diff --git "a/q2.5-ga/checkpoint-1200/trainer_state.json" "b/q2.5-ga/checkpoint-1200/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/q2.5-ga/checkpoint-1200/trainer_state.json"
@@ -0,0 +1,4241 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.42048092505803514,
+  "eval_steps": 500,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00035040077088169594,
+      "grad_norm": 6.5142412185668945,
+      "learning_rate": 0.0,
+      "loss": 5.324,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007008015417633919,
+      "grad_norm": 6.758334159851074,
+      "learning_rate": 6.993006993006994e-07,
+      "loss": 5.3405,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014016030835267838,
+      "grad_norm": 6.22674036026001,
+      "learning_rate": 2.0979020979020983e-06,
+      "loss": 5.3286,
+      "step": 4
+    },
+    {
+      "epoch": 0.0021024046252901755,
+      "grad_norm": 5.438386917114258,
+      "learning_rate": 3.496503496503497e-06,
+      "loss": 5.25,
+      "step": 6
+    },
+    {
+      "epoch": 0.0028032061670535675,
+      "grad_norm": 3.365504741668701,
+      "learning_rate": 4.895104895104895e-06,
+      "loss": 5.2821,
+      "step": 8
+    },
+    {
+      "epoch": 0.0035040077088169595,
+      "grad_norm": 7.186147212982178,
+      "learning_rate": 6.2937062937062944e-06,
+      "loss": 5.21,
+      "step": 10
+    },
+    {
+      "epoch": 0.004204809250580351,
+      "grad_norm": 4.960826396942139,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 5.0759,
+      "step": 12
+    },
+    {
+      "epoch": 0.004905610792343743,
+      "grad_norm": 4.001464366912842,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 5.1092,
+      "step": 14
+    },
+    {
+      "epoch": 0.005606412334107135,
+      "grad_norm": 3.2986342906951904,
+      "learning_rate": 1.048951048951049e-05,
+      "loss": 4.93,
+      "step": 16
+    },
+    {
+      "epoch": 0.006307213875870527,
+      "grad_norm": 2.5407276153564453,
+      "learning_rate": 1.188811188811189e-05,
+      "loss": 4.8535,
+      "step": 18
+    },
+    {
+      "epoch": 0.007008015417633919,
+      "grad_norm": 2.211754083633423,
+      "learning_rate": 1.3286713286713287e-05,
+      "loss": 4.74,
+      "step": 20
+    },
+    {
+      "epoch": 0.007708816959397311,
+      "grad_norm": 1.6710195541381836,
+      "learning_rate": 1.4685314685314686e-05,
+      "loss": 4.609,
+      "step": 22
+    },
+    {
+      "epoch": 0.008409618501160702,
+      "grad_norm": 1.280752182006836,
+      "learning_rate": 1.6083916083916083e-05,
+      "loss": 4.4879,
+      "step": 24
+    },
+    {
+      "epoch": 0.009110420042924094,
+      "grad_norm": 1.312186598777771,
+      "learning_rate": 1.7482517482517483e-05,
+      "loss": 4.3995,
+      "step": 26
+    },
+    {
+      "epoch": 0.009811221584687486,
+      "grad_norm": 1.3315190076828003,
+      "learning_rate": 1.888111888111888e-05,
+      "loss": 4.3005,
+      "step": 28
+    },
+    {
+      "epoch": 0.010512023126450878,
+      "grad_norm": 1.3252590894699097,
+      "learning_rate": 2.027972027972028e-05,
+      "loss": 4.1952,
+      "step": 30
+    },
+    {
+      "epoch": 0.01121282466821427,
+      "grad_norm": 1.3794758319854736,
+      "learning_rate": 2.1678321678321677e-05,
+      "loss": 4.1459,
+      "step": 32
+    },
+    {
+      "epoch": 0.011913626209977662,
+      "grad_norm": 1.1808068752288818,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 4.034,
+      "step": 34
+    },
+    {
+      "epoch": 0.012614427751741054,
+      "grad_norm": 1.31660795211792,
+      "learning_rate": 2.4475524475524478e-05,
+      "loss": 3.926,
+      "step": 36
+    },
+    {
+      "epoch": 0.013315229293504446,
+      "grad_norm": 1.0347495079040527,
+      "learning_rate": 2.5874125874125877e-05,
+      "loss": 3.8812,
+      "step": 38
+    },
+    {
+      "epoch": 0.014016030835267838,
+      "grad_norm": 1.050775408744812,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 3.7787,
+      "step": 40
+    },
+    {
+      "epoch": 0.01471683237703123,
+      "grad_norm": 0.9461761713027954,
+      "learning_rate": 2.8671328671328672e-05,
+      "loss": 3.6738,
+      "step": 42
+    },
+    {
+      "epoch": 0.015417633918794622,
+      "grad_norm": 1.0460454225540161,
+      "learning_rate": 3.0069930069930068e-05,
+      "loss": 3.6385,
+      "step": 44
+    },
+    {
+      "epoch": 0.016118435460558012,
+      "grad_norm": 1.0687191486358643,
+      "learning_rate": 3.146853146853147e-05,
+      "loss": 3.5701,
+      "step": 46
+    },
+    {
+      "epoch": 0.016819237002321404,
+      "grad_norm": 1.4722611904144287,
+      "learning_rate": 3.2867132867132866e-05,
+      "loss": 3.5438,
+      "step": 48
+    },
+    {
+      "epoch": 0.017520038544084796,
+      "grad_norm": 1.1305724382400513,
+      "learning_rate": 3.4265734265734265e-05,
+      "loss": 3.4694,
+      "step": 50
+    },
+    {
+      "epoch": 0.018220840085848188,
+      "grad_norm": 0.9322625994682312,
+      "learning_rate": 3.566433566433567e-05,
+      "loss": 3.4488,
+      "step": 52
+    },
+    {
+      "epoch": 0.01892164162761158,
+      "grad_norm": 1.2441555261611938,
+      "learning_rate": 3.7062937062937064e-05,
+      "loss": 3.4289,
+      "step": 54
+    },
+    {
+      "epoch": 0.019622443169374972,
+      "grad_norm": 0.9397731423377991,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 3.4021,
+      "step": 56
+    },
+    {
+      "epoch": 0.020323244711138364,
+      "grad_norm": 1.3261164426803589,
+      "learning_rate": 3.986013986013986e-05,
+      "loss": 3.3575,
+      "step": 58
+    },
+    {
+      "epoch": 0.021024046252901756,
+      "grad_norm": 1.08541738986969,
+      "learning_rate": 4.125874125874126e-05,
+      "loss": 3.3403,
+      "step": 60
+    },
+    {
+      "epoch": 0.021724847794665148,
+      "grad_norm": 0.8626166582107544,
+      "learning_rate": 4.265734265734266e-05,
+      "loss": 3.3306,
+      "step": 62
+    },
+    {
+      "epoch": 0.02242564933642854,
+      "grad_norm": 1.0596344470977783,
+      "learning_rate": 4.405594405594406e-05,
+      "loss": 3.2779,
+      "step": 64
+    },
+    {
+      "epoch": 0.023126450878191932,
+      "grad_norm": 1.511917233467102,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 3.2759,
+      "step": 66
+    },
+    {
+      "epoch": 0.023827252419955324,
+      "grad_norm": 1.2062046527862549,
+      "learning_rate": 4.685314685314686e-05,
+      "loss": 3.2545,
+      "step": 68
+    },
+    {
+      "epoch": 0.024528053961718716,
+      "grad_norm": 1.1399930715560913,
+      "learning_rate": 4.825174825174825e-05,
+      "loss": 3.2235,
+      "step": 70
+    },
+    {
+      "epoch": 0.025228855503482108,
+      "grad_norm": 0.8960133790969849,
+      "learning_rate": 4.9650349650349656e-05,
+      "loss": 3.2025,
+      "step": 72
+    },
+    {
+      "epoch": 0.0259296570452455,
+      "grad_norm": 1.3042056560516357,
+      "learning_rate": 5.1048951048951055e-05,
+      "loss": 3.1475,
+      "step": 74
+    },
+    {
+      "epoch": 0.026630458587008892,
+      "grad_norm": 1.186320424079895,
+      "learning_rate": 5.244755244755245e-05,
+      "loss": 3.1759,
+      "step": 76
+    },
+    {
+      "epoch": 0.027331260128772284,
+      "grad_norm": 1.2691158056259155,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 3.1296,
+      "step": 78
+    },
+    {
+      "epoch": 0.028032061670535676,
+      "grad_norm": 0.7816159129142761,
+      "learning_rate": 5.524475524475524e-05,
+      "loss": 3.1017,
+      "step": 80
+    },
+    {
+      "epoch": 0.028732863212299068,
+      "grad_norm": 1.1489295959472656,
+      "learning_rate": 5.664335664335665e-05,
+      "loss": 3.1151,
+      "step": 82
+    },
+    {
+      "epoch": 0.02943366475406246,
+      "grad_norm": 1.5686062574386597,
+      "learning_rate": 5.8041958041958044e-05,
+      "loss": 3.114,
+      "step": 84
+    },
+    {
+      "epoch": 0.030134466295825852,
+      "grad_norm": 1.4421433210372925,
+      "learning_rate": 5.944055944055944e-05,
+      "loss": 3.0946,
+      "step": 86
+    },
+    {
+      "epoch": 0.030835267837589244,
+      "grad_norm": 1.335250973701477,
+      "learning_rate": 6.083916083916085e-05,
+      "loss": 3.084,
+      "step": 88
+    },
+    {
+      "epoch": 0.03153606937935263,
+      "grad_norm": 0.970507800579071,
+      "learning_rate": 6.223776223776224e-05,
+      "loss": 3.1163,
+      "step": 90
+    },
+    {
+      "epoch": 0.032236870921116025,
+      "grad_norm": 1.2849407196044922,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 3.063,
+      "step": 92
+    },
+    {
+      "epoch": 0.032937672462879417,
+      "grad_norm": 1.0378247499465942,
+      "learning_rate": 6.503496503496504e-05,
+      "loss": 3.0223,
+      "step": 94
+    },
+    {
+      "epoch": 0.03363847400464281,
+      "grad_norm": 1.3139392137527466,
+      "learning_rate": 6.643356643356644e-05,
+      "loss": 3.0572,
+      "step": 96
+    },
+    {
+      "epoch": 0.0343392755464062,
+      "grad_norm": 1.254752278327942,
+      "learning_rate": 6.783216783216784e-05,
+      "loss": 3.0408,
+      "step": 98
+    },
+    {
+      "epoch": 0.03504007708816959,
+      "grad_norm": 1.3333168029785156,
+      "learning_rate": 6.923076923076924e-05,
+      "loss": 3.0185,
+      "step": 100
+    },
+    {
+      "epoch": 0.035740878629932984,
+      "grad_norm": 1.2795464992523193,
+      "learning_rate": 7.062937062937062e-05,
+      "loss": 3.0328,
+      "step": 102
+    },
+    {
+      "epoch": 0.036441680171696376,
+      "grad_norm": 1.2025645971298218,
+      "learning_rate": 7.202797202797204e-05,
+      "loss": 3.0303,
+      "step": 104
+    },
+    {
+      "epoch": 0.03714248171345977,
+      "grad_norm": 1.1741266250610352,
+      "learning_rate": 7.342657342657343e-05,
+      "loss": 3.0252,
+      "step": 106
+    },
+    {
+      "epoch": 0.03784328325522316,
+      "grad_norm": 1.2022653818130493,
+      "learning_rate": 7.482517482517482e-05,
+      "loss": 3.0183,
+      "step": 108
+    },
+    {
+      "epoch": 0.03854408479698655,
+      "grad_norm": 1.1950666904449463,
+      "learning_rate": 7.622377622377622e-05,
+      "loss": 2.9804,
+      "step": 110
+    },
+    {
+      "epoch": 0.039244886338749944,
+      "grad_norm": 1.5780822038650513,
+      "learning_rate": 7.762237762237763e-05,
+      "loss": 2.9804,
+      "step": 112
+    },
+    {
+      "epoch": 0.039945687880513336,
+      "grad_norm": 1.0478655099868774,
+      "learning_rate": 7.902097902097903e-05,
+      "loss": 2.9894,
+      "step": 114
+    },
+    {
+      "epoch": 0.04064648942227673,
+      "grad_norm": 1.1782268285751343,
+      "learning_rate": 8.041958041958042e-05,
+      "loss": 2.9717,
+      "step": 116
+    },
+    {
+      "epoch": 0.04134729096404012,
+      "grad_norm": 1.0321820974349976,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 2.9776,
+      "step": 118
+    },
+    {
+      "epoch": 0.04204809250580351,
+      "grad_norm": 0.9697206020355225,
+      "learning_rate": 8.321678321678323e-05,
+      "loss": 2.9804,
+      "step": 120
+    },
+    {
+      "epoch": 0.042748894047566904,
+      "grad_norm": 1.1984606981277466,
+      "learning_rate": 8.461538461538461e-05,
+      "loss": 2.9495,
+      "step": 122
+    },
+    {
+      "epoch": 0.043449695589330296,
+      "grad_norm": 0.9830178618431091,
+      "learning_rate": 8.601398601398601e-05,
+      "loss": 2.9656,
+      "step": 124
+    },
+    {
+      "epoch": 0.04415049713109369,
+      "grad_norm": 1.3105114698410034,
+      "learning_rate": 8.741258741258743e-05,
+      "loss": 2.9306,
+      "step": 126
+    },
+    {
+      "epoch": 0.04485129867285708,
+      "grad_norm": 1.3499157428741455,
+      "learning_rate": 8.881118881118881e-05,
+      "loss": 2.9381,
+      "step": 128
+    },
+    {
+      "epoch": 0.04555210021462047,
+      "grad_norm": 0.9977575540542603,
+      "learning_rate": 9.020979020979021e-05,
+      "loss": 2.907,
+      "step": 130
+    },
+    {
+      "epoch": 0.046252901756383864,
+      "grad_norm": 1.2331498861312866,
+      "learning_rate": 9.160839160839161e-05,
+      "loss": 2.9224,
+      "step": 132
+    },
+    {
+      "epoch": 0.046953703298147256,
+      "grad_norm": 1.451253890991211,
+      "learning_rate": 9.300699300699301e-05,
+      "loss": 2.9202,
+      "step": 134
+    },
+    {
+      "epoch": 0.04765450483991065,
+      "grad_norm": 1.2146471738815308,
+      "learning_rate": 9.440559440559441e-05,
+      "loss": 2.9098,
+      "step": 136
+    },
+    {
+      "epoch": 0.04835530638167404,
+      "grad_norm": 1.0873245000839233,
+      "learning_rate": 9.580419580419581e-05,
+      "loss": 2.9218,
+      "step": 138
+    },
+    {
+      "epoch": 0.04905610792343743,
+      "grad_norm": 1.276413083076477,
+      "learning_rate": 9.72027972027972e-05,
+      "loss": 2.8947,
+      "step": 140
+    },
+    {
+      "epoch": 0.049756909465200824,
+      "grad_norm": 1.126065731048584,
+      "learning_rate": 9.86013986013986e-05,
+      "loss": 2.8788,
+      "step": 142
+    },
+    {
+      "epoch": 0.050457711006964216,
+      "grad_norm": 1.5177017450332642,
+      "learning_rate": 0.0001,
+      "loss": 2.9043,
+      "step": 144
+    },
+    {
+      "epoch": 0.05115851254872761,
+      "grad_norm": 1.3744112253189087,
+      "learning_rate": 9.99998657109765e-05,
+      "loss": 2.888,
+      "step": 146
+    },
+    {
