05,
+      "loss": 1.4164,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12583442031765518,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.432160804020101e-05,
+      "loss": 1.4488,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1259367247244175,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.431658291457288e-05,
+      "loss": 1.6297,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12603902913117981,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.431155778894472e-05,
+      "loss": 1.3954,
+      "step": 1232
+    },
+    {
+      "epoch": 0.12614133353794216,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.430653266331659e-05,
+      "loss": 1.323,
+      "step": 1233
+    },
+    {
+      "epoch": 0.12624363794470447,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.430150753768845e-05,
+      "loss": 1.3743,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1263459423514668,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.429648241206031e-05,
+      "loss": 1.3475,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1264482467582291,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.429145728643216e-05,
+      "loss": 1.2728,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12655055116499142,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.428643216080403e-05,
+      "loss": 1.4278,
+      "step": 1237
+    },
+    {
+      "epoch": 0.12665285557175376,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.428140703517588e-05,
+      "loss": 1.5154,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12675515997851608,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.427638190954775e-05,
+      "loss": 1.5851,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1268574643852784,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.427135678391961e-05,
+      "loss": 1.5177,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1269597687920407,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.426633165829146e-05,
+      "loss": 1.4688,
+      "step": 1241
+    },
+    {
+      "epoch": 0.12706207319880303,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.426130653266332e-05,
+      "loss": 1.2734,
+      "step": 1242
+    },
+    {
+      "epoch": 0.12716437760556537,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.425628140703518e-05,
+      "loss": 1.2615,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1272666820123277,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.425125628140704e-05,
+      "loss": 1.3607,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12736898641909,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.424623115577889e-05,
+      "loss": 1.3513,
+      "step": 1245
+    },
+    {
+      "epoch": 0.12747129082585232,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.424120603015077e-05,
+      "loss": 1.3118,
+      "step": 1246
+    },
+    {
+      "epoch": 0.12757359523261463,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.423618090452261e-05,
+      "loss": 1.3949,
+      "step": 1247
+    },
+    {
+      "epoch": 0.12767589963937698,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.423115577889448e-05,
+      "loss": 1.5961,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1277782040461393,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.422613065326632e-05,
+      "loss": 1.5579,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1278805084529016,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.42211055276382e-05,
+      "loss": 1.3761,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12798281285966392,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.421608040201005e-05,
+      "loss": 1.2558,
+      "step": 1251
+    },
+    {
+      "epoch": 0.12808511726642624,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.421105527638191e-05,
+      "loss": 1.433,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12818742167318858,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.420603015075377e-05,
+      "loss": 1.6088,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1282897260799509,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.420100502512564e-05,
+      "loss": 1.4799,
+      "step": 1254
+    },
+    {
+      "epoch": 0.12839203048671322,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.419597989949748e-05,
+      "loss": 1.3946,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12849433489347553,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.419095477386936e-05,
+      "loss": 1.5135,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12859663930023785,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.418592964824121e-05,
+      "loss": 1.588,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1286989437070002,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.418090452261307e-05,
+      "loss": 1.4108,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1288012481137625,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.417587939698493e-05,
+      "loss": 1.461,
+      "step": 1259
+    },
+    {
+      "epoch": 0.12890355252052482,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.41708542713568e-05,
+      "loss": 1.329,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12900585692728714,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.416582914572864e-05,
+      "loss": 1.4846,
+      "step": 1261
+    },
+    {
+      "epoch": 0.12910816133404945,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.41608040201005e-05,
+      "loss": 1.5973,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1292104657408118,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.415577889447237e-05,
+      "loss": 1.3399,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1293127701475741,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.415075376884423e-05,
+      "loss": 1.3194,
+      "step": 1264
+    },
+    {
+      "epoch": 0.12941507455433643,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.414572864321608e-05,
+      "loss": 1.434,
+      "step": 1265
+    },
+    {
+      "epoch": 0.12951737896109874,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.414070351758795e-05,
+      "loss": 1.4753,
+      "step": 1266
+    },
+    {
+      "epoch": 0.12961968336786106,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.41356783919598e-05,
+      "loss": 1.4143,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1297219877746234,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.413065326633166e-05,
