diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11373 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "eval_steps": 500,
+  "global_step": 1620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.009259259259259259,
+      "grad_norm": 17.116676330566406,
+      "learning_rate": 2e-05,
+      "loss": 2.3189,
+      "step": 1
+    },
+    {
+      "epoch": 0.018518518518518517,
+      "grad_norm": 19.404932022094727,
+      "learning_rate": 4e-05,
+      "loss": 2.3402,
+      "step": 2
+    },
+    {
+      "epoch": 0.027777777777777776,
+      "grad_norm": 15.749666213989258,
+      "learning_rate": 6e-05,
+      "loss": 1.8406,
+      "step": 3
+    },
+    {
+      "epoch": 0.037037037037037035,
+      "grad_norm": 14.914640426635742,
+      "learning_rate": 8e-05,
+      "loss": 2.0528,
+      "step": 4
+    },
+    {
+      "epoch": 0.046296296296296294,
+      "grad_norm": 11.358346939086914,
+      "learning_rate": 0.0001,
+      "loss": 1.1907,
+      "step": 5
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 13.200481414794922,
+      "learning_rate": 0.00012,
+      "loss": 1.1732,
+      "step": 6
+    },
+    {
+      "epoch": 0.06481481481481481,
+      "grad_norm": 11.002130508422852,
+      "learning_rate": 0.00014,
+      "loss": 0.6805,
+      "step": 7
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 13.887001991271973,
+      "learning_rate": 0.00016,
+      "loss": 0.6608,
+      "step": 8
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 8.605554580688477,
+      "learning_rate": 0.00018,
+      "loss": 0.6947,
+      "step": 9
+    },
+    {
+      "epoch": 0.09259259259259259,
+      "grad_norm": 10.935001373291016,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 10
+    },
+    {
+      "epoch": 0.10185185185185185,
+      "grad_norm": 10.06436538696289,
+      "learning_rate": 0.00019987577639751553,
+      "loss": 0.9718,
+      "step": 11
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 11.505407333374023,
+      "learning_rate": 0.00019975155279503107,
+      "loss": 0.7152,
+      "step": 12
+    },
+    {
+      "epoch": 0.12037037037037036,
+      "grad_norm": 5.3426899909973145,
+      "learning_rate": 0.00019962732919254659,
+      "loss": 0.2549,
+      "step": 13
+    },
+    {
+      "epoch": 0.12962962962962962,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00019962732919254659,
+      "loss": 1.1325,
+      "step": 14
+    },
+    {
+      "epoch": 0.1388888888888889,
+      "grad_norm": 10.246034622192383,
+      "learning_rate": 0.0001995031055900621,
+      "loss": 1.418,
+      "step": 15
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 10.24806022644043,
+      "learning_rate": 0.00019937888198757767,
+      "loss": 0.6118,
+      "step": 16
+    },
+    {
+      "epoch": 0.1574074074074074,
+      "grad_norm": 13.599936485290527,
+      "learning_rate": 0.0001992546583850932,
+      "loss": 0.8439,
+      "step": 17
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 8.4192476272583,
+      "learning_rate": 0.0001991304347826087,
+      "loss": 0.7078,
+      "step": 18
+    },
+    {
+      "epoch": 0.17592592592592593,
+      "grad_norm": 7.700663089752197,
+      "learning_rate": 0.00019900621118012425,
+      "loss": 0.6345,
+      "step": 19
+    },
+    {
+      "epoch": 0.18518518518518517,
+      "grad_norm": 9.339193344116211,
+      "learning_rate": 0.00019888198757763977,
+      "loss": 0.4559,
+      "step": 20
+    },
+    {
+      "epoch": 0.19444444444444445,
+      "grad_norm": 16.562761306762695,
+      "learning_rate": 0.00019875776397515528,
+      "loss": 1.4581,
+      "step": 21
+    },
+    {
+      "epoch": 0.2037037037037037,
+      "grad_norm": 13.53317928314209,
+      "learning_rate": 0.00019863354037267082,
+      "loss": 1.3443,
+      "step": 22
+    },
+    {
+      "epoch": 0.21296296296296297,
+      "grad_norm": 9.449267387390137,
+      "learning_rate": 0.00019850931677018634,
+      "loss": 1.0892,
+      "step": 23
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 19.80406951904297,
+      "learning_rate": 0.00019838509316770186,
+      "loss": 1.1021,
+      "step": 24
+    },
+    {
+      "epoch": 0.23148148148148148,
+      "grad_norm": 9.569581985473633,
+      "learning_rate": 0.0001982608695652174,
+      "loss": 0.8147,
+      "step": 25
+    },
+    {
+      "epoch": 0.24074074074074073,
+      "grad_norm": 6.435489654541016,
+      "learning_rate": 0.00019813664596273294,
+      "loss": 0.4822,
+      "step": 26
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.351337432861328,
+      "learning_rate": 0.00019801242236024846,
+      "loss": 1.7875,
+      "step": 27
+    },
+    {
+      "epoch": 0.25925925925925924,
+      "grad_norm": 6.918874263763428,
+      "learning_rate": 0.00019788819875776398,
+      "loss": 1.0019,
+      "step": 28
+    },
+    {
+      "epoch": 0.26851851851851855,
+      "grad_norm": 7.385853290557861,
+      "learning_rate": 0.00019776397515527952,
+      "loss": 0.9338,
+      "step": 29
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 8.40998363494873,
+      "learning_rate": 0.00019763975155279504,
+      "loss": 0.4633,
+      "step": 30
+    },
+    {
+      "epoch": 0.28703703703703703,
+      "grad_norm": 12.839616775512695,
+      "learning_rate": 0.00019751552795031055,
+      "loss": 0.8168,
+      "step": 31
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 9.741667747497559,
+      "learning_rate": 0.0001973913043478261,
+      "loss": 0.7139,
+      "step": 32
+    },
+    {
+      "epoch": 0.3055555555555556,
+      "grad_norm": 10.23273754119873,
+      "learning_rate": 0.0001972670807453416,
+      "loss": 0.5477,
+      "step": 33
+    },
+    {
+      "epoch": 0.3148148148148148,
+      "grad_norm": 10.880776405334473,
+      "learning_rate": 0.00019714285714285716,
+      "loss": 1.6291,
+      "step": 34
+    },
+    {
+      "epoch": 0.32407407407407407,
+      "grad_norm": 11.059282302856445,
+      "learning_rate": 0.0001970186335403727,
+      "loss": 1.1802,
+      "step": 35
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 8.465683937072754,
+      "learning_rate": 0.00019689440993788822,
+      "loss": 1.5225,
+      "step": 36
+    },
+    {
+      "epoch": 0.3425925925925926,
+      "grad_norm": 8.37371826171875,
+      "learning_rate": 0.00019677018633540373,
+      "loss": 0.5846,
+      "step": 37
+    },
+    {
+      "epoch": 0.35185185185185186,
+      "grad_norm": 5.269484043121338,
+      "learning_rate": 0.00019664596273291928,
+      "loss": 0.3792,
+      "step": 38
+    },
+    {
+      "epoch": 0.3611111111111111,
+      "grad_norm": 5.26035213470459,
+      "learning_rate": 0.0001965217391304348,
+      "loss": 0.4575,
+      "step": 39
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 7.327594757080078,
+      "learning_rate": 0.0001963975155279503,
+      "loss": 0.6059,
+      "step": 40
+    },
+    {
+      "epoch": 0.37962962962962965,
+      "grad_norm": 10.461058616638184,
+      "learning_rate": 0.00019627329192546585,
+      "loss": 0.8153,
+      "step": 41
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 10.147456169128418,
+      "learning_rate": 0.00019614906832298137,
+      "loss": 0.4933,
+      "step": 42
+    },
+    {
+      "epoch": 0.39814814814814814,
+      "grad_norm": 6.203906536102295,
+      "learning_rate": 0.0001960248447204969,
+      "loss": 0.4631,
+      "step": 43
+    },
+    {
+      "epoch": 0.4074074074074074,
+      "grad_norm": 10.029688835144043,
+      "learning_rate": 0.00019590062111801243,
+      "loss": 1.1486,
+      "step": 44
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 8.837681770324707,
+      "learning_rate": 0.00019577639751552797,
+      "loss": 1.4024,
+      "step": 45
+    },
+    {
+      "epoch": 0.42592592592592593,
+      "grad_norm": 8.74865436553955,
+      "learning_rate": 0.0001956521739130435,
+      "loss": 1.5168,
+      "step": 46
+    },
+    {
+      "epoch": 0.4351851851851852,
+      "grad_norm": 5.9198479652404785,
+      "learning_rate": 0.000195527950310559,
+      "loss": 0.4223,
+      "step": 47
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 3.298370122909546,
+      "learning_rate": 0.00019540372670807455,
+      "loss": 0.1694,
+      "step": 48
+    },
+    {
+      "epoch": 0.4537037037037037,
+      "grad_norm": 8.61251163482666,
+      "learning_rate": 0.00019527950310559006,
+      "loss": 1.2626,
+      "step": 49
+    },
+    {
+      "epoch": 0.46296296296296297,
+      "grad_norm": 9.859038352966309,
+      "learning_rate": 0.00019515527950310558,
+      "loss": 1.4776,
+      "step": 50
+    },
+    {
+      "epoch": 0.4722222222222222,
+      "grad_norm": 10.356538772583008,
+      "learning_rate": 0.00019503105590062112,
+      "loss": 1.2718,
+      "step": 51
+    },
+    {
+      "epoch": 0.48148148148148145,
+      "grad_norm": 5.658873081207275,
+      "learning_rate": 0.00019490683229813667,
+      "loss": 0.1858,
+      "step": 52
+    },
+    {
+      "epoch": 0.49074074074074076,
+      "grad_norm": 10.131982803344727,
+      "learning_rate": 0.00019478260869565218,
+      "loss": 0.6037,
+      "step": 53
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.389909744262695,
+      "learning_rate": 0.00019465838509316773,
+      "loss": 1.4944,
+      "step": 54
+    },
+    {
+      "epoch": 0.5092592592592593,
+      "grad_norm": 12.465958595275879,
+      "learning_rate": 0.00019453416149068324,
+      "loss": 2.2114,
+      "step": 55
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 4.682929039001465,
+      "learning_rate": 0.00019440993788819876,
+      "loss": 0.1766,
+      "step": 56
+    },
+    {
+      "epoch": 0.5277777777777778,
+      "grad_norm": 6.179277420043945,
+      "learning_rate": 0.0001942857142857143,
+      "loss": 0.4002,
+      "step": 57
+    },
+    {
+      "epoch": 0.5370370370370371,
+      "grad_norm": 6.188724994659424,
+      "learning_rate": 0.00019416149068322982,
+      "loss": 0.4758,
+      "step": 58
+    },
+    {
+      "epoch": 0.5462962962962963,
+      "grad_norm": 8.731796264648438,
+      "learning_rate": 0.00019403726708074534,
+      "loss": 1.4991,
+      "step": 59
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 8.394841194152832,
+      "learning_rate": 0.00019391304347826088,
+      "loss": 0.506,
+      "step": 60
+    },
+    {
+      "epoch": 0.5648148148148148,
+      "grad_norm": 13.798285484313965,
+      "learning_rate": 0.00019378881987577642,
+      "loss": 2.0693,
+      "step": 61
+    },
+    {
+      "epoch": 0.5740740740740741,
+      "grad_norm": 6.169430255889893,
+      "learning_rate": 0.00019366459627329194,
+      "loss": 0.4175,
+      "step": 62
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 7.522382736206055,
+      "learning_rate": 0.00019354037267080746,
+      "loss": 0.6897,
+      "step": 63
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 5.0495219230651855,
+      "learning_rate": 0.000193416149068323,
+      "loss": 0.2197,
+      "step": 64
+    },
+    {
+      "epoch": 0.6018518518518519,
+      "grad_norm": 2.5212173461914062,
+      "learning_rate": 0.00019329192546583852,
+      "loss": 0.1039,
+      "step": 65
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 9.070035934448242,
+      "learning_rate": 0.00019316770186335403,
+      "loss": 1.0819,
+      "step": 66
+    },
+    {
+      "epoch": 0.6203703703703703,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00019316770186335403,
+      "loss": 0.7222,
+      "step": 67
+    },
+    {
+      "epoch": 0.6296296296296297,
+      "grad_norm": 8.20578384399414,
+      "learning_rate": 0.00019304347826086958,
+      "loss": 1.1301,
+      "step": 68
+    },
+    {
+      "epoch": 0.6388888888888888,
+      "grad_norm": 8.681023597717285,
+      "learning_rate": 0.0001929192546583851,
+      "loss": 1.3619,
+      "step": 69
+    },
+    {
+      "epoch": 0.6481481481481481,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001929192546583851,
+      "loss": 1.2046,
+      "step": 70
+    },
+    {
+      "epoch": 0.6574074074074074,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001929192546583851,
+      "loss": 0.6509,
+      "step": 71
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0001929192546583851,
+      "loss": 1.5734,
+      "step": 72
+    },
+    {
+      "epoch": 0.6759259259259259,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0001929192546583851,
+      "loss": 1.1999,
+      "step": 73
+    },
+    {
+      "epoch": 0.6851851851851852,
+      "grad_norm": 120.05274963378906,
+      "learning_rate": 0.0001927950310559006,
+      "loss": 1.0121,
+      "step": 74
+    },
+    {
+      "epoch": 0.6944444444444444,
+      "grad_norm": 6.013508319854736,
+      "learning_rate": 0.00019267080745341618,
+      "loss": 0.245,
+      "step": 75
+    },
+    {
+      "epoch": 0.7037037037037037,
+      "grad_norm": 10.709644317626953,
+      "learning_rate": 0.0001925465838509317,
+      "loss": 1.7426,
+      "step": 76
+    },
+    {
+      "epoch": 0.7129629629629629,
+      "grad_norm": 8.936174392700195,
+      "learning_rate": 0.0001924223602484472,
+      "loss": 2.0067,
+      "step": 77
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 10.918343544006348,
+      "learning_rate": 0.00019229813664596275,
+      "loss": 1.7653,
+      "step": 78
+    },
+    {
+      "epoch": 0.7314814814814815,
+      "grad_norm": 7.559531211853027,
+      "learning_rate": 0.00019217391304347827,
+      "loss": 0.7266,
+      "step": 79
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 5.118418216705322,
+      "learning_rate": 0.0001920496894409938,
+      "loss": 0.2286,
+      "step": 80
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.13199234008789,
+      "learning_rate": 0.00019192546583850933,
+      "loss": 1.4382,
+      "step": 81
+    },
+    {
+      "epoch": 0.7592592592592593,
+      "grad_norm": 5.355680465698242,
+      "learning_rate": 0.00019180124223602485,
+      "loss": 0.2843,
+      "step": 82
+    },
+    {
+      "epoch": 0.7685185185185185,
+      "grad_norm": 10.049469947814941,
+      "learning_rate": 0.00019167701863354036,
+      "loss": 0.9342,
+      "step": 83
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 5.9033918380737305,
+      "learning_rate": 0.0001915527950310559,
+      "loss": 0.3322,
+      "step": 84
+    },
+    {
+      "epoch": 0.7870370370370371,
+      "grad_norm": 11.78792953491211,
+      "learning_rate": 0.00019142857142857145,
+      "loss": 1.5358,
+      "step": 85
+    },
+    {
+      "epoch": 0.7962962962962963,
+      "grad_norm": 9.4061861038208,
+      "learning_rate": 0.00019130434782608697,
+      "loss": 1.7081,
+      "step": 86
+    },
+    {
+      "epoch": 0.8055555555555556,
+      "grad_norm": 9.609070777893066,
+      "learning_rate": 0.0001911801242236025,
+      "loss": 1.2752,
+      "step": 87
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 7.137119770050049,
+      "learning_rate": 0.00019105590062111803,
+      "loss": 0.6856,
+      "step": 88
+    },
+    {
+      "epoch": 0.8240740740740741,
+      "grad_norm": 8.402849197387695,
+      "learning_rate": 0.00019093167701863354,
+      "loss": 0.8608,
+      "step": 89
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 7.3328680992126465,
+      "learning_rate": 0.0001908074534161491,
+      "loss": 0.9065,
+      "step": 90
+    },
+    {
+      "epoch": 0.8425925925925926,
+      "grad_norm": 7.892049312591553,
+      "learning_rate": 0.0001906832298136646,
+      "loss": 0.7751,
+      "step": 91
+    },
+    {
+      "epoch": 0.8518518518518519,
+      "grad_norm": 7.179913520812988,
+      "learning_rate": 0.00019055900621118012,
+      "loss": 0.854,
+      "step": 92
+    },
+    {
+      "epoch": 0.8611111111111112,
+      "grad_norm": 8.245528221130371,
+      "learning_rate": 0.00019043478260869566,
+      "loss": 1.0387,
+      "step": 93
+    },
+    {
+      "epoch": 0.8703703703703703,
+      "grad_norm": 8.712199211120605,
+      "learning_rate": 0.0001903105590062112,
+      "loss": 1.2923,
+      "step": 94
+    },
+    {
+      "epoch": 0.8796296296296297,
+      "grad_norm": 4.641773223876953,
+      "learning_rate": 0.00019018633540372672,
+      "loss": 0.2925,
+      "step": 95
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 7.5324177742004395,
+      "learning_rate": 0.00019006211180124224,
+      "loss": 0.7769,
+      "step": 96
+    },
+    {
+      "epoch": 0.8981481481481481,
+      "grad_norm": 9.859864234924316,
+      "learning_rate": 0.00018993788819875778,
+      "loss": 1.7021,
+      "step": 97
+    },
+    {
+      "epoch": 0.9074074074074074,
+      "grad_norm": 9.314699172973633,
+      "learning_rate": 0.0001898136645962733,
+      "loss": 1.0719,
+      "step": 98
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 11.670292854309082,
+      "learning_rate": 0.00018968944099378881,
+      "loss": 1.6333,
+      "step": 99
+    },
+    {
+      "epoch": 0.9259259259259259,
+      "grad_norm": 12.171489715576172,
+      "learning_rate": 0.00018956521739130436,
+      "loss": 0.8229,
+      "step": 100
+    },
+    {
+      "epoch": 0.9351851851851852,
+      "grad_norm": 8.804001808166504,
+      "learning_rate": 0.00018944099378881987,
+      "loss": 1.3885,
+      "step": 101
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 8.818242073059082,
+      "learning_rate": 0.00018931677018633542,
+      "loss": 1.3888,
+      "step": 102
+    },
+    {
+      "epoch": 0.9537037037037037,
+      "grad_norm": 7.723133563995361,
+      "learning_rate": 0.00018919254658385096,
+      "loss": 0.8324,
+      "step": 103
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 8.167946815490723,
+      "learning_rate": 0.00018906832298136648,
+      "loss": 1.1976,
+      "step": 104
+    },
+    {
+      "epoch": 0.9722222222222222,
+      "grad_norm": 8.978958129882812,
+      "learning_rate": 0.000188944099378882,
+      "loss": 0.7289,
+      "step": 105
+    },
+    {
+      "epoch": 0.9814814814814815,
+      "grad_norm": 8.901488304138184,
+      "learning_rate": 0.00018881987577639754,
+      "loss": 1.2306,
+      "step": 106
+    },
+    {
+      "epoch": 0.9907407407407407,
+      "grad_norm": 5.3030877113342285,
+      "learning_rate": 0.00018869565217391305,
+      "loss": 0.2643,
+      "step": 107
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.44385814666748,
+      "learning_rate": 0.00018857142857142857,
+      "loss": 1.4146,
+      "step": 108
+    },
+    {
+      "epoch": 1.0092592592592593,
+      "grad_norm": 2.322072744369507,
+      "learning_rate": 0.00018844720496894411,
+      "loss": 0.0682,
+      "step": 109
+    },
+    {
+      "epoch": 1.0185185185185186,
+      "grad_norm": 9.836686134338379,
+      "learning_rate": 0.00018832298136645963,
+      "loss": 0.8827,
+      "step": 110
+    },
+    {
+      "epoch": 1.0277777777777777,
+      "grad_norm": 5.424779415130615,
+      "learning_rate": 0.00018819875776397517,
+      "loss": 0.385,
+      "step": 111
+    },
+    {
+      "epoch": 1.037037037037037,
+      "grad_norm": 8.20822525024414,
+      "learning_rate": 0.0001880745341614907,
+      "loss": 0.6324,
+      "step": 112
+    },
+    {
+      "epoch": 1.0462962962962963,
+      "grad_norm": 6.136359691619873,
+      "learning_rate": 0.00018795031055900623,
+      "loss": 0.3507,
+      "step": 113
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 12.00852108001709,
+      "learning_rate": 0.00018782608695652175,
+      "loss": 0.7757,
+      "step": 114
+    },
+    {
+      "epoch": 1.0648148148148149,
+      "grad_norm": 10.327400207519531,
+      "learning_rate": 0.00018770186335403727,
+      "loss": 0.6816,
+      "step": 115
+    },
+    {
+      "epoch": 1.074074074074074,
+      "grad_norm": 8.669729232788086,
+      "learning_rate": 0.0001875776397515528,
+      "loss": 0.5291,
+      "step": 116
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 7.052165985107422,
+      "learning_rate": 0.00018745341614906833,
+      "loss": 0.5359,
+      "step": 117
+    },
+    {
+      "epoch": 1.0925925925925926,
+      "grad_norm": 11.054783821105957,
+      "learning_rate": 0.00018732919254658384,
+      "loss": 0.8106,
+      "step": 118
+    },
+    {
+      "epoch": 1.1018518518518519,
+      "grad_norm": 10.095086097717285,
+      "learning_rate": 0.00018720496894409939,
+      "loss": 0.903,
+      "step": 119
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 23.136199951171875,
+      "learning_rate": 0.0001870807453416149,
+      "loss": 1.0593,
+      "step": 120
+    },
+    {
+      "epoch": 1.1203703703703705,
+      "grad_norm": 8.750686645507812,
+      "learning_rate": 0.00018695652173913045,
+      "loss": 0.2787,
+      "step": 121
+    },
+    {
+      "epoch": 1.1296296296296295,
+      "grad_norm": 7.291873931884766,
+      "learning_rate": 0.000186832298136646,
+      "loss": 0.5422,
+      "step": 122
+    },
+    {
+      "epoch": 1.1388888888888888,
+      "grad_norm": 9.136726379394531,
+      "learning_rate": 0.0001867080745341615,
+      "loss": 0.28,
+      "step": 123
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 7.925637245178223,
+      "learning_rate": 0.00018658385093167702,
+      "loss": 0.6995,
+      "step": 124
+    },
+    {
+      "epoch": 1.1574074074074074,
+      "grad_norm": 4.114215850830078,
+      "learning_rate": 0.00018645962732919257,
+      "loss": 0.1883,
+      "step": 125
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 8.64969539642334,
+      "learning_rate": 0.00018633540372670808,
+      "loss": 0.7698,
+      "step": 126
+    },
+    {
+      "epoch": 1.175925925925926,
+      "grad_norm": 10.038046836853027,
+      "learning_rate": 0.0001862111801242236,
+      "loss": 1.2307,
+      "step": 127
+    },
+    {
+      "epoch": 1.1851851851851851,
+      "grad_norm": 10.920683860778809,
+      "learning_rate": 0.00018608695652173914,
+      "loss": 1.0984,
+      "step": 128
+    },
+    {
+      "epoch": 1.1944444444444444,
+      "grad_norm": 5.161189556121826,
+      "learning_rate": 0.00018596273291925466,
+      "loss": 0.1952,
+      "step": 129
+    },
+    {
+      "epoch": 1.2037037037037037,
+      "grad_norm": 10.901082038879395,
+      "learning_rate": 0.0001858385093167702,
+      "loss": 0.5871,
+      "step": 130
+    },
+    {
+      "epoch": 1.212962962962963,
+      "grad_norm": 5.366225719451904,
+      "learning_rate": 0.00018571428571428572,
+      "loss": 0.1007,
+      "step": 131
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 7.6988959312438965,
+      "learning_rate": 0.00018559006211180126,
+      "loss": 0.6182,
+      "step": 132
+    },
+    {
+      "epoch": 1.2314814814814814,
+      "grad_norm": 5.929910659790039,
+      "learning_rate": 0.00018546583850931678,
+      "loss": 0.3714,
+      "step": 133
+    },
+    {
+      "epoch": 1.2407407407407407,
+      "grad_norm": 4.514689922332764,
+      "learning_rate": 0.0001853416149068323,
+      "loss": 0.1787,
+      "step": 134
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.674291610717773,
+      "learning_rate": 0.00018521739130434784,
+      "loss": 1.6911,
+      "step": 135
+    },
+    {
+      "epoch": 1.2592592592592593,
+      "grad_norm": 9.21278190612793,
+      "learning_rate": 0.00018509316770186335,
+      "loss": 0.7563,
+      "step": 136
+    },
+    {
+      "epoch": 1.2685185185185186,
+      "grad_norm": 8.567093849182129,
+      "learning_rate": 0.00018496894409937887,
+      "loss": 0.4734,
+      "step": 137
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 8.084855079650879,
+      "learning_rate": 0.0001848447204968944,
+      "loss": 1.0525,
+      "step": 138
+    },
+    {
+      "epoch": 1.287037037037037,
+      "grad_norm": 5.928450107574463,
+      "learning_rate": 0.00018472049689440996,
+      "loss": 0.5145,
+      "step": 139
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 7.130023956298828,
+      "learning_rate": 0.00018459627329192547,
+      "loss": 0.6849,
+      "step": 140
+    },
+    {
+      "epoch": 1.3055555555555556,
+      "grad_norm": 10.378190994262695,
+      "learning_rate": 0.00018447204968944102,
+      "loss": 0.986,
+      "step": 141
+    },
+    {
+      "epoch": 1.3148148148148149,
+      "grad_norm": 8.725743293762207,
+      "learning_rate": 0.00018434782608695653,
+      "loss": 0.8948,
+      "step": 142
+    },
+    {
+      "epoch": 1.324074074074074,
+      "grad_norm": 9.408001899719238,
+      "learning_rate": 0.00018422360248447205,
+      "loss": 1.6738,
+      "step": 143
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 8.074913024902344,
+      "learning_rate": 0.0001840993788819876,
+      "loss": 0.6298,
+      "step": 144
+    },
+    {
+      "epoch": 1.3425925925925926,
+      "grad_norm": 7.581278324127197,
+      "learning_rate": 0.0001839751552795031,
+      "loss": 0.3591,
+      "step": 145
+    },
+    {
+      "epoch": 1.3518518518518519,
+      "grad_norm": 6.366710186004639,
+      "learning_rate": 0.00018385093167701863,
+      "loss": 0.6695,
+      "step": 146
+    },
+    {
+      "epoch": 1.3611111111111112,
+      "grad_norm": 7.74171781539917,
+      "learning_rate": 0.00018372670807453417,
+      "loss": 0.4759,
+      "step": 147
+    },
+    {
+      "epoch": 1.3703703703703702,
+      "grad_norm": 7.979270935058594,
+      "learning_rate": 0.0001836024844720497,
+      "loss": 0.4445,
+      "step": 148
+    },
+    {
+      "epoch": 1.3796296296296298,
+      "grad_norm": 5.539051532745361,
+      "learning_rate": 0.00018347826086956523,
+      "loss": 0.2873,
+      "step": 149
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 8.086750030517578,
+      "learning_rate": 0.00018335403726708074,
+      "loss": 0.1622,
+      "step": 150
+    },
+    {
+      "epoch": 1.3981481481481481,
+      "grad_norm": 4.4308671951293945,
+      "learning_rate": 0.0001832298136645963,
+      "loss": 0.21,
+      "step": 151
+    },
+    {
+      "epoch": 1.4074074074074074,
+      "grad_norm": 8.82310676574707,
+      "learning_rate": 0.0001831055900621118,
+      "loss": 0.8503,
+      "step": 152
+    },
+    {
+      "epoch": 1.4166666666666667,
+      "grad_norm": 6.758204460144043,
+      "learning_rate": 0.00018298136645962735,
+      "loss": 0.3063,
+      "step": 153
+    },
+    {
+      "epoch": 1.425925925925926,
+      "grad_norm": 5.448921203613281,
+      "learning_rate": 0.00018285714285714286,
+      "loss": 0.2057,
+      "step": 154
+    },
+    {
+      "epoch": 1.4351851851851851,
+      "grad_norm": 5.913497447967529,
+      "learning_rate": 0.00018273291925465838,
+      "loss": 0.3591,
+      "step": 155
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 8.914529800415039,
+      "learning_rate": 0.00018260869565217392,
+      "loss": 0.8959,
+      "step": 156
+    },
+    {
+      "epoch": 1.4537037037037037,
+      "grad_norm": 10.08364486694336,
+      "learning_rate": 0.00018248447204968947,
+      "loss": 0.768,
+      "step": 157
+    },
+    {
+      "epoch": 1.462962962962963,
+      "grad_norm": 7.685210704803467,
+      "learning_rate": 0.00018236024844720498,
+      "loss": 0.8345,
+      "step": 158
+    },
+    {
+      "epoch": 1.4722222222222223,
+      "grad_norm": 4.343817710876465,
+      "learning_rate": 0.0001822360248447205,
+      "loss": 0.2126,
+      "step": 159
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 3.930588483810425,
+      "learning_rate": 0.00018211180124223604,
+      "loss": 0.1506,
+      "step": 160
+    },
+    {
+      "epoch": 1.4907407407407407,
+      "grad_norm": 6.4112958908081055,
+      "learning_rate": 0.00018198757763975156,
+      "loss": 0.2868,
+      "step": 161
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.307744026184082,
+      "learning_rate": 0.00018186335403726708,
+      "loss": 1.069,
+      "step": 162
+    },
+    {
+      "epoch": 1.5092592592592593,
+      "grad_norm": 11.457921028137207,
+      "learning_rate": 0.00018173913043478262,
+      "loss": 0.6819,
+      "step": 163
+    },
+    {
+      "epoch": 1.5185185185185186,
+      "grad_norm": 4.507647514343262,
+      "learning_rate": 0.00018161490683229814,
+      "loss": 0.292,
+      "step": 164
+    },
+    {
+      "epoch": 1.5277777777777777,
+      "grad_norm": 5.411582946777344,
+      "learning_rate": 0.00018149068322981365,
+      "loss": 0.2228,
+      "step": 165
+    },
+    {
+      "epoch": 1.5370370370370372,
+      "grad_norm": 9.850244522094727,
+      "learning_rate": 0.00018136645962732922,
+      "loss": 0.9859,
+      "step": 166
+    },
+    {
+      "epoch": 1.5462962962962963,
+      "grad_norm": 8.918268203735352,
+      "learning_rate": 0.00018124223602484474,
+      "loss": 0.7327,
+      "step": 167
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 13.585978507995605,
+      "learning_rate": 0.00018111801242236026,
+      "loss": 1.3928,
+      "step": 168
+    },
+    {
+      "epoch": 1.5648148148148149,
+      "grad_norm": 8.466153144836426,
+      "learning_rate": 0.0001809937888198758,
+      "loss": 0.8465,
+      "step": 169
+    },
+    {
+      "epoch": 1.574074074074074,
+      "grad_norm": 7.285000324249268,
+      "learning_rate": 0.00018086956521739132,
+      "loss": 0.5777,
+      "step": 170
+    },
+    {
+      "epoch": 1.5833333333333335,
+      "grad_norm": 3.1247687339782715,
+      "learning_rate": 0.00018074534161490683,
+      "loss": 0.1439,
+      "step": 171
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 7.981088161468506,
+      "learning_rate": 0.00018062111801242238,
+      "loss": 0.5934,
+      "step": 172
+    },
+    {
+      "epoch": 1.6018518518518519,
+      "grad_norm": 5.762646198272705,
+      "learning_rate": 0.0001804968944099379,
+      "loss": 0.1859,
+      "step": 173
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 3.394270181655884,
+      "learning_rate": 0.0001803726708074534,
+      "loss": 0.093,
+      "step": 174
+    },
+    {
+      "epoch": 1.6203703703703702,
+      "grad_norm": 8.87936019897461,
+      "learning_rate": 0.00018024844720496895,
+      "loss": 1.2075,
+      "step": 175
+    },
+    {
+      "epoch": 1.6296296296296298,
+      "grad_norm": 6.698455810546875,
+      "learning_rate": 0.0001801242236024845,
+      "loss": 0.2976,
+      "step": 176
+    },
+    {
+      "epoch": 1.6388888888888888,
+      "grad_norm": 6.693928241729736,
+      "learning_rate": 0.00018,
+      "loss": 0.478,
+      "step": 177
+    },
+    {
+      "epoch": 1.6481481481481481,
+      "grad_norm": 9.64937686920166,
+      "learning_rate": 0.00017987577639751553,
+      "loss": 1.1285,
+      "step": 178
+    },
+    {
+      "epoch": 1.6574074074074074,
+      "grad_norm": 7.241469860076904,
+      "learning_rate": 0.00017975155279503107,
+      "loss": 0.5239,
+      "step": 179
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 7.1578545570373535,
+      "learning_rate": 0.0001796273291925466,
+      "loss": 0.7933,
+      "step": 180
+    },
+    {
+      "epoch": 1.675925925925926,
+      "grad_norm": 6.9873046875,
+      "learning_rate": 0.0001795031055900621,
+      "loss": 0.4719,
+      "step": 181
+    },
+    {
+      "epoch": 1.6851851851851851,
+      "grad_norm": 6.540338516235352,
+      "learning_rate": 0.00017937888198757765,
+      "loss": 0.5329,
+      "step": 182
+    },
+    {
+      "epoch": 1.6944444444444444,
+      "grad_norm": 9.244357109069824,
+      "learning_rate": 0.00017925465838509316,
+      "loss": 1.0934,
+      "step": 183
+    },
+    {
+      "epoch": 1.7037037037037037,
+      "grad_norm": 8.517329216003418,
+      "learning_rate": 0.0001791304347826087,
+      "loss": 1.0569,
+      "step": 184
+    },
+    {
+      "epoch": 1.7129629629629628,
+      "grad_norm": 7.156026840209961,
+      "learning_rate": 0.00017900621118012425,
+      "loss": 0.4594,
+      "step": 185
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 9.52519702911377,
+      "learning_rate": 0.00017888198757763977,
+      "loss": 0.9583,
+      "step": 186
+    },
+    {
+      "epoch": 1.7314814814814814,
+      "grad_norm": 7.694867134094238,
+      "learning_rate": 0.00017875776397515528,
+      "loss": 1.013,
+      "step": 187
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 8.308947563171387,
+      "learning_rate": 0.00017863354037267083,
+      "loss": 0.8631,
+      "step": 188
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.8953022956848145,
+      "learning_rate": 0.00017850931677018634,
+      "loss": 0.1827,
+      "step": 189
+    },
+    {
+      "epoch": 1.7592592592592593,
+      "grad_norm": 5.308168411254883,
+      "learning_rate": 0.00017838509316770186,
+      "loss": 0.3858,
+      "step": 190
+    },
+    {
+      "epoch": 1.7685185185185186,
+      "grad_norm": 4.636107444763184,
+      "learning_rate": 0.0001782608695652174,
+      "loss": 0.2085,
+      "step": 191
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 11.884739875793457,
+      "learning_rate": 0.00017813664596273292,
+      "loss": 0.6365,
+      "step": 192
+    },
+    {
+      "epoch": 1.7870370370370372,
+      "grad_norm": 6.900138854980469,
+      "learning_rate": 0.00017801242236024846,
+      "loss": 0.3309,
+      "step": 193
+    },
+    {
+      "epoch": 1.7962962962962963,
+      "grad_norm": 5.4100494384765625,
+      "learning_rate": 0.00017788819875776398,
+      "loss": 0.315,
+      "step": 194
+    },
+    {
+      "epoch": 1.8055555555555556,
+      "grad_norm": 5.625341415405273,
+      "learning_rate": 0.00017776397515527952,
+      "loss": 0.2598,
+      "step": 195
+    },
+    {
+      "epoch": 1.8148148148148149,
+      "grad_norm": 8.942736625671387,
+      "learning_rate": 0.00017763975155279504,
+      "loss": 1.0631,
+      "step": 196
+    },
+    {
+      "epoch": 1.824074074074074,
+      "grad_norm": 7.598841190338135,
+      "learning_rate": 0.00017751552795031056,
+      "loss": 0.9039,
+      "step": 197
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 8.496445655822754,
+      "learning_rate": 0.0001773913043478261,
+      "loss": 1.4777,
+      "step": 198
+    },
+    {
+      "epoch": 1.8425925925925926,
+      "grad_norm": 7.111565113067627,
+      "learning_rate": 0.00017726708074534161,
+      "loss": 0.7539,
+      "step": 199
+    },
+    {
+      "epoch": 1.8518518518518519,
+      "grad_norm": 8.147820472717285,
+      "learning_rate": 0.00017714285714285713,
+      "loss": 0.8148,
+      "step": 200
+    },
+    {
+      "epoch": 1.8611111111111112,
+      "grad_norm": 9.675362586975098,
+      "learning_rate": 0.00017701863354037267,
+      "loss": 0.8903,
+      "step": 201
+    },
+    {
+      "epoch": 1.8703703703703702,
+      "grad_norm": 3.753589630126953,
+      "learning_rate": 0.00017689440993788822,
+      "loss": 0.1457,
+      "step": 202
+    },
+    {
+      "epoch": 1.8796296296296298,
+      "grad_norm": 4.512214183807373,
+      "learning_rate": 0.00017677018633540373,
+      "loss": 0.1948,
+      "step": 203
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 4.206667423248291,
+      "learning_rate": 0.00017664596273291928,
+      "loss": 0.2643,
+      "step": 204
+    },
+    {
+      "epoch": 1.8981481481481481,
+      "grad_norm": 8.376415252685547,
+      "learning_rate": 0.0001765217391304348,
+      "loss": 1.0019,
+      "step": 205
+    },
+    {
+      "epoch": 1.9074074074074074,
+      "grad_norm": 9.681005477905273,
+      "learning_rate": 0.0001763975155279503,
+      "loss": 1.0884,
+      "step": 206
+    },
+    {
+      "epoch": 1.9166666666666665,
+      "grad_norm": 4.311055660247803,
+      "learning_rate": 0.00017627329192546585,