+      "epoch": 0.051859314090491,
+      "grad_norm": 1.7921055555343628,
+      "learning_rate": 9.999946284462733e-05,
+      "loss": 2.8631,
+      "step": 148
+    },
+    {
+      "epoch": 0.05256011563225439,
+      "grad_norm": 1.1755317449569702,
+      "learning_rate": 9.999879140311652e-05,
+      "loss": 2.8735,
+      "step": 150
+    },
+    {
+      "epoch": 0.053260917174017784,
+      "grad_norm": 0.846362292766571,
+      "learning_rate": 9.999785139005073e-05,
+      "loss": 2.8768,
+      "step": 152
+    },
+    {
+      "epoch": 0.053961718715781176,
+      "grad_norm": 0.9867280721664429,
+      "learning_rate": 9.999664281047933e-05,
+      "loss": 2.8859,
+      "step": 154
+    },
+    {
+      "epoch": 0.05466252025754457,
+      "grad_norm": 0.9751666188240051,
+      "learning_rate": 9.999516567089429e-05,
+      "loss": 2.8497,
+      "step": 156
+    },
+    {
+      "epoch": 0.05536332179930796,
+      "grad_norm": 1.0603703260421753,
+      "learning_rate": 9.999341997923011e-05,
+      "loss": 2.8404,
+      "step": 158
+    },
+    {
+      "epoch": 0.05606412334107135,
+      "grad_norm": 1.0447975397109985,
+      "learning_rate": 9.999140574486392e-05,
+      "loss": 2.9092,
+      "step": 160
+    },
+    {
+      "epoch": 0.056764924882834744,
+      "grad_norm": 1.3046443462371826,
+      "learning_rate": 9.998912297861527e-05,
+      "loss": 2.8971,
+      "step": 162
+    },
+    {
+      "epoch": 0.057465726424598136,
+      "grad_norm": 1.1029243469238281,
+      "learning_rate": 9.998657169274622e-05,
+      "loss": 2.8834,
+      "step": 164
+    },
+    {
+      "epoch": 0.05816652796636153,
+      "grad_norm": 0.8594210743904114,
+      "learning_rate": 9.99837519009611e-05,
+      "loss": 2.8361,
+      "step": 166
+    },
+    {
+      "epoch": 0.05886732950812492,
+      "grad_norm": 0.8585363030433655,
+      "learning_rate": 9.998066361840665e-05,
+      "loss": 2.8782,
+      "step": 168
+    },
+    {
+      "epoch": 0.05956813104988831,
+      "grad_norm": 0.693467378616333,
+      "learning_rate": 9.997730686167173e-05,
+      "loss": 2.8537,
+      "step": 170
+    },
+    {
+      "epoch": 0.060268932591651704,
+      "grad_norm": 0.8418940305709839,
+      "learning_rate": 9.997368164878738e-05,
+      "loss": 2.8294,
+      "step": 172
+    },
+    {
+      "epoch": 0.060969734133415096,
+      "grad_norm": 0.9938271045684814,
+      "learning_rate": 9.996978799922665e-05,
+      "loss": 2.8458,
+      "step": 174
+    },
+    {
+      "epoch": 0.06167053567517849,
+      "grad_norm": 1.0347217321395874,
+      "learning_rate": 9.99656259339045e-05,
+      "loss": 2.8081,
+      "step": 176
+    },
+    {
+      "epoch": 0.06237133721694188,
+      "grad_norm": 0.9216743111610413,
+      "learning_rate": 9.996119547517775e-05,
+      "loss": 2.8655,
+      "step": 178
+    },
+    {
+      "epoch": 0.06307213875870527,
+      "grad_norm": 1.0579859018325806,
+      "learning_rate": 9.995649664684486e-05,
+      "loss": 2.823,
+      "step": 180
+    },
+    {
+      "epoch": 0.06377294030046866,
+      "grad_norm": 0.9864194393157959,
+      "learning_rate": 9.995152947414586e-05,
+      "loss": 2.8081,
+      "step": 182
+    },
+    {
+      "epoch": 0.06447374184223205,
+      "grad_norm": 0.8999143838882446,
+      "learning_rate": 9.994629398376226e-05,
+      "loss": 2.7947,
+      "step": 184
+    },
+    {
+      "epoch": 0.06517454338399545,
+      "grad_norm": 0.9121315479278564,
+      "learning_rate": 9.994079020381676e-05,
+      "loss": 2.8253,
+      "step": 186
+    },
+    {
+      "epoch": 0.06587534492575883,
+      "grad_norm": 0.8578842282295227,
+      "learning_rate": 9.993501816387329e-05,
+      "loss": 2.7548,
+      "step": 188
+    },
+    {
+      "epoch": 0.06657614646752223,
+      "grad_norm": 0.8564820289611816,
+      "learning_rate": 9.992897789493672e-05,
+      "loss": 2.8361,
+      "step": 190
+    },
+    {
+      "epoch": 0.06727694800928562,
+      "grad_norm": 0.8013344407081604,
+      "learning_rate": 9.992266942945269e-05,
+      "loss": 2.8606,
+      "step": 192
+    },
+    {
+      "epoch": 0.06797774955104902,
+      "grad_norm": 0.7343975901603699,
+      "learning_rate": 9.991609280130752e-05,
+      "loss": 2.7947,
+      "step": 194
+    },
+    {
+      "epoch": 0.0686785510928124,
+      "grad_norm": 0.7338536381721497,
+      "learning_rate": 9.990924804582797e-05,
+      "loss": 2.7492,
+      "step": 196
+    },
+    {
+      "epoch": 0.0693793526345758,
+      "grad_norm": 0.828781008720398,
+      "learning_rate": 9.990213519978109e-05,
+      "loss": 2.8013,
+      "step": 198
+    },
+    {
+      "epoch": 0.07008015417633918,
+      "grad_norm": 0.7156624794006348,
+      "learning_rate": 9.989475430137391e-05,
+      "loss": 2.7943,
+      "step": 200
+    },
+    {
+      "epoch": 0.07078095571810258,
+      "grad_norm": 0.6014353632926941,
+      "learning_rate": 9.988710539025341e-05,
+      "loss": 2.8099,
+      "step": 202
+    },
+    {
+      "epoch": 0.07148175725986597,
+      "grad_norm": 0.6569661498069763,
+      "learning_rate": 9.987918850750619e-05,
+      "loss": 2.8125,
+      "step": 204
+    },
+    {
+      "epoch": 0.07218255880162937,
+      "grad_norm": 0.6558775305747986,
+      "learning_rate": 9.987100369565825e-05,
+      "loss": 2.7487,
+      "step": 206
+    },
+    {
+      "epoch": 0.07288336034339275,
+      "grad_norm": 0.6454245448112488,
+      "learning_rate": 9.986255099867481e-05,
+      "loss": 2.7648,
+      "step": 208
+    },
+    {
+      "epoch": 0.07358416188515615,
+      "grad_norm": 0.5741921067237854,
+      "learning_rate": 9.985383046196004e-05,
+      "loss": 2.7743,
+      "step": 210
+    },
+    {
+      "epoch": 0.07428496342691954,
+      "grad_norm": 0.5875937938690186,
+      "learning_rate": 9.984484213235685e-05,
+      "loss": 2.7728,
+      "step": 212
+    },
+    {
+      "epoch": 0.07498576496868294,
+      "grad_norm": 0.6638422012329102,
+      "learning_rate": 9.98355860581466e-05,
+      "loss": 2.7504,
+      "step": 214
+    },
+    {
+      "epoch": 0.07568656651044632,
+      "grad_norm": 1.1614341735839844,
+      "learning_rate": 9.982606228904884e-05,
+      "loss": 2.7923,
+      "step": 216
+    },
+    {
+      "epoch": 0.07638736805220972,
+      "grad_norm": 1.005254864692688,
+      "learning_rate": 9.981627087622108e-05,
+      "loss": 2.76,
+      "step": 218
+    },
+    {
+      "epoch": 0.0770881695939731,
+      "grad_norm": 0.7738555669784546,
+      "learning_rate": 9.980621187225852e-05,
+      "loss": 2.7866,
+      "step": 220
+    },
+    {
+      "epoch": 0.0777889711357365,
+      "grad_norm": 0.9469527006149292,
+      "learning_rate": 9.979588533119367e-05,
+      "loss": 2.8012,
+      "step": 222
+    },
+    {
+      "epoch": 0.07848977267749989,
+      "grad_norm": 0.9031473398208618,
+      "learning_rate": 9.978529130849619e-05,
+      "loss": 2.7522,
+      "step": 224
+    },
+    {
+      "epoch": 0.07919057421926329,
+      "grad_norm": 0.9450514912605286,
+      "learning_rate": 9.977442986107252e-05,
+      "loss": 2.7791,
+      "step": 226
+    },
+    {
+      "epoch": 0.07989137576102667,
+      "grad_norm": 0.7259206771850586,
+      "learning_rate": 9.97633010472656e-05,
+      "loss": 2.7237,
+      "step": 228
+    },
+    {
+      "epoch": 0.08059217730279007,
+      "grad_norm": 0.6595309972763062,
+      "learning_rate": 9.975190492685451e-05,
+      "loss": 2.7284,
+      "step": 230
+    },
+    {
+      "epoch": 0.08129297884455346,
+      "grad_norm": 0.7696382999420166,
+      "learning_rate": 9.974024156105422e-05,
+      "loss": 2.7631,
+      "step": 232
+    },
+    {
+      "epoch": 0.08199378038631686,
+      "grad_norm": 0.7305110096931458,
+      "learning_rate": 9.972831101251521e-05,
+      "loss": 2.7793,
+      "step": 234
+    },
+    {
+      "epoch": 0.08269458192808024,
+      "grad_norm": 0.6039514541625977,
+      "learning_rate": 9.971611334532314e-05,
+      "loss": 2.7669,
+      "step": 236
+    },
+    {
+      "epoch": 0.08339538346984364,
+      "grad_norm": 0.5824711918830872,
+      "learning_rate": 9.970364862499852e-05,
+      "loss": 2.7476,
+      "step": 238
+    },
+    {
+      "epoch": 0.08409618501160702,
+      "grad_norm": 0.6831758618354797,
+      "learning_rate": 9.969091691849637e-05,
+      "loss": 2.7098,
+      "step": 240
+    },
+    {
+      "epoch": 0.08479698655337042,
+      "grad_norm": 0.6469074487686157,
+      "learning_rate": 9.967791829420581e-05,
+      "loss": 2.7609,
+      "step": 242
+    },
+    {
+      "epoch": 0.08549778809513381,
+      "grad_norm": 0.5876832604408264,
+      "learning_rate": 9.966465282194976e-05,
+      "loss": 2.7306,
+      "step": 244
+    },
+    {
+      "epoch": 0.08619858963689721,
+      "grad_norm": 0.6310129761695862,
+      "learning_rate": 9.965112057298451e-05,
+      "loss": 2.7283,
+      "step": 246
+    },
+    {
+      "epoch": 0.08689939117866059,
+      "grad_norm": 0.6113069653511047,
+      "learning_rate": 9.963732161999935e-05,
+      "loss": 2.7274,
+      "step": 248
+    },
+    {
+      "epoch": 0.08760019272042399,
+      "grad_norm": 1.0655111074447632,
+      "learning_rate": 9.96232560371162e-05,
+      "loss": 2.7022,
+      "step": 250
+    },
+    {
+      "epoch": 0.08830099426218738,
+      "grad_norm": 0.8412613272666931,
+      "learning_rate": 9.960892389988918e-05,
+      "loss": 2.7213,
+      "step": 252
+    },
+    {
+      "epoch": 0.08900179580395078,
+      "grad_norm": 0.7329776883125305,
+      "learning_rate": 9.959432528530428e-05,
+      "loss": 2.7343,
+      "step": 254
+    },
+    {
+      "epoch": 0.08970259734571416,
+      "grad_norm": 0.702498197555542,
+      "learning_rate": 9.95794602717788e-05,
+      "loss": 2.7642,
+      "step": 256
+    },
+    {
+      "epoch": 0.09040339888747755,
+      "grad_norm": 0.6936408281326294,
+      "learning_rate": 9.95643289391611e-05,
+      "loss": 2.7081,
+      "step": 258
+    },
+    {
+      "epoch": 0.09110420042924094,
+      "grad_norm": 0.664743959903717,
+      "learning_rate": 9.954893136873005e-05,
+      "loss": 2.7054,
+      "step": 260
+    },
+    {
+      "epoch": 0.09180500197100433,
+      "grad_norm": 0.5716791152954102,
+      "learning_rate": 9.953326764319463e-05,
+      "loss": 2.6751,
+      "step": 262
+    },
+    {
+      "epoch": 0.09250580351276773,
+      "grad_norm": 0.6207195520401001,
+      "learning_rate": 9.95173378466935e-05,
+      "loss": 2.6945,
+      "step": 264
+    },
+    {
+      "epoch": 0.09320660505453111,
+      "grad_norm": 0.6572092771530151,
+      "learning_rate": 9.950114206479453e-05,
+      "loss": 2.6989,
+      "step": 266
+    },
+    {
+      "epoch": 0.09390740659629451,
+      "grad_norm": 0.7676830887794495,
+      "learning_rate": 9.948468038449435e-05,
+      "loss": 2.7613,
+      "step": 268
+    },
+    {
+      "epoch": 0.0946082081380579,
+      "grad_norm": 0.5810503959655762,
+      "learning_rate": 9.946795289421787e-05,
+      "loss": 2.7234,
+      "step": 270
+    },
+    {
+      "epoch": 0.0953090096798213,
+      "grad_norm": 0.6459682583808899,
+      "learning_rate": 9.945095968381784e-05,
+      "loss": 2.717,
+      "step": 272
+    },
+    {
+      "epoch": 0.09600981122158468,
+      "grad_norm": 0.6498464345932007,
+      "learning_rate": 9.94337008445743e-05,
+      "loss": 2.7389,
+      "step": 274
+    },
+    {
+      "epoch": 0.09671061276334808,
+      "grad_norm": 0.6287350654602051,
+      "learning_rate": 9.941617646919421e-05,
+      "loss": 2.681,
+      "step": 276
+    },
+    {
+      "epoch": 0.09741141430511147,
+      "grad_norm": 0.7516258955001831,
+      "learning_rate": 9.939838665181076e-05,
+      "loss": 2.6696,
+      "step": 278
+    },
+    {
+      "epoch": 0.09811221584687486,
+      "grad_norm": 0.6962350606918335,
+      "learning_rate": 9.938033148798307e-05,
+      "loss": 2.6971,
+      "step": 280
+    },
+    {
+      "epoch": 0.09881301738863825,
+      "grad_norm": 0.6605144739151001,
+      "learning_rate": 9.936201107469555e-05,
+      "loss": 2.6999,
+      "step": 282
+    },
+    {
+      "epoch": 0.09951381893040165,
+      "grad_norm": 0.5991240739822388,
+      "learning_rate": 9.93434255103574e-05,
+      "loss": 2.6936,
+      "step": 284
+    },
+    {
+      "epoch": 0.10021462047216503,
+      "grad_norm": 0.5660961866378784,
+      "learning_rate": 9.932457489480213e-05,
+      "loss": 2.686,
+      "step": 286
+    },
+    {
+      "epoch": 0.10091542201392843,
+      "grad_norm": 0.690290093421936,
+      "learning_rate": 9.930545932928698e-05,
+      "loss": 2.6809,
+      "step": 288
+    },
+    {
+      "epoch": 0.10161622355569182,
+      "grad_norm": 0.7119167447090149,
+      "learning_rate": 9.928607891649234e-05,
+      "loss": 2.7221,
+      "step": 290
+    },
+    {
+      "epoch": 0.10231702509745522,
+      "grad_norm": 0.7049365639686584,
+      "learning_rate": 9.926643376052131e-05,
+      "loss": 2.6569,
+      "step": 292
+    },
+    {
+      "epoch": 0.1030178266392186,
+      "grad_norm": 0.6691743731498718,
+      "learning_rate": 9.924652396689902e-05,
+      "loss": 2.6751,
+      "step": 294
+    },
+    {
+      "epoch": 0.103718628180982,
+      "grad_norm": 0.5533433556556702,
+      "learning_rate": 9.922634964257215e-05,
+      "loss": 2.7064,
+      "step": 296
+    },
+    {
+      "epoch": 0.10441942972274539,
+      "grad_norm": 0.6669672727584839,