+      "loss": 1.4266,
+      "step": 1268
+    },
+    {
+      "epoch": 0.12982429218138572,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.412562814070352e-05,
+      "loss": 1.4237,
+      "step": 1269
+    },
+    {
+      "epoch": 0.12992659658814804,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.412060301507539e-05,
+      "loss": 1.5392,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13002890099491035,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.411557788944724e-05,
+      "loss": 1.3114,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13013120540167267,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.41105527638191e-05,
+      "loss": 1.351,
+      "step": 1272
+    },
+    {
+      "epoch": 0.130233509808435,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.410552763819096e-05,
+      "loss": 1.3402,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13033581421519733,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.410050251256282e-05,
+      "loss": 1.4906,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13043811862195964,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.409547738693468e-05,
+      "loss": 1.358,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13054042302872196,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.409045226130653e-05,
+      "loss": 1.5117,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13064272743548427,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.40854271356784e-05,
+      "loss": 1.3279,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13074503184224662,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.408040201005026e-05,
+      "loss": 1.2891,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13084733624900893,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.407537688442212e-05,
+      "loss": 1.418,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13094964065577125,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.407035175879397e-05,
+      "loss": 1.4411,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13105194506253356,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.406532663316583e-05,
+      "loss": 1.3356,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13115424946929588,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.406030150753769e-05,
+      "loss": 1.3191,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13125655387605822,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.405527638190955e-05,
+      "loss": 1.2795,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13135885828282054,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.40502512562814e-05,
+      "loss": 1.353,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13146116268958286,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.404522613065328e-05,
+      "loss": 1.36,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13156346709634517,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.404020100502513e-05,
+      "loss": 1.4578,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1316657715031075,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.403517587939699e-05,
+      "loss": 1.5062,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13176807590986983,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.403015075376885e-05,
+      "loss": 1.3796,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13187038031663215,
+      "grad_norm": 0.25,
+      "learning_rate": 9.402512562814071e-05,
+      "loss": 1.4375,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13197268472339446,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.402010050251256e-05,
+      "loss": 1.4054,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13207498913015678,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.401507537688444e-05,
+      "loss": 1.2487,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1321772935369191,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.401005025125628e-05,
+      "loss": 1.5881,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13227959794368144,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.400502512562815e-05,
+      "loss": 1.3529,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13238190235044375,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.4e-05,
+      "loss": 1.3694,
+      "step": 1294
+    },
+    {
+      "epoch": 0.13248420675720607,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.399497487437187e-05,
+      "loss": 1.3831,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13258651116396838,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.398994974874372e-05,
+      "loss": 1.4464,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1326888155707307,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.398492462311558e-05,
+      "loss": 1.4684,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13279111997749304,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.397989949748744e-05,
+      "loss": 1.3019,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13289342438425536,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.39748743718593e-05,
+      "loss": 1.4693,
+      "step": 1299
+    },
+    {
+      "epoch": 0.13299572879101768,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.396984924623115e-05,
+      "loss": 1.2702,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13309803319778,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.396482412060303e-05,
+      "loss": 1.4529,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1332003376045423,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.395979899497488e-05,
+      "loss": 1.2821,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13330264201130465,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.395477386934674e-05,
+      "loss": 1.3579,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13340494641806697,
+      "grad_norm": 0.25,
+      "learning_rate": 9.39497487437186e-05,
+      "loss": 1.4707,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13350725082482928,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.394472361809046e-05,
+      "loss": 1.3388,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1336095552315916,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.393969849246231e-05,
+      "loss": 1.3983,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1337118596383539,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.393467336683417e-05,