+      "loss": 0.1727,
+      "step": 207
+    },
+    {
+      "epoch": 1.925925925925926,
+      "grad_norm": 10.78451156616211,
+      "learning_rate": 0.00017614906832298137,
+      "loss": 0.3261,
+      "step": 208
+    },
+    {
+      "epoch": 1.9351851851851851,
+      "grad_norm": 3.912935972213745,
+      "learning_rate": 0.0001760248447204969,
+      "loss": 0.1821,
+      "step": 209
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 7.964453220367432,
+      "learning_rate": 0.00017590062111801243,
+      "loss": 0.6643,
+      "step": 210
+    },
+    {
+      "epoch": 1.9537037037037037,
+      "grad_norm": 5.46901273727417,
+      "learning_rate": 0.00017577639751552797,
+      "loss": 0.3486,
+      "step": 211
+    },
+    {
+      "epoch": 1.9629629629629628,
+      "grad_norm": 7.396164894104004,
+      "learning_rate": 0.0001756521739130435,
+      "loss": 0.7032,
+      "step": 212
+    },
+    {
+      "epoch": 1.9722222222222223,
+      "grad_norm": 3.847592830657959,
+      "learning_rate": 0.000175527950310559,
+      "loss": 0.0872,
+      "step": 213
+    },
+    {
+      "epoch": 1.9814814814814814,
+      "grad_norm": 7.062877178192139,
+      "learning_rate": 0.00017540372670807455,
+      "loss": 0.7447,
+      "step": 214
+    },
+    {
+      "epoch": 1.9907407407407407,
+      "grad_norm": 5.658631801605225,
+      "learning_rate": 0.00017527950310559007,
+      "loss": 0.3057,
+      "step": 215
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.587858200073242,
+      "learning_rate": 0.00017515527950310558,
+      "loss": 0.8977,
+      "step": 216
+    },
+    {
+      "epoch": 2.009259259259259,
+      "grad_norm": 9.634937286376953,
+      "learning_rate": 0.00017503105590062113,
+      "loss": 0.8675,
+      "step": 217
+    },
+    {
+      "epoch": 2.0185185185185186,
+      "grad_norm": 3.696709156036377,
+      "learning_rate": 0.00017490683229813664,
+      "loss": 0.1366,
+      "step": 218
+    },
+    {
+      "epoch": 2.0277777777777777,
+      "grad_norm": 5.684165000915527,
+      "learning_rate": 0.00017478260869565219,
+      "loss": 0.2566,
+      "step": 219
+    },
+    {
+      "epoch": 2.037037037037037,
+      "grad_norm": 7.722327709197998,
+      "learning_rate": 0.00017465838509316773,
+      "loss": 0.285,
+      "step": 220
+    },
+    {
+      "epoch": 2.0462962962962963,
+      "grad_norm": 4.990947246551514,
+      "learning_rate": 0.00017453416149068325,
+      "loss": 0.3135,
+      "step": 221
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 3.8842992782592773,
+      "learning_rate": 0.00017440993788819876,
+      "loss": 0.0931,
+      "step": 222
+    },
+    {
+      "epoch": 2.064814814814815,
+      "grad_norm": 3.498350143432617,
+      "learning_rate": 0.0001742857142857143,
+      "loss": 0.2054,
+      "step": 223
+    },
+    {
+      "epoch": 2.074074074074074,
+      "grad_norm": 7.796590805053711,
+      "learning_rate": 0.00017416149068322982,
+      "loss": 0.7749,
+      "step": 224
+    },
+    {
+      "epoch": 2.0833333333333335,
+      "grad_norm": 5.396276950836182,
+      "learning_rate": 0.00017403726708074534,
+      "loss": 0.3065,
+      "step": 225
+    },
+    {
+      "epoch": 2.0925925925925926,
+      "grad_norm": 5.663061618804932,
+      "learning_rate": 0.00017391304347826088,
+      "loss": 0.2401,
+      "step": 226
+    },
+    {
+      "epoch": 2.1018518518518516,
+      "grad_norm": 2.867725133895874,
+      "learning_rate": 0.0001737888198757764,
+      "loss": 0.0901,
+      "step": 227
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 4.327350616455078,
+      "learning_rate": 0.00017366459627329191,
+      "loss": 0.1796,
+      "step": 228
+    },
+    {
+      "epoch": 2.1203703703703702,
+      "grad_norm": 8.092239379882812,
+      "learning_rate": 0.00017354037267080748,
+      "loss": 0.4362,
+      "step": 229
+    },
+    {
+      "epoch": 2.1296296296296298,
+      "grad_norm": 8.841894149780273,
+      "learning_rate": 0.000173416149068323,
+      "loss": 0.4565,
+      "step": 230
+    },
+    {
+      "epoch": 2.138888888888889,
+      "grad_norm": 3.5413870811462402,
+      "learning_rate": 0.00017329192546583852,
+      "loss": 0.1747,
+      "step": 231
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 11.68136978149414,
+      "learning_rate": 0.00017316770186335406,
+      "loss": 0.2018,
+      "step": 232
+    },
+    {
+      "epoch": 2.1574074074074074,
+      "grad_norm": 10.268793106079102,
+      "learning_rate": 0.00017304347826086958,
+      "loss": 0.765,
+      "step": 233
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 9.073164939880371,
+      "learning_rate": 0.0001729192546583851,
+      "loss": 0.4725,
+      "step": 234
+    },
+    {
+      "epoch": 2.175925925925926,
+      "grad_norm": 3.81775164604187,
+      "learning_rate": 0.00017279503105590064,
+      "loss": 0.1851,
+      "step": 235
+    },
+    {
+      "epoch": 2.185185185185185,
+      "grad_norm": 7.331940174102783,
+      "learning_rate": 0.00017267080745341615,
+      "loss": 0.4361,
+      "step": 236
+    },
+    {
+      "epoch": 2.1944444444444446,
+      "grad_norm": 5.943619251251221,
+      "learning_rate": 0.00017254658385093167,
+      "loss": 0.3958,
+      "step": 237
+    },
+    {
+      "epoch": 2.2037037037037037,
+      "grad_norm": 9.69388484954834,
+      "learning_rate": 0.0001724223602484472,
+      "loss": 0.2486,
+      "step": 238
+    },
+    {
+      "epoch": 2.212962962962963,
+      "grad_norm": 9.289849281311035,
+      "learning_rate": 0.00017229813664596276,
+      "loss": 0.4469,
+      "step": 239
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 6.857956886291504,
+      "learning_rate": 0.00017217391304347827,
+      "loss": 0.3926,
+      "step": 240
+    },
+    {
+      "epoch": 2.2314814814814814,
+      "grad_norm": 10.366476058959961,
+      "learning_rate": 0.0001720496894409938,
+      "loss": 1.3493,
+      "step": 241
+    },
+    {
+      "epoch": 2.240740740740741,
+      "grad_norm": 5.455699443817139,
+      "learning_rate": 0.00017192546583850933,
+      "loss": 0.2597,
+      "step": 242
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.438530921936035,
+      "learning_rate": 0.00017180124223602485,
+      "loss": 0.5654,
+      "step": 243
+    },
+    {
+      "epoch": 2.259259259259259,
+      "grad_norm": 7.767147541046143,
+      "learning_rate": 0.00017167701863354037,
+      "loss": 0.5442,
+      "step": 244
+    },
+    {
+      "epoch": 2.2685185185185186,
+      "grad_norm": 9.872796058654785,
+      "learning_rate": 0.0001715527950310559,
+      "loss": 0.7294,
+      "step": 245
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 10.673378944396973,
+      "learning_rate": 0.00017142857142857143,
+      "loss": 1.1546,
+      "step": 246
+    },
+    {
+      "epoch": 2.287037037037037,
+      "grad_norm": 21.295475006103516,
+      "learning_rate": 0.00017130434782608697,
+      "loss": 0.2201,
+      "step": 247
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 8.360919952392578,
+      "learning_rate": 0.0001711801242236025,
+      "loss": 0.4692,
+      "step": 248
+    },
+    {
+      "epoch": 2.3055555555555554,
+      "grad_norm": 7.903579235076904,
+      "learning_rate": 0.00017105590062111803,
+      "loss": 0.2641,
+      "step": 249
+    },
+    {
+      "epoch": 2.314814814814815,
+      "grad_norm": 6.9065399169921875,
+      "learning_rate": 0.00017093167701863354,
+      "loss": 0.3394,
+      "step": 250
+    },
+    {
+      "epoch": 2.324074074074074,
+      "grad_norm": 4.162428379058838,
+      "learning_rate": 0.0001708074534161491,
+      "loss": 0.1332,
+      "step": 251
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 9.210295677185059,
+      "learning_rate": 0.0001706832298136646,
+      "loss": 0.6607,
+      "step": 252
+    },
+    {
+      "epoch": 2.3425925925925926,
+      "grad_norm": 10.019932746887207,
+      "learning_rate": 0.00017055900621118012,
+      "loss": 0.5042,
+      "step": 253
+    },
+    {
+      "epoch": 2.351851851851852,
+      "grad_norm": 4.602824687957764,
+      "learning_rate": 0.00017043478260869566,
+      "loss": 0.1234,
+      "step": 254
+    },
+    {
+      "epoch": 2.361111111111111,
+      "grad_norm": 10.391927719116211,
+      "learning_rate": 0.00017031055900621118,
+      "loss": 0.6339,
+      "step": 255
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 5.13262414932251,
+      "learning_rate": 0.00017018633540372672,
+      "loss": 0.2344,
+      "step": 256
+    },
+    {
+      "epoch": 2.3796296296296298,
+      "grad_norm": 8.059638023376465,
+      "learning_rate": 0.00017006211180124224,
+      "loss": 0.5158,
+      "step": 257
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 34.22102355957031,
+      "learning_rate": 0.00016993788819875778,
+      "loss": 0.5373,
+      "step": 258
+    },
+    {
+      "epoch": 2.398148148148148,
+      "grad_norm": 13.332140922546387,
+      "learning_rate": 0.0001698136645962733,
+      "loss": 0.7577,
+      "step": 259
+    },
+    {
+      "epoch": 2.4074074074074074,
+      "grad_norm": 4.515472412109375,
+      "learning_rate": 0.00016968944099378882,
+      "loss": 0.1429,
+      "step": 260
+    },
+    {
+      "epoch": 2.4166666666666665,
+      "grad_norm": 6.849452972412109,
+      "learning_rate": 0.00016956521739130436,
+      "loss": 0.1432,
+      "step": 261
+    },
+    {
+      "epoch": 2.425925925925926,
+      "grad_norm": 5.058333873748779,
+      "learning_rate": 0.00016944099378881988,
+      "loss": 0.2357,
+      "step": 262
+    },
+    {
+      "epoch": 2.435185185185185,
+      "grad_norm": 9.91899299621582,
+      "learning_rate": 0.0001693167701863354,
+      "loss": 0.3807,
+      "step": 263
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 10.223733901977539,
+      "learning_rate": 0.00016919254658385094,
+      "loss": 0.9076,
+      "step": 264
+    },
+    {
+      "epoch": 2.4537037037037037,
+      "grad_norm": 6.5338335037231445,
+      "learning_rate": 0.00016906832298136648,
+      "loss": 0.1497,
+      "step": 265
+    },
+    {
+      "epoch": 2.462962962962963,
+      "grad_norm": 5.8402299880981445,
+      "learning_rate": 0.000168944099378882,
+      "loss": 0.5041,
+      "step": 266
+    },
+    {
+      "epoch": 2.4722222222222223,
+      "grad_norm": 3.3992233276367188,
+      "learning_rate": 0.00016881987577639754,
+      "loss": 0.1547,
+      "step": 267
+    },
+    {
+      "epoch": 2.4814814814814814,
+      "grad_norm": 7.095219612121582,
+      "learning_rate": 0.00016869565217391306,
+      "loss": 0.5186,
+      "step": 268
+    },
+    {
+      "epoch": 2.490740740740741,
+      "grad_norm": 6.44674825668335,
+      "learning_rate": 0.00016857142857142857,
+      "loss": 0.1233,
+      "step": 269
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.471312522888184,
+      "learning_rate": 0.00016844720496894412,
+      "loss": 0.24,
+      "step": 270
+    },
+    {
+      "epoch": 2.5092592592592595,
+      "grad_norm": 5.365081310272217,
+      "learning_rate": 0.00016832298136645963,
+      "loss": 0.3773,
+      "step": 271
+    },
+    {
+      "epoch": 2.5185185185185186,
+      "grad_norm": 9.75694465637207,
+      "learning_rate": 0.00016819875776397515,
+      "loss": 0.5362,
+      "step": 272
+    },
+    {
+      "epoch": 2.5277777777777777,
+      "grad_norm": 8.214263916015625,
+      "learning_rate": 0.0001680745341614907,
+      "loss": 0.4716,
+      "step": 273
+    },
+    {
+      "epoch": 2.537037037037037,
+      "grad_norm": 10.777454376220703,
+      "learning_rate": 0.00016795031055900624,
+      "loss": 0.6536,
+      "step": 274
+    },
+    {
+      "epoch": 2.5462962962962963,
+      "grad_norm": 8.224162101745605,
+      "learning_rate": 0.00016782608695652175,
+      "loss": 0.5819,
+      "step": 275
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 10.33161449432373,
+      "learning_rate": 0.00016770186335403727,
+      "loss": 0.6333,
+      "step": 276
+    },
+    {
+      "epoch": 2.564814814814815,
+      "grad_norm": 6.827465057373047,
+      "learning_rate": 0.0001675776397515528,
+      "loss": 0.2474,
+      "step": 277
+    },
+    {
+      "epoch": 2.574074074074074,
+      "grad_norm": 7.896028995513916,
+      "learning_rate": 0.00016745341614906833,
+      "loss": 0.5632,
+      "step": 278
+    },
+    {
+      "epoch": 2.5833333333333335,
+      "grad_norm": 4.075222492218018,
+      "learning_rate": 0.00016732919254658384,
+      "loss": 0.1224,
+      "step": 279
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 4.907321929931641,
+      "learning_rate": 0.0001672049689440994,
+      "loss": 0.1726,
+      "step": 280
+    },
+    {
+      "epoch": 2.601851851851852,
+      "grad_norm": 8.992182731628418,
+      "learning_rate": 0.0001670807453416149,
+      "loss": 0.6303,
+      "step": 281
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 7.870790958404541,
+      "learning_rate": 0.00016695652173913042,
+      "loss": 0.4438,
+      "step": 282
+    },
+    {
+      "epoch": 2.6203703703703702,
+      "grad_norm": 5.292234420776367,
+      "learning_rate": 0.000166832298136646,
+      "loss": 0.1207,
+      "step": 283
+    },
+    {
+      "epoch": 2.6296296296296298,
+      "grad_norm": 1.3497140407562256,
+      "learning_rate": 0.0001667080745341615,
+      "loss": 0.0298,
+      "step": 284
+    },
+    {
+      "epoch": 2.638888888888889,
+      "grad_norm": 7.471672058105469,
+      "learning_rate": 0.00016658385093167702,
+      "loss": 0.5523,
+      "step": 285
+    },
+    {
+      "epoch": 2.648148148148148,
+      "grad_norm": 1.7936245203018188,
+      "learning_rate": 0.00016645962732919257,
+      "loss": 0.1268,
+      "step": 286
+    },
+    {
+      "epoch": 2.6574074074074074,
+      "grad_norm": 8.869959831237793,
+      "learning_rate": 0.00016633540372670808,
+      "loss": 0.7427,
+      "step": 287
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 7.612359523773193,
+      "learning_rate": 0.0001662111801242236,
+      "loss": 0.5962,
+      "step": 288
+    },
+    {
+      "epoch": 2.675925925925926,
+      "grad_norm": 3.3016366958618164,
+      "learning_rate": 0.00016608695652173914,
+      "loss": 0.124,
+      "step": 289
+    },
+    {
+      "epoch": 2.685185185185185,
+      "grad_norm": 4.570164680480957,
+      "learning_rate": 0.00016596273291925466,
+      "loss": 0.2078,
+      "step": 290
+    },
+    {
+      "epoch": 2.6944444444444446,
+      "grad_norm": 4.242143630981445,
+      "learning_rate": 0.00016583850931677018,
+      "loss": 0.1053,
+      "step": 291
+    },
+    {
+      "epoch": 2.7037037037037037,
+      "grad_norm": 3.7191407680511475,
+      "learning_rate": 0.00016571428571428575,
+      "loss": 0.1582,
+      "step": 292
+    },
+    {
+      "epoch": 2.712962962962963,
+      "grad_norm": 5.166412830352783,
+      "learning_rate": 0.00016559006211180126,
+      "loss": 0.2469,
+      "step": 293
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 4.837871551513672,
+      "learning_rate": 0.00016546583850931678,
+      "loss": 0.2485,
+      "step": 294
+    },
+    {
+      "epoch": 2.7314814814814814,
+      "grad_norm": 8.435837745666504,
+      "learning_rate": 0.00016534161490683232,
+      "loss": 0.5547,
+      "step": 295
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 13.024179458618164,
+      "learning_rate": 0.00016521739130434784,
+      "loss": 0.7232,
+      "step": 296
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.188488960266113,
+      "learning_rate": 0.00016509316770186336,
+      "loss": 0.9133,
+      "step": 297
+    },
+    {
+      "epoch": 2.7592592592592595,
+      "grad_norm": 6.474546432495117,
+      "learning_rate": 0.0001649689440993789,
+      "loss": 0.5251,
+      "step": 298
+    },
+    {
+      "epoch": 2.7685185185185186,
+      "grad_norm": 8.005181312561035,
+      "learning_rate": 0.00016484472049689441,
+      "loss": 0.6408,
+      "step": 299
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 7.383902549743652,
+      "learning_rate": 0.00016472049689440993,
+      "loss": 0.5634,
+      "step": 300
+    },
+    {
+      "epoch": 2.787037037037037,
+      "grad_norm": 5.368563652038574,
+      "learning_rate": 0.00016459627329192547,
+      "loss": 0.264,
+      "step": 301
+    },
+    {
+      "epoch": 2.7962962962962963,
+      "grad_norm": 3.9778330326080322,
+      "learning_rate": 0.00016447204968944102,
+      "loss": 0.1811,
+      "step": 302
+    },
+    {
+      "epoch": 2.8055555555555554,
+      "grad_norm": 6.301558971405029,
+      "learning_rate": 0.00016434782608695653,
+      "loss": 0.3352,
+      "step": 303
+    },
+    {
+      "epoch": 2.814814814814815,
+      "grad_norm": 11.769688606262207,
+      "learning_rate": 0.00016422360248447205,
+      "loss": 0.5181,
+      "step": 304
+    },
+    {
+      "epoch": 2.824074074074074,
+      "grad_norm": 11.051255226135254,
+      "learning_rate": 0.0001640993788819876,
+      "loss": 0.9957,
+      "step": 305
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 14.160080909729004,
+      "learning_rate": 0.0001639751552795031,
+      "loss": 0.3998,
+      "step": 306
+    },
+    {
+      "epoch": 2.8425925925925926,
+      "grad_norm": 5.162957668304443,
+      "learning_rate": 0.00016385093167701863,
+      "loss": 0.2078,
+      "step": 307
+    },
+    {
+      "epoch": 2.851851851851852,
+      "grad_norm": 4.017885208129883,
+      "learning_rate": 0.00016372670807453417,
+      "loss": 0.1804,
+      "step": 308
+    },
+    {
+      "epoch": 2.861111111111111,
+      "grad_norm": 8.482934951782227,
+      "learning_rate": 0.0001636024844720497,
+      "loss": 0.6282,
+      "step": 309
+    },
+    {
+      "epoch": 2.8703703703703702,
+      "grad_norm": 5.554912090301514,
+      "learning_rate": 0.00016347826086956523,
+      "loss": 0.2312,
+      "step": 310
+    },
+    {
+      "epoch": 2.8796296296296298,
+      "grad_norm": 9.502184867858887,
+      "learning_rate": 0.00016335403726708077,
+      "loss": 0.6199,
+      "step": 311
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 8.019617080688477,
+      "learning_rate": 0.0001632298136645963,
+      "loss": 0.6601,
+      "step": 312
+    },
+    {
+      "epoch": 2.898148148148148,
+      "grad_norm": 5.383131504058838,
+      "learning_rate": 0.0001631055900621118,
+      "loss": 0.3231,
+      "step": 313
+    },
+    {
+      "epoch": 2.9074074074074074,
+      "grad_norm": 5.896549224853516,
+      "learning_rate": 0.00016298136645962735,
+      "loss": 0.3919,
+      "step": 314
+    },
+    {
+      "epoch": 2.9166666666666665,
+      "grad_norm": 7.678750038146973,
+      "learning_rate": 0.00016285714285714287,
+      "loss": 0.4805,
+      "step": 315
+    },
+    {
+      "epoch": 2.925925925925926,
+      "grad_norm": 9.610578536987305,
+      "learning_rate": 0.00016273291925465838,
+      "loss": 0.9259,
+      "step": 316
+    },
+    {
+      "epoch": 2.935185185185185,
+      "grad_norm": 9.18185806274414,
+      "learning_rate": 0.00016260869565217393,
+      "loss": 1.0033,
+      "step": 317
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 3.519683599472046,
+      "learning_rate": 0.00016248447204968944,
+      "loss": 0.1465,
+      "step": 318
+    },
+    {
+      "epoch": 2.9537037037037037,
+      "grad_norm": 5.032393455505371,
+      "learning_rate": 0.00016236024844720496,
+      "loss": 0.2461,
+      "step": 319
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 7.032131671905518,
+      "learning_rate": 0.0001622360248447205,
+      "loss": 0.6732,
+      "step": 320
+    },
+    {
+      "epoch": 2.9722222222222223,
+      "grad_norm": 3.6258678436279297,
+      "learning_rate": 0.00016211180124223605,
+      "loss": 0.1176,
+      "step": 321
+    },
+    {
+      "epoch": 2.9814814814814814,
+      "grad_norm": 1.865622878074646,
+      "learning_rate": 0.00016198757763975156,
+      "loss": 0.0598,
+      "step": 322
+    },
+    {
+      "epoch": 2.9907407407407405,
+      "grad_norm": 6.845046043395996,
+      "learning_rate": 0.00016186335403726708,
+      "loss": 0.4092,
+      "step": 323
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.940382480621338,
+      "learning_rate": 0.00016173913043478262,
+      "loss": 0.1281,
+      "step": 324
+    },
+    {
+      "epoch": 3.009259259259259,
+      "grad_norm": 3.2817156314849854,
+      "learning_rate": 0.00016161490683229814,
+      "loss": 0.076,
+      "step": 325
+    },
+    {
+      "epoch": 3.0185185185185186,
+      "grad_norm": 4.320040225982666,
+      "learning_rate": 0.00016149068322981365,
+      "loss": 0.164,
+      "step": 326
+    },
+    {
+      "epoch": 3.0277777777777777,
+      "grad_norm": 6.783189296722412,
+      "learning_rate": 0.0001613664596273292,
+      "loss": 0.312,
+      "step": 327
+    },
+    {
+      "epoch": 3.037037037037037,
+      "grad_norm": 9.855510711669922,
+      "learning_rate": 0.00016124223602484471,
+      "loss": 0.7179,
+      "step": 328
+    },
+    {
+      "epoch": 3.0462962962962963,
+      "grad_norm": 11.480785369873047,
+      "learning_rate": 0.00016111801242236026,
+      "loss": 0.7069,
+      "step": 329
+    },
+    {
+      "epoch": 3.0555555555555554,
+      "grad_norm": 6.975889205932617,
+      "learning_rate": 0.0001609937888198758,
+      "loss": 0.3169,
+      "step": 330
+    },
+    {
+      "epoch": 3.064814814814815,
+      "grad_norm": 3.5186703205108643,
+      "learning_rate": 0.00016086956521739132,
+      "loss": 0.1183,
+      "step": 331
+    },
+    {
+      "epoch": 3.074074074074074,
+      "grad_norm": 9.784034729003906,
+      "learning_rate": 0.00016074534161490683,
+      "loss": 0.601,
+      "step": 332
+    },
+    {
+      "epoch": 3.0833333333333335,
+      "grad_norm": 2.755011796951294,
+      "learning_rate": 0.00016062111801242238,
+      "loss": 0.1034,
+      "step": 333
+    },
+    {
+      "epoch": 3.0925925925925926,
+      "grad_norm": 3.478494167327881,
+      "learning_rate": 0.0001604968944099379,
+      "loss": 0.1482,
+      "step": 334
+    },
+    {
+      "epoch": 3.1018518518518516,
+      "grad_norm": 5.404292583465576,
+      "learning_rate": 0.0001603726708074534,
+      "loss": 0.2123,
+      "step": 335
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 7.036523342132568,
+      "learning_rate": 0.00016024844720496895,
+      "loss": 0.4788,
+      "step": 336
+    },
+    {
+      "epoch": 3.1203703703703702,
+      "grad_norm": 4.069665908813477,
+      "learning_rate": 0.00016012422360248447,
+      "loss": 0.2323,
+      "step": 337
+    },
+    {
+      "epoch": 3.1296296296296298,
+      "grad_norm": 2.5665032863616943,
+      "learning_rate": 0.00016,
+      "loss": 0.0713,
+      "step": 338
+    },
+    {
+      "epoch": 3.138888888888889,
+      "grad_norm": 8.244379997253418,
+      "learning_rate": 0.00015987577639751553,
+      "loss": 0.4594,
+      "step": 339
+    },
+    {
+      "epoch": 3.148148148148148,
+      "grad_norm": 3.4311187267303467,
+      "learning_rate": 0.00015975155279503107,
+      "loss": 0.1345,
+      "step": 340
+    },
+    {
+      "epoch": 3.1574074074074074,
+      "grad_norm": 6.329025745391846,
+      "learning_rate": 0.0001596273291925466,
+      "loss": 0.2205,
+      "step": 341
+    },
+    {
+      "epoch": 3.1666666666666665,
+      "grad_norm": 8.342190742492676,
+      "learning_rate": 0.0001595031055900621,
+      "loss": 0.397,
+      "step": 342
+    },
+    {
+      "epoch": 3.175925925925926,
+      "grad_norm": 3.3707950115203857,
+      "learning_rate": 0.00015937888198757765,
+      "loss": 0.1535,
+      "step": 343
+    },
+    {
+      "epoch": 3.185185185185185,
+      "grad_norm": 2.9287197589874268,
+      "learning_rate": 0.00015925465838509317,
+      "loss": 0.114,
+      "step": 344
+    },
+    {
+      "epoch": 3.1944444444444446,
+      "grad_norm": 1.9324220418930054,
+      "learning_rate": 0.00015913043478260868,
+      "loss": 0.0537,
+      "step": 345
+    },
+    {
+      "epoch": 3.2037037037037037,
+      "grad_norm": 8.233223915100098,
+      "learning_rate": 0.00015900621118012423,
+      "loss": 0.2895,
+      "step": 346
+    },
+    {
+      "epoch": 3.212962962962963,
+      "grad_norm": 8.415253639221191,
+      "learning_rate": 0.00015888198757763977,
+      "loss": 0.2266,
+      "step": 347
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 7.05043888092041,
+      "learning_rate": 0.00015875776397515528,
+      "loss": 0.247,
+      "step": 348
+    },
+    {
+      "epoch": 3.2314814814814814,
+      "grad_norm": 7.294909477233887,
+      "learning_rate": 0.00015863354037267083,
+      "loss": 0.3165,
+      "step": 349
+    },
+    {
+      "epoch": 3.240740740740741,
+      "grad_norm": 4.877599716186523,
+      "learning_rate": 0.00015850931677018634,
+      "loss": 0.1593,
+      "step": 350
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.8372039794921875,
+      "learning_rate": 0.00015838509316770186,
+      "loss": 0.2842,
+      "step": 351
+    },
+    {
+      "epoch": 3.259259259259259,
+      "grad_norm": 5.851783275604248,
+      "learning_rate": 0.0001582608695652174,
+      "loss": 0.2469,
+      "step": 352
+    },
+    {
+      "epoch": 3.2685185185185186,
+      "grad_norm": 2.9726572036743164,
+      "learning_rate": 0.00015813664596273292,
+      "loss": 0.1536,
+      "step": 353
+    },
+    {
+      "epoch": 3.2777777777777777,
+      "grad_norm": 8.282976150512695,
+      "learning_rate": 0.00015801242236024844,
+      "loss": 0.3309,
+      "step": 354
+    },
+    {
+      "epoch": 3.287037037037037,
+      "grad_norm": 9.234270095825195,
+      "learning_rate": 0.00015788819875776398,
+      "loss": 0.4955,
+      "step": 355
+    },
+    {
+      "epoch": 3.2962962962962963,
+      "grad_norm": 9.033778190612793,
+      "learning_rate": 0.00015776397515527952,
+      "loss": 0.3431,
+      "step": 356
+    },
+    {
+      "epoch": 3.3055555555555554,
+      "grad_norm": 4.390086650848389,
+      "learning_rate": 0.00015763975155279504,
+      "loss": 0.2213,
+      "step": 357
+    },
+    {
+      "epoch": 3.314814814814815,
+      "grad_norm": 7.078484058380127,
+      "learning_rate": 0.00015751552795031058,
+      "loss": 0.4093,
+      "step": 358
+    },
+    {
+      "epoch": 3.324074074074074,
+      "grad_norm": 1.9521043300628662,
+      "learning_rate": 0.0001573913043478261,
+      "loss": 0.0501,
+      "step": 359
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 6.5932159423828125,
+      "learning_rate": 0.00015726708074534162,
+      "loss": 0.2902,
+      "step": 360
+    },
+    {
+      "epoch": 3.3425925925925926,
+      "grad_norm": 1.199304223060608,
+      "learning_rate": 0.00015714285714285716,
+      "loss": 0.0348,
+      "step": 361
+    },
+    {
+      "epoch": 3.351851851851852,
+      "grad_norm": 7.945189476013184,
+      "learning_rate": 0.00015701863354037268,
+      "loss": 0.4226,
+      "step": 362
+    },
+    {
+      "epoch": 3.361111111111111,
+      "grad_norm": 6.304150104522705,
+      "learning_rate": 0.0001568944099378882,
+      "loss": 0.3419,
+      "step": 363
+    },
+    {
+      "epoch": 3.3703703703703702,
+      "grad_norm": 0.575678825378418,
+      "learning_rate": 0.00015677018633540374,
+      "loss": 0.0143,
+      "step": 364
+    },
+    {
+      "epoch": 3.3796296296296298,
+      "grad_norm": 55.86056137084961,
+      "learning_rate": 0.00015664596273291928,
+      "loss": 1.0001,
+      "step": 365
+    },
+    {
+      "epoch": 3.388888888888889,
+      "grad_norm": 5.551873683929443,
+      "learning_rate": 0.0001565217391304348,
+      "loss": 0.316,
+      "step": 366
+    },
+    {
+      "epoch": 3.398148148148148,
+      "grad_norm": 4.200618743896484,
+      "learning_rate": 0.0001563975155279503,
+      "loss": 0.1263,
+      "step": 367
+    },
+    {
+      "epoch": 3.4074074074074074,
+      "grad_norm": 9.562318801879883,
+      "learning_rate": 0.00015627329192546586,
+      "loss": 0.8528,
+      "step": 368
+    },
+    {
+      "epoch": 3.4166666666666665,
+      "grad_norm": 3.7029190063476562,
+      "learning_rate": 0.00015614906832298137,
+      "loss": 0.1608,
+      "step": 369
+    },
+    {
+      "epoch": 3.425925925925926,
+      "grad_norm": 5.9052252769470215,
+      "learning_rate": 0.0001560248447204969,
+      "loss": 0.2919,
+      "step": 370
+    },
+    {
+      "epoch": 3.435185185185185,
+      "grad_norm": 5.278835773468018,
+      "learning_rate": 0.00015590062111801243,
+      "loss": 0.1226,
+      "step": 371
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 6.618929862976074,
+      "learning_rate": 0.00015577639751552795,
+      "loss": 0.3602,
+      "step": 372
+    },
+    {
+      "epoch": 3.4537037037037037,
+      "grad_norm": 3.45408296585083,
+      "learning_rate": 0.00015565217391304346,
+      "loss": 0.0969,
+      "step": 373
+    },
+    {
+      "epoch": 3.462962962962963,
+      "grad_norm": 4.552202224731445,
+      "learning_rate": 0.00015552795031055904,
+      "loss": 0.1777,
+      "step": 374
+    },
+    {
+      "epoch": 3.4722222222222223,
+      "grad_norm": 2.348261833190918,
+      "learning_rate": 0.00015540372670807455,
+      "loss": 0.0569,
+      "step": 375
+    },
+    {
+      "epoch": 3.4814814814814814,
+      "grad_norm": 1.7154549360275269,
+      "learning_rate": 0.00015527950310559007,
+      "loss": 0.0647,
+      "step": 376
+    },
+    {
+      "epoch": 3.490740740740741,
+      "grad_norm": 7.577757358551025,
+      "learning_rate": 0.0001551552795031056,
+      "loss": 0.4125,
+      "step": 377
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 6.152462005615234,
+      "learning_rate": 0.00015503105590062113,
+      "loss": 0.4456,
+      "step": 378
+    },
+    {
+      "epoch": 3.5092592592592595,
+      "grad_norm": 6.738099575042725,
+      "learning_rate": 0.00015490683229813664,
+      "loss": 0.4373,
+      "step": 379
+    },
+    {
+      "epoch": 3.5185185185185186,
+      "grad_norm": 6.085219860076904,
+      "learning_rate": 0.0001547826086956522,
+      "loss": 0.2929,
+      "step": 380
+    },
+    {
+      "epoch": 3.5277777777777777,
+      "grad_norm": 3.194652557373047,
+      "learning_rate": 0.0001546583850931677,
+      "loss": 0.2306,
+      "step": 381
+    },
+    {
+      "epoch": 3.537037037037037,
+      "grad_norm": 7.23938512802124,
+      "learning_rate": 0.00015453416149068322,
+      "loss": 0.5282,
+      "step": 382
+    },
+    {
+      "epoch": 3.5462962962962963,
+      "grad_norm": 5.027129173278809,
+      "learning_rate": 0.00015440993788819876,
+      "loss": 0.2126,
+      "step": 383
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 11.30113410949707,
+      "learning_rate": 0.0001542857142857143,
+      "loss": 0.4177,
+      "step": 384
+    },
+    {
+      "epoch": 3.564814814814815,
+      "grad_norm": 9.71517276763916,
+      "learning_rate": 0.00015416149068322982,
+      "loss": 0.4661,
+      "step": 385
+    },
+    {
+      "epoch": 3.574074074074074,
+      "grad_norm": 5.263823509216309,
+      "learning_rate": 0.00015403726708074534,
+      "loss": 0.0915,
+      "step": 386
+    },
+    {
+      "epoch": 3.5833333333333335,
+      "grad_norm": 8.799884796142578,
+      "learning_rate": 0.00015391304347826088,
+      "loss": 0.2908,
+      "step": 387
+    },
+    {
+      "epoch": 3.5925925925925926,
+      "grad_norm": 5.173797130584717,
+      "learning_rate": 0.0001537888198757764,
+      "loss": 0.2236,
+      "step": 388
+    },
+    {
+      "epoch": 3.601851851851852,
+      "grad_norm": 4.034602165222168,
+      "learning_rate": 0.00015366459627329192,
+      "loss": 0.2318,
+      "step": 389
+    },
+    {
+      "epoch": 3.611111111111111,
+      "grad_norm": 6.178292751312256,
+      "learning_rate": 0.00015354037267080746,
+      "loss": 0.4033,
+      "step": 390
+    },
+    {
+      "epoch": 3.6203703703703702,
+      "grad_norm": 6.233795166015625,
+      "learning_rate": 0.00015341614906832298,
+      "loss": 0.3785,
+      "step": 391
+    },
+    {
+      "epoch": 3.6296296296296298,
+      "grad_norm": 9.079377174377441,
+      "learning_rate": 0.00015329192546583852,
+      "loss": 0.2154,
+      "step": 392
+    },
+    {
+      "epoch": 3.638888888888889,
+      "grad_norm": 7.057285785675049,
+      "learning_rate": 0.00015316770186335406,
+      "loss": 0.4248,
+      "step": 393
+    },
+    {
+      "epoch": 3.648148148148148,
+      "grad_norm": 1.6093757152557373,
+      "learning_rate": 0.00015304347826086958,
+      "loss": 0.0697,
+      "step": 394
+    },
+    {
+      "epoch": 3.6574074074074074,
+      "grad_norm": 5.713193416595459,
+      "learning_rate": 0.0001529192546583851,
+      "loss": 0.304,
+      "step": 395
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 4.769229888916016,
+      "learning_rate": 0.00015279503105590064,
+      "loss": 0.1802,
+      "step": 396
+    },
+    {
+      "epoch": 3.675925925925926,
+      "grad_norm": 11.877578735351562,
+      "learning_rate": 0.00015267080745341616,
+      "loss": 0.952,
+      "step": 397
+    },
+    {
+      "epoch": 3.685185185185185,
+      "grad_norm": 3.5577473640441895,
+      "learning_rate": 0.00015254658385093167,
+      "loss": 0.1702,
+      "step": 398
+    },
+    {
+      "epoch": 3.6944444444444446,
+      "grad_norm": 4.835704803466797,
+      "learning_rate": 0.00015242236024844721,
+      "loss": 0.1728,
+      "step": 399
+    },
+    {
+      "epoch": 3.7037037037037037,
+      "grad_norm": 8.712724685668945,
+      "learning_rate": 0.00015229813664596273,
+      "loss": 0.5967,
+      "step": 400
+    },
+    {
+      "epoch": 3.712962962962963,
+      "grad_norm": 3.5652997493743896,
+      "learning_rate": 0.00015217391304347827,
+      "loss": 0.1029,
+      "step": 401
+    },
+    {
+      "epoch": 3.7222222222222223,
+      "grad_norm": 7.978386402130127,
+      "learning_rate": 0.0001520496894409938,
+      "loss": 0.456,
+      "step": 402
+    },
+    {
+      "epoch": 3.7314814814814814,
+      "grad_norm": 8.434885025024414,
+      "learning_rate": 0.00015192546583850933,
+      "loss": 0.4298,
+      "step": 403
+    },
+    {
+      "epoch": 3.7407407407407405,
+      "grad_norm": 5.78842306137085,
+      "learning_rate": 0.00015180124223602485,
+      "loss": 0.3563,
+      "step": 404
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.0681047439575195,
+      "learning_rate": 0.00015167701863354037,
+      "loss": 0.2031,
+      "step": 405
+    },
+    {
+      "epoch": 3.7592592592592595,
+      "grad_norm": 4.695619583129883,
+      "learning_rate": 0.0001515527950310559,
+      "loss": 0.1284,
+      "step": 406
+    },
+    {
+      "epoch": 3.7685185185185186,
+      "grad_norm": 4.437154769897461,
+      "learning_rate": 0.00015142857142857143,
+      "loss": 0.2112,
+      "step": 407
+    },
+    {
+      "epoch": 3.7777777777777777,
+      "grad_norm": 3.3746249675750732,
+      "learning_rate": 0.00015130434782608694,
+      "loss": 0.1581,
+      "step": 408
+    },
+    {
+      "epoch": 3.787037037037037,
+      "grad_norm": 7.2292585372924805,
+      "learning_rate": 0.0001511801242236025,
+      "loss": 0.4854,
+      "step": 409
+    },
+    {
+      "epoch": 3.7962962962962963,
+      "grad_norm": 7.788264751434326,
+      "learning_rate": 0.00015105590062111803,
+      "loss": 0.4422,
+      "step": 410
+    },
+    {
+      "epoch": 3.8055555555555554,
+      "grad_norm": 7.971909999847412,
+      "learning_rate": 0.00015093167701863355,