+      "learning_rate": 9.920591089590831e-05,
+      "loss": 2.687,
+      "step": 298
+    },
+    {
+      "epoch": 0.10512023126450878,
+      "grad_norm": 0.8539720773696899,
+      "learning_rate": 9.918520783669549e-05,
+      "loss": 2.6968,
+      "step": 300
+    },
+    {
+      "epoch": 0.10582103280627217,
+      "grad_norm": 0.827905535697937,
+      "learning_rate": 9.916424057614142e-05,
+      "loss": 2.7339,
+      "step": 302
+    },
+    {
+      "epoch": 0.10652183434803557,
+      "grad_norm": 0.7071542143821716,
+      "learning_rate": 9.9143009226873e-05,
+      "loss": 2.67,
+      "step": 304
+    },
+    {
+      "epoch": 0.10722263588979895,
+      "grad_norm": 0.6667853593826294,
+      "learning_rate": 9.912151390293575e-05,
+      "loss": 2.7113,
+      "step": 306
+    },
+    {
+      "epoch": 0.10792343743156235,
+      "grad_norm": 0.49210044741630554,
+      "learning_rate": 9.90997547197931e-05,
+      "loss": 2.7034,
+      "step": 308
+    },
+    {
+      "epoch": 0.10862423897332574,
+      "grad_norm": 0.5823047757148743,
+      "learning_rate": 9.907773179432581e-05,
+      "loss": 2.6815,
+      "step": 310
+    },
+    {
+      "epoch": 0.10932504051508914,
+      "grad_norm": 0.5159279704093933,
+      "learning_rate": 9.905544524483138e-05,
+      "loss": 2.7055,
+      "step": 312
+    },
+    {
+      "epoch": 0.11002584205685252,
+      "grad_norm": 0.5294278264045715,
+      "learning_rate": 9.903289519102338e-05,
+      "loss": 2.6821,
+      "step": 314
+    },
+    {
+      "epoch": 0.11072664359861592,
+      "grad_norm": 0.5865507125854492,
+      "learning_rate": 9.901008175403078e-05,
+      "loss": 2.698,
+      "step": 316
+    },
+    {
+      "epoch": 0.1114274451403793,
+      "grad_norm": 0.7102755904197693,
+      "learning_rate": 9.898700505639735e-05,
+      "loss": 2.693,
+      "step": 318
+    },
+    {
+      "epoch": 0.1121282466821427,
+      "grad_norm": 0.8151699900627136,
+      "learning_rate": 9.8963665222081e-05,
+      "loss": 2.6482,
+      "step": 320
+    },
+    {
+      "epoch": 0.11282904822390609,
+      "grad_norm": 0.5769193172454834,
+      "learning_rate": 9.894006237645304e-05,
+      "loss": 2.6893,
+      "step": 322
+    },
+    {
+      "epoch": 0.11352984976566949,
+      "grad_norm": 0.6606284976005554,
+      "learning_rate": 9.891619664629762e-05,
+      "loss": 2.6859,
+      "step": 324
+    },
+    {
+      "epoch": 0.11423065130743287,
+      "grad_norm": 0.5883016586303711,
+      "learning_rate": 9.889206815981094e-05,
+      "loss": 2.6622,
+      "step": 326
+    },
+    {
+      "epoch": 0.11493145284919627,
+      "grad_norm": 0.5413339734077454,
+      "learning_rate": 9.886767704660067e-05,
+      "loss": 2.6718,
+      "step": 328
+    },
+    {
+      "epoch": 0.11563225439095966,
+      "grad_norm": 0.7391770482063293,
+      "learning_rate": 9.884302343768512e-05,
+      "loss": 2.6695,
+      "step": 330
+    },
+    {
+      "epoch": 0.11633305593272306,
+      "grad_norm": 0.7529366612434387,
+      "learning_rate": 9.881810746549267e-05,
+      "loss": 2.7341,
+      "step": 332
+    },
+    {
+      "epoch": 0.11703385747448644,
+      "grad_norm": 0.6971571445465088,
+      "learning_rate": 9.8792929263861e-05,
+      "loss": 2.6444,
+      "step": 334
+    },
+    {
+      "epoch": 0.11773465901624984,
+      "grad_norm": 0.544129490852356,
+      "learning_rate": 9.876748896803633e-05,
+      "loss": 2.7351,
+      "step": 336
+    },
+    {
+      "epoch": 0.11843546055801323,
+      "grad_norm": 0.6561135649681091,
+      "learning_rate": 9.874178671467277e-05,
+      "loss": 2.6896,
+      "step": 338
+    },
+    {
+      "epoch": 0.11913626209977662,
+      "grad_norm": 0.6607089042663574,
+      "learning_rate": 9.871582264183155e-05,
+      "loss": 2.6664,
+      "step": 340
+    },
+    {
+      "epoch": 0.11983706364154001,
+      "grad_norm": 0.6727411150932312,
+      "learning_rate": 9.868959688898023e-05,
+      "loss": 2.68,
+      "step": 342
+    },
+    {
+      "epoch": 0.12053786518330341,
+      "grad_norm": 0.5672718286514282,
+      "learning_rate": 9.86631095969921e-05,
+      "loss": 2.6639,
+      "step": 344
+    },
+    {
+      "epoch": 0.1212386667250668,
+      "grad_norm": 0.7188961505889893,
+      "learning_rate": 9.86363609081452e-05,
+      "loss": 2.6604,
+      "step": 346
+    },
+    {
+      "epoch": 0.12193946826683019,
+      "grad_norm": 0.9785953760147095,
+      "learning_rate": 9.86093509661218e-05,
+      "loss": 2.6557,
+      "step": 348
+    },
+    {
+      "epoch": 0.12264026980859358,
+      "grad_norm": 0.7856999635696411,
+      "learning_rate": 9.85820799160074e-05,
+      "loss": 2.6418,
+      "step": 350
+    },
+    {
+      "epoch": 0.12334107135035698,
+      "grad_norm": 0.5956946015357971,
+      "learning_rate": 9.855454790429015e-05,
+      "loss": 2.658,
+      "step": 352
+    },
+    {
+      "epoch": 0.12404187289212036,
+      "grad_norm": 0.6523074507713318,
+      "learning_rate": 9.852675507885991e-05,
+      "loss": 2.6743,
+      "step": 354
+    },
+    {
+      "epoch": 0.12474267443388376,
+      "grad_norm": 0.71266108751297,
+      "learning_rate": 9.849870158900753e-05,
+      "loss": 2.6805,
+      "step": 356
+    },
+    {
+      "epoch": 0.12544347597564715,
+      "grad_norm": 0.5674154162406921,
+      "learning_rate": 9.847038758542404e-05,
+      "loss": 2.6678,
+      "step": 358
+    },
+    {
+      "epoch": 0.12614427751741053,
+      "grad_norm": 0.5430511236190796,
+      "learning_rate": 9.844181322019983e-05,
+      "loss": 2.643,
+      "step": 360
+    },
+    {
+      "epoch": 0.12684507905917394,
+      "grad_norm": 0.508791983127594,
+      "learning_rate": 9.841297864682388e-05,
+      "loss": 2.6524,
+      "step": 362
+    },
+    {
+      "epoch": 0.12754588060093733,
+      "grad_norm": 0.6082713603973389,
+      "learning_rate": 9.838388402018282e-05,
+      "loss": 2.6892,
+      "step": 364
+    },
+    {
+      "epoch": 0.1282466821427007,
+      "grad_norm": 0.6065689325332642,
+      "learning_rate": 9.835452949656022e-05,
+      "loss": 2.6083,
+      "step": 366
+    },
+    {
+      "epoch": 0.1289474836844641,
+      "grad_norm": 0.5220572352409363,
+      "learning_rate": 9.83249152336357e-05,
+      "loss": 2.6573,
+      "step": 368
+    },
+    {
+      "epoch": 0.1296482852262275,
+      "grad_norm": 0.568534791469574,
+      "learning_rate": 9.829504139048406e-05,
+      "loss": 2.6266,
+      "step": 370
+    },
+    {
+      "epoch": 0.1303490867679909,
+      "grad_norm": 0.6165401339530945,
+      "learning_rate": 9.826490812757452e-05,
+      "loss": 2.6928,
+      "step": 372
+    },
+    {
+      "epoch": 0.13104988830975428,
+      "grad_norm": 0.5951835513114929,
+      "learning_rate": 9.823451560676966e-05,
+      "loss": 2.6468,
+      "step": 374
+    },
+    {
+      "epoch": 0.13175068985151767,
+      "grad_norm": 0.4942519962787628,
+      "learning_rate": 9.820386399132482e-05,
+      "loss": 2.6493,
+      "step": 376
+    },
+    {
+      "epoch": 0.13245149139328108,
+      "grad_norm": 0.6185161471366882,
+      "learning_rate": 9.8172953445887e-05,
+      "loss": 2.6741,
+      "step": 378
+    },
+    {
+      "epoch": 0.13315229293504446,
+      "grad_norm": 0.5588895678520203,
+      "learning_rate": 9.814178413649407e-05,
+      "loss": 2.6393,
+      "step": 380
+    },
+    {
+      "epoch": 0.13385309447680785,
+      "grad_norm": 0.6289598941802979,
+      "learning_rate": 9.811035623057387e-05,
+      "loss": 2.6022,
+      "step": 382
+    },
+    {
+      "epoch": 0.13455389601857123,
+      "grad_norm": 0.6258370280265808,
+      "learning_rate": 9.807866989694334e-05,
+      "loss": 2.6033,
+      "step": 384
+    },
+    {
+      "epoch": 0.13525469756033462,
+      "grad_norm": 0.6390899419784546,
+      "learning_rate": 9.804672530580754e-05,
+      "loss": 2.6413,
+      "step": 386
+    },
+    {
+      "epoch": 0.13595549910209803,
+      "grad_norm": 0.6844115257263184,
+      "learning_rate": 9.801452262875877e-05,
+      "loss": 2.6339,
+      "step": 388
+    },
+    {
+      "epoch": 0.13665630064386142,
+      "grad_norm": 0.70540452003479,
+      "learning_rate": 9.798206203877569e-05,
+      "loss": 2.6471,
+      "step": 390
+    },
+    {
+      "epoch": 0.1373571021856248,
+      "grad_norm": 0.7336652278900146,
+      "learning_rate": 9.794934371022233e-05,
+      "loss": 2.6348,
+      "step": 392
+    },
+    {
+      "epoch": 0.1380579037273882,
+      "grad_norm": 0.7155029773712158,
+      "learning_rate": 9.79163678188472e-05,
+      "loss": 2.6128,
+      "step": 394
+    },
+    {
+      "epoch": 0.1387587052691516,
+      "grad_norm": 0.6354189515113831,
+      "learning_rate": 9.788313454178228e-05,
+      "loss": 2.6281,
+      "step": 396
+    },
+    {
+      "epoch": 0.13945950681091498,
+      "grad_norm": 0.596047043800354,
+      "learning_rate": 9.78496440575422e-05,
+      "loss": 2.6719,
+      "step": 398
+    },
+    {
+      "epoch": 0.14016030835267837,
+      "grad_norm": 0.6149719953536987,
+      "learning_rate": 9.781589654602306e-05,
+      "loss": 2.625,
+      "step": 400
+    },
+    {
+      "epoch": 0.14086110989444176,
+      "grad_norm": 0.6066911816596985,
+      "learning_rate": 9.778189218850174e-05,
+      "loss": 2.6193,
+      "step": 402
+    },
+    {
+      "epoch": 0.14156191143620517,
+      "grad_norm": 0.5690994262695312,
+      "learning_rate": 9.774763116763466e-05,
+      "loss": 2.6239,
+      "step": 404
+    },
+    {
+      "epoch": 0.14226271297796855,
+      "grad_norm": 0.532486081123352,
+      "learning_rate": 9.771311366745703e-05,
+      "loss": 2.6264,
+      "step": 406
+    },
+    {
+      "epoch": 0.14296351451973194,
+      "grad_norm": 0.5434598326683044,
+      "learning_rate": 9.767833987338171e-05,
+      "loss": 2.6534,
+      "step": 408
+    },
+    {
+      "epoch": 0.14366431606149532,
+      "grad_norm": 0.522413432598114,
+      "learning_rate": 9.764330997219822e-05,
+      "loss": 2.6468,
+      "step": 410
+    },
+    {
+      "epoch": 0.14436511760325874,
+      "grad_norm": 0.5612457990646362,
+      "learning_rate": 9.760802415207181e-05,
+      "loss": 2.6307,
+      "step": 412
+    },
+    {
+      "epoch": 0.14506591914502212,
+      "grad_norm": 0.5850318670272827,
+      "learning_rate": 9.757248260254244e-05,
+      "loss": 2.6324,
+      "step": 414
+    },
+    {
+      "epoch": 0.1457667206867855,
+      "grad_norm": 0.688555121421814,
+      "learning_rate": 9.753668551452368e-05,
+      "loss": 2.6066,
+      "step": 416
+    },
+    {
+      "epoch": 0.1464675222285489,
+      "grad_norm": 0.6506465077400208,
+      "learning_rate": 9.750063308030179e-05,
+      "loss": 2.5964,
+      "step": 418
+    },
+    {
+      "epoch": 0.1471683237703123,
+      "grad_norm": 0.6529019474983215,
+      "learning_rate": 9.746432549353462e-05,
+      "loss": 2.651,
+      "step": 420
+    },
+    {
+      "epoch": 0.1478691253120757,
+      "grad_norm": 0.5469995141029358,
+      "learning_rate": 9.742776294925058e-05,
+      "loss": 2.6129,
+      "step": 422
+    },
+    {
+      "epoch": 0.14856992685383907,
+      "grad_norm": 0.4992043673992157,
+      "learning_rate": 9.739094564384758e-05,
+      "loss": 2.6074,
+      "step": 424
+    },
+    {
+      "epoch": 0.14927072839560246,
+      "grad_norm": 0.5064156651496887,
+      "learning_rate": 9.735387377509206e-05,
+      "loss": 2.6408,
+      "step": 426
+    },
+    {
+      "epoch": 0.14997152993736587,
+      "grad_norm": 0.5961376428604126,
+      "learning_rate": 9.731654754211781e-05,
+      "loss": 2.615,
+      "step": 428
+    },
+    {
+      "epoch": 0.15067233147912926,
+      "grad_norm": 0.5533669590950012,
+      "learning_rate": 9.727896714542494e-05,
+      "loss": 2.6225,
+      "step": 430
+    },
+    {
+      "epoch": 0.15137313302089264,
+      "grad_norm": 0.5527905821800232,
+      "learning_rate": 9.724113278687888e-05,
+      "loss": 2.5836,
+      "step": 432
+    },
+    {
+      "epoch": 0.15207393456265603,
+      "grad_norm": 0.4616098701953888,
+      "learning_rate": 9.720304466970916e-05,
+      "loss": 2.6236,
+      "step": 434
+    },
+    {
+      "epoch": 0.15277473610441944,
+      "grad_norm": 0.5189539790153503,
+      "learning_rate": 9.716470299850844e-05,
+      "loss": 2.6364,
+      "step": 436
+    },
+    {
+      "epoch": 0.15347553764618282,
+      "grad_norm": 0.5303817987442017,
+      "learning_rate": 9.712610797923133e-05,
+      "loss": 2.6097,
+      "step": 438
+    },
+    {
+      "epoch": 0.1541763391879462,
+      "grad_norm": 0.5957894921302795,
+      "learning_rate": 9.708725981919333e-05,
+      "loss": 2.5749,
+      "step": 440
+    },
+    {
+      "epoch": 0.1548771407297096,
+      "grad_norm": 0.5686895251274109,
+      "learning_rate": 9.704815872706972e-05,
+      "loss": 2.6319,
+      "step": 442
+    },
+    {
+      "epoch": 0.155577942271473,
+      "grad_norm": 0.5570897459983826,
+      "learning_rate": 9.700880491289438e-05,
+      "loss": 2.6287,
+      "step": 444
+    },
+    {
+      "epoch": 0.1562787438132364,
+      "grad_norm": 0.5330969095230103,
+      "learning_rate": 9.696919858805873e-05,
+      "loss": 2.6014,
+      "step": 446
+    },
+    {
+      "epoch": 0.15697954535499978,
+      "grad_norm": 0.4891030192375183,
+      "learning_rate": 9.692933996531053e-05,
+      "loss": 2.6097,