+      "loss": 1.5177,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13381416404511623,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.392964824120604e-05,
+      "loss": 1.3884,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13391646845187857,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.39246231155779e-05,
+      "loss": 1.3248,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1340187728586409,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.391959798994975e-05,
+      "loss": 1.5042,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1341210772654032,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.391457286432161e-05,
+      "loss": 1.3493,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13422338167216552,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.390954773869347e-05,
+      "loss": 1.3112,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13432568607892784,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.390452261306533e-05,
+      "loss": 1.3269,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13442799048569018,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.38994974874372e-05,
+      "loss": 1.5885,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1345302948924525,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.389447236180904e-05,
+      "loss": 1.2694,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1346325992992148,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.38894472361809e-05,
+      "loss": 1.3555,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13473490370597713,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.388442211055277e-05,
+      "loss": 1.5271,
+      "step": 1317
+    },
+    {
+      "epoch": 0.13483720811273944,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.387939698492463e-05,
+      "loss": 1.6426,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1349395125195018,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.387437185929648e-05,
+      "loss": 1.2982,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1350418169262641,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.386934673366835e-05,
+      "loss": 2.4188,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13514412133302642,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.38643216080402e-05,
+      "loss": 1.2949,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13524642573978873,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.385929648241206e-05,
+      "loss": 1.517,
+      "step": 1322
+    },
+    {
+      "epoch": 0.13534873014655105,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.385427135678393e-05,
+      "loss": 1.51,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1354510345533134,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.384924623115579e-05,
+      "loss": 1.3594,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1355533389600757,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.384422110552764e-05,
+      "loss": 1.2651,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13565564336683802,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.38391959798995e-05,
+      "loss": 1.5951,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13575794777360034,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.383417085427136e-05,
+      "loss": 1.5322,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13586025218036266,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.382914572864322e-05,
+      "loss": 1.4998,
+      "step": 1328
+    },
+    {
+      "epoch": 0.135962556587125,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.382412060301507e-05,
+      "loss": 1.3811,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13606486099388732,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.381909547738695e-05,
+      "loss": 1.5584,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13616716540064963,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.38140703517588e-05,
+      "loss": 1.3378,
+      "step": 1331
+    },
+    {
+      "epoch": 0.13626946980741195,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.380904522613066e-05,
+      "loss": 1.3065,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13637177421417426,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.380402010050252e-05,
+      "loss": 1.3047,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1364740786209366,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.379899497487438e-05,
+      "loss": 1.4696,
+      "step": 1334
+    },
+    {
+      "epoch": 0.13657638302769892,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.379396984924623e-05,
+      "loss": 1.3529,
+      "step": 1335
+    },
+    {
+      "epoch": 0.13667868743446124,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.37889447236181e-05,
+      "loss": 1.1896,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13678099184122355,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.378391959798995e-05,
+      "loss": 1.3217,
+      "step": 1337
+    },
+    {
+      "epoch": 0.13688329624798587,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.377889447236182e-05,
+      "loss": 1.4287,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1369856006547482,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.377386934673368e-05,
+      "loss": 1.3681,
+      "step": 1339
+    },
+    {
+      "epoch": 0.13708790506151053,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.376884422110554e-05,
+      "loss": 1.4749,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13719020946827284,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.376381909547739e-05,
+      "loss": 1.4615,
+      "step": 1341
+    },
+    {
+      "epoch": 0.13729251387503516,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.375879396984925e-05,
+      "loss": 1.2923,
+      "step": 1342
+    },
+    {
+      "epoch": 0.13739481828179748,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.375376884422111e-05,
+      "loss": 1.3056,
+      "step": 1343
+    },
+    {
+      "epoch": 0.13749712268855982,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.374874371859297e-05,
+      "loss": 1.4948,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13759942709532214,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.374371859296482e-05,