+      "loss": 0.5586,
+      "step": 411
+    },
+    {
+      "epoch": 3.814814814814815,
+      "grad_norm": 7.098501682281494,
+      "learning_rate": 0.0001508074534161491,
+      "loss": 0.2834,
+      "step": 412
+    },
+    {
+      "epoch": 3.824074074074074,
+      "grad_norm": 3.429027795791626,
+      "learning_rate": 0.0001506832298136646,
+      "loss": 0.1501,
+      "step": 413
+    },
+    {
+      "epoch": 3.8333333333333335,
+      "grad_norm": 8.535176277160645,
+      "learning_rate": 0.00015055900621118012,
+      "loss": 0.3973,
+      "step": 414
+    },
+    {
+      "epoch": 3.8425925925925926,
+      "grad_norm": 7.709643363952637,
+      "learning_rate": 0.00015043478260869567,
+      "loss": 0.6272,
+      "step": 415
+    },
+    {
+      "epoch": 3.851851851851852,
+      "grad_norm": 5.064123630523682,
+      "learning_rate": 0.00015031055900621118,
+      "loss": 0.2087,
+      "step": 416
+    },
+    {
+      "epoch": 3.861111111111111,
+      "grad_norm": 2.5206053256988525,
+      "learning_rate": 0.0001501863354037267,
+      "loss": 0.095,
+      "step": 417
+    },
+    {
+      "epoch": 3.8703703703703702,
+      "grad_norm": 8.01754093170166,
+      "learning_rate": 0.00015006211180124224,
+      "loss": 0.2509,
+      "step": 418
+    },
+    {
+      "epoch": 3.8796296296296298,
+      "grad_norm": 3.0863702297210693,
+      "learning_rate": 0.00014993788819875779,
+      "loss": 0.1075,
+      "step": 419
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 4.70889139175415,
+      "learning_rate": 0.0001498136645962733,
+      "loss": 0.1924,
+      "step": 420
+    },
+    {
+      "epoch": 3.898148148148148,
+      "grad_norm": 7.025390148162842,
+      "learning_rate": 0.00014968944099378885,
+      "loss": 0.359,
+      "step": 421
+    },
+    {
+      "epoch": 3.9074074074074074,
+      "grad_norm": 5.652724266052246,
+      "learning_rate": 0.00014956521739130436,
+      "loss": 0.307,
+      "step": 422
+    },
+    {
+      "epoch": 3.9166666666666665,
+      "grad_norm": 6.922481536865234,
+      "learning_rate": 0.00014944099378881988,
+      "loss": 0.2555,
+      "step": 423
+    },
+    {
+      "epoch": 3.925925925925926,
+      "grad_norm": 3.9260189533233643,
+      "learning_rate": 0.00014931677018633542,
+      "loss": 0.2264,
+      "step": 424
+    },
+    {
+      "epoch": 3.935185185185185,
+      "grad_norm": 7.77918004989624,
+      "learning_rate": 0.00014919254658385094,
+      "loss": 0.5299,
+      "step": 425
+    },
+    {
+      "epoch": 3.9444444444444446,
+      "grad_norm": 8.371830940246582,
+      "learning_rate": 0.00014906832298136645,
+      "loss": 0.5051,
+      "step": 426
+    },
+    {
+      "epoch": 3.9537037037037037,
+      "grad_norm": 4.843681812286377,
+      "learning_rate": 0.000148944099378882,
+      "loss": 0.23,
+      "step": 427
+    },
+    {
+      "epoch": 3.962962962962963,
+      "grad_norm": 5.792315483093262,
+      "learning_rate": 0.00014881987577639754,
+      "loss": 0.2242,
+      "step": 428
+    },
+    {
+      "epoch": 3.9722222222222223,
+      "grad_norm": 3.9335224628448486,
+      "learning_rate": 0.00014869565217391306,
+      "loss": 0.1533,
+      "step": 429
+    },
+    {
+      "epoch": 3.9814814814814814,
+      "grad_norm": 12.831199645996094,
+      "learning_rate": 0.00014857142857142857,
+      "loss": 0.7754,
+      "step": 430
+    },
+    {
+      "epoch": 3.9907407407407405,
+      "grad_norm": 3.6635866165161133,
+      "learning_rate": 0.00014844720496894412,
+      "loss": 0.1512,
+      "step": 431
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 7.196122646331787,
+      "learning_rate": 0.00014832298136645963,
+      "loss": 0.3633,
+      "step": 432
+    },
+    {
+      "epoch": 4.0092592592592595,
+      "grad_norm": 3.1862425804138184,
+      "learning_rate": 0.00014819875776397515,
+      "loss": 0.1247,
+      "step": 433
+    },
+    {
+      "epoch": 4.018518518518518,
+      "grad_norm": 8.481317520141602,
+      "learning_rate": 0.0001480745341614907,
+      "loss": 0.4654,
+      "step": 434
+    },
+    {
+      "epoch": 4.027777777777778,
+      "grad_norm": 5.601878643035889,
+      "learning_rate": 0.0001479503105590062,
+      "loss": 0.3271,
+      "step": 435
+    },
+    {
+      "epoch": 4.037037037037037,
+      "grad_norm": 2.543224573135376,
+      "learning_rate": 0.00014782608695652173,
+      "loss": 0.0809,
+      "step": 436
+    },
+    {
+      "epoch": 4.046296296296297,
+      "grad_norm": 4.2349395751953125,
+      "learning_rate": 0.0001477018633540373,
+      "loss": 0.1481,
+      "step": 437
+    },
+    {
+      "epoch": 4.055555555555555,
+      "grad_norm": 1.989195704460144,
+      "learning_rate": 0.0001475776397515528,
+      "loss": 0.0472,
+      "step": 438
+    },
+    {
+      "epoch": 4.064814814814815,
+      "grad_norm": 3.653602123260498,
+      "learning_rate": 0.00014745341614906833,
+      "loss": 0.1803,
+      "step": 439
+    },
+    {
+      "epoch": 4.074074074074074,
+      "grad_norm": 5.280818939208984,
+      "learning_rate": 0.00014732919254658387,
+      "loss": 0.1864,
+      "step": 440
+    },
+    {
+      "epoch": 4.083333333333333,
+      "grad_norm": 2.3477160930633545,
+      "learning_rate": 0.0001472049689440994,
+      "loss": 0.0735,
+      "step": 441
+    },
+    {
+      "epoch": 4.092592592592593,
+      "grad_norm": 7.4617414474487305,
+      "learning_rate": 0.0001470807453416149,
+      "loss": 0.281,
+      "step": 442
+    },
+    {
+      "epoch": 4.101851851851852,
+      "grad_norm": 4.8935546875,
+      "learning_rate": 0.00014695652173913045,
+      "loss": 0.257,
+      "step": 443
+    },
+    {
+      "epoch": 4.111111111111111,
+      "grad_norm": 7.706578731536865,
+      "learning_rate": 0.00014683229813664597,
+      "loss": 0.3459,
+      "step": 444
+    },
+    {
+      "epoch": 4.12037037037037,
+      "grad_norm": 2.809396266937256,
+      "learning_rate": 0.00014670807453416148,
+      "loss": 0.0632,
+      "step": 445
+    },
+    {
+      "epoch": 4.12962962962963,
+      "grad_norm": 6.64403772354126,
+      "learning_rate": 0.00014658385093167703,
+      "loss": 0.3485,
+      "step": 446
+    },
+    {
+      "epoch": 4.138888888888889,
+      "grad_norm": 5.006530284881592,
+      "learning_rate": 0.00014645962732919257,
+      "loss": 0.2197,
+      "step": 447
+    },
+    {
+      "epoch": 4.148148148148148,
+      "grad_norm": 2.446791410446167,
+      "learning_rate": 0.00014633540372670808,
+      "loss": 0.0841,
+      "step": 448
+    },
+    {
+      "epoch": 4.157407407407407,
+      "grad_norm": 2.2801015377044678,
+      "learning_rate": 0.0001462111801242236,
+      "loss": 0.0983,
+      "step": 449
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 3.368048667907715,
+      "learning_rate": 0.00014608695652173914,
+      "loss": 0.1286,
+      "step": 450
+    },
+    {
+      "epoch": 4.175925925925926,
+      "grad_norm": 6.958519458770752,
+      "learning_rate": 0.00014596273291925466,
+      "loss": 0.2685,
+      "step": 451
+    },
+    {
+      "epoch": 4.185185185185185,
+      "grad_norm": 2.6293435096740723,
+      "learning_rate": 0.00014583850931677018,
+      "loss": 0.118,
+      "step": 452
+    },
+    {
+      "epoch": 4.194444444444445,
+      "grad_norm": 2.426443338394165,
+      "learning_rate": 0.00014571428571428572,
+      "loss": 0.0909,
+      "step": 453
+    },
+    {
+      "epoch": 4.203703703703703,
+      "grad_norm": 3.9959075450897217,
+      "learning_rate": 0.00014559006211180124,
+      "loss": 0.1808,
+      "step": 454
+    },
+    {
+      "epoch": 4.212962962962963,
+      "grad_norm": 4.146143436431885,
+      "learning_rate": 0.00014546583850931678,
+      "loss": 0.2056,
+      "step": 455
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 6.517556190490723,
+      "learning_rate": 0.00014534161490683232,
+      "loss": 0.2656,
+      "step": 456
+    },
+    {
+      "epoch": 4.231481481481482,
+      "grad_norm": 4.0375237464904785,
+      "learning_rate": 0.00014521739130434784,
+      "loss": 0.2048,
+      "step": 457
+    },
+    {
+      "epoch": 4.2407407407407405,
+      "grad_norm": 3.206602096557617,
+      "learning_rate": 0.00014509316770186336,
+      "loss": 0.1343,
+      "step": 458
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 4.9624786376953125,
+      "learning_rate": 0.0001449689440993789,
+      "loss": 0.2606,
+      "step": 459
+    },
+    {
+      "epoch": 4.2592592592592595,
+      "grad_norm": 4.6561455726623535,
+      "learning_rate": 0.00014484472049689442,
+      "loss": 0.1632,
+      "step": 460
+    },
+    {
+      "epoch": 4.268518518518518,
+      "grad_norm": 3.209395408630371,
+      "learning_rate": 0.00014472049689440993,
+      "loss": 0.1272,
+      "step": 461
+    },
+    {
+      "epoch": 4.277777777777778,
+      "grad_norm": 5.021265983581543,
+      "learning_rate": 0.00014459627329192548,
+      "loss": 0.2286,
+      "step": 462
+    },
+    {
+      "epoch": 4.287037037037037,
+      "grad_norm": 7.752100467681885,
+      "learning_rate": 0.000144472049689441,
+      "loss": 0.4169,
+      "step": 463
+    },
+    {
+      "epoch": 4.296296296296296,
+      "grad_norm": 7.6476850509643555,
+      "learning_rate": 0.00014434782608695654,
+      "loss": 0.276,
+      "step": 464
+    },
+    {
+      "epoch": 4.305555555555555,
+      "grad_norm": 4.68563175201416,
+      "learning_rate": 0.00014422360248447205,
+      "loss": 0.113,
+      "step": 465
+    },
+    {
+      "epoch": 4.314814814814815,
+      "grad_norm": 2.981081485748291,
+      "learning_rate": 0.0001440993788819876,
+      "loss": 0.1054,
+      "step": 466
+    },
+    {
+      "epoch": 4.324074074074074,
+      "grad_norm": 4.140884876251221,
+      "learning_rate": 0.0001439751552795031,
+      "loss": 0.1545,
+      "step": 467
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 4.200994968414307,
+      "learning_rate": 0.00014385093167701863,
+      "loss": 0.1325,
+      "step": 468
+    },
+    {
+      "epoch": 4.342592592592593,
+      "grad_norm": 4.808220386505127,
+      "learning_rate": 0.00014372670807453417,
+      "loss": 0.2498,
+      "step": 469
+    },
+    {
+      "epoch": 4.351851851851852,
+      "grad_norm": 3.549621820449829,
+      "learning_rate": 0.0001436024844720497,
+      "loss": 0.1205,
+      "step": 470
+    },
+    {
+      "epoch": 4.361111111111111,
+      "grad_norm": 4.576113224029541,
+      "learning_rate": 0.0001434782608695652,
+      "loss": 0.1623,
+      "step": 471
+    },
+    {
+      "epoch": 4.37037037037037,
+      "grad_norm": 9.293719291687012,
+      "learning_rate": 0.00014335403726708075,
+      "loss": 0.5254,
+      "step": 472
+    },
+    {
+      "epoch": 4.37962962962963,
+      "grad_norm": 4.397647380828857,
+      "learning_rate": 0.0001432298136645963,
+      "loss": 0.1521,
+      "step": 473
+    },
+    {
+      "epoch": 4.388888888888889,
+      "grad_norm": 3.319638252258301,
+      "learning_rate": 0.0001431055900621118,
+      "loss": 0.1145,
+      "step": 474
+    },
+    {
+      "epoch": 4.398148148148148,
+      "grad_norm": 6.22605562210083,
+      "learning_rate": 0.00014298136645962735,
+      "loss": 0.2767,
+      "step": 475
+    },
+    {
+      "epoch": 4.407407407407407,
+      "grad_norm": 8.628198623657227,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 0.502,
+      "step": 476
+    },
+    {
+      "epoch": 4.416666666666667,
+      "grad_norm": 5.40584659576416,
+      "learning_rate": 0.00014273291925465838,
+      "loss": 0.2086,
+      "step": 477
+    },
+    {
+      "epoch": 4.425925925925926,
+      "grad_norm": 7.446430683135986,
+      "learning_rate": 0.00014260869565217393,
+      "loss": 0.3835,
+      "step": 478
+    },
+    {
+      "epoch": 4.435185185185185,
+      "grad_norm": 3.44232439994812,
+      "learning_rate": 0.00014248447204968944,
+      "loss": 0.1593,
+      "step": 479
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 5.173421382904053,
+      "learning_rate": 0.00014236024844720496,
+      "loss": 0.1365,
+      "step": 480
+    },
+    {
+      "epoch": 4.453703703703704,
+      "grad_norm": 4.879393100738525,
+      "learning_rate": 0.0001422360248447205,
+      "loss": 0.1906,
+      "step": 481
+    },
+    {
+      "epoch": 4.462962962962963,
+      "grad_norm": 4.784718036651611,
+      "learning_rate": 0.00014211180124223605,
+      "loss": 0.2266,
+      "step": 482
+    },
+    {
+      "epoch": 4.472222222222222,
+      "grad_norm": 2.3085122108459473,
+      "learning_rate": 0.00014198757763975156,
+      "loss": 0.0864,
+      "step": 483
+    },
+    {
+      "epoch": 4.481481481481482,
+      "grad_norm": 6.118433475494385,
+      "learning_rate": 0.00014186335403726708,
+      "loss": 0.2416,
+      "step": 484
+    },
+    {
+      "epoch": 4.4907407407407405,
+      "grad_norm": 4.493291854858398,
+      "learning_rate": 0.00014173913043478262,
+      "loss": 0.174,
+      "step": 485
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 3.1890506744384766,
+      "learning_rate": 0.00014161490683229814,
+      "loss": 0.1136,
+      "step": 486
+    },
+    {
+      "epoch": 4.5092592592592595,
+      "grad_norm": 4.983994007110596,
+      "learning_rate": 0.00014149068322981368,
+      "loss": 0.2204,
+      "step": 487
+    },
+    {
+      "epoch": 4.518518518518518,
+      "grad_norm": 4.625283718109131,
+      "learning_rate": 0.0001413664596273292,
+      "loss": 0.2092,
+      "step": 488
+    },
+    {
+      "epoch": 4.527777777777778,
+      "grad_norm": 5.007232666015625,
+      "learning_rate": 0.00014124223602484472,
+      "loss": 0.1789,
+      "step": 489
+    },
+    {
+      "epoch": 4.537037037037037,
+      "grad_norm": 3.370143413543701,
+      "learning_rate": 0.00014111801242236026,
+      "loss": 0.1262,
+      "step": 490
+    },
+    {
+      "epoch": 4.546296296296296,
+      "grad_norm": 5.218165874481201,
+      "learning_rate": 0.0001409937888198758,
+      "loss": 0.1389,
+      "step": 491
+    },
+    {
+      "epoch": 4.555555555555555,
+      "grad_norm": 7.572610378265381,
+      "learning_rate": 0.00014086956521739132,
+      "loss": 0.4608,
+      "step": 492
+    },
+    {
+      "epoch": 4.564814814814815,
+      "grad_norm": 8.756231307983398,
+      "learning_rate": 0.00014074534161490684,
+      "loss": 0.4648,
+      "step": 493
+    },
+    {
+      "epoch": 4.574074074074074,
+      "grad_norm": 3.8681464195251465,
+      "learning_rate": 0.00014062111801242238,
+      "loss": 0.1241,
+      "step": 494
+    },
+    {
+      "epoch": 4.583333333333333,
+      "grad_norm": 2.857240915298462,
+      "learning_rate": 0.0001404968944099379,
+      "loss": 0.1154,
+      "step": 495
+    },
+    {
+      "epoch": 4.592592592592593,
+      "grad_norm": 3.749955654144287,
+      "learning_rate": 0.0001403726708074534,
+      "loss": 0.1234,
+      "step": 496
+    },
+    {
+      "epoch": 4.601851851851852,
+      "grad_norm": 7.389172077178955,
+      "learning_rate": 0.00014024844720496896,
+      "loss": 0.341,
+      "step": 497
+    },
+    {
+      "epoch": 4.611111111111111,
+      "grad_norm": 2.2326736450195312,
+      "learning_rate": 0.00014012422360248447,
+      "loss": 0.0911,
+      "step": 498
+    },
+    {
+      "epoch": 4.62037037037037,
+      "grad_norm": 5.606997489929199,
+      "learning_rate": 0.00014,
+      "loss": 0.2355,
+      "step": 499
+    },
+    {
+      "epoch": 4.62962962962963,
+      "grad_norm": 4.593019008636475,
+      "learning_rate": 0.00013987577639751556,
+      "loss": 0.2544,
+      "step": 500
+    },
+    {
+      "epoch": 4.638888888888889,
+      "grad_norm": 4.68584680557251,
+      "learning_rate": 0.00013975155279503107,
+      "loss": 0.1838,
+      "step": 501
+    },
+    {
+      "epoch": 4.648148148148148,
+      "grad_norm": 5.606670379638672,
+      "learning_rate": 0.0001396273291925466,
+      "loss": 0.2747,
+      "step": 502
+    },
+    {
+      "epoch": 4.657407407407407,
+      "grad_norm": 6.980198860168457,
+      "learning_rate": 0.00013950310559006213,
+      "loss": 0.3134,
+      "step": 503
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 5.77571439743042,
+      "learning_rate": 0.00013937888198757765,
+      "loss": 0.3051,
+      "step": 504
+    },
+    {
+      "epoch": 4.675925925925926,
+      "grad_norm": 8.558889389038086,
+      "learning_rate": 0.00013925465838509317,
+      "loss": 0.5136,
+      "step": 505
+    },
+    {
+      "epoch": 4.685185185185185,
+      "grad_norm": 6.231207847595215,
+      "learning_rate": 0.0001391304347826087,
+      "loss": 0.2484,
+      "step": 506
+    },
+    {
+      "epoch": 4.694444444444445,
+      "grad_norm": 4.745160102844238,
+      "learning_rate": 0.00013900621118012423,
+      "loss": 0.2174,
+      "step": 507
+    },
+    {
+      "epoch": 4.703703703703704,
+      "grad_norm": 6.144798278808594,
+      "learning_rate": 0.00013888198757763974,
+      "loss": 0.2192,
+      "step": 508
+    },
+    {
+      "epoch": 4.712962962962963,
+      "grad_norm": 6.802698612213135,
+      "learning_rate": 0.0001387577639751553,
+      "loss": 0.2729,
+      "step": 509
+    },
+    {
+      "epoch": 4.722222222222222,
+      "grad_norm": 5.089367389678955,
+      "learning_rate": 0.00013863354037267083,
+      "loss": 0.1986,
+      "step": 510
+    },
+    {
+      "epoch": 4.731481481481482,
+      "grad_norm": 2.46826434135437,
+      "learning_rate": 0.00013850931677018635,
+      "loss": 0.0684,
+      "step": 511
+    },
+    {
+      "epoch": 4.7407407407407405,
+      "grad_norm": 3.246137857437134,
+      "learning_rate": 0.00013838509316770186,
+      "loss": 0.106,
+      "step": 512
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 4.2868266105651855,
+      "learning_rate": 0.0001382608695652174,
+      "loss": 0.1894,
+      "step": 513
+    },
+    {
+      "epoch": 4.7592592592592595,
+      "grad_norm": 6.733697891235352,
+      "learning_rate": 0.00013813664596273292,
+      "loss": 0.2563,
+      "step": 514
+    },
+    {
+      "epoch": 4.768518518518518,
+      "grad_norm": 5.327560901641846,
+      "learning_rate": 0.00013801242236024844,
+      "loss": 0.2412,
+      "step": 515
+    },
+    {
+      "epoch": 4.777777777777778,
+      "grad_norm": 2.9705471992492676,
+      "learning_rate": 0.00013788819875776398,
+      "loss": 0.0845,
+      "step": 516
+    },
+    {
+      "epoch": 4.787037037037037,
+      "grad_norm": 2.2280571460723877,
+      "learning_rate": 0.0001377639751552795,
+      "loss": 0.0572,
+      "step": 517
+    },
+    {
+      "epoch": 4.796296296296296,
+      "grad_norm": 13.944962501525879,
+      "learning_rate": 0.00013763975155279504,
+      "loss": 0.4604,
+      "step": 518
+    },
+    {
+      "epoch": 4.805555555555555,
+      "grad_norm": 5.684059143066406,
+      "learning_rate": 0.00013751552795031059,
+      "loss": 0.184,
+      "step": 519
+    },
+    {
+      "epoch": 4.814814814814815,
+      "grad_norm": 3.9679698944091797,
+      "learning_rate": 0.0001373913043478261,
+      "loss": 0.1452,
+      "step": 520
+    },
+    {
+      "epoch": 4.824074074074074,
+      "grad_norm": 1.851914405822754,
+      "learning_rate": 0.00013726708074534162,
+      "loss": 0.065,
+      "step": 521
+    },
+    {
+      "epoch": 4.833333333333333,
+      "grad_norm": 6.1766133308410645,
+      "learning_rate": 0.00013714285714285716,
+      "loss": 0.2719,
+      "step": 522
+    },
+    {
+      "epoch": 4.842592592592593,
+      "grad_norm": 6.558730602264404,
+      "learning_rate": 0.00013701863354037268,
+      "loss": 0.3054,
+      "step": 523
+    },
+    {
+      "epoch": 4.851851851851852,
+      "grad_norm": 2.0593979358673096,
+      "learning_rate": 0.0001368944099378882,
+      "loss": 0.0425,
+      "step": 524
+    },
+    {
+      "epoch": 4.861111111111111,
+      "grad_norm": 4.957576274871826,
+      "learning_rate": 0.00013677018633540374,
+      "loss": 0.2348,
+      "step": 525
+    },
+    {
+      "epoch": 4.87037037037037,
+      "grad_norm": 5.931738376617432,
+      "learning_rate": 0.00013664596273291925,
+      "loss": 0.2931,
+      "step": 526
+    },
+    {
+      "epoch": 4.87962962962963,
+      "grad_norm": 8.989755630493164,
+      "learning_rate": 0.00013652173913043477,
+      "loss": 0.513,
+      "step": 527
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 6.920965671539307,
+      "learning_rate": 0.00013639751552795031,
+      "loss": 0.3348,
+      "step": 528
+    },
+    {
+      "epoch": 4.898148148148148,
+      "grad_norm": 2.210482358932495,
+      "learning_rate": 0.00013627329192546586,
+      "loss": 0.1061,
+      "step": 529
+    },
+    {
+      "epoch": 4.907407407407407,
+      "grad_norm": 5.41426420211792,
+      "learning_rate": 0.00013614906832298137,
+      "loss": 0.1837,
+      "step": 530
+    },
+    {
+      "epoch": 4.916666666666667,
+      "grad_norm": 8.46068000793457,
+      "learning_rate": 0.0001360248447204969,
+      "loss": 0.1896,
+      "step": 531
+    },
+    {
+      "epoch": 4.925925925925926,
+      "grad_norm": 5.2427144050598145,
+      "learning_rate": 0.00013590062111801243,
+      "loss": 0.2031,
+      "step": 532
+    },
+    {
+      "epoch": 4.935185185185185,
+      "grad_norm": 5.652583599090576,
+      "learning_rate": 0.00013577639751552795,
+      "loss": 0.209,
+      "step": 533
+    },
+    {
+      "epoch": 4.944444444444445,
+      "grad_norm": 3.267430543899536,
+      "learning_rate": 0.00013565217391304347,
+      "loss": 0.1083,
+      "step": 534
+    },
+    {
+      "epoch": 4.953703703703704,
+      "grad_norm": 7.333652973175049,
+      "learning_rate": 0.000135527950310559,
+      "loss": 0.191,
+      "step": 535
+    },
+    {
+      "epoch": 4.962962962962963,
+      "grad_norm": 2.8522582054138184,
+      "learning_rate": 0.00013540372670807453,
+      "loss": 0.0938,
+      "step": 536
+    },
+    {
+      "epoch": 4.972222222222222,
+      "grad_norm": 4.617889881134033,
+      "learning_rate": 0.00013527950310559007,
+      "loss": 0.1061,
+      "step": 537
+    },
+    {
+      "epoch": 4.981481481481482,
+      "grad_norm": 7.658900737762451,
+      "learning_rate": 0.0001351552795031056,
+      "loss": 0.4003,
+      "step": 538
+    },
+    {
+      "epoch": 4.9907407407407405,
+      "grad_norm": 7.5713653564453125,
+      "learning_rate": 0.00013503105590062113,
+      "loss": 0.3886,
+      "step": 539
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 9.257078170776367,
+      "learning_rate": 0.00013490683229813665,
+      "loss": 0.6839,
+      "step": 540
+    },
+    {
+      "epoch": 5.0092592592592595,
+      "grad_norm": 6.371257305145264,
+      "learning_rate": 0.0001347826086956522,
+      "loss": 0.314,
+      "step": 541
+    },
+    {
+      "epoch": 5.018518518518518,
+      "grad_norm": 2.907931327819824,
+      "learning_rate": 0.0001346583850931677,
+      "loss": 0.0629,
+      "step": 542
+    },
+    {
+      "epoch": 5.027777777777778,
+      "grad_norm": 4.068233966827393,
+      "learning_rate": 0.00013453416149068322,
+      "loss": 0.1179,
+      "step": 543
+    },
+    {
+      "epoch": 5.037037037037037,
+      "grad_norm": 2.4700989723205566,
+      "learning_rate": 0.00013440993788819877,
+      "loss": 0.0794,
+      "step": 544
+    },
+    {
+      "epoch": 5.046296296296297,
+      "grad_norm": 3.4381418228149414,
+      "learning_rate": 0.00013428571428571428,
+      "loss": 0.1431,
+      "step": 545
+    },
+    {
+      "epoch": 5.055555555555555,
+      "grad_norm": 0.883247971534729,
+      "learning_rate": 0.00013416149068322983,
+      "loss": 0.0141,
+      "step": 546
+    },
+    {
+      "epoch": 5.064814814814815,
+      "grad_norm": 4.4683518409729,
+      "learning_rate": 0.00013403726708074534,
+      "loss": 0.0748,
+      "step": 547
+    },
+    {
+      "epoch": 5.074074074074074,
+      "grad_norm": 5.847491264343262,
+      "learning_rate": 0.00013391304347826088,
+      "loss": 0.2232,
+      "step": 548
+    },
+    {
+      "epoch": 5.083333333333333,
+      "grad_norm": 1.6964125633239746,
+      "learning_rate": 0.0001337888198757764,
+      "loss": 0.0555,
+      "step": 549
+    },
+    {
+      "epoch": 5.092592592592593,
+      "grad_norm": 2.773496150970459,
+      "learning_rate": 0.00013366459627329192,
+      "loss": 0.0791,
+      "step": 550
+    },
+    {
+      "epoch": 5.101851851851852,
+      "grad_norm": 4.492348670959473,
+      "learning_rate": 0.00013354037267080746,
+      "loss": 0.1072,
+      "step": 551
+    },
+    {
+      "epoch": 5.111111111111111,
+      "grad_norm": 2.172301769256592,
+      "learning_rate": 0.00013341614906832298,
+      "loss": 0.0527,
+      "step": 552
+    },
+    {
+      "epoch": 5.12037037037037,
+      "grad_norm": 2.18371319770813,
+      "learning_rate": 0.00013329192546583852,
+      "loss": 0.0565,
+      "step": 553
+    },
+    {
+      "epoch": 5.12962962962963,
+      "grad_norm": 5.9107747077941895,
+      "learning_rate": 0.00013316770186335404,
+      "loss": 0.1583,
+      "step": 554
+    },
+    {
+      "epoch": 5.138888888888889,
+      "grad_norm": 4.741661548614502,
+      "learning_rate": 0.00013304347826086958,
+      "loss": 0.1513,
+      "step": 555
+    },
+    {
+      "epoch": 5.148148148148148,
+      "grad_norm": 4.259805202484131,
+      "learning_rate": 0.0001329192546583851,
+      "loss": 0.1908,
+      "step": 556
+    },
+    {
+      "epoch": 5.157407407407407,
+      "grad_norm": 6.191773414611816,
+      "learning_rate": 0.00013279503105590064,
+      "loss": 0.2065,
+      "step": 557
+    },
+    {
+      "epoch": 5.166666666666667,
+      "grad_norm": 2.9694433212280273,
+      "learning_rate": 0.00013267080745341616,
+      "loss": 0.0597,
+      "step": 558
+    },
+    {
+      "epoch": 5.175925925925926,
+      "grad_norm": 3.9127731323242188,
+      "learning_rate": 0.00013254658385093167,
+      "loss": 0.1278,
+      "step": 559
+    },
+    {
+      "epoch": 5.185185185185185,
+      "grad_norm": 4.475368499755859,
+      "learning_rate": 0.00013242236024844722,
+      "loss": 0.1604,
+      "step": 560
+    },
+    {
+      "epoch": 5.194444444444445,
+      "grad_norm": 8.596575736999512,
+      "learning_rate": 0.00013229813664596273,
+      "loss": 0.2838,
+      "step": 561
+    },
+    {
+      "epoch": 5.203703703703703,
+      "grad_norm": 3.6132686138153076,
+      "learning_rate": 0.00013217391304347825,
+      "loss": 0.0587,
+      "step": 562
+    },
+    {
+      "epoch": 5.212962962962963,
+      "grad_norm": 9.223053932189941,
+      "learning_rate": 0.0001320496894409938,
+      "loss": 0.2524,
+      "step": 563
+    },
+    {
+      "epoch": 5.222222222222222,
+      "grad_norm": 1.8150829076766968,
+      "learning_rate": 0.00013192546583850934,
+      "loss": 0.0615,
+      "step": 564
+    },
+    {
+      "epoch": 5.231481481481482,
+      "grad_norm": 4.499701976776123,
+      "learning_rate": 0.00013180124223602485,
+      "loss": 0.1889,
+      "step": 565
+    },
+    {
+      "epoch": 5.2407407407407405,
+      "grad_norm": 3.4126698970794678,
+      "learning_rate": 0.0001316770186335404,
+      "loss": 0.1661,
+      "step": 566
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.3917155265808105,
+      "learning_rate": 0.0001315527950310559,
+      "loss": 0.0479,
+      "step": 567
+    },
+    {
+      "epoch": 5.2592592592592595,
+      "grad_norm": 4.621674060821533,
+      "learning_rate": 0.00013142857142857143,
+      "loss": 0.1365,
+      "step": 568
+    },
+    {
+      "epoch": 5.268518518518518,
+      "grad_norm": 2.671595335006714,
+      "learning_rate": 0.00013130434782608697,
+      "loss": 0.0965,
+      "step": 569
+    },
+    {
+      "epoch": 5.277777777777778,
+      "grad_norm": 2.2948710918426514,
+      "learning_rate": 0.0001311801242236025,
+      "loss": 0.057,
+      "step": 570
+    },
+    {
+      "epoch": 5.287037037037037,
+      "grad_norm": 3.181098699569702,
+      "learning_rate": 0.000131055900621118,
+      "loss": 0.0844,
+      "step": 571
+    },
+    {
+      "epoch": 5.296296296296296,
+      "grad_norm": 1.5297895669937134,
+      "learning_rate": 0.00013093167701863355,
+      "loss": 0.09,
+      "step": 572
+    },
+    {
+      "epoch": 5.305555555555555,
+      "grad_norm": 5.39132022857666,
+      "learning_rate": 0.0001308074534161491,
+      "loss": 0.1515,
+      "step": 573
+    },
+    {
+      "epoch": 5.314814814814815,
+      "grad_norm": 3.3082971572875977,
+      "learning_rate": 0.0001306832298136646,
+      "loss": 0.1473,
+      "step": 574
+    },
+    {
+      "epoch": 5.324074074074074,
+      "grad_norm": 4.612159252166748,
+      "learning_rate": 0.00013055900621118012,
+      "loss": 0.2653,
+      "step": 575
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 8.018484115600586,
+      "learning_rate": 0.00013043478260869567,
+      "loss": 0.4067,
+      "step": 576
+    },
+    {
+      "epoch": 5.342592592592593,
+      "grad_norm": 2.2573232650756836,
+      "learning_rate": 0.00013031055900621118,
+      "loss": 0.0973,
+      "step": 577
+    },
+    {
+      "epoch": 5.351851851851852,
+      "grad_norm": 3.4523372650146484,
+      "learning_rate": 0.0001301863354037267,
+      "loss": 0.1417,
+      "step": 578
+    },
+    {
+      "epoch": 5.361111111111111,
+      "grad_norm": 3.2705371379852295,
+      "learning_rate": 0.00013006211180124224,
+      "loss": 0.1203,
+      "step": 579
+    },
+    {
+      "epoch": 5.37037037037037,
+      "grad_norm": 1.8731480836868286,
+      "learning_rate": 0.00012993788819875776,
+      "loss": 0.0674,
+      "step": 580
+    },
+    {
+      "epoch": 5.37962962962963,
+      "grad_norm": 4.4362101554870605,
+      "learning_rate": 0.00012981366459627328,
+      "loss": 0.1952,
+      "step": 581
+    },
+    {
+      "epoch": 5.388888888888889,
+      "grad_norm": 4.875374794006348,
+      "learning_rate": 0.00012968944099378885,
+      "loss": 0.1,
+      "step": 582
+    },
+    {
+      "epoch": 5.398148148148148,
+      "grad_norm": 5.201648235321045,
+      "learning_rate": 0.00012956521739130436,
+      "loss": 0.277,
+      "step": 583
+    },
+    {
+      "epoch": 5.407407407407407,
+      "grad_norm": 2.5301337242126465,
+      "learning_rate": 0.00012944099378881988,
+      "loss": 0.0705,
+      "step": 584
+    },
+    {
+      "epoch": 5.416666666666667,
+      "grad_norm": 3.144968271255493,
+      "learning_rate": 0.00012931677018633542,
+      "loss": 0.0667,
+      "step": 585
+    },
+    {
+      "epoch": 5.425925925925926,
+      "grad_norm": 2.509068489074707,
+      "learning_rate": 0.00012919254658385094,
+      "loss": 0.081,
+      "step": 586
+    },
+    {
+      "epoch": 5.435185185185185,
+      "grad_norm": 7.523172855377197,
+      "learning_rate": 0.00012906832298136646,
+      "loss": 0.2569,
+      "step": 587
+    },
+    {
+      "epoch": 5.444444444444445,
+      "grad_norm": 3.6031181812286377,
+      "learning_rate": 0.000128944099378882,
+      "loss": 0.1324,
+      "step": 588
+    },
+    {
+      "epoch": 5.453703703703704,
+      "grad_norm": 3.194385528564453,
+      "learning_rate": 0.00012881987577639752,
+      "loss": 0.113,
+      "step": 589
+    },
+    {
+      "epoch": 5.462962962962963,
+      "grad_norm": 1.4715389013290405,
+      "learning_rate": 0.00012869565217391303,
+      "loss": 0.073,
+      "step": 590
+    },
+    {
+      "epoch": 5.472222222222222,
+      "grad_norm": 2.3477694988250732,
+      "learning_rate": 0.00012857142857142858,
+      "loss": 0.0708,
+      "step": 591
+    },
+    {
+      "epoch": 5.481481481481482,
+      "grad_norm": 1.5223402976989746,
+      "learning_rate": 0.00012844720496894412,
+      "loss": 0.0464,
+      "step": 592
+    },
+    {
+      "epoch": 5.4907407407407405,
+      "grad_norm": 6.1966352462768555,
+      "learning_rate": 0.00012832298136645964,
+      "loss": 0.2479,
+      "step": 593
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 10.558868408203125,
+      "learning_rate": 0.00012819875776397515,
+      "loss": 0.5108,
+      "step": 594
+    },
+    {
+      "epoch": 5.5092592592592595,
+      "grad_norm": 5.110361576080322,
+      "learning_rate": 0.0001280745341614907,
+      "loss": 0.2486,
+      "step": 595
+    },
+    {
+      "epoch": 5.518518518518518,
+      "grad_norm": 1.4477581977844238,
+      "learning_rate": 0.0001279503105590062,
+      "loss": 0.0472,
+      "step": 596
+    },
+    {
+      "epoch": 5.527777777777778,
+      "grad_norm": 4.302581787109375,
+      "learning_rate": 0.00012782608695652173,
+      "loss": 0.2134,
+      "step": 597
+    },
+    {
+      "epoch": 5.537037037037037,
+      "grad_norm": 4.25509786605835,
+      "learning_rate": 0.00012770186335403727,
+      "loss": 0.1519,
+      "step": 598
+    },
+    {
+      "epoch": 5.546296296296296,
+      "grad_norm": 3.6477010250091553,
+      "learning_rate": 0.0001275776397515528,
+      "loss": 0.1756,
+      "step": 599
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 3.412436008453369,
+      "learning_rate": 0.00012745341614906833,
+      "loss": 0.1473,
+      "step": 600
+    },
+    {
+      "epoch": 5.564814814814815,
+      "grad_norm": 2.7615256309509277,
+      "learning_rate": 0.00012732919254658387,
+      "loss": 0.0985,
+      "step": 601
+    },
+    {
+      "epoch": 5.574074074074074,
+      "grad_norm": 4.826498985290527,
+      "learning_rate": 0.0001272049689440994,
+      "loss": 0.1564,
+      "step": 602
+    },
+    {
+      "epoch": 5.583333333333333,
+      "grad_norm": 4.768690586090088,
+      "learning_rate": 0.0001270807453416149,
+      "loss": 0.2187,
+      "step": 603
+    },
+    {
+      "epoch": 5.592592592592593,
+      "grad_norm": 4.710096836090088,
+      "learning_rate": 0.00012695652173913045,
+      "loss": 0.0982,
+      "step": 604
+    },
+    {
+      "epoch": 5.601851851851852,
+      "grad_norm": 3.721019744873047,
+      "learning_rate": 0.00012683229813664597,
+      "loss": 0.1331,
+      "step": 605
+    },
+    {
+      "epoch": 5.611111111111111,
+      "grad_norm": 10.823250770568848,
+      "learning_rate": 0.00012670807453416148,
+      "loss": 0.2592,
+      "step": 606
+    },
+    {
+      "epoch": 5.62037037037037,
+      "grad_norm": 2.030266761779785,
+      "learning_rate": 0.00012658385093167703,
+      "loss": 0.0532,
+      "step": 607
+    },
+    {
+      "epoch": 5.62962962962963,
+      "grad_norm": 4.477989673614502,
+      "learning_rate": 0.00012645962732919254,
+      "loss": 0.2144,
+      "step": 608
+    },
+    {
+      "epoch": 5.638888888888889,
+      "grad_norm": 4.200387954711914,
+      "learning_rate": 0.0001263354037267081,
+      "loss": 0.1709,
+      "step": 609
+    },
+    {
+      "epoch": 5.648148148148148,
+      "grad_norm": 6.254189968109131,
+      "learning_rate": 0.0001262111801242236,
+      "loss": 0.3155,
+      "step": 610
+    },
+    {
+      "epoch": 5.657407407407407,
+      "grad_norm": 8.545554161071777,
+      "learning_rate": 0.00012608695652173915,
+      "loss": 0.2297,
+      "step": 611
+    },
+    {
+      "epoch": 5.666666666666667,
+      "grad_norm": 6.469607830047607,
+      "learning_rate": 0.00012596273291925466,
+      "loss": 0.2023,
+      "step": 612
+    },
+    {
+      "epoch": 5.675925925925926,
+      "grad_norm": 4.819572448730469,
+      "learning_rate": 0.00012583850931677018,
+      "loss": 0.1118,
+      "step": 613
+    },
+    {
+      "epoch": 5.685185185185185,
