diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,19982 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9994373329582222,
+  "eval_steps": 500,
+  "global_step": 14216,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002813335208890139,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.625879043600562e-09,
+      "loss": 1.9967,
+      "step": 1
+    },
+    {
+      "epoch": 0.0014066676044450696,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8129395218002812e-08,
+      "loss": 1.834,
+      "step": 5
+    },
+    {
+      "epoch": 0.0028133352088901393,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.6258790436005624e-08,
+      "loss": 1.9701,
+      "step": 10
+    },
+    {
+      "epoch": 0.004220002813335209,
+      "grad_norm": 9.75,
+      "learning_rate": 8.438818565400843e-08,
+      "loss": 2.0358,
+      "step": 15
+    },
+    {
+      "epoch": 0.0056266704177802785,
+      "grad_norm": 9.25,
+      "learning_rate": 1.1251758087201125e-07,
+      "loss": 2.1257,
+      "step": 20
+    },
+    {
+      "epoch": 0.007033338022225348,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.4064697609001405e-07,
+      "loss": 2.1333,
+      "step": 25
+    },
+    {
+      "epoch": 0.008440005626670417,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6877637130801686e-07,
+      "loss": 2.0451,
+      "step": 30
+    },
+    {
+      "epoch": 0.009846673231115488,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9690576652601966e-07,
+      "loss": 2.114,
+      "step": 35
+    },
+    {
+      "epoch": 0.011253340835560557,
+      "grad_norm": 10.6875,
+      "learning_rate": 2.250351617440225e-07,
+      "loss": 2.0872,
+      "step": 40
+    },
+    {
+      "epoch": 0.012660008440005626,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.5316455696202533e-07,
+      "loss": 2.1654,
+      "step": 45
+    },
+    {
+      "epoch": 0.014066676044450697,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.812939521800281e-07,
+      "loss": 1.8529,
+      "step": 50
+    },
+    {
+      "epoch": 0.015473343648895766,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.0942334739803094e-07,
+      "loss": 2.0161,
+      "step": 55
+    },
+    {
+      "epoch": 0.016880011253340835,
+      "grad_norm": 8.875,
+      "learning_rate": 3.375527426160337e-07,
+      "loss": 1.9371,
+      "step": 60
+    },
+    {
+      "epoch": 0.018286678857785905,
+      "grad_norm": 10.0625,
+      "learning_rate": 3.6568213783403655e-07,
+      "loss": 1.7475,
+      "step": 65
+    },
+    {
+      "epoch": 0.019693346462230976,
+      "grad_norm": 8.625,
+      "learning_rate": 3.9381153305203933e-07,
+      "loss": 2.1096,
+      "step": 70
+    },
+    {
+      "epoch": 0.021100014066676043,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.2194092827004216e-07,
+      "loss": 2.1636,
+      "step": 75
+    },
+    {
+      "epoch": 0.022506681671121114,
+      "grad_norm": 18.375,
+      "learning_rate": 4.50070323488045e-07,
+      "loss": 1.888,
+      "step": 80
+    },
+    {
+      "epoch": 0.023913349275566185,
+      "grad_norm": 9.6875,
+      "learning_rate": 4.781997187060478e-07,
+      "loss": 2.1679,
+      "step": 85
+    },
+    {
+      "epoch": 0.025320016880011252,
+      "grad_norm": 8.5,
+      "learning_rate": 5.063291139240507e-07,
+      "loss": 1.8771,
+      "step": 90
+    },
+    {
+      "epoch": 0.026726684484456323,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.344585091420533e-07,
+      "loss": 1.9294,
+      "step": 95
+    },
+    {
+      "epoch": 0.028133352088901394,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.625879043600562e-07,
+      "loss": 1.8677,
+      "step": 100
+    },
+    {
+      "epoch": 0.02954001969334646,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.907172995780591e-07,
+      "loss": 1.9361,
+      "step": 105
+    },
+    {
+      "epoch": 0.03094668729779153,
+      "grad_norm": 10.3125,
+      "learning_rate": 6.188466947960619e-07,
+      "loss": 2.1789,
+      "step": 110
+    },
+    {
+      "epoch": 0.0323533549022366,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.469760900140648e-07,
+      "loss": 2.0211,
+      "step": 115
+    },
+    {
+      "epoch": 0.03376002250668167,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.751054852320674e-07,
+      "loss": 2.1231,
+      "step": 120
+    },
+    {
+      "epoch": 0.03516669011112674,
+      "grad_norm": 12.375,
+      "learning_rate": 7.032348804500703e-07,
+      "loss": 2.1527,
+      "step": 125
+    },
+    {
+      "epoch": 0.03657335771557181,
+      "grad_norm": 14.6875,
+      "learning_rate": 7.313642756680731e-07,
+      "loss": 1.7367,
+      "step": 130
+    },
+    {
+      "epoch": 0.03798002532001688,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.59493670886076e-07,
+      "loss": 2.2069,
+      "step": 135
+    },
+    {
+      "epoch": 0.03938669292446195,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.876230661040787e-07,
+      "loss": 2.1692,
+      "step": 140
+    },
+    {
+      "epoch": 0.040793360528907016,
+      "grad_norm": 10.1875,
+      "learning_rate": 8.157524613220815e-07,
+      "loss": 2.0095,
+      "step": 145
+    },
+    {
+      "epoch": 0.04220002813335209,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.438818565400843e-07,
+      "loss": 1.6281,
+      "step": 150
+    },
+    {
+      "epoch": 0.04360669573779716,
+      "grad_norm": 7.625,
+      "learning_rate": 8.720112517580872e-07,
+      "loss": 1.759,
+      "step": 155
+    },
+    {
+      "epoch": 0.04501336334224223,
+      "grad_norm": 8.75,
+      "learning_rate": 9.0014064697609e-07,
+      "loss": 1.9759,
+      "step": 160
+    },
+    {
+      "epoch": 0.0464200309466873,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.282700421940928e-07,
+      "loss": 2.0089,
+      "step": 165
+    },
+    {
+      "epoch": 0.04782669855113237,
+      "grad_norm": 15.375,
+      "learning_rate": 9.563994374120955e-07,
+      "loss": 1.8533,
+      "step": 170
+    },
+    {
+      "epoch": 0.04923336615557744,
+      "grad_norm": 11.625,
+      "learning_rate": 9.845288326300985e-07,
+      "loss": 2.0724,
+      "step": 175
+    },
+    {
+      "epoch": 0.050640033760022504,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.0126582278481013e-06,
+      "loss": 2.2394,
+      "step": 180
+    },
+    {
+      "epoch": 0.052046701364467575,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.040787623066104e-06,
+      "loss": 1.9738,
+      "step": 185
+    },
+    {
+      "epoch": 0.053453368968912646,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0689170182841067e-06,
+      "loss": 2.0486,
+      "step": 190
+    },
+    {
+      "epoch": 0.054860036573357716,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0970464135021096e-06,
+      "loss": 1.88,
+      "step": 195
+    },
+    {
+      "epoch": 0.05626670417780279,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1251758087201124e-06,
+      "loss": 1.9277,
+      "step": 200
+    },
+    {
+      "epoch": 0.05767337178224786,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1533052039381152e-06,
+      "loss": 1.8934,
+      "step": 205
+    },
+    {
+      "epoch": 0.05908003938669292,
+      "grad_norm": 7.0,
+      "learning_rate": 1.1814345991561182e-06,
+      "loss": 2.2194,
+      "step": 210
+    },
+    {
+      "epoch": 0.06048670699113799,
+      "grad_norm": 6.375,
+      "learning_rate": 1.209563994374121e-06,
+      "loss": 1.7994,
+      "step": 215
+    },
+    {
+      "epoch": 0.06189337459558306,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.2376933895921238e-06,
+      "loss": 1.7983,
+      "step": 220
+    },
+    {
+      "epoch": 0.06330004220002813,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.2658227848101265e-06,
+      "loss": 1.9125,
+      "step": 225
+    },
+    {
+      "epoch": 0.0647067098044732,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.2939521800281295e-06,
+      "loss": 1.9858,
+      "step": 230
+    },
+    {
+      "epoch": 0.06611337740891828,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.322081575246132e-06,
+      "loss": 2.0835,
+      "step": 235
+    },
+    {
+      "epoch": 0.06752004501336334,
+      "grad_norm": 5.625,
+      "learning_rate": 1.3502109704641349e-06,
+      "loss": 2.0088,
+      "step": 240
+    },
+    {
+      "epoch": 0.06892671261780842,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.3783403656821376e-06,
+      "loss": 2.0304,
+      "step": 245
+    },
+    {
+      "epoch": 0.07033338022225348,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.4064697609001406e-06,
+      "loss": 2.1096,
+      "step": 250
+    },
+    {
+      "epoch": 0.07174004782669856,
+      "grad_norm": 9.25,
+      "learning_rate": 1.4345991561181434e-06,
+      "loss": 1.982,
+      "step": 255
+    },
+    {
+      "epoch": 0.07314671543114362,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.4627285513361462e-06,
+      "loss": 1.9651,
+      "step": 260
+    },
+    {
+      "epoch": 0.07455338303558869,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.4908579465541492e-06,
+      "loss": 1.8912,
+      "step": 265
+    },
+    {
+      "epoch": 0.07596005064003376,
+      "grad_norm": 5.25,
+      "learning_rate": 1.518987341772152e-06,
+      "loss": 1.8922,
+      "step": 270
+    },
+    {
+      "epoch": 0.07736671824447883,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.5471167369901545e-06,
+      "loss": 2.033,
+      "step": 275
+    },
+    {
+      "epoch": 0.0787733858489239,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.5752461322081573e-06,
+      "loss": 1.8743,
+      "step": 280
+    },
+    {
+      "epoch": 0.08018005345336897,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6033755274261603e-06,
+      "loss": 1.9284,
+      "step": 285
+    },
+    {
+      "epoch": 0.08158672105781403,
+      "grad_norm": 4.5,
+      "learning_rate": 1.631504922644163e-06,
+      "loss": 1.8721,
+      "step": 290
+    },
+    {
+      "epoch": 0.08299338866225911,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.6596343178621659e-06,
+      "loss": 1.8106,
+      "step": 295
+    },
+    {
+      "epoch": 0.08440005626670417,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.6877637130801686e-06,
+      "loss": 1.8406,
+      "step": 300
+    },
+    {
+      "epoch": 0.08580672387114925,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.7158931082981716e-06,
+      "loss": 2.0746,
+      "step": 305
+    },
+    {
+      "epoch": 0.08721339147559432,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.7440225035161744e-06,
+      "loss": 1.7973,
+      "step": 310
+    },
+    {
+      "epoch": 0.08862005908003939,
+      "grad_norm": 5.25,
+      "learning_rate": 1.772151898734177e-06,
+      "loss": 1.9815,
+      "step": 315
+    },
+    {
+      "epoch": 0.09002672668448446,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.80028129395218e-06,
+      "loss": 1.9967,
+      "step": 320
+    },
+    {
+      "epoch": 0.09143339428892952,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.8284106891701827e-06,
+      "loss": 1.8916,
+      "step": 325
+    },
+    {
+      "epoch": 0.0928400618933746,
+      "grad_norm": 5.25,
+      "learning_rate": 1.8565400843881855e-06,
+      "loss": 1.686,
+      "step": 330
+    },
+    {
+      "epoch": 0.09424672949781966,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.8846694796061883e-06,
+      "loss": 2.0456,
+      "step": 335
+    },
+    {
+      "epoch": 0.09565339710226474,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.912798874824191e-06,
+      "loss": 2.115,
+      "step": 340
+    },
+    {
+      "epoch": 0.0970600647067098,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.940928270042194e-06,
+      "loss": 1.8874,
+      "step": 345
+    },
+    {
+      "epoch": 0.09846673231115488,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.969057665260197e-06,
+      "loss": 2.0844,
+      "step": 350
+    },
+    {
+      "epoch": 0.09987339991559994,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.9971870604782e-06,
+      "loss": 1.8196,
+      "step": 355
+    },
+    {
+      "epoch": 0.10128006752004501,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.0253164556962026e-06,
+      "loss": 1.9226,
+      "step": 360
+    },
+    {
+      "epoch": 0.10268673512449009,
+      "grad_norm": 4.75,
+      "learning_rate": 2.0534458509142054e-06,
+      "loss": 1.8721,
+      "step": 365
+    },
+    {
+      "epoch": 0.10409340272893515,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.081575246132208e-06,
+      "loss": 1.8233,
+      "step": 370
+    },
+    {
+      "epoch": 0.10550007033338023,
+      "grad_norm": 4.25,
+      "learning_rate": 2.109704641350211e-06,
+      "loss": 1.967,
+      "step": 375
+    },
+    {
+      "epoch": 0.10690673793782529,
+      "grad_norm": 3.625,
+      "learning_rate": 2.1378340365682133e-06,
+      "loss": 1.9782,
+      "step": 380
+    },
+    {
+      "epoch": 0.10831340554227036,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.1659634317862165e-06,
+      "loss": 2.1084,
+      "step": 385
+    },
+    {
+      "epoch": 0.10972007314671543,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.1940928270042193e-06,
+      "loss": 2.2131,
+      "step": 390
+    },
+    {
+      "epoch": 0.1111267407511605,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.8195,
+      "step": 395
+    },
+    {
+      "epoch": 0.11253340835560557,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.250351617440225e-06,
+      "loss": 1.9049,
+      "step": 400
+    },
+    {
+      "epoch": 0.11394007596005064,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.278481012658228e-06,
+      "loss": 1.9102,
+      "step": 405
+    },
+    {
+      "epoch": 0.11534674356449572,
+      "grad_norm": 5.875,
+      "learning_rate": 2.3066104078762304e-06,
+      "loss": 1.8624,
+      "step": 410
+    },
+    {
+      "epoch": 0.11675341116894078,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.3347398030942336e-06,
+      "loss": 1.7784,
+      "step": 415
+    },
+    {
+      "epoch": 0.11816007877338584,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3628691983122364e-06,
+      "loss": 1.9591,
+      "step": 420
+    },
+    {
+      "epoch": 0.11956674637783092,
+      "grad_norm": 4.3125,
+      "learning_rate": 2.3909985935302387e-06,
+      "loss": 1.7737,
+      "step": 425
+    },
+    {
+      "epoch": 0.12097341398227598,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.419127988748242e-06,
+      "loss": 1.7564,
+      "step": 430
+    },
+    {
+      "epoch": 0.12238008158672106,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.4472573839662443e-06,
+      "loss": 1.8436,
+      "step": 435
+    },
+    {
+      "epoch": 0.12378674919116613,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.4753867791842475e-06,
+      "loss": 1.9585,
+      "step": 440
+    },
+    {
+      "epoch": 0.1251934167956112,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.5035161744022503e-06,
+      "loss": 1.7232,
+      "step": 445
+    },
+    {
+      "epoch": 0.12660008440005627,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.531645569620253e-06,
+      "loss": 1.9946,
+      "step": 450
+    },
+    {
+      "epoch": 0.12800675200450135,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.559774964838256e-06,
+      "loss": 1.9282,
+      "step": 455
+    },
+    {
+      "epoch": 0.1294134196089464,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.587904360056259e-06,
+      "loss": 1.7587,
+      "step": 460
+    },
+    {
+      "epoch": 0.13082008721339147,
+      "grad_norm": 4.75,
+      "learning_rate": 2.6160337552742614e-06,
+      "loss": 1.785,
+      "step": 465
+    },
+    {
+      "epoch": 0.13222675481783655,
+      "grad_norm": 3.375,
+      "learning_rate": 2.644163150492264e-06,
+      "loss": 1.8986,
+      "step": 470
+    },
+    {
+      "epoch": 0.13363342242228163,
+      "grad_norm": 4.8125,
+      "learning_rate": 2.6722925457102674e-06,
+      "loss": 1.7295,
+      "step": 475
+    },
+    {
+      "epoch": 0.13504009002672668,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.7004219409282697e-06,
+      "loss": 1.9288,
+      "step": 480
+    },
+    {
+      "epoch": 0.13644675763117176,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.728551336146273e-06,
+      "loss": 1.7269,
+      "step": 485
+    },
+    {
+      "epoch": 0.13785342523561683,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7566807313642753e-06,
+      "loss": 1.4669,
+      "step": 490
+    },
+    {
+      "epoch": 0.13926009284006188,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.7848101265822785e-06,
+      "loss": 1.9001,
+      "step": 495
+    },
+    {
+      "epoch": 0.14066676044450696,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8129395218002813e-06,
+      "loss": 1.7307,
+      "step": 500
+    },
+    {
+      "epoch": 0.14207342804895204,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.8410689170182836e-06,
+      "loss": 1.8674,
+      "step": 505
+    },
+    {
+      "epoch": 0.14348009565339712,
+      "grad_norm": 12.375,
+      "learning_rate": 2.869198312236287e-06,
+      "loss": 1.9095,
+      "step": 510
+    },
+    {
+      "epoch": 0.14488676325784217,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.8973277074542896e-06,
+      "loss": 1.8201,
+      "step": 515
+    },
+    {
+      "epoch": 0.14629343086228724,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9254571026722924e-06,
+      "loss": 1.9156,
+      "step": 520
+    },
+    {
+      "epoch": 0.14770009846673232,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.953586497890295e-06,
+      "loss": 1.7277,
+      "step": 525
+    },
+    {
+      "epoch": 0.14910676607117737,
+      "grad_norm": 4.25,
+      "learning_rate": 2.9817158931082984e-06,
+      "loss": 1.7154,
+      "step": 530
+    },
+    {
+      "epoch": 0.15051343367562245,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.0098452883263007e-06,
+      "loss": 1.7168,
+      "step": 535
+    },
+    {
+      "epoch": 0.15192010128006753,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.037974683544304e-06,
+      "loss": 2.0013,
+      "step": 540
+    },
+    {
+      "epoch": 0.15332676888451258,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.0661040787623063e-06,
+      "loss": 1.8941,
+      "step": 545
+    },
+    {
+      "epoch": 0.15473343648895765,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.094233473980309e-06,
+      "loss": 1.682,
+      "step": 550
+    },
+    {
+      "epoch": 0.15614010409340273,
+      "grad_norm": 5.75,
+      "learning_rate": 3.1223628691983123e-06,
+      "loss": 1.7576,
+      "step": 555
+    },
+    {
+      "epoch": 0.1575467716978478,
+      "grad_norm": 4.71875,
+      "learning_rate": 3.1504922644163146e-06,
+      "loss": 1.8123,
+      "step": 560
+    },
+    {
+      "epoch": 0.15895343930229286,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.178621659634318e-06,
+      "loss": 1.6807,
+      "step": 565
+    },
+    {
+      "epoch": 0.16036010690673794,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.2067510548523206e-06,
+      "loss": 1.6907,
+      "step": 570
+    },
+    {
+      "epoch": 0.16176677451118301,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.2348804500703234e-06,
+      "loss": 1.892,
+      "step": 575
+    },
+    {
+      "epoch": 0.16317344211562806,
+      "grad_norm": 2.625,
+      "learning_rate": 3.263009845288326e-06,
+      "loss": 2.0725,
+      "step": 580
+    },
+    {
+      "epoch": 0.16458010972007314,
+      "grad_norm": 4.5,
+      "learning_rate": 3.2911392405063294e-06,
+      "loss": 1.8148,
+      "step": 585
+    },
+    {
+      "epoch": 0.16598677732451822,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.3192686357243317e-06,
+      "loss": 1.7494,
+      "step": 590
+    },
+    {
+      "epoch": 0.1673934449289633,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.3473980309423345e-06,
+      "loss": 1.8697,
+      "step": 595
+    },
+    {
+      "epoch": 0.16880011253340835,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.3755274261603373e-06,
+      "loss": 1.9731,
+      "step": 600
+    },
+    {
+      "epoch": 0.17020678013785343,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.40365682137834e-06,
+      "loss": 1.7131,
+      "step": 605
+    },
+    {
+      "epoch": 0.1716134477422985,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.4317862165963433e-06,
+      "loss": 1.8335,
+      "step": 610
+    },
+    {
+      "epoch": 0.17302011534674355,
+      "grad_norm": 2.625,
+      "learning_rate": 3.4599156118143456e-06,
+      "loss": 1.8878,
+      "step": 615
+    },
+    {
+      "epoch": 0.17442678295118863,
+      "grad_norm": 4.5,
+      "learning_rate": 3.488045007032349e-06,
+      "loss": 1.6365,
+      "step": 620
+    },
+    {
+      "epoch": 0.1758334505556337,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.5161744022503516e-06,
+      "loss": 1.8077,
+      "step": 625
+    },
+    {
+      "epoch": 0.17724011816007879,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.544303797468354e-06,
+      "loss": 1.9373,
+      "step": 630
+    },
+    {
+      "epoch": 0.17864678576452384,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.572433192686357e-06,
+      "loss": 1.8536,
+      "step": 635
+    },
+    {
+      "epoch": 0.1800534533689689,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.60056258790436e-06,
+      "loss": 1.7992,
+      "step": 640
+    },
+    {
+      "epoch": 0.181460120973414,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.6286919831223627e-06,
+      "loss": 1.7746,
+      "step": 645
+    },
+    {
+      "epoch": 0.18286678857785904,
+      "grad_norm": 3.125,
+      "learning_rate": 3.6568213783403655e-06,
+      "loss": 2.0007,
+      "step": 650
+    },
+    {
+      "epoch": 0.18427345618230412,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.6849507735583683e-06,
+      "loss": 1.8947,
+      "step": 655
+    },
+    {
+      "epoch": 0.1856801237867492,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.713080168776371e-06,
+      "loss": 1.8504,
+      "step": 660
+    },
+    {
+      "epoch": 0.18708679139119427,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.7412095639943743e-06,
+      "loss": 1.7796,
+      "step": 665
+    },
+    {
+      "epoch": 0.18849345899563932,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.7693389592123766e-06,
+      "loss": 1.7598,
+      "step": 670
+    },
+    {
+      "epoch": 0.1899001266000844,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.7974683544303794e-06,
+      "loss": 1.6422,
+      "step": 675
+    },
+    {
+      "epoch": 0.19130679420452948,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.825597749648382e-06,
+      "loss": 1.8552,
+      "step": 680
+    },
+    {
+      "epoch": 0.19271346180897453,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.853727144866385e-06,
+      "loss": 1.5753,
+      "step": 685
+    },
+    {
+      "epoch": 0.1941201294134196,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.881856540084388e-06,
+      "loss": 1.7207,
+      "step": 690
+    },
+    {
+      "epoch": 0.19552679701786468,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.909985935302391e-06,
+      "loss": 1.5458,
+      "step": 695
+    },
+    {
+      "epoch": 0.19693346462230976,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.938115330520394e-06,
+      "loss": 1.6871,
+      "step": 700
+    },
+    {
+      "epoch": 0.1983401322267548,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.9662447257383965e-06,
+      "loss": 1.9603,
+      "step": 705
+    },
+    {
+      "epoch": 0.1997467998311999,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.9943741209564e-06,
+      "loss": 1.6939,
+      "step": 710
+    },
+    {
+      "epoch": 0.20115346743564497,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.022503516174402e-06,
+      "loss": 1.7818,
+      "step": 715
+    },
+    {
+      "epoch": 0.20256013504009002,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.050632911392405e-06,
+      "loss": 1.7131,
+      "step": 720
+    },
+    {
+      "epoch": 0.2039668026445351,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.078762306610408e-06,
+      "loss": 1.8219,
+      "step": 725
+    },
+    {
+      "epoch": 0.20537347024898017,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.106891701828411e-06,
+      "loss": 1.6883,
+      "step": 730
+    },
+    {
+      "epoch": 0.20678013785342522,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.135021097046413e-06,
+      "loss": 1.8892,
+      "step": 735
+    },
+    {
+      "epoch": 0.2081868054578703,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.163150492264416e-06,
+      "loss": 1.6866,
+      "step": 740
+    },
+    {
+      "epoch": 0.20959347306231538,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.191279887482419e-06,
+      "loss": 1.4289,
+      "step": 745
+    },
+    {
+      "epoch": 0.21100014066676046,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.219409282700422e-06,
+      "loss": 1.5471,
+      "step": 750
+    },
+    {
+      "epoch": 0.2124068082712055,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.247538677918425e-06,
+      "loss": 1.9359,
+      "step": 755
+    },
+    {
+      "epoch": 0.21381347587565058,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.275668073136427e-06,
+      "loss": 1.9177,
+      "step": 760
+    },
+    {
+      "epoch": 0.21522014348009566,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.30379746835443e-06,
+      "loss": 1.9106,
+      "step": 765
+    },
+    {
+      "epoch": 0.2166268110845407,
+      "grad_norm": 3.375,
+      "learning_rate": 4.331926863572433e-06,
+      "loss": 1.9653,
+      "step": 770
+    },
+    {
+      "epoch": 0.2180334786889858,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.360056258790436e-06,
+      "loss": 1.7532,
+      "step": 775
+    },
+    {
+      "epoch": 0.21944014629343087,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.388185654008439e-06,
+      "loss": 1.8319,
+      "step": 780
+    },
+    {
+      "epoch": 0.22084681389787594,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.416315049226442e-06,
+      "loss": 2.0233,
+      "step": 785
+    },
+    {
+      "epoch": 0.222253481502321,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.9903,
+      "step": 790
+    },
+    {
+      "epoch": 0.22366014910676607,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.4725738396624465e-06,
+      "loss": 2.0019,
+      "step": 795
+    },
+    {
+      "epoch": 0.22506681671121115,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.50070323488045e-06,
+      "loss": 1.7976,
+      "step": 800
+    },
+    {
+      "epoch": 0.2264734843156562,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.528832630098453e-06,
+      "loss": 1.8314,
+      "step": 805
+    },
+    {
+      "epoch": 0.22788015192010128,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.556962025316456e-06,
+      "loss": 1.7646,
+      "step": 810
+    },
+    {
+      "epoch": 0.22928681952454635,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.585091420534458e-06,
+      "loss": 1.5575,
+      "step": 815
+    },
+    {
+      "epoch": 0.23069348712899143,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.613220815752461e-06,
+      "loss": 1.4394,
+      "step": 820
+    },
+    {
+      "epoch": 0.23210015473343648,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.641350210970464e-06,
+      "loss": 1.6944,
+      "step": 825
+    },
+    {
+      "epoch": 0.23350682233788156,
+      "grad_norm": 3.625,
+      "learning_rate": 4.669479606188467e-06,
+      "loss": 1.9698,
+      "step": 830
+    },
+    {
+      "epoch": 0.23491348994232664,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.69760900140647e-06,
+      "loss": 1.8369,
+      "step": 835
+    },
+    {
+      "epoch": 0.2363201575467717,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.725738396624473e-06,
+      "loss": 2.0049,
+      "step": 840
+    },
+    {
+      "epoch": 0.23772682515121676,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.753867791842475e-06,
+      "loss": 1.6301,
+      "step": 845
+    },
+    {
+      "epoch": 0.23913349275566184,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.7819971870604775e-06,
+      "loss": 2.0206,
+      "step": 850
+    },
+    {
+      "epoch": 0.24054016036010692,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.810126582278481e-06,
+      "loss": 1.9102,
+      "step": 855
+    },
+    {
+      "epoch": 0.24194682796455197,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.838255977496484e-06,
+      "loss": 1.5072,
+      "step": 860
+    },
+    {
+      "epoch": 0.24335349556899705,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.866385372714487e-06,
+      "loss": 1.5767,
+      "step": 865
+    },
+    {
+      "epoch": 0.24476016317344212,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.894514767932489e-06,
+      "loss": 1.979,
+      "step": 870
+    },
+    {
+      "epoch": 0.24616683077788717,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.922644163150492e-06,
+      "loss": 1.7614,
+      "step": 875
+    },
+    {
+      "epoch": 0.24757349838233225,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.950773558368495e-06,
+      "loss": 1.7441,
+      "step": 880
+    },
+    {
+      "epoch": 0.24898016598677733,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.978902953586497e-06,
+      "loss": 1.7426,
+      "step": 885
+    },
+    {
+      "epoch": 0.2503868335912224,
+      "grad_norm": 4.6875,
+      "learning_rate": 5.0070323488045006e-06,
+      "loss": 1.7325,
+      "step": 890
+    },
+    {
+      "epoch": 0.25179350119566746,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.035161744022504e-06,
+      "loss": 1.8025,
+      "step": 895
+    },
+    {
+      "epoch": 0.25320016880011254,
+      "grad_norm": 4.5,
+      "learning_rate": 5.063291139240506e-06,
+      "loss": 1.7743,
+      "step": 900
+    },
+    {
+      "epoch": 0.2546068364045576,
+      "grad_norm": 4.125,
+      "learning_rate": 5.0914205344585085e-06,
+      "loss": 1.5968,
+      "step": 905
+    },
+    {
+      "epoch": 0.2560135040090027,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.119549929676512e-06,
+      "loss": 1.7051,
+      "step": 910
+    },
+    {
+      "epoch": 0.25742017161344777,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.147679324894515e-06,
+      "loss": 1.7582,
+      "step": 915
+    },
+    {
+      "epoch": 0.2588268392178928,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.175808720112518e-06,
+      "loss": 1.7175,
+      "step": 920
+    },
+    {
+      "epoch": 0.26023350682233787,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.20393811533052e-06,
+      "loss": 1.7261,
+      "step": 925
+    },
+    {
+      "epoch": 0.26164017442678295,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.232067510548523e-06,
+      "loss": 1.7149,
+      "step": 930
+    },
+    {
+      "epoch": 0.263046842031228,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.260196905766526e-06,
+      "loss": 1.7658,
+      "step": 935
+    },
+    {
+      "epoch": 0.2644535096356731,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.288326300984528e-06,
+      "loss": 1.691,
+      "step": 940
+    },
+    {
+      "epoch": 0.2658601772401182,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.3164556962025316e-06,
+      "loss": 2.1113,
+      "step": 945
+    },
+    {
+      "epoch": 0.26726684484456326,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.344585091420535e-06,
+      "loss": 1.9024,
+      "step": 950
+    },
+    {
+      "epoch": 0.2686735124490083,
+      "grad_norm": 4.875,
+      "learning_rate": 5.372714486638537e-06,
+      "loss": 1.8391,
+      "step": 955
+    },
+    {
+      "epoch": 0.27008018005345336,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.4008438818565395e-06,
+      "loss": 1.607,
+      "step": 960
+    },
+    {
+      "epoch": 0.27148684765789843,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.428973277074543e-06,
+      "loss": 1.8196,
+      "step": 965
+    },
+    {
+      "epoch": 0.2728935152623435,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.457102672292546e-06,
+      "loss": 1.6765,
+      "step": 970
+    },
+    {
+      "epoch": 0.2743001828667886,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.485232067510548e-06,
+      "loss": 1.8108,
+      "step": 975
+    },
+    {
+      "epoch": 0.27570685047123367,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.513361462728551e-06,
+      "loss": 1.8147,
+      "step": 980
+    },
+    {
+      "epoch": 0.27711351807567874,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.541490857946554e-06,
+      "loss": 1.936,
+      "step": 985
+    },
+    {
+      "epoch": 0.27852018568012377,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.569620253164557e-06,
+      "loss": 1.781,
+      "step": 990
+    },
+    {
+      "epoch": 0.27992685328456884,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.597749648382559e-06,
+      "loss": 1.4884,
+      "step": 995
+    },
+    {
+      "epoch": 0.2813335208890139,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.6258790436005626e-06,
+      "loss": 1.472,
+      "step": 1000
+    },
+    {
+      "epoch": 0.282740188493459,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.654008438818566e-06,
+      "loss": 1.4928,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2841468560979041,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.682137834036567e-06,
+      "loss": 1.6619,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28555352370234915,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.7102672292545705e-06,
+      "loss": 2.0167,
+      "step": 1015
+    },
+    {
+      "epoch": 0.28696019130679423,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.738396624472574e-06,
+      "loss": 1.7161,
+      "step": 1020
+    },
+    {
+      "epoch": 0.28836685891123925,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.766526019690577e-06,
+      "loss": 1.9748,
+      "step": 1025
+    },
+    {
+      "epoch": 0.28977352651568433,
+      "grad_norm": 4.59375,
+      "learning_rate": 5.794655414908579e-06,
+      "loss": 1.7424,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2911801941201294,
+      "grad_norm": 2.625,
+      "learning_rate": 5.822784810126582e-06,
+      "loss": 1.6562,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2925868617245745,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.850914205344585e-06,
+      "loss": 1.7536,
+      "step": 1040
+    },
+    {
+      "epoch": 0.29399352932901957,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.879043600562588e-06,
+      "loss": 1.8031,
+      "step": 1045
+    },
+    {
+      "epoch": 0.29540019693346464,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.90717299578059e-06,
+      "loss": 1.6396,
+      "step": 1050
+    },
+    {
+      "epoch": 0.29680686453790966,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.9353023909985935e-06,
+      "loss": 1.7845,
+      "step": 1055
+    },
+    {
+      "epoch": 0.29821353214235474,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.963431786216597e-06,
+      "loss": 1.9526,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2996201997467998,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.991561181434598e-06,
+      "loss": 1.5986,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3010268673512449,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.0196905766526015e-06,
+      "loss": 1.7045,
+      "step": 1070
+    },
+    {
+      "epoch": 0.30243353495569,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.047819971870605e-06,
+      "loss": 1.6006,
+      "step": 1075
+    },
+    {
+      "epoch": 0.30384020256013505,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.075949367088608e-06,
+      "loss": 1.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.30524687016458013,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.10407876230661e-06,
+      "loss": 1.6308,
+      "step": 1085
+    },
+    {
+      "epoch": 0.30665353776902515,
+      "grad_norm": 2.375,
+      "learning_rate": 6.132208157524613e-06,
+      "loss": 1.7865,
+      "step": 1090
+    },
+    {
+      "epoch": 0.30806020537347023,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.160337552742616e-06,
+      "loss": 1.8331,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3094668729779153,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.188466947960618e-06,
+      "loss": 1.7786,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3108735405823604,
+      "grad_norm": 3.0,
+      "learning_rate": 6.216596343178621e-06,
+      "loss": 2.0003,
+      "step": 1105
+    },
+    {
+      "epoch": 0.31228020818680546,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.2447257383966245e-06,
+      "loss": 1.824,
+      "step": 1110
+    },
+    {
+      "epoch": 0.31368687579125054,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.272855133614628e-06,
+      "loss": 1.8362,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3150935433956956,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.300984528832629e-06,
+      "loss": 1.6952,
+      "step": 1120
+    },
+    {
+      "epoch": 0.31650021100014064,
+      "grad_norm": 5.84375,
+      "learning_rate": 6.3291139240506325e-06,
+      "loss": 1.5749,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3179068786045857,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.357243319268636e-06,
+      "loss": 1.5969,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3193135462090308,
+      "grad_norm": 4.375,
+      "learning_rate": 6.385372714486638e-06,
+      "loss": 1.5812,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3207202138134759,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.413502109704641e-06,
+      "loss": 1.9075,
+      "step": 1140
+    },
+    {
+      "epoch": 0.32212688141792095,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.4416315049226436e-06,
+      "loss": 1.9003,
+      "step": 1145
+    },
+    {
+      "epoch": 0.32353354902236603,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.469760900140647e-06,
+      "loss": 1.7614,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3249402166268111,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.497890295358649e-06,
+      "loss": 1.812,
+      "step": 1155
+    },
+    {
+      "epoch": 0.32634688423125613,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.526019690576652e-06,
+      "loss": 1.6565,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3277535518357012,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.5541490857946555e-06,
+      "loss": 1.9087,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3291602194401463,
+      "grad_norm": 3.375,
+      "learning_rate": 6.582278481012659e-06,
+      "loss": 1.6842,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33056688704459136,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.61040787623066e-06,
+      "loss": 1.4252,
+      "step": 1175
+    },
+    {
+      "epoch": 0.33197355464903644,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.6385372714486634e-06,
+      "loss": 1.6601,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3333802222534815,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.7866,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3347868898579266,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.694796061884669e-06,
+      "loss": 1.6193,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3361935574623716,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.722925457102672e-06,
+      "loss": 1.6185,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3376002250668167,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.7510548523206746e-06,
+      "loss": 1.7356,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3390068926712618,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.779184247538678e-06,
+      "loss": 1.8204,
+      "step": 1205
+    },
+    {
+      "epoch": 0.34041356027570685,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.80731364275668e-06,
+      "loss": 1.6249,
+      "step": 1210
+    },
+    {
+      "epoch": 0.34182022788015193,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.835443037974683e-06,
+      "loss": 1.8842,
+      "step": 1215
+    },
+    {
+      "epoch": 0.343226895484597,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.8635724331926865e-06,
+      "loss": 1.7801,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3446335630890421,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.891701828410689e-06,
+      "loss": 1.4851,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3460402306934871,
+      "grad_norm": 3.125,
+      "learning_rate": 6.919831223628691e-06,
+      "loss": 1.7178,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3474468982979322,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.9479606188466944e-06,
+      "loss": 1.6033,
+      "step": 1235
+    },
+    {
+      "epoch": 0.34885356590237726,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.976090014064698e-06,
+      "loss": 1.7969,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35026023350682234,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.0042194092827e-06,
+      "loss": 1.8187,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3516669011112674,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.032348804500703e-06,
+      "loss": 1.819,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3530735687157125,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.0604781997187056e-06,
+      "loss": 1.7712,
+      "step": 1255
+    },
+    {
+      "epoch": 0.35448023632015757,
+      "grad_norm": 5.375,
+      "learning_rate": 7.088607594936708e-06,
+      "loss": 1.7155,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3558869039246026,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.116736990154711e-06,
+      "loss": 1.6361,
+      "step": 1265
+    },
+    {
+      "epoch": 0.35729357152904767,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.144866385372714e-06,
+      "loss": 1.7478,
+      "step": 1270
+    },
+    {
+      "epoch": 0.35870023913349275,
+      "grad_norm": 3.0,
+      "learning_rate": 7.1729957805907175e-06,
+      "loss": 1.844,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3601069067379378,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.20112517580872e-06,
+      "loss": 1.6414,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3615135743423829,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.229254571026722e-06,
+      "loss": 1.7427,
+      "step": 1285
+    },
+    {
+      "epoch": 0.362920241946828,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.2573839662447254e-06,
+      "loss": 1.7395,
+      "step": 1290
+    },
+    {
+      "epoch": 0.36432690955127306,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.285513361462729e-06,
+      "loss": 1.8215,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3657335771557181,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.313642756680731e-06,
+      "loss": 1.7573,
+      "step": 1300
+    },
+    {
+      "epoch": 0.36714024476016316,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.341772151898734e-06,
+      "loss": 1.841,
+      "step": 1305
+    },
+    {
+      "epoch": 0.36854691236460824,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.3699015471167365e-06,