+      "step": 448
+    },
+    {
+      "epoch": 0.15768034689676316,
+      "grad_norm": 0.5465073585510254,
+      "learning_rate": 9.688922925875285e-05,
+      "loss": 2.6162,
+      "step": 450
+    },
+    {
+      "epoch": 0.15838114843852658,
+      "grad_norm": 0.5483290553092957,
+      "learning_rate": 9.684886668384277e-05,
+      "loss": 2.5999,
+      "step": 452
+    },
+    {
+      "epoch": 0.15908194998028996,
+      "grad_norm": 0.6061928868293762,
+      "learning_rate": 9.68082524573903e-05,
+      "loss": 2.6614,
+      "step": 454
+    },
+    {
+      "epoch": 0.15978275152205335,
+      "grad_norm": 0.5806353688240051,
+      "learning_rate": 9.676738679755726e-05,
+      "loss": 2.6039,
+      "step": 456
+    },
+    {
+      "epoch": 0.16048355306381673,
+      "grad_norm": 0.5722226500511169,
+      "learning_rate": 9.672626992385602e-05,
+      "loss": 2.6529,
+      "step": 458
+    },
+    {
+      "epoch": 0.16118435460558014,
+      "grad_norm": 0.5939204096794128,
+      "learning_rate": 9.668490205714839e-05,
+      "loss": 2.6314,
+      "step": 460
+    },
+    {
+      "epoch": 0.16188515614734353,
+      "grad_norm": 0.7260386943817139,
+      "learning_rate": 9.664328341964436e-05,
+      "loss": 2.6211,
+      "step": 462
+    },
+    {
+      "epoch": 0.1625859576891069,
+      "grad_norm": 0.8503554463386536,
+      "learning_rate": 9.6601414234901e-05,
+      "loss": 2.6134,
+      "step": 464
+    },
+    {
+      "epoch": 0.1632867592308703,
+      "grad_norm": 0.5818518996238708,
+      "learning_rate": 9.655929472782116e-05,
+      "loss": 2.5667,
+      "step": 466
+    },
+    {
+      "epoch": 0.1639875607726337,
+      "grad_norm": 0.5678598284721375,
+      "learning_rate": 9.651692512465239e-05,
+      "loss": 2.6153,
+      "step": 468
+    },
+    {
+      "epoch": 0.1646883623143971,
+      "grad_norm": 0.5939005613327026,
+      "learning_rate": 9.647430565298555e-05,
+      "loss": 2.6098,
+      "step": 470
+    },
+    {
+      "epoch": 0.16538916385616048,
+      "grad_norm": 0.5300047993659973,
+      "learning_rate": 9.643143654175373e-05,
+      "loss": 2.6167,
+      "step": 472
+    },
+    {
+      "epoch": 0.16608996539792387,
+      "grad_norm": 0.4946250319480896,
+      "learning_rate": 9.638831802123101e-05,
+      "loss": 2.581,
+      "step": 474
+    },
+    {
+      "epoch": 0.16679076693968728,
+      "grad_norm": 0.4555206000804901,
+      "learning_rate": 9.634495032303111e-05,
+      "loss": 2.588,
+      "step": 476
+    },
+    {
+      "epoch": 0.16749156848145066,
+      "grad_norm": 0.5159677267074585,
+      "learning_rate": 9.630133368010628e-05,
+      "loss": 2.5868,
+      "step": 478
+    },
+    {
+      "epoch": 0.16819237002321405,
+      "grad_norm": 0.5565433502197266,
+      "learning_rate": 9.625746832674597e-05,
+      "loss": 2.6185,
+      "step": 480
+    },
+    {
+      "epoch": 0.16889317156497743,
+      "grad_norm": 0.4775915741920471,
+      "learning_rate": 9.621335449857562e-05,
+      "loss": 2.5897,
+      "step": 482
+    },
+    {
+      "epoch": 0.16959397310674085,
+      "grad_norm": 0.5150102376937866,
+      "learning_rate": 9.616899243255532e-05,
+      "loss": 2.5478,
+      "step": 484
+    },
+    {
+      "epoch": 0.17029477464850423,
+      "grad_norm": 0.48455357551574707,
+      "learning_rate": 9.612438236697863e-05,
+      "loss": 2.5639,
+      "step": 486
+    },
+    {
+      "epoch": 0.17099557619026762,
+      "grad_norm": 0.5149878859519958,
+      "learning_rate": 9.607952454147121e-05,
+      "loss": 2.599,
+      "step": 488
+    },
+    {
+      "epoch": 0.171696377732031,
+      "grad_norm": 0.6969982385635376,
+      "learning_rate": 9.603441919698963e-05,
+      "loss": 2.5733,
+      "step": 490
+    },
+    {
+      "epoch": 0.17239717927379442,
+      "grad_norm": 0.57285475730896,
+      "learning_rate": 9.598906657582e-05,
+      "loss": 2.5791,
+      "step": 492
+    },
+    {
+      "epoch": 0.1730979808155578,
+      "grad_norm": 0.5704159140586853,
+      "learning_rate": 9.594346692157667e-05,
+      "loss": 2.5692,
+      "step": 494
+    },
+    {
+      "epoch": 0.17379878235732119,
+      "grad_norm": 0.681797444820404,
+      "learning_rate": 9.589762047920096e-05,
+      "loss": 2.5759,
+      "step": 496
+    },
+    {
+      "epoch": 0.17449958389908457,
+      "grad_norm": 0.49717003107070923,
+      "learning_rate": 9.585152749495984e-05,
+      "loss": 2.5848,
+      "step": 498
+    },
+    {
+      "epoch": 0.17520038544084798,
+      "grad_norm": 0.48680582642555237,
+      "learning_rate": 9.580518821644457e-05,
+      "loss": 2.5682,
+      "step": 500
+    },
+    {
+      "epoch": 0.17590118698261137,
+      "grad_norm": 0.5525830388069153,
+      "learning_rate": 9.575860289256943e-05,
+      "loss": 2.5894,
+      "step": 502
+    },
+    {
+      "epoch": 0.17660198852437475,
+      "grad_norm": 0.5562606453895569,
+      "learning_rate": 9.571177177357032e-05,
+      "loss": 2.5675,
+      "step": 504
+    },
+    {
+      "epoch": 0.17730279006613814,
+      "grad_norm": 0.5515877604484558,
+      "learning_rate": 9.566469511100345e-05,
+      "loss": 2.5877,
+      "step": 506
+    },
+    {
+      "epoch": 0.17800359160790155,
+      "grad_norm": 0.6816357970237732,
+      "learning_rate": 9.561737315774398e-05,
+      "loss": 2.596,
+      "step": 508
+    },
+    {
+      "epoch": 0.17870439314966494,
+      "grad_norm": 0.507437527179718,
+      "learning_rate": 9.556980616798463e-05,
+      "loss": 2.5721,
+      "step": 510
+    },
+    {
+      "epoch": 0.17940519469142832,
+      "grad_norm": 0.5275202989578247,
+      "learning_rate": 9.552199439723443e-05,
+      "loss": 2.568,
+      "step": 512
+    },
+    {
+      "epoch": 0.1801059962331917,
+      "grad_norm": 0.5467104911804199,
+      "learning_rate": 9.547393810231722e-05,
+      "loss": 2.5842,
+      "step": 514
+    },
+    {
+      "epoch": 0.1808067977749551,
+      "grad_norm": 0.5407027006149292,
+      "learning_rate": 9.542563754137031e-05,
+      "loss": 2.5891,
+      "step": 516
+    },
+    {
+      "epoch": 0.1815075993167185,
+      "grad_norm": 0.5731847882270813,
+      "learning_rate": 9.537709297384308e-05,
+      "loss": 2.6143,
+      "step": 518
+    },
+    {
+      "epoch": 0.1822084008584819,
+      "grad_norm": 0.566457986831665,
+      "learning_rate": 9.532830466049565e-05,
+      "loss": 2.5522,
+      "step": 520
+    },
+    {
+      "epoch": 0.18290920240024527,
+      "grad_norm": 0.4899183213710785,
+      "learning_rate": 9.527927286339744e-05,
+      "loss": 2.5961,
+      "step": 522
+    },
+    {
+      "epoch": 0.18361000394200866,
+      "grad_norm": 0.4883110523223877,
+      "learning_rate": 9.52299978459257e-05,
+      "loss": 2.5557,
+      "step": 524
+    },
+    {
+      "epoch": 0.18431080548377207,
+      "grad_norm": 0.5534235239028931,
+      "learning_rate": 9.518047987276421e-05,
+      "loss": 2.6452,
+      "step": 526
+    },
+    {
+      "epoch": 0.18501160702553546,
+      "grad_norm": 0.47292667627334595,
+      "learning_rate": 9.513071920990179e-05,
+      "loss": 2.5848,
+      "step": 528
+    },
+    {
+      "epoch": 0.18571240856729884,
+      "grad_norm": 0.5438964366912842,
+      "learning_rate": 9.508071612463086e-05,
+      "loss": 2.5332,
+      "step": 530
+    },
+    {
+      "epoch": 0.18641321010906223,
+      "grad_norm": 0.5318060517311096,
+      "learning_rate": 9.503047088554601e-05,
+      "loss": 2.585,
+      "step": 532
+    },
+    {
+      "epoch": 0.18711401165082564,
+      "grad_norm": 0.49279502034187317,
+      "learning_rate": 9.497998376254267e-05,
+      "loss": 2.5948,
+      "step": 534
+    },
+    {
+      "epoch": 0.18781481319258903,
+      "grad_norm": 0.5161717534065247,
+      "learning_rate": 9.492925502681545e-05,
+      "loss": 2.5644,
+      "step": 536
+    },
+    {
+      "epoch": 0.1885156147343524,
+      "grad_norm": 0.4586479663848877,
+      "learning_rate": 9.487828495085684e-05,
+      "loss": 2.5568,
+      "step": 538
+    },
+    {
+      "epoch": 0.1892164162761158,
+      "grad_norm": 0.4390322268009186,
+      "learning_rate": 9.482707380845573e-05,
+      "loss": 2.5938,
+      "step": 540
+    },
+    {
+      "epoch": 0.1899172178178792,
+      "grad_norm": 0.5253728628158569,
+      "learning_rate": 9.47756218746959e-05,
+      "loss": 2.5996,
+      "step": 542
+    },
+    {
+      "epoch": 0.1906180193596426,
+      "grad_norm": 0.4567623436450958,
+      "learning_rate": 9.472392942595454e-05,
+      "loss": 2.5576,
+      "step": 544
+    },
+    {
+      "epoch": 0.19131882090140598,
+      "grad_norm": 0.5091727375984192,
+      "learning_rate": 9.467199673990077e-05,
+      "loss": 2.5873,
+      "step": 546
+    },
+    {
+      "epoch": 0.19201962244316936,
+      "grad_norm": 0.4959392845630646,
+      "learning_rate": 9.46198240954942e-05,
+      "loss": 2.5291,
+      "step": 548
+    },
+    {
+      "epoch": 0.19272042398493278,
+      "grad_norm": 0.5150632262229919,
+      "learning_rate": 9.456741177298336e-05,
+      "loss": 2.5503,
+      "step": 550
+    },
+    {
+      "epoch": 0.19342122552669616,
+      "grad_norm": 0.4603368639945984,
+      "learning_rate": 9.451476005390422e-05,
+      "loss": 2.5785,
+      "step": 552
+    },
+    {
+      "epoch": 0.19412202706845955,
+      "grad_norm": 0.4441729784011841,
+      "learning_rate": 9.446186922107873e-05,
+      "loss": 2.5512,
+      "step": 554
+    },
+    {
+      "epoch": 0.19482282861022293,
+      "grad_norm": 0.5432455539703369,
+      "learning_rate": 9.44087395586132e-05,
+      "loss": 2.5741,
+      "step": 556
+    },
+    {
+      "epoch": 0.19552363015198634,
+      "grad_norm": 0.42969366908073425,
+      "learning_rate": 9.435537135189687e-05,
+      "loss": 2.5677,
+      "step": 558
+    },
+    {
+      "epoch": 0.19622443169374973,
+      "grad_norm": 0.5706619620323181,
+      "learning_rate": 9.430176488760027e-05,
+      "loss": 2.556,
+      "step": 560
+    },
+    {
+      "epoch": 0.19692523323551311,
+      "grad_norm": 0.7202513217926025,
+      "learning_rate": 9.424792045367383e-05,
+      "loss": 2.5435,
+      "step": 562
+    },
+    {
+      "epoch": 0.1976260347772765,
+      "grad_norm": 0.5471363663673401,
+      "learning_rate": 9.419383833934621e-05,
+      "loss": 2.572,
+      "step": 564
+    },
+    {
+      "epoch": 0.1983268363190399,
+      "grad_norm": 0.654058575630188,
+      "learning_rate": 9.413951883512275e-05,
+      "loss": 2.5432,
+      "step": 566
+    },
+    {
+      "epoch": 0.1990276378608033,
+      "grad_norm": 0.6124361157417297,
+      "learning_rate": 9.408496223278403e-05,
+      "loss": 2.5803,
+      "step": 568
+    },
+    {
+      "epoch": 0.19972843940256668,
+      "grad_norm": 0.5291132926940918,
+      "learning_rate": 9.403016882538408e-05,
+      "loss": 2.576,
+      "step": 570
+    },
+    {
+      "epoch": 0.20042924094433007,
+      "grad_norm": 0.6087374687194824,
+      "learning_rate": 9.397513890724911e-05,
+      "loss": 2.5171,
+      "step": 572
+    },
+    {
+      "epoch": 0.20113004248609348,
+      "grad_norm": 0.5776922106742859,
+      "learning_rate": 9.391987277397566e-05,
+      "loss": 2.6054,
+      "step": 574
+    },
+    {
+      "epoch": 0.20183084402785686,
+      "grad_norm": 0.544319748878479,
+      "learning_rate": 9.38643707224291e-05,
+      "loss": 2.548,
+      "step": 576
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.5210007429122925,
+      "learning_rate": 9.38086330507421e-05,
+      "loss": 2.6019,
+      "step": 578
+    },
+    {
+      "epoch": 0.20323244711138364,
+      "grad_norm": 0.5160629153251648,
+      "learning_rate": 9.375266005831297e-05,
+      "loss": 2.6046,
+      "step": 580
+    },
+    {
+      "epoch": 0.20393324865314705,
+      "grad_norm": 0.6452796459197998,
+      "learning_rate": 9.369645204580403e-05,
+      "loss": 2.566,
+      "step": 582
+    },
+    {
+      "epoch": 0.20463405019491043,
+      "grad_norm": 0.5813329815864563,
+      "learning_rate": 9.364000931514008e-05,
+      "loss": 2.5661,
+      "step": 584
+    },
+    {
+      "epoch": 0.20533485173667382,
+      "grad_norm": 0.5450593829154968,
+      "learning_rate": 9.358333216950664e-05,
+      "loss": 2.5769,
+      "step": 586
+    },
+    {
+      "epoch": 0.2060356532784372,
+      "grad_norm": 0.5340794324874878,
+      "learning_rate": 9.352642091334849e-05,
+      "loss": 2.5549,
+      "step": 588
+    },
+    {
+      "epoch": 0.20673645482020062,
+      "grad_norm": 0.5767348408699036,
+      "learning_rate": 9.34692758523679e-05,
+      "loss": 2.5604,
+      "step": 590
+    },
+    {
+      "epoch": 0.207437256361964,
+      "grad_norm": 0.6048093438148499,
+      "learning_rate": 9.341189729352302e-05,
+      "loss": 2.5929,
+      "step": 592
+    },
+    {
+      "epoch": 0.20813805790372739,
+      "grad_norm": 0.4430505335330963,
+      "learning_rate": 9.33542855450263e-05,
+      "loss": 2.5563,
+      "step": 594
+    },
+    {
+      "epoch": 0.20883885944549077,
+      "grad_norm": 0.49373888969421387,
+      "learning_rate": 9.329644091634278e-05,
+      "loss": 2.5517,
+      "step": 596
+    },
+    {
+      "epoch": 0.20953966098725418,
+      "grad_norm": 0.5227393507957458,
+      "learning_rate": 9.323836371818837e-05,
+      "loss": 2.5286,
+      "step": 598
+    },
+    {
+      "epoch": 0.21024046252901757,
+      "grad_norm": 0.497405081987381,