+      "grad_norm": 5.746372699737549,
+      "learning_rate": 0.00012571428571428572,
+      "loss": 0.2105,
+      "step": 614
+    },
+    {
+      "epoch": 5.694444444444445,
+      "grad_norm": 2.2594149112701416,
+      "learning_rate": 0.00012559006211180124,
+      "loss": 0.0692,
+      "step": 615
+    },
+    {
+      "epoch": 5.703703703703704,
+      "grad_norm": 3.391378402709961,
+      "learning_rate": 0.00012546583850931676,
+      "loss": 0.1631,
+      "step": 616
+    },
+    {
+      "epoch": 5.712962962962963,
+      "grad_norm": 2.416281223297119,
+      "learning_rate": 0.0001253416149068323,
+      "loss": 0.1129,
+      "step": 617
+    },
+    {
+      "epoch": 5.722222222222222,
+      "grad_norm": 6.688156604766846,
+      "learning_rate": 0.00012521739130434784,
+      "loss": 0.1804,
+      "step": 618
+    },
+    {
+      "epoch": 5.731481481481482,
+      "grad_norm": 5.007920265197754,
+      "learning_rate": 0.00012509316770186336,
+      "loss": 0.3592,
+      "step": 619
+    },
+    {
+      "epoch": 5.7407407407407405,
+      "grad_norm": 4.108309268951416,
+      "learning_rate": 0.0001249689440993789,
+      "loss": 0.1732,
+      "step": 620
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 3.446714401245117,
+      "learning_rate": 0.00012484472049689442,
+      "loss": 0.1708,
+      "step": 621
+    },
+    {
+      "epoch": 5.7592592592592595,
+      "grad_norm": 4.069822311401367,
+      "learning_rate": 0.00012472049689440993,
+      "loss": 0.198,
+      "step": 622
+    },
+    {
+      "epoch": 5.768518518518518,
+      "grad_norm": 4.49385929107666,
+      "learning_rate": 0.00012459627329192548,
+      "loss": 0.2133,
+      "step": 623
+    },
+    {
+      "epoch": 5.777777777777778,
+      "grad_norm": 2.8163678646087646,
+      "learning_rate": 0.000124472049689441,
+      "loss": 0.1077,
+      "step": 624
+    },
+    {
+      "epoch": 5.787037037037037,
+      "grad_norm": 2.6389107704162598,
+      "learning_rate": 0.0001243478260869565,
+      "loss": 0.087,
+      "step": 625
+    },
+    {
+      "epoch": 5.796296296296296,
+      "grad_norm": 4.10544490814209,
+      "learning_rate": 0.00012422360248447205,
+      "loss": 0.0834,
+      "step": 626
+    },
+    {
+      "epoch": 5.805555555555555,
+      "grad_norm": 1.4941080808639526,
+      "learning_rate": 0.0001240993788819876,
+      "loss": 0.0485,
+      "step": 627
+    },
+    {
+      "epoch": 5.814814814814815,
+      "grad_norm": 7.434000015258789,
+      "learning_rate": 0.00012397515527950311,
+      "loss": 0.3163,
+      "step": 628
+    },
+    {
+      "epoch": 5.824074074074074,
+      "grad_norm": 5.449849605560303,
+      "learning_rate": 0.00012385093167701866,
+      "loss": 0.1665,
+      "step": 629
+    },
+    {
+      "epoch": 5.833333333333333,
+      "grad_norm": 4.494096279144287,
+      "learning_rate": 0.00012372670807453417,
+      "loss": 0.1618,
+      "step": 630
+    },
+    {
+      "epoch": 5.842592592592593,
+      "grad_norm": 3.1159920692443848,
+      "learning_rate": 0.0001236024844720497,
+      "loss": 0.0993,
+      "step": 631
+    },
+    {
+      "epoch": 5.851851851851852,
+      "grad_norm": 7.402317523956299,
+      "learning_rate": 0.00012347826086956523,
+      "loss": 0.2546,
+      "step": 632
+    },
+    {
+      "epoch": 5.861111111111111,
+      "grad_norm": 5.94581937789917,
+      "learning_rate": 0.00012335403726708075,
+      "loss": 0.2239,
+      "step": 633
+    },
+    {
+      "epoch": 5.87037037037037,
+      "grad_norm": 1.6560418605804443,
+      "learning_rate": 0.00012322981366459627,
+      "loss": 0.0531,
+      "step": 634
+    },
+    {
+      "epoch": 5.87962962962963,
+      "grad_norm": 2.7259700298309326,
+      "learning_rate": 0.0001231055900621118,
+      "loss": 0.1182,
+      "step": 635
+    },
+    {
+      "epoch": 5.888888888888889,
+      "grad_norm": 6.228329181671143,
+      "learning_rate": 0.00012298136645962735,
+      "loss": 0.3486,
+      "step": 636
+    },
+    {
+      "epoch": 5.898148148148148,
+      "grad_norm": 1.6788215637207031,
+      "learning_rate": 0.00012285714285714287,
+      "loss": 0.0312,
+      "step": 637
+    },
+    {
+      "epoch": 5.907407407407407,
+      "grad_norm": 3.3095898628234863,
+      "learning_rate": 0.00012273291925465839,
+      "loss": 0.148,
+      "step": 638
+    },
+    {
+      "epoch": 5.916666666666667,
+      "grad_norm": 2.946584701538086,
+      "learning_rate": 0.00012260869565217393,
+      "loss": 0.0905,
+      "step": 639
+    },
+    {
+      "epoch": 5.925925925925926,
+      "grad_norm": 7.601954936981201,
+      "learning_rate": 0.00012248447204968945,
+      "loss": 0.3361,
+      "step": 640
+    },
+    {
+      "epoch": 5.935185185185185,
+      "grad_norm": 3.047450542449951,
+      "learning_rate": 0.00012236024844720496,
+      "loss": 0.08,
+      "step": 641
+    },
+    {
+      "epoch": 5.944444444444445,
+      "grad_norm": 2.289433717727661,
+      "learning_rate": 0.0001222360248447205,
+      "loss": 0.1263,
+      "step": 642
+    },
+    {
+      "epoch": 5.953703703703704,
+      "grad_norm": 3.0301482677459717,
+      "learning_rate": 0.00012211180124223602,
+      "loss": 0.1434,
+      "step": 643
+    },
+    {
+      "epoch": 5.962962962962963,
+      "grad_norm": 1.7519960403442383,
+      "learning_rate": 0.00012198757763975155,
+      "loss": 0.0757,
+      "step": 644
+    },
+    {
+      "epoch": 5.972222222222222,
+      "grad_norm": 1.5886623859405518,
+      "learning_rate": 0.0001218633540372671,
+      "loss": 0.0571,
+      "step": 645
+    },
+    {
+      "epoch": 5.981481481481482,
+      "grad_norm": 5.018250465393066,
+      "learning_rate": 0.00012173913043478263,
+      "loss": 0.3047,
+      "step": 646
+    },
+    {
+      "epoch": 5.9907407407407405,
+      "grad_norm": 2.7874503135681152,
+      "learning_rate": 0.00012161490683229814,
+      "loss": 0.1017,
+      "step": 647
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.445015907287598,
+      "learning_rate": 0.00012149068322981367,
+      "loss": 0.1223,
+      "step": 648
+    },
+    {
+      "epoch": 6.0092592592592595,
+      "grad_norm": 5.86106538772583,
+      "learning_rate": 0.0001213664596273292,
+      "loss": 0.1565,
+      "step": 649
+    },
+    {
+      "epoch": 6.018518518518518,
+      "grad_norm": 3.1399528980255127,
+      "learning_rate": 0.00012124223602484472,
+      "loss": 0.1554,
+      "step": 650
+    },
+    {
+      "epoch": 6.027777777777778,
+      "grad_norm": 3.606825828552246,
+      "learning_rate": 0.00012111801242236025,
+      "loss": 0.2018,
+      "step": 651
+    },
+    {
+      "epoch": 6.037037037037037,
+      "grad_norm": 2.6407530307769775,
+      "learning_rate": 0.00012099378881987578,
+      "loss": 0.125,
+      "step": 652
+    },
+    {
+      "epoch": 6.046296296296297,
+      "grad_norm": 2.7450759410858154,
+      "learning_rate": 0.00012086956521739131,
+      "loss": 0.0707,
+      "step": 653
+    },
+    {
+      "epoch": 6.055555555555555,
+      "grad_norm": 4.761390209197998,
+      "learning_rate": 0.00012074534161490685,
+      "loss": 0.1126,
+      "step": 654
+    },
+    {
+      "epoch": 6.064814814814815,
+      "grad_norm": 2.6973469257354736,
+      "learning_rate": 0.00012062111801242237,
+      "loss": 0.0593,
+      "step": 655
+    },
+    {
+      "epoch": 6.074074074074074,
+      "grad_norm": 3.7341206073760986,
+      "learning_rate": 0.0001204968944099379,
+      "loss": 0.132,
+      "step": 656
+    },
+    {
+      "epoch": 6.083333333333333,
+      "grad_norm": 8.52206039428711,
+      "learning_rate": 0.00012037267080745343,
+      "loss": 0.2624,
+      "step": 657
+    },
+    {
+      "epoch": 6.092592592592593,
+      "grad_norm": 2.5560076236724854,
+      "learning_rate": 0.00012024844720496896,
+      "loss": 0.0968,
+      "step": 658
+    },
+    {
+      "epoch": 6.101851851851852,
+      "grad_norm": 1.632588505744934,
+      "learning_rate": 0.00012012422360248447,
+      "loss": 0.0623,
+      "step": 659
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 3.6222176551818848,
+      "learning_rate": 0.00012,
+      "loss": 0.123,
+      "step": 660
+    },
+    {
+      "epoch": 6.12037037037037,
+      "grad_norm": 3.070725202560425,
+      "learning_rate": 0.00011987577639751553,
+      "loss": 0.1106,
+      "step": 661
+    },
+    {
+      "epoch": 6.12962962962963,
+      "grad_norm": 3.651214599609375,
+      "learning_rate": 0.00011975155279503105,
+      "loss": 0.1414,
+      "step": 662
+    },
+    {
+      "epoch": 6.138888888888889,
+      "grad_norm": 3.094923734664917,
+      "learning_rate": 0.0001196273291925466,
+      "loss": 0.1233,
+      "step": 663
+    },
+    {
+      "epoch": 6.148148148148148,
+      "grad_norm": 6.006683349609375,
+      "learning_rate": 0.00011950310559006212,
+      "loss": 0.1618,
+      "step": 664
+    },
+    {
+      "epoch": 6.157407407407407,
+      "grad_norm": 1.4269198179244995,
+      "learning_rate": 0.00011937888198757765,
+      "loss": 0.0483,
+      "step": 665
+    },
+    {
+      "epoch": 6.166666666666667,
+      "grad_norm": 3.9230058193206787,
+      "learning_rate": 0.00011925465838509318,
+      "loss": 0.0853,
+      "step": 666
+    },
+    {
+      "epoch": 6.175925925925926,
+      "grad_norm": 5.348933696746826,
+      "learning_rate": 0.0001191304347826087,
+      "loss": 0.1022,
+      "step": 667
+    },
+    {
+      "epoch": 6.185185185185185,
+      "grad_norm": 5.111001014709473,
+      "learning_rate": 0.00011900621118012423,
+      "loss": 0.1807,
+      "step": 668
+    },
+    {
+      "epoch": 6.194444444444445,
+      "grad_norm": 8.565694808959961,
+      "learning_rate": 0.00011888198757763976,
+      "loss": 0.0918,
+      "step": 669
+    },
+    {
+      "epoch": 6.203703703703703,
+      "grad_norm": 1.8871145248413086,
+      "learning_rate": 0.00011875776397515527,
+      "loss": 0.0277,
+      "step": 670
+    },
+    {
+      "epoch": 6.212962962962963,
+      "grad_norm": 2.728764772415161,
+      "learning_rate": 0.0001186335403726708,
+      "loss": 0.0934,
+      "step": 671
+    },
+    {
+      "epoch": 6.222222222222222,
+      "grad_norm": 4.937579154968262,
+      "learning_rate": 0.00011850931677018635,
+      "loss": 0.1346,
+      "step": 672
+    },
+    {
+      "epoch": 6.231481481481482,
+      "grad_norm": 6.841187953948975,
+      "learning_rate": 0.00011838509316770188,
+      "loss": 0.1542,
+      "step": 673
+    },
+    {
+      "epoch": 6.2407407407407405,
+      "grad_norm": 5.780396461486816,
+      "learning_rate": 0.00011826086956521741,
+      "loss": 0.2692,
+      "step": 674
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 4.7826080322265625,
+      "learning_rate": 0.00011813664596273292,
+      "loss": 0.1671,
+      "step": 675
+    },
+    {
+      "epoch": 6.2592592592592595,
+      "grad_norm": 4.660366535186768,
+      "learning_rate": 0.00011801242236024845,
+      "loss": 0.1156,
+      "step": 676
+    },
+    {
+      "epoch": 6.268518518518518,
+      "grad_norm": 2.414586067199707,
+      "learning_rate": 0.00011788819875776398,
+      "loss": 0.0844,
+      "step": 677
+    },
+    {
+      "epoch": 6.277777777777778,
+      "grad_norm": 2.0737192630767822,
+      "learning_rate": 0.0001177639751552795,
+      "loss": 0.0828,
+      "step": 678
+    },
+    {
+      "epoch": 6.287037037037037,
+      "grad_norm": 3.5586555004119873,
+      "learning_rate": 0.00011763975155279503,
+      "loss": 0.1531,
+      "step": 679
+    },
+    {
+      "epoch": 6.296296296296296,
+      "grad_norm": 5.451366901397705,
+      "learning_rate": 0.00011751552795031056,
+      "loss": 0.1401,
+      "step": 680
+    },
+    {
+      "epoch": 6.305555555555555,
+      "grad_norm": 2.353264570236206,
+      "learning_rate": 0.0001173913043478261,
+      "loss": 0.0969,
+      "step": 681
+    },
+    {
+      "epoch": 6.314814814814815,
+      "grad_norm": 3.3806886672973633,
+      "learning_rate": 0.00011726708074534163,
+      "loss": 0.0933,
+      "step": 682
+    },
+    {
+      "epoch": 6.324074074074074,
+      "grad_norm": 3.2616827487945557,
+      "learning_rate": 0.00011714285714285715,
+      "loss": 0.1034,
+      "step": 683
+    },
+    {
+      "epoch": 6.333333333333333,
+      "grad_norm": 2.7368857860565186,
+      "learning_rate": 0.00011701863354037268,
+      "loss": 0.0752,
+      "step": 684
+    },
+    {
+      "epoch": 6.342592592592593,
+      "grad_norm": 6.251594543457031,
+      "learning_rate": 0.00011689440993788821,
+      "loss": 0.1462,
+      "step": 685
+    },
+    {
+      "epoch": 6.351851851851852,
+      "grad_norm": 4.142698764801025,
+      "learning_rate": 0.00011677018633540373,
+      "loss": 0.1326,
+      "step": 686
+    },
+    {
+      "epoch": 6.361111111111111,
+      "grad_norm": 2.420140504837036,
+      "learning_rate": 0.00011664596273291926,
+      "loss": 0.0767,
+      "step": 687
+    },
+    {
+      "epoch": 6.37037037037037,
+      "grad_norm": 3.2684707641601562,
+      "learning_rate": 0.00011652173913043479,
+      "loss": 0.0698,
+      "step": 688
+    },
+    {
+      "epoch": 6.37962962962963,
+      "grad_norm": 2.8635385036468506,
+      "learning_rate": 0.0001163975155279503,
+      "loss": 0.1454,
+      "step": 689
+    },
+    {
+      "epoch": 6.388888888888889,
+      "grad_norm": 1.000563144683838,
+      "learning_rate": 0.00011627329192546586,
+      "loss": 0.037,
+      "step": 690
+    },
+    {
+      "epoch": 6.398148148148148,
+      "grad_norm": 5.726893424987793,
+      "learning_rate": 0.00011614906832298138,
+      "loss": 0.2738,
+      "step": 691
+    },
+    {
+      "epoch": 6.407407407407407,
+      "grad_norm": 1.0472891330718994,
+      "learning_rate": 0.0001160248447204969,
+      "loss": 0.0284,
+      "step": 692
+    },
+    {
+      "epoch": 6.416666666666667,
+      "grad_norm": 2.0266168117523193,
+      "learning_rate": 0.00011590062111801244,
+      "loss": 0.0703,
+      "step": 693
+    },
+    {
+      "epoch": 6.425925925925926,
+      "grad_norm": 4.494514465332031,
+      "learning_rate": 0.00011577639751552795,
+      "loss": 0.1956,
+      "step": 694
+    },
+    {
+      "epoch": 6.435185185185185,
+      "grad_norm": 5.9851837158203125,
+      "learning_rate": 0.00011565217391304348,
+      "loss": 0.1779,
+      "step": 695
+    },
+    {
+      "epoch": 6.444444444444445,
+      "grad_norm": 3.127121686935425,
+      "learning_rate": 0.00011552795031055901,
+      "loss": 0.0926,
+      "step": 696
+    },
+    {
+      "epoch": 6.453703703703704,
+      "grad_norm": 9.09554672241211,
+      "learning_rate": 0.00011540372670807453,
+      "loss": 0.236,
+      "step": 697
+    },
+    {
+      "epoch": 6.462962962962963,
+      "grad_norm": 6.604413032531738,
+      "learning_rate": 0.00011527950310559006,
+      "loss": 0.2466,
+      "step": 698
+    },
+    {
+      "epoch": 6.472222222222222,
+      "grad_norm": 1.7158292531967163,
+      "learning_rate": 0.0001151552795031056,
+      "loss": 0.0599,
+      "step": 699
+    },
+    {
+      "epoch": 6.481481481481482,
+      "grad_norm": 2.3201727867126465,
+      "learning_rate": 0.00011503105590062113,
+      "loss": 0.1221,
+      "step": 700
+    },
+    {
+      "epoch": 6.4907407407407405,
+      "grad_norm": 3.0037081241607666,
+      "learning_rate": 0.00011490683229813666,
+      "loss": 0.0752,
+      "step": 701
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 2.66445255279541,
+      "learning_rate": 0.00011478260869565218,
+      "loss": 0.0936,
+      "step": 702
+    },
+    {
+      "epoch": 6.5092592592592595,
+      "grad_norm": 2.3784914016723633,
+      "learning_rate": 0.00011465838509316771,
+      "loss": 0.1362,
+      "step": 703
+    },
+    {
+      "epoch": 6.518518518518518,
+      "grad_norm": 1.467207431793213,
+      "learning_rate": 0.00011453416149068324,
+      "loss": 0.033,
+      "step": 704
+    },
+    {
+      "epoch": 6.527777777777778,
+      "grad_norm": 4.558538436889648,
+      "learning_rate": 0.00011440993788819875,
+      "loss": 0.2072,
+      "step": 705
+    },
+    {
+      "epoch": 6.537037037037037,
+      "grad_norm": 2.80195951461792,
+      "learning_rate": 0.00011428571428571428,
+      "loss": 0.0687,
+      "step": 706
+    },
+    {
+      "epoch": 6.546296296296296,
+      "grad_norm": 8.565617561340332,
+      "learning_rate": 0.00011416149068322981,
+      "loss": 0.2849,
+      "step": 707
+    },
+    {
+      "epoch": 6.555555555555555,
+      "grad_norm": 1.682445764541626,
+      "learning_rate": 0.00011403726708074536,
+      "loss": 0.0688,
+      "step": 708
+    },
+    {
+      "epoch": 6.564814814814815,
+      "grad_norm": 2.2818193435668945,
+      "learning_rate": 0.00011391304347826089,
+      "loss": 0.0762,
+      "step": 709
+    },
+    {
+      "epoch": 6.574074074074074,
+      "grad_norm": 1.554701328277588,
+      "learning_rate": 0.0001137888198757764,
+      "loss": 0.0588,
+      "step": 710
+    },
+    {
+      "epoch": 6.583333333333333,
+      "grad_norm": 0.8565337061882019,
+      "learning_rate": 0.00011366459627329193,
+      "loss": 0.0296,
+      "step": 711
+    },
+    {
+      "epoch": 6.592592592592593,
+      "grad_norm": 3.192460060119629,
+      "learning_rate": 0.00011354037267080746,
+      "loss": 0.0978,
+      "step": 712
+    },
+    {
+      "epoch": 6.601851851851852,
+      "grad_norm": 1.5544946193695068,
+      "learning_rate": 0.00011341614906832298,
+      "loss": 0.0615,
+      "step": 713
+    },
+    {
+      "epoch": 6.611111111111111,
+      "grad_norm": 5.477461814880371,
+      "learning_rate": 0.00011329192546583851,
+      "loss": 0.2045,
+      "step": 714
+    },
+    {
+      "epoch": 6.62037037037037,
+      "grad_norm": 4.519773960113525,
+      "learning_rate": 0.00011316770186335404,
+      "loss": 0.2205,
+      "step": 715
+    },
+    {
+      "epoch": 6.62962962962963,
+      "grad_norm": 5.499269962310791,
+      "learning_rate": 0.00011304347826086956,
+      "loss": 0.1925,
+      "step": 716
+    },
+    {
+      "epoch": 6.638888888888889,
+      "grad_norm": 3.747269868850708,
+      "learning_rate": 0.00011291925465838511,
+      "loss": 0.1676,
+      "step": 717
+    },
+    {
+      "epoch": 6.648148148148148,
+      "grad_norm": 2.2231295108795166,
+      "learning_rate": 0.00011279503105590063,
+      "loss": 0.1009,
+      "step": 718
+    },
+    {
+      "epoch": 6.657407407407407,
+      "grad_norm": 2.7668187618255615,
+      "learning_rate": 0.00011267080745341616,
+      "loss": 0.075,
+      "step": 719
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 3.6396594047546387,
+      "learning_rate": 0.00011254658385093169,
+      "loss": 0.1634,
+      "step": 720
+    },
+    {
+      "epoch": 6.675925925925926,
+      "grad_norm": 2.8981094360351562,
+      "learning_rate": 0.00011242236024844722,
+      "loss": 0.167,
+      "step": 721
+    },
+    {
+      "epoch": 6.685185185185185,
+      "grad_norm": 1.7787268161773682,
+      "learning_rate": 0.00011229813664596273,
+      "loss": 0.0541,
+      "step": 722
+    },
+    {
+      "epoch": 6.694444444444445,
+      "grad_norm": 5.248884677886963,
+      "learning_rate": 0.00011217391304347826,
+      "loss": 0.1612,
+      "step": 723
+    },
+    {
+      "epoch": 6.703703703703704,
+      "grad_norm": 6.716445446014404,
+      "learning_rate": 0.0001120496894409938,
+      "loss": 0.232,
+      "step": 724
+    },
+    {
+      "epoch": 6.712962962962963,
+      "grad_norm": 4.387823581695557,
+      "learning_rate": 0.00011192546583850931,
+      "loss": 0.1634,
+      "step": 725
+    },
+    {
+      "epoch": 6.722222222222222,
+      "grad_norm": 2.278259515762329,
+      "learning_rate": 0.00011180124223602484,
+      "loss": 0.0771,
+      "step": 726
+    },
+    {
+      "epoch": 6.731481481481482,
+      "grad_norm": 3.915947198867798,
+      "learning_rate": 0.00011167701863354038,
+      "loss": 0.1595,
+      "step": 727
+    },
+    {
+      "epoch": 6.7407407407407405,
+      "grad_norm": 3.473524332046509,
+      "learning_rate": 0.00011155279503105591,
+      "loss": 0.1075,
+      "step": 728
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.3765463829040527,
+      "learning_rate": 0.00011142857142857144,
+      "loss": 0.0532,
+      "step": 729
+    },
+    {
+      "epoch": 6.7592592592592595,
+      "grad_norm": 3.4707448482513428,
+      "learning_rate": 0.00011130434782608696,
+      "loss": 0.1275,
+      "step": 730
+    },
+    {
+      "epoch": 6.768518518518518,
+      "grad_norm": 4.364494800567627,
+      "learning_rate": 0.00011118012422360249,
+      "loss": 0.1482,
+      "step": 731
+    },
+    {
+      "epoch": 6.777777777777778,
+      "grad_norm": 3.434342384338379,
+      "learning_rate": 0.00011105590062111802,
+      "loss": 0.091,
+      "step": 732
+    },
+    {
+      "epoch": 6.787037037037037,
+      "grad_norm": 2.428468942642212,
+      "learning_rate": 0.00011093167701863354,
+      "loss": 0.1011,
+      "step": 733
+    },
+    {
+      "epoch": 6.796296296296296,
+      "grad_norm": 4.1566057205200195,
+      "learning_rate": 0.00011080745341614907,
+      "loss": 0.1952,
+      "step": 734
+    },
+    {
+      "epoch": 6.805555555555555,
+      "grad_norm": 3.701270818710327,
+      "learning_rate": 0.0001106832298136646,
+      "loss": 0.1344,
+      "step": 735
+    },
+    {
+      "epoch": 6.814814814814815,
+      "grad_norm": 3.952190399169922,
+      "learning_rate": 0.00011055900621118014,
+      "loss": 0.1137,
+      "step": 736
+    },
+    {
+      "epoch": 6.824074074074074,
+      "grad_norm": 1.9077421426773071,
+      "learning_rate": 0.00011043478260869567,
+      "loss": 0.0495,
+      "step": 737
+    },
+    {
+      "epoch": 6.833333333333333,
+      "grad_norm": 7.015697956085205,
+      "learning_rate": 0.00011031055900621119,
+      "loss": 0.3821,
+      "step": 738
+    },
+    {
+      "epoch": 6.842592592592593,
+      "grad_norm": 2.9424688816070557,
+      "learning_rate": 0.00011018633540372672,
+      "loss": 0.0788,
+      "step": 739
+    },
+    {
+      "epoch": 6.851851851851852,
+      "grad_norm": 1.9170938730239868,
+      "learning_rate": 0.00011006211180124225,
+      "loss": 0.0735,
+      "step": 740
+    },
+    {
+      "epoch": 6.861111111111111,
+      "grad_norm": 3.8930490016937256,
+      "learning_rate": 0.00010993788819875776,
+      "loss": 0.0723,
+      "step": 741
+    },
+    {
+      "epoch": 6.87037037037037,
+      "grad_norm": 5.078344345092773,
+      "learning_rate": 0.00010981366459627329,
+      "loss": 0.1568,
+      "step": 742
+    },
+    {
+      "epoch": 6.87962962962963,
+      "grad_norm": 1.1494572162628174,
+      "learning_rate": 0.00010968944099378882,
+      "loss": 0.0333,
+      "step": 743
+    },
+    {
+      "epoch": 6.888888888888889,
+      "grad_norm": 4.472750663757324,
+      "learning_rate": 0.00010956521739130434,
+      "loss": 0.1611,
+      "step": 744
+    },
+    {
+      "epoch": 6.898148148148148,
+      "grad_norm": 2.977266311645508,
+      "learning_rate": 0.0001094409937888199,
+      "loss": 0.1003,
+      "step": 745
+    },
+    {
+      "epoch": 6.907407407407407,
+      "grad_norm": 2.7986366748809814,
+      "learning_rate": 0.00010931677018633541,
+      "loss": 0.104,
+      "step": 746
+    },
+    {
+      "epoch": 6.916666666666667,
+      "grad_norm": 3.25215482711792,
+      "learning_rate": 0.00010919254658385094,
+      "loss": 0.1337,
+      "step": 747
+    },
+    {
+      "epoch": 6.925925925925926,
+      "grad_norm": 4.609085559844971,
+      "learning_rate": 0.00010906832298136647,
+      "loss": 0.1346,
+      "step": 748
+    },
+    {
+      "epoch": 6.935185185185185,
+      "grad_norm": 2.8187437057495117,
+      "learning_rate": 0.00010894409937888199,
+      "loss": 0.0817,
+      "step": 749
+    },
+    {
+      "epoch": 6.944444444444445,
+      "grad_norm": 3.499835252761841,
+      "learning_rate": 0.00010881987577639752,
+      "loss": 0.1592,
+      "step": 750
+    },
+    {
+      "epoch": 6.953703703703704,
+      "grad_norm": 5.405245304107666,
+      "learning_rate": 0.00010869565217391305,
+      "loss": 0.1918,
+      "step": 751
+    },
+    {
+      "epoch": 6.962962962962963,
+      "grad_norm": 3.507462978363037,
+      "learning_rate": 0.00010857142857142856,
+      "loss": 0.1177,
+      "step": 752
+    },
+    {
+      "epoch": 6.972222222222222,
+      "grad_norm": 4.5707316398620605,
+      "learning_rate": 0.0001084472049689441,
+      "loss": 0.2059,
+      "step": 753
+    },
+    {
+      "epoch": 6.981481481481482,
+      "grad_norm": 4.220851898193359,
+      "learning_rate": 0.00010832298136645964,
+      "loss": 0.1032,
+      "step": 754
+    },
+    {
+      "epoch": 6.9907407407407405,
+      "grad_norm": 1.384399652481079,
+      "learning_rate": 0.00010819875776397517,
+      "loss": 0.0396,
+      "step": 755
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 4.726045608520508,
+      "learning_rate": 0.0001080745341614907,
+      "loss": 0.1665,
+      "step": 756
+    },
+    {
+      "epoch": 7.0092592592592595,
+      "grad_norm": 1.7653844356536865,
+      "learning_rate": 0.00010795031055900621,
+      "loss": 0.0523,
+      "step": 757
+    },
+    {
+      "epoch": 7.018518518518518,
+      "grad_norm": 3.113438606262207,
+      "learning_rate": 0.00010782608695652174,
+      "loss": 0.1037,
+      "step": 758
+    },
+    {
+      "epoch": 7.027777777777778,
+      "grad_norm": 0.7334030866622925,
+      "learning_rate": 0.00010770186335403727,
+      "loss": 0.018,
+      "step": 759
+    },
+    {
+      "epoch": 7.037037037037037,
+      "grad_norm": 4.1807403564453125,
+      "learning_rate": 0.00010757763975155279,
+      "loss": 0.1268,
+      "step": 760
+    },
+    {
+      "epoch": 7.046296296296297,
+      "grad_norm": 0.9362726807594299,
+      "learning_rate": 0.00010745341614906832,
+      "loss": 0.0297,
+      "step": 761
+    },
+    {
+      "epoch": 7.055555555555555,
+      "grad_norm": 4.32288932800293,
+      "learning_rate": 0.00010732919254658385,
+      "loss": 0.1442,
+      "step": 762
+    },
+    {
+      "epoch": 7.064814814814815,
+      "grad_norm": 5.067570209503174,
+      "learning_rate": 0.00010720496894409939,
+      "loss": 0.1025,
+      "step": 763
+    },
+    {
+      "epoch": 7.074074074074074,
+      "grad_norm": 4.261683464050293,
+      "learning_rate": 0.00010708074534161492,
+      "loss": 0.0737,
+      "step": 764
+    },
+    {
+      "epoch": 7.083333333333333,
+      "grad_norm": 1.640762448310852,
+      "learning_rate": 0.00010695652173913044,
+      "loss": 0.0539,
+      "step": 765
+    },
+    {
+      "epoch": 7.092592592592593,
+      "grad_norm": 4.321948051452637,
+      "learning_rate": 0.00010683229813664597,
+      "loss": 0.1349,
+      "step": 766
+    },
+    {
+      "epoch": 7.101851851851852,
+      "grad_norm": 3.3006460666656494,
+      "learning_rate": 0.0001067080745341615,
+      "loss": 0.1521,
+      "step": 767
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 4.936793804168701,
+      "learning_rate": 0.00010658385093167702,
+      "loss": 0.2134,
+      "step": 768
+    },
+    {
+      "epoch": 7.12037037037037,
+      "grad_norm": 4.625700950622559,
+      "learning_rate": 0.00010645962732919255,
+      "loss": 0.1357,
+      "step": 769
+    },
+    {
+      "epoch": 7.12962962962963,
+      "grad_norm": 3.6059927940368652,
+      "learning_rate": 0.00010633540372670807,
+      "loss": 0.1155,
+      "step": 770
+    },
+    {
+      "epoch": 7.138888888888889,
+      "grad_norm": 7.513297080993652,
+      "learning_rate": 0.00010621118012422359,
+      "loss": 0.2876,
+      "step": 771
+    },
+    {
+      "epoch": 7.148148148148148,
+      "grad_norm": 1.9764149188995361,
+      "learning_rate": 0.00010608695652173915,
+      "loss": 0.0494,
+      "step": 772
+    },
+    {
+      "epoch": 7.157407407407407,
+      "grad_norm": 6.305065631866455,
+      "learning_rate": 0.00010596273291925466,
+      "loss": 0.1551,
+      "step": 773
+    },
+    {
+      "epoch": 7.166666666666667,
+      "grad_norm": 7.061641216278076,
+      "learning_rate": 0.0001058385093167702,
+      "loss": 0.4477,
+      "step": 774
+    },
+    {
+      "epoch": 7.175925925925926,
+      "grad_norm": 2.5913658142089844,
+      "learning_rate": 0.00010571428571428572,
+      "loss": 0.0891,
+      "step": 775
+    },
+    {
+      "epoch": 7.185185185185185,
+      "grad_norm": 4.326108932495117,
+      "learning_rate": 0.00010559006211180124,
+      "loss": 0.1668,
+      "step": 776
+    },
+    {
+      "epoch": 7.194444444444445,
+      "grad_norm": 1.2704784870147705,
+      "learning_rate": 0.00010546583850931677,
+      "loss": 0.0479,
+      "step": 777
+    },
+    {
+      "epoch": 7.203703703703703,
+      "grad_norm": 4.410538673400879,
+      "learning_rate": 0.0001053416149068323,
+      "loss": 0.091,
+      "step": 778
+    },
+    {
+      "epoch": 7.212962962962963,
+      "grad_norm": 4.080562114715576,
+      "learning_rate": 0.00010521739130434782,
+      "loss": 0.0864,
+      "step": 779
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 2.016847848892212,
+      "learning_rate": 0.00010509316770186335,
+      "loss": 0.0642,
+      "step": 780
+    },
+    {
+      "epoch": 7.231481481481482,
+      "grad_norm": 1.3171155452728271,
+      "learning_rate": 0.00010496894409937889,
+      "loss": 0.0341,
+      "step": 781
+    },
+    {
+      "epoch": 7.2407407407407405,
+      "grad_norm": 3.395470380783081,
+      "learning_rate": 0.00010484472049689442,
+      "loss": 0.0992,
+      "step": 782
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 3.640408515930176,
+      "learning_rate": 0.00010472049689440995,
+      "loss": 0.1221,
+      "step": 783
+    },
+    {
+      "epoch": 7.2592592592592595,
+      "grad_norm": 2.7023701667785645,
+      "learning_rate": 0.00010459627329192547,
+      "loss": 0.0616,
+      "step": 784
+    },
+    {
+      "epoch": 7.268518518518518,
+      "grad_norm": 1.2496731281280518,
+      "learning_rate": 0.000104472049689441,
+      "loss": 0.0395,
+      "step": 785
+    },
+    {
+      "epoch": 7.277777777777778,
+      "grad_norm": 4.615080833435059,
+      "learning_rate": 0.00010434782608695653,
+      "loss": 0.1183,
+      "step": 786
+    },
+    {
+      "epoch": 7.287037037037037,
+      "grad_norm": 2.8553638458251953,
+      "learning_rate": 0.00010422360248447206,
+      "loss": 0.065,
+      "step": 787
+    },
+    {
+      "epoch": 7.296296296296296,
+      "grad_norm": 5.177341461181641,
+      "learning_rate": 0.00010409937888198757,
+      "loss": 0.1311,
+      "step": 788
+    },
+    {
+      "epoch": 7.305555555555555,
+      "grad_norm": 4.220547199249268,
+      "learning_rate": 0.0001039751552795031,
+      "loss": 0.0963,
+      "step": 789
+    },
+    {
+      "epoch": 7.314814814814815,
+      "grad_norm": 4.426061630249023,
+      "learning_rate": 0.00010385093167701865,
+      "loss": 0.0784,
+      "step": 790
+    },
+    {
+      "epoch": 7.324074074074074,
+      "grad_norm": 3.388397693634033,
+      "learning_rate": 0.00010372670807453418,
+      "loss": 0.0874,
+      "step": 791
+    },
+    {
+      "epoch": 7.333333333333333,
+      "grad_norm": 1.8489770889282227,
+      "learning_rate": 0.0001036024844720497,
+      "loss": 0.0512,
+      "step": 792
+    },
+    {
+      "epoch": 7.342592592592593,
+      "grad_norm": 3.316032886505127,
+      "learning_rate": 0.00010347826086956522,
+      "loss": 0.0538,
+      "step": 793
+    },
+    {
+      "epoch": 7.351851851851852,
+      "grad_norm": 2.2799925804138184,
+      "learning_rate": 0.00010335403726708075,
+      "loss": 0.0607,
+      "step": 794
+    },
+    {
+      "epoch": 7.361111111111111,
+      "grad_norm": 6.567706108093262,
+      "learning_rate": 0.00010322981366459628,
+      "loss": 0.2587,
+      "step": 795
+    },
+    {
+      "epoch": 7.37037037037037,
+      "grad_norm": 1.3475052118301392,
+      "learning_rate": 0.0001031055900621118,
+      "loss": 0.036,
+      "step": 796
+    },
+    {
+      "epoch": 7.37962962962963,
+      "grad_norm": 4.109946250915527,
+      "learning_rate": 0.00010298136645962733,
+      "loss": 0.1113,
+      "step": 797
+    },
+    {
+      "epoch": 7.388888888888889,
+      "grad_norm": 2.5367050170898438,
+      "learning_rate": 0.00010285714285714286,
+      "loss": 0.0818,
+      "step": 798
+    },
+    {
+      "epoch": 7.398148148148148,
+      "grad_norm": 3.9039812088012695,
+      "learning_rate": 0.0001027329192546584,
+      "loss": 0.1218,
+      "step": 799
+    },
+    {
+      "epoch": 7.407407407407407,
+      "grad_norm": 4.130105972290039,
+      "learning_rate": 0.00010260869565217393,
+      "loss": 0.111,
+      "step": 800
+    },
+    {
+      "epoch": 7.416666666666667,
+      "grad_norm": 3.941784381866455,
+      "learning_rate": 0.00010248447204968945,
+      "loss": 0.1325,
+      "step": 801
+    },
+    {
+      "epoch": 7.425925925925926,
+      "grad_norm": 1.690091848373413,
+      "learning_rate": 0.00010236024844720498,
+      "loss": 0.0302,
+      "step": 802
+    },
+    {
+      "epoch": 7.435185185185185,
+      "grad_norm": 3.6138830184936523,
+      "learning_rate": 0.00010223602484472051,
+      "loss": 0.0976,
+      "step": 803
+    },
+    {
+      "epoch": 7.444444444444445,
+      "grad_norm": 6.15052604675293,
+      "learning_rate": 0.00010211180124223602,
+      "loss": 0.2043,
+      "step": 804
+    },
+    {
+      "epoch": 7.453703703703704,
+      "grad_norm": 0.9923351407051086,
+      "learning_rate": 0.00010198757763975155,
+      "loss": 0.0339,
+      "step": 805
+    },
+    {
+      "epoch": 7.462962962962963,
+      "grad_norm": 1.4790726900100708,
+      "learning_rate": 0.00010186335403726708,
+      "loss": 0.0547,
+      "step": 806
+    },
+    {
+      "epoch": 7.472222222222222,
+      "grad_norm": 2.1284537315368652,
+      "learning_rate": 0.0001017391304347826,
+      "loss": 0.0825,
+      "step": 807
+    },
+    {
+      "epoch": 7.481481481481482,
+      "grad_norm": 6.377471923828125,
+      "learning_rate": 0.00010161490683229816,
+      "loss": 0.244,
+      "step": 808
+    },
+    {
+      "epoch": 7.4907407407407405,
+      "grad_norm": 1.6442444324493408,
+      "learning_rate": 0.00010149068322981367,
+      "loss": 0.0599,
+      "step": 809
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 10.18280029296875,
+      "learning_rate": 0.0001013664596273292,
+      "loss": 0.1209,
+      "step": 810
+    },
+    {
+      "epoch": 7.5092592592592595,
+      "grad_norm": 1.3371292352676392,
+      "learning_rate": 0.00010124223602484473,
+      "loss": 0.0328,
+      "step": 811
+    },
+    {
+      "epoch": 7.518518518518518,
+      "grad_norm": 1.0884917974472046,
+      "learning_rate": 0.00010111801242236025,
+      "loss": 0.0481,
+      "step": 812
+    },
+    {
+      "epoch": 7.527777777777778,
+      "grad_norm": 1.8597067594528198,
+      "learning_rate": 0.00010099378881987578,
+      "loss": 0.0612,
+      "step": 813
+    },
+    {
+      "epoch": 7.537037037037037,
+      "grad_norm": 1.331111192703247,
+      "learning_rate": 0.00010086956521739131,
+      "loss": 0.0415,
+      "step": 814
+    },
+    {
+      "epoch": 7.546296296296296,
+      "grad_norm": 5.1898322105407715,
+      "learning_rate": 0.00010074534161490683,
+      "loss": 0.2566,
+      "step": 815
+    },
+    {
+      "epoch": 7.555555555555555,