+      "loss": 1.3522,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13770173150208445,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.373869346733668e-05,
+      "loss": 1.4335,
+      "step": 1346
+    },
+    {
+      "epoch": 0.13780403590884677,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.373366834170855e-05,
+      "loss": 1.3943,
+      "step": 1347
+    },
+    {
+      "epoch": 0.13790634031560908,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.372864321608041e-05,
+      "loss": 1.316,
+      "step": 1348
+    },
+    {
+      "epoch": 0.13800864472237143,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.372361809045227e-05,
+      "loss": 1.4173,
+      "step": 1349
+    },
+    {
+      "epoch": 0.13811094912913374,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.371859296482412e-05,
+      "loss": 1.4768,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13821325353589606,
+      "grad_norm": 0.25,
+      "learning_rate": 9.371356783919598e-05,
+      "loss": 1.3711,
+      "step": 1351
+    },
+    {
+      "epoch": 0.13831555794265837,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.370854271356784e-05,
+      "loss": 1.5043,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1384178623494207,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.37035175879397e-05,
+      "loss": 1.4612,
+      "step": 1353
+    },
+    {
+      "epoch": 0.13852016675618303,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.369849246231155e-05,
+      "loss": 1.4977,
+      "step": 1354
+    },
+    {
+      "epoch": 0.13862247116294535,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.369346733668343e-05,
+      "loss": 1.444,
+      "step": 1355
+    },
+    {
+      "epoch": 0.13872477556970766,
+      "grad_norm": 0.25,
+      "learning_rate": 9.368844221105528e-05,
+      "loss": 1.3501,
+      "step": 1356
+    },
+    {
+      "epoch": 0.13882707997646998,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.368341708542714e-05,
+      "loss": 1.5193,
+      "step": 1357
+    },
+    {
+      "epoch": 0.1389293843832323,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.367839195979899e-05,
+      "loss": 1.3497,
+      "step": 1358
+    },
+    {
+      "epoch": 0.13903168878999464,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.367336683417086e-05,
+      "loss": 1.393,
+      "step": 1359
+    },
+    {
+      "epoch": 0.13913399319675696,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.366834170854271e-05,
+      "loss": 1.5311,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13923629760351927,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.366331658291457e-05,
+      "loss": 1.4558,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1393386020102816,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.365829145728644e-05,
+      "loss": 1.4907,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1394409064170439,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.36532663316583e-05,
+      "loss": 1.512,
+      "step": 1363
+    },
+    {
+      "epoch": 0.13954321082380625,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.364824120603015e-05,
+      "loss": 1.3416,
+      "step": 1364
+    },
+    {
+      "epoch": 0.13964551523056856,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.364321608040202e-05,
+      "loss": 1.1441,
+      "step": 1365
+    },
+    {
+      "epoch": 0.13974781963733088,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.363819095477387e-05,
+      "loss": 1.358,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1398501240440932,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.363316582914573e-05,
+      "loss": 1.504,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1399524284508555,
+      "grad_norm": 0.25,
+      "learning_rate": 9.36281407035176e-05,
+      "loss": 1.4762,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14005473285761785,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.362311557788946e-05,
+      "loss": 1.5369,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14015703726438017,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.36180904522613e-05,
+      "loss": 1.4866,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14025934167114248,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.361306532663318e-05,
+      "loss": 1.439,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1403616460779048,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.360804020100503e-05,
+      "loss": 1.3356,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14046395048466712,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.360301507537689e-05,
+      "loss": 1.4822,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14056625489142946,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.359798994974874e-05,
+      "loss": 1.3058,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14066855929819178,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.359296482412062e-05,
+      "loss": 1.4233,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1407708637049541,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.358793969849246e-05,
+      "loss": 1.3735,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1408731681117164,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.358291457286433e-05,
+      "loss": 1.5555,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14097547251847872,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.357788944723619e-05,
+      "loss": 1.466,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14107777692524107,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.357286432160805e-05,
+      "loss": 1.2837,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14118008133200338,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.35678391959799e-05,
+      "loss": 1.2846,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1412823857387657,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.356281407035176e-05,
+      "loss": 1.3052,
+      "step": 1381
+    },
+    {
+      "epoch": 0.141384690145528,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.355778894472362e-05,
+      "loss": 1.4061,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14148699455229033,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.355276381909549e-05,
+      "loss": 1.4703,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14158929895905267,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.354773869346735e-05,
+      "loss": 1.3716,
+      "step": 1384