+      "learning_rate": 9.318005426252832e-05,
+      "loss": 2.5638,
+      "step": 600
+    },
+    {
+      "epoch": 0.21094126407078095,
+      "grad_norm": 0.48721396923065186,
+      "learning_rate": 9.312151286257537e-05,
+      "loss": 2.5751,
+      "step": 602
+    },
+    {
+      "epoch": 0.21164206561254434,
+      "grad_norm": 0.4621741771697998,
+      "learning_rate": 9.306273983278825e-05,
+      "loss": 2.5654,
+      "step": 604
+    },
+    {
+      "epoch": 0.21234286715430775,
+      "grad_norm": 0.4756307005882263,
+      "learning_rate": 9.300373548886987e-05,
+      "loss": 2.5989,
+      "step": 606
+    },
+    {
+      "epoch": 0.21304366869607114,
+      "grad_norm": 0.42497771978378296,
+      "learning_rate": 9.294450014776566e-05,
+      "loss": 2.564,
+      "step": 608
+    },
+    {
+      "epoch": 0.21374447023783452,
+      "grad_norm": 0.5173219442367554,
+      "learning_rate": 9.288503412766185e-05,
+      "loss": 2.5296,
+      "step": 610
+    },
+    {
+      "epoch": 0.2144452717795979,
+      "grad_norm": 0.4622451066970825,
+      "learning_rate": 9.28253377479838e-05,
+      "loss": 2.5829,
+      "step": 612
+    },
+    {
+      "epoch": 0.21514607332136132,
+      "grad_norm": 0.5879294276237488,
+      "learning_rate": 9.276541132939428e-05,
+      "loss": 2.5462,
+      "step": 614
+    },
+    {
+      "epoch": 0.2158468748631247,
+      "grad_norm": 0.6237635612487793,
+      "learning_rate": 9.270525519379165e-05,
+      "loss": 2.6143,
+      "step": 616
+    },
+    {
+      "epoch": 0.2165476764048881,
+      "grad_norm": 0.5845280289649963,
+      "learning_rate": 9.264486966430829e-05,
+      "loss": 2.5272,
+      "step": 618
+    },
+    {
+      "epoch": 0.21724847794665147,
+      "grad_norm": 0.5140432715415955,
+      "learning_rate": 9.258425506530872e-05,
+      "loss": 2.5716,
+      "step": 620
+    },
+    {
+      "epoch": 0.2179492794884149,
+      "grad_norm": 0.5868300199508667,
+      "learning_rate": 9.2523411722388e-05,
+      "loss": 2.5699,
+      "step": 622
+    },
+    {
+      "epoch": 0.21865008103017827,
+      "grad_norm": 0.587374210357666,
+      "learning_rate": 9.246233996236983e-05,
+      "loss": 2.5335,
+      "step": 624
+    },
+    {
+      "epoch": 0.21935088257194166,
+      "grad_norm": 0.5000743865966797,
+      "learning_rate": 9.240104011330489e-05,
+      "loss": 2.5367,
+      "step": 626
+    },
+    {
+      "epoch": 0.22005168411370504,
+      "grad_norm": 0.5124289393424988,
+      "learning_rate": 9.233951250446902e-05,
+      "loss": 2.5598,
+      "step": 628
+    },
+    {
+      "epoch": 0.22075248565546846,
+      "grad_norm": 0.4815032482147217,
+      "learning_rate": 9.227775746636158e-05,
+      "loss": 2.5468,
+      "step": 630
+    },
+    {
+      "epoch": 0.22145328719723184,
+      "grad_norm": 0.5089353919029236,
+      "learning_rate": 9.22157753307035e-05,
+      "loss": 2.5482,
+      "step": 632
+    },
+    {
+      "epoch": 0.22215408873899523,
+      "grad_norm": 0.468841552734375,
+      "learning_rate": 9.215356643043559e-05,
+      "loss": 2.5138,
+      "step": 634
+    },
+    {
+      "epoch": 0.2228548902807586,
+      "grad_norm": 0.511968731880188,
+      "learning_rate": 9.209113109971676e-05,
+      "loss": 2.5481,
+      "step": 636
+    },
+    {
+      "epoch": 0.223555691822522,
+      "grad_norm": 0.6082082390785217,
+      "learning_rate": 9.202846967392217e-05,
+      "loss": 2.5459,
+      "step": 638
+    },
+    {
+      "epoch": 0.2242564933642854,
+      "grad_norm": 0.4931623637676239,
+      "learning_rate": 9.196558248964151e-05,
+      "loss": 2.5785,
+      "step": 640
+    },
+    {
+      "epoch": 0.2249572949060488,
+      "grad_norm": 0.5754916071891785,
+      "learning_rate": 9.190246988467712e-05,
+      "loss": 2.5166,
+      "step": 642
+    },
+    {
+      "epoch": 0.22565809644781218,
+      "grad_norm": 0.5335285067558289,
+      "learning_rate": 9.183913219804221e-05,
+      "loss": 2.4976,
+      "step": 644
+    },
+    {
+      "epoch": 0.22635889798957556,
+      "grad_norm": 0.4676333963871002,
+      "learning_rate": 9.1775569769959e-05,
+      "loss": 2.5361,
+      "step": 646
+    },
+    {
+      "epoch": 0.22705969953133898,
+      "grad_norm": 0.48826783895492554,
+      "learning_rate": 9.171178294185697e-05,
+      "loss": 2.5347,
+      "step": 648
+    },
+    {
+      "epoch": 0.22776050107310236,
+      "grad_norm": 0.509066104888916,
+      "learning_rate": 9.164777205637094e-05,
+      "loss": 2.5326,
+      "step": 650
+    },
+    {
+      "epoch": 0.22846130261486575,
+      "grad_norm": 0.5001896619796753,
+      "learning_rate": 9.158353745733927e-05,
+      "loss": 2.5605,
+      "step": 652
+    },
+    {
+      "epoch": 0.22916210415662913,
+      "grad_norm": 0.5497420430183411,
+      "learning_rate": 9.151907948980206e-05,
+      "loss": 2.5295,
+      "step": 654
+    },
+    {
+      "epoch": 0.22986290569839254,
+      "grad_norm": 0.43462875485420227,
+      "learning_rate": 9.145439849999919e-05,
+      "loss": 2.5358,
+      "step": 656
+    },
+    {
+      "epoch": 0.23056370724015593,
+      "grad_norm": 0.5398270487785339,
+      "learning_rate": 9.138949483536852e-05,
+      "loss": 2.5464,
+      "step": 658
+    },
+    {
+      "epoch": 0.23126450878191931,
+      "grad_norm": 0.5165109038352966,
+      "learning_rate": 9.132436884454408e-05,
+      "loss": 2.5043,
+      "step": 660
+    },
+    {
+      "epoch": 0.2319653103236827,
+      "grad_norm": 0.6717212200164795,
+      "learning_rate": 9.125902087735407e-05,
+      "loss": 2.547,
+      "step": 662
+    },
+    {
+      "epoch": 0.2326661118654461,
+      "grad_norm": 0.4584912061691284,
+      "learning_rate": 9.119345128481909e-05,
+      "loss": 2.5106,
+      "step": 664
+    },
+    {
+      "epoch": 0.2333669134072095,
+      "grad_norm": 0.5452204942703247,
+      "learning_rate": 9.112766041915019e-05,
+      "loss": 2.5189,
+      "step": 666
+    },
+    {
+      "epoch": 0.23406771494897288,
+      "grad_norm": 0.5055968761444092,
+      "learning_rate": 9.106164863374702e-05,
+      "loss": 2.4957,
+      "step": 668
+    },
+    {
+      "epoch": 0.23476851649073627,
+      "grad_norm": 0.4905461072921753,
+      "learning_rate": 9.099541628319592e-05,
+      "loss": 2.5523,
+      "step": 670
+    },
+    {
+      "epoch": 0.23546931803249968,
+      "grad_norm": 0.44840848445892334,
+      "learning_rate": 9.092896372326798e-05,
+      "loss": 2.4713,
+      "step": 672
+    },
+    {
+      "epoch": 0.23617011957426307,
+      "grad_norm": 0.46489134430885315,
+      "learning_rate": 9.086229131091717e-05,
+      "loss": 2.5071,
+      "step": 674
+    },
+    {
+      "epoch": 0.23687092111602645,
+      "grad_norm": 0.4460737705230713,
+      "learning_rate": 9.079539940427845e-05,
+      "loss": 2.5799,
+      "step": 676
+    },
+    {
+      "epoch": 0.23757172265778984,
+      "grad_norm": 0.5268511176109314,
+      "learning_rate": 9.072828836266574e-05,
+      "loss": 2.5574,
+      "step": 678
+    },
+    {
+      "epoch": 0.23827252419955325,
+      "grad_norm": 0.5001477003097534,
+      "learning_rate": 9.066095854657011e-05,
+      "loss": 2.5117,
+      "step": 680
+    },
+    {
+      "epoch": 0.23897332574131663,
+      "grad_norm": 0.5136899352073669,
+      "learning_rate": 9.059341031765773e-05,
+      "loss": 2.4855,
+      "step": 682
+    },
+    {
+      "epoch": 0.23967412728308002,
+      "grad_norm": 0.5532418489456177,
+      "learning_rate": 9.052564403876808e-05,
+      "loss": 2.5623,
+      "step": 684
+    },
+    {
+      "epoch": 0.2403749288248434,
+      "grad_norm": 0.4908037483692169,
+      "learning_rate": 9.045766007391185e-05,
+      "loss": 2.5248,
+      "step": 686
+    },
+    {
+      "epoch": 0.24107573036660682,
+      "grad_norm": 0.45994317531585693,
+      "learning_rate": 9.038945878826903e-05,
+      "loss": 2.5007,
+      "step": 688
+    },
+    {
+      "epoch": 0.2417765319083702,
+      "grad_norm": 0.5593565702438354,
+      "learning_rate": 9.032104054818698e-05,
+      "loss": 2.5759,
+      "step": 690
+    },
+    {
+      "epoch": 0.2424773334501336,
+      "grad_norm": 0.5076695084571838,
+      "learning_rate": 9.025240572117846e-05,
+      "loss": 2.5272,
+      "step": 692
+    },
+    {
+      "epoch": 0.24317813499189697,
+      "grad_norm": 0.3996141850948334,
+      "learning_rate": 9.018355467591962e-05,
+      "loss": 2.5317,
+      "step": 694
+    },
+    {
+      "epoch": 0.24387893653366038,
+      "grad_norm": 0.49347859621047974,
+      "learning_rate": 9.011448778224802e-05,
+      "loss": 2.5186,
+      "step": 696
+    },
+    {
+      "epoch": 0.24457973807542377,
+      "grad_norm": 0.5040503144264221,
+      "learning_rate": 9.004520541116075e-05,
+      "loss": 2.5015,
+      "step": 698
+    },
+    {
+      "epoch": 0.24528053961718715,
+      "grad_norm": 0.4658913016319275,
+      "learning_rate": 8.997570793481223e-05,
+      "loss": 2.5481,
+      "step": 700
+    },
+    {
+      "epoch": 0.24598134115895054,
+      "grad_norm": 0.47850051522254944,
+      "learning_rate": 8.990599572651242e-05,
+      "loss": 2.5505,
+      "step": 702
+    },
+    {
+      "epoch": 0.24668214270071395,
+      "grad_norm": 0.48090964555740356,
+      "learning_rate": 8.983606916072469e-05,
+      "loss": 2.5669,
+      "step": 704
+    },
+    {
+      "epoch": 0.24738294424247734,
+      "grad_norm": 0.5716775059700012,
+      "learning_rate": 8.976592861306384e-05,
+      "loss": 2.523,
+      "step": 706
+    },
+    {
+      "epoch": 0.24808374578424072,
+      "grad_norm": 0.49985334277153015,
+      "learning_rate": 8.969557446029409e-05,
+      "loss": 2.5439,
+      "step": 708
+    },
+    {
+      "epoch": 0.2487845473260041,
+      "grad_norm": 0.6331408023834229,
+      "learning_rate": 8.962500708032708e-05,
+      "loss": 2.5601,
+      "step": 710
+    },
+    {
+      "epoch": 0.24948534886776752,
+      "grad_norm": 0.5418590307235718,
+      "learning_rate": 8.955422685221979e-05,
+      "loss": 2.5495,
+      "step": 712
+    },
+    {
+      "epoch": 0.2501861504095309,
+      "grad_norm": 0.5396260619163513,
+      "learning_rate": 8.948323415617253e-05,
+      "loss": 2.5151,
+      "step": 714
+    },
+    {
+      "epoch": 0.2508869519512943,
+      "grad_norm": 0.5641499161720276,
+      "learning_rate": 8.941202937352686e-05,
+      "loss": 2.4895,
+      "step": 716
+    },
+    {
+      "epoch": 0.2515877534930577,
+      "grad_norm": 0.47651517391204834,
+      "learning_rate": 8.934061288676365e-05,
+      "loss": 2.5634,
+      "step": 718
+    },
+    {
+      "epoch": 0.25228855503482106,
+      "grad_norm": 0.5351449251174927,
+      "learning_rate": 8.92689850795009e-05,
+      "loss": 2.4804,
+      "step": 720
+    },
+    {
+      "epoch": 0.25298935657658445,
+      "grad_norm": 0.5856335759162903,
+      "learning_rate": 8.919714633649172e-05,
+      "loss": 2.5304,
+      "step": 722
+    },
+    {
+      "epoch": 0.2536901581183479,
+      "grad_norm": 0.4513723850250244,
+      "learning_rate": 8.912509704362232e-05,
+      "loss": 2.5369,
+      "step": 724
+    },
+    {
+      "epoch": 0.25439095966011127,
+      "grad_norm": 0.4676707983016968,
+      "learning_rate": 8.905283758790985e-05,
+      "loss": 2.5589,
+      "step": 726
+    },
+    {
+      "epoch": 0.25509176120187466,
+      "grad_norm": 0.5069173574447632,
+      "learning_rate": 8.89803683575004e-05,
+      "loss": 2.4958,
+      "step": 728
+    },
+    {
+      "epoch": 0.25579256274363804,
+      "grad_norm": 0.4774676263332367,
+      "learning_rate": 8.890768974166685e-05,
+      "loss": 2.5229,
+      "step": 730
+    },
+    {
+      "epoch": 0.2564933642854014,
+      "grad_norm": 0.548409104347229,
+      "learning_rate": 8.883480213080681e-05,
+      "loss": 2.4815,
+      "step": 732
+    },
+    {
+      "epoch": 0.2571941658271648,
+      "grad_norm": 0.4854792356491089,
+      "learning_rate": 8.876170591644054e-05,
+      "loss": 2.5118,
+      "step": 734
+    },
+    {
+      "epoch": 0.2578949673689282,
+      "grad_norm": 0.4988788664340973,
+      "learning_rate": 8.868840149120876e-05,
+      "loss": 2.5073,
+      "step": 736
+    },
+    {
+      "epoch": 0.2585957689106916,
+      "grad_norm": 0.4614211618900299,
+      "learning_rate": 8.861488924887071e-05,
+      "loss": 2.4866,
+      "step": 738
+    },
+    {
+      "epoch": 0.259296570452455,
+      "grad_norm": 0.4878149926662445,
+      "learning_rate": 8.854116958430185e-05,
+      "loss": 2.5315,
+      "step": 740
+    },
+    {
+      "epoch": 0.2599973719942184,
+      "grad_norm": 0.47185149788856506,
+      "learning_rate": 8.846724289349189e-05,
+      "loss": 2.4766,
+      "step": 742
+    },
+    {
+      "epoch": 0.2606981735359818,
+      "grad_norm": 0.446411669254303,
+      "learning_rate": 8.839310957354249e-05,
+      "loss": 2.5278,
+      "step": 744
+    },
+    {
+      "epoch": 0.2613989750777452,
+      "grad_norm": 0.45869573950767517,
+      "learning_rate": 8.831877002266536e-05,
+      "loss": 2.5051,
+      "step": 746
+    },
+    {
+      "epoch": 0.26209977661950856,
+      "grad_norm": 0.4578917920589447,
+      "learning_rate": 8.82442246401799e-05,
+      "loss": 2.4903,
+      "step": 748
+    },
+    {
+      "epoch": 0.26280057816127195,
+      "grad_norm": 0.4389136731624603,
+      "learning_rate": 8.816947382651116e-05,