+      "grad_norm": 3.672396183013916,
+      "learning_rate": 0.00010062111801242236,
+      "loss": 0.0728,
+      "step": 816
+    },
+    {
+      "epoch": 7.564814814814815,
+      "grad_norm": 3.882545232772827,
+      "learning_rate": 0.0001004968944099379,
+      "loss": 0.1063,
+      "step": 817
+    },
+    {
+      "epoch": 7.574074074074074,
+      "grad_norm": 4.563115119934082,
+      "learning_rate": 0.00010037267080745343,
+      "loss": 0.1335,
+      "step": 818
+    },
+    {
+      "epoch": 7.583333333333333,
+      "grad_norm": 2.913468599319458,
+      "learning_rate": 0.00010024844720496896,
+      "loss": 0.1307,
+      "step": 819
+    },
+    {
+      "epoch": 7.592592592592593,
+      "grad_norm": 5.023247241973877,
+      "learning_rate": 0.00010012422360248447,
+      "loss": 0.1065,
+      "step": 820
+    },
+    {
+      "epoch": 7.601851851851852,
+      "grad_norm": 3.0828170776367188,
+      "learning_rate": 0.0001,
+      "loss": 0.0745,
+      "step": 821
+    },
+    {
+      "epoch": 7.611111111111111,
+      "grad_norm": 1.2979748249053955,
+      "learning_rate": 9.987577639751553e-05,
+      "loss": 0.0521,
+      "step": 822
+    },
+    {
+      "epoch": 7.62037037037037,
+      "grad_norm": 3.1594204902648926,
+      "learning_rate": 9.975155279503105e-05,
+      "loss": 0.1559,
+      "step": 823
+    },
+    {
+      "epoch": 7.62962962962963,
+      "grad_norm": 5.367359638214111,
+      "learning_rate": 9.96273291925466e-05,
+      "loss": 0.2003,
+      "step": 824
+    },
+    {
+      "epoch": 7.638888888888889,
+      "grad_norm": 4.4314284324646,
+      "learning_rate": 9.950310559006212e-05,
+      "loss": 0.1119,
+      "step": 825
+    },
+    {
+      "epoch": 7.648148148148148,
+      "grad_norm": 2.753791570663452,
+      "learning_rate": 9.937888198757764e-05,
+      "loss": 0.05,
+      "step": 826
+    },
+    {
+      "epoch": 7.657407407407407,
+      "grad_norm": 2.6093661785125732,
+      "learning_rate": 9.925465838509317e-05,
+      "loss": 0.0753,
+      "step": 827
+    },
+    {
+      "epoch": 7.666666666666667,
+      "grad_norm": 3.5111002922058105,
+      "learning_rate": 9.91304347826087e-05,
+      "loss": 0.0774,
+      "step": 828
+    },
+    {
+      "epoch": 7.675925925925926,
+      "grad_norm": 4.576236724853516,
+      "learning_rate": 9.900621118012423e-05,
+      "loss": 0.1454,
+      "step": 829
+    },
+    {
+      "epoch": 7.685185185185185,
+      "grad_norm": 1.4670056104660034,
+      "learning_rate": 9.888198757763976e-05,
+      "loss": 0.059,
+      "step": 830
+    },
+    {
+      "epoch": 7.694444444444445,
+      "grad_norm": 2.6183865070343018,
+      "learning_rate": 9.875776397515528e-05,
+      "loss": 0.0755,
+      "step": 831
+    },
+    {
+      "epoch": 7.703703703703704,
+      "grad_norm": 2.1917359828948975,
+      "learning_rate": 9.86335403726708e-05,
+      "loss": 0.054,
+      "step": 832
+    },
+    {
+      "epoch": 7.712962962962963,
+      "grad_norm": 1.6838301420211792,
+      "learning_rate": 9.850931677018635e-05,
+      "loss": 0.0606,
+      "step": 833
+    },
+    {
+      "epoch": 7.722222222222222,
+      "grad_norm": 3.6826465129852295,
+      "learning_rate": 9.838509316770187e-05,
+      "loss": 0.1208,
+      "step": 834
+    },
+    {
+      "epoch": 7.731481481481482,
+      "grad_norm": 1.396178126335144,
+      "learning_rate": 9.82608695652174e-05,
+      "loss": 0.0475,
+      "step": 835
+    },
+    {
+      "epoch": 7.7407407407407405,
+      "grad_norm": 0.8179628849029541,
+      "learning_rate": 9.813664596273293e-05,
+      "loss": 0.0337,
+      "step": 836
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 1.4039196968078613,
+      "learning_rate": 9.801242236024846e-05,
+      "loss": 0.045,
+      "step": 837
+    },
+    {
+      "epoch": 7.7592592592592595,
+      "grad_norm": 1.8538316488265991,
+      "learning_rate": 9.788819875776399e-05,
+      "loss": 0.0504,
+      "step": 838
+    },
+    {
+      "epoch": 7.768518518518518,
+      "grad_norm": 3.978010654449463,
+      "learning_rate": 9.77639751552795e-05,
+      "loss": 0.1365,
+      "step": 839
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 6.751378059387207,
+      "learning_rate": 9.763975155279503e-05,
+      "loss": 0.2244,
+      "step": 840
+    },
+    {
+      "epoch": 7.787037037037037,
+      "grad_norm": 1.3808748722076416,
+      "learning_rate": 9.751552795031056e-05,
+      "loss": 0.0637,
+      "step": 841
+    },
+    {
+      "epoch": 7.796296296296296,
+      "grad_norm": 3.026418685913086,
+      "learning_rate": 9.739130434782609e-05,
+      "loss": 0.0923,
+      "step": 842
+    },
+    {
+      "epoch": 7.805555555555555,
+      "grad_norm": 3.5097484588623047,
+      "learning_rate": 9.726708074534162e-05,
+      "loss": 0.1476,
+      "step": 843
+    },
+    {
+      "epoch": 7.814814814814815,
+      "grad_norm": 1.7709219455718994,
+      "learning_rate": 9.714285714285715e-05,
+      "loss": 0.0732,
+      "step": 844
+    },
+    {
+      "epoch": 7.824074074074074,
+      "grad_norm": 1.2974692583084106,
+      "learning_rate": 9.701863354037267e-05,
+      "loss": 0.0752,
+      "step": 845
+    },
+    {
+      "epoch": 7.833333333333333,
+      "grad_norm": 0.6633884310722351,
+      "learning_rate": 9.689440993788821e-05,
+      "loss": 0.021,
+      "step": 846
+    },
+    {
+      "epoch": 7.842592592592593,
+      "grad_norm": 2.9434750080108643,
+      "learning_rate": 9.677018633540373e-05,
+      "loss": 0.0862,
+      "step": 847
+    },
+    {
+      "epoch": 7.851851851851852,
+      "grad_norm": 2.329890727996826,
+      "learning_rate": 9.664596273291926e-05,
+      "loss": 0.0689,
+      "step": 848
+    },
+    {
+      "epoch": 7.861111111111111,
+      "grad_norm": 4.351027965545654,
+      "learning_rate": 9.652173913043479e-05,
+      "loss": 0.0473,
+      "step": 849
+    },
+    {
+      "epoch": 7.87037037037037,
+      "grad_norm": 2.6241369247436523,
+      "learning_rate": 9.63975155279503e-05,
+      "loss": 0.0887,
+      "step": 850
+    },
+    {
+      "epoch": 7.87962962962963,
+      "grad_norm": 1.7621710300445557,
+      "learning_rate": 9.627329192546585e-05,
+      "loss": 0.0686,
+      "step": 851
+    },
+    {
+      "epoch": 7.888888888888889,
+      "grad_norm": 0.9145649075508118,
+      "learning_rate": 9.614906832298138e-05,
+      "loss": 0.0286,
+      "step": 852
+    },
+    {
+      "epoch": 7.898148148148148,
+      "grad_norm": 5.4611101150512695,
+      "learning_rate": 9.60248447204969e-05,
+      "loss": 0.1921,
+      "step": 853
+    },
+    {
+      "epoch": 7.907407407407407,
+      "grad_norm": 8.37093448638916,
+      "learning_rate": 9.590062111801242e-05,
+      "loss": 0.183,
+      "step": 854
+    },
+    {
+      "epoch": 7.916666666666667,
+      "grad_norm": 1.575817346572876,
+      "learning_rate": 9.577639751552795e-05,
+      "loss": 0.0464,
+      "step": 855
+    },
+    {
+      "epoch": 7.925925925925926,
+      "grad_norm": 5.839998722076416,
+      "learning_rate": 9.565217391304348e-05,
+      "loss": 0.1948,
+      "step": 856
+    },
+    {
+      "epoch": 7.935185185185185,
+      "grad_norm": 3.813732624053955,
+      "learning_rate": 9.552795031055901e-05,
+      "loss": 0.1637,
+      "step": 857
+    },
+    {
+      "epoch": 7.944444444444445,
+      "grad_norm": 6.688395023345947,
+      "learning_rate": 9.540372670807454e-05,
+      "loss": 0.1742,
+      "step": 858
+    },
+    {
+      "epoch": 7.953703703703704,
+      "grad_norm": 1.701634168624878,
+      "learning_rate": 9.527950310559006e-05,
+      "loss": 0.0734,
+      "step": 859
+    },
+    {
+      "epoch": 7.962962962962963,
+      "grad_norm": 6.329746246337891,
+      "learning_rate": 9.51552795031056e-05,
+      "loss": 0.4215,
+      "step": 860
+    },
+    {
+      "epoch": 7.972222222222222,
+      "grad_norm": 1.4201380014419556,
+      "learning_rate": 9.503105590062112e-05,
+      "loss": 0.0552,
+      "step": 861
+    },
+    {
+      "epoch": 7.981481481481482,
+      "grad_norm": 3.697819709777832,
+      "learning_rate": 9.490683229813665e-05,
+      "loss": 0.1865,
+      "step": 862
+    },
+    {
+      "epoch": 7.9907407407407405,
+      "grad_norm": 2.9667913913726807,
+      "learning_rate": 9.478260869565218e-05,
+      "loss": 0.1707,
+      "step": 863
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 4.044425010681152,
+      "learning_rate": 9.465838509316771e-05,
+      "loss": 0.147,
+      "step": 864
+    },
+    {
+      "epoch": 8.00925925925926,
+      "grad_norm": 3.552003860473633,
+      "learning_rate": 9.453416149068324e-05,
+      "loss": 0.2017,
+      "step": 865
+    },
+    {
+      "epoch": 8.018518518518519,
+      "grad_norm": 1.0738625526428223,
+      "learning_rate": 9.440993788819877e-05,
+      "loss": 0.0394,
+      "step": 866
+    },
+    {
+      "epoch": 8.027777777777779,
+      "grad_norm": 2.0488643646240234,
+      "learning_rate": 9.428571428571429e-05,
+      "loss": 0.0338,
+      "step": 867
+    },
+    {
+      "epoch": 8.037037037037036,
+      "grad_norm": 0.6434540152549744,
+      "learning_rate": 9.416149068322982e-05,
+      "loss": 0.0203,
+      "step": 868
+    },
+    {
+      "epoch": 8.046296296296296,
+      "grad_norm": 2.2168703079223633,
+      "learning_rate": 9.403726708074534e-05,
+      "loss": 0.0658,
+      "step": 869
+    },
+    {
+      "epoch": 8.055555555555555,
+      "grad_norm": 4.93209171295166,
+      "learning_rate": 9.391304347826087e-05,
+      "loss": 0.0787,
+      "step": 870
+    },
+    {
+      "epoch": 8.064814814814815,
+      "grad_norm": 1.0295863151550293,
+      "learning_rate": 9.37888198757764e-05,
+      "loss": 0.0422,
+      "step": 871
+    },
+    {
+      "epoch": 8.074074074074074,
+      "grad_norm": 4.721346855163574,
+      "learning_rate": 9.366459627329192e-05,
+      "loss": 0.1582,
+      "step": 872
+    },
+    {
+      "epoch": 8.083333333333334,
+      "grad_norm": 1.6571961641311646,
+      "learning_rate": 9.354037267080745e-05,
+      "loss": 0.0674,
+      "step": 873
+    },
+    {
+      "epoch": 8.092592592592593,
+      "grad_norm": 1.5981183052062988,
+      "learning_rate": 9.3416149068323e-05,
+      "loss": 0.0582,
+      "step": 874
+    },
+    {
+      "epoch": 8.101851851851851,
+      "grad_norm": 1.2692313194274902,
+      "learning_rate": 9.329192546583851e-05,
+      "loss": 0.0342,
+      "step": 875
+    },
+    {
+      "epoch": 8.11111111111111,
+      "grad_norm": 1.5510468482971191,
+      "learning_rate": 9.316770186335404e-05,
+      "loss": 0.0303,
+      "step": 876
+    },
+    {
+      "epoch": 8.12037037037037,
+      "grad_norm": 3.052006959915161,
+      "learning_rate": 9.304347826086957e-05,
+      "loss": 0.0613,
+      "step": 877
+    },
+    {
+      "epoch": 8.12962962962963,
+      "grad_norm": 0.9419259428977966,
+      "learning_rate": 9.29192546583851e-05,
+      "loss": 0.0336,
+      "step": 878
+    },
+    {
+      "epoch": 8.13888888888889,
+      "grad_norm": 1.4912816286087036,
+      "learning_rate": 9.279503105590063e-05,
+      "loss": 0.0579,
+      "step": 879
+    },
+    {
+      "epoch": 8.148148148148149,
+      "grad_norm": 1.7151224613189697,
+      "learning_rate": 9.267080745341615e-05,
+      "loss": 0.0427,
+      "step": 880
+    },
+    {
+      "epoch": 8.157407407407407,
+      "grad_norm": 2.1876168251037598,
+      "learning_rate": 9.254658385093168e-05,
+      "loss": 0.1089,
+      "step": 881
+    },
+    {
+      "epoch": 8.166666666666666,
+      "grad_norm": 2.9372048377990723,
+      "learning_rate": 9.24223602484472e-05,
+      "loss": 0.0714,
+      "step": 882
+    },
+    {
+      "epoch": 8.175925925925926,
+      "grad_norm": 1.4971363544464111,
+      "learning_rate": 9.229813664596274e-05,
+      "loss": 0.065,
+      "step": 883
+    },
+    {
+      "epoch": 8.185185185185185,
+      "grad_norm": 0.9148528575897217,
+      "learning_rate": 9.217391304347827e-05,
+      "loss": 0.0339,
+      "step": 884
+    },
+    {
+      "epoch": 8.194444444444445,
+      "grad_norm": 5.527726650238037,
+      "learning_rate": 9.20496894409938e-05,
+      "loss": 0.2355,
+      "step": 885
+    },
+    {
+      "epoch": 8.203703703703704,
+      "grad_norm": 1.7691618204116821,
+      "learning_rate": 9.192546583850931e-05,
+      "loss": 0.0613,
+      "step": 886
+    },
+    {
+      "epoch": 8.212962962962964,
+      "grad_norm": 5.282628536224365,
+      "learning_rate": 9.180124223602486e-05,
+      "loss": 0.2057,
+      "step": 887
+    },
+    {
+      "epoch": 8.222222222222221,
+      "grad_norm": 4.678130626678467,
+      "learning_rate": 9.167701863354037e-05,
+      "loss": 0.0744,
+      "step": 888
+    },
+    {
+      "epoch": 8.231481481481481,
+      "grad_norm": 1.7943284511566162,
+      "learning_rate": 9.15527950310559e-05,
+      "loss": 0.0748,
+      "step": 889
+    },
+    {
+      "epoch": 8.24074074074074,
+      "grad_norm": 0.601938784122467,
+      "learning_rate": 9.142857142857143e-05,
+      "loss": 0.0185,
+      "step": 890
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 3.1283457279205322,
+      "learning_rate": 9.130434782608696e-05,
+      "loss": 0.0937,
+      "step": 891
+    },
+    {
+      "epoch": 8.25925925925926,
+      "grad_norm": 3.6321136951446533,
+      "learning_rate": 9.118012422360249e-05,
+      "loss": 0.1203,
+      "step": 892
+    },
+    {
+      "epoch": 8.268518518518519,
+      "grad_norm": 4.4610185623168945,
+      "learning_rate": 9.105590062111802e-05,
+      "loss": 0.2488,
+      "step": 893
+    },
+    {
+      "epoch": 8.277777777777779,
+      "grad_norm": 1.2480530738830566,
+      "learning_rate": 9.093167701863354e-05,
+      "loss": 0.0595,
+      "step": 894
+    },
+    {
+      "epoch": 8.287037037037036,
+      "grad_norm": 3.5497171878814697,
+      "learning_rate": 9.080745341614907e-05,
+      "loss": 0.1169,
+      "step": 895
+    },
+    {
+      "epoch": 8.296296296296296,
+      "grad_norm": 1.2633136510849,
+      "learning_rate": 9.068322981366461e-05,
+      "loss": 0.0435,
+      "step": 896
+    },
+    {
+      "epoch": 8.305555555555555,
+      "grad_norm": 0.9569624066352844,
+      "learning_rate": 9.055900621118013e-05,
+      "loss": 0.0457,
+      "step": 897
+    },
+    {
+      "epoch": 8.314814814814815,
+      "grad_norm": 1.081705927848816,
+      "learning_rate": 9.043478260869566e-05,
+      "loss": 0.0537,
+      "step": 898
+    },
+    {
+      "epoch": 8.324074074074074,
+      "grad_norm": 1.474846601486206,
+      "learning_rate": 9.031055900621119e-05,
+      "loss": 0.0465,
+      "step": 899
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 1.3447438478469849,
+      "learning_rate": 9.01863354037267e-05,
+      "loss": 0.0534,
+      "step": 900
+    },
+    {
+      "epoch": 8.342592592592593,
+      "grad_norm": 4.491641998291016,
+      "learning_rate": 9.006211180124225e-05,
+      "loss": 0.1298,
+      "step": 901
+    },
+    {
+      "epoch": 8.351851851851851,
+      "grad_norm": 0.8619861602783203,
+      "learning_rate": 8.993788819875776e-05,
+      "loss": 0.0237,
+      "step": 902
+    },
+    {
+      "epoch": 8.36111111111111,
+      "grad_norm": 1.2432044744491577,
+      "learning_rate": 8.98136645962733e-05,
+      "loss": 0.0499,
+      "step": 903
+    },
+    {
+      "epoch": 8.37037037037037,
+      "grad_norm": 2.284329891204834,
+      "learning_rate": 8.968944099378882e-05,
+      "loss": 0.0876,
+      "step": 904
+    },
+    {
+      "epoch": 8.37962962962963,
+      "grad_norm": 4.381556034088135,
+      "learning_rate": 8.956521739130435e-05,
+      "loss": 0.1647,
+      "step": 905
+    },
+    {
+      "epoch": 8.38888888888889,
+      "grad_norm": 1.341193675994873,
+      "learning_rate": 8.944099378881988e-05,
+      "loss": 0.0628,
+      "step": 906
+    },
+    {
+      "epoch": 8.398148148148149,
+      "grad_norm": 1.5119465589523315,
+      "learning_rate": 8.931677018633541e-05,
+      "loss": 0.0429,
+      "step": 907
+    },
+    {
+      "epoch": 8.407407407407407,
+      "grad_norm": 1.5674864053726196,
+      "learning_rate": 8.919254658385093e-05,
+      "loss": 0.066,
+      "step": 908
+    },
+    {
+      "epoch": 8.416666666666666,
+      "grad_norm": 2.23675537109375,
+      "learning_rate": 8.906832298136646e-05,
+      "loss": 0.0567,
+      "step": 909
+    },
+    {
+      "epoch": 8.425925925925926,
+      "grad_norm": 1.609828233718872,
+      "learning_rate": 8.894409937888199e-05,
+      "loss": 0.053,
+      "step": 910
+    },
+    {
+      "epoch": 8.435185185185185,
+      "grad_norm": 2.384761095046997,
+      "learning_rate": 8.881987577639752e-05,
+      "loss": 0.0779,
+      "step": 911
+    },
+    {
+      "epoch": 8.444444444444445,
+      "grad_norm": 1.0519819259643555,
+      "learning_rate": 8.869565217391305e-05,
+      "loss": 0.0365,
+      "step": 912
+    },
+    {
+      "epoch": 8.453703703703704,
+      "grad_norm": 2.049403429031372,
+      "learning_rate": 8.857142857142857e-05,
+      "loss": 0.0595,
+      "step": 913
+    },
+    {
+      "epoch": 8.462962962962964,
+      "grad_norm": 3.0430166721343994,
+      "learning_rate": 8.844720496894411e-05,
+      "loss": 0.1162,
+      "step": 914
+    },
+    {
+      "epoch": 8.472222222222221,
+      "grad_norm": 2.900303602218628,
+      "learning_rate": 8.832298136645964e-05,
+      "loss": 0.0528,
+      "step": 915
+    },
+    {
+      "epoch": 8.481481481481481,
+      "grad_norm": 3.444995164871216,
+      "learning_rate": 8.819875776397516e-05,
+      "loss": 0.0581,
+      "step": 916
+    },
+    {
+      "epoch": 8.49074074074074,
+      "grad_norm": 3.1520862579345703,
+      "learning_rate": 8.807453416149069e-05,
+      "loss": 0.0796,
+      "step": 917
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 2.5778634548187256,
+      "learning_rate": 8.795031055900622e-05,
+      "loss": 0.0574,
+      "step": 918
+    },
+    {
+      "epoch": 8.50925925925926,
+      "grad_norm": 1.2131502628326416,
+      "learning_rate": 8.782608695652174e-05,
+      "loss": 0.0589,
+      "step": 919
+    },
+    {
+      "epoch": 8.518518518518519,
+      "grad_norm": 1.663877248764038,
+      "learning_rate": 8.770186335403727e-05,
+      "loss": 0.0636,
+      "step": 920
+    },
+    {
+      "epoch": 8.527777777777779,
+      "grad_norm": 1.3171827793121338,
+      "learning_rate": 8.757763975155279e-05,
+      "loss": 0.0412,
+      "step": 921
+    },
+    {
+      "epoch": 8.537037037037036,
+      "grad_norm": 3.0393033027648926,
+      "learning_rate": 8.745341614906832e-05,
+      "loss": 0.0926,
+      "step": 922
+    },
+    {
+      "epoch": 8.546296296296296,
+      "grad_norm": 3.5446856021881104,
+      "learning_rate": 8.732919254658386e-05,
+      "loss": 0.0902,
+      "step": 923
+    },
+    {
+      "epoch": 8.555555555555555,
+      "grad_norm": 4.914732933044434,
+      "learning_rate": 8.720496894409938e-05,
+      "loss": 0.1173,
+      "step": 924
+    },
+    {
+      "epoch": 8.564814814814815,
+      "grad_norm": 6.190938949584961,
+      "learning_rate": 8.708074534161491e-05,
+      "loss": 0.1385,
+      "step": 925
+    },
+    {
+      "epoch": 8.574074074074074,
+      "grad_norm": 1.2483842372894287,
+      "learning_rate": 8.695652173913044e-05,
+      "loss": 0.0517,
+      "step": 926
+    },
+    {
+      "epoch": 8.583333333333334,
+      "grad_norm": 2.203432083129883,
+      "learning_rate": 8.683229813664596e-05,
+      "loss": 0.1012,
+      "step": 927
+    },
+    {
+      "epoch": 8.592592592592592,
+      "grad_norm": 1.5815249681472778,
+      "learning_rate": 8.67080745341615e-05,
+      "loss": 0.0935,
+      "step": 928
+    },
+    {
+      "epoch": 8.601851851851851,
+      "grad_norm": 6.097355842590332,
+      "learning_rate": 8.658385093167703e-05,
+      "loss": 0.1332,
+      "step": 929
+    },
+    {
+      "epoch": 8.61111111111111,
+      "grad_norm": 1.781144618988037,
+      "learning_rate": 8.645962732919255e-05,
+      "loss": 0.0739,
+      "step": 930
+    },
+    {
+      "epoch": 8.62037037037037,
+      "grad_norm": 10.972087860107422,
+      "learning_rate": 8.633540372670808e-05,
+      "loss": 0.2456,
+      "step": 931
+    },
+    {
+      "epoch": 8.62962962962963,
+      "grad_norm": 2.080036163330078,
+      "learning_rate": 8.62111801242236e-05,
+      "loss": 0.0516,
+      "step": 932
+    },
+    {
+      "epoch": 8.63888888888889,
+      "grad_norm": 3.2868905067443848,
+      "learning_rate": 8.608695652173914e-05,
+      "loss": 0.0682,
+      "step": 933
+    },
+    {
+      "epoch": 8.648148148148149,
+      "grad_norm": 1.0348496437072754,
+      "learning_rate": 8.596273291925467e-05,
+      "loss": 0.0435,
+      "step": 934
+    },
+    {
+      "epoch": 8.657407407407408,
+      "grad_norm": 2.882920026779175,
+      "learning_rate": 8.583850931677018e-05,
+      "loss": 0.0909,
+      "step": 935
+    },
+    {
+      "epoch": 8.666666666666666,
+      "grad_norm": 2.1770124435424805,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 0.0953,
+      "step": 936
+    },
+    {
+      "epoch": 8.675925925925926,
+      "grad_norm": 1.0724560022354126,
+      "learning_rate": 8.559006211180126e-05,
+      "loss": 0.0421,
+      "step": 937
+    },
+    {
+      "epoch": 8.685185185185185,
+      "grad_norm": 1.1822600364685059,
+      "learning_rate": 8.546583850931677e-05,
+      "loss": 0.0382,
+      "step": 938
+    },
+    {
+      "epoch": 8.694444444444445,
+      "grad_norm": 1.566285490989685,
+      "learning_rate": 8.53416149068323e-05,
+      "loss": 0.0548,
+      "step": 939
+    },
+    {
+      "epoch": 8.703703703703704,
+      "grad_norm": 3.2085978984832764,
+      "learning_rate": 8.521739130434783e-05,
+      "loss": 0.0455,
+      "step": 940
+    },
+    {
+      "epoch": 8.712962962962964,
+      "grad_norm": 1.2845832109451294,
+      "learning_rate": 8.509316770186336e-05,
+      "loss": 0.0439,
+      "step": 941
+    },
+    {
+      "epoch": 8.722222222222221,
+      "grad_norm": 1.9509384632110596,
+      "learning_rate": 8.496894409937889e-05,
+      "loss": 0.0869,
+      "step": 942
+    },
+    {
+      "epoch": 8.731481481481481,
+      "grad_norm": 1.4507602453231812,
+      "learning_rate": 8.484472049689441e-05,
+      "loss": 0.0512,
+      "step": 943
+    },
+    {
+      "epoch": 8.74074074074074,
+      "grad_norm": 5.29793119430542,
+      "learning_rate": 8.472049689440994e-05,
+      "loss": 0.1188,
+      "step": 944
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.0561885833740234,
+      "learning_rate": 8.459627329192547e-05,
+      "loss": 0.0446,
+      "step": 945
+    },
+    {
+      "epoch": 8.75925925925926,
+      "grad_norm": 0.9933180212974548,
+      "learning_rate": 8.4472049689441e-05,
+      "loss": 0.0251,
+      "step": 946
+    },
+    {
+      "epoch": 8.768518518518519,
+      "grad_norm": 8.12339973449707,
+      "learning_rate": 8.434782608695653e-05,
+      "loss": 0.0529,
+      "step": 947
+    },
+    {
+      "epoch": 8.777777777777779,
+      "grad_norm": 1.750126600265503,
+      "learning_rate": 8.422360248447206e-05,
+      "loss": 0.0537,
+      "step": 948
+    },
+    {
+      "epoch": 8.787037037037036,
+      "grad_norm": 2.373239278793335,
+      "learning_rate": 8.409937888198757e-05,
+      "loss": 0.1339,
+      "step": 949
+    },
+    {
+      "epoch": 8.796296296296296,
+      "grad_norm": 1.7778215408325195,
+      "learning_rate": 8.397515527950312e-05,
+      "loss": 0.0794,
+      "step": 950
+    },
+    {
+      "epoch": 8.805555555555555,
+      "grad_norm": 4.6339430809021,
+      "learning_rate": 8.385093167701863e-05,
+      "loss": 0.0935,
+      "step": 951
+    },
+    {
+      "epoch": 8.814814814814815,
+      "grad_norm": 2.691041946411133,
+      "learning_rate": 8.372670807453416e-05,
+      "loss": 0.0981,
+      "step": 952
+    },
+    {
+      "epoch": 8.824074074074074,
+      "grad_norm": 1.270849347114563,
+      "learning_rate": 8.36024844720497e-05,
+      "loss": 0.0512,
+      "step": 953
+    },
+    {
+      "epoch": 8.833333333333334,
+      "grad_norm": 0.8391767144203186,
+      "learning_rate": 8.347826086956521e-05,
+      "loss": 0.0233,
+      "step": 954
+    },
+    {
+      "epoch": 8.842592592592592,
+      "grad_norm": 2.3432822227478027,
+      "learning_rate": 8.335403726708075e-05,
+      "loss": 0.0834,
+      "step": 955
+    },
+    {
+      "epoch": 8.851851851851851,
+      "grad_norm": 1.2851802110671997,
+      "learning_rate": 8.322981366459628e-05,
+      "loss": 0.0468,
+      "step": 956
+    },
+    {
+      "epoch": 8.86111111111111,
+      "grad_norm": 1.9697790145874023,
+      "learning_rate": 8.31055900621118e-05,
+      "loss": 0.0759,
+      "step": 957
+    },
+    {
+      "epoch": 8.87037037037037,
+      "grad_norm": 2.5893197059631348,
+      "learning_rate": 8.298136645962733e-05,
+      "loss": 0.1465,
+      "step": 958
+    },
+    {
+      "epoch": 8.87962962962963,
+      "grad_norm": 2.174973249435425,
+      "learning_rate": 8.285714285714287e-05,
+      "loss": 0.0834,
+      "step": 959
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.8256950974464417,
+      "learning_rate": 8.273291925465839e-05,
+      "loss": 0.0349,
+      "step": 960
+    },
+    {
+      "epoch": 8.898148148148149,
+      "grad_norm": 1.0629161596298218,
+      "learning_rate": 8.260869565217392e-05,
+      "loss": 0.0447,
+      "step": 961
+    },
+    {
+      "epoch": 8.907407407407408,
+      "grad_norm": 1.6304731369018555,
+      "learning_rate": 8.248447204968945e-05,
+      "loss": 0.0446,
+      "step": 962
+    },
+    {
+      "epoch": 8.916666666666666,
+      "grad_norm": 4.195140361785889,
+      "learning_rate": 8.236024844720497e-05,
+      "loss": 0.1002,
+      "step": 963
+    },
+    {
+      "epoch": 8.925925925925926,
+      "grad_norm": 0.7746004462242126,
+      "learning_rate": 8.223602484472051e-05,
+      "loss": 0.0266,
+      "step": 964
+    },
+    {
+      "epoch": 8.935185185185185,
+      "grad_norm": 0.6275144815444946,
+      "learning_rate": 8.211180124223603e-05,
+      "loss": 0.0238,
+      "step": 965
+    },
+    {
+      "epoch": 8.944444444444445,
+      "grad_norm": 2.2046384811401367,
+      "learning_rate": 8.198757763975156e-05,
+      "loss": 0.0676,
+      "step": 966
+    },
+    {
+      "epoch": 8.953703703703704,
+      "grad_norm": 4.583156585693359,
+      "learning_rate": 8.186335403726709e-05,
+      "loss": 0.0885,
+      "step": 967
+    },
+    {
+      "epoch": 8.962962962962964,
+      "grad_norm": 3.3135993480682373,
+      "learning_rate": 8.173913043478262e-05,
+      "loss": 0.0971,
+      "step": 968
+    },
+    {
+      "epoch": 8.972222222222221,
+      "grad_norm": 1.8723350763320923,
+      "learning_rate": 8.161490683229814e-05,
+      "loss": 0.0651,
+      "step": 969
+    },
+    {
+      "epoch": 8.981481481481481,
+      "grad_norm": 1.9317792654037476,
+      "learning_rate": 8.149068322981367e-05,
+      "loss": 0.0656,
+      "step": 970
+    },
+    {
+      "epoch": 8.99074074074074,
+      "grad_norm": 1.3685884475708008,
+      "learning_rate": 8.136645962732919e-05,
+      "loss": 0.0618,
+      "step": 971
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.9626964330673218,
+      "learning_rate": 8.124223602484472e-05,
+      "loss": 0.0594,
+      "step": 972
+    },
+    {
+      "epoch": 9.00925925925926,
+      "grad_norm": 2.1750247478485107,
+      "learning_rate": 8.111801242236025e-05,
+      "loss": 0.0503,
+      "step": 973
+    },
+    {
+      "epoch": 9.018518518518519,
+      "grad_norm": 1.803802251815796,
+      "learning_rate": 8.099378881987578e-05,
+      "loss": 0.043,
+      "step": 974
+    },
+    {
+      "epoch": 9.027777777777779,
+      "grad_norm": 3.9391958713531494,
+      "learning_rate": 8.086956521739131e-05,
+      "loss": 0.0847,
+      "step": 975
+    },
+    {
+      "epoch": 9.037037037037036,
+      "grad_norm": 1.5313109159469604,
+      "learning_rate": 8.074534161490683e-05,
+      "loss": 0.0599,
+      "step": 976
+    },
+    {
+      "epoch": 9.046296296296296,
+      "grad_norm": 3.568204164505005,
+      "learning_rate": 8.062111801242236e-05,
+      "loss": 0.1605,
+      "step": 977
+    },
+    {
+      "epoch": 9.055555555555555,
+      "grad_norm": 0.7717824578285217,
+      "learning_rate": 8.04968944099379e-05,
+      "loss": 0.0243,
+      "step": 978
+    },
+    {
+      "epoch": 9.064814814814815,
+      "grad_norm": 1.0113966464996338,
+      "learning_rate": 8.037267080745342e-05,
+      "loss": 0.0401,
+      "step": 979
+    },
+    {
+      "epoch": 9.074074074074074,
+      "grad_norm": 1.8266679048538208,
+      "learning_rate": 8.024844720496895e-05,
+      "loss": 0.0533,
+      "step": 980
+    },
+    {
+      "epoch": 9.083333333333334,
+      "grad_norm": 2.636436700820923,
+      "learning_rate": 8.012422360248448e-05,
+      "loss": 0.0923,
+      "step": 981
+    },
+    {
+      "epoch": 9.092592592592593,
+      "grad_norm": 2.6462342739105225,
+      "learning_rate": 8e-05,
+      "loss": 0.0497,
+      "step": 982
+    },
+    {
+      "epoch": 9.101851851851851,
+      "grad_norm": 1.8009244203567505,
+      "learning_rate": 7.987577639751554e-05,
+      "loss": 0.0718,
+      "step": 983
+    },
+    {
+      "epoch": 9.11111111111111,
+      "grad_norm": 1.0373586416244507,
+      "learning_rate": 7.975155279503105e-05,
+      "loss": 0.0269,
+      "step": 984
+    },
+    {
+      "epoch": 9.12037037037037,
+      "grad_norm": 1.1197093725204468,
+      "learning_rate": 7.962732919254658e-05,
+      "loss": 0.0375,
+      "step": 985
+    },
+    {
+      "epoch": 9.12962962962963,
+      "grad_norm": 0.7301254272460938,
+      "learning_rate": 7.950310559006211e-05,
+      "loss": 0.0335,
+      "step": 986
+    },
+    {
+      "epoch": 9.13888888888889,
+      "grad_norm": 1.0725760459899902,
+      "learning_rate": 7.937888198757764e-05,
+      "loss": 0.0392,
+      "step": 987
+    },
+    {
+      "epoch": 9.148148148148149,
+      "grad_norm": 0.7969951033592224,
+      "learning_rate": 7.925465838509317e-05,
+      "loss": 0.0259,
+      "step": 988
+    },
+    {
+      "epoch": 9.157407407407407,
+      "grad_norm": 1.0825790166854858,
+      "learning_rate": 7.91304347826087e-05,
+      "loss": 0.0346,
+      "step": 989
+    },
+    {
+      "epoch": 9.166666666666666,
+      "grad_norm": 2.6907694339752197,
+      "learning_rate": 7.900621118012422e-05,
+      "loss": 0.0462,
+      "step": 990
+    },
+    {
+      "epoch": 9.175925925925926,
+      "grad_norm": 1.146231770515442,
+      "learning_rate": 7.888198757763976e-05,
+      "loss": 0.0422,
+      "step": 991
+    },
+    {
+      "epoch": 9.185185185185185,
+      "grad_norm": 0.8455992341041565,
+      "learning_rate": 7.875776397515529e-05,
+      "loss": 0.0356,
+      "step": 992
+    },
+    {
+      "epoch": 9.194444444444445,
+      "grad_norm": 1.0626360177993774,
+      "learning_rate": 7.863354037267081e-05,
+      "loss": 0.0415,
+      "step": 993
+    },
+    {
+      "epoch": 9.203703703703704,
+      "grad_norm": 1.5166810750961304,
+      "learning_rate": 7.850931677018634e-05,
+      "loss": 0.0477,
+      "step": 994
+    },
+    {
+      "epoch": 9.212962962962964,
+      "grad_norm": 3.8809115886688232,
+      "learning_rate": 7.838509316770187e-05,
+      "loss": 0.0976,
+      "step": 995
+    },
+    {
+      "epoch": 9.222222222222221,
+      "grad_norm": 2.389176368713379,
+      "learning_rate": 7.82608695652174e-05,
+      "loss": 0.0649,
+      "step": 996
+    },
+    {
+      "epoch": 9.231481481481481,
+      "grad_norm": 0.7285131812095642,
+      "learning_rate": 7.813664596273293e-05,
+      "loss": 0.0192,
+      "step": 997
+    },
+    {
+      "epoch": 9.24074074074074,
+      "grad_norm": 2.0538992881774902,
+      "learning_rate": 7.801242236024844e-05,
+      "loss": 0.0743,
+      "step": 998
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 3.144094944000244,
+      "learning_rate": 7.788819875776397e-05,
+      "loss": 0.0677,
+      "step": 999
+    },
+    {
+      "epoch": 9.25925925925926,
+      "grad_norm": 1.6257270574569702,
+      "learning_rate": 7.776397515527952e-05,
+      "loss": 0.0739,
+      "step": 1000
+    },
+    {
+      "epoch": 9.268518518518519,
+      "grad_norm": 3.6802332401275635,
+      "learning_rate": 7.763975155279503e-05,
+      "loss": 0.1243,
+      "step": 1001
+    },
+    {
+      "epoch": 9.277777777777779,
+      "grad_norm": 0.8263121247291565,
+      "learning_rate": 7.751552795031056e-05,
+      "loss": 0.0376,
+      "step": 1002
+    },
+    {
+      "epoch": 9.287037037037036,
+      "grad_norm": 2.52786922454834,
+      "learning_rate": 7.73913043478261e-05,
+      "loss": 0.0423,
+      "step": 1003
+    },
+    {
+      "epoch": 9.296296296296296,
+      "grad_norm": 1.0051758289337158,
+      "learning_rate": 7.726708074534161e-05,
+      "loss": 0.0281,
+      "step": 1004
+    },
+    {
+      "epoch": 9.305555555555555,
+      "grad_norm": 3.8244469165802,
+      "learning_rate": 7.714285714285715e-05,
+      "loss": 0.1488,
+      "step": 1005
+    },
+    {
+      "epoch": 9.314814814814815,
+      "grad_norm": 5.202383518218994,
+      "learning_rate": 7.701863354037267e-05,
+      "loss": 0.1224,
+      "step": 1006
+    },
+    {
+      "epoch": 9.324074074074074,
+      "grad_norm": 0.9186612963676453,
+      "learning_rate": 7.68944099378882e-05,
+      "loss": 0.0289,
+      "step": 1007
+    },
+    {
+      "epoch": 9.333333333333334,
+      "grad_norm": 1.926774263381958,
+      "learning_rate": 7.677018633540373e-05,
+      "loss": 0.0516,
+      "step": 1008
+    },
+    {
+      "epoch": 9.342592592592593,
+      "grad_norm": 4.624532699584961,
+      "learning_rate": 7.664596273291926e-05,
+      "loss": 0.2248,
+      "step": 1009
+    },
+    {
+      "epoch": 9.351851851851851,
+      "grad_norm": 1.0460689067840576,
+      "learning_rate": 7.652173913043479e-05,
+      "loss": 0.0344,
+      "step": 1010
+    },
+    {
+      "epoch": 9.36111111111111,
+      "grad_norm": 1.524003267288208,
+      "learning_rate": 7.639751552795032e-05,
+      "loss": 0.0372,
+      "step": 1011
+    },
+    {
+      "epoch": 9.37037037037037,
+      "grad_norm": 2.419135570526123,
+      "learning_rate": 7.627329192546584e-05,
+      "loss": 0.0545,
+      "step": 1012
+    },
+    {
+      "epoch": 9.37962962962963,
+      "grad_norm": 2.654794454574585,
+      "learning_rate": 7.614906832298137e-05,
+      "loss": 0.0649,
+      "step": 1013
+    },
+    {
+      "epoch": 9.38888888888889,
+      "grad_norm": 2.0655901432037354,
+      "learning_rate": 7.60248447204969e-05,
+      "loss": 0.0731,
+      "step": 1014
+    },
+    {
+      "epoch": 9.398148148148149,
+      "grad_norm": 2.0986382961273193,
+      "learning_rate": 7.590062111801243e-05,
+      "loss": 0.073,
+      "step": 1015
+    },
+    {
+      "epoch": 9.407407407407407,
+      "grad_norm": 3.522204637527466,
+      "learning_rate": 7.577639751552796e-05,
+      "loss": 0.0648,
+      "step": 1016
+    },
+    {
+      "epoch": 9.416666666666666,
+      "grad_norm": 1.7459393739700317,
+      "learning_rate": 7.565217391304347e-05,
+      "loss": 0.0258,
+      "step": 1017
+    },
+    {
+      "epoch": 9.425925925925926,
+      "grad_norm": 2.7577157020568848,
+      "learning_rate": 7.552795031055902e-05,
+      "loss": 0.0928,
+      "step": 1018
+    },
+    {