+      "loss": 1.842,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3699535799690533,
+      "grad_norm": 2.875,
+      "learning_rate": 7.398030942334739e-06,
+      "loss": 2.0199,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3713602475734984,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.426160337552742e-06,
+      "loss": 1.8881,
+      "step": 1320
+    },
+    {
+      "epoch": 0.37276691517794347,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.454289732770745e-06,
+      "loss": 1.7692,
+      "step": 1325
+    },
+    {
+      "epoch": 0.37417358278238855,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.4824191279887485e-06,
+      "loss": 1.8311,
+      "step": 1330
+    },
+    {
+      "epoch": 0.37558025038683357,
+      "grad_norm": 5.34375,
+      "learning_rate": 7.510548523206751e-06,
+      "loss": 1.7703,
+      "step": 1335
+    },
+    {
+      "epoch": 0.37698691799127865,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.538677918424753e-06,
+      "loss": 1.7416,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3783935855957237,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.566807313642756e-06,
+      "loss": 1.7398,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3798002532001688,
+      "grad_norm": 2.875,
+      "learning_rate": 7.594936708860759e-06,
+      "loss": 1.571,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3812069208046139,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.623066104078762e-06,
+      "loss": 1.4582,
+      "step": 1355
+    },
+    {
+      "epoch": 0.38261358840905896,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.651195499296764e-06,
+      "loss": 1.981,
+      "step": 1360
+    },
+    {
+      "epoch": 0.38402025601350404,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.679324894514768e-06,
+      "loss": 1.611,
+      "step": 1365
+    },
+    {
+      "epoch": 0.38542692361794906,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.70745428973277e-06,
+      "loss": 1.7369,
+      "step": 1370
+    },
+    {
+      "epoch": 0.38683359122239414,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.735583684950773e-06,
+      "loss": 2.081,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3882402588268392,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.763713080168775e-06,
+      "loss": 1.5745,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3896469264312843,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.791842475386778e-06,
+      "loss": 1.7473,
+      "step": 1385
+    },
+    {
+      "epoch": 0.39105359403572937,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.819971870604782e-06,
+      "loss": 1.7226,
+      "step": 1390
+    },
+    {
+      "epoch": 0.39246026164017445,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.848101265822784e-06,
+      "loss": 1.5063,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3938669292446195,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.876230661040788e-06,
+      "loss": 1.5973,
+      "step": 1400
+    },
+    {
+      "epoch": 0.39527359684906455,
+      "grad_norm": 3.625,
+      "learning_rate": 7.904360056258789e-06,
+      "loss": 1.6063,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3966802644535096,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.932489451476793e-06,
+      "loss": 1.889,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3980869320579547,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.960618846694795e-06,
+      "loss": 1.65,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3994935996623998,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.9887482419128e-06,
+      "loss": 1.6546,
+      "step": 1420
+    },
+    {
+      "epoch": 0.40090026726684486,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.999998914675671e-06,
+      "loss": 1.6465,
+      "step": 1425
+    },
+    {
+      "epoch": 0.40230693487128993,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.999992282140243e-06,
+      "loss": 1.7975,
+      "step": 1430
+    },
+    {
+      "epoch": 0.40371360247573496,
+      "grad_norm": 3.625,
+      "learning_rate": 7.999979620037334e-06,
+      "loss": 1.7769,
+      "step": 1435
+    },
+    {
+      "epoch": 0.40512027008018003,
+      "grad_norm": 3.25,
+      "learning_rate": 7.999960928386025e-06,
+      "loss": 1.6168,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4065269376846251,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.9999362072145e-06,
+      "loss": 1.8668,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4079336052890702,
+      "grad_norm": 2.875,
+      "learning_rate": 7.999905456560018e-06,
+      "loss": 1.8308,
+      "step": 1450
+    },
+    {
+      "epoch": 0.40934027289351527,
+      "grad_norm": 3.984375,
+      "learning_rate": 7.999868676468933e-06,
+      "loss": 1.7166,
+      "step": 1455
+    },
+    {
+      "epoch": 0.41074694049796034,
+      "grad_norm": 3.125,
+      "learning_rate": 7.99982586699669e-06,
+      "loss": 1.9376,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4121536081024054,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.999777028207818e-06,
+      "loss": 1.9246,
+      "step": 1465
+    },
+    {
+      "epoch": 0.41356027570685044,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.999722160175935e-06,
+      "loss": 1.8283,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4149669433112955,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.99966126298375e-06,
+      "loss": 1.5573,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4163736109157406,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.99959433672306e-06,
+      "loss": 1.7425,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4177802785201857,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.999521381494747e-06,
+      "loss": 1.468,
+      "step": 1485
+    },
+    {
+      "epoch": 0.41918694612463075,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.999442397408785e-06,
+      "loss": 2.0143,
+      "step": 1490
+    },
+    {
+      "epoch": 0.42059361372907583,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.999357384584235e-06,
+      "loss": 1.5066,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4220002813335209,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.999266343149242e-06,
+      "loss": 1.5112,
+      "step": 1500
+    },
+    {
+      "epoch": 0.42340694893796593,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.999169273241046e-06,
+      "loss": 1.6816,
+      "step": 1505
+    },
+    {
+      "epoch": 0.424813616542411,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.999066175005965e-06,
+      "loss": 1.7814,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4262202841468561,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.99895704859941e-06,
+      "loss": 1.5777,
+      "step": 1515
+    },
+    {
+      "epoch": 0.42762695175130117,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.99884189418588e-06,
+      "loss": 1.614,
+      "step": 1520
+    },
+    {
+      "epoch": 0.42903361935574624,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.998720711938954e-06,
+      "loss": 1.3149,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4304402869601913,
+      "grad_norm": 3.125,
+      "learning_rate": 7.998593502041306e-06,
+      "loss": 1.7651,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4318469545646364,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.998460264684688e-06,
+      "loss": 1.5757,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4332536221690814,
+      "grad_norm": 2.625,
+      "learning_rate": 7.998321000069943e-06,
+      "loss": 1.9167,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4346602897735265,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.998175708406999e-06,
+      "loss": 1.388,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4360669573779716,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.998024389914864e-06,
+      "loss": 1.7055,
+      "step": 1550
+    },
+    {
+      "epoch": 0.43747362498241665,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.997867044821638e-06,
+      "loss": 1.5819,
+      "step": 1555
+    },
+    {
+      "epoch": 0.43888029258686173,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.997703673364501e-06,
+      "loss": 1.7198,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4402869601913068,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.997534275789718e-06,
+      "loss": 1.894,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4416936277957519,
+      "grad_norm": 3.0,
+      "learning_rate": 7.99735885235264e-06,
+      "loss": 1.5067,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4431002954001969,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.997177403317696e-06,
+      "loss": 1.6449,
+      "step": 1575
+    },
+    {
+      "epoch": 0.444506963004642,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.996989928958404e-06,
+      "loss": 1.8517,
+      "step": 1580
+    },
+    {
+      "epoch": 0.44591363060908706,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.996796429557362e-06,
+      "loss": 1.7391,
+      "step": 1585
+    },
+    {
+      "epoch": 0.44732029821353214,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.996596905406248e-06,
+      "loss": 1.8785,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4487269658179772,
+      "grad_norm": 2.625,
+      "learning_rate": 7.996391356805825e-06,
+      "loss": 1.6024,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4501336334224223,
+      "grad_norm": 6.21875,
+      "learning_rate": 7.996179784065935e-06,
+      "loss": 1.6681,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4515403010268674,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.995962187505502e-06,
+      "loss": 1.742,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4529469686313124,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.995738567452531e-06,
+      "loss": 1.477,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4543536362357575,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.995508924244104e-06,
+      "loss": 1.7455,
+      "step": 1615
+    },
+    {
+      "epoch": 0.45576030384020255,
+      "grad_norm": 2.625,
+      "learning_rate": 7.995273258226387e-06,
+      "loss": 1.7959,
+      "step": 1620
+    },
+    {
+      "epoch": 0.45716697144464763,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.995031569754617e-06,
+      "loss": 1.7619,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4585736390490927,
+      "grad_norm": 3.25,
+      "learning_rate": 7.994783859193119e-06,
+      "loss": 1.6018,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4599803066535378,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.994530126915285e-06,
+      "loss": 1.7328,
+      "step": 1635
+    },
+    {
+      "epoch": 0.46138697425798286,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.994270373303593e-06,
+      "loss": 1.6123,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4627936418624279,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.994004598749597e-06,
+      "loss": 1.6376,
+      "step": 1645
+    },
+    {
+      "epoch": 0.46420030946687296,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.99373280365392e-06,
+      "loss": 1.5974,
+      "step": 1650
+    },
+    {
+      "epoch": 0.46560697707131804,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.993454988426265e-06,
+      "loss": 1.3981,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4670136446757631,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.993171153485412e-06,
+      "loss": 1.7091,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4684203122802082,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.992881299259208e-06,
+      "loss": 1.4902,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4698269798846533,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.99258542618458e-06,
+      "loss": 1.8131,
+      "step": 1670
+    },
+    {
+      "epoch": 0.47123364748909835,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.992283534707527e-06,
+      "loss": 1.5423,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4726403150935434,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.991975625283116e-06,
+      "loss": 1.7866,
+      "step": 1680
+    },
+    {
+      "epoch": 0.47404698269798845,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.991661698375489e-06,
+      "loss": 1.4981,
+      "step": 1685
+    },
+    {
+      "epoch": 0.47545365030243353,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.991341754457858e-06,
+      "loss": 1.6244,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4768603179068786,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.991015794012506e-06,
+      "loss": 1.6531,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4782669855113237,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.990683817530783e-06,
+      "loss": 1.6086,
+      "step": 1700
+    },
+    {
+      "epoch": 0.47967365311576876,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.990345825513106e-06,
+      "loss": 1.3878,
+      "step": 1705
+    },
+    {
+      "epoch": 0.48108032072021384,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.990001818468968e-06,
+      "loss": 1.658,
+      "step": 1710
+    },
+    {
+      "epoch": 0.48248698832465886,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.989651796916918e-06,
+      "loss": 1.9873,
+      "step": 1715
+    },
+    {
+      "epoch": 0.48389365592910394,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.98929576138458e-06,
+      "loss": 1.75,
+      "step": 1720
+    },
+    {
+      "epoch": 0.485300323533549,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.98893371240864e-06,
+      "loss": 1.5848,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4867069911379941,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.988565650534847e-06,
+      "loss": 1.6315,
+      "step": 1730
+    },
+    {
+      "epoch": 0.48811365874243917,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.988191576318015e-06,
+      "loss": 1.613,
+      "step": 1735
+    },
+    {
+      "epoch": 0.48952032634688425,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.987811490322025e-06,
+      "loss": 1.4464,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4909269939513293,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.987425393119813e-06,
+      "loss": 1.7572,
+      "step": 1745
+    },
+    {
+      "epoch": 0.49233366155577435,
+      "grad_norm": 3.625,
+      "learning_rate": 7.987033285293382e-06,
+      "loss": 1.5372,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4937403291602194,
+      "grad_norm": 9.3125,
+      "learning_rate": 7.986635167433794e-06,
+      "loss": 1.8296,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4951469967646645,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.986231040141167e-06,
+      "loss": 1.8108,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4965536643691096,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.985820904024682e-06,
+      "loss": 1.6946,
+      "step": 1765
+    },
+    {
+      "epoch": 0.49796033197355466,
+      "grad_norm": 3.25,
+      "learning_rate": 7.985404759702576e-06,
+      "loss": 1.5829,
+      "step": 1770
+    },
+    {
+      "epoch": 0.49936699957799974,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.984982607802143e-06,
+      "loss": 1.7967,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5007736671824448,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.984554448959733e-06,
+      "loss": 1.7127,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5021803347868898,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.984120283820747e-06,
+      "loss": 1.7665,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5035870023913349,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.983680113039648e-06,
+      "loss": 1.6801,
+      "step": 1790
+    },
+    {
+      "epoch": 0.50499366999578,
+      "grad_norm": 5.15625,
+      "learning_rate": 7.983233937279946e-06,
+      "loss": 1.7679,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5064003376002251,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.982781757214201e-06,
+      "loss": 1.5918,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5078070052046701,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.982323573524031e-06,
+      "loss": 1.5204,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5092136728091152,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.981859386900095e-06,
+      "loss": 1.791,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5106203404135603,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.98138919804211e-06,
+      "loss": 1.6106,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5120270080180054,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.980913007658834e-06,
+      "loss": 1.6606,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5134336756224505,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.980430816468074e-06,
+      "loss": 1.5026,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5148403432268955,
+      "grad_norm": 9.0,
+      "learning_rate": 7.979942625196683e-06,
+      "loss": 1.4711,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5162470108313405,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.979448434580558e-06,
+      "loss": 1.8645,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5176536784357856,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.978948245364639e-06,
+      "loss": 1.7167,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5190603460402307,
+      "grad_norm": 4.125,
+      "learning_rate": 7.97844205830291e-06,
+      "loss": 1.6777,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5204670136446757,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.977929874158391e-06,
+      "loss": 1.7533,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5218736812491208,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.97741169370315e-06,
+      "loss": 1.7477,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5232803488535659,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.976887517718287e-06,
+      "loss": 1.8136,
+      "step": 1860
+    },
+    {
+      "epoch": 0.524687016458011,
+      "grad_norm": 3.25,
+      "learning_rate": 7.976357346993943e-06,
+      "loss": 1.5982,
+      "step": 1865
+    },
+    {
+      "epoch": 0.526093684062456,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.975821182329293e-06,
+      "loss": 1.6659,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5275003516669011,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.975279024532551e-06,
+      "loss": 1.7181,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5289070192713462,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.974730874420964e-06,
+      "loss": 1.7149,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5303136868757913,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.974176732820807e-06,
+      "loss": 1.5102,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5317203544802364,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.973616600567391e-06,
+      "loss": 1.7282,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5331270220846814,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.973050478505058e-06,
+      "loss": 1.5252,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5345336896891265,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.972478367487176e-06,
+      "loss": 1.6819,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5359403572935715,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.971900268376144e-06,
+      "loss": 1.5836,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5373470248980166,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.971316182043384e-06,
+      "loss": 1.6865,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5387536925024616,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.970726109369344e-06,
+      "loss": 1.6588,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5401603601069067,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.970130051243498e-06,
+      "loss": 1.7915,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5415670277113518,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.969528008564342e-06,
+      "loss": 1.7502,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5429736953157969,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.96891998223939e-06,
+      "loss": 1.5522,
+      "step": 1930
+    },
+    {
+      "epoch": 0.544380362920242,
+      "grad_norm": 3.625,
+      "learning_rate": 7.968305973185177e-06,
+      "loss": 1.8124,
+      "step": 1935
+    },
+    {
+      "epoch": 0.545787030524687,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.96768598232726e-06,
+      "loss": 1.7465,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5471936981291321,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.967060010600207e-06,
+      "loss": 1.7834,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5486003657335772,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.966428058947607e-06,
+      "loss": 1.5552,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5500070333380223,
+      "grad_norm": 5.0625,
+      "learning_rate": 7.965790128322056e-06,
+      "loss": 1.6003,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5514137009424673,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.965146219685173e-06,
+      "loss": 1.6883,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5528203685469124,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.96449633400758e-06,
+      "loss": 1.5172,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5542270361513575,
+      "grad_norm": 2.75,
+      "learning_rate": 7.963840472268913e-06,
+      "loss": 1.5894,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5556337037558025,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.963178635457812e-06,
+      "loss": 1.5496,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5570403713602475,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.962510824571927e-06,
+      "loss": 1.8202,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5584470389646926,
+      "grad_norm": 3.0,
+      "learning_rate": 7.961837040617912e-06,
+      "loss": 1.6368,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5598537065691377,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.961157284611427e-06,
+      "loss": 1.7324,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5612603741735828,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.960471557577132e-06,
+      "loss": 1.5617,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5626670417780278,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.959779860548688e-06,
+      "loss": 1.7674,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5640737093824729,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.959082194568757e-06,
+      "loss": 1.8521,
+      "step": 2005
+    },
+    {
+      "epoch": 0.565480376986918,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.958378560688997e-06,
+      "loss": 1.605,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5668870445913631,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.957668959970058e-06,
+      "loss": 1.6868,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5682937121958082,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.956953393481593e-06,
+      "loss": 1.9079,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5697003798002532,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.956231862302242e-06,
+      "loss": 1.6886,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5711070474046983,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.955504367519637e-06,
+      "loss": 1.6482,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5725137150091434,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.954770910230399e-06,
+      "loss": 1.9038,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5739203826135885,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.954031491540138e-06,
+      "loss": 1.7288,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5753270502180334,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.953286112563452e-06,
+      "loss": 1.6836,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5767337178224785,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.952534774423918e-06,
+      "loss": 1.6717,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5781403854269236,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.951777478254102e-06,
+      "loss": 1.6014,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5795470530313687,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.951014225195548e-06,
+      "loss": 1.4636,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5809537206358137,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.950245016398778e-06,
+      "loss": 1.4488,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5823603882402588,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.949469853023294e-06,
+      "loss": 1.5397,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5837670558447039,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.948688736237573e-06,
+      "loss": 1.751,
+      "step": 2075
+    },
+    {
+      "epoch": 0.585173723449149,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.947901667219067e-06,
+      "loss": 1.7123,
+      "step": 2080
+    },
+    {
+      "epoch": 0.586580391053594,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.9471086471542e-06,
+      "loss": 1.7476,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5879870586580391,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.946309677238364e-06,
+      "loss": 1.8185,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5893937262624842,
+      "grad_norm": 3.625,
+      "learning_rate": 7.945504758675926e-06,
+      "loss": 1.5302,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5908003938669293,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.944693892680213e-06,
+      "loss": 1.4795,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5922070614713744,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.943877080473521e-06,
+      "loss": 1.4504,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5936137290758193,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.94305432328711e-06,
+      "loss": 1.636,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5950203966802644,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.942225622361197e-06,
+      "loss": 1.6305,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5964270642847095,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.941390978944963e-06,
+      "loss": 1.7123,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5978337318891546,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.940550394296545e-06,
+      "loss": 1.7594,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5992403994935996,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.939703869683038e-06,
+      "loss": 1.5839,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6006470670980447,
+      "grad_norm": 2.875,
+      "learning_rate": 7.938851406380484e-06,
+      "loss": 1.5178,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6020537347024898,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.937993005673886e-06,
+      "loss": 1.7266,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6034604023069349,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.93712866885719e-06,
+      "loss": 1.5003,
+      "step": 2145
+    },
+    {
+      "epoch": 0.60486706991138,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.936258397233296e-06,
+      "loss": 1.6785,
+      "step": 2150
+    },
+    {
+      "epoch": 0.606273737515825,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.935382192114043e-06,
+      "loss": 1.7834,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6076804051202701,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.93450005482022e-06,
+      "loss": 1.7403,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6090870727247152,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.933611986681556e-06,
+      "loss": 1.7666,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6104937403291603,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.93271798903672e-06,
+      "loss": 1.5542,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6119004079336053,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.931818063233322e-06,
+      "loss": 1.8542,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6133070755380503,
+      "grad_norm": 4.75,
+      "learning_rate": 7.930912210627902e-06,
+      "loss": 1.5718,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6147137431424954,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.930000432585939e-06,
+      "loss": 1.5713,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6161204107469405,
+      "grad_norm": 2.625,
+      "learning_rate": 7.929082730481841e-06,
+      "loss": 1.8829,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6175270783513855,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.928159105698949e-06,
+      "loss": 1.9,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6189337459558306,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.927229559629529e-06,
+      "loss": 1.6255,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6203404135602757,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.926294093674777e-06,
+      "loss": 1.4732,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6217470811647208,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.925352709244804e-06,
+      "loss": 1.6324,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6231537487691658,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.924405407758654e-06,
+      "loss": 1.5333,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6245604163736109,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.923452190644279e-06,
+      "loss": 1.7322,
+      "step": 2220
+    },
+    {
+      "epoch": 0.625967083978056,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.922493059338556e-06,
+      "loss": 1.7649,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6273737515825011,
+      "grad_norm": 2.75,
+      "learning_rate": 7.921528015287276e-06,
+      "loss": 1.6691,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6287804191869462,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.920557059945137e-06,
+      "loss": 1.7656,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6301870867913912,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.919580194775758e-06,
+      "loss": 1.7602,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6315937543958363,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.918597421251656e-06,
+      "loss": 1.7364,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6330004220002813,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.917608740854259e-06,
+      "loss": 1.6754,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6344070896047264,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.9166141550739e-06,
+      "loss": 1.5354,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6358137572091714,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.915613665409813e-06,
+      "loss": 1.532,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6372204248136165,
+      "grad_norm": 2.25,
+      "learning_rate": 7.914607273370129e-06,
+      "loss": 1.6626,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6386270924180616,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.913594980471877e-06,
+      "loss": 1.7334,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6400337600225067,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.912576788240987e-06,
+      "loss": 1.537,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6414404276269517,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.911552698212271e-06,
+      "loss": 1.7401,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6428470952313968,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.910522711929444e-06,
+      "loss": 1.7289,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6442537628358419,
+      "grad_norm": 3.125,
+      "learning_rate": 7.909486830945092e-06,
+      "loss": 1.5732,
+      "step": 2290
+    },
+    {
+      "epoch": 0.645660430440287,
+      "grad_norm": 2.875,
+      "learning_rate": 7.908445056820707e-06,
+      "loss": 1.7419,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6470670980447321,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.907397391126647e-06,
+      "loss": 1.6438,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6484737656491771,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.906343835442159e-06,
+      "loss": 1.3731,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6498804332536222,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.90528439135537e-06,
+      "loss": 1.6436,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6512871008580673,
+      "grad_norm": 4.125,
+      "learning_rate": 7.904219060463277e-06,
+      "loss": 1.8662,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6526937684625123,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.903147844371757e-06,
+      "loss": 1.7982,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6541004360669573,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.902070744695553e-06,
+      "loss": 1.6941,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6555071036714024,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.900987763058281e-06,
+      "loss": 1.8189,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6569137712758475,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.899898901092425e-06,
+      "loss": 1.6437,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6583204388802926,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.898804160439322e-06,
+      "loss": 1.5489,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6597271064847376,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.897703542749186e-06,
+      "loss": 1.5735,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6611337740891827,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.896597049681078e-06,
+      "loss": 1.561,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6625404416936278,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.895484682902921e-06,
+      "loss": 1.8226,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6639471092980729,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.89436644409149e-06,
+      "loss": 1.6574,
+      "step": 2360
+    },
+    {
+      "epoch": 0.665353776902518,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.893242334932415e-06,
+      "loss": 1.4988,
+      "step": 2365
+    },
+    {
+      "epoch": 0.666760444506963,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.892112357120171e-06,
+      "loss": 1.7978,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6681671121114081,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.890976512358079e-06,
+      "loss": 1.6548,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6695737797158532,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.889834802358309e-06,
+      "loss": 1.6971,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6709804473202983,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.888687228841864e-06,
+      "loss": 1.5706,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6723871149247432,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.887533793538594e-06,
+      "loss": 1.4289,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6737937825291883,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.886374498187178e-06,
+      "loss": 1.7071,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6752004501336334,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.885209344535135e-06,
+      "loss": 1.8025,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6766071177380785,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.884038334338812e-06,
+      "loss": 1.6936,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6780137853425235,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.88286146936338e-06,
+      "loss": 1.8952,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6794204529469686,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.881678751382842e-06,
+      "loss": 1.5186,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6808271205514137,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.880490182180022e-06,
+      "loss": 1.9398,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6822337881558588,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.879295763546558e-06,
+      "loss": 1.7953,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6836404557603039,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.878095497282916e-06,
+      "loss": 1.8955,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6850471233647489,
+      "grad_norm": 3.25,
+      "learning_rate": 7.876889385198367e-06,
+      "loss": 1.5763,
+      "step": 2435
+    },
+    {
+      "epoch": 0.686453790969194,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.875677429111e-06,
+      "loss": 1.5909,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6878604585736391,
+      "grad_norm": 5.9375,
+      "learning_rate": 7.874459630847711e-06,
+      "loss": 1.6029,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6892671261780842,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.873235992244203e-06,
+      "loss": 1.4505,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6906737937825291,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.872006515144983e-06,
+      "loss": 1.8632,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6920804613869742,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.870771201403356e-06,
+      "loss": 1.6993,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6934871289914193,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.86953005288143e-06,
+      "loss": 1.803,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6948937965958644,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.868283071450105e-06,
+      "loss": 1.7066,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6963004642003094,
+      "grad_norm": 3.875,
+      "learning_rate": 7.867030258989072e-06,
+      "loss": 1.6787,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6977071318047545,
+      "grad_norm": 3.375,
+      "learning_rate": 7.865771617386817e-06,
+      "loss": 1.5385,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6991137994091996,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.86450714854061e-06,
+      "loss": 1.6667,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7005204670136447,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.863236854356502e-06,
+      "loss": 1.6079,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7019271346180898,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.861960736749331e-06,
+      "loss": 1.7048,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7033338022225348,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.860678797642707e-06,
+      "loss": 1.855,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7047404698269799,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.859391038969021e-06,
+      "loss": 1.8016,
+      "step": 2505
+    },
+    {
+      "epoch": 0.706147137431425,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.858097462669432e-06,
+      "loss": 1.8948,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7075538050358701,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.85679807069387e-06,
+      "loss": 1.5018,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7089604726403151,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.855492865001033e-06,
+      "loss": 1.7206,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7103671402447601,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.85418184755838e-06,
+      "loss": 1.5677,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7117738078492052,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.852865020342133e-06,
+      "loss": 1.7892,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7131804754536503,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.851542385337269e-06,
+      "loss": 1.3885,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7145871430580953,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.850213944537522e-06,
+      "loss": 1.6664,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7159938106625404,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.848879699945377e-06,
+      "loss": 1.5967,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7174004782669855,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.847539653572066e-06,
+      "loss": 1.5588,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7188071458714306,
+      "grad_norm": 4.9375,
+      "learning_rate": 7.846193807437571e-06,
+      "loss": 1.7426,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7202138134758757,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.84484216357061e-06,
+      "loss": 1.7578,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7216204810803207,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.843484724008645e-06,
+      "loss": 1.5375,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7230271486847658,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.842121490797876e-06,
+      "loss": 1.4915,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7244338162892109,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.840752465993228e-06,
+      "loss": 1.385,
+      "step": 2575
+    },
+    {
+      "epoch": 0.725840483893656,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.839377651658368e-06,
+      "loss": 1.3509,
+      "step": 2580
+    },
+    {
+      "epoch": 0.727247151498101,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.837997049865677e-06,
+      "loss": 1.6331,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7286538191025461,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.836610662696273e-06,
+      "loss": 1.785,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7300604867069911,
+      "grad_norm": 4.0,
+      "learning_rate": 7.835218492239987e-06,
+      "loss": 1.7578,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7314671543114362,
+      "grad_norm": 3.875,
+      "learning_rate": 7.833820540595369e-06,
+      "loss": 1.7416,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7328738219158812,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.832416809869684e-06,
+      "loss": 1.7128,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7342804895203263,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.831007302178908e-06,
+      "loss": 1.7317,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7356871571247714,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.829592019647729e-06,
+      "loss": 1.365,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7370938247292165,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.82817096440953e-06,
+      "loss": 1.7279,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7385004923336616,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.826744138606408e-06,
+      "loss": 1.2845,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7399071599381066,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.825311544389149e-06,
+      "loss": 1.5838,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7413138275425517,
+      "grad_norm": 2.75,