+      "loss": 2.519,
+      "step": 750
+    },
+    {
+      "epoch": 0.26350137970303533,
+      "grad_norm": 0.4686265289783478,
+      "learning_rate": 8.80945179831877e-05,
+      "loss": 2.5537,
+      "step": 752
+    },
+    {
+      "epoch": 0.2642021812447987,
+      "grad_norm": 0.49357905983924866,
+      "learning_rate": 8.801935751283944e-05,
+      "loss": 2.4971,
+      "step": 754
+    },
+    {
+      "epoch": 0.26490298278656216,
+      "grad_norm": 0.5659007430076599,
+      "learning_rate": 8.794399281919537e-05,
+      "loss": 2.5291,
+      "step": 756
+    },
+    {
+      "epoch": 0.26560378432832554,
+      "grad_norm": 0.5637578964233398,
+      "learning_rate": 8.786842430708157e-05,
+      "loss": 2.5335,
+      "step": 758
+    },
+    {
+      "epoch": 0.26630458587008893,
+      "grad_norm": 0.47859886288642883,
+      "learning_rate": 8.779265238241888e-05,
+      "loss": 2.5104,
+      "step": 760
+    },
+    {
+      "epoch": 0.2670053874118523,
+      "grad_norm": 0.5444939732551575,
+      "learning_rate": 8.771667745222082e-05,
+      "loss": 2.4823,
+      "step": 762
+    },
+    {
+      "epoch": 0.2677061889536157,
+      "grad_norm": 0.5456621050834656,
+      "learning_rate": 8.76404999245914e-05,
+      "loss": 2.5027,
+      "step": 764
+    },
+    {
+      "epoch": 0.2684069904953791,
+      "grad_norm": 0.5168180465698242,
+      "learning_rate": 8.75641202087228e-05,
+      "loss": 2.5562,
+      "step": 766
+    },
+    {
+      "epoch": 0.26910779203714247,
+      "grad_norm": 0.5675712823867798,
+      "learning_rate": 8.748753871489333e-05,
+      "loss": 2.5195,
+      "step": 768
+    },
+    {
+      "epoch": 0.26980859357890585,
+      "grad_norm": 0.4084811806678772,
+      "learning_rate": 8.741075585446514e-05,
+      "loss": 2.4853,
+      "step": 770
+    },
+    {
+      "epoch": 0.27050939512066924,
+      "grad_norm": 0.4109669327735901,
+      "learning_rate": 8.733377203988208e-05,
+      "loss": 2.5186,
+      "step": 772
+    },
+    {
+      "epoch": 0.2712101966624327,
+      "grad_norm": 0.5689636468887329,
+      "learning_rate": 8.725658768466738e-05,
+      "loss": 2.5106,
+      "step": 774
+    },
+    {
+      "epoch": 0.27191099820419606,
+      "grad_norm": 0.4750414192676544,
+      "learning_rate": 8.71792032034215e-05,
+      "loss": 2.4927,
+      "step": 776
+    },
+    {
+      "epoch": 0.27261179974595945,
+      "grad_norm": 0.4577466547489166,
+      "learning_rate": 8.710161901181993e-05,
+      "loss": 2.5005,
+      "step": 778
+    },
+    {
+      "epoch": 0.27331260128772283,
+      "grad_norm": 0.4786745011806488,
+      "learning_rate": 8.702383552661081e-05,
+      "loss": 2.5099,
+      "step": 780
+    },
+    {
+      "epoch": 0.2740134028294862,
+      "grad_norm": 0.508456289768219,
+      "learning_rate": 8.694585316561296e-05,
+      "loss": 2.5377,
+      "step": 782
+    },
+    {
+      "epoch": 0.2747142043712496,
+      "grad_norm": 0.49584171175956726,
+      "learning_rate": 8.686767234771333e-05,
+      "loss": 2.5208,
+      "step": 784
+    },
+    {
+      "epoch": 0.275415005913013,
+      "grad_norm": 0.4523308575153351,
+      "learning_rate": 8.678929349286498e-05,
+      "loss": 2.5663,
+      "step": 786
+    },
+    {
+      "epoch": 0.2761158074547764,
+      "grad_norm": 0.411276638507843,
+      "learning_rate": 8.671071702208467e-05,
+      "loss": 2.5076,
+      "step": 788
+    },
+    {
+      "epoch": 0.2768166089965398,
+      "grad_norm": 0.47366130352020264,
+      "learning_rate": 8.663194335745071e-05,
+      "loss": 2.4725,
+      "step": 790
+    },
+    {
+      "epoch": 0.2775174105383032,
+      "grad_norm": 0.44845113158226013,
+      "learning_rate": 8.655297292210067e-05,
+      "loss": 2.5204,
+      "step": 792
+    },
+    {
+      "epoch": 0.2782182120800666,
+      "grad_norm": 0.4630947709083557,
+      "learning_rate": 8.647380614022902e-05,
+      "loss": 2.4848,
+      "step": 794
+    },
+    {
+      "epoch": 0.27891901362182997,
+      "grad_norm": 0.4739050567150116,
+      "learning_rate": 8.639444343708496e-05,
+      "loss": 2.4975,
+      "step": 796
+    },
+    {
+      "epoch": 0.27961981516359335,
+      "grad_norm": 0.41872844099998474,
+      "learning_rate": 8.631488523897011e-05,
+      "loss": 2.5105,
+      "step": 798
+    },
+    {
+      "epoch": 0.28032061670535674,
+      "grad_norm": 0.5174891948699951,
+      "learning_rate": 8.623513197323615e-05,
+      "loss": 2.4428,
+      "step": 800
+    },
+    {
+      "epoch": 0.2810214182471201,
+      "grad_norm": 0.4543634057044983,
+      "learning_rate": 8.615518406828262e-05,
+      "loss": 2.5248,
+      "step": 802
+    },
+    {
+      "epoch": 0.2817222197888835,
+      "grad_norm": 0.433250367641449,
+      "learning_rate": 8.607504195355458e-05,
+      "loss": 2.4887,
+      "step": 804
+    },
+    {
+      "epoch": 0.28242302133064695,
+      "grad_norm": 0.47642698884010315,
+      "learning_rate": 8.599470605954025e-05,
+      "loss": 2.5391,
+      "step": 806
+    },
+    {
+      "epoch": 0.28312382287241034,
+      "grad_norm": 0.45496654510498047,
+      "learning_rate": 8.59141768177688e-05,
+      "loss": 2.5444,
+      "step": 808
+    },
+    {
+      "epoch": 0.2838246244141737,
+      "grad_norm": 0.4619695544242859,
+      "learning_rate": 8.583345466080796e-05,
+      "loss": 2.504,
+      "step": 810
+    },
+    {
+      "epoch": 0.2845254259559371,
+      "grad_norm": 0.4610481262207031,
+      "learning_rate": 8.575254002226173e-05,
+      "loss": 2.4904,
+      "step": 812
+    },
+    {
+      "epoch": 0.2852262274977005,
+      "grad_norm": 0.4597660005092621,
+      "learning_rate": 8.5671433336768e-05,
+      "loss": 2.4923,
+      "step": 814
+    },
+    {
+      "epoch": 0.2859270290394639,
+      "grad_norm": 0.5440905094146729,
+      "learning_rate": 8.559013503999626e-05,
+      "loss": 2.4806,
+      "step": 816
+    },
+    {
+      "epoch": 0.28662783058122726,
+      "grad_norm": 0.4667718708515167,
+      "learning_rate": 8.550864556864529e-05,
+      "loss": 2.5595,
+      "step": 818
+    },
+    {
+      "epoch": 0.28732863212299065,
+      "grad_norm": 0.47145599126815796,
+      "learning_rate": 8.542696536044075e-05,
+      "loss": 2.4813,
+      "step": 820
+    },
+    {
+      "epoch": 0.2880294336647541,
+      "grad_norm": 0.4581964313983917,
+      "learning_rate": 8.534509485413284e-05,
+      "loss": 2.5467,
+      "step": 822
+    },
+    {
+      "epoch": 0.28873023520651747,
+      "grad_norm": 0.5127134919166565,
+      "learning_rate": 8.5263034489494e-05,
+      "loss": 2.5067,
+      "step": 824
+    },
+    {
+      "epoch": 0.28943103674828086,
+      "grad_norm": 0.5416949391365051,
+      "learning_rate": 8.518078470731644e-05,
+      "loss": 2.4669,
+      "step": 826
+    },
+    {
+      "epoch": 0.29013183829004424,
+      "grad_norm": 0.442828506231308,
+      "learning_rate": 8.509834594940991e-05,
+      "loss": 2.4708,
+      "step": 828
+    },
+    {
+      "epoch": 0.2908326398318076,
+      "grad_norm": 0.4708557426929474,
+      "learning_rate": 8.501571865859924e-05,
+      "loss": 2.5192,
+      "step": 830
+    },
+    {
+      "epoch": 0.291533441373571,
+      "grad_norm": 0.4371870458126068,
+      "learning_rate": 8.49329032787219e-05,
+      "loss": 2.4778,
+      "step": 832
+    },
+    {
+      "epoch": 0.2922342429153344,
+      "grad_norm": 0.48408806324005127,
+      "learning_rate": 8.48499002546258e-05,
+      "loss": 2.4868,
+      "step": 834
+    },
+    {
+      "epoch": 0.2929350444570978,
+      "grad_norm": 0.45126622915267944,
+      "learning_rate": 8.47667100321667e-05,
+      "loss": 2.4999,
+      "step": 836
+    },
+    {
+      "epoch": 0.2936358459988612,
+      "grad_norm": 0.4448654353618622,
+      "learning_rate": 8.468333305820599e-05,
+      "loss": 2.4848,
+      "step": 838
+    },
+    {
+      "epoch": 0.2943366475406246,
+      "grad_norm": 0.47776126861572266,
+      "learning_rate": 8.459976978060815e-05,
+      "loss": 2.5515,
+      "step": 840
+    },
+    {
+      "epoch": 0.295037449082388,
+      "grad_norm": 0.4572128653526306,
+      "learning_rate": 8.45160206482384e-05,
+      "loss": 2.5172,
+      "step": 842
+    },
+    {
+      "epoch": 0.2957382506241514,
+      "grad_norm": 0.4419424831867218,
+      "learning_rate": 8.443208611096036e-05,
+      "loss": 2.5035,
+      "step": 844
+    },
+    {
+      "epoch": 0.29643905216591476,
+      "grad_norm": 0.42213693261146545,
+      "learning_rate": 8.434796661963344e-05,
+      "loss": 2.542,
+      "step": 846
+    },
+    {
+      "epoch": 0.29713985370767815,
+      "grad_norm": 0.446344792842865,
+      "learning_rate": 8.426366262611067e-05,
+      "loss": 2.5119,
+      "step": 848
+    },
+    {
+      "epoch": 0.29784065524944153,
+      "grad_norm": 0.44233253598213196,
+      "learning_rate": 8.417917458323607e-05,
+      "loss": 2.4985,
+      "step": 850
+    },
+    {
+      "epoch": 0.2985414567912049,
+      "grad_norm": 0.492471843957901,
+      "learning_rate": 8.40945029448423e-05,
+      "loss": 2.4553,
+      "step": 852
+    },
+    {
+      "epoch": 0.29924225833296836,
+      "grad_norm": 0.4490063488483429,
+      "learning_rate": 8.400964816574826e-05,
+      "loss": 2.5389,
+      "step": 854
+    },
+    {
+      "epoch": 0.29994305987473174,
+      "grad_norm": 0.5494585633277893,
+      "learning_rate": 8.392461070175652e-05,
+      "loss": 2.5163,
+      "step": 856
+    },
+    {
+      "epoch": 0.30064386141649513,
+      "grad_norm": 0.4822872281074524,
+      "learning_rate": 8.383939100965103e-05,
+      "loss": 2.504,
+      "step": 858
+    },
+    {
+      "epoch": 0.3013446629582585,
+      "grad_norm": 0.5434439778327942,
+      "learning_rate": 8.375398954719456e-05,
+      "loss": 2.4841,
+      "step": 860
+    },
+    {
+      "epoch": 0.3020454645000219,
+      "grad_norm": 0.5055859088897705,
+      "learning_rate": 8.366840677312626e-05,
+      "loss": 2.4985,
+      "step": 862
+    },
+    {
+      "epoch": 0.3027462660417853,
+      "grad_norm": 0.44319674372673035,
+      "learning_rate": 8.358264314715923e-05,
+      "loss": 2.4661,
+      "step": 864
+    },
+    {
+      "epoch": 0.30344706758354867,
+      "grad_norm": 0.5121539235115051,
+      "learning_rate": 8.349669912997799e-05,
+      "loss": 2.4797,
+      "step": 866
+    },
+    {
+      "epoch": 0.30414786912531205,
+      "grad_norm": 0.4748767912387848,
+      "learning_rate": 8.341057518323607e-05,
+      "loss": 2.5009,
+      "step": 868
+    },
+    {
+      "epoch": 0.3048486706670755,
+      "grad_norm": 0.4823194742202759,
+      "learning_rate": 8.332427176955353e-05,
+      "loss": 2.4798,
+      "step": 870
+    },
+    {
+      "epoch": 0.3055494722088389,
+      "grad_norm": 0.4242302477359772,
+      "learning_rate": 8.323778935251437e-05,
+      "loss": 2.4764,
+      "step": 872
+    },
+    {
+      "epoch": 0.30625027375060226,
+      "grad_norm": 0.46324998140335083,
+      "learning_rate": 8.31511283966642e-05,
+      "loss": 2.509,
+      "step": 874
+    },
+    {
+      "epoch": 0.30695107529236565,
+      "grad_norm": 0.4894976317882538,
+      "learning_rate": 8.30642893675076e-05,
+      "loss": 2.498,
+      "step": 876
+    },
+    {
+      "epoch": 0.30765187683412903,
+      "grad_norm": 0.4574197232723236,
+      "learning_rate": 8.297727273150573e-05,
+      "loss": 2.48,
+      "step": 878
+    },
+    {
+      "epoch": 0.3083526783758924,
+      "grad_norm": 0.44225645065307617,
+      "learning_rate": 8.289007895607375e-05,
+      "loss": 2.502,
+      "step": 880
+    },
+    {
+      "epoch": 0.3090534799176558,
+      "grad_norm": 0.47749781608581543,
+      "learning_rate": 8.28027085095783e-05,
+      "loss": 2.5043,
+      "step": 882
+    },
+    {
+      "epoch": 0.3097542814594192,
+      "grad_norm": 0.4569682478904724,
+      "learning_rate": 8.271516186133511e-05,
+      "loss": 2.4454,
+      "step": 884
+    },
+    {
+      "epoch": 0.31045508300118263,
+      "grad_norm": 0.4561903178691864,
+      "learning_rate": 8.262743948160632e-05,
+      "loss": 2.4826,
+      "step": 886
+    },
+    {
+      "epoch": 0.311155884542946,
+      "grad_norm": 0.4749627411365509,
+      "learning_rate": 8.253954184159803e-05,
+      "loss": 2.4707,
+      "step": 888
+    },
+    {
+      "epoch": 0.3118566860847094,
+      "grad_norm": 0.4455653131008148,
+      "learning_rate": 8.245146941345774e-05,
+      "loss": 2.4647,
+      "step": 890
+    },
+    {
+      "epoch": 0.3125574876264728,
+      "grad_norm": 0.4758734405040741,
+      "learning_rate": 8.236322267027193e-05,
+      "loss": 2.4885,
+      "step": 892
+    },
+    {
+      "epoch": 0.31325828916823617,
+      "grad_norm": 0.45016252994537354,
+      "learning_rate": 8.227480208606333e-05,
+      "loss": 2.4993,
+      "step": 894
+    },
+    {
+      "epoch": 0.31395909070999956,
+      "grad_norm": 0.48177486658096313,
+      "learning_rate": 8.218620813578847e-05,
+      "loss": 2.4838,
+      "step": 896
+    },
+    {
+      "epoch": 0.31465989225176294,
+      "grad_norm": 0.4863053858280182,
+      "learning_rate": 8.209744129533519e-05,
+      "loss": 2.5381,
+      "step": 898
+    },
+    {
+      "epoch": 0.3153606937935263,
+      "grad_norm": 0.49010857939720154,
+      "learning_rate": 8.200850204151995e-05,
+      "loss": 2.5721,
+      "step": 900
+    },
+    {
+      "epoch": 0.3160614953352897,