+      "epoch": 9.435185185185185,
+      "grad_norm": 2.878016710281372,
+      "learning_rate": 7.540372670807454e-05,
+      "loss": 0.0556,
+      "step": 1019
+    },
+    {
+      "epoch": 9.444444444444445,
+      "grad_norm": 2.508056640625,
+      "learning_rate": 7.527950310559006e-05,
+      "loss": 0.1017,
+      "step": 1020
+    },
+    {
+      "epoch": 9.453703703703704,
+      "grad_norm": 2.9259493350982666,
+      "learning_rate": 7.515527950310559e-05,
+      "loss": 0.1015,
+      "step": 1021
+    },
+    {
+      "epoch": 9.462962962962964,
+      "grad_norm": 1.2046016454696655,
+      "learning_rate": 7.503105590062112e-05,
+      "loss": 0.052,
+      "step": 1022
+    },
+    {
+      "epoch": 9.472222222222221,
+      "grad_norm": 0.83958500623703,
+      "learning_rate": 7.490683229813665e-05,
+      "loss": 0.0264,
+      "step": 1023
+    },
+    {
+      "epoch": 9.481481481481481,
+      "grad_norm": 2.824143171310425,
+      "learning_rate": 7.478260869565218e-05,
+      "loss": 0.0801,
+      "step": 1024
+    },
+    {
+      "epoch": 9.49074074074074,
+      "grad_norm": 3.1509366035461426,
+      "learning_rate": 7.465838509316771e-05,
+      "loss": 0.0832,
+      "step": 1025
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 2.522878885269165,
+      "learning_rate": 7.453416149068323e-05,
+      "loss": 0.0588,
+      "step": 1026
+    },
+    {
+      "epoch": 9.50925925925926,
+      "grad_norm": 0.79912930727005,
+      "learning_rate": 7.440993788819877e-05,
+      "loss": 0.0423,
+      "step": 1027
+    },
+    {
+      "epoch": 9.518518518518519,
+      "grad_norm": 2.4580273628234863,
+      "learning_rate": 7.428571428571429e-05,
+      "loss": 0.0771,
+      "step": 1028
+    },
+    {
+      "epoch": 9.527777777777779,
+      "grad_norm": 2.6916251182556152,
+      "learning_rate": 7.416149068322982e-05,
+      "loss": 0.0663,
+      "step": 1029
+    },
+    {
+      "epoch": 9.537037037037036,
+      "grad_norm": 0.7641211748123169,
+      "learning_rate": 7.403726708074535e-05,
+      "loss": 0.0276,
+      "step": 1030
+    },
+    {
+      "epoch": 9.546296296296296,
+      "grad_norm": 2.337360382080078,
+      "learning_rate": 7.391304347826086e-05,
+      "loss": 0.0539,
+      "step": 1031
+    },
+    {
+      "epoch": 9.555555555555555,
+      "grad_norm": 1.247176170349121,
+      "learning_rate": 7.37888198757764e-05,
+      "loss": 0.0633,
+      "step": 1032
+    },
+    {
+      "epoch": 9.564814814814815,
+      "grad_norm": 0.9180092811584473,
+      "learning_rate": 7.366459627329194e-05,
+      "loss": 0.0353,
+      "step": 1033
+    },
+    {
+      "epoch": 9.574074074074074,
+      "grad_norm": 3.825732707977295,
+      "learning_rate": 7.354037267080745e-05,
+      "loss": 0.0991,
+      "step": 1034
+    },
+    {
+      "epoch": 9.583333333333334,
+      "grad_norm": 4.522989273071289,
+      "learning_rate": 7.341614906832298e-05,
+      "loss": 0.3405,
+      "step": 1035
+    },
+    {
+      "epoch": 9.592592592592592,
+      "grad_norm": 4.249613285064697,
+      "learning_rate": 7.329192546583851e-05,
+      "loss": 0.1238,
+      "step": 1036
+    },
+    {
+      "epoch": 9.601851851851851,
+      "grad_norm": 1.2015916109085083,
+      "learning_rate": 7.316770186335404e-05,
+      "loss": 0.0475,
+      "step": 1037
+    },
+    {
+      "epoch": 9.61111111111111,
+      "grad_norm": 2.1656363010406494,
+      "learning_rate": 7.304347826086957e-05,
+      "loss": 0.0668,
+      "step": 1038
+    },
+    {
+      "epoch": 9.62037037037037,
+      "grad_norm": 1.3153462409973145,
+      "learning_rate": 7.291925465838509e-05,
+      "loss": 0.0747,
+      "step": 1039
+    },
+    {
+      "epoch": 9.62962962962963,
+      "grad_norm": 0.8693362474441528,
+      "learning_rate": 7.279503105590062e-05,
+      "loss": 0.0263,
+      "step": 1040
+    },
+    {
+      "epoch": 9.63888888888889,
+      "grad_norm": 0.9345487952232361,
+      "learning_rate": 7.267080745341616e-05,
+      "loss": 0.031,
+      "step": 1041
+    },
+    {
+      "epoch": 9.648148148148149,
+      "grad_norm": 3.389455795288086,
+      "learning_rate": 7.254658385093168e-05,
+      "loss": 0.1584,
+      "step": 1042
+    },
+    {
+      "epoch": 9.657407407407408,
+      "grad_norm": 2.5855634212493896,
+      "learning_rate": 7.242236024844721e-05,
+      "loss": 0.0557,
+      "step": 1043
+    },
+    {
+      "epoch": 9.666666666666666,
+      "grad_norm": 1.2336293458938599,
+      "learning_rate": 7.229813664596274e-05,
+      "loss": 0.0438,
+      "step": 1044
+    },
+    {
+      "epoch": 9.675925925925926,
+      "grad_norm": 4.778486728668213,
+      "learning_rate": 7.217391304347827e-05,
+      "loss": 0.2042,
+      "step": 1045
+    },
+    {
+      "epoch": 9.685185185185185,
+      "grad_norm": 1.5163942575454712,
+      "learning_rate": 7.20496894409938e-05,
+      "loss": 0.0564,
+      "step": 1046
+    },
+    {
+      "epoch": 9.694444444444445,
+      "grad_norm": 0.900606632232666,
+      "learning_rate": 7.192546583850931e-05,
+      "loss": 0.0381,
+      "step": 1047
+    },
+    {
+      "epoch": 9.703703703703704,
+      "grad_norm": 1.1757590770721436,
+      "learning_rate": 7.180124223602484e-05,
+      "loss": 0.0412,
+      "step": 1048
+    },
+    {
+      "epoch": 9.712962962962964,
+      "grad_norm": 3.0201354026794434,
+      "learning_rate": 7.167701863354037e-05,
+      "loss": 0.0674,
+      "step": 1049
+    },
+    {
+      "epoch": 9.722222222222221,
+      "grad_norm": 1.2021557092666626,
+      "learning_rate": 7.15527950310559e-05,
+      "loss": 0.0485,
+      "step": 1050
+    },
+    {
+      "epoch": 9.731481481481481,
+      "grad_norm": 0.8072912096977234,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.0317,
+      "step": 1051
+    },
+    {
+      "epoch": 9.74074074074074,
+      "grad_norm": 3.1552536487579346,
+      "learning_rate": 7.130434782608696e-05,
+      "loss": 0.0777,
+      "step": 1052
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 1.6829980611801147,
+      "learning_rate": 7.118012422360248e-05,
+      "loss": 0.1083,
+      "step": 1053
+    },
+    {
+      "epoch": 9.75925925925926,
+      "grad_norm": 1.2367135286331177,
+      "learning_rate": 7.105590062111802e-05,
+      "loss": 0.061,
+      "step": 1054
+    },
+    {
+      "epoch": 9.768518518518519,
+      "grad_norm": 2.295994997024536,
+      "learning_rate": 7.093167701863354e-05,
+      "loss": 0.1145,
+      "step": 1055
+    },
+    {
+      "epoch": 9.777777777777779,
+      "grad_norm": 1.0769095420837402,
+      "learning_rate": 7.080745341614907e-05,
+      "loss": 0.0448,
+      "step": 1056
+    },
+    {
+      "epoch": 9.787037037037036,
+      "grad_norm": 0.814005434513092,
+      "learning_rate": 7.06832298136646e-05,
+      "loss": 0.0279,
+      "step": 1057
+    },
+    {
+      "epoch": 9.796296296296296,
+      "grad_norm": 2.416013240814209,
+      "learning_rate": 7.055900621118013e-05,
+      "loss": 0.1144,
+      "step": 1058
+    },
+    {
+      "epoch": 9.805555555555555,
+      "grad_norm": 1.4185346364974976,
+      "learning_rate": 7.043478260869566e-05,
+      "loss": 0.0632,
+      "step": 1059
+    },
+    {
+      "epoch": 9.814814814814815,
+      "grad_norm": 1.7376627922058105,
+      "learning_rate": 7.031055900621119e-05,
+      "loss": 0.0452,
+      "step": 1060
+    },
+    {
+      "epoch": 9.824074074074074,
+      "grad_norm": 0.911581814289093,
+      "learning_rate": 7.01863354037267e-05,
+      "loss": 0.0404,
+      "step": 1061
+    },
+    {
+      "epoch": 9.833333333333334,
+      "grad_norm": 0.774612545967102,
+      "learning_rate": 7.006211180124224e-05,
+      "loss": 0.0418,
+      "step": 1062
+    },
+    {
+      "epoch": 9.842592592592592,
+      "grad_norm": 4.290308475494385,
+      "learning_rate": 6.993788819875778e-05,
+      "loss": 0.1378,
+      "step": 1063
+    },
+    {
+      "epoch": 9.851851851851851,
+      "grad_norm": 2.3391199111938477,
+      "learning_rate": 6.98136645962733e-05,
+      "loss": 0.075,
+      "step": 1064
+    },
+    {
+      "epoch": 9.86111111111111,
+      "grad_norm": 2.517441511154175,
+      "learning_rate": 6.968944099378883e-05,
+      "loss": 0.0692,
+      "step": 1065
+    },
+    {
+      "epoch": 9.87037037037037,
+      "grad_norm": 2.425095558166504,
+      "learning_rate": 6.956521739130436e-05,
+      "loss": 0.1089,
+      "step": 1066
+    },
+    {
+      "epoch": 9.87962962962963,
+      "grad_norm": 1.5073820352554321,
+      "learning_rate": 6.944099378881987e-05,
+      "loss": 0.045,
+      "step": 1067
+    },
+    {
+      "epoch": 9.88888888888889,
+      "grad_norm": 1.071384072303772,
+      "learning_rate": 6.931677018633542e-05,
+      "loss": 0.0336,
+      "step": 1068
+    },
+    {
+      "epoch": 9.898148148148149,
+      "grad_norm": 1.1007689237594604,
+      "learning_rate": 6.919254658385093e-05,
+      "loss": 0.0302,
+      "step": 1069
+    },
+    {
+      "epoch": 9.907407407407408,
+      "grad_norm": 4.436883926391602,
+      "learning_rate": 6.906832298136646e-05,
+      "loss": 0.0853,
+      "step": 1070
+    },
+    {
+      "epoch": 9.916666666666666,
+      "grad_norm": 0.7881323099136353,
+      "learning_rate": 6.894409937888199e-05,
+      "loss": 0.0275,
+      "step": 1071
+    },
+    {
+      "epoch": 9.925925925925926,
+      "grad_norm": 4.63049840927124,
+      "learning_rate": 6.881987577639752e-05,
+      "loss": 0.1019,
+      "step": 1072
+    },
+    {
+      "epoch": 9.935185185185185,
+      "grad_norm": 1.3890912532806396,
+      "learning_rate": 6.869565217391305e-05,
+      "loss": 0.0554,
+      "step": 1073
+    },
+    {
+      "epoch": 9.944444444444445,
+      "grad_norm": 1.3758798837661743,
+      "learning_rate": 6.857142857142858e-05,
+      "loss": 0.0474,
+      "step": 1074
+    },
+    {
+      "epoch": 9.953703703703704,
+      "grad_norm": 3.138035774230957,
+      "learning_rate": 6.84472049689441e-05,
+      "loss": 0.1148,
+      "step": 1075
+    },
+    {
+      "epoch": 9.962962962962964,
+      "grad_norm": 3.5089962482452393,
+      "learning_rate": 6.832298136645963e-05,
+      "loss": 0.1058,
+      "step": 1076
+    },
+    {
+      "epoch": 9.972222222222221,
+      "grad_norm": 2.708378314971924,
+      "learning_rate": 6.819875776397516e-05,
+      "loss": 0.1467,
+      "step": 1077
+    },
+    {
+      "epoch": 9.981481481481481,
+      "grad_norm": 1.7580418586730957,
+      "learning_rate": 6.807453416149069e-05,
+      "loss": 0.1067,
+      "step": 1078
+    },
+    {
+      "epoch": 9.99074074074074,
+      "grad_norm": 1.5734944343566895,
+      "learning_rate": 6.795031055900622e-05,
+      "loss": 0.0604,
+      "step": 1079
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.7437562942504883,
+      "learning_rate": 6.782608695652173e-05,
+      "loss": 0.078,
+      "step": 1080
+    },
+    {
+      "epoch": 10.00925925925926,
+      "grad_norm": 1.2374075651168823,
+      "learning_rate": 6.770186335403726e-05,
+      "loss": 0.0517,
+      "step": 1081
+    },
+    {
+      "epoch": 10.018518518518519,
+      "grad_norm": 3.730271816253662,
+      "learning_rate": 6.75776397515528e-05,
+      "loss": 0.0712,
+      "step": 1082
+    },
+    {
+      "epoch": 10.027777777777779,
+      "grad_norm": 1.618425965309143,
+      "learning_rate": 6.745341614906832e-05,
+      "loss": 0.0471,
+      "step": 1083
+    },
+    {
+      "epoch": 10.037037037037036,
+      "grad_norm": 2.101398468017578,
+      "learning_rate": 6.732919254658385e-05,
+      "loss": 0.0438,
+      "step": 1084
+    },
+    {
+      "epoch": 10.046296296296296,
+      "grad_norm": 2.635305166244507,
+      "learning_rate": 6.720496894409938e-05,
+      "loss": 0.1445,
+      "step": 1085
+    },
+    {
+      "epoch": 10.055555555555555,
+      "grad_norm": 1.2977538108825684,
+      "learning_rate": 6.708074534161491e-05,
+      "loss": 0.0367,
+      "step": 1086
+    },
+    {
+      "epoch": 10.064814814814815,
+      "grad_norm": 2.383920192718506,
+      "learning_rate": 6.695652173913044e-05,
+      "loss": 0.066,
+      "step": 1087
+    },
+    {
+      "epoch": 10.074074074074074,
+      "grad_norm": 2.3182485103607178,
+      "learning_rate": 6.683229813664596e-05,
+      "loss": 0.0399,
+      "step": 1088
+    },
+    {
+      "epoch": 10.083333333333334,
+      "grad_norm": 2.031402826309204,
+      "learning_rate": 6.670807453416149e-05,
+      "loss": 0.037,
+      "step": 1089
+    },
+    {
+      "epoch": 10.092592592592593,
+      "grad_norm": 4.036484718322754,
+      "learning_rate": 6.658385093167702e-05,
+      "loss": 0.0377,
+      "step": 1090
+    },
+    {
+      "epoch": 10.101851851851851,
+      "grad_norm": 2.554116725921631,
+      "learning_rate": 6.645962732919255e-05,
+      "loss": 0.1237,
+      "step": 1091
+    },
+    {
+      "epoch": 10.11111111111111,
+      "grad_norm": 2.1571269035339355,
+      "learning_rate": 6.633540372670808e-05,
+      "loss": 0.0415,
+      "step": 1092
+    },
+    {
+      "epoch": 10.12037037037037,
+      "grad_norm": 0.994016706943512,
+      "learning_rate": 6.621118012422361e-05,
+      "loss": 0.0474,
+      "step": 1093
+    },
+    {
+      "epoch": 10.12962962962963,
+      "grad_norm": 0.7662463784217834,
+      "learning_rate": 6.608695652173912e-05,
+      "loss": 0.0263,
+      "step": 1094
+    },
+    {
+      "epoch": 10.13888888888889,
+      "grad_norm": 1.1817699670791626,
+      "learning_rate": 6.596273291925467e-05,
+      "loss": 0.0419,
+      "step": 1095
+    },
+    {
+      "epoch": 10.148148148148149,
+      "grad_norm": 1.2163158655166626,
+      "learning_rate": 6.58385093167702e-05,
+      "loss": 0.0463,
+      "step": 1096
+    },
+    {
+      "epoch": 10.157407407407407,
+      "grad_norm": 0.646700382232666,
+      "learning_rate": 6.571428571428571e-05,
+      "loss": 0.0178,
+      "step": 1097
+    },
+    {
+      "epoch": 10.166666666666666,
+      "grad_norm": 0.7059783935546875,
+      "learning_rate": 6.559006211180124e-05,
+      "loss": 0.0289,
+      "step": 1098
+    },
+    {
+      "epoch": 10.175925925925926,
+      "grad_norm": 4.205524921417236,
+      "learning_rate": 6.546583850931677e-05,
+      "loss": 0.1329,
+      "step": 1099
+    },
+    {
+      "epoch": 10.185185185185185,
+      "grad_norm": 1.418074131011963,
+      "learning_rate": 6.53416149068323e-05,
+      "loss": 0.0412,
+      "step": 1100
+    },
+    {
+      "epoch": 10.194444444444445,
+      "grad_norm": 0.9323771595954895,
+      "learning_rate": 6.521739130434783e-05,
+      "loss": 0.043,
+      "step": 1101
+    },
+    {
+      "epoch": 10.203703703703704,
+      "grad_norm": 1.6822084188461304,
+      "learning_rate": 6.509316770186335e-05,
+      "loss": 0.0402,
+      "step": 1102
+    },
+    {
+      "epoch": 10.212962962962964,
+      "grad_norm": 1.1018824577331543,
+      "learning_rate": 6.496894409937888e-05,
+      "loss": 0.0424,
+      "step": 1103
+    },
+    {
+      "epoch": 10.222222222222221,
+      "grad_norm": 0.921039342880249,
+      "learning_rate": 6.484472049689442e-05,
+      "loss": 0.0251,
+      "step": 1104
+    },
+    {
+      "epoch": 10.231481481481481,
+      "grad_norm": 2.9697976112365723,
+      "learning_rate": 6.472049689440994e-05,
+      "loss": 0.0627,
+      "step": 1105
+    },
+    {
+      "epoch": 10.24074074074074,
+      "grad_norm": 1.5444436073303223,
+      "learning_rate": 6.459627329192547e-05,
+      "loss": 0.0361,
+      "step": 1106
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.545248031616211,
+      "learning_rate": 6.4472049689441e-05,
+      "loss": 0.0571,
+      "step": 1107
+    },
+    {
+      "epoch": 10.25925925925926,
+      "grad_norm": 2.4655678272247314,
+      "learning_rate": 6.434782608695652e-05,
+      "loss": 0.0929,
+      "step": 1108
+    },
+    {
+      "epoch": 10.268518518518519,
+      "grad_norm": 1.0610233545303345,
+      "learning_rate": 6.422360248447206e-05,
+      "loss": 0.0431,
+      "step": 1109
+    },
+    {
+      "epoch": 10.277777777777779,
+      "grad_norm": 1.7690080404281616,
+      "learning_rate": 6.409937888198758e-05,
+      "loss": 0.0542,
+      "step": 1110
+    },
+    {
+      "epoch": 10.287037037037036,
+      "grad_norm": 1.6236170530319214,
+      "learning_rate": 6.39751552795031e-05,
+      "loss": 0.0465,
+      "step": 1111
+    },
+    {
+      "epoch": 10.296296296296296,
+      "grad_norm": 2.925307035446167,
+      "learning_rate": 6.385093167701864e-05,
+      "loss": 0.0608,
+      "step": 1112
+    },
+    {
+      "epoch": 10.305555555555555,
+      "grad_norm": 1.1506783962249756,
+      "learning_rate": 6.372670807453417e-05,
+      "loss": 0.0494,
+      "step": 1113
+    },
+    {
+      "epoch": 10.314814814814815,
+      "grad_norm": 1.6719893217086792,
+      "learning_rate": 6.36024844720497e-05,
+      "loss": 0.0527,
+      "step": 1114
+    },
+    {
+      "epoch": 10.324074074074074,
+      "grad_norm": 1.290839433670044,
+      "learning_rate": 6.347826086956523e-05,
+      "loss": 0.0655,
+      "step": 1115
+    },
+    {
+      "epoch": 10.333333333333334,
+      "grad_norm": 2.2520599365234375,
+      "learning_rate": 6.335403726708074e-05,
+      "loss": 0.0639,
+      "step": 1116
+    },
+    {
+      "epoch": 10.342592592592593,
+      "grad_norm": 7.430666446685791,
+      "learning_rate": 6.322981366459627e-05,
+      "loss": 0.0843,
+      "step": 1117
+    },
+    {
+      "epoch": 10.351851851851851,
+      "grad_norm": 1.3490774631500244,
+      "learning_rate": 6.31055900621118e-05,
+      "loss": 0.0582,
+      "step": 1118
+    },
+    {
+      "epoch": 10.36111111111111,
+      "grad_norm": 2.280447244644165,
+      "learning_rate": 6.298136645962733e-05,
+      "loss": 0.0675,
+      "step": 1119
+    },
+    {
+      "epoch": 10.37037037037037,
+      "grad_norm": 0.784392237663269,
+      "learning_rate": 6.285714285714286e-05,
+      "loss": 0.0369,
+      "step": 1120
+    },
+    {
+      "epoch": 10.37962962962963,
+      "grad_norm": 2.3288660049438477,
+      "learning_rate": 6.273291925465838e-05,
+      "loss": 0.0548,
+      "step": 1121
+    },
+    {
+      "epoch": 10.38888888888889,
+      "grad_norm": 0.6705870628356934,
+      "learning_rate": 6.260869565217392e-05,
+      "loss": 0.0257,
+      "step": 1122
+    },
+    {
+      "epoch": 10.398148148148149,
+      "grad_norm": 1.070380449295044,
+      "learning_rate": 6.248447204968945e-05,
+      "loss": 0.0457,
+      "step": 1123
+    },
+    {
+      "epoch": 10.407407407407407,
+      "grad_norm": 1.0458325147628784,
+      "learning_rate": 6.236024844720497e-05,
+      "loss": 0.0443,
+      "step": 1124
+    },
+    {
+      "epoch": 10.416666666666666,
+      "grad_norm": 2.423663377761841,
+      "learning_rate": 6.22360248447205e-05,
+      "loss": 0.0768,
+      "step": 1125
+    },
+    {
+      "epoch": 10.425925925925926,
+      "grad_norm": 0.707668662071228,
+      "learning_rate": 6.211180124223603e-05,
+      "loss": 0.0245,
+      "step": 1126
+    },
+    {
+      "epoch": 10.435185185185185,
+      "grad_norm": 0.9784311056137085,
+      "learning_rate": 6.198757763975156e-05,
+      "loss": 0.0356,
+      "step": 1127
+    },
+    {
+      "epoch": 10.444444444444445,
+      "grad_norm": 1.0185497999191284,
+      "learning_rate": 6.186335403726709e-05,
+      "loss": 0.0408,
+      "step": 1128
+    },
+    {
+      "epoch": 10.453703703703704,
+      "grad_norm": 2.9687418937683105,
+      "learning_rate": 6.173913043478262e-05,
+      "loss": 0.0602,
+      "step": 1129
+    },
+    {
+      "epoch": 10.462962962962964,
+      "grad_norm": 0.9770888090133667,
+      "learning_rate": 6.161490683229813e-05,
+      "loss": 0.0404,
+      "step": 1130
+    },
+    {
+      "epoch": 10.472222222222221,
+      "grad_norm": 1.217678427696228,
+      "learning_rate": 6.149068322981368e-05,
+      "loss": 0.0381,
+      "step": 1131
+    },
+    {
+      "epoch": 10.481481481481481,
+      "grad_norm": 0.7180342674255371,
+      "learning_rate": 6.136645962732919e-05,
+      "loss": 0.0266,
+      "step": 1132
+    },
+    {
+      "epoch": 10.49074074074074,
+      "grad_norm": 3.5172603130340576,
+      "learning_rate": 6.124223602484472e-05,
+      "loss": 0.0743,
+      "step": 1133
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 1.9250236749649048,
+      "learning_rate": 6.111801242236025e-05,
+      "loss": 0.0557,
+      "step": 1134
+    },
+    {
+      "epoch": 10.50925925925926,
+      "grad_norm": 3.3495516777038574,
+      "learning_rate": 6.0993788819875776e-05,
+      "loss": 0.0715,
+      "step": 1135
+    },
+    {
+      "epoch": 10.518518518518519,
+      "grad_norm": 1.3475662469863892,
+      "learning_rate": 6.086956521739131e-05,
+      "loss": 0.0692,
+      "step": 1136
+    },
+    {
+      "epoch": 10.527777777777779,
+      "grad_norm": 0.7849929928779602,
+      "learning_rate": 6.0745341614906836e-05,
+      "loss": 0.0263,
+      "step": 1137
+    },
+    {
+      "epoch": 10.537037037037036,
+      "grad_norm": 1.3163514137268066,
+      "learning_rate": 6.062111801242236e-05,
+      "loss": 0.0406,
+      "step": 1138
+    },
+    {
+      "epoch": 10.546296296296296,
+      "grad_norm": 0.7919923663139343,
+      "learning_rate": 6.049689440993789e-05,
+      "loss": 0.0361,
+      "step": 1139
+    },
+    {
+      "epoch": 10.555555555555555,
+      "grad_norm": 1.334409475326538,
+      "learning_rate": 6.0372670807453425e-05,
+      "loss": 0.0446,
+      "step": 1140
+    },
+    {
+      "epoch": 10.564814814814815,
+      "grad_norm": 3.7500407695770264,
+      "learning_rate": 6.024844720496895e-05,
+      "loss": 0.0969,
+      "step": 1141
+    },
+    {
+      "epoch": 10.574074074074074,
+      "grad_norm": 1.9532806873321533,
+      "learning_rate": 6.012422360248448e-05,
+      "loss": 0.0426,
+      "step": 1142
+    },
+    {
+      "epoch": 10.583333333333334,
+      "grad_norm": 1.429587483406067,
+      "learning_rate": 6e-05,
+      "loss": 0.0647,
+      "step": 1143
+    },
+    {
+      "epoch": 10.592592592592592,
+      "grad_norm": 0.8298918604850769,
+      "learning_rate": 5.9875776397515525e-05,
+      "loss": 0.0336,
+      "step": 1144
+    },
+    {
+      "epoch": 10.601851851851851,
+      "grad_norm": 2.578657388687134,
+      "learning_rate": 5.975155279503106e-05,
+      "loss": 0.1264,
+      "step": 1145
+    },
+    {
+      "epoch": 10.61111111111111,
+      "grad_norm": 0.9398471117019653,
+      "learning_rate": 5.962732919254659e-05,
+      "loss": 0.0354,
+      "step": 1146
+    },
+    {
+      "epoch": 10.62037037037037,
+      "grad_norm": 2.5847291946411133,
+      "learning_rate": 5.9503105590062114e-05,
+      "loss": 0.0763,
+      "step": 1147
+    },
+    {
+      "epoch": 10.62962962962963,
+      "grad_norm": 1.649775505065918,
+      "learning_rate": 5.937888198757764e-05,
+      "loss": 0.0329,
+      "step": 1148
+    },
+    {
+      "epoch": 10.63888888888889,
+      "grad_norm": 0.9102606177330017,
+      "learning_rate": 5.9254658385093174e-05,
+      "loss": 0.0396,
+      "step": 1149
+    },
+    {
+      "epoch": 10.648148148148149,
+      "grad_norm": 0.9106652736663818,
+      "learning_rate": 5.9130434782608704e-05,
+      "loss": 0.0326,
+      "step": 1150
+    },
+    {
+      "epoch": 10.657407407407408,
+      "grad_norm": 1.1053392887115479,
+      "learning_rate": 5.900621118012423e-05,
+      "loss": 0.0484,
+      "step": 1151
+    },
+    {
+      "epoch": 10.666666666666666,
+      "grad_norm": 0.9101239442825317,
+      "learning_rate": 5.888198757763975e-05,
+      "loss": 0.0368,
+      "step": 1152
+    },
+    {
+      "epoch": 10.675925925925926,
+      "grad_norm": 0.8673473000526428,
+      "learning_rate": 5.875776397515528e-05,
+      "loss": 0.0308,
+      "step": 1153
+    },
+    {
+      "epoch": 10.685185185185185,
+      "grad_norm": 0.9925929307937622,
+      "learning_rate": 5.863354037267082e-05,
+      "loss": 0.0415,
+      "step": 1154
+    },
+    {
+      "epoch": 10.694444444444445,
+      "grad_norm": 0.7710978388786316,
+      "learning_rate": 5.850931677018634e-05,
+      "loss": 0.0279,
+      "step": 1155
+    },
+    {
+      "epoch": 10.703703703703704,
+      "grad_norm": 1.1936169862747192,
+      "learning_rate": 5.838509316770186e-05,
+      "loss": 0.0393,
+      "step": 1156
+    },
+    {
+      "epoch": 10.712962962962964,
+      "grad_norm": 1.6933677196502686,
+      "learning_rate": 5.826086956521739e-05,
+      "loss": 0.0745,
+      "step": 1157
+    },
+    {
+      "epoch": 10.722222222222221,
+      "grad_norm": 1.3728935718536377,
+      "learning_rate": 5.813664596273293e-05,
+      "loss": 0.0541,
+      "step": 1158
+    },
+    {
+      "epoch": 10.731481481481481,
+      "grad_norm": 0.8559166193008423,
+      "learning_rate": 5.801242236024845e-05,
+      "loss": 0.0298,
+      "step": 1159
+    },
+    {
+      "epoch": 10.74074074074074,
+      "grad_norm": 0.8479191660881042,
+      "learning_rate": 5.7888198757763976e-05,
+      "loss": 0.0291,
+      "step": 1160
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 3.2553343772888184,
+      "learning_rate": 5.7763975155279506e-05,
+      "loss": 0.0713,
+      "step": 1161
+    },
+    {
+      "epoch": 10.75925925925926,
+      "grad_norm": 0.7682176828384399,
+      "learning_rate": 5.763975155279503e-05,
+      "loss": 0.0244,
+      "step": 1162
+    },
+    {
+      "epoch": 10.768518518518519,
+      "grad_norm": 0.8638216257095337,
+      "learning_rate": 5.7515527950310566e-05,
+      "loss": 0.0238,
+      "step": 1163
+    },
+    {
+      "epoch": 10.777777777777779,
+      "grad_norm": 1.0196231603622437,
+      "learning_rate": 5.739130434782609e-05,
+      "loss": 0.0309,
+      "step": 1164
+    },
+    {
+      "epoch": 10.787037037037036,
+      "grad_norm": 0.9703147411346436,
+      "learning_rate": 5.726708074534162e-05,
+      "loss": 0.0308,
+      "step": 1165
+    },
+    {
+      "epoch": 10.796296296296296,
+      "grad_norm": 0.8336690068244934,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.0272,
+      "step": 1166
+    },
+    {
+      "epoch": 10.805555555555555,
+      "grad_norm": 0.5598458051681519,
+      "learning_rate": 5.701863354037268e-05,
+      "loss": 0.0182,
+      "step": 1167
+    },
+    {
+      "epoch": 10.814814814814815,
+      "grad_norm": 0.8961607813835144,
+      "learning_rate": 5.68944099378882e-05,
+      "loss": 0.0313,
+      "step": 1168
+    },
+    {
+      "epoch": 10.824074074074074,
+      "grad_norm": 4.9854207038879395,
+      "learning_rate": 5.677018633540373e-05,
+      "loss": 0.1544,
+      "step": 1169
+    },
+    {
+      "epoch": 10.833333333333334,
+      "grad_norm": 1.967529058456421,
+      "learning_rate": 5.6645962732919255e-05,
+      "loss": 0.0656,
+      "step": 1170
+    },
+    {
+      "epoch": 10.842592592592592,
+      "grad_norm": 1.4599827527999878,
+      "learning_rate": 5.652173913043478e-05,
+      "loss": 0.0564,
+      "step": 1171
+    },
+    {
+      "epoch": 10.851851851851851,
+      "grad_norm": 0.9949111342430115,
+      "learning_rate": 5.6397515527950314e-05,
+      "loss": 0.0314,
+      "step": 1172
+    },
+    {
+      "epoch": 10.86111111111111,
+      "grad_norm": 1.4174683094024658,
+      "learning_rate": 5.6273291925465844e-05,
+      "loss": 0.0552,
+      "step": 1173
+    },
+    {
+      "epoch": 10.87037037037037,
+      "grad_norm": 1.6052273511886597,
+      "learning_rate": 5.614906832298137e-05,
+      "loss": 0.0557,
+      "step": 1174
+    },
+    {
+      "epoch": 10.87962962962963,
+      "grad_norm": 1.5254520177841187,
+      "learning_rate": 5.60248447204969e-05,
+      "loss": 0.0595,
+      "step": 1175
+    },
+    {
+      "epoch": 10.88888888888889,
+      "grad_norm": 1.0831390619277954,
+      "learning_rate": 5.590062111801242e-05,
+      "loss": 0.0364,
+      "step": 1176
+    },
+    {
+      "epoch": 10.898148148148149,
+      "grad_norm": 2.996645212173462,
+      "learning_rate": 5.577639751552796e-05,
+      "loss": 0.1694,
+      "step": 1177
+    },
+    {
+      "epoch": 10.907407407407408,
+      "grad_norm": 1.1233099699020386,
+      "learning_rate": 5.565217391304348e-05,
+      "loss": 0.0422,
+      "step": 1178
+    },
+    {
+      "epoch": 10.916666666666666,
+      "grad_norm": 1.1308964490890503,
+      "learning_rate": 5.552795031055901e-05,
+      "loss": 0.0414,
+      "step": 1179
+    },
+    {
+      "epoch": 10.925925925925926,
+      "grad_norm": 0.7682093381881714,
+      "learning_rate": 5.540372670807453e-05,
+      "loss": 0.0219,
+      "step": 1180
+    },
+    {
+      "epoch": 10.935185185185185,
+      "grad_norm": 1.0228471755981445,
+      "learning_rate": 5.527950310559007e-05,
+      "loss": 0.0451,
+      "step": 1181
+    },
+    {
+      "epoch": 10.944444444444445,
+      "grad_norm": 0.8916832208633423,
+      "learning_rate": 5.515527950310559e-05,
+      "loss": 0.0315,
+      "step": 1182
+    },
+    {
+      "epoch": 10.953703703703704,
+      "grad_norm": 2.371509313583374,
+      "learning_rate": 5.503105590062112e-05,
+      "loss": 0.0667,
+      "step": 1183
+    },
+    {
+      "epoch": 10.962962962962964,
+      "grad_norm": 1.1845130920410156,
+      "learning_rate": 5.4906832298136646e-05,
+      "loss": 0.0391,
+      "step": 1184
+    },
+    {
+      "epoch": 10.972222222222221,
+      "grad_norm": 1.607428789138794,
+      "learning_rate": 5.478260869565217e-05,
+      "loss": 0.0506,
+      "step": 1185
+    },
+    {
+      "epoch": 10.981481481481481,
+      "grad_norm": 3.609421968460083,
+      "learning_rate": 5.4658385093167706e-05,
+      "loss": 0.1342,
+      "step": 1186
+    },
+    {
+      "epoch": 10.99074074074074,
+      "grad_norm": 2.8664135932922363,
+      "learning_rate": 5.4534161490683236e-05,
+      "loss": 0.1277,
+      "step": 1187
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 3.0398714542388916,
+      "learning_rate": 5.440993788819876e-05,
+      "loss": 0.102,
+      "step": 1188
+    },
+    {
+      "epoch": 11.00925925925926,
+      "grad_norm": 0.8366669416427612,
+      "learning_rate": 5.428571428571428e-05,
+      "loss": 0.0293,
+      "step": 1189
+    },
+    {
+      "epoch": 11.018518518518519,
+      "grad_norm": 0.8883122205734253,
+      "learning_rate": 5.416149068322982e-05,
+      "loss": 0.0401,
+      "step": 1190
+    },
+    {
+      "epoch": 11.027777777777779,
+      "grad_norm": 0.9267494678497314,
+      "learning_rate": 5.403726708074535e-05,
+      "loss": 0.0403,
+      "step": 1191
+    },
+    {
+      "epoch": 11.037037037037036,
+      "grad_norm": 0.8485496044158936,
+      "learning_rate": 5.391304347826087e-05,
+      "loss": 0.0272,
+      "step": 1192
+    },
+    {
+      "epoch": 11.046296296296296,
+      "grad_norm": 0.8455182909965515,
+      "learning_rate": 5.3788819875776395e-05,
+      "loss": 0.0277,
+      "step": 1193
+    },
+    {
+      "epoch": 11.055555555555555,
+      "grad_norm": 0.47915375232696533,
+      "learning_rate": 5.3664596273291925e-05,
+      "loss": 0.0132,
+      "step": 1194
+    },
+    {
+      "epoch": 11.064814814814815,
+      "grad_norm": 1.502824068069458,
+      "learning_rate": 5.354037267080746e-05,
+      "loss": 0.0531,
+      "step": 1195
+    },
+    {
+      "epoch": 11.074074074074074,
+      "grad_norm": 0.8204081654548645,
+      "learning_rate": 5.3416149068322984e-05,
+      "loss": 0.0349,
+      "step": 1196
+    },
+    {
+      "epoch": 11.083333333333334,
+      "grad_norm": 0.6424718499183655,
+      "learning_rate": 5.329192546583851e-05,
+      "loss": 0.0163,
+      "step": 1197
+    },
+    {
+      "epoch": 11.092592592592593,
+      "grad_norm": 1.192146897315979,
+      "learning_rate": 5.316770186335404e-05,
+      "loss": 0.0575,
+      "step": 1198
+    },
+    {
+      "epoch": 11.101851851851851,
+      "grad_norm": 0.9034618735313416,
+      "learning_rate": 5.3043478260869574e-05,
+      "loss": 0.0339,
+      "step": 1199
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 0.9357187151908875,
+      "learning_rate": 5.29192546583851e-05,
+      "loss": 0.0301,
+      "step": 1200
+    },
+    {
+      "epoch": 11.12037037037037,
+      "grad_norm": 0.802025318145752,
+      "learning_rate": 5.279503105590062e-05,
+      "loss": 0.033,
+      "step": 1201
+    },
+    {
+      "epoch": 11.12962962962963,
+      "grad_norm": 0.8969548940658569,
+      "learning_rate": 5.267080745341615e-05,
+      "loss": 0.038,
+      "step": 1202
+    },
+    {
+      "epoch": 11.13888888888889,
+      "grad_norm": 0.8712778091430664,
+      "learning_rate": 5.254658385093167e-05,
+      "loss": 0.0417,
+      "step": 1203
+    },
+    {
+      "epoch": 11.148148148148149,
+      "grad_norm": 1.1687285900115967,
+      "learning_rate": 5.242236024844721e-05,
+      "loss": 0.0394,
+      "step": 1204
+    },
+    {
+      "epoch": 11.157407407407407,
+      "grad_norm": 0.8223320841789246,
+      "learning_rate": 5.229813664596273e-05,
+      "loss": 0.0245,
+      "step": 1205
+    },
+    {
+      "epoch": 11.166666666666666,
+      "grad_norm": 1.6229560375213623,
+      "learning_rate": 5.217391304347826e-05,
+      "loss": 0.0486,
+      "step": 1206
+    },
+    {
+      "epoch": 11.175925925925926,
+      "grad_norm": 0.4597468376159668,
+      "learning_rate": 5.2049689440993786e-05,
+      "loss": 0.0116,
+      "step": 1207
+    },
+    {
+      "epoch": 11.185185185185185,
+      "grad_norm": 0.9737257957458496,
+      "learning_rate": 5.192546583850932e-05,
+      "loss": 0.0296,
+      "step": 1208
+    },
+    {
+      "epoch": 11.194444444444445,
+      "grad_norm": 1.6504062414169312,
+      "learning_rate": 5.180124223602485e-05,
+      "loss": 0.0711,
+      "step": 1209
+    },
+    {
+      "epoch": 11.203703703703704,
+      "grad_norm": 1.0058763027191162,
+      "learning_rate": 5.1677018633540376e-05,
+      "loss": 0.0392,
+      "step": 1210
+    },
+    {
+      "epoch": 11.212962962962964,
+      "grad_norm": 0.9670447707176208,
+      "learning_rate": 5.15527950310559e-05,
+      "loss": 0.0411,
+      "step": 1211
+    },
+    {
+      "epoch": 11.222222222222221,
+      "grad_norm": 0.8737079501152039,
+      "learning_rate": 5.142857142857143e-05,
+      "loss": 0.0258,
+      "step": 1212
+    },
+    {
+      "epoch": 11.231481481481481,
+      "grad_norm": 0.8489564061164856,
+      "learning_rate": 5.1304347826086966e-05,
+      "loss": 0.0355,
+      "step": 1213
+    },
+    {
+      "epoch": 11.24074074074074,
+      "grad_norm": 0.8969002366065979,
+      "learning_rate": 5.118012422360249e-05,
+      "loss": 0.0415,
+      "step": 1214
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 0.8433552384376526,
+      "learning_rate": 5.105590062111801e-05,
+      "loss": 0.0359,
+      "step": 1215
+    },
+    {
+      "epoch": 11.25925925925926,
+      "grad_norm": 1.3878159523010254,
+      "learning_rate": 5.093167701863354e-05,
+      "loss": 0.055,
+      "step": 1216
+    },
+    {
+      "epoch": 11.268518518518519,
+      "grad_norm": 1.1686030626296997,
+      "learning_rate": 5.080745341614908e-05,
+      "loss": 0.0403,
+      "step": 1217
+    },
+    {
+      "epoch": 11.277777777777779,
+      "grad_norm": 0.8048990964889526,
+      "learning_rate": 5.06832298136646e-05,
+      "loss": 0.0322,
+      "step": 1218
+    },
+    {
+      "epoch": 11.287037037037036,
+      "grad_norm": 0.661250114440918,
+      "learning_rate": 5.0559006211180125e-05,
+      "loss": 0.0259,