+    },
+    {
+      "epoch": 0.141691603365815,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.35427135678392e-05,
+      "loss": 1.4987,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1417939077725773,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.353768844221106e-05,
+      "loss": 1.4073,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14189621217933962,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.353266331658292e-05,
+      "loss": 1.3411,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14199851658610194,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.352763819095478e-05,
+      "loss": 1.5503,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14210082099286428,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.352261306532663e-05,
+      "loss": 1.4155,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1422031253996266,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.351758793969849e-05,
+      "loss": 1.3624,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1423054298063889,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.351256281407035e-05,
+      "loss": 1.3099,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14240773421315123,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.350753768844222e-05,
+      "loss": 1.4846,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14251003861991354,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.350251256281406e-05,
+      "loss": 1.4683,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1426123430266759,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.349748743718594e-05,
+      "loss": 1.5472,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1427146474334382,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.349246231155779e-05,
+      "loss": 1.1725,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14281695184020052,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.348743718592965e-05,
+      "loss": 1.3884,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14291925624696283,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.348241206030151e-05,
+      "loss": 1.528,
+      "step": 1397
+    },
+    {
+      "epoch": 0.14302156065372515,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.347738693467338e-05,
+      "loss": 1.6301,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1431238650604875,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.347236180904522e-05,
+      "loss": 1.2984,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1432261694672498,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.34673366834171e-05,
+      "loss": 1.6307,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14332847387401212,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.346231155778895e-05,
+      "loss": 1.5485,
+      "step": 1401
+    },
+    {
+      "epoch": 0.14343077828077444,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.345728643216081e-05,
+      "loss": 1.5383,
+      "step": 1402
+    },
+    {
+      "epoch": 0.14353308268753676,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.345226130653266e-05,
+      "loss": 1.5429,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1436353870942991,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.344723618090453e-05,
+      "loss": 1.6136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14373769150106142,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.344221105527638e-05,
+      "loss": 1.4403,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14383999590782373,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.343718592964824e-05,
+      "loss": 1.3496,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14394230031458605,
+      "grad_norm": 0.25,
+      "learning_rate": 9.34321608040201e-05,
+      "loss": 1.2438,
+      "step": 1407
+    },
+    {
+      "epoch": 0.14404460472134836,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.342713567839197e-05,
+      "loss": 1.3352,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1441469091281107,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.342211055276382e-05,
+      "loss": 1.42,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14424921353487302,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.341708542713569e-05,
+      "loss": 1.4237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14435151794163534,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.341206030150754e-05,
+      "loss": 1.3476,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14445382234839765,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.34070351758794e-05,
+      "loss": 1.5361,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14455612675515997,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.340201005025126e-05,
+      "loss": 1.334,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1446584311619223,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.339698492462313e-05,
+      "loss": 1.3794,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14476073556868463,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.339195979899498e-05,
+      "loss": 1.5126,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14486303997544694,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.338693467336684e-05,
+      "loss": 1.4047,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14496534438220926,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.33819095477387e-05,
+      "loss": 1.3488,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14506764878897158,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.337688442211056e-05,
+      "loss": 1.5726,
+      "step": 1418
+    },
+    {
+      "epoch": 0.14516995319573392,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.337185929648241e-05,
+      "loss": 1.2858,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14527225760249624,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.336683417085427e-05,
+      "loss": 1.4068,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14537456200925855,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.336180904522613e-05,
+      "loss": 1.3946,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14547686641602087,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.3356783919598e-05,
+      "loss": 1.5393,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14557917082278318,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.335175879396986e-05,
+      "loss": 1.461,