+      "learning_rate": 7.82387318391724e-06,
+      "loss": 1.6606,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7427204951469968,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.822429059358859e-06,
+      "loss": 1.767,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7441271627514419,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.820979172890869e-06,
+      "loss": 1.6674,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7455338303558869,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.819523526698824e-06,
+      "loss": 1.7634,
+      "step": 2650
+    },
+    {
+      "epoch": 0.746940497960332,
+      "grad_norm": 3.25,
+      "learning_rate": 7.818062122976954e-06,
+      "loss": 1.6022,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7483471655647771,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.816594963928176e-06,
+      "loss": 1.6332,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7497538331692221,
+      "grad_norm": 5.9375,
+      "learning_rate": 7.815122051764075e-06,
+      "loss": 1.711,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7511605007736671,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.813643388704912e-06,
+      "loss": 1.4206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7525671683781122,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.812158976979614e-06,
+      "loss": 1.5857,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7539738359825573,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.810668818825778e-06,
+      "loss": 1.6887,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7553805035870024,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.80917291648966e-06,
+      "loss": 1.4489,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7567871711914474,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.807671272226175e-06,
+      "loss": 1.7821,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7581938387958925,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.806163888298894e-06,
+      "loss": 1.449,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7596005064003376,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.80465076698004e-06,
+      "loss": 1.7329,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7610071740047827,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.80313191055048e-06,
+      "loss": 1.5321,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7624138416092278,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.801607321299738e-06,
+      "loss": 1.3949,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7638205092136728,
+      "grad_norm": 3.25,
+      "learning_rate": 7.800077001525966e-06,
+      "loss": 1.6693,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7652271768181179,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.798540953535962e-06,
+      "loss": 1.3889,
+      "step": 2720
+    },
+    {
+      "epoch": 0.766633844422563,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.796999179645157e-06,
+      "loss": 1.8232,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7680405120270081,
+      "grad_norm": 3.0,
+      "learning_rate": 7.795451682177613e-06,
+      "loss": 1.5121,
+      "step": 2730
+    },
+    {
+      "epoch": 0.769447179631453,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.793898463466018e-06,
+      "loss": 1.7762,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7708538472358981,
+      "grad_norm": 3.375,
+      "learning_rate": 7.792339525851686e-06,
+      "loss": 1.6207,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7722605148403432,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.790774871684554e-06,
+      "loss": 1.6523,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7736671824447883,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.789204503323172e-06,
+      "loss": 1.8693,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7750738500492333,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.787628423134702e-06,
+      "loss": 1.5125,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7764805176536784,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.786046633494924e-06,
+      "loss": 1.4513,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7778871852581235,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.784459136788217e-06,
+      "loss": 1.7672,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7792938528625686,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.782865935407566e-06,
+      "loss": 1.6418,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7807005204670137,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.781267031754553e-06,
+      "loss": 1.6256,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7821071880714587,
+      "grad_norm": 3.25,
+      "learning_rate": 7.779662428239359e-06,
+      "loss": 1.4991,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7835138556759038,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.778052127280754e-06,
+      "loss": 1.7964,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7849205232803489,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.776436131306096e-06,
+      "loss": 1.5225,
+      "step": 2790
+    },
+    {
+      "epoch": 0.786327190884794,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.774814442751332e-06,
+      "loss": 1.5578,
+      "step": 2795
+    },
+    {
+      "epoch": 0.787733858489239,
+      "grad_norm": 3.984375,
+      "learning_rate": 7.773187064060981e-06,
+      "loss": 1.6137,
+      "step": 2800
+    },
+    {
+      "epoch": 0.789140526093684,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.771553997688153e-06,
+      "loss": 1.604,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7905471936981291,
+      "grad_norm": 3.0,
+      "learning_rate": 7.769915246094519e-06,
+      "loss": 1.583,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7919538613025742,
+      "grad_norm": 5.0,
+      "learning_rate": 7.768270811750326e-06,
+      "loss": 1.613,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7933605289070192,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.766620697134385e-06,
+      "loss": 1.6581,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7947671965114643,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.76496490473407e-06,
+      "loss": 1.5417,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7961738641159094,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.763303437045313e-06,
+      "loss": 1.9133,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7975805317203545,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.761636296572605e-06,
+      "loss": 1.8703,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7989871993247996,
+      "grad_norm": 4.9375,
+      "learning_rate": 7.759963485828982e-06,
+      "loss": 1.5487,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8003938669292446,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.75828500733603e-06,
+      "loss": 1.7283,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8018005345336897,
+      "grad_norm": 3.875,
+      "learning_rate": 7.75660086362388e-06,
+      "loss": 1.4787,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8032072021381348,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.754911057231202e-06,
+      "loss": 1.6337,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8046138697425799,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.7532155907052e-06,
+      "loss": 1.6622,
+      "step": 2860
+    },
+    {
+      "epoch": 0.806020537347025,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.751514466601611e-06,
+      "loss": 1.8183,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8074272049514699,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.749807687484702e-06,
+      "loss": 1.5107,
+      "step": 2870
+    },
+    {
+      "epoch": 0.808833872555915,
+      "grad_norm": 3.5,
+      "learning_rate": 7.748095255927262e-06,
+      "loss": 1.3176,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8102405401603601,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.746377174510603e-06,
+      "loss": 1.7498,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8116472077648051,
+      "grad_norm": 2.75,
+      "learning_rate": 7.74465344582455e-06,
+      "loss": 1.7046,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8130538753692502,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.742924072467442e-06,
+      "loss": 1.6646,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8144605429736953,
+      "grad_norm": 3.25,
+      "learning_rate": 7.74118905704613e-06,
+      "loss": 1.5339,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8158672105781404,
+      "grad_norm": 3.375,
+      "learning_rate": 7.739448402175967e-06,
+      "loss": 1.7851,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8172738781825855,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.737702110480804e-06,
+      "loss": 1.7214,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8186805457870305,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.735950184592994e-06,
+      "loss": 1.7116,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8200872133914756,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.734192627153382e-06,
+      "loss": 1.761,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8214938809959207,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.732429440811297e-06,
+      "loss": 1.6123,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8229005486003658,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.730660628224563e-06,
+      "loss": 1.6235,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8243072162048108,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.728886192059474e-06,
+      "loss": 1.8165,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8257138838092559,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.727106134990808e-06,
+      "loss": 1.5903,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8271205514137009,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.725320459701813e-06,
+      "loss": 1.3788,
+      "step": 2940
+    },
+    {
+      "epoch": 0.828527219018146,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.723529168884205e-06,
+      "loss": 1.6623,
+      "step": 2945
+    },
+    {
+      "epoch": 0.829933886622591,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.72173226523817e-06,
+      "loss": 1.6399,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8313405542270361,
+      "grad_norm": 3.375,
+      "learning_rate": 7.719929751472348e-06,
+      "loss": 1.677,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8327472218314812,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.71812163030384e-06,
+      "loss": 1.3748,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8341538894359263,
+      "grad_norm": 5.59375,
+      "learning_rate": 7.7163079044582e-06,
+      "loss": 1.7135,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8355605570403714,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.714488576669427e-06,
+      "loss": 1.7007,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8369672246448164,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.712663649679966e-06,
+      "loss": 1.5469,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8383738922492615,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.710833126240702e-06,
+      "loss": 1.5403,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8397805598537066,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.70899700911096e-06,
+      "loss": 1.5021,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8411872274581517,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.707155301058488e-06,
+      "loss": 1.7806,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8425938950625967,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.705308004859471e-06,
+      "loss": 1.5206,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8440005626670418,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.703455123298512e-06,
+      "loss": 1.5729,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8454072302714869,
+      "grad_norm": 3.75,
+      "learning_rate": 7.701596659168637e-06,
+      "loss": 1.6561,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8468138978759319,
+      "grad_norm": 4.125,
+      "learning_rate": 7.699732615271283e-06,
+      "loss": 1.4492,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8482205654803769,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.697862994416301e-06,
+      "loss": 1.7029,
+      "step": 3015
+    },
+    {
+      "epoch": 0.849627233084822,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.695987799421947e-06,
+      "loss": 1.4796,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8510339006892671,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.694107033114882e-06,
+      "loss": 1.6716,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8524405682937122,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.692220698330161e-06,
+      "loss": 1.6725,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8538472358981573,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.690328797911235e-06,
+      "loss": 1.7419,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8552539035026023,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.688431334709947e-06,
+      "loss": 1.4105,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8566605711070474,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.686528311586523e-06,
+      "loss": 1.2964,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8580672387114925,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.684619731409566e-06,
+      "loss": 1.4006,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8594739063159376,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.682705597056066e-06,
+      "loss": 1.6018,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8608805739203826,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.680785911411375e-06,
+      "loss": 1.8649,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8622872415248277,
+      "grad_norm": 3.125,
+      "learning_rate": 7.678860677369218e-06,
+      "loss": 1.5029,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8636939091292728,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.676929897831684e-06,
+      "loss": 1.4425,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8651005767337179,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.674993575709218e-06,
+      "loss": 1.7859,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8665072443381628,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.673051713920624e-06,
+      "loss": 1.7595,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8679139119426079,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.671104315393053e-06,
+      "loss": 1.3791,
+      "step": 3085
+    },
+    {
+      "epoch": 0.869320579547053,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.669151383062003e-06,
+      "loss": 1.8302,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8707272471514981,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.667192919871313e-06,
+      "loss": 1.6029,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8721339147559432,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.665228928773164e-06,
+      "loss": 1.7232,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8735405823603882,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.663259412728062e-06,
+      "loss": 1.6841,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8749472499648333,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.661284374704848e-06,
+      "loss": 1.5463,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8763539175692784,
+      "grad_norm": 3.0,
+      "learning_rate": 7.659303817680682e-06,
+      "loss": 1.6332,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8777605851737235,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.657317744641047e-06,
+      "loss": 1.9633,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8791672527781685,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.655326158579739e-06,
+      "loss": 1.6777,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8805739203826136,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.65332906249886e-06,
+      "loss": 1.5168,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8819805879870587,
+      "grad_norm": 3.0,
+      "learning_rate": 7.65132645940883e-06,
+      "loss": 1.8697,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8833872555915038,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.649318352328356e-06,
+      "loss": 1.3569,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8847939231959489,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.647304744284452e-06,
+      "loss": 1.6584,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8862005908003938,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.645285638312418e-06,
+      "loss": 1.6259,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8876072584048389,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.643261037455844e-06,
+      "loss": 1.7632,
+      "step": 3155
+    },
+    {
+      "epoch": 0.889013926009284,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.641230944766605e-06,
+      "loss": 1.5457,
+      "step": 3160
+    },
+    {
+      "epoch": 0.890420593613729,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.63919536330485e-06,
+      "loss": 1.599,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8918272612181741,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.637154296139003e-06,
+      "loss": 1.733,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8932339288226192,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.63510774634576e-06,
+      "loss": 1.6175,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8946405964270643,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.633055717010078e-06,
+      "loss": 1.7642,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8960472640315094,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.630998211225177e-06,
+      "loss": 1.5858,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8974539316359544,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.62893523209253e-06,
+      "loss": 1.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8988605992403995,
+      "grad_norm": 2.75,
+      "learning_rate": 7.62686678272186e-06,
+      "loss": 1.7159,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9002672668448446,
+      "grad_norm": 4.5,
+      "learning_rate": 7.624792866231137e-06,
+      "loss": 1.4671,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9016739344492897,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.622713485746573e-06,
+      "loss": 1.6495,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9030806020537347,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.620628644402613e-06,
+      "loss": 1.4105,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9044872696581798,
+      "grad_norm": 3.5,
+      "learning_rate": 7.618538345341938e-06,
+      "loss": 1.55,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9058939372626248,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.6164425917154545e-06,
+      "loss": 1.4803,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9073006048670699,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.614341386682289e-06,
+      "loss": 1.4704,
+      "step": 3225
+    },
+    {
+      "epoch": 0.908707272471515,
+      "grad_norm": 6.6875,
+      "learning_rate": 7.612234733409786e-06,
+      "loss": 1.7306,
+      "step": 3230
+    },
+    {
+      "epoch": 0.91011394007596,
+      "grad_norm": 4.25,
+      "learning_rate": 7.610122635073507e-06,
+      "loss": 1.7463,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9115206076804051,
+      "grad_norm": 3.5,
+      "learning_rate": 7.608005094857213e-06,
+      "loss": 1.7264,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9129272752848502,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.60588211595288e-06,
+      "loss": 1.5772,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9143339428892953,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.603753701560669e-06,
+      "loss": 1.7315,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9157406104937403,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.6016198548889446e-06,
+      "loss": 1.5967,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9171472780981854,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.599480579154253e-06,
+      "loss": 1.5748,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9185539457026305,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.59733587758133e-06,
+      "loss": 1.6969,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9199606133070756,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.595185753403086e-06,
+      "loss": 1.3966,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9213672809115206,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.593030209860608e-06,
+      "loss": 1.8503,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9227739485159657,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.590869250203151e-06,
+      "loss": 1.4599,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9241806161204107,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.588702877688133e-06,
+      "loss": 1.7539,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9255872837248558,
+      "grad_norm": 4.75,
+      "learning_rate": 7.586531095581135e-06,
+      "loss": 1.6305,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9269939513293008,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.584353907155886e-06,
+      "loss": 1.9388,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9284006189337459,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.5821713156942725e-06,
+      "loss": 1.6272,
+      "step": 3300
+    },
+    {
+      "epoch": 0.929807286538191,
+      "grad_norm": 3.75,
+      "learning_rate": 7.57998332448632e-06,
+      "loss": 1.6722,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9312139541426361,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.577789936830194e-06,
+      "loss": 1.533,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9326206217470812,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.575591156032198e-06,
+      "loss": 1.6827,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9340272893515262,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.573386985406761e-06,
+      "loss": 1.5372,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9354339569559713,
+      "grad_norm": 2.75,
+      "learning_rate": 7.571177428276439e-06,
+      "loss": 1.749,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9368406245604164,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.568962487971905e-06,
+      "loss": 1.5339,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9382472921648615,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.56674216783195e-06,
+      "loss": 1.416,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9396539597693065,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.564516471203474e-06,
+      "loss": 1.6917,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9410606273737516,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.562285401441478e-06,
+      "loss": 1.6447,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9424672949781967,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.560048961909068e-06,
+      "loss": 1.7682,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9438739625826417,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.5578071559774384e-06,
+      "loss": 1.5933,
+      "step": 3355
+    },
+    {
+      "epoch": 0.9452806301870867,
+      "grad_norm": 3.5,
+      "learning_rate": 7.555559987025878e-06,
+      "loss": 1.7298,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9466872977915318,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.553307458441755e-06,
+      "loss": 1.7833,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9480939653959769,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.551049573620521e-06,
+      "loss": 1.7445,
+      "step": 3370
+    },
+    {
+      "epoch": 0.949500633000422,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.5487863359656994e-06,
+      "loss": 1.3276,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9509073006048671,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.546517748888882e-06,
+      "loss": 1.7944,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9523139682093121,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.544243815809729e-06,
+      "loss": 1.6007,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9537206358137572,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.54196454015595e-06,
+      "loss": 1.6207,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9551273034182023,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.539679925363316e-06,
+      "loss": 1.534,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9565339710226474,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.5373899748756435e-06,
+      "loss": 1.5586,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9579406386270924,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.53509469214479e-06,
+      "loss": 1.3578,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9593473062315375,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.532794080630655e-06,
+      "loss": 1.7003,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9607539738359826,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.530488143801166e-06,
+      "loss": 1.6732,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9621606414404277,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.528176885132283e-06,
+      "loss": 1.675,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9635673090448726,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.525860308107983e-06,
+      "loss": 1.6125,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9649739766493177,
+      "grad_norm": 2.875,
+      "learning_rate": 7.523538416220264e-06,
+      "loss": 1.621,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9663806442537628,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.52121121296913e-06,
+      "loss": 1.7444,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9677873118582079,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.518878701862599e-06,
+      "loss": 1.8158,
+      "step": 3440
+    },
+    {
+      "epoch": 0.969193979462653,
+      "grad_norm": 3.0,
+      "learning_rate": 7.5165408864166845e-06,
+      "loss": 1.6573,
+      "step": 3445
+    },
+    {
+      "epoch": 0.970600647067098,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.514197770155398e-06,
+      "loss": 1.5889,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9720073146715431,
+      "grad_norm": 3.0,
+      "learning_rate": 7.511849356610738e-06,
+      "loss": 1.3515,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9734139822759882,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.5094956493226955e-06,
+      "loss": 1.6675,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9748206498804333,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.507136651839233e-06,
+      "loss": 1.5557,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9762273174848783,
+      "grad_norm": 3.0,
+      "learning_rate": 7.504772367716292e-06,
+      "loss": 1.4733,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9776339850893234,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.5024028005177814e-06,
+      "loss": 1.6102,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9790406526937685,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.500027953815577e-06,
+      "loss": 1.5866,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9804473202982136,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.497647831189506e-06,
+      "loss": 1.5003,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9818539879026587,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.495262436227356e-06,
+      "loss": 1.6196,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9832606555071036,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.492871772524859e-06,
+      "loss": 1.6242,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9846673231115487,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.490475843685686e-06,
+      "loss": 1.635,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9860739907159938,
+      "grad_norm": 3.25,
+      "learning_rate": 7.488074653321452e-06,
+      "loss": 1.5958,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9874806583204389,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.485668205051696e-06,
+      "loss": 1.5429,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9888873259248839,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.4832565025038855e-06,
+      "loss": 1.8323,
+      "step": 3515
+    },
+    {
+      "epoch": 0.990293993529329,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.480839549313409e-06,
+      "loss": 1.6328,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9917006611337741,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.478417349123569e-06,
+      "loss": 1.6957,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9931073287382192,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.475989905585578e-06,
+      "loss": 1.4479,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9945139963426642,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.473557222358551e-06,
+      "loss": 1.5085,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9959206639471093,
+      "grad_norm": 4.375,
+      "learning_rate": 7.471119303109502e-06,
+      "loss": 1.5398,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9973273315515544,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.468676151513339e-06,
+      "loss": 1.503,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9987339991559995,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.4662277712528536e-06,
+      "loss": 1.6194,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9998593332395554,
+      "eval_loss": 1.611290454864502,
+      "eval_runtime": 329.7515,
+      "eval_samples_per_second": 9.577,
+      "eval_steps_per_second": 4.788,
+      "step": 3554
+    },
+    {
+      "epoch": 1.0001406667604444,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.463774166018723e-06,
+      "loss": 1.6976,
+      "step": 3555
+    },
+    {
+      "epoch": 1.0015473343648895,
+      "grad_norm": 3.25,
+      "learning_rate": 7.461315339509499e-06,
+      "loss": 1.6923,
+      "step": 3560
+    },
+    {
+      "epoch": 1.0029540019693346,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.458851295431601e-06,
+      "loss": 1.5505,
+      "step": 3565
+    },
+    {
+      "epoch": 1.0043606695737797,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.456382037499322e-06,
+      "loss": 1.5004,
+      "step": 3570
+    },
+    {
+      "epoch": 1.0057673371782248,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.453907569434804e-06,
+      "loss": 1.5441,
+      "step": 3575
+    },
+    {
+      "epoch": 1.0071740047826698,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.451427894968049e-06,
+      "loss": 1.4524,
+      "step": 3580
+    },
+    {
+      "epoch": 1.008580672387115,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.448943017836903e-06,
+      "loss": 1.5271,
+      "step": 3585
+    },
+    {
+      "epoch": 1.00998733999156,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.44645294178706e-06,
+      "loss": 1.5947,
+      "step": 3590
+    },
+    {
+      "epoch": 1.011394007596005,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.443957670572046e-06,
+      "loss": 1.7221,
+      "step": 3595
+    },
+    {
+      "epoch": 1.0128006752004501,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.4414572079532205e-06,
+      "loss": 1.538,
+      "step": 3600
+    },
+    {
+      "epoch": 1.0142073428048952,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.438951557699767e-06,
+      "loss": 1.4715,
+      "step": 3605
+    },
+    {
+      "epoch": 1.0156140104093403,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.436440723588688e-06,
+      "loss": 1.5015,
+      "step": 3610
+    },
+    {
+      "epoch": 1.0170206780137854,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.433924709404806e-06,
+      "loss": 1.5526,
+      "step": 3615
+    },
+    {
+      "epoch": 1.0184273456182305,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.4314035189407436e-06,
+      "loss": 1.2446,
+      "step": 3620
+    },
+    {
+      "epoch": 1.0198340132226755,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.428877155996934e-06,
+      "loss": 1.6687,
+      "step": 3625
+    },
+    {
+      "epoch": 1.0212406808271206,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.4263456243816e-06,
+      "loss": 1.3963,
+      "step": 3630
+    },
+    {
+      "epoch": 1.0226473484315657,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.42380892791076e-06,
+      "loss": 1.606,
+      "step": 3635
+    },
+    {
+      "epoch": 1.0240540160360108,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.421267070408218e-06,
+      "loss": 1.6143,
+      "step": 3640
+    },
+    {
+      "epoch": 1.0254606836404558,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.418720055705556e-06,
+      "loss": 1.6649,
+      "step": 3645
+    },
+    {
+      "epoch": 1.026867351244901,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.416167887642132e-06,
+      "loss": 1.6411,
+      "step": 3650
+    },
+    {
+      "epoch": 1.028274018849346,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.413610570065069e-06,
+      "loss": 1.7296,
+      "step": 3655
+    },
+    {
+      "epoch": 1.029680686453791,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.411048106829253e-06,
+      "loss": 1.6742,
+      "step": 3660
+    },
+    {
+      "epoch": 1.031087354058236,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.408480501797333e-06,
+      "loss": 1.4126,
+      "step": 3665
+    },
+    {
+      "epoch": 1.032494021662681,
+      "grad_norm": 5.0625,
+      "learning_rate": 7.405907758839698e-06,
+      "loss": 1.6467,
+      "step": 3670
+    },
+    {
+      "epoch": 1.033900689267126,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.403329881834489e-06,
+      "loss": 1.4774,
+      "step": 3675
+    },
+    {
+      "epoch": 1.0353073568715712,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.400746874667586e-06,
+      "loss": 1.6924,
+      "step": 3680
+    },
+    {
+      "epoch": 1.0367140244760162,
+      "grad_norm": 3.375,
+      "learning_rate": 7.398158741232598e-06,
+      "loss": 1.635,
+      "step": 3685
+    },
+    {
+      "epoch": 1.0381206920804613,
+      "grad_norm": 5.84375,
+      "learning_rate": 7.395565485430866e-06,
+      "loss": 1.4753,
+      "step": 3690
+    },
+    {
+      "epoch": 1.0395273596849064,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.392967111171448e-06,
+      "loss": 1.4399,
+      "step": 3695
+    },
+    {
+      "epoch": 1.0409340272893515,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.390363622371122e-06,
+      "loss": 1.6359,
+      "step": 3700
+    },
+    {
+      "epoch": 1.0423406948937965,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.387755022954373e-06,
+      "loss": 1.6028,
+      "step": 3705
+    },
+    {
+      "epoch": 1.0437473624982416,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.385141316853388e-06,
+      "loss": 1.4053,
+      "step": 3710
+    },
+    {
+      "epoch": 1.0451540301026867,
+      "grad_norm": 2.875,
+      "learning_rate": 7.382522508008056e-06,
+      "loss": 1.5238,
+      "step": 3715
+    },
+    {
+      "epoch": 1.0465606977071318,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.379898600365956e-06,
+      "loss": 1.7097,
+      "step": 3720
+    },
+    {
+      "epoch": 1.0479673653115769,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.377269597882351e-06,
+      "loss": 1.5788,
+      "step": 3725
+    },
+    {
+      "epoch": 1.049374032916022,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.374635504520186e-06,
+      "loss": 1.6983,
+      "step": 3730
+    },
+    {
+      "epoch": 1.050780700520467,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.371996324250083e-06,
+      "loss": 1.5622,
+      "step": 3735
+    },
+    {
+      "epoch": 1.052187368124912,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.369352061050324e-06,
+      "loss": 1.4207,
+      "step": 3740
+    },
+    {
+      "epoch": 1.0535940357293572,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.366702718906859e-06,
+      "loss": 1.5624,
+      "step": 3745
+    },
+    {
+      "epoch": 1.0550007033338022,
+      "grad_norm": 3.75,
+      "learning_rate": 7.364048301813293e-06,
+      "loss": 1.3799,
+      "step": 3750
+    },
+    {
+      "epoch": 1.0564073709382473,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.361388813770881e-06,
+      "loss": 1.781,
+      "step": 3755
+    },
+    {
+      "epoch": 1.0578140385426924,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.35872425878852e-06,
+      "loss": 1.5601,
+      "step": 3760
+    },
+    {
+      "epoch": 1.0592207061471375,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.356054640882747e-06,
+      "loss": 1.5782,
+      "step": 3765
+    },
+    {
+      "epoch": 1.0606273737515826,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.35337996407773e-06,
+      "loss": 1.7065,
+      "step": 3770
+    },
+    {
+      "epoch": 1.0620340413560276,
+      "grad_norm": 3.5,
+      "learning_rate": 7.350700232405263e-06,
+      "loss": 1.7547,
+      "step": 3775
+    },
+    {
+      "epoch": 1.0634407089604727,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.3480154499047585e-06,
+      "loss": 1.797,
+      "step": 3780
+    },
+    {
+      "epoch": 1.0648473765649178,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.345325620623246e-06,
+      "loss": 1.6371,
+      "step": 3785
+    },
+    {
+      "epoch": 1.0662540441693629,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.3426307486153575e-06,
+      "loss": 1.6329,
+      "step": 3790
+    },
+    {
+      "epoch": 1.067660711773808,
+      "grad_norm": 2.625,
+      "learning_rate": 7.339930837943331e-06,
+      "loss": 1.5983,
+      "step": 3795
+    },
+    {
+      "epoch": 1.0690673793782528,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.337225892676997e-06,
+      "loss": 1.6337,
+      "step": 3800
+    },
+    {
+      "epoch": 1.070474046982698,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.334515916893774e-06,
+      "loss": 1.6965,
+      "step": 3805
+    },
+    {
+      "epoch": 1.071880714587143,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.3318009146786695e-06,
+      "loss": 1.5862,
+      "step": 3810
+    },
+    {
+      "epoch": 1.073287382191588,
+      "grad_norm": 3.25,
+      "learning_rate": 7.3290808901242595e-06,
+      "loss": 1.5898,
+      "step": 3815
+    },
+    {
+      "epoch": 1.0746940497960331,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.326355847330698e-06,
+      "loss": 1.5154,
+      "step": 3820
+    },
+    {
+      "epoch": 1.0761007174004782,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.323625790405698e-06,
+      "loss": 1.6193,
+      "step": 3825
+    },
+    {
+      "epoch": 1.0775073850049233,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.320890723464535e-06,
+      "loss": 1.7274,
+      "step": 3830
+    },
+    {
+      "epoch": 1.0789140526093683,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.3181506506300324e-06,
+      "loss": 1.418,
+      "step": 3835
+    },
+    {
+      "epoch": 1.0803207202138134,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.315405576032563e-06,
+      "loss": 1.6552,
+      "step": 3840
+    },
+    {
+      "epoch": 1.0817273878182585,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.3126555038100374e-06,
+      "loss": 1.5523,
+      "step": 3845
+    },
+    {
+      "epoch": 1.0831340554227036,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.3099004381079e-06,
+      "loss": 1.4922,
+      "step": 3850
+    },
+    {
+      "epoch": 1.0845407230271487,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.307140383079125e-06,
+      "loss": 1.5572,
+      "step": 3855
+    },
+    {
+      "epoch": 1.0859473906315937,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.304375342884201e-06,
+      "loss": 1.4289,
+      "step": 3860
+    },
+    {
+      "epoch": 1.0873540582360388,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.301605321691138e-06,
+      "loss": 1.7568,
+      "step": 3865
+    },
+    {
+      "epoch": 1.088760725840484,
+      "grad_norm": 3.125,
+      "learning_rate": 7.2988303236754515e-06,
+      "loss": 1.4492,
+      "step": 3870
+    },
+    {
+      "epoch": 1.090167393444929,
+      "grad_norm": 3.125,
+      "learning_rate": 7.296050353020156e-06,
+      "loss": 1.8061,
+      "step": 3875
+    },
+    {
+      "epoch": 1.091574061049374,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.293265413915767e-06,
+      "loss": 1.775,
+      "step": 3880
+    },
+    {
+      "epoch": 1.0929807286538191,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.290475510560288e-06,
+      "loss": 1.6663,
+      "step": 3885
+    },
+    {
+      "epoch": 1.0943873962582642,
+      "grad_norm": 2.75,
+      "learning_rate": 7.287680647159202e-06,
+      "loss": 1.6116,
+      "step": 3890
+    },
+    {
+      "epoch": 1.0957940638627093,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.2848808279254745e-06,
+      "loss": 1.4655,
+      "step": 3895
+    },
+    {
+      "epoch": 1.0972007314671544,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.282076057079537e-06,
+      "loss": 1.7702,
+      "step": 3900
+    },
+    {
+      "epoch": 1.0986073990715994,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.2792663388492865e-06,
+      "loss": 1.4311,
+      "step": 3905
+    },
+    {
+      "epoch": 1.1000140666760445,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.2764516774700775e-06,
+      "loss": 1.6944,
+      "step": 3910
+    },
+    {
+      "epoch": 1.1014207342804896,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.273632077184716e-06,
+      "loss": 1.7071,
+      "step": 3915
+    },
+    {
+      "epoch": 1.1028274018849347,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.270807542243453e-06,
+      "loss": 1.5593,
+      "step": 3920
+    },
+    {
+      "epoch": 1.1042340694893797,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.2679780769039775e-06,
+      "loss": 1.5341,
+      "step": 3925
+    },
+    {
+      "epoch": 1.1056407370938248,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.26514368543141e-06,
+      "loss": 1.8932,
+      "step": 3930
+    },
+    {
+      "epoch": 1.10704740469827,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.262304372098299e-06,
+      "loss": 1.7494,
+      "step": 3935
+    },
+    {
+      "epoch": 1.108454072302715,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.259460141184609e-06,
+      "loss": 1.6125,
+      "step": 3940
+    },
+    {
+      "epoch": 1.1098607399071598,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.25661099697772e-06,
+      "loss": 1.7168,
+      "step": 3945
+    },
+    {
+      "epoch": 1.111267407511605,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.253756943772416e-06,
+      "loss": 1.7773,
+      "step": 3950
+    },
+    {