+      "step": 1219
+    },
+    {
+      "epoch": 11.296296296296296,
+      "grad_norm": 0.7250069975852966,
+      "learning_rate": 5.0434782608695655e-05,
+      "loss": 0.0239,
+      "step": 1220
+    },
+    {
+      "epoch": 11.305555555555555,
+      "grad_norm": 1.1156517267227173,
+      "learning_rate": 5.031055900621118e-05,
+      "loss": 0.0347,
+      "step": 1221
+    },
+    {
+      "epoch": 11.314814814814815,
+      "grad_norm": 0.941217303276062,
+      "learning_rate": 5.0186335403726714e-05,
+      "loss": 0.0318,
+      "step": 1222
+    },
+    {
+      "epoch": 11.324074074074074,
+      "grad_norm": 4.407271385192871,
+      "learning_rate": 5.006211180124224e-05,
+      "loss": 0.0731,
+      "step": 1223
+    },
+    {
+      "epoch": 11.333333333333334,
+      "grad_norm": 1.0587486028671265,
+      "learning_rate": 4.993788819875777e-05,
+      "loss": 0.042,
+      "step": 1224
+    },
+    {
+      "epoch": 11.342592592592593,
+      "grad_norm": 1.2566124200820923,
+      "learning_rate": 4.98136645962733e-05,
+      "loss": 0.0387,
+      "step": 1225
+    },
+    {
+      "epoch": 11.351851851851851,
+      "grad_norm": 0.9527769684791565,
+      "learning_rate": 4.968944099378882e-05,
+      "loss": 0.0415,
+      "step": 1226
+    },
+    {
+      "epoch": 11.36111111111111,
+      "grad_norm": 1.3963499069213867,
+      "learning_rate": 4.956521739130435e-05,
+      "loss": 0.0543,
+      "step": 1227
+    },
+    {
+      "epoch": 11.37037037037037,
+      "grad_norm": 0.9101176261901855,
+      "learning_rate": 4.944099378881988e-05,
+      "loss": 0.0387,
+      "step": 1228
+    },
+    {
+      "epoch": 11.37962962962963,
+      "grad_norm": 1.1519862413406372,
+      "learning_rate": 4.93167701863354e-05,
+      "loss": 0.0381,
+      "step": 1229
+    },
+    {
+      "epoch": 11.38888888888889,
+      "grad_norm": 1.6292718648910522,
+      "learning_rate": 4.919254658385093e-05,
+      "loss": 0.0682,
+      "step": 1230
+    },
+    {
+      "epoch": 11.398148148148149,
+      "grad_norm": 0.717383623123169,
+      "learning_rate": 4.906832298136646e-05,
+      "loss": 0.0306,
+      "step": 1231
+    },
+    {
+      "epoch": 11.407407407407407,
+      "grad_norm": 1.2665631771087646,
+      "learning_rate": 4.894409937888199e-05,
+      "loss": 0.064,
+      "step": 1232
+    },
+    {
+      "epoch": 11.416666666666666,
+      "grad_norm": 0.914641261100769,
+      "learning_rate": 4.8819875776397516e-05,
+      "loss": 0.0319,
+      "step": 1233
+    },
+    {
+      "epoch": 11.425925925925926,
+      "grad_norm": 3.159534454345703,
+      "learning_rate": 4.8695652173913046e-05,
+      "loss": 0.1141,
+      "step": 1234
+    },
+    {
+      "epoch": 11.435185185185185,
+      "grad_norm": 0.8833385109901428,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.0276,
+      "step": 1235
+    },
+    {
+      "epoch": 11.444444444444445,
+      "grad_norm": 0.916418194770813,
+      "learning_rate": 4.8447204968944106e-05,
+      "loss": 0.0311,
+      "step": 1236
+    },
+    {
+      "epoch": 11.453703703703704,
+      "grad_norm": 1.092687964439392,
+      "learning_rate": 4.832298136645963e-05,
+      "loss": 0.048,
+      "step": 1237
+    },
+    {
+      "epoch": 11.462962962962964,
+      "grad_norm": 0.7379303574562073,
+      "learning_rate": 4.819875776397515e-05,
+      "loss": 0.0241,
+      "step": 1238
+    },
+    {
+      "epoch": 11.472222222222221,
+      "grad_norm": 1.1241528987884521,
+      "learning_rate": 4.807453416149069e-05,
+      "loss": 0.0429,
+      "step": 1239
+    },
+    {
+      "epoch": 11.481481481481481,
+      "grad_norm": 2.250471591949463,
+      "learning_rate": 4.795031055900621e-05,
+      "loss": 0.0408,
+      "step": 1240
+    },
+    {
+      "epoch": 11.49074074074074,
+      "grad_norm": 1.1494890451431274,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.0501,
+      "step": 1241
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 0.8903128504753113,
+      "learning_rate": 4.770186335403727e-05,
+      "loss": 0.0281,
+      "step": 1242
+    },
+    {
+      "epoch": 11.50925925925926,
+      "grad_norm": 0.6605562567710876,
+      "learning_rate": 4.75776397515528e-05,
+      "loss": 0.0235,
+      "step": 1243
+    },
+    {
+      "epoch": 11.518518518518519,
+      "grad_norm": 1.227002739906311,
+      "learning_rate": 4.7453416149068325e-05,
+      "loss": 0.0506,
+      "step": 1244
+    },
+    {
+      "epoch": 11.527777777777779,
+      "grad_norm": 1.0969829559326172,
+      "learning_rate": 4.7329192546583855e-05,
+      "loss": 0.0435,
+      "step": 1245
+    },
+    {
+      "epoch": 11.537037037037036,
+      "grad_norm": 1.4205132722854614,
+      "learning_rate": 4.7204968944099384e-05,
+      "loss": 0.0762,
+      "step": 1246
+    },
+    {
+      "epoch": 11.546296296296296,
+      "grad_norm": 1.1668907403945923,
+      "learning_rate": 4.708074534161491e-05,
+      "loss": 0.0478,
+      "step": 1247
+    },
+    {
+      "epoch": 11.555555555555555,
+      "grad_norm": 0.9908825159072876,
+      "learning_rate": 4.695652173913044e-05,
+      "loss": 0.0437,
+      "step": 1248
+    },
+    {
+      "epoch": 11.564814814814815,
+      "grad_norm": 1.2143094539642334,
+      "learning_rate": 4.683229813664596e-05,
+      "loss": 0.0452,
+      "step": 1249
+    },
+    {
+      "epoch": 11.574074074074074,
+      "grad_norm": 0.7645125389099121,
+      "learning_rate": 4.67080745341615e-05,
+      "loss": 0.0297,
+      "step": 1250
+    },
+    {
+      "epoch": 11.583333333333334,
+      "grad_norm": 0.8744638562202454,
+      "learning_rate": 4.658385093167702e-05,
+      "loss": 0.0342,
+      "step": 1251
+    },
+    {
+      "epoch": 11.592592592592592,
+      "grad_norm": 0.9421960115432739,
+      "learning_rate": 4.645962732919255e-05,
+      "loss": 0.0422,
+      "step": 1252
+    },
+    {
+      "epoch": 11.601851851851851,
+      "grad_norm": 0.9777348637580872,
+      "learning_rate": 4.633540372670807e-05,
+      "loss": 0.022,
+      "step": 1253
+    },
+    {
+      "epoch": 11.61111111111111,
+      "grad_norm": 1.009817361831665,
+      "learning_rate": 4.62111801242236e-05,
+      "loss": 0.0458,
+      "step": 1254
+    },
+    {
+      "epoch": 11.62037037037037,
+      "grad_norm": 1.0724588632583618,
+      "learning_rate": 4.608695652173913e-05,
+      "loss": 0.0441,
+      "step": 1255
+    },
+    {
+      "epoch": 11.62962962962963,
+      "grad_norm": 0.9254492521286011,
+      "learning_rate": 4.5962732919254656e-05,
+      "loss": 0.037,
+      "step": 1256
+    },
+    {
+      "epoch": 11.63888888888889,
+      "grad_norm": 1.018061637878418,
+      "learning_rate": 4.5838509316770186e-05,
+      "loss": 0.0378,
+      "step": 1257
+    },
+    {
+      "epoch": 11.648148148148149,
+      "grad_norm": 0.8017347455024719,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.0369,
+      "step": 1258
+    },
+    {
+      "epoch": 11.657407407407408,
+      "grad_norm": 1.1639477014541626,
+      "learning_rate": 4.5590062111801246e-05,
+      "loss": 0.0428,
+      "step": 1259
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 1.008543610572815,
+      "learning_rate": 4.546583850931677e-05,
+      "loss": 0.0424,
+      "step": 1260
+    },
+    {
+      "epoch": 11.675925925925926,
+      "grad_norm": 1.6025316715240479,
+      "learning_rate": 4.5341614906832306e-05,
+      "loss": 0.0614,
+      "step": 1261
+    },
+    {
+      "epoch": 11.685185185185185,
+      "grad_norm": 0.9679161310195923,
+      "learning_rate": 4.521739130434783e-05,
+      "loss": 0.0328,
+      "step": 1262
+    },
+    {
+      "epoch": 11.694444444444445,
+      "grad_norm": 1.2503728866577148,
+      "learning_rate": 4.509316770186335e-05,
+      "loss": 0.0471,
+      "step": 1263
+    },
+    {
+      "epoch": 11.703703703703704,
+      "grad_norm": 1.0044113397598267,
+      "learning_rate": 4.496894409937888e-05,
+      "loss": 0.0361,
+      "step": 1264
+    },
+    {
+      "epoch": 11.712962962962964,
+      "grad_norm": 0.9651626348495483,
+      "learning_rate": 4.484472049689441e-05,
+      "loss": 0.0413,
+      "step": 1265
+    },
+    {
+      "epoch": 11.722222222222221,
+      "grad_norm": 2.5001490116119385,
+      "learning_rate": 4.472049689440994e-05,
+      "loss": 0.0573,
+      "step": 1266
+    },
+    {
+      "epoch": 11.731481481481481,
+      "grad_norm": 0.8396499156951904,
+      "learning_rate": 4.4596273291925465e-05,
+      "loss": 0.033,
+      "step": 1267
+    },
+    {
+      "epoch": 11.74074074074074,
+      "grad_norm": 0.7739114165306091,
+      "learning_rate": 4.4472049689440995e-05,
+      "loss": 0.0297,
+      "step": 1268
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 0.7845550775527954,
+      "learning_rate": 4.4347826086956525e-05,
+      "loss": 0.0384,
+      "step": 1269
+    },
+    {
+      "epoch": 11.75925925925926,
+      "grad_norm": 0.9752318859100342,
+      "learning_rate": 4.4223602484472055e-05,
+      "loss": 0.0359,
+      "step": 1270
+    },
+    {
+      "epoch": 11.768518518518519,
+      "grad_norm": 0.8686448335647583,
+      "learning_rate": 4.409937888198758e-05,
+      "loss": 0.0333,
+      "step": 1271
+    },
+    {
+      "epoch": 11.777777777777779,
+      "grad_norm": 1.0658447742462158,
+      "learning_rate": 4.397515527950311e-05,
+      "loss": 0.0446,
+      "step": 1272
+    },
+    {
+      "epoch": 11.787037037037036,
+      "grad_norm": 2.433940887451172,
+      "learning_rate": 4.385093167701864e-05,
+      "loss": 0.0614,
+      "step": 1273
+    },
+    {
+      "epoch": 11.796296296296296,
+      "grad_norm": 1.705237865447998,
+      "learning_rate": 4.372670807453416e-05,
+      "loss": 0.0731,
+      "step": 1274
+    },
+    {
+      "epoch": 11.805555555555555,
+      "grad_norm": 0.6588522791862488,
+      "learning_rate": 4.360248447204969e-05,
+      "loss": 0.0384,
+      "step": 1275
+    },
+    {
+      "epoch": 11.814814814814815,
+      "grad_norm": 1.0283806324005127,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.0339,
+      "step": 1276
+    },
+    {
+      "epoch": 11.824074074074074,
+      "grad_norm": 1.315185308456421,
+      "learning_rate": 4.335403726708075e-05,
+      "loss": 0.0578,
+      "step": 1277
+    },
+    {
+      "epoch": 11.833333333333334,
+      "grad_norm": 0.930867612361908,
+      "learning_rate": 4.322981366459627e-05,
+      "loss": 0.0354,
+      "step": 1278
+    },
+    {
+      "epoch": 11.842592592592592,
+      "grad_norm": 1.5080018043518066,
+      "learning_rate": 4.31055900621118e-05,
+      "loss": 0.0446,
+      "step": 1279
+    },
+    {
+      "epoch": 11.851851851851851,
+      "grad_norm": 0.7906777262687683,
+      "learning_rate": 4.298136645962733e-05,
+      "loss": 0.0353,
+      "step": 1280
+    },
+    {
+      "epoch": 11.86111111111111,
+      "grad_norm": 1.961120843887329,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.0355,
+      "step": 1281
+    },
+    {
+      "epoch": 11.87037037037037,
+      "grad_norm": 1.2218685150146484,
+      "learning_rate": 4.2732919254658386e-05,
+      "loss": 0.048,
+      "step": 1282
+    },
+    {
+      "epoch": 11.87962962962963,
+      "grad_norm": 1.000490427017212,
+      "learning_rate": 4.2608695652173916e-05,
+      "loss": 0.0402,
+      "step": 1283
+    },
+    {
+      "epoch": 11.88888888888889,
+      "grad_norm": 0.762229859828949,
+      "learning_rate": 4.2484472049689446e-05,
+      "loss": 0.0322,
+      "step": 1284
+    },
+    {
+      "epoch": 11.898148148148149,
+      "grad_norm": 0.7762311100959778,
+      "learning_rate": 4.236024844720497e-05,
+      "loss": 0.0324,
+      "step": 1285
+    },
+    {
+      "epoch": 11.907407407407408,
+      "grad_norm": 0.8216007947921753,
+      "learning_rate": 4.22360248447205e-05,
+      "loss": 0.0281,
+      "step": 1286
+    },
+    {
+      "epoch": 11.916666666666666,
+      "grad_norm": 0.8371996879577637,
+      "learning_rate": 4.211180124223603e-05,
+      "loss": 0.0316,
+      "step": 1287
+    },
+    {
+      "epoch": 11.925925925925926,
+      "grad_norm": 1.3675734996795654,
+      "learning_rate": 4.198757763975156e-05,
+      "loss": 0.063,
+      "step": 1288
+    },
+    {
+      "epoch": 11.935185185185185,
+      "grad_norm": 0.5560950636863708,
+      "learning_rate": 4.186335403726708e-05,
+      "loss": 0.0242,
+      "step": 1289
+    },
+    {
+      "epoch": 11.944444444444445,
+      "grad_norm": 0.775168240070343,
+      "learning_rate": 4.1739130434782605e-05,
+      "loss": 0.0339,
+      "step": 1290
+    },
+    {
+      "epoch": 11.953703703703704,
+      "grad_norm": 1.0372427701950073,
+      "learning_rate": 4.161490683229814e-05,
+      "loss": 0.0398,
+      "step": 1291
+    },
+    {
+      "epoch": 11.962962962962964,
+      "grad_norm": 2.2468655109405518,
+      "learning_rate": 4.1490683229813665e-05,
+      "loss": 0.0525,
+      "step": 1292
+    },
+    {
+      "epoch": 11.972222222222221,
+      "grad_norm": 0.867691695690155,
+      "learning_rate": 4.1366459627329195e-05,
+      "loss": 0.0443,
+      "step": 1293
+    },
+    {
+      "epoch": 11.981481481481481,
+      "grad_norm": 0.6092666387557983,
+      "learning_rate": 4.1242236024844725e-05,
+      "loss": 0.0258,
+      "step": 1294
+    },
+    {
+      "epoch": 11.99074074074074,
+      "grad_norm": 0.9719471335411072,
+      "learning_rate": 4.1118012422360255e-05,
+      "loss": 0.0413,
+      "step": 1295
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.8400522470474243,
+      "learning_rate": 4.099378881987578e-05,
+      "loss": 0.0342,
+      "step": 1296
+    },
+    {
+      "epoch": 12.00925925925926,
+      "grad_norm": 0.9376289248466492,
+      "learning_rate": 4.086956521739131e-05,
+      "loss": 0.0373,
+      "step": 1297
+    },
+    {
+      "epoch": 12.018518518518519,
+      "grad_norm": 1.0187164545059204,
+      "learning_rate": 4.074534161490684e-05,
+      "loss": 0.0377,
+      "step": 1298
+    },
+    {
+      "epoch": 12.027777777777779,
+      "grad_norm": 1.1379069089889526,
+      "learning_rate": 4.062111801242236e-05,
+      "loss": 0.0572,
+      "step": 1299
+    },
+    {
+      "epoch": 12.037037037037036,
+      "grad_norm": 0.9860619902610779,
+      "learning_rate": 4.049689440993789e-05,
+      "loss": 0.0426,
+      "step": 1300
+    },
+    {
+      "epoch": 12.046296296296296,
+      "grad_norm": 0.9077693223953247,
+      "learning_rate": 4.0372670807453414e-05,
+      "loss": 0.036,
+      "step": 1301
+    },
+    {
+      "epoch": 12.055555555555555,
+      "grad_norm": 0.5804708003997803,
+      "learning_rate": 4.024844720496895e-05,
+      "loss": 0.0247,
+      "step": 1302
+    },
+    {
+      "epoch": 12.064814814814815,
+      "grad_norm": 0.7517374753952026,
+      "learning_rate": 4.012422360248447e-05,
+      "loss": 0.0259,
+      "step": 1303
+    },
+    {
+      "epoch": 12.074074074074074,
+      "grad_norm": 0.9251551032066345,
+      "learning_rate": 4e-05,
+      "loss": 0.0379,
+      "step": 1304
+    },
+    {
+      "epoch": 12.083333333333334,
+      "grad_norm": 0.7882636189460754,
+      "learning_rate": 3.9875776397515526e-05,
+      "loss": 0.0287,
+      "step": 1305
+    },
+    {
+      "epoch": 12.092592592592593,
+      "grad_norm": 0.6092426180839539,
+      "learning_rate": 3.9751552795031056e-05,
+      "loss": 0.0211,
+      "step": 1306
+    },
+    {
+      "epoch": 12.101851851851851,
+      "grad_norm": 0.8440184593200684,
+      "learning_rate": 3.9627329192546586e-05,
+      "loss": 0.0282,
+      "step": 1307
+    },
+    {
+      "epoch": 12.11111111111111,
+      "grad_norm": 0.9278969168663025,
+      "learning_rate": 3.950310559006211e-05,
+      "loss": 0.0304,
+      "step": 1308
+    },
+    {
+      "epoch": 12.12037037037037,
+      "grad_norm": 0.6380085945129395,
+      "learning_rate": 3.9378881987577646e-05,
+      "loss": 0.0223,
+      "step": 1309
+    },
+    {
+      "epoch": 12.12962962962963,
+      "grad_norm": 0.7007056474685669,
+      "learning_rate": 3.925465838509317e-05,
+      "loss": 0.0241,
+      "step": 1310
+    },
+    {
+      "epoch": 12.13888888888889,
+      "grad_norm": 0.7670966982841492,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.0283,
+      "step": 1311
+    },
+    {
+      "epoch": 12.148148148148149,
+      "grad_norm": 0.8422584533691406,
+      "learning_rate": 3.900621118012422e-05,
+      "loss": 0.0291,
+      "step": 1312
+    },
+    {
+      "epoch": 12.157407407407407,
+      "grad_norm": 0.6846734881401062,
+      "learning_rate": 3.888198757763976e-05,
+      "loss": 0.0232,
+      "step": 1313
+    },
+    {
+      "epoch": 12.166666666666666,
+      "grad_norm": 0.900596559047699,
+      "learning_rate": 3.875776397515528e-05,
+      "loss": 0.0295,
+      "step": 1314
+    },
+    {
+      "epoch": 12.175925925925926,
+      "grad_norm": 0.6619224548339844,
+      "learning_rate": 3.8633540372670805e-05,
+      "loss": 0.0243,
+      "step": 1315
+    },
+    {
+      "epoch": 12.185185185185185,
+      "grad_norm": 0.431782603263855,
+      "learning_rate": 3.8509316770186335e-05,
+      "loss": 0.0174,
+      "step": 1316
+    },
+    {
+      "epoch": 12.194444444444445,
+      "grad_norm": 1.0050886869430542,
+      "learning_rate": 3.8385093167701865e-05,
+      "loss": 0.0391,
+      "step": 1317
+    },
+    {
+      "epoch": 12.203703703703704,
+      "grad_norm": 0.8259010910987854,
+      "learning_rate": 3.8260869565217395e-05,
+      "loss": 0.0259,
+      "step": 1318
+    },
+    {
+      "epoch": 12.212962962962964,
+      "grad_norm": 1.4165301322937012,
+      "learning_rate": 3.813664596273292e-05,
+      "loss": 0.0446,
+      "step": 1319
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 0.7267447113990784,
+      "learning_rate": 3.801242236024845e-05,
+      "loss": 0.022,
+      "step": 1320
+    },
+    {
+      "epoch": 12.231481481481481,
+      "grad_norm": 0.8817240595817566,
+      "learning_rate": 3.788819875776398e-05,
+      "loss": 0.0349,
+      "step": 1321
+    },
+    {
+      "epoch": 12.24074074074074,
+      "grad_norm": 0.815268337726593,
+      "learning_rate": 3.776397515527951e-05,
+      "loss": 0.027,
+      "step": 1322
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 0.6415846347808838,
+      "learning_rate": 3.763975155279503e-05,
+      "loss": 0.0212,
+      "step": 1323
+    },
+    {
+      "epoch": 12.25925925925926,
+      "grad_norm": 0.7012866139411926,
+      "learning_rate": 3.751552795031056e-05,
+      "loss": 0.0219,
+      "step": 1324
+    },
+    {
+      "epoch": 12.268518518518519,
+      "grad_norm": 0.9753007888793945,
+      "learning_rate": 3.739130434782609e-05,
+      "loss": 0.0339,
+      "step": 1325
+    },
+    {
+      "epoch": 12.277777777777779,
+      "grad_norm": 0.8093515038490295,
+      "learning_rate": 3.7267080745341614e-05,
+      "loss": 0.0277,
+      "step": 1326
+    },
+    {
+      "epoch": 12.287037037037036,
+      "grad_norm": 0.8891710638999939,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.0413,
+      "step": 1327
+    },
+    {
+      "epoch": 12.296296296296296,
+      "grad_norm": 0.5199636816978455,
+      "learning_rate": 3.701863354037267e-05,
+      "loss": 0.0169,
+      "step": 1328
+    },
+    {
+      "epoch": 12.305555555555555,
+      "grad_norm": 0.8058778643608093,
+      "learning_rate": 3.68944099378882e-05,
+      "loss": 0.0324,
+      "step": 1329
+    },
+    {
+      "epoch": 12.314814814814815,
+      "grad_norm": 0.7521347403526306,
+      "learning_rate": 3.6770186335403726e-05,
+      "loss": 0.0225,
+      "step": 1330
+    },
+    {
+      "epoch": 12.324074074074074,
+      "grad_norm": 0.9405580163002014,
+      "learning_rate": 3.6645962732919256e-05,
+      "loss": 0.0321,
+      "step": 1331
+    },
+    {
+      "epoch": 12.333333333333334,
+      "grad_norm": 0.6070406436920166,
+      "learning_rate": 3.6521739130434786e-05,
+      "loss": 0.0192,
+      "step": 1332
+    },
+    {
+      "epoch": 12.342592592592593,
+      "grad_norm": 0.9515055418014526,
+      "learning_rate": 3.639751552795031e-05,
+      "loss": 0.0273,
+      "step": 1333
+    },
+    {
+      "epoch": 12.351851851851851,
+      "grad_norm": 0.5170685052871704,
+      "learning_rate": 3.627329192546584e-05,
+      "loss": 0.0264,
+      "step": 1334
+    },
+    {
+      "epoch": 12.36111111111111,
+      "grad_norm": 1.2589349746704102,
+      "learning_rate": 3.614906832298137e-05,
+      "loss": 0.0593,
+      "step": 1335
+    },
+    {
+      "epoch": 12.37037037037037,
+      "grad_norm": 1.3736205101013184,
+      "learning_rate": 3.60248447204969e-05,
+      "loss": 0.038,
+      "step": 1336
+    },
+    {
+      "epoch": 12.37962962962963,
+      "grad_norm": 0.7936609387397766,
+      "learning_rate": 3.590062111801242e-05,
+      "loss": 0.029,
+      "step": 1337
+    },
+    {
+      "epoch": 12.38888888888889,
+      "grad_norm": 0.908733606338501,
+      "learning_rate": 3.577639751552795e-05,
+      "loss": 0.0343,
+      "step": 1338
+    },
+    {
+      "epoch": 12.398148148148149,
+      "grad_norm": 0.8484895825386047,
+      "learning_rate": 3.565217391304348e-05,
+      "loss": 0.0275,
+      "step": 1339
+    },
+    {
+      "epoch": 12.407407407407407,
+      "grad_norm": 1.0681447982788086,
+      "learning_rate": 3.552795031055901e-05,
+      "loss": 0.0328,
+      "step": 1340
+    },
+    {
+      "epoch": 12.416666666666666,
+      "grad_norm": 1.075875997543335,
+      "learning_rate": 3.5403726708074535e-05,
+      "loss": 0.0418,
+      "step": 1341
+    },
+    {
+      "epoch": 12.425925925925926,
+      "grad_norm": 0.6161890029907227,
+      "learning_rate": 3.5279503105590065e-05,
+      "loss": 0.0296,
+      "step": 1342
+    },
+    {
+      "epoch": 12.435185185185185,
+      "grad_norm": 1.1916195154190063,
+      "learning_rate": 3.5155279503105595e-05,
+      "loss": 0.0486,
+      "step": 1343
+    },
+    {
+      "epoch": 12.444444444444445,
+      "grad_norm": 0.6621356010437012,
+      "learning_rate": 3.503105590062112e-05,
+      "loss": 0.0221,
+      "step": 1344
+    },
+    {
+      "epoch": 12.453703703703704,
+      "grad_norm": 1.6666837930679321,
+      "learning_rate": 3.490683229813665e-05,
+      "loss": 0.0459,
+      "step": 1345
+    },
+    {
+      "epoch": 12.462962962962964,
+      "grad_norm": 0.7422270178794861,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.0249,
+      "step": 1346
+    },
+    {
+      "epoch": 12.472222222222221,
+      "grad_norm": 0.7906531095504761,
+      "learning_rate": 3.465838509316771e-05,
+      "loss": 0.038,
+      "step": 1347
+    },
+    {
+      "epoch": 12.481481481481481,
+      "grad_norm": 0.9974073767662048,
+      "learning_rate": 3.453416149068323e-05,
+      "loss": 0.0406,
+      "step": 1348
+    },
+    {
+      "epoch": 12.49074074074074,
+      "grad_norm": 0.6397982835769653,
+      "learning_rate": 3.440993788819876e-05,
+      "loss": 0.0222,
+      "step": 1349
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 0.682849645614624,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.0298,
+      "step": 1350
+    },
+    {
+      "epoch": 12.50925925925926,
+      "grad_norm": 1.4077941179275513,
+      "learning_rate": 3.4161490683229814e-05,
+      "loss": 0.0511,
+      "step": 1351
+    },
+    {
+      "epoch": 12.518518518518519,
+      "grad_norm": 0.7027615308761597,
+      "learning_rate": 3.4037267080745343e-05,
+      "loss": 0.0203,
+      "step": 1352
+    },
+    {
+      "epoch": 12.527777777777779,
+      "grad_norm": 1.0765162706375122,
+      "learning_rate": 3.3913043478260867e-05,
+      "loss": 0.0379,
+      "step": 1353
+    },
+    {
+      "epoch": 12.537037037037036,
+      "grad_norm": 0.7544326782226562,
+      "learning_rate": 3.37888198757764e-05,
+      "loss": 0.0278,
+      "step": 1354
+    },
+    {
+      "epoch": 12.546296296296296,
+      "grad_norm": 1.0442858934402466,
+      "learning_rate": 3.3664596273291926e-05,
+      "loss": 0.0441,
+      "step": 1355
+    },
+    {
+      "epoch": 12.555555555555555,
+      "grad_norm": 1.0142107009887695,
+      "learning_rate": 3.3540372670807456e-05,
+      "loss": 0.0446,
+      "step": 1356
+    },
+    {
+      "epoch": 12.564814814814815,
+      "grad_norm": 1.299988865852356,
+      "learning_rate": 3.341614906832298e-05,
+      "loss": 0.0447,
+      "step": 1357
+    },
+    {
+      "epoch": 12.574074074074074,
+      "grad_norm": 0.6750516295433044,
+      "learning_rate": 3.329192546583851e-05,
+      "loss": 0.03,
+      "step": 1358
+    },
+    {
+      "epoch": 12.583333333333334,
+      "grad_norm": 0.8981741070747375,
+      "learning_rate": 3.316770186335404e-05,
+      "loss": 0.0315,
+      "step": 1359
+    },
+    {
+      "epoch": 12.592592592592592,
+      "grad_norm": 1.8234065771102905,
+      "learning_rate": 3.304347826086956e-05,
+      "loss": 0.0857,
+      "step": 1360
+    },
+    {
+      "epoch": 12.601851851851851,
+      "grad_norm": 1.1053465604782104,
+      "learning_rate": 3.29192546583851e-05,
+      "loss": 0.0436,
+      "step": 1361
+    },
+    {
+      "epoch": 12.61111111111111,
+      "grad_norm": 1.0005968809127808,
+      "learning_rate": 3.279503105590062e-05,
+      "loss": 0.042,
+      "step": 1362
+    },
+    {
+      "epoch": 12.62037037037037,
+      "grad_norm": 0.952779233455658,
+      "learning_rate": 3.267080745341615e-05,
+      "loss": 0.0314,
+      "step": 1363
+    },
+    {
+      "epoch": 12.62962962962963,
+      "grad_norm": 2.057094097137451,
+      "learning_rate": 3.2546583850931675e-05,
+      "loss": 0.0484,
+      "step": 1364
+    },
+    {
+      "epoch": 12.63888888888889,
+      "grad_norm": 0.7208571434020996,
+      "learning_rate": 3.242236024844721e-05,
+      "loss": 0.0309,
+      "step": 1365
+    },
+    {
+      "epoch": 12.648148148148149,
+      "grad_norm": 0.8814190626144409,
+      "learning_rate": 3.2298136645962735e-05,
+      "loss": 0.0373,
+      "step": 1366
+    },
+    {
+      "epoch": 12.657407407407408,
+      "grad_norm": 0.7553854584693909,
+      "learning_rate": 3.217391304347826e-05,
+      "loss": 0.0259,
+      "step": 1367
+    },
+    {
+      "epoch": 12.666666666666666,
+      "grad_norm": 0.8120577335357666,
+      "learning_rate": 3.204968944099379e-05,
+      "loss": 0.0332,
+      "step": 1368
+    },
+    {
+      "epoch": 12.675925925925926,
+      "grad_norm": 0.8838486075401306,
+      "learning_rate": 3.192546583850932e-05,
+      "loss": 0.0324,
+      "step": 1369
+    },
+    {
+      "epoch": 12.685185185185185,
+      "grad_norm": 0.7664040923118591,
+      "learning_rate": 3.180124223602485e-05,
+      "loss": 0.0312,
+      "step": 1370
+    },
+    {
+      "epoch": 12.694444444444445,
+      "grad_norm": 0.7717130780220032,
+      "learning_rate": 3.167701863354037e-05,
+      "loss": 0.0352,
+      "step": 1371
+    },
+    {
+      "epoch": 12.703703703703704,
+      "grad_norm": 3.366727590560913,
+      "learning_rate": 3.15527950310559e-05,
+      "loss": 0.1035,
+      "step": 1372
+    },
+    {
+      "epoch": 12.712962962962964,
+      "grad_norm": 0.723676323890686,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.0251,
+      "step": 1373
+    },
+    {
+      "epoch": 12.722222222222221,
+      "grad_norm": 0.7675268650054932,
+      "learning_rate": 3.130434782608696e-05,
+      "loss": 0.0313,
+      "step": 1374
+    },
+    {
+      "epoch": 12.731481481481481,
+      "grad_norm": 0.6413992643356323,
+      "learning_rate": 3.1180124223602484e-05,
+      "loss": 0.0287,
+      "step": 1375
+    },
+    {
+      "epoch": 12.74074074074074,
+      "grad_norm": 0.8057203888893127,
+      "learning_rate": 3.1055900621118014e-05,
+      "loss": 0.0252,
+      "step": 1376
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 0.9161250591278076,
+      "learning_rate": 3.0931677018633543e-05,
+      "loss": 0.0329,
+      "step": 1377
+    },
+    {
+      "epoch": 12.75925925925926,
+      "grad_norm": 0.7578794956207275,
+      "learning_rate": 3.0807453416149067e-05,
+      "loss": 0.031,
+      "step": 1378
+    },
+    {
+      "epoch": 12.768518518518519,
+      "grad_norm": 0.7553548216819763,
+      "learning_rate": 3.0683229813664596e-05,
+      "loss": 0.0279,
+      "step": 1379
+    },
+    {
+      "epoch": 12.777777777777779,
+      "grad_norm": 0.9717443585395813,
+      "learning_rate": 3.0559006211180126e-05,
+      "loss": 0.0333,
+      "step": 1380
+    },
+    {
+      "epoch": 12.787037037037036,
+      "grad_norm": 0.7987967133522034,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.0317,
+      "step": 1381
+    },
+    {
+      "epoch": 12.796296296296296,
+      "grad_norm": 1.3905160427093506,
+      "learning_rate": 3.031055900621118e-05,
+      "loss": 0.0558,
+      "step": 1382
+    },
+    {
+      "epoch": 12.805555555555555,
+      "grad_norm": 0.7816559076309204,
+      "learning_rate": 3.0186335403726713e-05,
+      "loss": 0.036,
+      "step": 1383
+    },
+    {
+      "epoch": 12.814814814814815,
+      "grad_norm": 0.6633138060569763,
+      "learning_rate": 3.006211180124224e-05,
+      "loss": 0.0287,
+      "step": 1384
+    },
+    {
+      "epoch": 12.824074074074074,
+      "grad_norm": 0.6967059373855591,
+      "learning_rate": 2.9937888198757762e-05,
+      "loss": 0.0237,
+      "step": 1385
+    },
+    {
+      "epoch": 12.833333333333334,
+      "grad_norm": 0.9407333135604858,
+      "learning_rate": 2.9813664596273296e-05,
+      "loss": 0.0354,
+      "step": 1386
+    },
+    {
+      "epoch": 12.842592592592592,
+      "grad_norm": 1.1632462739944458,
+      "learning_rate": 2.968944099378882e-05,
+      "loss": 0.0468,
+      "step": 1387
+    },
+    {
+      "epoch": 12.851851851851851,
+      "grad_norm": 0.5927537679672241,
+      "learning_rate": 2.9565217391304352e-05,
+      "loss": 0.0226,
+      "step": 1388
+    },
+    {
+      "epoch": 12.86111111111111,
+      "grad_norm": 0.7235986590385437,
+      "learning_rate": 2.9440993788819875e-05,
+      "loss": 0.0229,
+      "step": 1389
+    },
+    {
+      "epoch": 12.87037037037037,
+      "grad_norm": 0.7975889444351196,
+      "learning_rate": 2.931677018633541e-05,
+      "loss": 0.0293,
+      "step": 1390
+    },
+    {
+      "epoch": 12.87962962962963,
+      "grad_norm": 0.8041378259658813,
+      "learning_rate": 2.919254658385093e-05,
+      "loss": 0.0276,
+      "step": 1391
+    },
+    {
+      "epoch": 12.88888888888889,
+      "grad_norm": 0.5750847458839417,
+      "learning_rate": 2.9068322981366465e-05,
+      "loss": 0.024,
+      "step": 1392
+    },
+    {
+      "epoch": 12.898148148148149,
+      "grad_norm": 0.8676141500473022,
+      "learning_rate": 2.8944099378881988e-05,
+      "loss": 0.0302,
+      "step": 1393
+    },
+    {
+      "epoch": 12.907407407407408,
+      "grad_norm": 0.9506133794784546,
+      "learning_rate": 2.8819875776397514e-05,
+      "loss": 0.0317,
+      "step": 1394
+    },
+    {
+      "epoch": 12.916666666666666,
+      "grad_norm": 1.1184026002883911,
+      "learning_rate": 2.8695652173913044e-05,
+      "loss": 0.0448,
+      "step": 1395
+    },
+    {
+      "epoch": 12.925925925925926,
+      "grad_norm": 1.0374799966812134,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.0428,
+      "step": 1396
+    },
+    {
+      "epoch": 12.935185185185185,
+      "grad_norm": 1.3991471529006958,
+      "learning_rate": 2.84472049689441e-05,
+      "loss": 0.0761,
+      "step": 1397
+    },
+    {
+      "epoch": 12.944444444444445,
+      "grad_norm": 1.7948813438415527,
+      "learning_rate": 2.8322981366459627e-05,
+      "loss": 0.0657,
+      "step": 1398
+    },
+    {
+      "epoch": 12.953703703703704,
+      "grad_norm": 0.7981337308883667,
+      "learning_rate": 2.8198757763975157e-05,
+      "loss": 0.0256,
+      "step": 1399
+    },
+    {
+      "epoch": 12.962962962962964,
+      "grad_norm": 1.0698860883712769,
+      "learning_rate": 2.8074534161490684e-05,
+      "loss": 0.0428,
+      "step": 1400
+    },
+    {
+      "epoch": 12.972222222222221,
+      "grad_norm": 1.00368332862854,
+      "learning_rate": 2.795031055900621e-05,
+      "loss": 0.0353,
+      "step": 1401
+    },
+    {
+      "epoch": 12.981481481481481,
+      "grad_norm": 0.9316563606262207,
+      "learning_rate": 2.782608695652174e-05,
+      "loss": 0.0312,
+      "step": 1402
+    },
+    {
+      "epoch": 12.99074074074074,
+      "grad_norm": 0.9837682843208313,
+      "learning_rate": 2.7701863354037267e-05,
+      "loss": 0.039,
+      "step": 1403
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.0201270580291748,
+      "learning_rate": 2.7577639751552796e-05,
+      "loss": 0.0391,
+      "step": 1404
+    },
+    {
+      "epoch": 13.00925925925926,
+      "grad_norm": 0.7964289784431458,
+      "learning_rate": 2.7453416149068323e-05,
+      "loss": 0.0294,
+      "step": 1405
+    },
+    {
+      "epoch": 13.018518518518519,
+      "grad_norm": 0.5249894261360168,
+      "learning_rate": 2.7329192546583853e-05,
+      "loss": 0.0237,
+      "step": 1406
+    },
+    {
+      "epoch": 13.027777777777779,
+      "grad_norm": 0.9201458096504211,
+      "learning_rate": 2.720496894409938e-05,
+      "loss": 0.0257,
+      "step": 1407
+    },
+    {
+      "epoch": 13.037037037037036,
+      "grad_norm": 0.6795076131820679,
+      "learning_rate": 2.708074534161491e-05,
+      "loss": 0.0207,
+      "step": 1408
+    },
+    {
+      "epoch": 13.046296296296296,
+      "grad_norm": 1.0227335691452026,
+      "learning_rate": 2.6956521739130436e-05,
+      "loss": 0.044,
+      "step": 1409
+    },
+    {
+      "epoch": 13.055555555555555,
+      "grad_norm": 0.9504062533378601,
+      "learning_rate": 2.6832298136645962e-05,
+      "loss": 0.029,
+      "step": 1410
+    },
+    {
+      "epoch": 13.064814814814815,
+      "grad_norm": 0.8904616832733154,
+      "learning_rate": 2.6708074534161492e-05,
+      "loss": 0.0286,
+      "step": 1411
+    },
+    {
+      "epoch": 13.074074074074074,
+      "grad_norm": 1.118471384048462,
+      "learning_rate": 2.658385093167702e-05,
+      "loss": 0.0352,
+      "step": 1412
+    },
+    {
+      "epoch": 13.083333333333334,
+      "grad_norm": 0.6861169934272766,
+      "learning_rate": 2.645962732919255e-05,
+      "loss": 0.0218,
+      "step": 1413
+    },
+    {
+      "epoch": 13.092592592592593,
+      "grad_norm": 0.802804172039032,
+      "learning_rate": 2.6335403726708075e-05,
+      "loss": 0.0302,
+      "step": 1414
+    },
+    {
+      "epoch": 13.101851851851851,
+      "grad_norm": 0.7278661727905273,
+      "learning_rate": 2.6211180124223605e-05,
+      "loss": 0.0251,
+      "step": 1415
+    },
+    {
+      "epoch": 13.11111111111111,
+      "grad_norm": 0.9800047874450684,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.0274,
+      "step": 1416
+    },
+    {
+      "epoch": 13.12037037037037,
+      "grad_norm": 0.8877214193344116,
+      "learning_rate": 2.596273291925466e-05,
+      "loss": 0.0292,
+      "step": 1417
+    },
+    {
+      "epoch": 13.12962962962963,
+      "grad_norm": 0.8059203624725342,
+      "learning_rate": 2.5838509316770188e-05,
+      "loss": 0.0197,
+      "step": 1418
+    },
+    {
+      "epoch": 13.13888888888889,
+      "grad_norm": 0.5480836033821106,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.0181,
+      "step": 1419
+    },
+    {
+      "epoch": 13.148148148148149,
+      "grad_norm": 0.649368166923523,