+      "step": 1423
+    },
+    {
+      "epoch": 0.14568147522954553,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.33467336683417e-05,
+      "loss": 1.3556,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14578377963630784,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.334170854271357e-05,
+      "loss": 1.3817,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14588608404307016,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.333668341708543e-05,
+      "loss": 1.2861,
+      "step": 1426
+    },
+    {
+      "epoch": 0.14598838844983247,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.333165829145729e-05,
+      "loss": 1.5393,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1460906928565948,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.332663316582914e-05,
+      "loss": 1.3029,
+      "step": 1428
+    },
+    {
+      "epoch": 0.14619299726335713,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.332160804020102e-05,
+      "loss": 1.5126,
+      "step": 1429
+    },
+    {
+      "epoch": 0.14629530167011945,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.331658291457287e-05,
+      "loss": 1.2109,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14639760607688176,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.331155778894473e-05,
+      "loss": 1.3133,
+      "step": 1431
+    },
+    {
+      "epoch": 0.14649991048364408,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.330653266331659e-05,
+      "loss": 1.6092,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1466022148904064,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.330150753768845e-05,
+      "loss": 1.4411,
+      "step": 1433
+    },
+    {
+      "epoch": 0.1467045192971687,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.32964824120603e-05,
+      "loss": 1.3302,
+      "step": 1434
+    },
+    {
+      "epoch": 0.14680682370393106,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.329145728643216e-05,
+      "loss": 1.3564,
+      "step": 1435
+    },
+    {
+      "epoch": 0.14690912811069337,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.328643216080402e-05,
+      "loss": 1.4569,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1470114325174557,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.328140703517589e-05,
+      "loss": 1.3947,
+      "step": 1437
+    },
+    {
+      "epoch": 0.147113736924218,
+      "grad_norm": 0.25,
+      "learning_rate": 9.327638190954773e-05,
+      "loss": 1.3733,
+      "step": 1438
+    },
+    {
+      "epoch": 0.14721604133098032,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.327135678391961e-05,
+      "loss": 1.2267,
+      "step": 1439
+    },
+    {
+      "epoch": 0.14731834573774266,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.326633165829146e-05,
+      "loss": 1.451,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14742065014450498,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.326130653266332e-05,
+      "loss": 1.4144,
+      "step": 1441
+    },
+    {
+      "epoch": 0.1475229545512673,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.325628140703518e-05,
+      "loss": 1.2605,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1476252589580296,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.325125628140704e-05,
+      "loss": 1.4911,
+      "step": 1443
+    },
+    {
+      "epoch": 0.14772756336479193,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.324623115577889e-05,
+      "loss": 1.4002,
+      "step": 1444
+    },
+    {
+      "epoch": 0.14782986777155427,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.324120603015077e-05,
+      "loss": 1.3487,
+      "step": 1445
+    },
+    {
+      "epoch": 0.14793217217831658,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.323618090452262e-05,
+      "loss": 1.547,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1480344765850789,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.323115577889448e-05,
+      "loss": 1.404,
+      "step": 1447
+    },
+    {
+      "epoch": 0.14813678099184122,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.322613065326634e-05,
+      "loss": 1.3566,
+      "step": 1448
+    },
+    {
+      "epoch": 0.14823908539860353,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.32211055276382e-05,
+      "loss": 1.3179,
+      "step": 1449
+    },
+    {
+      "epoch": 0.14834138980536588,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.321608040201005e-05,
+      "loss": 1.3029,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1484436942121282,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.321105527638191e-05,
+      "loss": 1.5071,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1485459986188905,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.320603015075378e-05,
+      "loss": 1.5372,
+      "step": 1452
+    },
+    {
+      "epoch": 0.14864830302565282,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.320100502512564e-05,
+      "loss": 1.4287,
+      "step": 1453
+    },
+    {
+      "epoch": 0.14875060743241514,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.319597989949749e-05,
+      "loss": 1.3861,
+      "step": 1454
+    },
+    {
+      "epoch": 0.14885291183917748,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.319095477386935e-05,
+      "loss": 1.3681,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1489552162459398,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.318592964824121e-05,
+      "loss": 1.4694,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1490575206527021,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.318090452261307e-05,
+      "loss": 1.4062,
+      "step": 1457
+    },
+    {
+      "epoch": 0.14915982505946443,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.317587939698493e-05,
+      "loss": 1.4852,
+      "step": 1458
+    },
+    {
+      "epoch": 0.14926212946622675,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.317085427135678e-05,
+      "loss": 1.4286,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1493644338729891,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.316582914572864e-05,
+      "loss": 1.3055,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1494667382797514,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.316080402010051e-05,
+      "loss": 1.5557,
+      "step": 1461
+    },
+    {
+      "epoch": 0.14956904268651372,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.315577889447237e-05,