+      "epoch": 1.11267407511605,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.250897985870884e-06,
+      "loss": 1.6824,
+      "step": 3955
+    },
+    {
+      "epoch": 1.114080742720495,
+      "grad_norm": 2.625,
+      "learning_rate": 7.248034127582698e-06,
+      "loss": 1.6164,
+      "step": 3960
+    },
+    {
+      "epoch": 1.1154874103249401,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.245165373224829e-06,
+      "loss": 1.6013,
+      "step": 3965
+    },
+    {
+      "epoch": 1.1168940779293852,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.242291727121617e-06,
+      "loss": 1.6286,
+      "step": 3970
+    },
+    {
+      "epoch": 1.1183007455338303,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.2394131936047845e-06,
+      "loss": 1.5001,
+      "step": 3975
+    },
+    {
+      "epoch": 1.1197074131382754,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.236529777013416e-06,
+      "loss": 1.5944,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1211140807427205,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.233641481693959e-06,
+      "loss": 1.5331,
+      "step": 3985
+    },
+    {
+      "epoch": 1.1225207483471655,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.230748312000216e-06,
+      "loss": 1.772,
+      "step": 3990
+    },
+    {
+      "epoch": 1.1239274159516106,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.227850272293334e-06,
+      "loss": 1.5381,
+      "step": 3995
+    },
+    {
+      "epoch": 1.1253340835560557,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.224947366941805e-06,
+      "loss": 1.7734,
+      "step": 4000
+    },
+    {
+      "epoch": 1.1267407511605008,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.2220396003214525e-06,
+      "loss": 1.703,
+      "step": 4005
+    },
+    {
+      "epoch": 1.1281474187649458,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.219126976815427e-06,
+      "loss": 1.3594,
+      "step": 4010
+    },
+    {
+      "epoch": 1.129554086369391,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.216209500814205e-06,
+      "loss": 1.3872,
+      "step": 4015
+    },
+    {
+      "epoch": 1.130960753973836,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.213287176715571e-06,
+      "loss": 1.5567,
+      "step": 4020
+    },
+    {
+      "epoch": 1.132367421578281,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.210360008924625e-06,
+      "loss": 1.5977,
+      "step": 4025
+    },
+    {
+      "epoch": 1.1337740891827262,
+      "grad_norm": 3.375,
+      "learning_rate": 7.207428001853762e-06,
+      "loss": 1.7796,
+      "step": 4030
+    },
+    {
+      "epoch": 1.1351807567871712,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.204491159922675e-06,
+      "loss": 1.7479,
+      "step": 4035
+    },
+    {
+      "epoch": 1.1365874243916163,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.201549487558344e-06,
+      "loss": 1.5466,
+      "step": 4040
+    },
+    {
+      "epoch": 1.1379940919960614,
+      "grad_norm": 2.375,
+      "learning_rate": 7.198602989195029e-06,
+      "loss": 1.5066,
+      "step": 4045
+    },
+    {
+      "epoch": 1.1394007596005065,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.19565166927427e-06,
+      "loss": 1.3287,
+      "step": 4050
+    },
+    {
+      "epoch": 1.1408074272049515,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.192695532244867e-06,
+      "loss": 1.4833,
+      "step": 4055
+    },
+    {
+      "epoch": 1.1422140948093966,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.1897345825628875e-06,
+      "loss": 1.5049,
+      "step": 4060
+    },
+    {
+      "epoch": 1.1436207624138417,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.186768824691652e-06,
+      "loss": 1.7101,
+      "step": 4065
+    },
+    {
+      "epoch": 1.1450274300182868,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.183798263101729e-06,
+      "loss": 1.6311,
+      "step": 4070
+    },
+    {
+      "epoch": 1.1464340976227319,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.180822902270926e-06,
+      "loss": 1.2967,
+      "step": 4075
+    },
+    {
+      "epoch": 1.1478407652271767,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.177842746684287e-06,
+      "loss": 1.3902,
+      "step": 4080
+    },
+    {
+      "epoch": 1.149247432831622,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.174857800834083e-06,
+      "loss": 1.5712,
+      "step": 4085
+    },
+    {
+      "epoch": 1.1506541004360669,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.171868069219804e-06,
+      "loss": 1.5564,
+      "step": 4090
+    },
+    {
+      "epoch": 1.152060768040512,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.16887355634816e-06,
+      "loss": 1.435,
+      "step": 4095
+    },
+    {
+      "epoch": 1.153467435644957,
+      "grad_norm": 3.875,
+      "learning_rate": 7.16587426673306e-06,
+      "loss": 1.6016,
+      "step": 4100
+    },
+    {
+      "epoch": 1.154874103249402,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.16287020489562e-06,
+      "loss": 1.6038,
+      "step": 4105
+    },
+    {
+      "epoch": 1.1562807708538472,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.159861375364146e-06,
+      "loss": 1.5753,
+      "step": 4110
+    },
+    {
+      "epoch": 1.1576874384582923,
+      "grad_norm": 4.5,
+      "learning_rate": 7.156847782674132e-06,
+      "loss": 1.6043,
+      "step": 4115
+    },
+    {
+      "epoch": 1.1590941060627373,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.153829431368252e-06,
+      "loss": 1.5679,
+      "step": 4120
+    },
+    {
+      "epoch": 1.1605007736671824,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.150806325996354e-06,
+      "loss": 1.5646,
+      "step": 4125
+    },
+    {
+      "epoch": 1.1619074412716275,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.147778471115449e-06,
+      "loss": 1.693,
+      "step": 4130
+    },
+    {
+      "epoch": 1.1633141088760726,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.144745871289711e-06,
+      "loss": 1.6215,
+      "step": 4135
+    },
+    {
+      "epoch": 1.1647207764805176,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.141708531090467e-06,
+      "loss": 1.5527,
+      "step": 4140
+    },
+    {
+      "epoch": 1.1661274440849627,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.138666455096183e-06,
+      "loss": 1.8448,
+      "step": 4145
+    },
+    {
+      "epoch": 1.1675341116894078,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.1356196478924734e-06,
+      "loss": 1.7495,
+      "step": 4150
+    },
+    {
+      "epoch": 1.1689407792938529,
+      "grad_norm": 3.625,
+      "learning_rate": 7.132568114072077e-06,
+      "loss": 1.4359,
+      "step": 4155
+    },
+    {
+      "epoch": 1.170347446898298,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.12951185823486e-06,
+      "loss": 1.6285,
+      "step": 4160
+    },
+    {
+      "epoch": 1.171754114502743,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.126450884987807e-06,
+      "loss": 1.7377,
+      "step": 4165
+    },
+    {
+      "epoch": 1.173160782107188,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.123385198945012e-06,
+      "loss": 1.7262,
+      "step": 4170
+    },
+    {
+      "epoch": 1.1745674497116332,
+      "grad_norm": 3.375,
+      "learning_rate": 7.120314804727676e-06,
+      "loss": 1.5923,
+      "step": 4175
+    },
+    {
+      "epoch": 1.1759741173160783,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.117239706964094e-06,
+      "loss": 1.7389,
+      "step": 4180
+    },
+    {
+      "epoch": 1.1773807849205233,
+      "grad_norm": 3.125,
+      "learning_rate": 7.114159910289652e-06,
+      "loss": 1.7112,
+      "step": 4185
+    },
+    {
+      "epoch": 1.1787874525249684,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.111075419346821e-06,
+      "loss": 1.7166,
+      "step": 4190
+    },
+    {
+      "epoch": 1.1801941201294135,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.107986238785145e-06,
+      "loss": 1.4257,
+      "step": 4195
+    },
+    {
+      "epoch": 1.1816007877338586,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.10489237326124e-06,
+      "loss": 1.6122,
+      "step": 4200
+    },
+    {
+      "epoch": 1.1830074553383036,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.101793827438781e-06,
+      "loss": 1.415,
+      "step": 4205
+    },
+    {
+      "epoch": 1.1844141229427487,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.098690605988501e-06,
+      "loss": 1.5566,
+      "step": 4210
+    },
+    {
+      "epoch": 1.1858207905471936,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.095582713588179e-06,
+      "loss": 1.3807,
+      "step": 4215
+    },
+    {
+      "epoch": 1.1872274581516389,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.092470154922638e-06,
+      "loss": 1.7086,
+      "step": 4220
+    },
+    {
+      "epoch": 1.1886341257560837,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.089352934683729e-06,
+      "loss": 1.5553,
+      "step": 4225
+    },
+    {
+      "epoch": 1.1900407933605288,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.086231057570337e-06,
+      "loss": 1.3568,
+      "step": 4230
+    },
+    {
+      "epoch": 1.191447460964974,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.083104528288361e-06,
+      "loss": 1.5711,
+      "step": 4235
+    },
+    {
+      "epoch": 1.192854128569419,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.079973351550716e-06,
+      "loss": 1.5155,
+      "step": 4240
+    },
+    {
+      "epoch": 1.194260796173864,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.076837532077321e-06,
+      "loss": 1.5077,
+      "step": 4245
+    },
+    {
+      "epoch": 1.1956674637783091,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.073697074595095e-06,
+      "loss": 1.5776,
+      "step": 4250
+    },
+    {
+      "epoch": 1.1970741313827542,
+      "grad_norm": 4.25,
+      "learning_rate": 7.070551983837945e-06,
+      "loss": 1.5348,
+      "step": 4255
+    },
+    {
+      "epoch": 1.1984807989871993,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.067402264546766e-06,
+      "loss": 1.7168,
+      "step": 4260
+    },
+    {
+      "epoch": 1.1998874665916444,
+      "grad_norm": 3.875,
+      "learning_rate": 7.064247921469429e-06,
+      "loss": 1.6378,
+      "step": 4265
+    },
+    {
+      "epoch": 1.2012941341960894,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.061088959360772e-06,
+      "loss": 1.4526,
+      "step": 4270
+    },
+    {
+      "epoch": 1.2027008018005345,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.0579253829826e-06,
+      "loss": 1.6888,
+      "step": 4275
+    },
+    {
+      "epoch": 1.2041074694049796,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.05475719710367e-06,
+      "loss": 1.6771,
+      "step": 4280
+    },
+    {
+      "epoch": 1.2055141370094247,
+      "grad_norm": 3.125,
+      "learning_rate": 7.051584406499691e-06,
+      "loss": 1.7519,
+      "step": 4285
+    },
+    {
+      "epoch": 1.2069208046138697,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.048407015953309e-06,
+      "loss": 1.6397,
+      "step": 4290
+    },
+    {
+      "epoch": 1.2083274722183148,
+      "grad_norm": 4.75,
+      "learning_rate": 7.045225030254107e-06,
+      "loss": 1.6319,
+      "step": 4295
+    },
+    {
+      "epoch": 1.20973413982276,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.042038454198593e-06,
+      "loss": 1.5327,
+      "step": 4300
+    },
+    {
+      "epoch": 1.211140807427205,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.038847292590196e-06,
+      "loss": 1.4904,
+      "step": 4305
+    },
+    {
+      "epoch": 1.21254747503165,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.0356515502392555e-06,
+      "loss": 1.4568,
+      "step": 4310
+    },
+    {
+      "epoch": 1.2139541426360951,
+      "grad_norm": 3.625,
+      "learning_rate": 7.032451231963016e-06,
+      "loss": 1.5869,
+      "step": 4315
+    },
+    {
+      "epoch": 1.2153608102405402,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.0292463425856235e-06,
+      "loss": 1.6826,
+      "step": 4320
+    },
+    {
+      "epoch": 1.2167674778449853,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.026036886938108e-06,
+      "loss": 1.5787,
+      "step": 4325
+    },
+    {
+      "epoch": 1.2181741454494304,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.022822869858389e-06,
+      "loss": 1.575,
+      "step": 4330
+    },
+    {
+      "epoch": 1.2195808130538754,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.0196042961912575e-06,
+      "loss": 1.5395,
+      "step": 4335
+    },
+    {
+      "epoch": 1.2209874806583205,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.016381170788375e-06,
+      "loss": 1.6192,
+      "step": 4340
+    },
+    {
+      "epoch": 1.2223941482627656,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.013153498508263e-06,
+      "loss": 1.6263,
+      "step": 4345
+    },
+    {
+      "epoch": 1.2238008158672105,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.009921284216299e-06,
+      "loss": 1.4436,
+      "step": 4350
+    },
+    {
+      "epoch": 1.2252074834716558,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.006684532784707e-06,
+      "loss": 1.425,
+      "step": 4355
+    },
+    {
+      "epoch": 1.2266141510761006,
+      "grad_norm": 4.375,
+      "learning_rate": 7.003443249092547e-06,
+      "loss": 1.3797,
+      "step": 4360
+    },
+    {
+      "epoch": 1.228020818680546,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.000197438025715e-06,
+      "loss": 1.7619,
+      "step": 4365
+    },
+    {
+      "epoch": 1.2294274862849908,
+      "grad_norm": 2.75,
+      "learning_rate": 6.9969471044769275e-06,
+      "loss": 1.6214,
+      "step": 4370
+    },
+    {
+      "epoch": 1.2308341538894358,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.993692253345722e-06,
+      "loss": 1.7026,
+      "step": 4375
+    },
+    {
+      "epoch": 1.232240821493881,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.990432889538444e-06,
+      "loss": 1.5261,
+      "step": 4380
+    },
+    {
+      "epoch": 1.233647489098326,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.98716901796824e-06,
+      "loss": 1.6586,
+      "step": 4385
+    },
+    {
+      "epoch": 1.235054156702771,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.983900643555056e-06,
+      "loss": 1.5024,
+      "step": 4390
+    },
+    {
+      "epoch": 1.2364608243072162,
+      "grad_norm": 3.875,
+      "learning_rate": 6.980627771225618e-06,
+      "loss": 1.5133,
+      "step": 4395
+    },
+    {
+      "epoch": 1.2378674919116612,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.977350405913442e-06,
+      "loss": 1.4485,
+      "step": 4400
+    },
+    {
+      "epoch": 1.2392741595161063,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.974068552558806e-06,
+      "loss": 1.6037,
+      "step": 4405
+    },
+    {
+      "epoch": 1.2406808271205514,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.970782216108764e-06,
+      "loss": 1.833,
+      "step": 4410
+    },
+    {
+      "epoch": 1.2420874947249965,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.967491401517118e-06,
+      "loss": 1.6397,
+      "step": 4415
+    },
+    {
+      "epoch": 1.2434941623294415,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.964196113744427e-06,
+      "loss": 1.6655,
+      "step": 4420
+    },
+    {
+      "epoch": 1.2449008299338866,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.960896357757989e-06,
+      "loss": 1.588,
+      "step": 4425
+    },
+    {
+      "epoch": 1.2463074975383317,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.957592138531841e-06,
+      "loss": 1.7372,
+      "step": 4430
+    },
+    {
+      "epoch": 1.2477141651427768,
+      "grad_norm": 3.375,
+      "learning_rate": 6.954283461046744e-06,
+      "loss": 1.459,
+      "step": 4435
+    },
+    {
+      "epoch": 1.2491208327472219,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.950970330290182e-06,
+      "loss": 1.5329,
+      "step": 4440
+    },
+    {
+      "epoch": 1.250527500351667,
+      "grad_norm": 4.0,
+      "learning_rate": 6.947652751256351e-06,
+      "loss": 1.6937,
+      "step": 4445
+    },
+    {
+      "epoch": 1.251934167956112,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.944330728946153e-06,
+      "loss": 1.5424,
+      "step": 4450
+    },
+    {
+      "epoch": 1.253340835560557,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.941004268367185e-06,
+      "loss": 1.4694,
+      "step": 4455
+    },
+    {
+      "epoch": 1.2547475031650022,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.937673374533738e-06,
+      "loss": 1.5039,
+      "step": 4460
+    },
+    {
+      "epoch": 1.2561541707694472,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.934338052466785e-06,
+      "loss": 1.5262,
+      "step": 4465
+    },
+    {
+      "epoch": 1.2575608383738923,
+      "grad_norm": 4.3125,
+      "learning_rate": 6.93099830719397e-06,
+      "loss": 1.5302,
+      "step": 4470
+    },
+    {
+      "epoch": 1.2589675059783374,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.92765414374961e-06,
+      "loss": 1.6213,
+      "step": 4475
+    },
+    {
+      "epoch": 1.2603741735827825,
+      "grad_norm": 4.25,
+      "learning_rate": 6.924305567174678e-06,
+      "loss": 1.4612,
+      "step": 4480
+    },
+    {
+      "epoch": 1.2617808411872273,
+      "grad_norm": 3.75,
+      "learning_rate": 6.920952582516802e-06,
+      "loss": 1.5592,
+      "step": 4485
+    },
+    {
+      "epoch": 1.2631875087916726,
+      "grad_norm": 5.59375,
+      "learning_rate": 6.917595194830253e-06,
+      "loss": 1.709,
+      "step": 4490
+    },
+    {
+      "epoch": 1.2645941763961175,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.91423340917594e-06,
+      "loss": 1.4438,
+      "step": 4495
+    },
+    {
+      "epoch": 1.2660008440005628,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.9108672306214e-06,
+      "loss": 1.9044,
+      "step": 4500
+    },
+    {
+      "epoch": 1.2674075116050076,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.907496664240796e-06,
+      "loss": 1.6878,
+      "step": 4505
+    },
+    {
+      "epoch": 1.268814179209453,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.9041217151149e-06,
+      "loss": 1.529,
+      "step": 4510
+    },
+    {
+      "epoch": 1.2702208468138978,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.900742388331091e-06,
+      "loss": 1.5458,
+      "step": 4515
+    },
+    {
+      "epoch": 1.2716275144183429,
+      "grad_norm": 3.5,
+      "learning_rate": 6.897358688983351e-06,
+      "loss": 1.5387,
+      "step": 4520
+    },
+    {
+      "epoch": 1.273034182022788,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.893970622172251e-06,
+      "loss": 1.6777,
+      "step": 4525
+    },
+    {
+      "epoch": 1.274440849627233,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.890578193004944e-06,
+      "loss": 1.654,
+      "step": 4530
+    },
+    {
+      "epoch": 1.275847517231678,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.88718140659516e-06,
+      "loss": 1.7976,
+      "step": 4535
+    },
+    {
+      "epoch": 1.2772541848361232,
+      "grad_norm": 2.625,
+      "learning_rate": 6.883780268063198e-06,
+      "loss": 1.4941,
+      "step": 4540
+    },
+    {
+      "epoch": 1.2786608524405683,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.880374782535915e-06,
+      "loss": 1.6671,
+      "step": 4545
+    },
+    {
+      "epoch": 1.2800675200450133,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.8769649551467235e-06,
+      "loss": 1.3581,
+      "step": 4550
+    },
+    {
+      "epoch": 1.2814741876494584,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.87355079103558e-06,
+      "loss": 1.4248,
+      "step": 4555
+    },
+    {
+      "epoch": 1.2828808552539035,
+      "grad_norm": 2.25,
+      "learning_rate": 6.8701322953489755e-06,
+      "loss": 1.7387,
+      "step": 4560
+    },
+    {
+      "epoch": 1.2842875228583486,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.866709473239932e-06,
+      "loss": 1.6799,
+      "step": 4565
+    },
+    {
+      "epoch": 1.2856941904627937,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.8632823298679985e-06,
+      "loss": 1.492,
+      "step": 4570
+    },
+    {
+      "epoch": 1.2871008580672387,
+      "grad_norm": 3.5,
+      "learning_rate": 6.859850870399229e-06,
+      "loss": 1.7618,
+      "step": 4575
+    },
+    {
+      "epoch": 1.2885075256716838,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.856415100006188e-06,
+      "loss": 1.6899,
+      "step": 4580
+    },
+    {
+      "epoch": 1.2899141932761289,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.852975023867939e-06,
+      "loss": 1.3663,
+      "step": 4585
+    },
+    {
+      "epoch": 1.291320860880574,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.849530647170033e-06,
+      "loss": 1.5616,
+      "step": 4590
+    },
+    {
+      "epoch": 1.292727528485019,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.846081975104507e-06,
+      "loss": 1.8242,
+      "step": 4595
+    },
+    {
+      "epoch": 1.2941341960894641,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.842629012869872e-06,
+      "loss": 1.7529,
+      "step": 4600
+    },
+    {
+      "epoch": 1.2955408636939092,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.839171765671104e-06,
+      "loss": 1.6587,
+      "step": 4605
+    },
+    {
+      "epoch": 1.2969475312983543,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.835710238719638e-06,
+      "loss": 1.7018,
+      "step": 4610
+    },
+    {
+      "epoch": 1.2983541989027994,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.832244437233364e-06,
+      "loss": 1.6016,
+      "step": 4615
+    },
+    {
+      "epoch": 1.2997608665072442,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.828774366436613e-06,
+      "loss": 1.8004,
+      "step": 4620
+    },
+    {
+      "epoch": 1.3011675341116895,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.82530003156015e-06,
+      "loss": 1.5459,
+      "step": 4625
+    },
+    {
+      "epoch": 1.3025742017161344,
+      "grad_norm": 3.125,
+      "learning_rate": 6.82182143784117e-06,
+      "loss": 1.3069,
+      "step": 4630
+    },
+    {
+      "epoch": 1.3039808693205797,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.818338590523288e-06,
+      "loss": 1.7134,
+      "step": 4635
+    },
+    {
+      "epoch": 1.3053875369250245,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.8148514948565275e-06,
+      "loss": 1.3821,
+      "step": 4640
+    },
+    {
+      "epoch": 1.3067942045294698,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.81136015609732e-06,
+      "loss": 1.5428,
+      "step": 4645
+    },
+    {
+      "epoch": 1.3082008721339147,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.8078645795084925e-06,
+      "loss": 1.8607,
+      "step": 4650
+    },
+    {
+      "epoch": 1.3096075397383597,
+      "grad_norm": 6.96875,
+      "learning_rate": 6.804364770359257e-06,
+      "loss": 1.6046,
+      "step": 4655
+    },
+    {
+      "epoch": 1.3110142073428048,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.8008607339252075e-06,
+      "loss": 1.6103,
+      "step": 4660
+    },
+    {
+      "epoch": 1.31242087494725,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.797352475488311e-06,
+      "loss": 1.3723,
+      "step": 4665
+    },
+    {
+      "epoch": 1.313827542551695,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.7938400003368975e-06,
+      "loss": 1.6436,
+      "step": 4670
+    },
+    {
+      "epoch": 1.31523421015614,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.790323313765654e-06,
+      "loss": 1.4495,
+      "step": 4675
+    },
+    {
+      "epoch": 1.3166408777605851,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.786802421075615e-06,
+      "loss": 1.7251,
+      "step": 4680
+    },
+    {
+      "epoch": 1.3180475453650302,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.783277327574156e-06,
+      "loss": 1.6689,
+      "step": 4685
+    },
+    {
+      "epoch": 1.3194542129694753,
+      "grad_norm": 4.0,
+      "learning_rate": 6.779748038574986e-06,
+      "loss": 1.821,
+      "step": 4690
+    },
+    {
+      "epoch": 1.3208608805739204,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.776214559398134e-06,
+      "loss": 1.7114,
+      "step": 4695
+    },
+    {
+      "epoch": 1.3222675481783654,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.772676895369951e-06,
+      "loss": 1.6903,
+      "step": 4700
+    },
+    {
+      "epoch": 1.3236742157828105,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.769135051823092e-06,
+      "loss": 1.5268,
+      "step": 4705
+    },
+    {
+      "epoch": 1.3250808833872556,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.7655890340965125e-06,
+      "loss": 1.8307,
+      "step": 4710
+    },
+    {
+      "epoch": 1.3264875509917007,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.762038847535461e-06,
+      "loss": 1.6079,
+      "step": 4715
+    },
+    {
+      "epoch": 1.3278942185961458,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.758484497491473e-06,
+      "loss": 1.5382,
+      "step": 4720
+    },
+    {
+      "epoch": 1.3293008862005908,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.754925989322353e-06,
+      "loss": 1.5711,
+      "step": 4725
+    },
+    {
+      "epoch": 1.330707553805036,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.751363328392182e-06,
+      "loss": 1.7392,
+      "step": 4730
+    },
+    {
+      "epoch": 1.332114221409481,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.747796520071293e-06,
+      "loss": 1.4317,
+      "step": 4735
+    },
+    {
+      "epoch": 1.333520889013926,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.744225569736276e-06,
+      "loss": 1.7177,
+      "step": 4740
+    },
+    {
+      "epoch": 1.3349275566183711,
+      "grad_norm": 1.96875,
+      "learning_rate": 6.740650482769963e-06,
+      "loss": 1.7749,
+      "step": 4745
+    },
+    {
+      "epoch": 1.3363342242228162,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.737071264561421e-06,
+      "loss": 1.6097,
+      "step": 4750
+    },
+    {
+      "epoch": 1.337740891827261,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.733487920505945e-06,
+      "loss": 1.8134,
+      "step": 4755
+    },
+    {
+      "epoch": 1.3391475594317064,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.729900456005049e-06,
+      "loss": 1.7359,
+      "step": 4760
+    },
+    {
+      "epoch": 1.3405542270361512,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.7263088764664575e-06,
+      "loss": 1.5704,
+      "step": 4765
+    },
+    {
+      "epoch": 1.3419608946405965,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.7227131873041e-06,
+      "loss": 1.4039,
+      "step": 4770
+    },
+    {
+      "epoch": 1.3433675622450414,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.719113393938099e-06,
+      "loss": 1.5993,
+      "step": 4775
+    },
+    {
+      "epoch": 1.3447742298494867,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.715509501794763e-06,
+      "loss": 1.7964,
+      "step": 4780
+    },
+    {
+      "epoch": 1.3461808974539315,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.711901516306583e-06,
+      "loss": 1.4551,
+      "step": 4785
+    },
+    {
+      "epoch": 1.3475875650583766,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.708289442912216e-06,
+      "loss": 1.4375,
+      "step": 4790
+    },
+    {
+      "epoch": 1.3489942326628217,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.7046732870564816e-06,
+      "loss": 1.4929,
+      "step": 4795
+    },
+    {
+      "epoch": 1.3504009002672668,
+      "grad_norm": 4.5,
+      "learning_rate": 6.7010530541903565e-06,
+      "loss": 1.5064,
+      "step": 4800
+    },
+    {
+      "epoch": 1.3518075678717119,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.697428749770958e-06,
+      "loss": 1.6775,
+      "step": 4805
+    },
+    {
+      "epoch": 1.353214235476157,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.693800379261546e-06,
+      "loss": 1.6935,
+      "step": 4810
+    },
+    {
+      "epoch": 1.354620903080602,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.690167948131506e-06,
+      "loss": 1.5423,
+      "step": 4815
+    },
+    {
+      "epoch": 1.356027570685047,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.686531461856345e-06,
+      "loss": 1.644,
+      "step": 4820
+    },
+    {
+      "epoch": 1.3574342382894922,
+      "grad_norm": 3.625,
+      "learning_rate": 6.6828909259176865e-06,
+      "loss": 1.7919,
+      "step": 4825
+    },
+    {
+      "epoch": 1.3588409058939372,
+      "grad_norm": 3.5,
+      "learning_rate": 6.6792463458032534e-06,
+      "loss": 1.5362,
+      "step": 4830
+    },
+    {
+      "epoch": 1.3602475734983823,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.675597727006866e-06,
+      "loss": 1.7212,
+      "step": 4835
+    },
+    {
+      "epoch": 1.3616542411028274,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.671945075028434e-06,
+      "loss": 1.6932,
+      "step": 4840
+    },
+    {
+      "epoch": 1.3630609087072725,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.668288395373946e-06,
+      "loss": 1.594,
+      "step": 4845
+    },
+    {
+      "epoch": 1.3644675763117176,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.664627693555462e-06,
+      "loss": 1.8579,
+      "step": 4850
+    },
+    {
+      "epoch": 1.3658742439161626,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.660962975091104e-06,
+      "loss": 1.6345,
+      "step": 4855
+    },
+    {
+      "epoch": 1.3672809115206077,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.657294245505051e-06,
+      "loss": 1.6163,
+      "step": 4860
+    },
+    {
+      "epoch": 1.3686875791250528,
+      "grad_norm": 5.0625,
+      "learning_rate": 6.653621510327525e-06,
+      "loss": 1.6408,
+      "step": 4865
+    },
+    {
+      "epoch": 1.3700942467294979,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.64994477509479e-06,
+      "loss": 1.7848,
+      "step": 4870
+    },
+    {
+      "epoch": 1.371500914333943,
+      "grad_norm": 6.28125,
+      "learning_rate": 6.646264045349134e-06,
+      "loss": 1.6636,
+      "step": 4875
+    },
+    {
+      "epoch": 1.372907581938388,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.642579326638872e-06,
+      "loss": 1.4139,
+      "step": 4880
+    },
+    {
+      "epoch": 1.374314249542833,
+      "grad_norm": 4.125,
+      "learning_rate": 6.638890624518332e-06,
+      "loss": 1.5858,
+      "step": 4885
+    },
+    {
+      "epoch": 1.3757209171472782,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.63519794454784e-06,
+      "loss": 1.5255,
+      "step": 4890
+    },
+    {
+      "epoch": 1.3771275847517233,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.631501292293725e-06,
+      "loss": 1.7696,
+      "step": 4895
+    },
+    {
+      "epoch": 1.378534252356168,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.627800673328302e-06,
+      "loss": 1.6701,
+      "step": 4900
+    },
+    {
+      "epoch": 1.3799409199606134,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.624096093229863e-06,
+      "loss": 1.6558,
+      "step": 4905
+    },
+    {
+      "epoch": 1.3813475875650583,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.620387557582672e-06,
+      "loss": 1.531,
+      "step": 4910
+    },
+    {
+      "epoch": 1.3827542551695036,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.616675071976958e-06,
+      "loss": 1.4897,
+      "step": 4915
+    },
+    {
+      "epoch": 1.3841609227739484,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.612958642008904e-06,
+      "loss": 1.6086,
+      "step": 4920
+    },
+    {
+      "epoch": 1.3855675903783937,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.609238273280633e-06,
+      "loss": 1.496,
+      "step": 4925
+    },
+    {
+      "epoch": 1.3869742579828386,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.605513971400212e-06,
+      "loss": 1.7276,
+      "step": 4930
+    },
+    {
+      "epoch": 1.3883809255872837,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.601785741981634e-06,
+      "loss": 1.634,
+      "step": 4935
+    },
+    {
+      "epoch": 1.3897875931917287,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.5980535906448114e-06,
+      "loss": 1.5213,
+      "step": 4940
+    },
+    {
+      "epoch": 1.3911942607961738,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.594317523015571e-06,
+      "loss": 1.6396,
+      "step": 4945
+    },
+    {
+      "epoch": 1.3926009284006189,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.590577544725642e-06,
+      "loss": 1.4338,
+      "step": 4950
+    },
+    {
+      "epoch": 1.394007596005064,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.586833661412646e-06,
+      "loss": 1.8134,
+      "step": 4955
+    },
+    {
+      "epoch": 1.395414263609509,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.583085878720095e-06,
+      "loss": 1.5643,
+      "step": 4960
+    },
+    {
+      "epoch": 1.3968209312139541,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.579334202297376e-06,
+      "loss": 1.5369,
+      "step": 4965
+    },
+    {
+      "epoch": 1.3982275988183992,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.575578637799747e-06,
+      "loss": 1.6658,
+      "step": 4970
+    },
+    {
+      "epoch": 1.3996342664228443,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.5718191908883265e-06,
+      "loss": 1.7028,
+      "step": 4975
+    },
+    {
+      "epoch": 1.4010409340272894,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.568055867230086e-06,
+      "loss": 1.4661,
+      "step": 4980
+    },
+    {
+      "epoch": 1.4024476016317344,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.564288672497838e-06,
+      "loss": 1.4148,
+      "step": 4985
+    },
+    {
+      "epoch": 1.4038542692361795,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.560517612370232e-06,
+      "loss": 1.4908,
+      "step": 4990
+    },
+    {
+      "epoch": 1.4052609368406246,
+      "grad_norm": 3.5,
+      "learning_rate": 6.556742692531747e-06,
+      "loss": 1.7999,
+      "step": 4995
+    },
+    {
+      "epoch": 1.4066676044450697,
+      "grad_norm": 3.25,
+      "learning_rate": 6.552963918672675e-06,
+      "loss": 1.5283,
+      "step": 5000
+    },
+    {
+      "epoch": 1.4080742720495147,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.549181296489121e-06,
+      "loss": 1.6524,
+      "step": 5005
+    },
+    {
+      "epoch": 1.4094809396539598,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.545394831682989e-06,
+      "loss": 1.5902,
+      "step": 5010
+    },
+    {
+      "epoch": 1.410887607258405,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.541604529961978e-06,
+      "loss": 1.6948,
+      "step": 5015
+    },
+    {
+      "epoch": 1.41229427486285,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.537810397039568e-06,
+      "loss": 1.4649,
+      "step": 5020
+    },
+    {
+      "epoch": 1.413700942467295,
+      "grad_norm": 4.0,
+      "learning_rate": 6.534012438635015e-06,
+      "loss": 1.5802,
+      "step": 5025
+    },
+    {
+      "epoch": 1.4151076100717401,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.530210660473341e-06,
+      "loss": 1.6675,
+      "step": 5030
+    },
+    {
+      "epoch": 1.416514277676185,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.526405068285329e-06,
+      "loss": 1.7445,
+      "step": 5035
+    },
+    {
+      "epoch": 1.4179209452806303,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.522595667807506e-06,
+      "loss": 1.6208,
+      "step": 5040
+    },
+    {
+      "epoch": 1.4193276128850751,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.518782464782144e-06,
+      "loss": 1.6644,
+      "step": 5045
+    },
+    {
+      "epoch": 1.4207342804895204,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.514965464957246e-06,
+      "loss": 1.4335,
+      "step": 5050
+    },
+    {
+      "epoch": 1.4221409480939653,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.511144674086536e-06,
+      "loss": 1.5942,
+      "step": 5055
+    },
+    {
+      "epoch": 1.4235476156984106,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.507320097929453e-06,
+      "loss": 1.5058,
+      "step": 5060
+    },
+    {
+      "epoch": 1.4249542833028555,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.5034917422511465e-06,
+      "loss": 1.6495,
+      "step": 5065
+    },
+    {
+      "epoch": 1.4263609509073005,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.499659612822458e-06,
+      "loss": 1.6415,
+      "step": 5070
+    },
+    {
+      "epoch": 1.4277676185117456,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.49582371541992e-06,
+      "loss": 1.5523,
+      "step": 5075
+    },
+    {
+      "epoch": 1.4291742861161907,
+      "grad_norm": 2.875,
+      "learning_rate": 6.491984055825744e-06,
+      "loss": 1.5582,
+      "step": 5080
+    },
+    {
+      "epoch": 1.4305809537206358,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.488140639827812e-06,
+      "loss": 1.4445,
+      "step": 5085
+    },
+    {
+      "epoch": 1.4319876213250808,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.484293473219671e-06,
+      "loss": 1.8517,
+      "step": 5090
+    },
+    {
+      "epoch": 1.433394288929526,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.480442561800517e-06,
+      "loss": 1.753,
+      "step": 5095
+    },
+    {
+      "epoch": 1.434800956533971,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.4765879113751965e-06,
+      "loss": 1.6244,
+      "step": 5100
+    },
+    {
+      "epoch": 1.436207624138416,
+      "grad_norm": 3.375,
+      "learning_rate": 6.472729527754188e-06,
+      "loss": 1.6974,
+      "step": 5105
+    },
+    {
+      "epoch": 1.4376142917428612,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.4688674167536e-06,
+      "loss": 1.5025,
+      "step": 5110
+    },
+    {
+      "epoch": 1.4390209593473062,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.465001584195157e-06,
+      "loss": 1.3674,
+      "step": 5115
+    },
+    {
+      "epoch": 1.4404276269517513,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.461132035906196e-06,
+      "loss": 1.627,
+      "step": 5120
+    },
+    {
+      "epoch": 1.4418342945561964,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.4572587777196534e-06,
+      "loss": 1.6016,
+      "step": 5125
+    },
+    {
+      "epoch": 1.4432409621606415,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.453381815474059e-06,
+      "loss": 1.5537,
+      "step": 5130
+    },
+    {
+      "epoch": 1.4446476297650865,
+      "grad_norm": 5.5,
+      "learning_rate": 6.4495011550135245e-06,
+      "loss": 1.7066,
+      "step": 5135
+    },
+    {
+      "epoch": 1.4460542973695316,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.4456168021877376e-06,
+      "loss": 1.4199,
+      "step": 5140
+    },
+    {
+      "epoch": 1.4474609649739767,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.4417287628519504e-06,
+      "loss": 1.6078,
+      "step": 5145
+    },
+    {
+      "epoch": 1.4488676325784218,
+      "grad_norm": 5.9375,
+      "learning_rate": 6.437837042866975e-06,
+      "loss": 1.6382,
+      "step": 5150
+    },
+    {
+      "epoch": 1.4502743001828668,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.43394164809917e-06,
+      "loss": 1.7017,
+      "step": 5155
+    },
+    {
+      "epoch": 1.451680967787312,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.4300425844204305e-06,
+      "loss": 1.4553,
+      "step": 5160
+    },
+    {
+      "epoch": 1.453087635391757,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.426139857708187e-06,
+      "loss": 1.4535,
+      "step": 5165
+    },
+    {
+      "epoch": 1.4544943029962019,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.422233473845388e-06,
+      "loss": 1.578,
+      "step": 5170
+    },
+    {
+      "epoch": 1.4559009706006472,
+      "grad_norm": 2.875,
+      "learning_rate": 6.418323438720497e-06,
+      "loss": 1.5562,
+      "step": 5175
+    },
+    {
+      "epoch": 1.457307638205092,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.414409758227482e-06,
+      "loss": 1.4599,
+      "step": 5180
+    },
+    {
+      "epoch": 1.4587143058095373,
+      "grad_norm": 3.875,
+      "learning_rate": 6.4104924382657995e-06,
+      "loss": 1.5497,
+      "step": 5185
+    },
+    {
+      "epoch": 1.4601209734139822,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.4065714847404035e-06,
+      "loss": 1.4751,
+      "step": 5190
+    },
+    {
+      "epoch": 1.4615276410184275,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.402646903561715e-06,
+      "loss": 1.7455,
+      "step": 5195
+    },
+    {
+      "epoch": 1.4629343086228723,
+      "grad_norm": 4.5625,
+      "learning_rate": 6.398718700645628e-06,
+      "loss": 1.4352,
+      "step": 5200
+    },
+    {
+      "epoch": 1.4643409762273174,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.394786881913496e-06,
+      "loss": 1.6347,
+      "step": 5205
+    },
+    {
+      "epoch": 1.4657476438317625,
+      "grad_norm": 8.8125,
+      "learning_rate": 6.39085145329212e-06,
+      "loss": 1.4356,
+      "step": 5210
+    },
+    {
+      "epoch": 1.4671543114362076,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.386912420713746e-06,
+      "loss": 1.6925,
+      "step": 5215
+    },
+    {
+      "epoch": 1.4685609790406526,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.382969790116052e-06,
+      "loss": 1.6203,
+      "step": 5220
+    },
+    {
+      "epoch": 1.4699676466450977,
+      "grad_norm": 4.0,
+      "learning_rate": 6.379023567442136e-06,
+      "loss": 1.5191,
+      "step": 5225
+    },
+    {
+      "epoch": 1.4713743142495428,
+      "grad_norm": 4.5625,
+      "learning_rate": 6.375073758640516e-06,
+      "loss": 1.1834,
+      "step": 5230
+    },
+    {
+      "epoch": 1.4727809818539879,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.371120369665112e-06,
+      "loss": 1.4152,
+      "step": 5235
+    },
+    {
+      "epoch": 1.474187649458433,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.3671634064752425e-06,
+      "loss": 1.3893,
+      "step": 5240
+    },
+    {
+      "epoch": 1.475594317062878,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.3632028750356125e-06,
+      "loss": 1.7045,
+      "step": 5245
+    },
+    {
+      "epoch": 1.477000984667323,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.359238781316307e-06,
+      "loss": 1.653,
+      "step": 5250
+    },
+    {
+      "epoch": 1.4784076522717682,
+      "grad_norm": 3.25,
+      "learning_rate": 6.35527113129278e-06,
+      "loss": 1.6699,
+      "step": 5255
+    },
+    {
+      "epoch": 1.4798143198762133,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.351299930945846e-06,
+      "loss": 1.5831,
+      "step": 5260
+    },
+    {
+      "epoch": 1.4812209874806583,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.347325186261672e-06,
+      "loss": 1.5442,
+      "step": 5265
+    },
+    {
+      "epoch": 1.4826276550851034,
+      "grad_norm": 4.0,
+      "learning_rate": 6.343346903231769e-06,
+      "loss": 1.5064,
+      "step": 5270