+      "learning_rate": 2.5590062111801244e-05,
+      "loss": 0.0182,
+      "step": 1420
+    },
+    {
+      "epoch": 13.157407407407407,
+      "grad_norm": 0.8103269934654236,
+      "learning_rate": 2.546583850931677e-05,
+      "loss": 0.0229,
+      "step": 1421
+    },
+    {
+      "epoch": 13.166666666666666,
+      "grad_norm": 0.8214313983917236,
+      "learning_rate": 2.53416149068323e-05,
+      "loss": 0.0293,
+      "step": 1422
+    },
+    {
+      "epoch": 13.175925925925926,
+      "grad_norm": 0.7193138003349304,
+      "learning_rate": 2.5217391304347827e-05,
+      "loss": 0.0253,
+      "step": 1423
+    },
+    {
+      "epoch": 13.185185185185185,
+      "grad_norm": 0.7240740656852722,
+      "learning_rate": 2.5093167701863357e-05,
+      "loss": 0.0251,
+      "step": 1424
+    },
+    {
+      "epoch": 13.194444444444445,
+      "grad_norm": 0.720065176486969,
+      "learning_rate": 2.4968944099378884e-05,
+      "loss": 0.0207,
+      "step": 1425
+    },
+    {
+      "epoch": 13.203703703703704,
+      "grad_norm": 0.49787089228630066,
+      "learning_rate": 2.484472049689441e-05,
+      "loss": 0.0184,
+      "step": 1426
+    },
+    {
+      "epoch": 13.212962962962964,
+      "grad_norm": 0.836642861366272,
+      "learning_rate": 2.472049689440994e-05,
+      "loss": 0.0318,
+      "step": 1427
+    },
+    {
+      "epoch": 13.222222222222221,
+      "grad_norm": 0.7410364151000977,
+      "learning_rate": 2.4596273291925467e-05,
+      "loss": 0.0254,
+      "step": 1428
+    },
+    {
+      "epoch": 13.231481481481481,
+      "grad_norm": 0.7820467352867126,
+      "learning_rate": 2.4472049689440996e-05,
+      "loss": 0.0288,
+      "step": 1429
+    },
+    {
+      "epoch": 13.24074074074074,
+      "grad_norm": 0.8425258994102478,
+      "learning_rate": 2.4347826086956523e-05,
+      "loss": 0.0349,
+      "step": 1430
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 0.891117513179779,
+      "learning_rate": 2.4223602484472053e-05,
+      "loss": 0.0277,
+      "step": 1431
+    },
+    {
+      "epoch": 13.25925925925926,
+      "grad_norm": 0.9912608861923218,
+      "learning_rate": 2.4099378881987576e-05,
+      "loss": 0.0338,
+      "step": 1432
+    },
+    {
+      "epoch": 13.268518518518519,
+      "grad_norm": 0.8655397295951843,
+      "learning_rate": 2.3975155279503106e-05,
+      "loss": 0.0341,
+      "step": 1433
+    },
+    {
+      "epoch": 13.277777777777779,
+      "grad_norm": 0.8703702688217163,
+      "learning_rate": 2.3850931677018636e-05,
+      "loss": 0.0301,
+      "step": 1434
+    },
+    {
+      "epoch": 13.287037037037036,
+      "grad_norm": 0.816727340221405,
+      "learning_rate": 2.3726708074534162e-05,
+      "loss": 0.0268,
+      "step": 1435
+    },
+    {
+      "epoch": 13.296296296296296,
+      "grad_norm": 0.6376839280128479,
+      "learning_rate": 2.3602484472049692e-05,
+      "loss": 0.0251,
+      "step": 1436
+    },
+    {
+      "epoch": 13.305555555555555,
+      "grad_norm": 1.3493186235427856,
+      "learning_rate": 2.347826086956522e-05,
+      "loss": 0.046,
+      "step": 1437
+    },
+    {
+      "epoch": 13.314814814814815,
+      "grad_norm": 1.1794747114181519,
+      "learning_rate": 2.335403726708075e-05,
+      "loss": 0.0458,
+      "step": 1438
+    },
+    {
+      "epoch": 13.324074074074074,
+      "grad_norm": 0.8929194211959839,
+      "learning_rate": 2.3229813664596275e-05,
+      "loss": 0.0345,
+      "step": 1439
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 0.7308538556098938,
+      "learning_rate": 2.31055900621118e-05,
+      "loss": 0.0246,
+      "step": 1440
+    },
+    {
+      "epoch": 13.342592592592593,
+      "grad_norm": 0.701432466506958,
+      "learning_rate": 2.2981366459627328e-05,
+      "loss": 0.0233,
+      "step": 1441
+    },
+    {
+      "epoch": 13.351851851851851,
+      "grad_norm": 0.7073217630386353,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.0255,
+      "step": 1442
+    },
+    {
+      "epoch": 13.36111111111111,
+      "grad_norm": 0.8080083727836609,
+      "learning_rate": 2.2732919254658385e-05,
+      "loss": 0.0322,
+      "step": 1443
+    },
+    {
+      "epoch": 13.37037037037037,
+      "grad_norm": 0.8558483719825745,
+      "learning_rate": 2.2608695652173914e-05,
+      "loss": 0.0289,
+      "step": 1444
+    },
+    {
+      "epoch": 13.37962962962963,
+      "grad_norm": 0.7872486710548401,
+      "learning_rate": 2.248447204968944e-05,
+      "loss": 0.0253,
+      "step": 1445
+    },
+    {
+      "epoch": 13.38888888888889,
+      "grad_norm": 0.6285867094993591,
+      "learning_rate": 2.236024844720497e-05,
+      "loss": 0.0205,
+      "step": 1446
+    },
+    {
+      "epoch": 13.398148148148149,
+      "grad_norm": 0.9971258044242859,
+      "learning_rate": 2.2236024844720497e-05,
+      "loss": 0.0332,
+      "step": 1447
+    },
+    {
+      "epoch": 13.407407407407407,
+      "grad_norm": 1.0363770723342896,
+      "learning_rate": 2.2111801242236027e-05,
+      "loss": 0.0383,
+      "step": 1448
+    },
+    {
+      "epoch": 13.416666666666666,
+      "grad_norm": 1.03330397605896,
+      "learning_rate": 2.1987577639751554e-05,
+      "loss": 0.0288,
+      "step": 1449
+    },
+    {
+      "epoch": 13.425925925925926,
+      "grad_norm": 0.7237390279769897,
+      "learning_rate": 2.186335403726708e-05,
+      "loss": 0.0265,
+      "step": 1450
+    },
+    {
+      "epoch": 13.435185185185185,
+      "grad_norm": 0.8273149132728577,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.024,
+      "step": 1451
+    },
+    {
+      "epoch": 13.444444444444445,
+      "grad_norm": 0.7203170657157898,
+      "learning_rate": 2.1614906832298137e-05,
+      "loss": 0.0253,
+      "step": 1452
+    },
+    {
+      "epoch": 13.453703703703704,
+      "grad_norm": 0.970585823059082,
+      "learning_rate": 2.1490683229813667e-05,
+      "loss": 0.0339,
+      "step": 1453
+    },
+    {
+      "epoch": 13.462962962962964,
+      "grad_norm": 0.783500075340271,
+      "learning_rate": 2.1366459627329193e-05,
+      "loss": 0.024,
+      "step": 1454
+    },
+    {
+      "epoch": 13.472222222222221,
+      "grad_norm": 0.7558131814002991,
+      "learning_rate": 2.1242236024844723e-05,
+      "loss": 0.0225,
+      "step": 1455
+    },
+    {
+      "epoch": 13.481481481481481,
+      "grad_norm": 1.1136420965194702,
+      "learning_rate": 2.111801242236025e-05,
+      "loss": 0.0336,
+      "step": 1456
+    },
+    {
+      "epoch": 13.49074074074074,
+      "grad_norm": 0.7491280436515808,
+      "learning_rate": 2.099378881987578e-05,
+      "loss": 0.0282,
+      "step": 1457
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 0.8197646737098694,
+      "learning_rate": 2.0869565217391303e-05,
+      "loss": 0.0323,
+      "step": 1458
+    },
+    {
+      "epoch": 13.50925925925926,
+      "grad_norm": 1.0424060821533203,
+      "learning_rate": 2.0745341614906832e-05,
+      "loss": 0.0343,
+      "step": 1459
+    },
+    {
+      "epoch": 13.518518518518519,
+      "grad_norm": 0.8803132176399231,
+      "learning_rate": 2.0621118012422362e-05,
+      "loss": 0.0321,
+      "step": 1460
+    },
+    {
+      "epoch": 13.527777777777779,
+      "grad_norm": 0.8206008672714233,
+      "learning_rate": 2.049689440993789e-05,
+      "loss": 0.0266,
+      "step": 1461
+    },
+    {
+      "epoch": 13.537037037037036,
+      "grad_norm": 0.9388163685798645,
+      "learning_rate": 2.037267080745342e-05,
+      "loss": 0.034,
+      "step": 1462
+    },
+    {
+      "epoch": 13.546296296296296,
+      "grad_norm": 0.7545292973518372,
+      "learning_rate": 2.0248447204968945e-05,
+      "loss": 0.0308,
+      "step": 1463
+    },
+    {
+      "epoch": 13.555555555555555,
+      "grad_norm": 0.9023415446281433,
+      "learning_rate": 2.0124223602484475e-05,
+      "loss": 0.0329,
+      "step": 1464
+    },
+    {
+      "epoch": 13.564814814814815,
+      "grad_norm": 0.8173019886016846,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 1465
+    },
+    {
+      "epoch": 13.574074074074074,
+      "grad_norm": 1.1476924419403076,
+      "learning_rate": 1.9875776397515528e-05,
+      "loss": 0.0428,
+      "step": 1466
+    },
+    {
+      "epoch": 13.583333333333334,
+      "grad_norm": 1.10149085521698,
+      "learning_rate": 1.9751552795031055e-05,
+      "loss": 0.0377,
+      "step": 1467
+    },
+    {
+      "epoch": 13.592592592592592,
+      "grad_norm": 0.8061996102333069,
+      "learning_rate": 1.9627329192546585e-05,
+      "loss": 0.031,
+      "step": 1468
+    },
+    {
+      "epoch": 13.601851851851851,
+      "grad_norm": 0.982791543006897,
+      "learning_rate": 1.950310559006211e-05,
+      "loss": 0.0341,
+      "step": 1469
+    },
+    {
+      "epoch": 13.61111111111111,
+      "grad_norm": 0.7261083126068115,
+      "learning_rate": 1.937888198757764e-05,
+      "loss": 0.0269,
+      "step": 1470
+    },
+    {
+      "epoch": 13.62037037037037,
+      "grad_norm": 1.003458857536316,
+      "learning_rate": 1.9254658385093167e-05,
+      "loss": 0.0344,
+      "step": 1471
+    },
+    {
+      "epoch": 13.62962962962963,
+      "grad_norm": 0.9957963228225708,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 0.0359,
+      "step": 1472
+    },
+    {
+      "epoch": 13.63888888888889,
+      "grad_norm": 0.9977306127548218,
+      "learning_rate": 1.9006211180124224e-05,
+      "loss": 0.0398,
+      "step": 1473
+    },
+    {
+      "epoch": 13.648148148148149,
+      "grad_norm": 0.6655192375183105,
+      "learning_rate": 1.8881987577639754e-05,
+      "loss": 0.0252,
+      "step": 1474
+    },
+    {
+      "epoch": 13.657407407407408,
+      "grad_norm": 0.9726771712303162,
+      "learning_rate": 1.875776397515528e-05,
+      "loss": 0.0345,
+      "step": 1475
+    },
+    {
+      "epoch": 13.666666666666666,
+      "grad_norm": 0.9408605694770813,
+      "learning_rate": 1.8633540372670807e-05,
+      "loss": 0.0295,
+      "step": 1476
+    },
+    {
+      "epoch": 13.675925925925926,
+      "grad_norm": 0.9074751138687134,
+      "learning_rate": 1.8509316770186337e-05,
+      "loss": 0.0294,
+      "step": 1477
+    },
+    {
+      "epoch": 13.685185185185185,
+      "grad_norm": 0.5638766884803772,
+      "learning_rate": 1.8385093167701863e-05,
+      "loss": 0.0215,
+      "step": 1478
+    },
+    {
+      "epoch": 13.694444444444445,
+      "grad_norm": 1.9929472208023071,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 0.0977,
+      "step": 1479
+    },
+    {
+      "epoch": 13.703703703703704,
+      "grad_norm": 0.645876944065094,
+      "learning_rate": 1.813664596273292e-05,
+      "loss": 0.0319,
+      "step": 1480
+    },
+    {
+      "epoch": 13.712962962962964,
+      "grad_norm": 0.8106043338775635,
+      "learning_rate": 1.801242236024845e-05,
+      "loss": 0.0333,
+      "step": 1481
+    },
+    {
+      "epoch": 13.722222222222221,
+      "grad_norm": 0.9891499280929565,
+      "learning_rate": 1.7888198757763976e-05,
+      "loss": 0.0348,
+      "step": 1482
+    },
+    {
+      "epoch": 13.731481481481481,
+      "grad_norm": 0.7550816535949707,
+      "learning_rate": 1.7763975155279506e-05,
+      "loss": 0.0309,
+      "step": 1483
+    },
+    {
+      "epoch": 13.74074074074074,
+      "grad_norm": 0.7451123595237732,
+      "learning_rate": 1.7639751552795032e-05,
+      "loss": 0.0278,
+      "step": 1484
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 0.8352628946304321,
+      "learning_rate": 1.751552795031056e-05,
+      "loss": 0.0274,
+      "step": 1485
+    },
+    {
+      "epoch": 13.75925925925926,
+      "grad_norm": 0.8166201710700989,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.026,
+      "step": 1486
+    },
+    {
+      "epoch": 13.768518518518519,
+      "grad_norm": 1.422154426574707,
+      "learning_rate": 1.7267080745341615e-05,
+      "loss": 0.0392,
+      "step": 1487
+    },
+    {
+      "epoch": 13.777777777777779,
+      "grad_norm": 0.8086047172546387,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.0286,
+      "step": 1488
+    },
+    {
+      "epoch": 13.787037037037036,
+      "grad_norm": 0.9935398101806641,
+      "learning_rate": 1.7018633540372672e-05,
+      "loss": 0.0326,
+      "step": 1489
+    },
+    {
+      "epoch": 13.796296296296296,
+      "grad_norm": 0.8330230712890625,
+      "learning_rate": 1.68944099378882e-05,
+      "loss": 0.0268,
+      "step": 1490
+    },
+    {
+      "epoch": 13.805555555555555,
+      "grad_norm": 0.8333765268325806,
+      "learning_rate": 1.6770186335403728e-05,
+      "loss": 0.0286,
+      "step": 1491
+    },
+    {
+      "epoch": 13.814814814814815,
+      "grad_norm": 0.641851007938385,
+      "learning_rate": 1.6645962732919255e-05,
+      "loss": 0.0285,
+      "step": 1492
+    },
+    {
+      "epoch": 13.824074074074074,
+      "grad_norm": 0.789412260055542,
+      "learning_rate": 1.652173913043478e-05,
+      "loss": 0.0288,
+      "step": 1493
+    },
+    {
+      "epoch": 13.833333333333334,
+      "grad_norm": 0.9055834412574768,
+      "learning_rate": 1.639751552795031e-05,
+      "loss": 0.0278,
+      "step": 1494
+    },
+    {
+      "epoch": 13.842592592592592,
+      "grad_norm": 0.7134174108505249,
+      "learning_rate": 1.6273291925465838e-05,
+      "loss": 0.0292,
+      "step": 1495
+    },
+    {
+      "epoch": 13.851851851851851,
+      "grad_norm": 1.001650333404541,
+      "learning_rate": 1.6149068322981367e-05,
+      "loss": 0.0345,
+      "step": 1496
+    },
+    {
+      "epoch": 13.86111111111111,
+      "grad_norm": 0.800206184387207,
+      "learning_rate": 1.6024844720496894e-05,
+      "loss": 0.0304,
+      "step": 1497
+    },
+    {
+      "epoch": 13.87037037037037,
+      "grad_norm": 0.8903914093971252,
+      "learning_rate": 1.5900621118012424e-05,
+      "loss": 0.0362,
+      "step": 1498
+    },
+    {
+      "epoch": 13.87962962962963,
+      "grad_norm": 0.7213857769966125,
+      "learning_rate": 1.577639751552795e-05,
+      "loss": 0.0317,
+      "step": 1499
+    },
+    {
+      "epoch": 13.88888888888889,
+      "grad_norm": 0.7966362237930298,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 0.0238,
+      "step": 1500
+    },
+    {
+      "epoch": 13.898148148148149,
+      "grad_norm": 0.877825140953064,
+      "learning_rate": 1.5527950310559007e-05,
+      "loss": 0.0293,
+      "step": 1501
+    },
+    {
+      "epoch": 13.907407407407408,
+      "grad_norm": 1.0245945453643799,
+      "learning_rate": 1.5403726708074533e-05,
+      "loss": 0.038,
+      "step": 1502
+    },
+    {
+      "epoch": 13.916666666666666,
+      "grad_norm": 0.9006460309028625,
+      "learning_rate": 1.5279503105590063e-05,
+      "loss": 0.0375,
+      "step": 1503
+    },
+    {
+      "epoch": 13.925925925925926,
+      "grad_norm": 1.0009137392044067,
+      "learning_rate": 1.515527950310559e-05,
+      "loss": 0.036,
+      "step": 1504
+    },
+    {
+      "epoch": 13.935185185185185,
+      "grad_norm": 1.1247698068618774,
+      "learning_rate": 1.503105590062112e-05,
+      "loss": 0.0462,
+      "step": 1505
+    },
+    {
+      "epoch": 13.944444444444445,
+      "grad_norm": 1.1003761291503906,
+      "learning_rate": 1.4906832298136648e-05,
+      "loss": 0.0386,
+      "step": 1506
+    },
+    {
+      "epoch": 13.953703703703704,
+      "grad_norm": 0.9153029918670654,
+      "learning_rate": 1.4782608695652176e-05,
+      "loss": 0.0305,
+      "step": 1507
+    },
+    {
+      "epoch": 13.962962962962964,
+      "grad_norm": 0.6955792903900146,
+      "learning_rate": 1.4658385093167704e-05,
+      "loss": 0.0211,
+      "step": 1508
+    },
+    {
+      "epoch": 13.972222222222221,
+      "grad_norm": 0.918514609336853,
+      "learning_rate": 1.4534161490683232e-05,
+      "loss": 0.0296,
+      "step": 1509
+    },
+    {
+      "epoch": 13.981481481481481,
+      "grad_norm": 0.7454483509063721,
+      "learning_rate": 1.4409937888198757e-05,
+      "loss": 0.029,
+      "step": 1510
+    },
+    {
+      "epoch": 13.99074074074074,
+      "grad_norm": 0.7576525211334229,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.0293,
+      "step": 1511
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.9357626438140869,
+      "learning_rate": 1.4161490683229814e-05,
+      "loss": 0.0339,
+      "step": 1512
+    },
+    {
+      "epoch": 14.00925925925926,
+      "grad_norm": 1.0074899196624756,
+      "learning_rate": 1.4037267080745342e-05,
+      "loss": 0.0338,
+      "step": 1513
+    },
+    {
+      "epoch": 14.018518518518519,
+      "grad_norm": 0.5901816487312317,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 0.0201,
+      "step": 1514
+    },
+    {
+      "epoch": 14.027777777777779,
+      "grad_norm": 0.6215850710868835,
+      "learning_rate": 1.3788819875776398e-05,
+      "loss": 0.0256,
+      "step": 1515
+    },
+    {
+      "epoch": 14.037037037037036,
+      "grad_norm": 0.7354216575622559,
+      "learning_rate": 1.3664596273291926e-05,
+      "loss": 0.0289,
+      "step": 1516
+    },
+    {
+      "epoch": 14.046296296296296,
+      "grad_norm": 0.5978808403015137,
+      "learning_rate": 1.3540372670807455e-05,
+      "loss": 0.0202,
+      "step": 1517
+    },
+    {
+      "epoch": 14.055555555555555,
+      "grad_norm": 0.7752763628959656,
+      "learning_rate": 1.3416149068322981e-05,
+      "loss": 0.0275,
+      "step": 1518
+    },
+    {
+      "epoch": 14.064814814814815,
+      "grad_norm": 0.8024547100067139,
+      "learning_rate": 1.329192546583851e-05,
+      "loss": 0.0254,
+      "step": 1519
+    },
+    {
+      "epoch": 14.074074074074074,
+      "grad_norm": 0.7298763394355774,
+      "learning_rate": 1.3167701863354038e-05,
+      "loss": 0.028,
+      "step": 1520
+    },
+    {
+      "epoch": 14.083333333333334,
+      "grad_norm": 0.4958341419696808,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.0221,
+      "step": 1521
+    },
+    {
+      "epoch": 14.092592592592593,
+      "grad_norm": 0.9592882990837097,
+      "learning_rate": 1.2919254658385094e-05,
+      "loss": 0.03,
+      "step": 1522
+    },
+    {
+      "epoch": 14.101851851851851,
+      "grad_norm": 0.9992770552635193,
+      "learning_rate": 1.2795031055900622e-05,
+      "loss": 0.0291,
+      "step": 1523
+    },
+    {
+      "epoch": 14.11111111111111,
+      "grad_norm": 0.7901219129562378,
+      "learning_rate": 1.267080745341615e-05,
+      "loss": 0.0263,
+      "step": 1524
+    },
+    {
+      "epoch": 14.12037037037037,
+      "grad_norm": 0.7581226825714111,
+      "learning_rate": 1.2546583850931679e-05,
+      "loss": 0.0241,
+      "step": 1525
+    },
+    {
+      "epoch": 14.12962962962963,
+      "grad_norm": 0.9516726732254028,
+      "learning_rate": 1.2422360248447205e-05,
+      "loss": 0.0302,
+      "step": 1526
+    },
+    {
+      "epoch": 14.13888888888889,
+      "grad_norm": 0.791049063205719,
+      "learning_rate": 1.2298136645962733e-05,
+      "loss": 0.0224,
+      "step": 1527
+    },
+    {
+      "epoch": 14.148148148148149,
+      "grad_norm": 0.8296195268630981,
+      "learning_rate": 1.2173913043478261e-05,
+      "loss": 0.0274,
+      "step": 1528
+    },
+    {
+      "epoch": 14.157407407407407,
+      "grad_norm": 0.6744861602783203,
+      "learning_rate": 1.2049689440993788e-05,
+      "loss": 0.0216,
+      "step": 1529
+    },
+    {
+      "epoch": 14.166666666666666,
+      "grad_norm": 0.7672494649887085,
+      "learning_rate": 1.1925465838509318e-05,
+      "loss": 0.0228,
+      "step": 1530
+    },
+    {
+      "epoch": 14.175925925925926,
+      "grad_norm": 0.7237677574157715,
+      "learning_rate": 1.1801242236024846e-05,
+      "loss": 0.0228,
+      "step": 1531
+    },
+    {
+      "epoch": 14.185185185185185,
+      "grad_norm": 1.0658645629882812,
+      "learning_rate": 1.1677018633540374e-05,
+      "loss": 0.0349,
+      "step": 1532
+    },
+    {
+      "epoch": 14.194444444444445,
+      "grad_norm": 0.8038221001625061,
+      "learning_rate": 1.15527950310559e-05,
+      "loss": 0.0258,
+      "step": 1533
+    },
+    {
+      "epoch": 14.203703703703704,
+      "grad_norm": 0.7512486577033997,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.027,
+      "step": 1534
+    },
+    {
+      "epoch": 14.212962962962964,
+      "grad_norm": 0.7751714587211609,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 0.0275,
+      "step": 1535
+    },
+    {
+      "epoch": 14.222222222222221,
+      "grad_norm": 0.8435266017913818,
+      "learning_rate": 1.1180124223602485e-05,
+      "loss": 0.0288,
+      "step": 1536
+    },
+    {
+      "epoch": 14.231481481481481,
+      "grad_norm": 0.7947162985801697,
+      "learning_rate": 1.1055900621118014e-05,
+      "loss": 0.0274,
+      "step": 1537
+    },
+    {
+      "epoch": 14.24074074074074,
+      "grad_norm": 0.6701507568359375,
+      "learning_rate": 1.093167701863354e-05,
+      "loss": 0.0198,
+      "step": 1538
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 0.6321646571159363,
+      "learning_rate": 1.0807453416149068e-05,
+      "loss": 0.0212,
+      "step": 1539
+    },
+    {
+      "epoch": 14.25925925925926,
+      "grad_norm": 0.6891692280769348,
+      "learning_rate": 1.0683229813664597e-05,
+      "loss": 0.0232,
+      "step": 1540
+    },
+    {
+      "epoch": 14.268518518518519,
+      "grad_norm": 0.9563750624656677,
+      "learning_rate": 1.0559006211180125e-05,
+      "loss": 0.0315,
+      "step": 1541
+    },
+    {
+      "epoch": 14.277777777777779,
+      "grad_norm": 0.7779302597045898,
+      "learning_rate": 1.0434782608695651e-05,
+      "loss": 0.0244,
+      "step": 1542
+    },
+    {
+      "epoch": 14.287037037037036,
+      "grad_norm": 0.8213071823120117,
+      "learning_rate": 1.0310559006211181e-05,
+      "loss": 0.0267,
+      "step": 1543
+    },
+    {
+      "epoch": 14.296296296296296,
+      "grad_norm": 0.7205976843833923,
+      "learning_rate": 1.018633540372671e-05,
+      "loss": 0.0237,
+      "step": 1544
+    },
+    {
+      "epoch": 14.305555555555555,
+      "grad_norm": 0.8843669295310974,
+      "learning_rate": 1.0062111801242238e-05,
+      "loss": 0.0264,
+      "step": 1545
+    },
+    {
+      "epoch": 14.314814814814815,
+      "grad_norm": 1.0456665754318237,
+      "learning_rate": 9.937888198757764e-06,
+      "loss": 0.0357,
+      "step": 1546
+    },
+    {
+      "epoch": 14.324074074074074,
+      "grad_norm": 1.090004324913025,
+      "learning_rate": 9.813664596273292e-06,
+      "loss": 0.0351,
+      "step": 1547
+    },
+    {
+      "epoch": 14.333333333333334,
+      "grad_norm": 0.9663504958152771,
+      "learning_rate": 9.68944099378882e-06,
+      "loss": 0.0285,
+      "step": 1548
+    },
+    {
+      "epoch": 14.342592592592593,
+      "grad_norm": 0.6734511256217957,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.023,
+      "step": 1549
+    },
+    {
+      "epoch": 14.351851851851851,
+      "grad_norm": 1.1340980529785156,
+      "learning_rate": 9.440993788819877e-06,
+      "loss": 0.035,
+      "step": 1550
+    },
+    {
+      "epoch": 14.36111111111111,
+      "grad_norm": 0.6549594402313232,
+      "learning_rate": 9.316770186335403e-06,
+      "loss": 0.0184,
+      "step": 1551
+    },
+    {
+      "epoch": 14.37037037037037,
+      "grad_norm": 0.732521116733551,
+      "learning_rate": 9.192546583850932e-06,
+      "loss": 0.0247,
+      "step": 1552
+    },
+    {
+      "epoch": 14.37962962962963,
+      "grad_norm": 0.8055844306945801,
+      "learning_rate": 9.06832298136646e-06,
+      "loss": 0.0307,
+      "step": 1553
+    },
+    {
+      "epoch": 14.38888888888889,
+      "grad_norm": 0.8509539365768433,
+      "learning_rate": 8.944099378881988e-06,
+      "loss": 0.0283,
+      "step": 1554
+    },
+    {
+      "epoch": 14.398148148148149,
+      "grad_norm": 0.6740679144859314,
+      "learning_rate": 8.819875776397516e-06,
+      "loss": 0.0235,
+      "step": 1555
+    },
+    {
+      "epoch": 14.407407407407407,
+      "grad_norm": 0.8333950638771057,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.0296,
+      "step": 1556
+    },
+    {
+      "epoch": 14.416666666666666,
+      "grad_norm": 1.1533918380737305,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.0308,
+      "step": 1557
+    },
+    {
+      "epoch": 14.425925925925926,
+      "grad_norm": 0.7060422301292419,
+      "learning_rate": 8.4472049689441e-06,
+      "loss": 0.0244,
+      "step": 1558
+    },
+    {
+      "epoch": 14.435185185185185,
+      "grad_norm": 0.9376555681228638,
+      "learning_rate": 8.322981366459627e-06,
+      "loss": 0.027,
+      "step": 1559
+    },
+    {
+      "epoch": 14.444444444444445,
+      "grad_norm": 1.0092798471450806,
+      "learning_rate": 8.198757763975156e-06,
+      "loss": 0.0291,
+      "step": 1560
+    },
+    {
+      "epoch": 14.453703703703704,
+      "grad_norm": 0.8119834661483765,
+      "learning_rate": 8.074534161490684e-06,
+      "loss": 0.0254,
+      "step": 1561
+    },
+    {
+      "epoch": 14.462962962962964,
+      "grad_norm": 1.0415433645248413,
+      "learning_rate": 7.950310559006212e-06,
+      "loss": 0.032,
+      "step": 1562
+    },
+    {
+      "epoch": 14.472222222222221,
+      "grad_norm": 0.9587968587875366,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.0321,
+      "step": 1563
+    },
+    {
+      "epoch": 14.481481481481481,
+      "grad_norm": 0.9727386832237244,
+      "learning_rate": 7.701863354037267e-06,
+      "loss": 0.0344,
+      "step": 1564
+    },
+    {
+      "epoch": 14.49074074074074,
+      "grad_norm": 0.8076003193855286,
+      "learning_rate": 7.577639751552795e-06,
+      "loss": 0.0255,
+      "step": 1565
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 0.7089889049530029,
+      "learning_rate": 7.453416149068324e-06,
+      "loss": 0.0239,
+      "step": 1566
+    },
+    {
+      "epoch": 14.50925925925926,
+      "grad_norm": 0.6875848174095154,
+      "learning_rate": 7.329192546583852e-06,
+      "loss": 0.0257,
+      "step": 1567
+    },
+    {
+      "epoch": 14.518518518518519,
+      "grad_norm": 0.7025448679924011,
+      "learning_rate": 7.204968944099379e-06,
+      "loss": 0.0218,
+      "step": 1568
+    },
+    {
+      "epoch": 14.527777777777779,
+      "grad_norm": 0.686010479927063,
+      "learning_rate": 7.080745341614907e-06,
+      "loss": 0.0204,
+      "step": 1569
+    },
+    {
+      "epoch": 14.537037037037036,
+      "grad_norm": 0.8934335708618164,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.0291,
+      "step": 1570
+    },
+    {
+      "epoch": 14.546296296296296,
+      "grad_norm": 0.7375936508178711,
+      "learning_rate": 6.832298136645963e-06,
+      "loss": 0.024,
+      "step": 1571
+    },
+    {
+      "epoch": 14.555555555555555,
+      "grad_norm": 0.865662693977356,
+      "learning_rate": 6.708074534161491e-06,
+      "loss": 0.0263,
+      "step": 1572
+    },
+    {
+      "epoch": 14.564814814814815,
+      "grad_norm": 0.6800843477249146,
+      "learning_rate": 6.583850931677019e-06,
+      "loss": 0.0247,
+      "step": 1573
+    },
+    {
+      "epoch": 14.574074074074074,
+      "grad_norm": 0.8272568583488464,
+      "learning_rate": 6.459627329192547e-06,
+      "loss": 0.0316,
+      "step": 1574
+    },
+    {
+      "epoch": 14.583333333333334,
+      "grad_norm": 0.5861930251121521,
+      "learning_rate": 6.335403726708075e-06,
+      "loss": 0.0217,
+      "step": 1575
+    },
+    {
+      "epoch": 14.592592592592592,
+      "grad_norm": 0.6214504241943359,
+      "learning_rate": 6.2111801242236025e-06,
+      "loss": 0.0212,
+      "step": 1576
+    },
+    {
+      "epoch": 14.601851851851851,
+      "grad_norm": 0.5459976196289062,
+      "learning_rate": 6.086956521739131e-06,
+      "loss": 0.0189,
+      "step": 1577
+    },
+    {
+      "epoch": 14.61111111111111,
+      "grad_norm": 1.0046416521072388,
+      "learning_rate": 5.962732919254659e-06,
+      "loss": 0.0337,
+      "step": 1578
+    },
+    {
+      "epoch": 14.62037037037037,
+      "grad_norm": 0.7500995993614197,
+      "learning_rate": 5.838509316770187e-06,
+      "loss": 0.0259,
+      "step": 1579
+    },
+    {
+      "epoch": 14.62962962962963,
+      "grad_norm": 0.8355032801628113,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.0255,
+      "step": 1580
+    },
+    {
+      "epoch": 14.63888888888889,
+      "grad_norm": 2.543530225753784,
+      "learning_rate": 5.590062111801243e-06,
+      "loss": 0.0551,
+      "step": 1581
+    },
+    {
+      "epoch": 14.648148148148149,
+      "grad_norm": 0.7404892444610596,
+      "learning_rate": 5.46583850931677e-06,
+      "loss": 0.0234,
+      "step": 1582
+    },
+    {
+      "epoch": 14.657407407407408,
+      "grad_norm": 0.8910460472106934,
+      "learning_rate": 5.341614906832298e-06,
+      "loss": 0.026,
+      "step": 1583
+    },
+    {
+      "epoch": 14.666666666666666,
+      "grad_norm": 0.7894750237464905,
+      "learning_rate": 5.217391304347826e-06,
+      "loss": 0.031,
+      "step": 1584
+    },
+    {
+      "epoch": 14.675925925925926,
+      "grad_norm": 0.719997763633728,
+      "learning_rate": 5.093167701863355e-06,
+      "loss": 0.0236,
+      "step": 1585
+    },
+    {
+      "epoch": 14.685185185185185,
+      "grad_norm": 0.8099305629730225,
+      "learning_rate": 4.968944099378882e-06,
+      "loss": 0.0272,
+      "step": 1586
+    },
+    {
+      "epoch": 14.694444444444445,
+      "grad_norm": 0.9355957508087158,
+      "learning_rate": 4.84472049689441e-06,
+      "loss": 0.0279,
+      "step": 1587
+    },
+    {
+      "epoch": 14.703703703703704,
+      "grad_norm": 0.9016175270080566,
+      "learning_rate": 4.7204968944099384e-06,
+      "loss": 0.0275,
+      "step": 1588
+    },
+    {
+      "epoch": 14.712962962962964,
+      "grad_norm": 0.8064716458320618,
+      "learning_rate": 4.596273291925466e-06,
+      "loss": 0.0275,
+      "step": 1589
+    },
+    {
+      "epoch": 14.722222222222221,
+      "grad_norm": 0.7592180967330933,
+      "learning_rate": 4.472049689440994e-06,
+      "loss": 0.0246,
+      "step": 1590
+    },
+    {
+      "epoch": 14.731481481481481,
+      "grad_norm": 0.8269107341766357,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.0275,
+      "step": 1591
+    },
+    {
+      "epoch": 14.74074074074074,
+      "grad_norm": 0.8514775633811951,
+      "learning_rate": 4.22360248447205e-06,
+      "loss": 0.0282,
+      "step": 1592
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 0.7846024632453918,
+      "learning_rate": 4.099378881987578e-06,
+      "loss": 0.0235,
+      "step": 1593
+    },
+    {
+      "epoch": 14.75925925925926,
+      "grad_norm": 0.8372007012367249,
+      "learning_rate": 3.975155279503106e-06,
+      "loss": 0.0286,
+      "step": 1594
+    },
+    {
+      "epoch": 14.768518518518519,
+      "grad_norm": 0.965676486492157,
+      "learning_rate": 3.850931677018633e-06,
+      "loss": 0.0316,
+      "step": 1595
+    },
+    {
+      "epoch": 14.777777777777779,
+      "grad_norm": 0.925922691822052,
+      "learning_rate": 3.726708074534162e-06,
+      "loss": 0.0281,
+      "step": 1596
+    },
+    {
+      "epoch": 14.787037037037036,
+      "grad_norm": 1.0515544414520264,
+      "learning_rate": 3.6024844720496893e-06,
+      "loss": 0.031,
+      "step": 1597
+    },
+    {
+      "epoch": 14.796296296296296,
+      "grad_norm": 0.88392573595047,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.0282,
+      "step": 1598
+    },
+    {
+      "epoch": 14.805555555555555,
+      "grad_norm": 0.984674334526062,
+      "learning_rate": 3.3540372670807453e-06,
+      "loss": 0.0273,
+      "step": 1599
+    },
+    {
+      "epoch": 14.814814814814815,
+      "grad_norm": 0.7821362018585205,
+      "learning_rate": 3.2298136645962735e-06,
+      "loss": 0.0295,
+      "step": 1600
+    },
+    {
+      "epoch": 14.824074074074074,
+      "grad_norm": 0.9542562961578369,
+      "learning_rate": 3.1055900621118013e-06,
+      "loss": 0.0296,
+      "step": 1601
+    },
+    {
+      "epoch": 14.833333333333334,
+      "grad_norm": 1.00540292263031,
+      "learning_rate": 2.9813664596273295e-06,
+      "loss": 0.0302,
+      "step": 1602
+    },
+    {
+      "epoch": 14.842592592592592,
+      "grad_norm": 0.8111598491668701,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.0231,
+      "step": 1603
+    },
+    {
+      "epoch": 14.851851851851851,
+      "grad_norm": 0.7784263491630554,
+      "learning_rate": 2.732919254658385e-06,
+      "loss": 0.0276,
+      "step": 1604
+    },
+    {
+      "epoch": 14.86111111111111,
+      "grad_norm": 0.6490851044654846,
+      "learning_rate": 2.608695652173913e-06,
+      "loss": 0.0196,
+      "step": 1605
+    },
+    {
+      "epoch": 14.87037037037037,
+      "grad_norm": 0.8383249640464783,
+      "learning_rate": 2.484472049689441e-06,
+      "loss": 0.0244,
+      "step": 1606
+    },
+    {
+      "epoch": 14.87962962962963,
+      "grad_norm": 0.7719290852546692,
+      "learning_rate": 2.3602484472049692e-06,
+      "loss": 0.0288,
+      "step": 1607
+    },
+    {
+      "epoch": 14.88888888888889,
+      "grad_norm": 0.9684656858444214,
+      "learning_rate": 2.236024844720497e-06,
+      "loss": 0.0296,
+      "step": 1608
+    },
+    {
+      "epoch": 14.898148148148149,
+      "grad_norm": 0.8234147429466248,
+      "learning_rate": 2.111801242236025e-06,
+      "loss": 0.0237,
+      "step": 1609
+    },
+    {
+      "epoch": 14.907407407407408,
+      "grad_norm": 0.7515906095504761,
+      "learning_rate": 1.987577639751553e-06,
+      "loss": 0.0281,
+      "step": 1610
+    },
+    {
+      "epoch": 14.916666666666666,
+      "grad_norm": 0.7840167284011841,
+      "learning_rate": 1.863354037267081e-06,
+      "loss": 0.0239,
+      "step": 1611
+    },
+    {
+      "epoch": 14.925925925925926,
+      "grad_norm": 0.720740020275116,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.0249,
+      "step": 1612
+    },
+    {
+      "epoch": 14.935185185185185,
+      "grad_norm": 0.9448007941246033,
+      "learning_rate": 1.6149068322981367e-06,
+      "loss": 0.0301,
+      "step": 1613
+    },
+    {
+      "epoch": 14.944444444444445,
+      "grad_norm": 0.5331466197967529,
+      "learning_rate": 1.4906832298136647e-06,
+      "loss": 0.0187,
+      "step": 1614
+    },
+    {
+      "epoch": 14.953703703703704,
+      "grad_norm": 0.8185423016548157,
+      "learning_rate": 1.3664596273291925e-06,
+      "loss": 0.0309,
+      "step": 1615
+    },
+    {
+      "epoch": 14.962962962962964,
+      "grad_norm": 0.684502124786377,
+      "learning_rate": 1.2422360248447205e-06,
+      "loss": 0.0211,
+      "step": 1616
+    },
+    {
+      "epoch": 14.972222222222221,
+      "grad_norm": 0.8936087489128113,
+      "learning_rate": 1.1180124223602485e-06,
+      "loss": 0.0252,
+      "step": 1617
+    },
+    {
+      "epoch": 14.981481481481481,
+      "grad_norm": 1.2246485948562622,
+      "learning_rate": 9.937888198757765e-07,
+      "loss": 0.0386,
+      "step": 1618
+    },
+    {
+      "epoch": 14.99074074074074,
+      "grad_norm": 0.8689442873001099,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 0.0265,
+      "step": 1619
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.6490731239318848,
+      "learning_rate": 7.453416149068324e-07,
+      "loss": 0.0299,
+      "step": 1620
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.598422110437376e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}