+      "grad_norm": 0.43315884470939636,
+      "learning_rate": 8.191939085208542e-05,
+      "loss": 2.4976,
+      "step": 902
+    },
+    {
+      "epoch": 0.31676229687705315,
+      "grad_norm": 0.4580542743206024,
+      "learning_rate": 8.183010820569775e-05,
+      "loss": 2.4885,
+      "step": 904
+    },
+    {
+      "epoch": 0.31746309841881654,
+      "grad_norm": 0.40409061312675476,
+      "learning_rate": 8.17406545819441e-05,
+      "loss": 2.4872,
+      "step": 906
+    },
+    {
+      "epoch": 0.3181638999605799,
+      "grad_norm": 0.5763331055641174,
+      "learning_rate": 8.16510304613301e-05,
+      "loss": 2.4991,
+      "step": 908
+    },
+    {
+      "epoch": 0.3188647015023433,
+      "grad_norm": 0.4705376625061035,
+      "learning_rate": 8.156123632527714e-05,
+      "loss": 2.5071,
+      "step": 910
+    },
+    {
+      "epoch": 0.3195655030441067,
+      "grad_norm": 0.42091286182403564,
+      "learning_rate": 8.147127265611991e-05,
+      "loss": 2.4639,
+      "step": 912
+    },
+    {
+      "epoch": 0.3202663045858701,
+      "grad_norm": 0.4637336730957031,
+      "learning_rate": 8.138113993710377e-05,
+      "loss": 2.5368,
+      "step": 914
+    },
+    {
+      "epoch": 0.32096710612763346,
+      "grad_norm": 0.48139557242393494,
+      "learning_rate": 8.129083865238207e-05,
+      "loss": 2.4677,
+      "step": 916
+    },
+    {
+      "epoch": 0.32166790766939685,
+      "grad_norm": 0.4305325448513031,
+      "learning_rate": 8.120036928701367e-05,
+      "loss": 2.5125,
+      "step": 918
+    },
+    {
+      "epoch": 0.3223687092111603,
+      "grad_norm": 0.4576675295829773,
+      "learning_rate": 8.110973232696029e-05,
+      "loss": 2.4559,
+      "step": 920
+    },
+    {
+      "epoch": 0.32306951075292367,
+      "grad_norm": 0.4506300687789917,
+      "learning_rate": 8.10189282590839e-05,
+      "loss": 2.46,
+      "step": 922
+    },
+    {
+      "epoch": 0.32377031229468706,
+      "grad_norm": 0.4221961200237274,
+      "learning_rate": 8.092795757114405e-05,
+      "loss": 2.5269,
+      "step": 924
+    },
+    {
+      "epoch": 0.32447111383645044,
+      "grad_norm": 0.4183507561683655,
+      "learning_rate": 8.083682075179535e-05,
+      "loss": 2.4963,
+      "step": 926
+    },
+    {
+      "epoch": 0.3251719153782138,
+      "grad_norm": 0.40157100558280945,
+      "learning_rate": 8.07455182905848e-05,
+      "loss": 2.497,
+      "step": 928
+    },
+    {
+      "epoch": 0.3258727169199772,
+      "grad_norm": 0.41130340099334717,
+      "learning_rate": 8.06540506779491e-05,
+      "loss": 2.4894,
+      "step": 930
+    },
+    {
+      "epoch": 0.3265735184617406,
+      "grad_norm": 0.45491817593574524,
+      "learning_rate": 8.056241840521212e-05,
+      "loss": 2.4991,
+      "step": 932
+    },
+    {
+      "epoch": 0.327274320003504,
+      "grad_norm": 0.5285101532936096,
+      "learning_rate": 8.047062196458222e-05,
+      "loss": 2.4956,
+      "step": 934
+    },
+    {
+      "epoch": 0.3279751215452674,
+      "grad_norm": 0.4162616729736328,
+      "learning_rate": 8.037866184914952e-05,
+      "loss": 2.4349,
+      "step": 936
+    },
+    {
+      "epoch": 0.3286759230870308,
+      "grad_norm": 0.47431331872940063,
+      "learning_rate": 8.028653855288342e-05,
+      "loss": 2.4502,
+      "step": 938
+    },
+    {
+      "epoch": 0.3293767246287942,
+      "grad_norm": 0.516228437423706,
+      "learning_rate": 8.019425257062983e-05,
+      "loss": 2.463,
+      "step": 940
+    },
+    {
+      "epoch": 0.3300775261705576,
+      "grad_norm": 0.4729955196380615,
+      "learning_rate": 8.010180439810852e-05,
+      "loss": 2.4764,
+      "step": 942
+    },
+    {
+      "epoch": 0.33077832771232096,
+      "grad_norm": 0.4590371549129486,
+      "learning_rate": 8.000919453191046e-05,
+      "loss": 2.4677,
+      "step": 944
+    },
+    {
+      "epoch": 0.33147912925408435,
+      "grad_norm": 0.4180818498134613,
+      "learning_rate": 7.99164234694952e-05,
+      "loss": 2.4478,
+      "step": 946
+    },
+    {
+      "epoch": 0.33217993079584773,
+      "grad_norm": 0.459830641746521,
+      "learning_rate": 7.982349170918819e-05,
+      "loss": 2.5018,
+      "step": 948
+    },
+    {
+      "epoch": 0.3328807323376111,
+      "grad_norm": 0.42037877440452576,
+      "learning_rate": 7.9730399750178e-05,
+      "loss": 2.5018,
+      "step": 950
+    },
+    {
+      "epoch": 0.33358153387937456,
+      "grad_norm": 0.48284855484962463,
+      "learning_rate": 7.963714809251375e-05,
+      "loss": 2.4853,
+      "step": 952
+    },
+    {
+      "epoch": 0.33428233542113794,
+      "grad_norm": 0.4275258779525757,
+      "learning_rate": 7.954373723710247e-05,
+      "loss": 2.4716,
+      "step": 954
+    },
+    {
+      "epoch": 0.33498313696290133,
+      "grad_norm": 0.4475047290325165,
+      "learning_rate": 7.945016768570619e-05,
+      "loss": 2.5075,
+      "step": 956
+    },
+    {
+      "epoch": 0.3356839385046647,
+      "grad_norm": 0.4026126265525818,
+      "learning_rate": 7.93564399409395e-05,
+      "loss": 2.4381,
+      "step": 958
+    },
+    {
+      "epoch": 0.3363847400464281,
+      "grad_norm": 0.4063897132873535,
+      "learning_rate": 7.926255450626668e-05,
+      "loss": 2.4607,
+      "step": 960
+    },
+    {
+      "epoch": 0.3370855415881915,
+      "grad_norm": 0.4034193158149719,
+      "learning_rate": 7.916851188599908e-05,
+      "loss": 2.4615,
+      "step": 962
+    },
+    {
+      "epoch": 0.33778634312995487,
+      "grad_norm": 0.449735552072525,
+      "learning_rate": 7.907431258529232e-05,
+      "loss": 2.447,
+      "step": 964
+    },
+    {
+      "epoch": 0.33848714467171825,
+      "grad_norm": 0.46238863468170166,
+      "learning_rate": 7.897995711014373e-05,
+      "loss": 2.5097,
+      "step": 966
+    },
+    {
+      "epoch": 0.3391879462134817,
+      "grad_norm": 0.4304943084716797,
+      "learning_rate": 7.88854459673895e-05,
+      "loss": 2.48,
+      "step": 968
+    },
+    {
+      "epoch": 0.3398887477552451,
+      "grad_norm": 0.4926084578037262,
+      "learning_rate": 7.879077966470194e-05,
+      "loss": 2.4716,
+      "step": 970
+    },
+    {
+      "epoch": 0.34058954929700846,
+      "grad_norm": 0.4950031638145447,
+      "learning_rate": 7.86959587105869e-05,
+      "loss": 2.4648,
+      "step": 972
+    },
+    {
+      "epoch": 0.34129035083877185,
+      "grad_norm": 0.4255363345146179,
+      "learning_rate": 7.860098361438092e-05,
+      "loss": 2.4388,
+      "step": 974
+    },
+    {
+      "epoch": 0.34199115238053523,
+      "grad_norm": 0.48036453127861023,
+      "learning_rate": 7.85058548862485e-05,
+      "loss": 2.4286,
+      "step": 976
+    },
+    {
+      "epoch": 0.3426919539222986,
+      "grad_norm": 0.43235254287719727,
+      "learning_rate": 7.84105730371794e-05,
+      "loss": 2.5079,
+      "step": 978
+    },
+    {
+      "epoch": 0.343392755464062,
+      "grad_norm": 0.4415871500968933,
+      "learning_rate": 7.831513857898589e-05,
+      "loss": 2.4995,
+      "step": 980
+    },
+    {
+      "epoch": 0.3440935570058254,
+      "grad_norm": 0.4261731207370758,
+      "learning_rate": 7.821955202429997e-05,
+      "loss": 2.4732,
+      "step": 982
+    },
+    {
+      "epoch": 0.34479435854758883,
+      "grad_norm": 0.4182513654232025,
+      "learning_rate": 7.812381388657066e-05,
+      "loss": 2.4424,
+      "step": 984
+    },
+    {
+      "epoch": 0.3454951600893522,
+      "grad_norm": 0.38975790143013,
+      "learning_rate": 7.802792468006119e-05,
+      "loss": 2.4672,
+      "step": 986
+    },
+    {
+      "epoch": 0.3461959616311156,
+      "grad_norm": 0.4010710120201111,
+      "learning_rate": 7.793188491984626e-05,
+      "loss": 2.4458,
+      "step": 988
+    },
+    {
+      "epoch": 0.346896763172879,
+      "grad_norm": 0.4334274232387543,
+      "learning_rate": 7.783569512180933e-05,
+      "loss": 2.4296,
+      "step": 990
+    },
+    {
+      "epoch": 0.34759756471464237,
+      "grad_norm": 0.42379283905029297,
+      "learning_rate": 7.77393558026397e-05,
+      "loss": 2.4737,
+      "step": 992
+    },
+    {
+      "epoch": 0.34829836625640576,
+      "grad_norm": 0.3837527632713318,
+      "learning_rate": 7.764286747982998e-05,
+      "loss": 2.5038,
+      "step": 994
+    },
+    {
+      "epoch": 0.34899916779816914,
+      "grad_norm": 0.3951718211174011,
+      "learning_rate": 7.754623067167301e-05,
+      "loss": 2.4743,
+      "step": 996
+    },
+    {
+      "epoch": 0.3496999693399325,
+      "grad_norm": 0.42385029792785645,
+      "learning_rate": 7.744944589725931e-05,
+      "loss": 2.4839,
+      "step": 998
+    },
+    {
+      "epoch": 0.35040077088169597,
+      "grad_norm": 0.35979923605918884,
+      "learning_rate": 7.73525136764742e-05,
+      "loss": 2.466,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35110157242345935,
+      "grad_norm": 0.3813255727291107,
+      "learning_rate": 7.725543452999501e-05,
+      "loss": 2.4657,
+      "step": 1002
+    },
+    {
+      "epoch": 0.35180237396522274,
+      "grad_norm": 0.44004762172698975,
+      "learning_rate": 7.715820897928831e-05,
+      "loss": 2.4969,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3525031755069861,
+      "grad_norm": 0.41002413630485535,
+      "learning_rate": 7.706083754660704e-05,
+      "loss": 2.4762,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3532039770487495,
+      "grad_norm": 0.39008840918540955,
+      "learning_rate": 7.696332075498778e-05,
+      "loss": 2.4857,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3539047785905129,
+      "grad_norm": 0.38840317726135254,
+      "learning_rate": 7.686565912824797e-05,
+      "loss": 2.4517,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3546055801322763,
+      "grad_norm": 0.4124884605407715,
+      "learning_rate": 7.676785319098292e-05,
+      "loss": 2.5003,
+      "step": 1012
+    },
+    {
+      "epoch": 0.35530638167403966,
+      "grad_norm": 0.39664867520332336,
+      "learning_rate": 7.666990346856323e-05,
+      "loss": 2.4567,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3560071832158031,
+      "grad_norm": 0.41662028431892395,
+      "learning_rate": 7.65718104871318e-05,
+      "loss": 2.4603,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3567079847575665,
+      "grad_norm": 0.41218042373657227,
+      "learning_rate": 7.647357477360103e-05,
+      "loss": 2.4775,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3574087862993299,
+      "grad_norm": 0.4241933226585388,
+      "learning_rate": 7.637519685565007e-05,
+      "loss": 2.4581,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35810958784109326,
+      "grad_norm": 0.4417102336883545,
+      "learning_rate": 7.627667726172188e-05,
+      "loss": 2.4707,
+      "step": 1022
+    },
+    {
+      "epoch": 0.35881038938285664,
+      "grad_norm": 0.4153529405593872,
+      "learning_rate": 7.617801652102047e-05,
+      "loss": 2.4899,
+      "step": 1024
+    },
+    {
+      "epoch": 0.35951119092462,
+      "grad_norm": 0.42329278588294983,
+      "learning_rate": 7.607921516350805e-05,
+      "loss": 2.4753,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3602119924663834,
+      "grad_norm": 0.4246659576892853,
+      "learning_rate": 7.598027371990209e-05,
+      "loss": 2.4887,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3609127940081468,
+      "grad_norm": 0.4006962180137634,
+      "learning_rate": 7.58811927216726e-05,
+      "loss": 2.4446,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3616135955499102,
+      "grad_norm": 0.4364110231399536,
+      "learning_rate": 7.578197270103922e-05,
+      "loss": 2.4661,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3623143970916736,
+      "grad_norm": 0.41410303115844727,
+      "learning_rate": 7.568261419096829e-05,
+      "loss": 2.5071,
+      "step": 1034
+    },
+    {
+      "epoch": 0.363015198633437,
+      "grad_norm": 0.4187104403972626,
+      "learning_rate": 7.558311772517015e-05,
+      "loss": 2.4325,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3637160001752004,
+      "grad_norm": 0.3853437304496765,
+      "learning_rate": 7.548348383809609e-05,
+      "loss": 2.5366,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3644168017169638,
+      "grad_norm": 0.4048047363758087,
+      "learning_rate": 7.538371306493565e-05,
+      "loss": 2.5017,
+      "step": 1040
+    },
+    {
+      "epoch": 0.36511760325872716,
+      "grad_norm": 0.5083643198013306,
+      "learning_rate": 7.528380594161357e-05,
+      "loss": 2.4921,
+      "step": 1042
+    },
+    {
+      "epoch": 0.36581840480049055,
+      "grad_norm": 0.48616695404052734,
+      "learning_rate": 7.51837630047871e-05,
+      "loss": 2.4939,
+      "step": 1044
+    },
+    {
+      "epoch": 0.36651920634225393,
+      "grad_norm": 0.4780706763267517,
+      "learning_rate": 7.508358479184292e-05,
+      "loss": 2.4355,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3672200078840173,
+      "grad_norm": 0.48671412467956543,
+      "learning_rate": 7.498327184089444e-05,
+      "loss": 2.4921,
+      "step": 1048
+    },
+    {
+      "epoch": 0.36792080942578076,
+      "grad_norm": 0.4375866651535034,
+      "learning_rate": 7.488282469077878e-05,
+      "loss": 2.47,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36862161096754414,
+      "grad_norm": 0.40635204315185547,