+    },
+    {
+      "epoch": 1.4840343226895485,
+      "grad_norm": 3.0,
+      "learning_rate": 6.339365087852977e-06,
+      "loss": 1.7953,
+      "step": 5275
+    },
+    {
+      "epoch": 1.4854409902939936,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.335379746127465e-06,
+      "loss": 1.7672,
+      "step": 5280
+    },
+    {
+      "epoch": 1.4868476578984386,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.3313908840627165e-06,
+      "loss": 1.4996,
+      "step": 5285
+    },
+    {
+      "epoch": 1.4882543255028837,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.327398507671523e-06,
+      "loss": 1.7181,
+      "step": 5290
+    },
+    {
+      "epoch": 1.4896609931073288,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.3234026229719685e-06,
+      "loss": 1.7623,
+      "step": 5295
+    },
+    {
+      "epoch": 1.4910676607117739,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.319403235987431e-06,
+      "loss": 1.9328,
+      "step": 5300
+    },
+    {
+      "epoch": 1.492474328316219,
+      "grad_norm": 3.625,
+      "learning_rate": 6.315400352746566e-06,
+      "loss": 1.6441,
+      "step": 5305
+    },
+    {
+      "epoch": 1.493880995920664,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.311393979283296e-06,
+      "loss": 1.5421,
+      "step": 5310
+    },
+    {
+      "epoch": 1.495287663525109,
+      "grad_norm": 4.46875,
+      "learning_rate": 6.307384121636811e-06,
+      "loss": 1.6609,
+      "step": 5315
+    },
+    {
+      "epoch": 1.4966943311295542,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.303370785851545e-06,
+      "loss": 1.7574,
+      "step": 5320
+    },
+    {
+      "epoch": 1.498100998733999,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.299353977977184e-06,
+      "loss": 1.7455,
+      "step": 5325
+    },
+    {
+      "epoch": 1.4995076663384443,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.295333704068641e-06,
+      "loss": 1.6189,
+      "step": 5330
+    },
+    {
+      "epoch": 1.5009143339428892,
+      "grad_norm": 4.25,
+      "learning_rate": 6.2913099701860565e-06,
+      "loss": 1.5672,
+      "step": 5335
+    },
+    {
+      "epoch": 1.5023210015473345,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.287282782394786e-06,
+      "loss": 1.6904,
+      "step": 5340
+    },
+    {
+      "epoch": 1.5037276691517794,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.283252146765391e-06,
+      "loss": 1.5382,
+      "step": 5345
+    },
+    {
+      "epoch": 1.5051343367562247,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.279218069373631e-06,
+      "loss": 1.7194,
+      "step": 5350
+    },
+    {
+      "epoch": 1.5065410043606695,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.275180556300452e-06,
+      "loss": 1.5181,
+      "step": 5355
+    },
+    {
+      "epoch": 1.5079476719651146,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.27113961363198e-06,
+      "loss": 1.5198,
+      "step": 5360
+    },
+    {
+      "epoch": 1.5093543395695597,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.267095247459514e-06,
+      "loss": 1.7598,
+      "step": 5365
+    },
+    {
+      "epoch": 1.5107610071740047,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.263047463879506e-06,
+      "loss": 1.7883,
+      "step": 5370
+    },
+    {
+      "epoch": 1.5121676747784498,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.258996268993568e-06,
+      "loss": 1.4889,
+      "step": 5375
+    },
+    {
+      "epoch": 1.513574342382895,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.254941668908447e-06,
+      "loss": 1.5787,
+      "step": 5380
+    },
+    {
+      "epoch": 1.51498100998734,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.250883669736028e-06,
+      "loss": 1.4366,
+      "step": 5385
+    },
+    {
+      "epoch": 1.516387677591785,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.246822277593317e-06,
+      "loss": 1.5613,
+      "step": 5390
+    },
+    {
+      "epoch": 1.5177943451962301,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.242757498602435e-06,
+      "loss": 1.7432,
+      "step": 5395
+    },
+    {
+      "epoch": 1.5192010128006752,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.238689338890608e-06,
+      "loss": 1.4916,
+      "step": 5400
+    },
+    {
+      "epoch": 1.5206076804051203,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.23461780459016e-06,
+      "loss": 1.4807,
+      "step": 5405
+    },
+    {
+      "epoch": 1.5220143480095654,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.2305429018385e-06,
+      "loss": 1.7498,
+      "step": 5410
+    },
+    {
+      "epoch": 1.5234210156140104,
+      "grad_norm": 3.0,
+      "learning_rate": 6.226464636778116e-06,
+      "loss": 1.6306,
+      "step": 5415
+    },
+    {
+      "epoch": 1.5248276832184555,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.222383015556562e-06,
+      "loss": 1.5031,
+      "step": 5420
+    },
+    {
+      "epoch": 1.5262343508229006,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.2182980443264545e-06,
+      "loss": 1.607,
+      "step": 5425
+    },
+    {
+      "epoch": 1.5276410184273455,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.2142097292454555e-06,
+      "loss": 1.5335,
+      "step": 5430
+    },
+    {
+      "epoch": 1.5290476860317908,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.210118076476271e-06,
+      "loss": 1.6874,
+      "step": 5435
+    },
+    {
+      "epoch": 1.5304543536362356,
+      "grad_norm": 3.125,
+      "learning_rate": 6.206023092186637e-06,
+      "loss": 1.707,
+      "step": 5440
+    },
+    {
+      "epoch": 1.531861021240681,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.20192478254931e-06,
+      "loss": 1.5466,
+      "step": 5445
+    },
+    {
+      "epoch": 1.5332676888451258,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.197823153742064e-06,
+      "loss": 1.4202,
+      "step": 5450
+    },
+    {
+      "epoch": 1.534674356449571,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.19371821194767e-06,
+      "loss": 1.7848,
+      "step": 5455
+    },
+    {
+      "epoch": 1.536081024054016,
+      "grad_norm": 3.625,
+      "learning_rate": 6.189609963353897e-06,
+      "loss": 1.6941,
+      "step": 5460
+    },
+    {
+      "epoch": 1.5374876916584612,
+      "grad_norm": 2.5,
+      "learning_rate": 6.185498414153494e-06,
+      "loss": 1.6015,
+      "step": 5465
+    },
+    {
+      "epoch": 1.538894359262906,
+      "grad_norm": 3.75,
+      "learning_rate": 6.181383570544195e-06,
+      "loss": 1.6375,
+      "step": 5470
+    },
+    {
+      "epoch": 1.5403010268673514,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.17726543872869e-06,
+      "loss": 1.3859,
+      "step": 5475
+    },
+    {
+      "epoch": 1.5417076944717962,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.1731440249146286e-06,
+      "loss": 1.6703,
+      "step": 5480
+    },
+    {
+      "epoch": 1.5431143620762415,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.169019335314612e-06,
+      "loss": 1.6369,
+      "step": 5485
+    },
+    {
+      "epoch": 1.5445210296806864,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.164891376146173e-06,
+      "loss": 1.488,
+      "step": 5490
+    },
+    {
+      "epoch": 1.5459276972851317,
+      "grad_norm": 2.625,
+      "learning_rate": 6.160760153631775e-06,
+      "loss": 1.4559,
+      "step": 5495
+    },
+    {
+      "epoch": 1.5473343648895765,
+      "grad_norm": 3.25,
+      "learning_rate": 6.156625673998804e-06,
+      "loss": 1.5765,
+      "step": 5500
+    },
+    {
+      "epoch": 1.5487410324940216,
+      "grad_norm": 5.75,
+      "learning_rate": 6.152487943479551e-06,
+      "loss": 1.6279,
+      "step": 5505
+    },
+    {
+      "epoch": 1.5501477000984667,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.14834696831121e-06,
+      "loss": 1.4578,
+      "step": 5510
+    },
+    {
+      "epoch": 1.5515543677029118,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.144202754735866e-06,
+      "loss": 1.4965,
+      "step": 5515
+    },
+    {
+      "epoch": 1.5529610353073569,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.140055309000482e-06,
+      "loss": 1.4934,
+      "step": 5520
+    },
+    {
+      "epoch": 1.554367702911802,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.135904637356901e-06,
+      "loss": 1.2053,
+      "step": 5525
+    },
+    {
+      "epoch": 1.555774370516247,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.13175074606182e-06,
+      "loss": 1.9499,
+      "step": 5530
+    },
+    {
+      "epoch": 1.557181038120692,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.127593641376793e-06,
+      "loss": 1.681,
+      "step": 5535
+    },
+    {
+      "epoch": 1.5585877057251372,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.12343332956822e-06,
+      "loss": 1.5296,
+      "step": 5540
+    },
+    {
+      "epoch": 1.5599943733295822,
+      "grad_norm": 3.5,
+      "learning_rate": 6.119269816907332e-06,
+      "loss": 1.4921,
+      "step": 5545
+    },
+    {
+      "epoch": 1.5614010409340273,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.115103109670187e-06,
+      "loss": 1.4377,
+      "step": 5550
+    },
+    {
+      "epoch": 1.5628077085384724,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.110933214137657e-06,
+      "loss": 1.3758,
+      "step": 5555
+    },
+    {
+      "epoch": 1.5642143761429175,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.10676013659542e-06,
+      "loss": 1.3795,
+      "step": 5560
+    },
+    {
+      "epoch": 1.5656210437473626,
+      "grad_norm": 3.125,
+      "learning_rate": 6.1025838833339545e-06,
+      "loss": 1.4445,
+      "step": 5565
+    },
+    {
+      "epoch": 1.5670277113518076,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.0984044606485185e-06,
+      "loss": 1.6025,
+      "step": 5570
+    },
+    {
+      "epoch": 1.5684343789562525,
+      "grad_norm": 4.21875,
+      "learning_rate": 6.094221874839157e-06,
+      "loss": 1.6672,
+      "step": 5575
+    },
+    {
+      "epoch": 1.5698410465606978,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.090036132210673e-06,
+      "loss": 1.7502,
+      "step": 5580
+    },
+    {
+      "epoch": 1.5712477141651426,
+      "grad_norm": 3.25,
+      "learning_rate": 6.085847239072634e-06,
+      "loss": 1.5658,
+      "step": 5585
+    },
+    {
+      "epoch": 1.572654381769588,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.081655201739359e-06,
+      "loss": 1.5319,
+      "step": 5590
+    },
+    {
+      "epoch": 1.5740610493740328,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.077460026529901e-06,
+      "loss": 1.4186,
+      "step": 5595
+    },
+    {
+      "epoch": 1.575467716978478,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.073261719768044e-06,
+      "loss": 1.4423,
+      "step": 5600
+    },
+    {
+      "epoch": 1.576874384582923,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.069060287782296e-06,
+      "loss": 1.6494,
+      "step": 5605
+    },
+    {
+      "epoch": 1.5782810521873682,
+      "grad_norm": 4.90625,
+      "learning_rate": 6.064855736905872e-06,
+      "loss": 1.5472,
+      "step": 5610
+    },
+    {
+      "epoch": 1.579687719791813,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.060648073476691e-06,
+      "loss": 1.7602,
+      "step": 5615
+    },
+    {
+      "epoch": 1.5810943873962584,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.056437303837362e-06,
+      "loss": 1.5036,
+      "step": 5620
+    },
+    {
+      "epoch": 1.5825010550007033,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.052223434335179e-06,
+      "loss": 1.7547,
+      "step": 5625
+    },
+    {
+      "epoch": 1.5839077226051486,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.0480064713221036e-06,
+      "loss": 1.3458,
+      "step": 5630
+    },
+    {
+      "epoch": 1.5853143902095934,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.043786421154767e-06,
+      "loss": 1.6531,
+      "step": 5635
+    },
+    {
+      "epoch": 1.5867210578140385,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.0395632901944485e-06,
+      "loss": 1.5451,
+      "step": 5640
+    },
+    {
+      "epoch": 1.5881277254184836,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.035337084807077e-06,
+      "loss": 1.5324,
+      "step": 5645
+    },
+    {
+      "epoch": 1.5895343930229286,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.031107811363208e-06,
+      "loss": 1.6149,
+      "step": 5650
+    },
+    {
+      "epoch": 1.5909410606273737,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.026875476238031e-06,
+      "loss": 1.7126,
+      "step": 5655
+    },
+    {
+      "epoch": 1.5923477282318188,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.022640085811341e-06,
+      "loss": 1.4252,
+      "step": 5660
+    },
+    {
+      "epoch": 1.5937543958362639,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.018401646467546e-06,
+      "loss": 1.5583,
+      "step": 5665
+    },
+    {
+      "epoch": 1.595161063440709,
+      "grad_norm": 3.125,
+      "learning_rate": 6.014160164595648e-06,
+      "loss": 1.6051,
+      "step": 5670
+    },
+    {
+      "epoch": 1.596567731045154,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.009915646589231e-06,
+      "loss": 1.5316,
+      "step": 5675
+    },
+    {
+      "epoch": 1.5979743986495991,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.005668098846465e-06,
+      "loss": 1.711,
+      "step": 5680
+    },
+    {
+      "epoch": 1.5993810662540442,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.001417527770076e-06,
+      "loss": 1.5814,
+      "step": 5685
+    },
+    {
+      "epoch": 1.6007877338584893,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.9971639397673565e-06,
+      "loss": 1.768,
+      "step": 5690
+    },
+    {
+      "epoch": 1.6021944014629343,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.992907341250142e-06,
+      "loss": 1.6234,
+      "step": 5695
+    },
+    {
+      "epoch": 1.6036010690673794,
+      "grad_norm": 2.75,
+      "learning_rate": 5.988647738634803e-06,
+      "loss": 1.5887,
+      "step": 5700
+    },
+    {
+      "epoch": 1.6050077366718245,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.984385138342248e-06,
+      "loss": 1.7158,
+      "step": 5705
+    },
+    {
+      "epoch": 1.6064144042762694,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.980119546797895e-06,
+      "loss": 1.4228,
+      "step": 5710
+    },
+    {
+      "epoch": 1.6078210718807147,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.975850970431675e-06,
+      "loss": 1.6542,
+      "step": 5715
+    },
+    {
+      "epoch": 1.6092277394851595,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.971579415678018e-06,
+      "loss": 1.616,
+      "step": 5720
+    },
+    {
+      "epoch": 1.6106344070896048,
+      "grad_norm": 2.625,
+      "learning_rate": 5.967304888975844e-06,
+      "loss": 1.6365,
+      "step": 5725
+    },
+    {
+      "epoch": 1.6120410746940497,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.9630273967685505e-06,
+      "loss": 1.5371,
+      "step": 5730
+    },
+    {
+      "epoch": 1.613447742298495,
+      "grad_norm": 2.875,
+      "learning_rate": 5.958746945504009e-06,
+      "loss": 1.8092,
+      "step": 5735
+    },
+    {
+      "epoch": 1.6148544099029398,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.954463541634547e-06,
+      "loss": 1.4343,
+      "step": 5740
+    },
+    {
+      "epoch": 1.6162610775073851,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.950177191616946e-06,
+      "loss": 1.5462,
+      "step": 5745
+    },
+    {
+      "epoch": 1.61766774511183,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.94588790191243e-06,
+      "loss": 1.3474,
+      "step": 5750
+    },
+    {
+      "epoch": 1.6190744127162753,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.941595678986648e-06,
+      "loss": 1.4929,
+      "step": 5755
+    },
+    {
+      "epoch": 1.6204810803207201,
+      "grad_norm": 3.0,
+      "learning_rate": 5.937300529309677e-06,
+      "loss": 1.6187,
+      "step": 5760
+    },
+    {
+      "epoch": 1.6218877479251654,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.933002459356004e-06,
+      "loss": 1.5226,
+      "step": 5765
+    },
+    {
+      "epoch": 1.6232944155296103,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.928701475604515e-06,
+      "loss": 1.7075,
+      "step": 5770
+    },
+    {
+      "epoch": 1.6247010831340554,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.924397584538491e-06,
+      "loss": 1.6602,
+      "step": 5775
+    },
+    {
+      "epoch": 1.6261077507385004,
+      "grad_norm": 9.3125,
+      "learning_rate": 5.920090792645595e-06,
+      "loss": 1.6669,
+      "step": 5780
+    },
+    {
+      "epoch": 1.6275144183429455,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.915781106417863e-06,
+      "loss": 1.4352,
+      "step": 5785
+    },
+    {
+      "epoch": 1.6289210859473906,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.911468532351694e-06,
+      "loss": 1.5932,
+      "step": 5790
+    },
+    {
+      "epoch": 1.6303277535518357,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.907153076947839e-06,
+      "loss": 1.422,
+      "step": 5795
+    },
+    {
+      "epoch": 1.6317344211562808,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.9028347467113926e-06,
+      "loss": 1.4173,
+      "step": 5800
+    },
+    {
+      "epoch": 1.6331410887607258,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.898513548151782e-06,
+      "loss": 1.6911,
+      "step": 5805
+    },
+    {
+      "epoch": 1.634547756365171,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.894189487782763e-06,
+      "loss": 1.7045,
+      "step": 5810
+    },
+    {
+      "epoch": 1.635954423969616,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.889862572122399e-06,
+      "loss": 1.4862,
+      "step": 5815
+    },
+    {
+      "epoch": 1.637361091574061,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.88553280769306e-06,
+      "loss": 1.3872,
+      "step": 5820
+    },
+    {
+      "epoch": 1.6387677591785061,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.88120020102141e-06,
+      "loss": 1.5883,
+      "step": 5825
+    },
+    {
+      "epoch": 1.6401744267829512,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.876864758638401e-06,
+      "loss": 1.4336,
+      "step": 5830
+    },
+    {
+      "epoch": 1.6415810943873963,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.872526487079253e-06,
+      "loss": 1.528,
+      "step": 5835
+    },
+    {
+      "epoch": 1.6429877619918414,
+      "grad_norm": 4.9375,
+      "learning_rate": 5.868185392883454e-06,
+      "loss": 1.6604,
+      "step": 5840
+    },
+    {
+      "epoch": 1.6443944295962862,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.8638414825947476e-06,
+      "loss": 1.6125,
+      "step": 5845
+    },
+    {
+      "epoch": 1.6458010972007315,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.859494762761122e-06,
+      "loss": 1.3778,
+      "step": 5850
+    },
+    {
+      "epoch": 1.6472077648051764,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.855145239934797e-06,
+      "loss": 1.7263,
+      "step": 5855
+    },
+    {
+      "epoch": 1.6486144324096217,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.850792920672225e-06,
+      "loss": 1.3598,
+      "step": 5860
+    },
+    {
+      "epoch": 1.6500211000140665,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.846437811534068e-06,
+      "loss": 1.5234,
+      "step": 5865
+    },
+    {
+      "epoch": 1.6514277676185118,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.842079919085192e-06,
+      "loss": 1.6599,
+      "step": 5870
+    },
+    {
+      "epoch": 1.6528344352229567,
+      "grad_norm": 5.46875,
+      "learning_rate": 5.837719249894665e-06,
+      "loss": 1.7368,
+      "step": 5875
+    },
+    {
+      "epoch": 1.654241102827402,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.833355810535734e-06,
+      "loss": 1.6333,
+      "step": 5880
+    },
+    {
+      "epoch": 1.6556477704318469,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.8289896075858255e-06,
+      "loss": 1.7083,
+      "step": 5885
+    },
+    {
+      "epoch": 1.6570544380362922,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.824620647626533e-06,
+      "loss": 1.7947,
+      "step": 5890
+    },
+    {
+      "epoch": 1.658461105640737,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.820248937243602e-06,
+      "loss": 1.5477,
+      "step": 5895
+    },
+    {
+      "epoch": 1.6598677732451823,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.815874483026926e-06,
+      "loss": 1.7948,
+      "step": 5900
+    },
+    {
+      "epoch": 1.6612744408496272,
+      "grad_norm": 3.75,
+      "learning_rate": 5.811497291570535e-06,
+      "loss": 1.2987,
+      "step": 5905
+    },
+    {
+      "epoch": 1.6626811084540725,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.807117369472585e-06,
+      "loss": 1.6014,
+      "step": 5910
+    },
+    {
+      "epoch": 1.6640877760585173,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.8027347233353465e-06,
+      "loss": 1.5098,
+      "step": 5915
+    },
+    {
+      "epoch": 1.6654944436629624,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.798349359765198e-06,
+      "loss": 1.5863,
+      "step": 5920
+    },
+    {
+      "epoch": 1.6669011112674075,
+      "grad_norm": 3.875,
+      "learning_rate": 5.793961285372614e-06,
+      "loss": 1.5017,
+      "step": 5925
+    },
+    {
+      "epoch": 1.6683077788718526,
+      "grad_norm": 5.0,
+      "learning_rate": 5.789570506772154e-06,
+      "loss": 1.4812,
+      "step": 5930
+    },
+    {
+      "epoch": 1.6697144464762976,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.785177030582455e-06,
+      "loss": 1.613,
+      "step": 5935
+    },
+    {
+      "epoch": 1.6711211140807427,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.7807808634262205e-06,
+      "loss": 1.4484,
+      "step": 5940
+    },
+    {
+      "epoch": 1.6725277816851878,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.776382011930211e-06,
+      "loss": 1.5913,
+      "step": 5945
+    },
+    {
+      "epoch": 1.6739344492896329,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.77198048272523e-06,
+      "loss": 1.5753,
+      "step": 5950
+    },
+    {
+      "epoch": 1.675341116894078,
+      "grad_norm": 3.375,
+      "learning_rate": 5.767576282446121e-06,
+      "loss": 1.6776,
+      "step": 5955
+    },
+    {
+      "epoch": 1.676747784498523,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.763169417731751e-06,
+      "loss": 1.5151,
+      "step": 5960
+    },
+    {
+      "epoch": 1.678154452102968,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.758759895225008e-06,
+      "loss": 1.2589,
+      "step": 5965
+    },
+    {
+      "epoch": 1.6795611197074132,
+      "grad_norm": 3.375,
+      "learning_rate": 5.75434772157278e-06,
+      "loss": 1.5056,
+      "step": 5970
+    },
+    {
+      "epoch": 1.6809677873118583,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.749932903425957e-06,
+      "loss": 1.7067,
+      "step": 5975
+    },
+    {
+      "epoch": 1.6823744549163033,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.745515447439411e-06,
+      "loss": 1.4572,
+      "step": 5980
+    },
+    {
+      "epoch": 1.6837811225207484,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.741095360271992e-06,
+      "loss": 1.6003,
+      "step": 5985
+    },
+    {
+      "epoch": 1.6851877901251933,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.736672648586518e-06,
+      "loss": 1.5607,
+      "step": 5990
+    },
+    {
+      "epoch": 1.6865944577296386,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.732247319049761e-06,
+      "loss": 1.6931,
+      "step": 5995
+    },
+    {
+      "epoch": 1.6880011253340834,
+      "grad_norm": 3.375,
+      "learning_rate": 5.727819378332437e-06,
+      "loss": 1.4869,
+      "step": 6000
+    },
+    {
+      "epoch": 1.6894077929385287,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.723388833109205e-06,
+      "loss": 1.6263,
+      "step": 6005
+    },
+    {
+      "epoch": 1.6908144605429736,
+      "grad_norm": 4.59375,
+      "learning_rate": 5.718955690058644e-06,
+      "loss": 1.4953,
+      "step": 6010
+    },
+    {
+      "epoch": 1.6922211281474189,
+      "grad_norm": 4.46875,
+      "learning_rate": 5.714519955863249e-06,
+      "loss": 1.378,
+      "step": 6015
+    },
+    {
+      "epoch": 1.6936277957518637,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.710081637209425e-06,
+      "loss": 1.6132,
+      "step": 6020
+    },
+    {
+      "epoch": 1.695034463356309,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.705640740787467e-06,
+      "loss": 1.6166,
+      "step": 6025
+    },
+    {
+      "epoch": 1.6964411309607539,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.701197273291563e-06,
+      "loss": 1.4201,
+      "step": 6030
+    },
+    {
+      "epoch": 1.6978477985651992,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.696751241419771e-06,
+      "loss": 1.6117,
+      "step": 6035
+    },
+    {
+      "epoch": 1.699254466169644,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.692302651874016e-06,
+      "loss": 1.6267,
+      "step": 6040
+    },
+    {
+      "epoch": 1.7006611337740893,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.68785151136008e-06,
+      "loss": 1.6265,
+      "step": 6045
+    },
+    {
+      "epoch": 1.7020678013785342,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.683397826587586e-06,
+      "loss": 1.502,
+      "step": 6050
+    },
+    {
+      "epoch": 1.7034744689829793,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.678941604269999e-06,
+      "loss": 1.66,
+      "step": 6055
+    },
+    {
+      "epoch": 1.7048811365874244,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.674482851124603e-06,
+      "loss": 1.6555,
+      "step": 6060
+    },
+    {
+      "epoch": 1.7062878041918694,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.670021573872498e-06,
+      "loss": 1.6171,
+      "step": 6065
+    },
+    {
+      "epoch": 1.7076944717963145,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.665557779238593e-06,
+      "loss": 1.723,
+      "step": 6070
+    },
+    {
+      "epoch": 1.7091011394007596,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.661091473951587e-06,
+      "loss": 1.5578,
+      "step": 6075
+    },
+    {
+      "epoch": 1.7105078070052047,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.656622664743965e-06,
+      "loss": 1.6143,
+      "step": 6080
+    },
+    {
+      "epoch": 1.7119144746096497,
+      "grad_norm": 4.71875,
+      "learning_rate": 5.652151358351988e-06,
+      "loss": 1.5817,
+      "step": 6085
+    },
+    {
+      "epoch": 1.7133211422140948,
+      "grad_norm": 3.375,
+      "learning_rate": 5.64767756151568e-06,
+      "loss": 1.5214,
+      "step": 6090
+    },
+    {
+      "epoch": 1.71472780981854,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.643201280978816e-06,
+      "loss": 1.6993,
+      "step": 6095
+    },
+    {
+      "epoch": 1.716134477422985,
+      "grad_norm": 3.375,
+      "learning_rate": 5.638722523488921e-06,
+      "loss": 1.6905,
+      "step": 6100
+    },
+    {
+      "epoch": 1.71754114502743,
+      "grad_norm": 3.125,
+      "learning_rate": 5.63424129579725e-06,
+      "loss": 1.7052,
+      "step": 6105
+    },
+    {
+      "epoch": 1.7189478126318751,
+      "grad_norm": 2.875,
+      "learning_rate": 5.629757604658781e-06,
+      "loss": 1.6538,
+      "step": 6110
+    },
+    {
+      "epoch": 1.7203544802363202,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.625271456832209e-06,
+      "loss": 1.4619,
+      "step": 6115
+    },
+    {
+      "epoch": 1.7217611478407653,
+      "grad_norm": 3.25,
+      "learning_rate": 5.620782859079929e-06,
+      "loss": 1.7455,
+      "step": 6120
+    },
+    {
+      "epoch": 1.7231678154452101,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.6162918181680264e-06,
+      "loss": 1.5303,
+      "step": 6125
+    },
+    {
+      "epoch": 1.7245744830496554,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.611798340866278e-06,
+      "loss": 1.6425,
+      "step": 6130
+    },
+    {
+      "epoch": 1.7259811506541003,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.607302433948126e-06,
+      "loss": 1.7051,
+      "step": 6135
+    },
+    {
+      "epoch": 1.7273878182585456,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.602804104190674e-06,
+      "loss": 1.5346,
+      "step": 6140
+    },
+    {
+      "epoch": 1.7287944858629904,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.598303358374686e-06,
+      "loss": 1.5561,
+      "step": 6145
+    },
+    {
+      "epoch": 1.7302011534674357,
+      "grad_norm": 3.75,
+      "learning_rate": 5.5938002032845596e-06,
+      "loss": 1.6081,
+      "step": 6150
+    },
+    {
+      "epoch": 1.7316078210718806,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.589294645708326e-06,
+      "loss": 1.7825,
+      "step": 6155
+    },
+    {
+      "epoch": 1.733014488676326,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.584786692437644e-06,
+      "loss": 1.4638,
+      "step": 6160
+    },
+    {
+      "epoch": 1.7344211562807708,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.580276350267774e-06,
+      "loss": 1.8662,
+      "step": 6165
+    },
+    {
+      "epoch": 1.735827823885216,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.575763625997584e-06,
+      "loss": 1.8925,
+      "step": 6170
+    },
+    {
+      "epoch": 1.737234491489661,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.5712485264295314e-06,
+      "loss": 1.4838,
+      "step": 6175
+    },
+    {
+      "epoch": 1.7386411590941062,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.566731058369655e-06,
+      "loss": 1.6893,
+      "step": 6180
+    },
+    {
+      "epoch": 1.740047826698551,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.562211228627559e-06,
+      "loss": 1.426,
+      "step": 6185
+    },
+    {
+      "epoch": 1.7414544943029961,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.557689044016414e-06,
+      "loss": 1.4707,
+      "step": 6190
+    },
+    {
+      "epoch": 1.7428611619074412,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.553164511352936e-06,
+      "loss": 1.5854,
+      "step": 6195
+    },
+    {
+      "epoch": 1.7442678295118863,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.548637637457383e-06,
+      "loss": 1.7302,
+      "step": 6200
+    },
+    {
+      "epoch": 1.7456744971163314,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.544108429153541e-06,
+      "loss": 1.5283,
+      "step": 6205
+    },
+    {
+      "epoch": 1.7470811647207765,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.539576893268714e-06,
+      "loss": 1.8094,
+      "step": 6210
+    },
+    {
+      "epoch": 1.7484878323252215,
+      "grad_norm": 2.875,
+      "learning_rate": 5.535043036633716e-06,
+      "loss": 1.4656,
+      "step": 6215
+    },
+    {
+      "epoch": 1.7498944999296666,
+      "grad_norm": 3.75,
+      "learning_rate": 5.530506866082858e-06,
+      "loss": 1.5639,
+      "step": 6220
+    },
+    {
+      "epoch": 1.7513011675341117,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.525968388453943e-06,
+      "loss": 1.6606,
+      "step": 6225
+    },
+    {
+      "epoch": 1.7527078351385568,
+      "grad_norm": 2.625,
+      "learning_rate": 5.521427610588246e-06,
+      "loss": 1.5511,
+      "step": 6230
+    },
+    {
+      "epoch": 1.7541145027430018,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.51688453933051e-06,
+      "loss": 1.3522,
+      "step": 6235
+    },
+    {
+      "epoch": 1.755521170347447,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.51233918152894e-06,
+      "loss": 1.5404,
+      "step": 6240
+    },
+    {
+      "epoch": 1.756927837951892,
+      "grad_norm": 3.125,
+      "learning_rate": 5.507791544035183e-06,
+      "loss": 1.4346,
+      "step": 6245
+    },
+    {
+      "epoch": 1.758334505556337,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.5032416337043255e-06,
+      "loss": 1.5443,
+      "step": 6250
+    },
+    {
+      "epoch": 1.7597411731607822,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.498689457394877e-06,
+      "loss": 1.5366,
+      "step": 6255
+    },
+    {
+      "epoch": 1.761147840765227,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.494135021968766e-06,
+      "loss": 1.5194,
+      "step": 6260
+    },
+    {
+      "epoch": 1.7625545083696723,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.489578334291323e-06,
+      "loss": 1.6533,
+      "step": 6265
+    },
+    {
+      "epoch": 1.7639611759741172,
+      "grad_norm": 3.125,
+      "learning_rate": 5.485019401231275e-06,
+      "loss": 1.5755,
+      "step": 6270
+    },
+    {
+      "epoch": 1.7653678435785625,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.480458229660736e-06,
+      "loss": 1.5212,
+      "step": 6275
+    },
+    {
+      "epoch": 1.7667745111830073,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.4758948264551905e-06,
+      "loss": 1.3376,
+      "step": 6280
+    },
+    {
+      "epoch": 1.7681811787874526,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.471329198493489e-06,
+      "loss": 1.4777,
+      "step": 6285
+    },
+    {
+      "epoch": 1.7695878463918975,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.466761352657836e-06,
+      "loss": 1.5956,
+      "step": 6290
+    },
+    {
+      "epoch": 1.7709945139963428,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.462191295833777e-06,
+      "loss": 1.6315,
+      "step": 6295
+    },
+    {
+      "epoch": 1.7724011816007876,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.457619034910193e-06,
+      "loss": 1.4154,
+      "step": 6300
+    },
+    {
+      "epoch": 1.773807849205233,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.453044576779286e-06,
+      "loss": 1.5447,
+      "step": 6305
+    },
+    {
+      "epoch": 1.7752145168096778,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.448467928336571e-06,
+      "loss": 1.4752,
+      "step": 6310
+    },
+    {
+      "epoch": 1.776621184414123,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.4438890964808605e-06,
+      "loss": 1.5837,
+      "step": 6315
+    },
+    {
+      "epoch": 1.778027852018568,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.439308088114267e-06,
+      "loss": 1.3614,
+      "step": 6320
+    },
+    {
+      "epoch": 1.7794345196230132,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.434724910142175e-06,
+      "loss": 1.4906,
+      "step": 6325
+    },
+    {
+      "epoch": 1.780841187227458,
+      "grad_norm": 5.59375,
+      "learning_rate": 5.430139569473244e-06,
+      "loss": 1.4475,
+      "step": 6330
+    },
+    {
+      "epoch": 1.7822478548319032,
+      "grad_norm": 2.75,
+      "learning_rate": 5.425552073019392e-06,
+      "loss": 1.6079,
+      "step": 6335
+    },
+    {
+      "epoch": 1.7836545224363483,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.420962427695789e-06,
+      "loss": 1.6304,
+      "step": 6340
+    },
+    {
+      "epoch": 1.7850611900407933,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.416370640420842e-06,
+      "loss": 1.8129,
+      "step": 6345
+    },
+    {
+      "epoch": 1.7864678576452384,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.411776718116185e-06,
+      "loss": 1.6699,
+      "step": 6350
+    },
+    {
+      "epoch": 1.7878745252496835,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.4071806677066744e-06,
+      "loss": 1.5165,
+      "step": 6355
+    },
+    {
+      "epoch": 1.7892811928541286,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.402582496120372e-06,
+      "loss": 1.6656,
+      "step": 6360
+    },
+    {
+      "epoch": 1.7906878604585736,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.397982210288536e-06,
+      "loss": 1.6198,
+      "step": 6365
+    },
+    {
+      "epoch": 1.7920945280630187,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.393379817145617e-06,
+      "loss": 1.8459,
+      "step": 6370
+    },
+    {
+      "epoch": 1.7935011956674638,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.388775323629236e-06,
+      "loss": 1.4617,
+      "step": 6375
+    },
+    {
+      "epoch": 1.7949078632719089,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.384168736680182e-06,
+      "loss": 1.6081,
+      "step": 6380
+    },
+    {
+      "epoch": 1.796314530876354,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.379560063242403e-06,
+      "loss": 1.5674,
+      "step": 6385
+    },
+    {
+      "epoch": 1.797721198480799,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.374949310262985e-06,
+      "loss": 1.3943,
+      "step": 6390
+    },
+    {
+      "epoch": 1.799127866085244,
+      "grad_norm": 4.25,
+      "learning_rate": 5.370336484692156e-06,
+      "loss": 1.5251,
+      "step": 6395
+    },
+    {
+      "epoch": 1.8005345336896892,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.3657215934832645e-06,
+      "loss": 1.5846,
+      "step": 6400
+    },
+    {
+      "epoch": 1.801941201294134,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.361104643592773e-06,
+      "loss": 1.7833,
+      "step": 6405
+    },
+    {
+      "epoch": 1.8033478688985793,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.356485641980249e-06,
+      "loss": 1.2723,
+      "step": 6410
+    },
+    {
+      "epoch": 1.8047545365030242,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.351864595608349e-06,
+      "loss": 1.4335,
+      "step": 6415
+    },
+    {
+      "epoch": 1.8061612041074695,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.347241511442816e-06,
+      "loss": 1.4807,
+      "step": 6420
+    },
+    {
+      "epoch": 1.8075678717119144,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.342616396452463e-06,
+      "loss": 1.4345,
+      "step": 6425
+    },
+    {
+      "epoch": 1.8089745393163597,
+      "grad_norm": 5.28125,
+      "learning_rate": 5.337989257609163e-06,
+      "loss": 1.2621,
+      "step": 6430
+    },
+    {
+      "epoch": 1.8103812069208045,
+      "grad_norm": 3.0,
+      "learning_rate": 5.333360101887843e-06,
+      "loss": 1.6888,
+      "step": 6435
+    },
+    {
+      "epoch": 1.8117878745252498,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.328728936266466e-06,
+      "loss": 1.5794,
+      "step": 6440
+    },
+    {
+      "epoch": 1.8131945421296947,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.324095767726027e-06,
+      "loss": 1.5931,
+      "step": 6445
+    },
+    {
+      "epoch": 1.81460120973414,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.319460603250541e-06,
+      "loss": 1.6198,
+      "step": 6450
+    },
+    {
+      "epoch": 1.8160078773385848,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.314823449827031e-06,
+      "loss": 1.5342,
+      "step": 6455
+    },
+    {
+      "epoch": 1.8174145449430301,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.310184314445515e-06,
+      "loss": 1.7853,
+      "step": 6460
+    },
+    {
+      "epoch": 1.818821212547475,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.305543204099006e-06,
+      "loss": 1.7108,
+      "step": 6465
+    },
+    {
+      "epoch": 1.82022788015192,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.3009001257834875e-06,
+      "loss": 1.6419,
+      "step": 6470
+    },
+    {
+      "epoch": 1.8216345477563651,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.29625508649791e-06,
+      "loss": 1.6769,
+      "step": 6475
+    },
+    {
+      "epoch": 1.8230412153608102,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.291608093244183e-06,
+      "loss": 1.659,
+      "step": 6480
+    },
+    {
+      "epoch": 1.8244478829652553,
+      "grad_norm": 3.125,
+      "learning_rate": 5.286959153027162e-06,
+      "loss": 1.8752,
+      "step": 6485
+    },
+    {
+      "epoch": 1.8258545505697004,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.28230827285463e-06,
+      "loss": 1.6693,
+      "step": 6490
+    },
+    {
+      "epoch": 1.8272612181741454,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.277655459737303e-06,
+      "loss": 1.4653,
+      "step": 6495
+    },
+    {
+      "epoch": 1.8286678857785905,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.2730007206888074e-06,
+      "loss": 1.5908,
+      "step": 6500
+    },
+    {
+      "epoch": 1.8300745533830356,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.268344062725671e-06,
+      "loss": 1.6227,
+      "step": 6505
+    },
+    {
+      "epoch": 1.8314812209874807,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.263685492867317e-06,
+      "loss": 1.4596,
+      "step": 6510
+    },
+    {
+      "epoch": 1.8328878885919258,
+      "grad_norm": 3.0,
+      "learning_rate": 5.259025018136049e-06,
+      "loss": 1.5202,
+      "step": 6515
+    },
+    {
+      "epoch": 1.8342945561963708,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.25436264555704e-06,
+      "loss": 1.6152,
+      "step": 6520
+    },
+    {
+      "epoch": 1.835701223800816,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.249698382158329e-06,
+      "loss": 1.7461,
+      "step": 6525
+    },
+    {
+      "epoch": 1.837107891405261,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.245032234970801e-06,
+      "loss": 1.6038,
+      "step": 6530
+    },
+    {
+      "epoch": 1.838514559009706,
+      "grad_norm": 3.625,
+      "learning_rate": 5.240364211028183e-06,
+      "loss": 1.7761,
+      "step": 6535
+    },
+    {
+      "epoch": 1.839921226614151,
+      "grad_norm": 4.5,
+      "learning_rate": 5.235694317367028e-06,
+      "loss": 1.682,
+      "step": 6540
+    },
+    {
+      "epoch": 1.8413278942185962,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.231022561026712e-06,
+      "loss": 1.7722,
+      "step": 6545
+    },
+    {
+      "epoch": 1.842734561823041,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.226348949049414e-06,
+      "loss": 1.4845,
+      "step": 6550
+    },
+    {
+      "epoch": 1.8441412294274864,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.2216734884801126e-06,
+      "loss": 1.6831,
+      "step": 6555
+    },
+    {
+      "epoch": 1.8455478970319312,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.216996186366573e-06,
+      "loss": 1.6796,
+      "step": 6560
+    },
+    {
+      "epoch": 1.8469545646363765,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.212317049759336e-06,
+      "loss": 1.6936,
+      "step": 6565
+    },
+    {
+      "epoch": 1.8483612322408214,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.207636085711707e-06,
+      "loss": 1.8625,
+      "step": 6570
+    },
+    {
+      "epoch": 1.8497678998452667,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.202953301279748e-06,
+      "loss": 1.7222,
+      "step": 6575
+    },
+    {
+      "epoch": 1.8511745674497115,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.198268703522263e-06,
+      "loss": 1.437,
+      "step": 6580
+    },
+    {
+      "epoch": 1.8525812350541568,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.1935822995007896e-06,
+      "loss": 1.6653,
+      "step": 6585
+    },
+    {
+      "epoch": 1.8539879026586017,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.188894096279591e-06,