+      "loss": 1.3931,
+      "step": 1462
+    },
+    {
+      "epoch": 0.14967134709327604,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.315075376884422e-05,
+      "loss": 1.4205,
+      "step": 1463
+    },
+    {
+      "epoch": 0.14977365150003835,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.314572864321609e-05,
+      "loss": 1.3247,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1498759559068007,
+      "grad_norm": 0.25,
+      "learning_rate": 9.314070351758794e-05,
+      "loss": 1.3802,
+      "step": 1465
+    },
+    {
+      "epoch": 0.149978260313563,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.31356783919598e-05,
+      "loss": 1.4823,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15008056472032533,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.313065326633165e-05,
+      "loss": 1.4942,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15018286912708764,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.312562814070353e-05,
+      "loss": 1.3978,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15028517353384996,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.312060301507538e-05,
+      "loss": 1.4395,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1503874779406123,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.311557788944724e-05,
+      "loss": 1.1709,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15048978234737462,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.31105527638191e-05,
+      "loss": 1.3122,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15059208675413693,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.310552763819096e-05,
+      "loss": 1.2706,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15069439116089925,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.310050251256281e-05,
+      "loss": 1.4891,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15079669556766157,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.309547738693469e-05,
+      "loss": 1.4163,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1508989999744239,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.309045226130653e-05,
+      "loss": 1.4613,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15100130438118622,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.30854271356784e-05,
+      "loss": 1.3649,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15110360878794854,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.308040201005026e-05,
+      "loss": 1.3,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15120591319471086,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.307537688442212e-05,
+      "loss": 1.2663,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15130821760147317,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.307035175879397e-05,
+      "loss": 1.35,
+      "step": 1479
+    },
+    {
+      "epoch": 0.15141052200823552,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.306532663316585e-05,
+      "loss": 1.4805,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15151282641499783,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.30603015075377e-05,
+      "loss": 1.311,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15161513082176015,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.305527638190956e-05,
+      "loss": 1.3754,
+      "step": 1482
+    },
+    {
+      "epoch": 0.15171743522852246,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.30502512562814e-05,
+      "loss": 1.551,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15181973963528478,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.304522613065328e-05,
+      "loss": 1.1387,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15192204404204712,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.304020100502513e-05,
+      "loss": 1.4253,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15202434844880944,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.303517587939699e-05,
+      "loss": 1.5667,
+      "step": 1486
+    },
+    {
+      "epoch": 0.15212665285557175,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.303015075376885e-05,
+      "loss": 1.5913,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15222895726233407,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.302512562814071e-05,
+      "loss": 1.4776,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15233126166909638,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.302010050251256e-05,
+      "loss": 1.578,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15243356607585873,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.301507537688442e-05,
+      "loss": 1.3873,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15253587048262104,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.301005025125629e-05,
+      "loss": 1.3191,
+      "step": 1491
+    },
+    {
+      "epoch": 0.15263817488938336,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.300502512562815e-05,
+      "loss": 1.4192,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15274047929614568,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.3095,
+      "step": 1493
+    },
+    {
+      "epoch": 0.152842783702908,
+      "grad_norm": 0.25,
+      "learning_rate": 9.299497487437186e-05,
+      "loss": 1.3828,
+      "step": 1494
+    },
+    {
+      "epoch": 0.15294508810967034,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.298994974874372e-05,
+      "loss": 1.2199,
+      "step": 1495
+    },
+    {
+      "epoch": 0.15304739251643265,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.298492462311558e-05,
+      "loss": 1.4219,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15314969692319497,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.297989949748745e-05,
+      "loss": 1.4654,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15325200132995728,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.29748743718593e-05,
+      "loss": 1.3656,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1533543057367196,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.296984924623116e-05,
+      "loss": 1.5602,
+      "step": 1499
+    },
+    {
+      "epoch": 0.15345661014348194,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.296482412060302e-05,
+      "loss": 1.3809,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.87471749677056e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}