+      "learning_rate": 7.478224388105395e-05,
+      "loss": 2.4903,
+      "step": 1052
+    },
+    {
+      "epoch": 0.36932241250930753,
+      "grad_norm": 0.4381403923034668,
+      "learning_rate": 7.468152995199586e-05,
+      "loss": 2.4512,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3700232140510709,
+      "grad_norm": 0.4581202566623688,
+      "learning_rate": 7.458068344459556e-05,
+      "loss": 2.461,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3707240155928343,
+      "grad_norm": 0.41336745023727417,
+      "learning_rate": 7.447970490055615e-05,
+      "loss": 2.5116,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3714248171345977,
+      "grad_norm": 0.44782060384750366,
+      "learning_rate": 7.437859486229008e-05,
+      "loss": 2.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37212561867636107,
+      "grad_norm": 0.39710888266563416,
+      "learning_rate": 7.42773538729161e-05,
+      "loss": 2.4415,
+      "step": 1062
+    },
+    {
+      "epoch": 0.37282642021812445,
+      "grad_norm": 0.4727705717086792,
+      "learning_rate": 7.41759824762563e-05,
+      "loss": 2.5044,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3735272217598879,
+      "grad_norm": 0.4554997384548187,
+      "learning_rate": 7.407448121683334e-05,
+      "loss": 2.458,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3742280233016513,
+      "grad_norm": 0.3889988958835602,
+      "learning_rate": 7.397285063986743e-05,
+      "loss": 2.3949,
+      "step": 1068
+    },
+    {
+      "epoch": 0.37492882484341467,
+      "grad_norm": 0.4311942160129547,
+      "learning_rate": 7.387109129127338e-05,
+      "loss": 2.49,
+      "step": 1070
+    },
+    {
+      "epoch": 0.37562962638517805,
+      "grad_norm": 0.5033543705940247,
+      "learning_rate": 7.376920371765778e-05,
+      "loss": 2.5222,
+      "step": 1072
+    },
+    {
+      "epoch": 0.37633042792694144,
+      "grad_norm": 0.44039902091026306,
+      "learning_rate": 7.366718846631589e-05,
+      "loss": 2.4676,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3770312294687048,
+      "grad_norm": 0.4856110215187073,
+      "learning_rate": 7.356504608522886e-05,
+      "loss": 2.475,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3777320310104682,
+      "grad_norm": 0.4543730914592743,
+      "learning_rate": 7.346277712306074e-05,
+      "loss": 2.4278,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3784328325522316,
+      "grad_norm": 0.44085338711738586,
+      "learning_rate": 7.336038212915547e-05,
+      "loss": 2.4556,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37913363409399503,
+      "grad_norm": 0.46556994318962097,
+      "learning_rate": 7.325786165353403e-05,
+      "loss": 2.5022,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3798344356357584,
+      "grad_norm": 0.49558380246162415,
+      "learning_rate": 7.315521624689135e-05,
+      "loss": 2.441,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3805352371775218,
+      "grad_norm": 0.41412246227264404,
+      "learning_rate": 7.305244646059353e-05,
+      "loss": 2.4467,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3812360387192852,
+      "grad_norm": 0.40206030011177063,
+      "learning_rate": 7.294955284667473e-05,
+      "loss": 2.4546,
+      "step": 1088
+    },
+    {
+      "epoch": 0.38193684026104857,
+      "grad_norm": 0.4202801585197449,
+      "learning_rate": 7.284653595783427e-05,
+      "loss": 2.4638,
+      "step": 1090
+    },
+    {
+      "epoch": 0.38263764180281196,
+      "grad_norm": 0.41482287645339966,
+      "learning_rate": 7.274339634743364e-05,
+      "loss": 2.4295,
+      "step": 1092
+    },
+    {
+      "epoch": 0.38333844334457534,
+      "grad_norm": 0.4058382511138916,
+      "learning_rate": 7.264013456949352e-05,
+      "loss": 2.4885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3840392448863387,
+      "grad_norm": 0.3752575218677521,
+      "learning_rate": 7.253675117869088e-05,
+      "loss": 2.4247,
+      "step": 1096
+    },
+    {
+      "epoch": 0.38474004642810217,
+      "grad_norm": 0.4093259572982788,
+      "learning_rate": 7.24332467303559e-05,
+      "loss": 2.4403,
+      "step": 1098
+    },
+    {
+      "epoch": 0.38544084796986555,
+      "grad_norm": 0.4372352063655853,
+      "learning_rate": 7.232962178046901e-05,
+      "loss": 2.4606,
+      "step": 1100
+    },
+    {
+      "epoch": 0.38614164951162894,
+      "grad_norm": 0.4315595328807831,
+      "learning_rate": 7.222587688565796e-05,
+      "loss": 2.4756,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3868424510533923,
+      "grad_norm": 0.4413852095603943,
+      "learning_rate": 7.212201260319477e-05,
+      "loss": 2.4678,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3875432525951557,
+      "grad_norm": 0.41132184863090515,
+      "learning_rate": 7.201802949099275e-05,
+      "loss": 2.4659,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3882440541369191,
+      "grad_norm": 0.3732544183731079,
+      "learning_rate": 7.191392810760356e-05,
+      "loss": 2.4209,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3889448556786825,
+      "grad_norm": 0.4892706871032715,
+      "learning_rate": 7.180970901221408e-05,
+      "loss": 2.4762,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38964565722044586,
+      "grad_norm": 0.42454981803894043,
+      "learning_rate": 7.170537276464355e-05,
+      "loss": 2.4781,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3903464587622093,
+      "grad_norm": 0.4017874598503113,
+      "learning_rate": 7.160091992534051e-05,
+      "loss": 2.4563,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3910472603039727,
+      "grad_norm": 0.4679020643234253,
+      "learning_rate": 7.14963510553797e-05,
+      "loss": 2.4337,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3917480618457361,
+      "grad_norm": 0.4594592750072479,
+      "learning_rate": 7.139166671645923e-05,
+      "loss": 2.4431,
+      "step": 1118
+    },
+    {
+      "epoch": 0.39244886338749946,
+      "grad_norm": 0.4678356349468231,
+      "learning_rate": 7.128686747089737e-05,
+      "loss": 2.4513,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39314966492926284,
+      "grad_norm": 0.462960422039032,
+      "learning_rate": 7.118195388162966e-05,
+      "loss": 2.4636,
+      "step": 1122
+    },
+    {
+      "epoch": 0.39385046647102623,
+      "grad_norm": 0.4102267920970917,
+      "learning_rate": 7.107692651220585e-05,
+      "loss": 2.4249,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3945512680127896,
+      "grad_norm": 0.4165082573890686,
+      "learning_rate": 7.097178592678683e-05,
+      "loss": 2.4793,
+      "step": 1126
+    },
+    {
+      "epoch": 0.395252069554553,
+      "grad_norm": 0.438484787940979,
+      "learning_rate": 7.086653269014171e-05,
+      "loss": 2.4437,
+      "step": 1128
+    },
+    {
+      "epoch": 0.39595287109631644,
+      "grad_norm": 0.4235878884792328,
+      "learning_rate": 7.076116736764461e-05,
+      "loss": 2.4276,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3966536726380798,
+      "grad_norm": 0.4382783770561218,
+      "learning_rate": 7.065569052527182e-05,
+      "loss": 2.4572,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3973544741798432,
+      "grad_norm": 0.4001896381378174,
+      "learning_rate": 7.055010272959861e-05,
+      "loss": 2.4796,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3980552757216066,
+      "grad_norm": 0.5236755609512329,
+      "learning_rate": 7.044440454779625e-05,
+      "loss": 2.395,
+      "step": 1136
+    },
+    {
+      "epoch": 0.39875607726337,
+      "grad_norm": 0.4379066824913025,
+      "learning_rate": 7.033859654762898e-05,
+      "loss": 2.4206,
+      "step": 1138
+    },
+    {
+      "epoch": 0.39945687880513336,
+      "grad_norm": 0.44132131338119507,
+      "learning_rate": 7.023267929745092e-05,
+      "loss": 2.4723,
+      "step": 1140
+    },
+    {
+      "epoch": 0.40015768034689675,
+      "grad_norm": 0.4132694602012634,
+      "learning_rate": 7.012665336620303e-05,
+      "loss": 2.4371,
+      "step": 1142
+    },
+    {
+      "epoch": 0.40085848188866013,
+      "grad_norm": 0.4186249077320099,
+      "learning_rate": 7.002051932341008e-05,
+      "loss": 2.4136,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4015592834304235,
+      "grad_norm": 0.43945756554603577,
+      "learning_rate": 6.99142777391775e-05,
+      "loss": 2.3931,
+      "step": 1146
+    },
+    {
+      "epoch": 0.40226008497218696,
+      "grad_norm": 0.402310311794281,
+      "learning_rate": 6.980792918418849e-05,
+      "loss": 2.4078,
+      "step": 1148
+    },
+    {
+      "epoch": 0.40296088651395034,
+      "grad_norm": 0.4048164486885071,
+      "learning_rate": 6.970147422970074e-05,
+      "loss": 2.44,
+      "step": 1150
+    },
+    {
+      "epoch": 0.40366168805571373,
+      "grad_norm": 0.4304601550102234,
+      "learning_rate": 6.959491344754357e-05,
+      "loss": 2.481,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4043624895974771,
+      "grad_norm": 0.41222673654556274,
+      "learning_rate": 6.94882474101147e-05,
+      "loss": 2.4581,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.4438692033290863,
+      "learning_rate": 6.938147669037724e-05,
+      "loss": 2.4474,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4057640926810039,
+      "grad_norm": 0.41356995701789856,
+      "learning_rate": 6.927460186185663e-05,
+      "loss": 2.4394,
+      "step": 1158
+    },
+    {
+      "epoch": 0.40646489422276727,
+      "grad_norm": 0.43938153982162476,
+      "learning_rate": 6.916762349863755e-05,
+      "loss": 2.4529,
+      "step": 1160
+    },
+    {
+      "epoch": 0.40716569576453066,
+      "grad_norm": 0.4264746904373169,
+      "learning_rate": 6.906054217536074e-05,
+      "loss": 2.4103,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4078664973062941,
+      "grad_norm": 0.40205949544906616,
+      "learning_rate": 6.895335846722007e-05,
+      "loss": 2.4533,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4085672988480575,
+      "grad_norm": 0.3951698839664459,
+      "learning_rate": 6.88460729499594e-05,
+      "loss": 2.4059,
+      "step": 1166
+    },
+    {
+      "epoch": 0.40926810038982087,
+      "grad_norm": 0.40137380361557007,
+      "learning_rate": 6.873868619986938e-05,
+      "loss": 2.4282,
+      "step": 1168
+    },
+    {
+      "epoch": 0.40996890193158425,
+      "grad_norm": 0.39662086963653564,
+      "learning_rate": 6.863119879378451e-05,
+      "loss": 2.404,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41066970347334764,
+      "grad_norm": 0.40667644143104553,
+      "learning_rate": 6.852361130907992e-05,
+      "loss": 2.4205,
+      "step": 1172
+    },
+    {
+      "epoch": 0.411370505015111,
+      "grad_norm": 0.4023439288139343,
+      "learning_rate": 6.841592432366835e-05,
+      "loss": 2.4271,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4120713065568744,
+      "grad_norm": 0.4581812620162964,
+      "learning_rate": 6.830813841599702e-05,
+      "loss": 2.4589,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4127721080986378,
+      "grad_norm": 0.40035149455070496,
+      "learning_rate": 6.820025416504447e-05,
+      "loss": 2.4299,
+      "step": 1178
+    },
+    {
+      "epoch": 0.41347290964040123,
+      "grad_norm": 0.40674343705177307,
+      "learning_rate": 6.809227215031757e-05,
+      "loss": 2.4479,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4141737111821646,
+      "grad_norm": 0.3893290162086487,
+      "learning_rate": 6.798419295184823e-05,
+      "loss": 2.4555,
+      "step": 1182
+    },
+    {
+      "epoch": 0.414874512723928,
+      "grad_norm": 0.4277731776237488,
+      "learning_rate": 6.787601715019051e-05,
+      "loss": 2.426,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4155753142656914,
+      "grad_norm": 0.3891987204551697,
+      "learning_rate": 6.776774532641731e-05,
+      "loss": 2.455,
+      "step": 1186
+    },
+    {
+      "epoch": 0.41627611580745477,
+      "grad_norm": 0.42866095900535583,
+      "learning_rate": 6.765937806211731e-05,
+      "loss": 2.4278,
+      "step": 1188
+    },
+    {
+      "epoch": 0.41697691734921816,
+      "grad_norm": 0.4561346769332886,
+      "learning_rate": 6.75509159393919e-05,
+      "loss": 2.4655,
+      "step": 1190
+    },
+    {
+      "epoch": 0.41767771889098154,
+      "grad_norm": 0.4195440709590912,
+      "learning_rate": 6.744235954085193e-05,
+      "loss": 2.4429,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4183785204327449,
+      "grad_norm": 0.43625494837760925,
+      "learning_rate": 6.733370944961476e-05,
+      "loss": 2.4733,
+      "step": 1194
+    },
+    {
+      "epoch": 0.41907932197450837,
+      "grad_norm": 0.45825034379959106,
+      "learning_rate": 6.7224966249301e-05,
+      "loss": 2.4267,
+      "step": 1196
+    },
+    {
+      "epoch": 0.41978012351627175,
+      "grad_norm": 0.4522963762283325,
+      "learning_rate": 6.711613052403129e-05,
+      "loss": 2.4268,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42048092505803514,
+      "grad_norm": 0.43499064445495605,
+      "learning_rate": 6.700720285842344e-05,
+      "loss": 2.4429,
+      "step": 1200
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 2854,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3510201344786432e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}