+      "loss": 1.5089,
+      "step": 6590
+    },
+    {
+      "epoch": 1.855394570263047,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.184204100925639e-06,
+      "loss": 1.4483,
+      "step": 6595
+    },
+    {
+      "epoch": 1.8568012378674918,
+      "grad_norm": 6.53125,
+      "learning_rate": 5.179512320508606e-06,
+      "loss": 1.4933,
+      "step": 6600
+    },
+    {
+      "epoch": 1.858207905471937,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.17481876210086e-06,
+      "loss": 1.4914,
+      "step": 6605
+    },
+    {
+      "epoch": 1.859614573076382,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.170123432777446e-06,
+      "loss": 1.6377,
+      "step": 6610
+    },
+    {
+      "epoch": 1.861021240680827,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.165426339616078e-06,
+      "loss": 1.6628,
+      "step": 6615
+    },
+    {
+      "epoch": 1.8624279082852722,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.160727489697131e-06,
+      "loss": 1.6961,
+      "step": 6620
+    },
+    {
+      "epoch": 1.8638345758897172,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.156026890103626e-06,
+      "loss": 1.5194,
+      "step": 6625
+    },
+    {
+      "epoch": 1.8652412434941623,
+      "grad_norm": 4.25,
+      "learning_rate": 5.1513245479212215e-06,
+      "loss": 1.5143,
+      "step": 6630
+    },
+    {
+      "epoch": 1.8666479110986074,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.146620470238205e-06,
+      "loss": 1.6946,
+      "step": 6635
+    },
+    {
+      "epoch": 1.8680545787030525,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.1419146641454784e-06,
+      "loss": 1.413,
+      "step": 6640
+    },
+    {
+      "epoch": 1.8694612463074975,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.137207136736549e-06,
+      "loss": 1.8252,
+      "step": 6645
+    },
+    {
+      "epoch": 1.8708679139119426,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.132497895107518e-06,
+      "loss": 1.5078,
+      "step": 6650
+    },
+    {
+      "epoch": 1.8722745815163877,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.127786946357074e-06,
+      "loss": 1.6362,
+      "step": 6655
+    },
+    {
+      "epoch": 1.8736812491208328,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.123074297586475e-06,
+      "loss": 1.6801,
+      "step": 6660
+    },
+    {
+      "epoch": 1.8750879167252779,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.118359955899542e-06,
+      "loss": 1.4851,
+      "step": 6665
+    },
+    {
+      "epoch": 1.876494584329723,
+      "grad_norm": 3.375,
+      "learning_rate": 5.113643928402651e-06,
+      "loss": 1.5629,
+      "step": 6670
+    },
+    {
+      "epoch": 1.8779012519341678,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.108926222204716e-06,
+      "loss": 1.7819,
+      "step": 6675
+    },
+    {
+      "epoch": 1.879307919538613,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.104206844417184e-06,
+      "loss": 1.6994,
+      "step": 6680
+    },
+    {
+      "epoch": 1.880714587143058,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.099485802154019e-06,
+      "loss": 1.5289,
+      "step": 6685
+    },
+    {
+      "epoch": 1.8821212547475032,
+      "grad_norm": 2.875,
+      "learning_rate": 5.094763102531697e-06,
+      "loss": 1.4957,
+      "step": 6690
+    },
+    {
+      "epoch": 1.883527922351948,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.09003875266919e-06,
+      "loss": 1.6463,
+      "step": 6695
+    },
+    {
+      "epoch": 1.8849345899563934,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.085312759687958e-06,
+      "loss": 1.5889,
+      "step": 6700
+    },
+    {
+      "epoch": 1.8863412575608383,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.080585130711938e-06,
+      "loss": 1.79,
+      "step": 6705
+    },
+    {
+      "epoch": 1.8877479251652836,
+      "grad_norm": 3.75,
+      "learning_rate": 5.0758558728675345e-06,
+      "loss": 1.8305,
+      "step": 6710
+    },
+    {
+      "epoch": 1.8891545927697284,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.0711249932836035e-06,
+      "loss": 1.4388,
+      "step": 6715
+    },
+    {
+      "epoch": 1.8905612603741737,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.066392499091451e-06,
+      "loss": 1.7699,
+      "step": 6720
+    },
+    {
+      "epoch": 1.8919679279786186,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.061658397424814e-06,
+      "loss": 1.7132,
+      "step": 6725
+    },
+    {
+      "epoch": 1.8933745955830639,
+      "grad_norm": 4.59375,
+      "learning_rate": 5.056922695419849e-06,
+      "loss": 1.3894,
+      "step": 6730
+    },
+    {
+      "epoch": 1.8947812631875087,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.052185400215134e-06,
+      "loss": 1.476,
+      "step": 6735
+    },
+    {
+      "epoch": 1.896187930791954,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.047446518951638e-06,
+      "loss": 1.61,
+      "step": 6740
+    },
+    {
+      "epoch": 1.8975945983963989,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.042706058772728e-06,
+      "loss": 1.6619,
+      "step": 6745
+    },
+    {
+      "epoch": 1.899001266000844,
+      "grad_norm": 5.875,
+      "learning_rate": 5.037964026824148e-06,
+      "loss": 1.4887,
+      "step": 6750
+    },
+    {
+      "epoch": 1.900407933605289,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.033220430254015e-06,
+      "loss": 1.5818,
+      "step": 6755
+    },
+    {
+      "epoch": 1.9018146012097341,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.0284752762128e-06,
+      "loss": 1.4241,
+      "step": 6760
+    },
+    {
+      "epoch": 1.9032212688141792,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.023728571853322e-06,
+      "loss": 1.6242,
+      "step": 6765
+    },
+    {
+      "epoch": 1.9046279364186243,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.018980324330741e-06,
+      "loss": 1.7302,
+      "step": 6770
+    },
+    {
+      "epoch": 1.9060346040230693,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.014230540802538e-06,
+      "loss": 1.7645,
+      "step": 6775
+    },
+    {
+      "epoch": 1.9074412716275144,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.009479228428513e-06,
+      "loss": 1.4511,
+      "step": 6780
+    },
+    {
+      "epoch": 1.9088479392319595,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.00472639437077e-06,
+      "loss": 1.5802,
+      "step": 6785
+    },
+    {
+      "epoch": 1.9102546068364046,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.999972045793705e-06,
+      "loss": 1.5002,
+      "step": 6790
+    },
+    {
+      "epoch": 1.9116612744408497,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.995216189863999e-06,
+      "loss": 1.3874,
+      "step": 6795
+    },
+    {
+      "epoch": 1.9130679420452947,
+      "grad_norm": 3.25,
+      "learning_rate": 4.990458833750606e-06,
+      "loss": 1.5605,
+      "step": 6800
+    },
+    {
+      "epoch": 1.9144746096497398,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.985699984624736e-06,
+      "loss": 1.6657,
+      "step": 6805
+    },
+    {
+      "epoch": 1.915881277254185,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.980939649659856e-06,
+      "loss": 1.574,
+      "step": 6810
+    },
+    {
+      "epoch": 1.91728794485863,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.976177836031669e-06,
+      "loss": 1.5493,
+      "step": 6815
+    },
+    {
+      "epoch": 1.9186946124630748,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.97141455091811e-06,
+      "loss": 1.7608,
+      "step": 6820
+    },
+    {
+      "epoch": 1.9201012800675201,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.966649801499327e-06,
+      "loss": 1.4364,
+      "step": 6825
+    },
+    {
+      "epoch": 1.921507947671965,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.961883594957681e-06,
+      "loss": 1.2724,
+      "step": 6830
+    },
+    {
+      "epoch": 1.9229146152764103,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.957115938477726e-06,
+      "loss": 1.4279,
+      "step": 6835
+    },
+    {
+      "epoch": 1.9243212828808551,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.952346839246202e-06,
+      "loss": 1.414,
+      "step": 6840
+    },
+    {
+      "epoch": 1.9257279504853004,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.947576304452025e-06,
+      "loss": 1.6354,
+      "step": 6845
+    },
+    {
+      "epoch": 1.9271346180897453,
+      "grad_norm": 3.875,
+      "learning_rate": 4.942804341286274e-06,
+      "loss": 1.4963,
+      "step": 6850
+    },
+    {
+      "epoch": 1.9285412856941906,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.938030956942181e-06,
+      "loss": 1.6742,
+      "step": 6855
+    },
+    {
+      "epoch": 1.9299479532986354,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.933256158615121e-06,
+      "loss": 1.5579,
+      "step": 6860
+    },
+    {
+      "epoch": 1.9313546209030807,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9284799535026e-06,
+      "loss": 1.3593,
+      "step": 6865
+    },
+    {
+      "epoch": 1.9327612885075256,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.923702348804244e-06,
+      "loss": 1.397,
+      "step": 6870
+    },
+    {
+      "epoch": 1.934167956111971,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.918923351721791e-06,
+      "loss": 1.652,
+      "step": 6875
+    },
+    {
+      "epoch": 1.9355746237164158,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9141429694590745e-06,
+      "loss": 1.6749,
+      "step": 6880
+    },
+    {
+      "epoch": 1.9369812913208608,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.909361209222018e-06,
+      "loss": 1.7391,
+      "step": 6885
+    },
+    {
+      "epoch": 1.938387958925306,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.9045780782186225e-06,
+      "loss": 1.8264,
+      "step": 6890
+    },
+    {
+      "epoch": 1.939794626529751,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.899793583658955e-06,
+      "loss": 1.4543,
+      "step": 6895
+    },
+    {
+      "epoch": 1.941201294134196,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.895007732755138e-06,
+      "loss": 1.6126,
+      "step": 6900
+    },
+    {
+      "epoch": 1.9426079617386411,
+      "grad_norm": 4.875,
+      "learning_rate": 4.890220532721336e-06,
+      "loss": 1.8178,
+      "step": 6905
+    },
+    {
+      "epoch": 1.9440146293430862,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.885431990773752e-06,
+      "loss": 1.574,
+      "step": 6910
+    },
+    {
+      "epoch": 1.9454212969475313,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.880642114130609e-06,
+      "loss": 1.3841,
+      "step": 6915
+    },
+    {
+      "epoch": 1.9468279645519764,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.875850910012138e-06,
+      "loss": 1.3911,
+      "step": 6920
+    },
+    {
+      "epoch": 1.9482346321564215,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.87105838564058e-06,
+      "loss": 1.791,
+      "step": 6925
+    },
+    {
+      "epoch": 1.9496412997608665,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8662645482401584e-06,
+      "loss": 1.6071,
+      "step": 6930
+    },
+    {
+      "epoch": 1.9510479673653116,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.861469405037079e-06,
+      "loss": 1.3506,
+      "step": 6935
+    },
+    {
+      "epoch": 1.9524546349697567,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.856672963259518e-06,
+      "loss": 1.2333,
+      "step": 6940
+    },
+    {
+      "epoch": 1.9538613025742018,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.851875230137603e-06,
+      "loss": 1.6049,
+      "step": 6945
+    },
+    {
+      "epoch": 1.9552679701786468,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.847076212903414e-06,
+      "loss": 1.8223,
+      "step": 6950
+    },
+    {
+      "epoch": 1.9566746377830917,
+      "grad_norm": 4.25,
+      "learning_rate": 4.842275918790965e-06,
+      "loss": 1.5588,
+      "step": 6955
+    },
+    {
+      "epoch": 1.958081305387537,
+      "grad_norm": 2.625,
+      "learning_rate": 4.837474355036191e-06,
+      "loss": 1.6483,
+      "step": 6960
+    },
+    {
+      "epoch": 1.9594879729919819,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.83267152887695e-06,
+      "loss": 1.6539,
+      "step": 6965
+    },
+    {
+      "epoch": 1.9608946405964272,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.8278674475529915e-06,
+      "loss": 1.5487,
+      "step": 6970
+    },
+    {
+      "epoch": 1.962301308200872,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.823062118305966e-06,
+      "loss": 1.3615,
+      "step": 6975
+    },
+    {
+      "epoch": 1.9637079758053173,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8182555483794e-06,
+      "loss": 1.6147,
+      "step": 6980
+    },
+    {
+      "epoch": 1.9651146434097622,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.813447745018692e-06,
+      "loss": 1.7156,
+      "step": 6985
+    },
+    {
+      "epoch": 1.9665213110142075,
+      "grad_norm": 3.125,
+      "learning_rate": 4.808638715471101e-06,
+      "loss": 1.5149,
+      "step": 6990
+    },
+    {
+      "epoch": 1.9679279786186523,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.803828466985732e-06,
+      "loss": 1.2237,
+      "step": 6995
+    },
+    {
+      "epoch": 1.9693346462230976,
+      "grad_norm": 3.0,
+      "learning_rate": 4.799017006813527e-06,
+      "loss": 1.5996,
+      "step": 7000
+    },
+    {
+      "epoch": 1.9707413138275425,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.794204342207259e-06,
+      "loss": 1.5614,
+      "step": 7005
+    },
+    {
+      "epoch": 1.9721479814319878,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.789390480421512e-06,
+      "loss": 1.65,
+      "step": 7010
+    },
+    {
+      "epoch": 1.9735546490364326,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.784575428712676e-06,
+      "loss": 1.5145,
+      "step": 7015
+    },
+    {
+      "epoch": 1.9749613166408777,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.7797591943389355e-06,
+      "loss": 1.6309,
+      "step": 7020
+    },
+    {
+      "epoch": 1.9763679842453228,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.774941784560256e-06,
+      "loss": 1.7972,
+      "step": 7025
+    },
+    {
+      "epoch": 1.9777746518497679,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.770123206638376e-06,
+      "loss": 1.6059,
+      "step": 7030
+    },
+    {
+      "epoch": 1.979181319454213,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.765303467836794e-06,
+      "loss": 1.6813,
+      "step": 7035
+    },
+    {
+      "epoch": 1.980587987058658,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.760482575420762e-06,
+      "loss": 1.6176,
+      "step": 7040
+    },
+    {
+      "epoch": 1.981994654663103,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.755660536657266e-06,
+      "loss": 1.689,
+      "step": 7045
+    },
+    {
+      "epoch": 1.9834013222675482,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.7508373588150216e-06,
+      "loss": 1.3935,
+      "step": 7050
+    },
+    {
+      "epoch": 1.9848079898719933,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.746013049164463e-06,
+      "loss": 1.4812,
+      "step": 7055
+    },
+    {
+      "epoch": 1.9862146574764383,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.74118761497773e-06,
+      "loss": 1.7011,
+      "step": 7060
+    },
+    {
+      "epoch": 1.9876213250808834,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7363610635286536e-06,
+      "loss": 1.6525,
+      "step": 7065
+    },
+    {
+      "epoch": 1.9890279926853285,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.731533402092756e-06,
+      "loss": 1.7934,
+      "step": 7070
+    },
+    {
+      "epoch": 1.9904346602897736,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.726704637947228e-06,
+      "loss": 1.2473,
+      "step": 7075
+    },
+    {
+      "epoch": 1.9918413278942186,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.721874778370921e-06,
+      "loss": 1.2767,
+      "step": 7080
+    },
+    {
+      "epoch": 1.9932479954986637,
+      "grad_norm": 4.625,
+      "learning_rate": 4.717043830644344e-06,
+      "loss": 1.5313,
+      "step": 7085
+    },
+    {
+      "epoch": 1.9946546631031086,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.7122118020496385e-06,
+      "loss": 1.6033,
+      "step": 7090
+    },
+    {
+      "epoch": 1.9960613307075539,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.707378699870582e-06,
+      "loss": 1.7435,
+      "step": 7095
+    },
+    {
+      "epoch": 1.9974679983119987,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.702544531392565e-06,
+      "loss": 1.6058,
+      "step": 7100
+    },
+    {
+      "epoch": 1.998874665916444,
+      "grad_norm": 3.875,
+      "learning_rate": 4.697709303902592e-06,
+      "loss": 1.7475,
+      "step": 7105
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.5795072317123413,
+      "eval_runtime": 330.9822,
+      "eval_samples_per_second": 9.541,
+      "eval_steps_per_second": 4.771,
+      "step": 7109
+    },
+    {
+      "epoch": 2.000281333520889,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6928730246892536e-06,
+      "loss": 1.5257,
+      "step": 7110
+    },
+    {
+      "epoch": 2.001688001125334,
+      "grad_norm": 4.125,
+      "learning_rate": 4.6880357010427375e-06,
+      "loss": 1.5725,
+      "step": 7115
+    },
+    {
+      "epoch": 2.003094668729779,
+      "grad_norm": 3.875,
+      "learning_rate": 4.683197340254798e-06,
+      "loss": 1.3852,
+      "step": 7120
+    },
+    {
+      "epoch": 2.0045013363342243,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.678357949618754e-06,
+      "loss": 1.6272,
+      "step": 7125
+    },
+    {
+      "epoch": 2.005908003938669,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.673517536429479e-06,
+      "loss": 1.4325,
+      "step": 7130
+    },
+    {
+      "epoch": 2.0073146715431145,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.6686761079833855e-06,
+      "loss": 1.5526,
+      "step": 7135
+    },
+    {
+      "epoch": 2.0087213391475593,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.663833671578418e-06,
+      "loss": 1.7071,
+      "step": 7140
+    },
+    {
+      "epoch": 2.0101280067520046,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.6589902345140394e-06,
+      "loss": 1.2759,
+      "step": 7145
+    },
+    {
+      "epoch": 2.0115346743564495,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.654145804091223e-06,
+      "loss": 1.6467,
+      "step": 7150
+    },
+    {
+      "epoch": 2.012941341960895,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.649300387612436e-06,
+      "loss": 1.4278,
+      "step": 7155
+    },
+    {
+      "epoch": 2.0143480095653397,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.644453992381633e-06,
+      "loss": 1.4842,
+      "step": 7160
+    },
+    {
+      "epoch": 2.015754677169785,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.639606625704249e-06,
+      "loss": 1.9023,
+      "step": 7165
+    },
+    {
+      "epoch": 2.01716134477423,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.634758294887175e-06,
+      "loss": 1.6057,
+      "step": 7170
+    },
+    {
+      "epoch": 2.018568012378675,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.629909007238762e-06,
+      "loss": 1.7244,
+      "step": 7175
+    },
+    {
+      "epoch": 2.01997467998312,
+      "grad_norm": 4.0,
+      "learning_rate": 4.6250587700688e-06,
+      "loss": 1.8344,
+      "step": 7180
+    },
+    {
+      "epoch": 2.0213813475875653,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.620207590688512e-06,
+      "loss": 1.5223,
+      "step": 7185
+    },
+    {
+      "epoch": 2.02278801519201,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.61535547641054e-06,
+      "loss": 1.7127,
+      "step": 7190
+    },
+    {
+      "epoch": 2.024194682796455,
+      "grad_norm": 3.25,
+      "learning_rate": 4.610502434548934e-06,
+      "loss": 1.8786,
+      "step": 7195
+    },
+    {
+      "epoch": 2.0256013504009003,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.6056484724191476e-06,
+      "loss": 1.5157,
+      "step": 7200
+    },
+    {
+      "epoch": 2.027008018005345,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.600793597338015e-06,
+      "loss": 1.624,
+      "step": 7205
+    },
+    {
+      "epoch": 2.0284146856097904,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.59593781662375e-06,
+      "loss": 1.6772,
+      "step": 7210
+    },
+    {
+      "epoch": 2.0298213532142353,
+      "grad_norm": 3.375,
+      "learning_rate": 4.591081137595933e-06,
+      "loss": 1.3959,
+      "step": 7215
+    },
+    {
+      "epoch": 2.0312280208186806,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5862235675754935e-06,
+      "loss": 1.5757,
+      "step": 7220
+    },
+    {
+      "epoch": 2.0326346884231254,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.58136511388471e-06,
+      "loss": 1.3308,
+      "step": 7225
+    },
+    {
+      "epoch": 2.0340413560275707,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.5765057838471884e-06,
+      "loss": 1.685,
+      "step": 7230
+    },
+    {
+      "epoch": 2.0354480236320156,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.571645584787858e-06,
+      "loss": 1.6677,
+      "step": 7235
+    },
+    {
+      "epoch": 2.036854691236461,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.566784524032958e-06,
+      "loss": 1.5998,
+      "step": 7240
+    },
+    {
+      "epoch": 2.0382613588409058,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.561922608910025e-06,
+      "loss": 1.4078,
+      "step": 7245
+    },
+    {
+      "epoch": 2.039668026445351,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.557059846747886e-06,
+      "loss": 1.5471,
+      "step": 7250
+    },
+    {
+      "epoch": 2.041074694049796,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.5521962448766416e-06,
+      "loss": 1.4645,
+      "step": 7255
+    },
+    {
+      "epoch": 2.042481361654241,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.547331810627661e-06,
+      "loss": 1.4794,
+      "step": 7260
+    },
+    {
+      "epoch": 2.043888029258686,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.542466551333568e-06,
+      "loss": 1.672,
+      "step": 7265
+    },
+    {
+      "epoch": 2.0452946968631314,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.5376004743282255e-06,
+      "loss": 1.573,
+      "step": 7270
+    },
+    {
+      "epoch": 2.0467013644675762,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.532733586946736e-06,
+      "loss": 1.6004,
+      "step": 7275
+    },
+    {
+      "epoch": 2.0481080320720215,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.527865896525419e-06,
+      "loss": 1.7544,
+      "step": 7280
+    },
+    {
+      "epoch": 2.0495146996764664,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.522997410401805e-06,
+      "loss": 1.4427,
+      "step": 7285
+    },
+    {
+      "epoch": 2.0509213672809117,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.518128135914625e-06,
+      "loss": 1.5777,
+      "step": 7290
+    },
+    {
+      "epoch": 2.0523280348853565,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.5132580804037984e-06,
+      "loss": 1.3477,
+      "step": 7295
+    },
+    {
+      "epoch": 2.053734702489802,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.50838725121042e-06,
+      "loss": 1.6699,
+      "step": 7300
+    },
+    {
+      "epoch": 2.0551413700942467,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.5035156556767555e-06,
+      "loss": 1.5229,
+      "step": 7305
+    },
+    {
+      "epoch": 2.056548037698692,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.498643301146219e-06,
+      "loss": 1.7442,
+      "step": 7310
+    },
+    {
+      "epoch": 2.057954705303137,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.493770194963374e-06,
+      "loss": 1.4243,
+      "step": 7315
+    },
+    {
+      "epoch": 2.059361372907582,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.488896344473914e-06,
+      "loss": 1.653,
+      "step": 7320
+    },
+    {
+      "epoch": 2.060768040512027,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.484021757024658e-06,
+      "loss": 1.6292,
+      "step": 7325
+    },
+    {
+      "epoch": 2.062174708116472,
+      "grad_norm": 3.0,
+      "learning_rate": 4.479146439963533e-06,
+      "loss": 1.4996,
+      "step": 7330
+    },
+    {
+      "epoch": 2.063581375720917,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.474270400639565e-06,
+      "loss": 1.7773,
+      "step": 7335
+    },
+    {
+      "epoch": 2.064988043325362,
+      "grad_norm": 2.25,
+      "learning_rate": 4.469393646402872e-06,
+      "loss": 1.7348,
+      "step": 7340
+    },
+    {
+      "epoch": 2.0663947109298073,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.4645161846046465e-06,
+      "loss": 1.351,
+      "step": 7345
+    },
+    {
+      "epoch": 2.067801378534252,
+      "grad_norm": 3.875,
+      "learning_rate": 4.459638022597149e-06,
+      "loss": 1.1414,
+      "step": 7350
+    },
+    {
+      "epoch": 2.0692080461386975,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.454759167733697e-06,
+      "loss": 1.4292,
+      "step": 7355
+    },
+    {
+      "epoch": 2.0706147137431423,
+      "grad_norm": 3.375,
+      "learning_rate": 4.449879627368649e-06,
+      "loss": 1.608,
+      "step": 7360
+    },
+    {
+      "epoch": 2.0720213813475876,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.4449994088574e-06,
+      "loss": 1.5892,
+      "step": 7365
+    },
+    {
+      "epoch": 2.0734280489520325,
+      "grad_norm": 4.875,
+      "learning_rate": 4.440118519556366e-06,
+      "loss": 1.5495,
+      "step": 7370
+    },
+    {
+      "epoch": 2.0748347165564778,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.435236966822972e-06,
+      "loss": 1.7177,
+      "step": 7375
+    },
+    {
+      "epoch": 2.0762413841609226,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.430354758015648e-06,
+      "loss": 1.4,
+      "step": 7380
+    },
+    {
+      "epoch": 2.077648051765368,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.425471900493806e-06,
+      "loss": 1.6965,
+      "step": 7385
+    },
+    {
+      "epoch": 2.079054719369813,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.420588401617845e-06,
+      "loss": 1.4151,
+      "step": 7390
+    },
+    {
+      "epoch": 2.080461386974258,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.415704268749123e-06,
+      "loss": 1.5518,
+      "step": 7395
+    },
+    {
+      "epoch": 2.081868054578703,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.410819509249956e-06,
+      "loss": 1.4657,
+      "step": 7400
+    },
+    {
+      "epoch": 2.0832747221831482,
+      "grad_norm": 3.25,
+      "learning_rate": 4.405934130483606e-06,
+      "loss": 1.5352,
+      "step": 7405
+    },
+    {
+      "epoch": 2.084681389787593,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.401048139814268e-06,
+      "loss": 1.6914,
+      "step": 7410
+    },
+    {
+      "epoch": 2.0860880573920384,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3961615446070564e-06,
+      "loss": 1.5646,
+      "step": 7415
+    },
+    {
+      "epoch": 2.0874947249964833,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.391274352228002e-06,
+      "loss": 1.5388,
+      "step": 7420
+    },
+    {
+      "epoch": 2.0889013926009286,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.3863865700440316e-06,
+      "loss": 1.5648,
+      "step": 7425
+    },
+    {
+      "epoch": 2.0903080602053734,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.3814982054229604e-06,
+      "loss": 1.455,
+      "step": 7430
+    },
+    {
+      "epoch": 2.0917147278098187,
+      "grad_norm": 4.125,
+      "learning_rate": 4.37660926573349e-06,
+      "loss": 1.5373,
+      "step": 7435
+    },
+    {
+      "epoch": 2.0931213954142636,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.371719758345176e-06,
+      "loss": 1.5629,
+      "step": 7440
+    },
+    {
+      "epoch": 2.094528063018709,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.366829690628439e-06,
+      "loss": 1.7871,
+      "step": 7445
+    },
+    {
+      "epoch": 2.0959347306231537,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.3619390699545425e-06,
+      "loss": 1.4415,
+      "step": 7450
+    },
+    {
+      "epoch": 2.097341398227599,
+      "grad_norm": 3.0,
+      "learning_rate": 4.357047903695582e-06,
+      "loss": 1.4951,
+      "step": 7455
+    },
+    {
+      "epoch": 2.098748065832044,
+      "grad_norm": 3.125,
+      "learning_rate": 4.352156199224474e-06,
+      "loss": 1.7415,
+      "step": 7460
+    },
+    {
+      "epoch": 2.100154733436489,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.347263963914951e-06,
+      "loss": 1.5393,
+      "step": 7465
+    },
+    {
+      "epoch": 2.101561401040934,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.3423712051415415e-06,
+      "loss": 1.5584,
+      "step": 7470
+    },
+    {
+      "epoch": 2.102968068645379,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.337477930279565e-06,
+      "loss": 1.4683,
+      "step": 7475
+    },
+    {
+      "epoch": 2.104374736249824,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.332584146705119e-06,
+      "loss": 1.7479,
+      "step": 7480
+    },
+    {
+      "epoch": 2.105781403854269,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.327689861795066e-06,
+      "loss": 1.6146,
+      "step": 7485
+    },
+    {
+      "epoch": 2.1071880714587143,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.322795082927027e-06,
+      "loss": 1.5614,
+      "step": 7490
+    },
+    {
+      "epoch": 2.108594739063159,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.317899817479363e-06,
+      "loss": 1.4357,
+      "step": 7495
+    },
+    {
+      "epoch": 2.1100014066676045,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.313004072831177e-06,
+      "loss": 1.6396,
+      "step": 7500
+    },
+    {
+      "epoch": 2.1114080742720494,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.308107856362284e-06,
+      "loss": 1.4383,
+      "step": 7505
+    },
+    {
+      "epoch": 2.1128147418764947,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.303211175453216e-06,
+      "loss": 1.3966,
+      "step": 7510
+    },
+    {
+      "epoch": 2.1142214094809395,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.2983140374852076e-06,
+      "loss": 1.3908,
+      "step": 7515
+    },
+    {
+      "epoch": 2.115628077085385,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.293416449840175e-06,
+      "loss": 1.6366,
+      "step": 7520
+    },
+    {
+      "epoch": 2.1170347446898297,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.288518419900718e-06,
+      "loss": 1.596,
+      "step": 7525
+    },
+    {
+      "epoch": 2.118441412294275,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.2836199550501e-06,
+      "loss": 1.7149,
+      "step": 7530
+    },
+    {
+      "epoch": 2.11984807989872,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.278721062672244e-06,
+      "loss": 1.4818,
+      "step": 7535
+    },
+    {
+      "epoch": 2.121254747503165,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.273821750151712e-06,
+      "loss": 1.4746,
+      "step": 7540
+    },
+    {
+      "epoch": 2.12266141510761,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.268922024873705e-06,
+      "loss": 1.7616,
+      "step": 7545
+    },
+    {
+      "epoch": 2.1240680827120553,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.264021894224042e-06,
+      "loss": 1.6073,
+      "step": 7550
+    },
+    {
+      "epoch": 2.1254747503165,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.259121365589152e-06,
+      "loss": 1.4738,
+      "step": 7555
+    },
+    {
+      "epoch": 2.1268814179209454,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.25422044635607e-06,
+      "loss": 1.7426,
+      "step": 7560
+    },
+    {
+      "epoch": 2.1282880855253903,
+      "grad_norm": 4.25,
+      "learning_rate": 4.249319143912415e-06,
+      "loss": 1.595,
+      "step": 7565
+    },
+    {
+      "epoch": 2.1296947531298356,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.244417465646382e-06,
+      "loss": 1.7493,
+      "step": 7570
+    },
+    {
+      "epoch": 2.1311014207342804,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.239515418946739e-06,
+      "loss": 1.5547,
+      "step": 7575
+    },
+    {
+      "epoch": 2.1325080883387257,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.234613011202804e-06,
+      "loss": 1.4594,
+      "step": 7580
+    },
+    {
+      "epoch": 2.1339147559431706,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.22971024980444e-06,
+      "loss": 1.7132,
+      "step": 7585
+    },
+    {
+      "epoch": 2.135321423547616,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.2248071421420445e-06,
+      "loss": 1.5946,
+      "step": 7590
+    },
+    {
+      "epoch": 2.1367280911520607,
+      "grad_norm": 3.5,
+      "learning_rate": 4.219903695606538e-06,
+      "loss": 1.3981,
+      "step": 7595
+    },
+    {
+      "epoch": 2.1381347587565056,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.214999917589347e-06,
+      "loss": 1.463,
+      "step": 7600
+    },
+    {
+      "epoch": 2.139541426360951,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.210095815482404e-06,
+      "loss": 1.3421,
+      "step": 7605
+    },
+    {
+      "epoch": 2.140948093965396,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.205191396678126e-06,
+      "loss": 1.2721,
+      "step": 7610
+    },
+    {
+      "epoch": 2.142354761569841,
+      "grad_norm": 3.75,
+      "learning_rate": 4.200286668569407e-06,
+      "loss": 1.6508,
+      "step": 7615
+    },
+    {
+      "epoch": 2.143761429174286,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.195381638549609e-06,
+      "loss": 1.569,
+      "step": 7620
+    },
+    {
+      "epoch": 2.145168096778731,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.190476314012551e-06,
+      "loss": 1.5305,
+      "step": 7625
+    },
+    {
+      "epoch": 2.146574764383176,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.185570702352491e-06,
+      "loss": 1.4491,
+      "step": 7630
+    },
+    {
+      "epoch": 2.1479814319876214,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.180664810964121e-06,
+      "loss": 1.656,
+      "step": 7635
+    },
+    {
+      "epoch": 2.1493880995920662,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.175758647242561e-06,
+      "loss": 1.4181,
+      "step": 7640
+    },
+    {
+      "epoch": 2.1507947671965115,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.170852218583333e-06,
+      "loss": 1.3543,
+      "step": 7645
+    },
+    {
+      "epoch": 2.1522014348009564,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.1659455323823615e-06,
+      "loss": 1.6262,
+      "step": 7650
+    },
+    {
+      "epoch": 2.1536081024054017,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.161038596035963e-06,
+      "loss": 1.6097,
+      "step": 7655
+    },
+    {
+      "epoch": 2.1550147700098465,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.156131416940824e-06,
+      "loss": 1.6947,
+      "step": 7660
+    },
+    {
+      "epoch": 2.156421437614292,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.151224002494002e-06,
+      "loss": 1.5401,
+      "step": 7665
+    },
+    {
+      "epoch": 2.1578281052187367,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.146316360092909e-06,
+      "loss": 1.6713,
+      "step": 7670
+    },
+    {
+      "epoch": 2.159234772823182,
+      "grad_norm": 4.0,
+      "learning_rate": 4.141408497135299e-06,
+      "loss": 1.5029,
+      "step": 7675
+    },
+    {
+      "epoch": 2.160641440427627,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.136500421019258e-06,
+      "loss": 1.5932,
+      "step": 7680
+    },
+    {
+      "epoch": 2.162048108032072,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.131592139143195e-06,
+      "loss": 1.6317,
+      "step": 7685
+    },
+    {
+      "epoch": 2.163454775636517,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.126683658905829e-06,
+      "loss": 1.603,
+      "step": 7690
+    },
+    {
+      "epoch": 2.1648614432409623,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.121774987706177e-06,
+      "loss": 1.5748,
+      "step": 7695
+    },
+    {
+      "epoch": 2.166268110845407,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.116866132943544e-06,
+      "loss": 1.4496,
+      "step": 7700
+    },
+    {
+      "epoch": 2.1676747784498525,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.111957102017513e-06,
+      "loss": 1.3999,
+      "step": 7705
+    },
+    {
+      "epoch": 2.1690814460542973,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.10704790232793e-06,
+      "loss": 1.6404,
+      "step": 7710
+    },
+    {
+      "epoch": 2.1704881136587426,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.102138541274898e-06,
+      "loss": 1.6583,
+      "step": 7715
+    },
+    {
+      "epoch": 2.1718947812631875,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.097229026258762e-06,
+      "loss": 1.6281,
+      "step": 7720
+    },
+    {
+      "epoch": 2.1733014488676328,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.092319364680101e-06,
+      "loss": 1.6493,
+      "step": 7725
+    },
+    {
+      "epoch": 2.1747081164720776,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.08740956393971e-06,
+      "loss": 1.7934,
+      "step": 7730
+    },
+    {
+      "epoch": 2.176114784076523,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.082499631438599e-06,
+      "loss": 1.6603,
+      "step": 7735
+    },
+    {
+      "epoch": 2.177521451680968,
+      "grad_norm": 4.25,
+      "learning_rate": 4.077589574577975e-06,
+      "loss": 1.5522,
+      "step": 7740
+    },
+    {
+      "epoch": 2.1789281192854126,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.07267940075923e-06,
+      "loss": 1.6806,
+      "step": 7745
+    },
+    {
+      "epoch": 2.180334786889858,
+      "grad_norm": 4.375,
+      "learning_rate": 4.067769117383936e-06,
+      "loss": 1.5831,
+      "step": 7750
+    },
+    {
+      "epoch": 2.181741454494303,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.0628587318538295e-06,
+      "loss": 1.5568,
+      "step": 7755
+    },
+    {
+      "epoch": 2.183148122098748,
+      "grad_norm": 2.875,
+      "learning_rate": 4.057948251570798e-06,
+      "loss": 1.6057,
+      "step": 7760
+    },
+    {
+      "epoch": 2.184554789703193,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.053037683936875e-06,
+      "loss": 1.3439,
+      "step": 7765
+    },
+    {
+      "epoch": 2.1859614573076382,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.048127036354224e-06,
+      "loss": 1.4215,
+      "step": 7770
+    },
+    {
+      "epoch": 2.187368124912083,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.0432163162251295e-06,
+      "loss": 1.4706,
+      "step": 7775
+    },
+    {
+      "epoch": 2.1887747925165284,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.038305530951986e-06,
+      "loss": 1.5168,
+      "step": 7780
+    },
+    {
+      "epoch": 2.1901814601209733,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.033394687937284e-06,
+      "loss": 1.7687,
+      "step": 7785
+    },
+    {
+      "epoch": 2.1915881277254186,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.028483794583606e-06,
+      "loss": 1.6538,
+      "step": 7790
+    },
+    {
+      "epoch": 2.1929947953298634,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.023572858293602e-06,
+      "loss": 1.6807,
+      "step": 7795
+    },
+    {
+      "epoch": 2.1944014629343087,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.018661886469996e-06,
+      "loss": 1.5059,
+      "step": 7800
+    },
+    {
+      "epoch": 2.1958081305387536,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.01375088651556e-06,
+      "loss": 1.5728,
+      "step": 7805
+    },
+    {
+      "epoch": 2.197214798143199,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.008839865833108e-06,
+      "loss": 1.5053,
+      "step": 7810
+    },
+    {
+      "epoch": 2.1986214657476437,
+      "grad_norm": 3.0,
+      "learning_rate": 4.0039288318254895e-06,
+      "loss": 1.544,
+      "step": 7815
+    },
+    {
+      "epoch": 2.200028133352089,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.999017791895571e-06,
+      "loss": 1.4862,
+      "step": 7820
+    },
+    {
+      "epoch": 2.201434800956534,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.994106753446225e-06,
+      "loss": 1.6569,
+      "step": 7825
+    },
+    {
+      "epoch": 2.202841468560979,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.989195723880332e-06,
+      "loss": 1.484,
+      "step": 7830
+    },
+    {
+      "epoch": 2.204248136165424,
+      "grad_norm": 2.375,
+      "learning_rate": 3.984284710600746e-06,
+      "loss": 1.4287,
+      "step": 7835
+    },
+    {
+      "epoch": 2.2056548037698693,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.979373721010306e-06,
+      "loss": 1.7917,
+      "step": 7840
+    },
+    {
+      "epoch": 2.207061471374314,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.97446276251181e-06,
+      "loss": 1.5958,
+      "step": 7845
+    },
+    {
+      "epoch": 2.2084681389787595,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.969551842508014e-06,
+      "loss": 1.7798,
+      "step": 7850
+    },
+    {
+      "epoch": 2.2098748065832043,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.964640968401612e-06,
+      "loss": 1.6493,
+      "step": 7855
+    },
+    {
+      "epoch": 2.2112814741876496,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.959730147595228e-06,
+      "loss": 1.3387,
+      "step": 7860
+    },
+    {
+      "epoch": 2.2126881417920945,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.954819387491411e-06,
+      "loss": 1.6065,
+      "step": 7865
+    },
+    {
+      "epoch": 2.21409480939654,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.949908695492612e-06,
+      "loss": 1.4857,
+      "step": 7870
+    },
+    {
+      "epoch": 2.2155014770009847,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.944998079001185e-06,
+      "loss": 1.6135,
+      "step": 7875
+    },
+    {
+      "epoch": 2.21690814460543,
+      "grad_norm": 3.625,
+      "learning_rate": 3.940087545419365e-06,
+      "loss": 1.4992,
+      "step": 7880
+    },
+    {
+      "epoch": 2.218314812209875,
+      "grad_norm": 5.03125,
+      "learning_rate": 3.9351771021492686e-06,
+      "loss": 1.4218,
+      "step": 7885
+    },
+    {
+      "epoch": 2.2197214798143197,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.9302667565928676e-06,
+      "loss": 1.6094,
+      "step": 7890
+    },
+    {
+      "epoch": 2.221128147418765,
+      "grad_norm": 2.375,
+      "learning_rate": 3.925356516151996e-06,
+      "loss": 1.4345,
+      "step": 7895
+    },
+    {
+      "epoch": 2.22253481502321,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.920446388228319e-06,
+      "loss": 1.3853,
+      "step": 7900
+    },
+    {
+      "epoch": 2.223941482627655,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.915536380223344e-06,
+      "loss": 1.6986,
+      "step": 7905
+    },
+    {
+      "epoch": 2.2253481502321,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.910626499538387e-06,
+      "loss": 1.8287,
+      "step": 7910
+    },
+    {
+      "epoch": 2.2267548178365453,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.9057167535745795e-06,
+      "loss": 1.453,
+      "step": 7915
+    },
+    {
+      "epoch": 2.22816148544099,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.900807149732843e-06,
+      "loss": 1.5434,
+      "step": 7920
+    },
+    {
+      "epoch": 2.2295681530454354,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.895897695413892e-06,
+      "loss": 1.6757,
+      "step": 7925
+    },
+    {
+      "epoch": 2.2309748206498803,
+      "grad_norm": 3.25,
+      "learning_rate": 3.890988398018212e-06,
+      "loss": 1.4654,
+      "step": 7930
+    },
+    {
+      "epoch": 2.2323814882543256,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.886079264946052e-06,
+      "loss": 1.4485,
+      "step": 7935
+    },
+    {
+      "epoch": 2.2337881558587704,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.881170303597412e-06,
+      "loss": 1.5495,
+      "step": 7940
+    },
+    {
+      "epoch": 2.2351948234632157,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.8762615213720365e-06,
+      "loss": 1.4623,
+      "step": 7945
+    },
+    {
+      "epoch": 2.2366014910676606,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.871352925669398e-06,
+      "loss": 1.7347,
+      "step": 7950
+    },
+    {
+      "epoch": 2.238008158672106,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.866444523888687e-06,
+      "loss": 1.5133,
+      "step": 7955
+    },
+    {
+      "epoch": 2.2394148262765508,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.861536323428805e-06,
+      "loss": 1.6838,
+      "step": 7960
+    },
+    {
+      "epoch": 2.240821493880996,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.856628331688346e-06,
+      "loss": 1.3434,
+      "step": 7965
+    },
+    {
+      "epoch": 2.242228161485441,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.8517205560655895e-06,
+      "loss": 1.6505,
+      "step": 7970
+    },
+    {
+      "epoch": 2.243634829089886,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.846813003958493e-06,
+      "loss": 1.5625,
+      "step": 7975
+    },
+    {
+      "epoch": 2.245041496694331,
+      "grad_norm": 2.375,
+      "learning_rate": 3.841905682764676e-06,
+      "loss": 1.6773,
+      "step": 7980
+    },
+    {
+      "epoch": 2.2464481642987764,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.836998599881406e-06,
+      "loss": 1.6122,
+      "step": 7985
+    },
+    {
+      "epoch": 2.247854831903221,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.832091762705595e-06,
+      "loss": 1.705,
+      "step": 7990
+    },
+    {
+      "epoch": 2.2492614995076665,
+      "grad_norm": 3.0,
+      "learning_rate": 3.827185178633787e-06,
+      "loss": 1.5123,
+      "step": 7995
+    },
+    {
+      "epoch": 2.2506681671121114,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.822278855062136e-06,
+      "loss": 1.4593,
+      "step": 8000
+    },
+    {
+      "epoch": 2.2520748347165567,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.8173727993864115e-06,
+      "loss": 1.4751,
+      "step": 8005
+    },
+    {
+      "epoch": 2.2534815023210015,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.8124670190019755e-06,
+      "loss": 1.3753,
+      "step": 8010
+    },
+    {
+      "epoch": 2.2548881699254464,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.807561521303777e-06,
+      "loss": 1.6663,
+      "step": 8015
+    },
+    {
+      "epoch": 2.2562948375298917,
+      "grad_norm": 3.0,
+      "learning_rate": 3.802656313686336e-06,
+      "loss": 1.5228,
+      "step": 8020
+    },
+    {
+      "epoch": 2.257701505134337,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.7977514035437383e-06,
+      "loss": 1.6107,
+      "step": 8025
+    },
+    {
+      "epoch": 2.259108172738782,
+      "grad_norm": 3.25,
+      "learning_rate": 3.7928467982696174e-06,
+      "loss": 1.6885,
+      "step": 8030
+    },
+    {
+      "epoch": 2.2605148403432267,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.7879425052571525e-06,
+      "loss": 1.5311,
+      "step": 8035
+    },
+    {
+      "epoch": 2.261921507947672,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.783038531899047e-06,
+      "loss": 1.681,
+      "step": 8040
+    },
+    {
+      "epoch": 2.263328175552117,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.7781348855875263e-06,
+      "loss": 1.1735,
+      "step": 8045
+    },
+    {
+      "epoch": 2.264734843156562,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.7732315737143205e-06,
+      "loss": 1.615,
+      "step": 8050
+    },
+    {
+      "epoch": 2.266141510761007,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.768328603670658e-06,
+      "loss": 1.4219,
+      "step": 8055
+    },
+    {
+      "epoch": 2.2675481783654523,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.7634259828472467e-06,
+      "loss": 1.5921,
+      "step": 8060
+    },
+    {
+      "epoch": 2.268954845969897,
+      "grad_norm": 4.125,
+      "learning_rate": 3.7585237186342743e-06,
+      "loss": 1.6932,
+      "step": 8065
+    },
+    {
+      "epoch": 2.2703615135743425,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.753621818421388e-06,
+      "loss": 1.7495,
+      "step": 8070
+    },
+    {
+      "epoch": 2.2717681811787873,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.7487202895976864e-06,
+      "loss": 1.6282,
+      "step": 8075
+    },
+    {
+      "epoch": 2.2731748487832326,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.743819139551708e-06,
+      "loss": 1.6892,
+      "step": 8080
+    },
+    {
+      "epoch": 2.2745815163876775,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.7389183756714207e-06,
+      "loss": 1.5165,
+      "step": 8085
+    },
+    {
+      "epoch": 2.2759881839921228,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.7340180053442127e-06,
+      "loss": 1.6025,
+      "step": 8090
+    },
+    {
+      "epoch": 2.2773948515965676,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.7291180359568735e-06,
+      "loss": 1.7114,
+      "step": 8095
+    },
+    {
+      "epoch": 2.278801519201013,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.724218474895593e-06,
+      "loss": 1.6153,
+      "step": 8100
+    },
+    {
+      "epoch": 2.280208186805458,
+      "grad_norm": 2.875,
+      "learning_rate": 3.719319329545943e-06,
+      "loss": 1.5714,
+      "step": 8105
+    },
+    {
+      "epoch": 2.281614854409903,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.7144206072928704e-06,
+      "loss": 1.7677,
+      "step": 8110
+    },
+    {
+      "epoch": 2.283021522014348,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.709522315520683e-06,
+      "loss": 1.4741,
+      "step": 8115
+    },
+    {
+      "epoch": 2.2844281896187932,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.704624461613043e-06,
+      "loss": 1.7835,
+      "step": 8120
+    },
+    {
+      "epoch": 2.285834857223238,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.6997270529529445e-06,
+      "loss": 1.8235,
+      "step": 8125
+    },
+    {
+      "epoch": 2.2872415248276834,
+      "grad_norm": 4.25,
+      "learning_rate": 3.69483009692272e-06,
+      "loss": 1.5382,
+      "step": 8130
+    },
+    {
+      "epoch": 2.2886481924321282,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.6899336009040132e-06,
+      "loss": 1.5006,
+      "step": 8135
+    },
+    {
+      "epoch": 2.2900548600365735,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.685037572277778e-06,
+      "loss": 1.8186,
+      "step": 8140
+    },
+    {
+      "epoch": 2.2914615276410184,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6801420184242626e-06,
+      "loss": 1.7269,
+      "step": 8145
+    },
+    {
+      "epoch": 2.2928681952454637,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.6752469467229975e-06,
+      "loss": 1.5189,
+      "step": 8150
+    },
+    {
+      "epoch": 2.2942748628499086,
+      "grad_norm": 3.75,
+      "learning_rate": 3.6703523645527915e-06,
+      "loss": 1.6616,
+      "step": 8155
+    },
+    {
+      "epoch": 2.2956815304543534,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.6654582792917074e-06,
+      "loss": 1.6282,
+      "step": 8160
+    },
+    {
+      "epoch": 2.2970881980587987,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.660564698317069e-06,
+      "loss": 1.5604,
+      "step": 8165
+    },
+    {
+      "epoch": 2.298494865663244,
+      "grad_norm": 4.5625,
+      "learning_rate": 3.6556716290054306e-06,
+      "loss": 1.4532,
+      "step": 8170
+    },
+    {
+      "epoch": 2.299901533267689,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.650779078732582e-06,
+      "loss": 1.7224,
+      "step": 8175
+    },
+    {
+      "epoch": 2.3013082008721337,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.6458870548735255e-06,
+      "loss": 1.6006,
+      "step": 8180
+    },
+    {
+      "epoch": 2.302714868476579,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.6409955648024756e-06,
+      "loss": 1.3733,
+      "step": 8185
+    },
+    {
+      "epoch": 2.304121536081024,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.6361046158928343e-06,
+      "loss": 1.6563,
+      "step": 8190
+    },
+    {
+      "epoch": 2.305528203685469,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.631214215517198e-06,
+      "loss": 1.7411,
+      "step": 8195
+    },
+    {
+      "epoch": 2.306934871289914,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.6263243710473258e-06,
+      "loss": 1.5237,
+      "step": 8200
+    },
+    {
+      "epoch": 2.3083415388943593,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.621435089854146e-06,
+      "loss": 1.685,
+      "step": 8205
+    },
+    {
+      "epoch": 2.309748206498804,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.616546379307736e-06,
+      "loss": 1.3735,
+      "step": 8210
+    },
+    {
+      "epoch": 2.3111548741032495,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.611658246777311e-06,
+      "loss": 1.5922,
+      "step": 8215
+    },
+    {
+      "epoch": 2.3125615417076943,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.6067706996312196e-06,
+      "loss": 1.6493,
+      "step": 8220
+    },
+    {
+      "epoch": 2.3139682093121396,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.601883745236919e-06,
+      "loss": 1.3882,
+      "step": 8225
+    },
+    {
+      "epoch": 2.3153748769165845,
+      "grad_norm": 4.46875,
+      "learning_rate": 3.5969973909609857e-06,
+      "loss": 1.483,
+      "step": 8230
+    },
+    {
+      "epoch": 2.31678154452103,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.592111644169079e-06,
+      "loss": 1.3885,
+      "step": 8235
+    },
+    {
+      "epoch": 2.3181882121254747,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.5872265122259517e-06,
+      "loss": 1.4347,
+      "step": 8240
+    },
+    {
+      "epoch": 2.31959487972992,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.5823420024954233e-06,
+      "loss": 1.7898,
+      "step": 8245
+    },
+    {
+      "epoch": 2.321001547334365,
+      "grad_norm": 3.125,
+      "learning_rate": 3.577458122340382e-06,
+      "loss": 1.4103,
+      "step": 8250
+    },
+    {
+      "epoch": 2.32240821493881,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.572574879122758e-06,
+      "loss": 1.644,
+      "step": 8255
+    },
+    {
+      "epoch": 2.323814882543255,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.5676922802035324e-06,
+      "loss": 1.5583,
+      "step": 8260
+    },
+    {
+      "epoch": 2.3252215501477003,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.562810332942705e-06,
+      "loss": 1.4209,
+      "step": 8265
+    },
+    {
+      "epoch": 2.326628217752145,
+      "grad_norm": 4.75,
+      "learning_rate": 3.5579290446992996e-06,
+      "loss": 1.476,
+      "step": 8270
+    },
+    {
+      "epoch": 2.3280348853565904,
+      "grad_norm": 3.875,
+      "learning_rate": 3.553048422831344e-06,
+      "loss": 1.5637,
+      "step": 8275
+    },
+    {
+      "epoch": 2.3294415529610353,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.548168474695862e-06,
+      "loss": 1.4261,
+      "step": 8280
+    },
+    {
+      "epoch": 2.33084822056548,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.5432892076488636e-06,
+      "loss": 1.4481,
+      "step": 8285
+    },
+    {
+      "epoch": 2.3322548881699254,
+      "grad_norm": 3.25,
+      "learning_rate": 3.5384106290453275e-06,
+      "loss": 1.6372,
+      "step": 8290
+    },
+    {
+      "epoch": 2.3336615557743707,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.5335327462392014e-06,
+      "loss": 1.6052,
+      "step": 8295
+    },
+    {
+      "epoch": 2.3350682233788156,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.5286555665833763e-06,
+      "loss": 1.5632,
+      "step": 8300
+    },
+    {
+      "epoch": 2.3364748909832604,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.52377909742969e-06,
+      "loss": 1.6063,
+      "step": 8305
+    },
+    {
+      "epoch": 2.3378815585877057,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.5189033461289057e-06,
+      "loss": 1.1783,
+      "step": 8310
+    },
+    {
+      "epoch": 2.339288226192151,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.514028320030706e-06,
+      "loss": 1.6411,
+      "step": 8315
+    },
+    {
+      "epoch": 2.340694893796596,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5091540264836788e-06,
+      "loss": 1.5097,
+      "step": 8320
+    },
+    {
+      "epoch": 2.3421015614010408,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.5042804728353112e-06,
+      "loss": 1.6188,
+      "step": 8325
+    },
+    {
+      "epoch": 2.343508229005486,
+      "grad_norm": 3.875,
+      "learning_rate": 3.499407666431969e-06,
+      "loss": 1.5711,
+      "step": 8330
+    },
+    {
+      "epoch": 2.344914896609931,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.4945356146188977e-06,
+      "loss": 1.2004,
+      "step": 8335
+    },
+    {
+      "epoch": 2.346321564214376,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.489664324740201e-06,
+      "loss": 1.7899,
+      "step": 8340
+    },
+    {
+      "epoch": 2.347728231818821,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.4847938041388376e-06,
+      "loss": 1.6106,
+      "step": 8345
+    },
+    {
+      "epoch": 2.3491348994232664,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4799240601566036e-06,
+      "loss": 1.4558,
+      "step": 8350
+    },
+    {
+      "epoch": 2.350541567027711,
+      "grad_norm": 3.25,
+      "learning_rate": 3.4750551001341257e-06,
+      "loss": 1.4972,
+      "step": 8355
+    },
+    {
+      "epoch": 2.3519482346321565,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.4701869314108503e-06,
+      "loss": 1.4276,
+      "step": 8360
+    },
+    {
+      "epoch": 2.3533549022366014,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.465319561325027e-06,
+      "loss": 1.6449,
+      "step": 8365
+    },
+    {
+      "epoch": 2.3547615698410467,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.460452997213707e-06,
+      "loss": 1.6125,
+      "step": 8370
+    },
+    {
+      "epoch": 2.3561682374454915,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.4555872464127207e-06,
+      "loss": 1.443,
+      "step": 8375
+    },
+    {
+      "epoch": 2.357574905049937,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.4507223162566776e-06,
+      "loss": 1.5425,
+      "step": 8380
+    },
+    {
+      "epoch": 2.3589815726543817,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.445858214078946e-06,
+      "loss": 1.5239,
+      "step": 8385
+    },
+    {
+      "epoch": 2.360388240258827,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.440994947211652e-06,
+      "loss": 1.3468,
+      "step": 8390
+    },
+    {
+      "epoch": 2.361794907863272,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.4361325229856537e-06,
+      "loss": 1.7437,
+      "step": 8395
+    },
+    {
+      "epoch": 2.363201575467717,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.4312709487305474e-06,
+      "loss": 1.6956,
+      "step": 8400
+    },
+    {
+      "epoch": 2.364608243072162,
+      "grad_norm": 5.5,
+      "learning_rate": 3.4264102317746424e-06,
+      "loss": 1.631,
+      "step": 8405
+    },
+    {
+      "epoch": 2.3660149106766073,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.4215503794449613e-06,
+      "loss": 1.6996,
+      "step": 8410
+    },
+    {
+      "epoch": 2.367421578281052,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.416691399067217e-06,
+      "loss": 1.6483,
+      "step": 8415
+    },
+    {
+      "epoch": 2.3688282458854975,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.4118332979658116e-06,
+      "loss": 1.5546,
+      "step": 8420
+    },
+    {
+      "epoch": 2.3702349134899423,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.406976083463824e-06,
+      "loss": 1.7836,
+      "step": 8425
+    },
+    {
+      "epoch": 2.371641581094387,
+      "grad_norm": 4.21875,
+      "learning_rate": 3.4021197628829902e-06,
+      "loss": 1.5147,
+      "step": 8430
+    },
+    {
+      "epoch": 2.3730482486988325,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.3972643435437062e-06,
+      "loss": 1.6279,
+      "step": 8435
+    },
+    {
+      "epoch": 2.3744549163032778,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.392409832765002e-06,
+      "loss": 1.8905,
+      "step": 8440
+    },
+    {
+      "epoch": 2.3758615839077226,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.387556237864545e-06,
+      "loss": 1.6089,
+      "step": 8445
+    },
+    {
+      "epoch": 2.3772682515121675,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.3827035661586165e-06,
+      "loss": 1.3313,
+      "step": 8450
+    },
+    {
+      "epoch": 2.3786749191166128,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.3778518249621117e-06,
+      "loss": 1.5232,
+      "step": 8455
+    },
+    {
+      "epoch": 2.3800815867210576,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.3730010215885155e-06,
+      "loss": 1.6252,
+      "step": 8460
+    },
+    {
+      "epoch": 2.381488254325503,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.368151163349907e-06,
+      "loss": 1.397,
+      "step": 8465
+    },
+    {
+      "epoch": 2.382894921929948,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.363302257556935e-06,
+      "loss": 1.6473,
+      "step": 8470
+    },
+    {
+      "epoch": 2.384301589534393,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.3584543115188167e-06,
+      "loss": 1.6578,
+      "step": 8475
+    },
+    {
+      "epoch": 2.385708257138838,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.353607332543319e-06,
+      "loss": 1.4938,
+      "step": 8480
+    },
+    {
+      "epoch": 2.3871149247432832,
+      "grad_norm": 2.375,
+      "learning_rate": 3.348761327936755e-06,
+      "loss": 1.5833,
+      "step": 8485
+    },
+    {
+      "epoch": 2.388521592347728,
+      "grad_norm": 3.125,
+      "learning_rate": 3.3439163050039637e-06,
+      "loss": 1.5118,
+      "step": 8490
+    },
+    {
+      "epoch": 2.3899282599521734,
+      "grad_norm": 5.875,
+      "learning_rate": 3.339072271048308e-06,
+      "loss": 1.635,
+      "step": 8495
+    },
+    {
+      "epoch": 2.3913349275566183,
+      "grad_norm": 3.0,
+      "learning_rate": 3.3342292333716626e-06,
+      "loss": 1.5718,
+      "step": 8500
+    },
+    {
+      "epoch": 2.3927415951610636,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.3293871992743935e-06,
+      "loss": 1.6902,
+      "step": 8505
+    },
+    {
+      "epoch": 2.3941482627655084,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.32454617605536e-06,
+      "loss": 1.6936,
+      "step": 8510
+    },
+    {
+      "epoch": 2.3955549303699537,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.3197061710118926e-06,
+      "loss": 1.6683,
+      "step": 8515
+    },
+    {
+      "epoch": 2.3969615979743986,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.314867191439794e-06,
+      "loss": 1.5576,
+      "step": 8520
+    },
+    {
+      "epoch": 2.398368265578844,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.3100292446333103e-06,
+      "loss": 1.737,
+      "step": 8525
+    },
+    {
+      "epoch": 2.3997749331832887,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.305192337885144e-06,
+      "loss": 1.6273,
+      "step": 8530
+    },
+    {
+      "epoch": 2.401181600787734,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.3003564784864185e-06,
+      "loss": 1.8476,
+      "step": 8535
+    },
+    {
+      "epoch": 2.402588268392179,
+      "grad_norm": 2.75,
+      "learning_rate": 3.2955216737266854e-06,
+      "loss": 1.6457,
+      "step": 8540
+    },
+    {
+      "epoch": 2.403994935996624,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.2906879308939024e-06,
+      "loss": 1.6365,
+      "step": 8545
+    },
+    {
+      "epoch": 2.405401603601069,
+      "grad_norm": 2.75,
+      "learning_rate": 3.2858552572744306e-06,
+      "loss": 1.7155,
+      "step": 8550
+    },
+    {
+      "epoch": 2.4068082712055143,
+      "grad_norm": 4.40625,
+      "learning_rate": 3.2810236601530134e-06,
+      "loss": 1.7192,
+      "step": 8555
+    },
+    {
+      "epoch": 2.408214938809959,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.2761931468127777e-06,
+      "loss": 1.5524,
+      "step": 8560
+    },
+    {
+      "epoch": 2.4096216064144045,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.2713637245352154e-06,
+      "loss": 1.4309,
+      "step": 8565
+    },
+    {
+      "epoch": 2.4110282740188493,
+      "grad_norm": 9.5625,
+      "learning_rate": 3.2665354006001687e-06,
+      "loss": 1.5507,
+      "step": 8570
+    },
+    {
+      "epoch": 2.412434941623294,
+      "grad_norm": 4.375,
+      "learning_rate": 3.2617081822858303e-06,
+      "loss": 1.7103,
+      "step": 8575
+    },
+    {
+      "epoch": 2.4138416092277395,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.256882076868723e-06,
+      "loss": 1.2869,
+      "step": 8580
+    },
+    {
+      "epoch": 2.415248276832185,
+      "grad_norm": 4.5,
+      "learning_rate": 3.252057091623695e-06,
+      "loss": 1.554,
+      "step": 8585
+    },
+    {
+      "epoch": 2.4166549444366296,
+      "grad_norm": 4.21875,
+      "learning_rate": 3.2472332338238994e-06,
+      "loss": 1.5645,
+      "step": 8590
+    },
+    {
+      "epoch": 2.4180616120410745,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.2424105107407996e-06,
+      "loss": 1.6879,
+      "step": 8595
+    },
+    {
+      "epoch": 2.41946827964552,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.237588929644139e-06,
+      "loss": 1.507,
+      "step": 8600
+    },
+    {
+      "epoch": 2.4208749472499647,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.2327684978019464e-06,
+      "loss": 1.4155,
+      "step": 8605
+    },
+    {
+      "epoch": 2.42228161485441,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.227949222480513e-06,
+      "loss": 1.5608,
+      "step": 8610
+    },
+    {
+      "epoch": 2.423688282458855,
+      "grad_norm": 3.375,
+      "learning_rate": 3.223131110944393e-06,
+      "loss": 1.6058,
+      "step": 8615
+    },
+    {
+      "epoch": 2.4250949500633,
+      "grad_norm": 4.125,
+      "learning_rate": 3.218314170456378e-06,
+      "loss": 1.5107,
+      "step": 8620
+    },
+    {
+      "epoch": 2.426501617667745,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2134984082775036e-06,
+      "loss": 1.86,
+      "step": 8625
+    },
+    {
+      "epoch": 2.4279082852721903,
+      "grad_norm": 2.75,
+      "learning_rate": 3.2086838316670204e-06,
+      "loss": 1.5441,
+      "step": 8630
+    },
+    {
+      "epoch": 2.429314952876635,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.2038704478823983e-06,
+      "loss": 1.6657,
+      "step": 8635
+    },
+    {
+      "epoch": 2.4307216204810804,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.1990582641793078e-06,
+      "loss": 1.6059,
+      "step": 8640
+    },
+    {
+      "epoch": 2.4321282880855253,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.1942472878116066e-06,
+      "loss": 1.5391,
+      "step": 8645
+    },
+    {
+      "epoch": 2.4335349556899706,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.1894375260313384e-06,
+      "loss": 1.4821,
+      "step": 8650
+    },
+    {
+      "epoch": 2.4349416232944154,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.18462898608871e-06,
+      "loss": 1.581,
+      "step": 8655
+    },
+    {
+      "epoch": 2.4363482908988607,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.1798216752320934e-06,
+      "loss": 1.6751,
+      "step": 8660
+    },
+    {
+      "epoch": 2.4377549585033056,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.175015600707999e-06,
+      "loss": 1.4178,
+      "step": 8665
+    },
+    {
+      "epoch": 2.439161626107751,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.1702107697610825e-06,
+      "loss": 1.4115,
+      "step": 8670
+    },
+    {
+      "epoch": 2.4405682937121957,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.1654071896341184e-06,
+      "loss": 1.354,
+      "step": 8675
+    },
+    {
+      "epoch": 2.441974961316641,
+      "grad_norm": 3.0,
+      "learning_rate": 3.1606048675680002e-06,
+      "loss": 1.7495,
+      "step": 8680
+    },
+    {
+      "epoch": 2.443381628921086,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.1558038108017213e-06,
+      "loss": 1.6442,
+      "step": 8685
+    },
+    {
+      "epoch": 2.444788296525531,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.151004026572372e-06,
+      "loss": 1.5592,
+      "step": 8690
+    },
+    {
+      "epoch": 2.446194964129976,
+      "grad_norm": 2.75,
+      "learning_rate": 3.146205522115119e-06,
+      "loss": 1.7754,
+      "step": 8695
+    },
+    {
+      "epoch": 2.447601631734421,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.141408304663205e-06,
+      "loss": 1.461,
+      "step": 8700
+    },
+    {
+      "epoch": 2.449008299338866,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.1366123814479293e-06,
+      "loss": 1.3845,
+      "step": 8705
+    },
+    {
+      "epoch": 2.4504149669433115,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.1318177596986425e-06,
+      "loss": 1.4798,
+      "step": 8710
+    },
+    {
+      "epoch": 2.4518216345477564,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.127024446642732e-06,
+      "loss": 1.702,
+      "step": 8715
+    },
+    {
+      "epoch": 2.4532283021522012,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.1222324495056124e-06,
+      "loss": 1.8042,
+      "step": 8720
+    },
+    {
+      "epoch": 2.4546349697566465,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.1174417755107177e-06,
+      "loss": 1.5827,
+      "step": 8725
+    },
+    {
+      "epoch": 2.456041637361092,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.112652431879481e-06,
+      "loss": 1.7237,
+      "step": 8730
+    },
+    {
+      "epoch": 2.4574483049655367,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.1078644258313365e-06,
+      "loss": 1.5673,
+      "step": 8735
+    },
+    {
+      "epoch": 2.4588549725699815,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.1030777645836974e-06,
+      "loss": 1.5594,
+      "step": 8740
+    },
+    {
+      "epoch": 2.460261640174427,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.0982924553519548e-06,
+      "loss": 1.4208,
+      "step": 8745
+    },
+    {
+      "epoch": 2.4616683077788717,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.0935085053494557e-06,
+      "loss": 1.5434,
+      "step": 8750
+    },
+    {
+      "epoch": 2.463074975383317,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.088725921787505e-06,
+      "loss": 1.6963,
+      "step": 8755
+    },
+    {
+      "epoch": 2.464481642987762,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.0839447118753407e-06,
+      "loss": 1.5322,
+      "step": 8760
+    },
+    {
+      "epoch": 2.465888310592207,
+      "grad_norm": 4.75,
+      "learning_rate": 3.0791648828201354e-06,
+      "loss": 1.6891,
+      "step": 8765
+    },
+    {
+      "epoch": 2.467294978196652,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.0743864418269777e-06,
+      "loss": 1.652,
+      "step": 8770
+    },
+    {
+      "epoch": 2.4687016458010973,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.069609396098865e-06,
+      "loss": 1.5267,
+      "step": 8775
+    },
+    {
+      "epoch": 2.470108313405542,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.064833752836692e-06,
+      "loss": 1.4356,
+      "step": 8780
+    },
+    {
+      "epoch": 2.4715149810099875,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.0600595192392364e-06,
+      "loss": 1.6106,
+      "step": 8785
+    },
+    {
+      "epoch": 2.4729216486144323,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.055286702503156e-06,
+      "loss": 1.4059,
+      "step": 8790
+    },
+    {
+      "epoch": 2.4743283162188776,
+      "grad_norm": 3.875,
+      "learning_rate": 3.050515309822966e-06,
+      "loss": 1.5044,
+      "step": 8795
+    },
+    {
+      "epoch": 2.4757349838233225,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.0457453483910417e-06,
+      "loss": 1.5502,
+      "step": 8800
+    },
+    {
+      "epoch": 2.4771416514277678,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.0409768253975967e-06,
+      "loss": 1.6319,
+      "step": 8805
+    },
+    {
+      "epoch": 2.4785483190322126,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.0362097480306787e-06,
+      "loss": 1.4717,
+      "step": 8810
+    },
+    {
+      "epoch": 2.479954986636658,
+      "grad_norm": 3.0,
+      "learning_rate": 3.031444123476154e-06,
+      "loss": 1.702,
+      "step": 8815
+    },
+    {
+      "epoch": 2.4813616542411028,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.0266799589177023e-06,
+      "loss": 1.7863,
+      "step": 8820
+    },
+    {
+      "epoch": 2.482768321845548,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.021917261536797e-06,
+      "loss": 1.1797,
+      "step": 8825
+    },
+    {
+      "epoch": 2.484174989449993,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.0171560385127066e-06,
+      "loss": 1.608,
+      "step": 8830
+    },
+    {
+      "epoch": 2.4855816570544382,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.012396297022471e-06,
+      "loss": 1.2161,
+      "step": 8835
+    },
+    {
+      "epoch": 2.486988324658883,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.0076380442409023e-06,
+      "loss": 1.6565,
+      "step": 8840
+    },
+    {
+      "epoch": 2.488394992263328,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0028812873405636e-06,
+      "loss": 1.745,
+      "step": 8845
+    },
+    {
+      "epoch": 2.4898016598677732,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9981260334917666e-06,
+      "loss": 1.6035,
+      "step": 8850
+    },
+    {
+      "epoch": 2.4912083274722185,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9933722898625575e-06,
+      "loss": 1.558,
+      "step": 8855
+    },
+    {
+      "epoch": 2.4926149950766634,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.988620063618701e-06,
+      "loss": 1.6606,
+      "step": 8860
+    },
+    {
+      "epoch": 2.4940216626811083,
+      "grad_norm": 4.25,
+      "learning_rate": 2.9838693619236823e-06,
+      "loss": 1.5781,
+      "step": 8865
+    },
+    {
+      "epoch": 2.4954283302855536,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9791201919386807e-06,
+      "loss": 1.6853,
+      "step": 8870
+    },
+    {
+      "epoch": 2.4968349978899984,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.974372560822573e-06,
+      "loss": 1.5243,
+      "step": 8875
+    },
+    {
+      "epoch": 2.4982416654944437,
+      "grad_norm": 4.8125,
+      "learning_rate": 2.9696264757319113e-06,
+      "loss": 1.5861,
+      "step": 8880
+    },
+    {
+      "epoch": 2.4996483330988886,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.9648819438209228e-06,
+      "loss": 1.6032,
+      "step": 8885
+    },
+    {
+      "epoch": 2.501055000703334,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.960138972241485e-06,
+      "loss": 1.6981,
+      "step": 8890
+    },
+    {
+      "epoch": 2.5024616683077787,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.955397568143134e-06,
+      "loss": 1.5882,
+      "step": 8895
+    },
+    {
+      "epoch": 2.503868335912224,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.950657738673033e-06,
+      "loss": 1.5093,
+      "step": 8900
+    },
+    {
+      "epoch": 2.505275003516669,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.945919490975979e-06,
+      "loss": 1.7241,
+      "step": 8905
+    },
+    {
+      "epoch": 2.506681671121114,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9411828321943804e-06,
+      "loss": 1.546,
+      "step": 8910
+    },
+    {
+      "epoch": 2.508088338725559,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9364477694682546e-06,
+      "loss": 1.6981,
+      "step": 8915
+    },
+    {
+      "epoch": 2.5094950063300043,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9317143099352056e-06,
+      "loss": 1.5693,
+      "step": 8920
+    },
+    {
+      "epoch": 2.510901673934449,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.926982460730429e-06,
+      "loss": 1.436,
+      "step": 8925
+    },
+    {
+      "epoch": 2.5123083415388945,
+      "grad_norm": 3.625,
+      "learning_rate": 2.922252228986691e-06,
+      "loss": 1.7639,
+      "step": 8930
+    },
+    {
+      "epoch": 2.5137150091433393,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.917523621834314e-06,
+      "loss": 1.568,
+      "step": 8935
+    },
+    {
+      "epoch": 2.5151216767477846,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9127966464011787e-06,
+      "loss": 1.7766,
+      "step": 8940
+    },
+    {
+      "epoch": 2.5165283443522295,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.908071309812702e-06,
+      "loss": 1.4208,
+      "step": 8945
+    },
+    {
+      "epoch": 2.517935011956675,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9033476191918338e-06,
+      "loss": 1.7117,
+      "step": 8950
+    },
+    {
+      "epoch": 2.5193416795611197,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8986255816590365e-06,
+      "loss": 1.8485,
+      "step": 8955
+    },
+    {
+      "epoch": 2.520748347165565,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8939052043322895e-06,
+      "loss": 1.6935,
+      "step": 8960
+    },
+    {
+      "epoch": 2.52215501477001,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.8891864943270603e-06,
+      "loss": 1.6999,
+      "step": 8965
+    },
+    {
+      "epoch": 2.5235616823744547,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.884469458756312e-06,
+      "loss": 1.4383,
+      "step": 8970
+    },
+    {
+      "epoch": 2.5249683499789,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8797541047304764e-06,
+      "loss": 1.5988,
+      "step": 8975
+    },
+    {
+      "epoch": 2.5263750175833453,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.875040439357456e-06,
+      "loss": 1.6122,
+      "step": 8980
+    },
+    {
+      "epoch": 2.52778168518779,
+      "grad_norm": 4.5625,
+      "learning_rate": 2.8703284697426015e-06,
+      "loss": 1.5492,
+      "step": 8985
+    },
+    {
+      "epoch": 2.529188352792235,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.8656182029887148e-06,
+      "loss": 1.2839,
+      "step": 8990
+    },
+    {
+      "epoch": 2.5305950203966803,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8609096461960276e-06,
+      "loss": 1.8007,
+      "step": 8995
+    },
+    {
+      "epoch": 2.5320016880011256,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.8562028064621917e-06,
+      "loss": 1.4047,
+      "step": 9000
+    },
+    {
+      "epoch": 2.5334083556055704,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.851497690882274e-06,
+      "loss": 1.5034,
+      "step": 9005
+    },
+    {
+      "epoch": 2.5348150232100153,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.84679430654874e-06,
+      "loss": 1.6142,
+      "step": 9010
+    },
+    {
+      "epoch": 2.5362216908144606,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.842092660551448e-06,
+      "loss": 1.3421,
+      "step": 9015
+    },
+    {
+      "epoch": 2.537628358418906,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.837392759977634e-06,
+      "loss": 1.6882,
+      "step": 9020
+    },
+    {
+      "epoch": 2.5390350260233507,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.832694611911905e-06,
+      "loss": 1.4796,
+      "step": 9025
+    },
+    {
+      "epoch": 2.5404416936277956,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.8279982234362223e-06,
+      "loss": 1.4512,
+      "step": 9030
+    },
+    {
+      "epoch": 2.541848361232241,
+      "grad_norm": 5.0,
+      "learning_rate": 2.8233036016299e-06,
+      "loss": 1.6511,
+      "step": 9035
+    },
+    {
+      "epoch": 2.5432550288366857,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.818610753569583e-06,
+      "loss": 1.5149,
+      "step": 9040
+    },
+    {
+      "epoch": 2.544661696441131,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8139196863292497e-06,
+      "loss": 1.4205,
+      "step": 9045
+    },
+    {
+      "epoch": 2.546068364045576,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.8092304069801875e-06,
+      "loss": 1.4803,
+      "step": 9050
+    },
+    {
+      "epoch": 2.547475031650021,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.8045429225909953e-06,
+      "loss": 1.5202,
+      "step": 9055
+    },
+    {
+      "epoch": 2.548881699254466,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.799857240227558e-06,
+      "loss": 1.5887,
+      "step": 9060
+    },
+    {
+      "epoch": 2.5502883668589114,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.795173366953051e-06,
+      "loss": 1.6058,
+      "step": 9065
+    },
+    {
+      "epoch": 2.551695034463356,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.7904913098279213e-06,
+      "loss": 1.6265,
+      "step": 9070
+    },
+    {
+      "epoch": 2.5531017020678015,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.7858110759098753e-06,
+      "loss": 1.4793,
+      "step": 9075
+    },
+    {
+      "epoch": 2.5545083696722464,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.7811326722538755e-06,
+      "loss": 1.5202,
+      "step": 9080
+    },
+    {
+      "epoch": 2.5559150372766917,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.776456105912121e-06,
+      "loss": 1.6495,
+      "step": 9085
+    },
+    {
+      "epoch": 2.5573217048811365,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.771781383934046e-06,
+      "loss": 1.6192,
+      "step": 9090
+    },
+    {
+      "epoch": 2.5587283724855814,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.767108513366299e-06,
+      "loss": 1.4899,
+      "step": 9095
+    },
+    {
+      "epoch": 2.5601350400900267,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7624375012527423e-06,
+      "loss": 1.6238,
+      "step": 9100
+    },
+    {
+      "epoch": 2.561541707694472,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.757768354634435e-06,
+      "loss": 1.142,
+      "step": 9105
+    },
+    {
+      "epoch": 2.562948375298917,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.7531010805496245e-06,
+      "loss": 1.5075,
+      "step": 9110
+    },
+    {
+      "epoch": 2.5643550429033617,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.748435686033735e-06,
+      "loss": 1.5947,
+      "step": 9115
+    },
+    {
+      "epoch": 2.565761710507807,
+      "grad_norm": 4.71875,
+      "learning_rate": 2.7437721781193596e-06,
+      "loss": 1.5573,
+      "step": 9120
+    },
+    {
+      "epoch": 2.5671683781122523,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7391105638362422e-06,
+      "loss": 1.4712,
+      "step": 9125
+    },
+    {
+      "epoch": 2.568575045716697,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.734450850211278e-06,
+      "loss": 1.5379,
+      "step": 9130
+    },
+    {
+      "epoch": 2.569981713321142,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7297930442684958e-06,
+      "loss": 1.2337,
+      "step": 9135
+    },
+    {
+      "epoch": 2.5713883809255873,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7251371530290464e-06,
+      "loss": 1.5338,
+      "step": 9140
+    },
+    {
+      "epoch": 2.5727950485300326,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.720483183511197e-06,
+      "loss": 1.3647,
+      "step": 9145
+    },
+    {
+      "epoch": 2.5742017161344775,
+      "grad_norm": 3.625,
+      "learning_rate": 2.715831142730316e-06,
+      "loss": 1.4631,
+      "step": 9150
+    },
+    {
+      "epoch": 2.5756083837389223,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.711181037698867e-06,
+      "loss": 1.4525,
+      "step": 9155
+    },
+    {
+      "epoch": 2.5770150513433676,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.706532875426392e-06,
+      "loss": 1.6094,
+      "step": 9160
+    },
+    {
+      "epoch": 2.5784217189478125,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.7018866629195077e-06,
+      "loss": 1.4726,
+      "step": 9165
+    },
+    {
+      "epoch": 2.5798283865522578,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.69724240718189e-06,
+      "loss": 1.507,
+      "step": 9170
+    },
+    {
+      "epoch": 2.5812350541567026,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.692600115214267e-06,
+      "loss": 1.586,
+      "step": 9175
+    },
+    {
+      "epoch": 2.582641721761148,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.6879597940144038e-06,
+      "loss": 1.5724,
+      "step": 9180
+    },
+    {
+      "epoch": 2.584048389365593,
+      "grad_norm": 4.5625,
+      "learning_rate": 2.683321450577098e-06,
+      "loss": 1.6215,
+      "step": 9185
+    },
+    {
+      "epoch": 2.585455056970038,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.678685091894162e-06,
+      "loss": 1.6285,
+      "step": 9190
+    },
+    {
+      "epoch": 2.586861724574483,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.674050724954421e-06,
+      "loss": 1.4316,
+      "step": 9195
+    },
+    {
+      "epoch": 2.5882683921789282,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.6694183567436936e-06,
+      "loss": 1.4817,
+      "step": 9200
+    },
+    {
+      "epoch": 2.589675059783373,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.664787994244788e-06,
+      "loss": 1.5046,
+      "step": 9205
+    },
+    {
+      "epoch": 2.5910817273878184,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.66015964443749e-06,
+      "loss": 1.4744,
+      "step": 9210
+    },
+    {
+      "epoch": 2.5924883949922632,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.655533314298548e-06,
+      "loss": 1.6461,
+      "step": 9215
+    },
+    {
+      "epoch": 2.5938950625967085,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6509090108016707e-06,
+      "loss": 1.6825,
+      "step": 9220
+    },
+    {
+      "epoch": 2.5953017302011534,
+      "grad_norm": 4.3125,
+      "learning_rate": 2.646286740917504e-06,
+      "loss": 1.5424,
+      "step": 9225
+    },
+    {
+      "epoch": 2.5967083978055987,
+      "grad_norm": 10.125,
+      "learning_rate": 2.641666511613639e-06,
+      "loss": 1.5151,
+      "step": 9230
+    },
+    {
+      "epoch": 2.5981150654100436,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.637048329854581e-06,
+      "loss": 1.6234,
+      "step": 9235
+    },
+    {
+      "epoch": 2.5995217330144884,
+      "grad_norm": 3.0,
+      "learning_rate": 2.632432202601755e-06,
+      "loss": 1.7191,
+      "step": 9240
+    },
+    {
+      "epoch": 2.6009284006189337,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.6278181368134873e-06,
+      "loss": 1.7231,
+      "step": 9245
+    },
+    {
+      "epoch": 2.602335068223379,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.623206139444997e-06,
+      "loss": 1.6375,
+      "step": 9250
+    },
+    {
+      "epoch": 2.603741735827824,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.6185962174483815e-06,
+      "loss": 1.4945,
+      "step": 9255
+    },
+    {
+      "epoch": 2.6051484034322687,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.6139883777726185e-06,
+      "loss": 1.7805,
+      "step": 9260
+    },
+    {
+      "epoch": 2.606555071036714,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.609382627363536e-06,
+      "loss": 1.6192,
+      "step": 9265
+    },
+    {
+      "epoch": 2.6079617386411593,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.6047789731638224e-06,
+      "loss": 1.4136,
+      "step": 9270
+    },
+    {
+      "epoch": 2.609368406245604,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.600177422112999e-06,
+      "loss": 1.5048,
+      "step": 9275
+    },
+    {
+      "epoch": 2.610775073850049,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.5955779811474213e-06,
+      "loss": 1.6703,
+      "step": 9280
+    },
+    {
+      "epoch": 2.6121817414544943,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.5909806572002634e-06,
+      "loss": 1.6786,
+      "step": 9285
+    },
+    {
+      "epoch": 2.6135884090589396,
+      "grad_norm": 4.875,
+      "learning_rate": 2.5863854572015057e-06,
+      "loss": 1.6156,
+      "step": 9290
+    },
+    {
+      "epoch": 2.6149950766633845,
+      "grad_norm": 3.0,
+      "learning_rate": 2.5817923880779308e-06,
+      "loss": 1.6928,
+      "step": 9295
+    },
+    {
+      "epoch": 2.6164017442678293,
+      "grad_norm": 2.875,
+      "learning_rate": 2.577201456753104e-06,
+      "loss": 1.6488,
+      "step": 9300
+    },
+    {
+      "epoch": 2.6178084118722746,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.572612670147374e-06,
+      "loss": 1.6353,
+      "step": 9305
+    },
+    {
+      "epoch": 2.6192150794767195,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5680260351778523e-06,
+      "loss": 1.4599,
+      "step": 9310
+    },
+    {
+      "epoch": 2.620621747081165,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.56344155875841e-06,
+      "loss": 1.4838,
+      "step": 9315
+    },
+    {
+      "epoch": 2.6220284146856097,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.55885924779966e-06,
+      "loss": 1.5812,
+      "step": 9320
+    },
+    {
+      "epoch": 2.623435082290055,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5542791092089586e-06,
+      "loss": 1.4303,
+      "step": 9325
+    },
+    {
+      "epoch": 2.6248417498945,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.549701149890377e-06,
+      "loss": 1.6144,
+      "step": 9330
+    },
+    {
+      "epoch": 2.626248417498945,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.545125376744712e-06,
+      "loss": 1.588,
+      "step": 9335
+    },
+    {
+      "epoch": 2.62765508510339,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.540551796669457e-06,
+      "loss": 1.6687,
+      "step": 9340
+    },
+    {
+      "epoch": 2.6290617527078353,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.535980416558804e-06,
+      "loss": 1.5527,
+      "step": 9345
+    },
+    {
+      "epoch": 2.63046842031228,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.531411243303629e-06,
+      "loss": 1.429,
+      "step": 9350
+    },
+    {
+      "epoch": 2.6318750879167254,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.526844283791477e-06,
+      "loss": 1.4638,
+      "step": 9355
+    },
+    {
+      "epoch": 2.6332817555211703,
+      "grad_norm": 4.0,
+      "learning_rate": 2.5222795449065623e-06,
+      "loss": 1.8405,
+      "step": 9360
+    },
+    {
+      "epoch": 2.6346884231256156,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.5177170335297445e-06,
+      "loss": 1.6266,
+      "step": 9365
+    },
+    {
+      "epoch": 2.6360950907300604,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.5131567565385327e-06,
+      "loss": 1.5076,
+      "step": 9370
+    },
+    {
+      "epoch": 2.6375017583345057,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.5085987208070628e-06,
+      "loss": 1.6445,
+      "step": 9375
+    },
+    {
+      "epoch": 2.6389084259389506,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.5040429332060953e-06,
+      "loss": 1.6817,
+      "step": 9380
+    },
+    {
+      "epoch": 2.6403150935433954,
+      "grad_norm": 3.375,
+      "learning_rate": 2.499489400602999e-06,
+      "loss": 1.7791,
+      "step": 9385
+    },
+    {
+      "epoch": 2.6417217611478407,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4949381298617478e-06,
+      "loss": 1.5416,
+      "step": 9390
+    },
+    {
+      "epoch": 2.643128428752286,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.4903891278429002e-06,
+      "loss": 1.7715,
+      "step": 9395
+    },
+    {
+      "epoch": 2.644535096356731,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.485842401403601e-06,
+      "loss": 1.6767,
+      "step": 9400
+    },
+    {
+      "epoch": 2.6459417639611758,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.4812979573975595e-06,
+      "loss": 1.7507,
+      "step": 9405
+    },
+    {
+      "epoch": 2.647348431565621,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.476755802675049e-06,
+      "loss": 1.608,
+      "step": 9410
+    },
+    {
+      "epoch": 2.6487550991700664,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4722159440828877e-06,
+      "loss": 1.5923,
+      "step": 9415
+    },
+    {
+      "epoch": 2.650161766774511,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.467678388464436e-06,
+      "loss": 1.4541,
+      "step": 9420
+    },
+    {
+      "epoch": 2.651568434378956,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.4631431426595826e-06,
+      "loss": 1.6113,
+      "step": 9425
+    },
+    {
+      "epoch": 2.6529751019834014,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.4586102135047314e-06,
+      "loss": 1.6558,
+      "step": 9430
+    },
+    {
+      "epoch": 2.6543817695878467,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.4540796078327966e-06,
+      "loss": 1.7034,
+      "step": 9435
+    },
+    {
+      "epoch": 2.6557884371922915,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.4495513324731897e-06,
+      "loss": 1.68,
+      "step": 9440
+    },
+    {
+      "epoch": 2.6571951047967364,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.4450253942518105e-06,
+      "loss": 1.675,
+      "step": 9445
+    },
+    {
+      "epoch": 2.6586017724011817,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.4405017999910324e-06,
+      "loss": 1.6803,
+      "step": 9450
+    },
+    {
+      "epoch": 2.6600084400056265,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.4359805565097006e-06,
+      "loss": 1.5876,
+      "step": 9455
+    },
+    {
+      "epoch": 2.661415107610072,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.431461670623111e-06,
+      "loss": 1.7244,
+      "step": 9460
+    },
+    {
+      "epoch": 2.6628217752145167,
+      "grad_norm": 3.0,
+      "learning_rate": 2.4269451491430103e-06,
+      "loss": 1.7627,
+      "step": 9465
+    },
+    {
+      "epoch": 2.664228442818962,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.422430998877578e-06,
+      "loss": 1.4034,
+      "step": 9470
+    },
+    {
+      "epoch": 2.665635110423407,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.417919226631423e-06,
+      "loss": 1.5969,
+      "step": 9475
+    },
+    {
+      "epoch": 2.667041778027852,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.413409839205565e-06,
+      "loss": 1.4532,
+      "step": 9480
+    },
+    {
+      "epoch": 2.668448445632297,
+      "grad_norm": 2.5,
+      "learning_rate": 2.4089028433974335e-06,
+      "loss": 1.346,
+      "step": 9485
+    },
+    {
+      "epoch": 2.6698551132367423,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.4043982460008466e-06,
+      "loss": 1.3301,
+      "step": 9490
+    },
+    {
+      "epoch": 2.671261780841187,
+      "grad_norm": 3.875,
+      "learning_rate": 2.3998960538060138e-06,
+      "loss": 1.5903,
+      "step": 9495
+    },
+    {
+      "epoch": 2.6726684484456324,
+      "grad_norm": 4.65625,
+      "learning_rate": 2.3953962735995167e-06,
+      "loss": 1.4014,
+      "step": 9500
+    },
+    {
+      "epoch": 2.6740751160500773,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.390898912164298e-06,
+      "loss": 1.8069,
+      "step": 9505
+    },
+    {
+      "epoch": 2.675481783654522,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.3864039762796583e-06,
+      "loss": 1.5562,
+      "step": 9510
+    },
+    {
+      "epoch": 2.6768884512589675,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.38191147272124e-06,
+      "loss": 1.7332,
+      "step": 9515
+    },
+    {
+      "epoch": 2.6782951188634128,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.3774214082610217e-06,
+      "loss": 1.5451,
+      "step": 9520
+    },
+    {
+      "epoch": 2.6797017864678576,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.3729337896672996e-06,
+      "loss": 1.3584,
+      "step": 9525
+    },
+    {
+      "epoch": 2.6811084540723025,
+      "grad_norm": 4.8125,
+      "learning_rate": 2.3684486237046886e-06,
+      "loss": 1.5757,
+      "step": 9530
+    },
+    {
+      "epoch": 2.6825151216767478,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3639659171341036e-06,
+      "loss": 1.617,
+      "step": 9535
+    },
+    {
+      "epoch": 2.683921789281193,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3594856767127542e-06,
+      "loss": 1.6402,
+      "step": 9540
+    },
+    {
+      "epoch": 2.685328456885638,
+      "grad_norm": 3.375,
+      "learning_rate": 2.35500790919413e-06,
+      "loss": 1.7177,
+      "step": 9545
+    },
+    {
+      "epoch": 2.686735124490083,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.3505326213279964e-06,
+      "loss": 1.5701,
+      "step": 9550
+    },
+    {
+      "epoch": 2.688141792094528,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.346059819860376e-06,
+      "loss": 1.5764,
+      "step": 9555
+    },
+    {
+      "epoch": 2.6895484596989734,
+      "grad_norm": 2.5,
+      "learning_rate": 2.3415895115335477e-06,
+      "loss": 1.7206,
+      "step": 9560
+    },
+    {
+      "epoch": 2.6909551273034182,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.3371217030860337e-06,
+      "loss": 1.7077,
+      "step": 9565
+    },
+    {
+      "epoch": 2.692361794907863,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.3326564012525804e-06,
+      "loss": 1.6132,
+      "step": 9570
+    },
+    {
+      "epoch": 2.6937684625123084,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.3281936127641644e-06,
+      "loss": 1.5924,
+      "step": 9575
+    },
+    {
+      "epoch": 2.6951751301167532,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3237333443479676e-06,
+      "loss": 1.3612,
+      "step": 9580
+    },
+    {
+      "epoch": 2.6965817977211985,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3192756027273766e-06,
+      "loss": 1.6446,
+      "step": 9585
+    },
+    {
+      "epoch": 2.6979884653256434,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.3148203946219644e-06,
+      "loss": 1.9269,
+      "step": 9590
+    },
+    {
+      "epoch": 2.6993951329300887,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3103677267474934e-06,
+      "loss": 1.5727,
+      "step": 9595
+    },
+    {
+      "epoch": 2.7008018005345336,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.3059176058158897e-06,
+      "loss": 1.6316,
+      "step": 9600
+    },
+    {
+      "epoch": 2.702208468138979,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.3014700385352425e-06,
+      "loss": 1.2899,
+      "step": 9605
+    },
+    {
+      "epoch": 2.7036151357434237,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.2970250316097914e-06,
+      "loss": 1.5956,
+      "step": 9610
+    },
+    {
+      "epoch": 2.705021803347869,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.292582591739916e-06,
+      "loss": 1.4859,
+      "step": 9615
+    },
+    {
+      "epoch": 2.706428470952314,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.2881427256221263e-06,
+      "loss": 1.5503,
+      "step": 9620
+    },
+    {
+      "epoch": 2.707835138556759,
+      "grad_norm": 2.5,
+      "learning_rate": 2.283705439949056e-06,
+      "loss": 1.4448,
+      "step": 9625
+    },
+    {
+      "epoch": 2.709241806161204,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.2792707414094447e-06,
+      "loss": 1.5078,
+      "step": 9630
+    },
+    {
+      "epoch": 2.7106484737656493,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.2748386366881327e-06,
+      "loss": 1.5553,
+      "step": 9635
+    },
+    {
+      "epoch": 2.712055141370094,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.2704091324660557e-06,
+      "loss": 1.6245,
+      "step": 9640
+    },
+    {
+      "epoch": 2.7134618089745395,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.26598223542022e-06,
+      "loss": 1.4752,
+      "step": 9645
+    },
+    {
+      "epoch": 2.7148684765789843,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.2615579522237103e-06,
+      "loss": 1.6861,
+      "step": 9650
+    },
+    {
+      "epoch": 2.716275144183429,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.2571362895456673e-06,
+      "loss": 1.3737,
+      "step": 9655
+    },
+    {
+      "epoch": 2.7176818117878745,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.2527172540512817e-06,
+      "loss": 1.7256,
+      "step": 9660
+    },
+    {
+      "epoch": 2.71908847939232,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.248300852401784e-06,
+      "loss": 1.4467,
+      "step": 9665
+    },
+    {
+      "epoch": 2.7204951469967646,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.2438870912544386e-06,
+      "loss": 1.7083,
+      "step": 9670
+    },
+    {
+      "epoch": 2.7219018146012095,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.239475977262521e-06,
+      "loss": 1.6565,
+      "step": 9675
+    },
+    {
+      "epoch": 2.723308482205655,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.2350675170753247e-06,
+      "loss": 1.4021,
+      "step": 9680
+    },
+    {
+      "epoch": 2.7247151498101,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.230661717338138e-06,
+      "loss": 1.5423,
+      "step": 9685
+    },
+    {
+      "epoch": 2.726121817414545,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.2262585846922418e-06,
+      "loss": 1.4289,
+      "step": 9690
+    },
+    {
+      "epoch": 2.72752848501899,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.2218581257748927e-06,
+      "loss": 1.4817,
+      "step": 9695
+    },
+    {
+      "epoch": 2.728935152623435,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.2174603472193224e-06,
+      "loss": 1.3842,
+      "step": 9700
+    },
+    {
+      "epoch": 2.7303418202278804,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.213065255654719e-06,
+      "loss": 1.5652,
+      "step": 9705
+    },
+    {
+      "epoch": 2.7317484878323253,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.2086728577062178e-06,
+      "loss": 1.3789,
+      "step": 9710
+    },
+    {
+      "epoch": 2.73315515543677,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.204283159994902e-06,
+      "loss": 1.3959,
+      "step": 9715
+    },
+    {
+      "epoch": 2.7345618230412154,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.199896169137772e-06,
+      "loss": 1.7312,
+      "step": 9720
+    },
+    {
+      "epoch": 2.7359684906456603,
+      "grad_norm": 4.5,
+      "learning_rate": 2.19551189174776e-06,
+      "loss": 1.4778,
+      "step": 9725
+    },
+    {
+      "epoch": 2.7373751582501056,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.1911303344337014e-06,
+      "loss": 1.4665,
+      "step": 9730
+    },
+    {
+      "epoch": 2.7387818258545504,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.186751503800332e-06,
+      "loss": 1.5714,
+      "step": 9735
+    },
+    {
+      "epoch": 2.7401884934589957,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.1823754064482786e-06,
+      "loss": 1.5682,
+      "step": 9740
+    },
+    {
+      "epoch": 2.7415951610634406,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.1780020489740506e-06,
+      "loss": 1.4708,
+      "step": 9745
+    },
+    {
+      "epoch": 2.743001828667886,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.1736314379700177e-06,
+      "loss": 1.5423,
+      "step": 9750
+    },
+    {
+      "epoch": 2.7444084962723307,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.1692635800244222e-06,
+      "loss": 1.7049,
+      "step": 9755
+    },
+    {
+      "epoch": 2.745815163876776,
+      "grad_norm": 4.625,
+      "learning_rate": 2.164898481721348e-06,
+      "loss": 1.4801,
+      "step": 9760
+    },
+    {
+      "epoch": 2.747221831481221,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.160536149640721e-06,
+      "loss": 1.6165,
+      "step": 9765
+    },
+    {
+      "epoch": 2.748628499085666,
+      "grad_norm": 4.90625,
+      "learning_rate": 2.1561765903582985e-06,
+      "loss": 1.4022,
+      "step": 9770
+    },
+    {
+      "epoch": 2.750035166690111,
+      "grad_norm": 6.375,
+      "learning_rate": 2.151819810445656e-06,
+      "loss": 1.3853,
+      "step": 9775
+    },
+    {
+      "epoch": 2.7514418342945564,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.147465816470183e-06,
+      "loss": 1.5349,
+      "step": 9780
+    },
+    {
+      "epoch": 2.752848501899001,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.1431146149950673e-06,
+      "loss": 1.575,
+      "step": 9785
+    },
+    {
+      "epoch": 2.7542551695034465,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.138766212579286e-06,
+      "loss": 1.4022,
+      "step": 9790
+    },
+    {
+      "epoch": 2.7556618371078914,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1344206157775963e-06,
+      "loss": 1.5342,
+      "step": 9795
+    },
+    {
+      "epoch": 2.757068504712336,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.130077831140534e-06,
+      "loss": 1.6823,
+      "step": 9800
+    },
+    {
+      "epoch": 2.7584751723167815,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.125737865214383e-06,
+      "loss": 1.5458,
+      "step": 9805
+    },
+    {
+      "epoch": 2.759881839921227,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.12140072454119e-06,
+      "loss": 1.5436,
+      "step": 9810
+    },
+    {
+      "epoch": 2.7612885075256717,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.1170664156587374e-06,
+      "loss": 1.4426,
+      "step": 9815
+    },
+    {
+      "epoch": 2.7626951751301165,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.1127349451005387e-06,
+      "loss": 1.6613,
+      "step": 9820
+    },
+    {
+      "epoch": 2.764101842734562,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.1084063193958292e-06,
+      "loss": 1.522,
+      "step": 9825
+    },
+    {
+      "epoch": 2.765508510339007,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.104080545069561e-06,
+      "loss": 1.6962,
+      "step": 9830
+    },
+    {
+      "epoch": 2.766915177943452,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.0997576286423773e-06,
+      "loss": 1.5804,
+      "step": 9835
+    },
+    {
+      "epoch": 2.768321845547897,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0954375766306256e-06,
+      "loss": 1.5818,
+      "step": 9840
+    },
+    {
+      "epoch": 2.769728513152342,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0911203955463262e-06,
+      "loss": 1.7594,
+      "step": 9845
+    },
+    {
+      "epoch": 2.7711351807567874,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0868060918971754e-06,
+      "loss": 1.541,
+      "step": 9850
+    },
+    {
+      "epoch": 2.7725418483612323,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.082494672186535e-06,
+      "loss": 1.6876,
+      "step": 9855
+    },
+    {
+      "epoch": 2.773948515965677,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.078186142913414e-06,
+      "loss": 1.4451,
+      "step": 9860
+    },
+    {
+      "epoch": 2.7753551835701225,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.0738805105724676e-06,
+      "loss": 1.8509,
+      "step": 9865
+    },
+    {
+      "epoch": 2.7767618511745673,
+      "grad_norm": 5.75,
+      "learning_rate": 2.069577781653982e-06,
+      "loss": 1.648,
+      "step": 9870
+    },
+    {
+      "epoch": 2.7781685187790126,
+      "grad_norm": 5.15625,
+      "learning_rate": 2.065277962643873e-06,
+      "loss": 1.5538,
+      "step": 9875
+    },
+    {
+      "epoch": 2.7795751863834575,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0609810600236586e-06,
+      "loss": 1.4278,
+      "step": 9880
+    },
+    {
+      "epoch": 2.7809818539879028,
+      "grad_norm": 3.75,
+      "learning_rate": 2.056687080270473e-06,
+      "loss": 1.5183,
+      "step": 9885
+    },
+    {
+      "epoch": 2.7823885215923476,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.0523960298570368e-06,
+      "loss": 1.4949,
+      "step": 9890
+    },
+    {
+      "epoch": 2.783795189196793,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.0481079152516564e-06,
+      "loss": 1.5487,
+      "step": 9895
+    },
+    {
+      "epoch": 2.7852018568012378,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.043822742918212e-06,
+      "loss": 1.6569,
+      "step": 9900
+    },
+    {
+      "epoch": 2.786608524405683,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.0395405193161557e-06,
+      "loss": 1.5651,
+      "step": 9905
+    },
+    {
+      "epoch": 2.788015192010128,
+      "grad_norm": 3.625,
+      "learning_rate": 2.0352612509004816e-06,
+      "loss": 1.7325,
+      "step": 9910
+    },
+    {
+      "epoch": 2.7894218596145732,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.030984944121742e-06,
+      "loss": 1.6342,
+      "step": 9915
+    },
+    {
+      "epoch": 2.790828527219018,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.0267116054260174e-06,
+      "loss": 1.6308,
+      "step": 9920
+    },
+    {
+      "epoch": 2.792235194823463,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.0224412412549153e-06,
+      "loss": 1.5591,
+      "step": 9925
+    },
+    {
+      "epoch": 2.7936418624279082,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.0181738580455626e-06,
+      "loss": 1.6227,
+      "step": 9930
+    },
+    {
+      "epoch": 2.7950485300323535,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.013909462230589e-06,
+      "loss": 1.6837,
+      "step": 9935
+    },
+    {
+      "epoch": 2.7964551976367984,
+      "grad_norm": 5.125,
+      "learning_rate": 2.009648060238123e-06,
+      "loss": 1.5786,
+      "step": 9940
+    },
+    {
+      "epoch": 2.7978618652412433,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0053896584917804e-06,
+      "loss": 1.5091,
+      "step": 9945
+    },
+    {
+      "epoch": 2.7992685328456886,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.001134263410652e-06,
+      "loss": 1.6791,
+      "step": 9950
+    },
+    {
+      "epoch": 2.800675200450134,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.9968818814092975e-06,
+      "loss": 1.6616,
+      "step": 9955
+    },
+    {
+      "epoch": 2.8020818680545787,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9926325188977382e-06,
+      "loss": 1.5081,
+      "step": 9960
+    },
+    {
+      "epoch": 2.8034885356590236,
+      "grad_norm": 3.625,
+      "learning_rate": 1.98838618228144e-06,
+      "loss": 1.5836,
+      "step": 9965
+    },
+    {
+      "epoch": 2.804895203263469,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.9841428779613085e-06,
+      "loss": 1.8027,
+      "step": 9970
+    },
+    {
+      "epoch": 2.806301870867914,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.979902612333678e-06,
+      "loss": 1.6843,
+      "step": 9975
+    },
+    {
+      "epoch": 2.807708538472359,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.9756653917903024e-06,
+      "loss": 1.6057,
+      "step": 9980
+    },
+    {
+      "epoch": 2.809115206076804,
+      "grad_norm": 3.625,
+      "learning_rate": 1.9714312227183448e-06,
+      "loss": 1.4281,
+      "step": 9985
+    },
+    {
+      "epoch": 2.810521873681249,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.9672001115003734e-06,
+      "loss": 1.5506,
+      "step": 9990
+    },
+    {
+      "epoch": 2.811928541285694,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.96297206451434e-06,
+      "loss": 1.4472,
+      "step": 9995
+    },
+    {
+      "epoch": 2.8133352088901393,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.95874708813358e-06,
+      "loss": 1.7408,
+      "step": 10000
+    },
+    {
+      "epoch": 2.814741876494584,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.9545251887268055e-06,
+      "loss": 1.8876,
+      "step": 10005
+    },
+    {
+      "epoch": 2.8161485440990295,
+      "grad_norm": 3.125,
+      "learning_rate": 1.9503063726580794e-06,
+      "loss": 1.6134,
+      "step": 10010
+    },
+    {
+      "epoch": 2.8175552117034743,
+      "grad_norm": 4.125,
+      "learning_rate": 1.9460906462868266e-06,
+      "loss": 1.3413,
+      "step": 10015
+    },
+    {
+      "epoch": 2.8189618793079196,
+      "grad_norm": 3.875,
+      "learning_rate": 1.941878015967811e-06,
+      "loss": 1.5506,
+      "step": 10020
+    },
+    {
+      "epoch": 2.8203685469123645,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9376684880511283e-06,
+      "loss": 1.7976,
+      "step": 10025
+    },
+    {
+      "epoch": 2.82177521451681,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9334620688821986e-06,
+      "loss": 1.5492,
+      "step": 10030
+    },
+    {
+      "epoch": 2.8231818821212546,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.9292587648017597e-06,
+      "loss": 1.6118,
+      "step": 10035
+    },
+    {
+      "epoch": 2.8245885497257,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.925058582145844e-06,
+      "loss": 1.4627,
+      "step": 10040
+    },
+    {
+      "epoch": 2.825995217330145,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.9208615272457907e-06,
+      "loss": 1.9225,
+      "step": 10045
+    },
+    {
+      "epoch": 2.82740188493459,
+      "grad_norm": 3.375,
+      "learning_rate": 1.916667606428216e-06,
+      "loss": 1.4313,
+      "step": 10050
+    },
+    {
+      "epoch": 2.828808552539035,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9124768260150144e-06,
+      "loss": 1.7128,
+      "step": 10055
+    },
+    {
+      "epoch": 2.8302152201434803,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.9082891923233453e-06,
+      "loss": 1.4724,
+      "step": 10060
+    },
+    {
+      "epoch": 2.831621887747925,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.9041047116656279e-06,
+      "loss": 1.5864,
+      "step": 10065
+    },
+    {
+      "epoch": 2.83302855535237,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.8999233903495262e-06,
+      "loss": 1.5653,
+      "step": 10070
+    },
+    {
+      "epoch": 2.8344352229568153,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8957452346779399e-06,
+      "loss": 1.681,
+      "step": 10075
+    },
+    {
+      "epoch": 2.8358418905612606,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.8915702509490035e-06,
+      "loss": 1.5478,
+      "step": 10080
+    },
+    {
+      "epoch": 2.8372485581657054,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.88739844545606e-06,
+      "loss": 1.6975,
+      "step": 10085
+    },
+    {
+      "epoch": 2.8386552257701503,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.8832298244876718e-06,
+      "loss": 1.618,
+      "step": 10090
+    },
+    {
+      "epoch": 2.8400618933745956,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.8790643943275946e-06,
+      "loss": 1.4996,
+      "step": 10095
+    },
+    {
+      "epoch": 2.841468560979041,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.8749021612547762e-06,
+      "loss": 1.6171,
+      "step": 10100
+    },
+    {
+      "epoch": 2.8428752285834857,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8707431315433433e-06,
+      "loss": 1.4352,
+      "step": 10105
+    },
+    {
+      "epoch": 2.8442818961879306,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8665873114626001e-06,
+      "loss": 1.4832,
+      "step": 10110
+    },
+    {
+      "epoch": 2.845688563792376,
+      "grad_norm": 2.5,
+      "learning_rate": 1.8624347072770026e-06,
+      "loss": 1.6562,
+      "step": 10115
+    },
+    {
+      "epoch": 2.847095231396821,
+      "grad_norm": 4.0,
+      "learning_rate": 1.8582853252461686e-06,
+      "loss": 1.5724,
+      "step": 10120
+    },
+    {
+      "epoch": 2.848501899001266,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8541391716248533e-06,
+      "loss": 1.6138,
+      "step": 10125
+    },
+    {
+      "epoch": 2.849908566605711,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.849996252662946e-06,
+      "loss": 1.6924,
+      "step": 10130
+    },
+    {
+      "epoch": 2.851315234210156,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.8458565746054657e-06,
+      "loss": 1.5711,
+      "step": 10135
+    },
+    {
+      "epoch": 2.852721901814601,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.8417201436925352e-06,
+      "loss": 1.399,
+      "step": 10140
+    },
+    {
+      "epoch": 2.8541285694190464,
+      "grad_norm": 5.0,
+      "learning_rate": 1.8375869661593933e-06,
+      "loss": 1.5267,
+      "step": 10145
+    },
+    {
+      "epoch": 2.855535237023491,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.8334570482363687e-06,
+      "loss": 1.6057,
+      "step": 10150
+    },
+    {
+      "epoch": 2.8569419046279365,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.8293303961488783e-06,
+      "loss": 1.5148,
+      "step": 10155
+    },
+    {
+      "epoch": 2.8583485722323814,
+      "grad_norm": 3.125,
+      "learning_rate": 1.8252070161174142e-06,
+      "loss": 1.6119,
+      "step": 10160
+    },
+    {
+      "epoch": 2.8597552398368267,
+      "grad_norm": 4.25,
+      "learning_rate": 1.8210869143575432e-06,
+      "loss": 1.5562,
+      "step": 10165
+    },
+    {
+      "epoch": 2.8611619074412715,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.8169700970798777e-06,
+      "loss": 1.6375,
+      "step": 10170
+    },
+    {
+      "epoch": 2.862568575045717,
+      "grad_norm": 4.0,
+      "learning_rate": 1.8128565704900925e-06,
+      "loss": 1.536,
+      "step": 10175
+    },
+    {
+      "epoch": 2.8639752426501617,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.8087463407888942e-06,
+      "loss": 1.7519,
+      "step": 10180
+    },
+    {
+      "epoch": 2.865381910254607,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.8046394141720208e-06,
+      "loss": 1.4386,
+      "step": 10185
+    },
+    {
+      "epoch": 2.866788577859052,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.8005357968302318e-06,
+      "loss": 1.7448,
+      "step": 10190
+    },
+    {
+      "epoch": 2.868195245463497,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.796435494949302e-06,
+      "loss": 1.4544,
+      "step": 10195
+    },
+    {
+      "epoch": 2.869601913067942,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.7923385147099999e-06,
+      "loss": 1.4657,
+      "step": 10200
+    },
+    {
+      "epoch": 2.8710085806723873,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.7882448622880943e-06,
+      "loss": 1.6822,
+      "step": 10205
+    },
+    {
+      "epoch": 2.872415248276832,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.7841545438543392e-06,
+      "loss": 1.5488,
+      "step": 10210
+    },
+    {
+      "epoch": 2.873821915881277,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.7800675655744528e-06,
+      "loss": 1.5145,
+      "step": 10215
+    },
+    {
+      "epoch": 2.8752285834857223,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.7759839336091296e-06,
+      "loss": 1.6704,
+      "step": 10220
+    },
+    {
+      "epoch": 2.8766352510901676,
+      "grad_norm": 2.875,
+      "learning_rate": 1.771903654114013e-06,
+      "loss": 1.5645,
+      "step": 10225
+    },
+    {
+      "epoch": 2.8780419186946125,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.7678267332396958e-06,
+      "loss": 1.727,
+      "step": 10230
+    },
+    {
+      "epoch": 2.8794485862990573,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7637531771317056e-06,
+      "loss": 1.647,
+      "step": 10235
+    },
+    {
+      "epoch": 2.8808552539035026,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7596829919305037e-06,
+      "loss": 1.4213,
+      "step": 10240
+    },
+    {
+      "epoch": 2.882261921507948,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.7556161837714606e-06,
+      "loss": 1.4697,
+      "step": 10245
+    },
+    {
+      "epoch": 2.8836685891123928,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.7515527587848652e-06,
+      "loss": 1.6989,
+      "step": 10250
+    },
+    {
+      "epoch": 2.8850752567168376,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.7474927230959025e-06,
+      "loss": 1.7053,
+      "step": 10255
+    },
+    {
+      "epoch": 2.886481924321283,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.7434360828246488e-06,
+      "loss": 1.4548,
+      "step": 10260
+    },
+    {
+      "epoch": 2.887888591925728,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.7393828440860613e-06,
+      "loss": 1.575,
+      "step": 10265
+    },
+    {
+      "epoch": 2.889295259530173,
+      "grad_norm": 5.0,
+      "learning_rate": 1.7353330129899736e-06,
+      "loss": 1.7149,
+      "step": 10270
+    },
+    {
+      "epoch": 2.890701927134618,
+      "grad_norm": 3.5,
+      "learning_rate": 1.731286595641078e-06,
+      "loss": 1.7091,
+      "step": 10275
+    },
+    {
+      "epoch": 2.8921085947390632,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7272435981389237e-06,
+      "loss": 1.5249,
+      "step": 10280
+    },
+    {
+      "epoch": 2.893515262343508,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.7232040265779038e-06,
+      "loss": 1.5553,
+      "step": 10285
+    },
+    {
+      "epoch": 2.8949219299479534,
+      "grad_norm": 4.875,
+      "learning_rate": 1.7191678870472459e-06,
+      "loss": 1.3392,
+      "step": 10290
+    },
+    {
+      "epoch": 2.8963285975523982,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.715135185631008e-06,
+      "loss": 1.6271,
+      "step": 10295
+    },
+    {
+      "epoch": 2.8977352651568435,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.7111059284080627e-06,
+      "loss": 1.7061,
+      "step": 10300
+    },
+    {
+      "epoch": 2.8991419327612884,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.7070801214520904e-06,
+      "loss": 1.4828,
+      "step": 10305
+    },
+    {
+      "epoch": 2.9005486003657337,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.703057770831572e-06,
+      "loss": 1.6617,
+      "step": 10310
+    },
+    {
+      "epoch": 2.9019552679701786,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.699038882609778e-06,
+      "loss": 1.6372,
+      "step": 10315
+    },
+    {
+      "epoch": 2.903361935574624,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.695023462844757e-06,
+      "loss": 1.6865,
+      "step": 10320
+    },
+    {
+      "epoch": 2.9047686031790687,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6910115175893362e-06,
+      "loss": 1.5584,
+      "step": 10325
+    },
+    {
+      "epoch": 2.906175270783514,
+      "grad_norm": 3.125,
+      "learning_rate": 1.6870030528910983e-06,
+      "loss": 1.6264,
+      "step": 10330
+    },
+    {
+      "epoch": 2.907581938387959,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.6829980747923828e-06,
+      "loss": 1.6237,
+      "step": 10335
+    },
+    {
+      "epoch": 2.9089886059924037,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.6789965893302723e-06,
+      "loss": 1.7661,
+      "step": 10340
+    },
+    {
+      "epoch": 2.910395273596849,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.6749986025365836e-06,
+      "loss": 1.5507,
+      "step": 10345
+    },
+    {
+      "epoch": 2.9118019412012943,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.6710041204378649e-06,
+      "loss": 1.525,
+      "step": 10350
+    },
+    {
+      "epoch": 2.913208608805739,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.667013149055375e-06,
+      "loss": 1.4026,
+      "step": 10355
+    },
+    {
+      "epoch": 2.914615276410184,
+      "grad_norm": 3.25,
+      "learning_rate": 1.6630256944050842e-06,
+      "loss": 1.7627,
+      "step": 10360
+    },
+    {
+      "epoch": 2.9160219440146293,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.659041762497659e-06,
+      "loss": 1.6459,
+      "step": 10365
+    },
+    {
+      "epoch": 2.9174286116190746,
+      "grad_norm": 3.25,
+      "learning_rate": 1.6550613593384614e-06,
+      "loss": 1.6321,
+      "step": 10370
+    },
+    {
+      "epoch": 2.9188352792235195,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.6510844909275257e-06,
+      "loss": 1.5042,
+      "step": 10375
+    },
+    {
+      "epoch": 2.9202419468279643,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.6471111632595665e-06,
+      "loss": 1.4446,
+      "step": 10380
+    },
+    {
+      "epoch": 2.9216486144324096,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6431413823239551e-06,
+      "loss": 1.6827,
+      "step": 10385
+    },
+    {
+      "epoch": 2.923055282036855,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.6391751541047189e-06,
+      "loss": 1.4172,
+      "step": 10390
+    },
+    {
+      "epoch": 2.9244619496413,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6352124845805286e-06,
+      "loss": 1.4812,
+      "step": 10395
+    },
+    {
+      "epoch": 2.9258686172457447,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.6312533797246957e-06,
+      "loss": 1.4395,
+      "step": 10400
+    },
+    {
+      "epoch": 2.92727528485019,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.627297845505148e-06,
+      "loss": 1.4606,
+      "step": 10405
+    },
+    {
+      "epoch": 2.928681952454635,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.6233458878844418e-06,
+      "loss": 1.7314,
+      "step": 10410
+    },
+    {
+      "epoch": 2.93008862005908,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.6193975128197356e-06,
+      "loss": 1.7334,
+      "step": 10415
+    },
+    {
+      "epoch": 2.931495287663525,
+      "grad_norm": 4.75,
+      "learning_rate": 1.6154527262627889e-06,
+      "loss": 1.5945,
+      "step": 10420
+    },
+    {
+      "epoch": 2.9329019552679703,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.6115115341599542e-06,
+      "loss": 1.5073,
+      "step": 10425
+    },
+    {
+      "epoch": 2.934308622872415,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.6075739424521623e-06,
+      "loss": 1.8863,
+      "step": 10430
+    },
+    {
+      "epoch": 2.9357152904768604,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6036399570749194e-06,
+      "loss": 1.7013,
+      "step": 10435
+    },
+    {
+      "epoch": 2.9371219580813053,
+      "grad_norm": 2.625,
+      "learning_rate": 1.5997095839582927e-06,
+      "loss": 1.7134,
+      "step": 10440
+    },
+    {
+      "epoch": 2.9385286256857506,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.59578282902691e-06,
+      "loss": 1.5742,
+      "step": 10445
+    },
+    {
+      "epoch": 2.9399352932901954,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.5918596981999359e-06,
+      "loss": 1.5151,
+      "step": 10450
+    },
+    {
+      "epoch": 2.9413419608946407,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.5879401973910813e-06,
+      "loss": 1.6329,
+      "step": 10455
+    },
+    {
+      "epoch": 2.9427486284990856,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.58402433250858e-06,
+      "loss": 1.4513,
+      "step": 10460
+    },
+    {
+      "epoch": 2.944155296103531,
+      "grad_norm": 2.875,
+      "learning_rate": 1.5801121094551863e-06,
+      "loss": 1.5627,
+      "step": 10465
+    },
+    {
+      "epoch": 2.9455619637079757,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5762035341281634e-06,
+      "loss": 1.7602,
+      "step": 10470
+    },
+    {
+      "epoch": 2.946968631312421,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5722986124192813e-06,
+      "loss": 1.486,
+      "step": 10475
+    },
+    {
+      "epoch": 2.948375298916866,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.5683973502147936e-06,
+      "loss": 1.6301,
+      "step": 10480
+    },
+    {
+      "epoch": 2.9497819665213107,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.564499753395446e-06,
+      "loss": 1.4134,
+      "step": 10485
+    },
+    {
+      "epoch": 2.951188634125756,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.5606058278364546e-06,
+      "loss": 1.5875,
+      "step": 10490
+    },
+    {
+      "epoch": 2.9525953017302013,
+      "grad_norm": 2.625,
+      "learning_rate": 1.5567155794075016e-06,
+      "loss": 1.6243,
+      "step": 10495
+    },
+    {
+      "epoch": 2.954001969334646,
+      "grad_norm": 3.25,
+      "learning_rate": 1.55282901397273e-06,
+      "loss": 1.5925,
+      "step": 10500
+    },
+    {
+      "epoch": 2.955408636939091,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.548946137390724e-06,
+      "loss": 1.4384,
+      "step": 10505
+    },
+    {
+      "epoch": 2.9568153045435364,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.5450669555145153e-06,
+      "loss": 1.4555,
+      "step": 10510
+    },
+    {
+      "epoch": 2.9582219721479817,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.54119147419156e-06,
+      "loss": 1.7412,
+      "step": 10515
+    },
+    {
+      "epoch": 2.9596286397524265,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.5373196992637403e-06,
+      "loss": 1.3866,
+      "step": 10520
+    },
+    {
+      "epoch": 2.9610353073568714,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5334516365673462e-06,
+      "loss": 1.6634,
+      "step": 10525
+    },
+    {
+      "epoch": 2.9624419749613167,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.529587291933081e-06,
+      "loss": 1.5429,
+      "step": 10530
+    },
+    {
+      "epoch": 2.963848642565762,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.5257266711860308e-06,
+      "loss": 1.6514,
+      "step": 10535
+    },
+    {
+      "epoch": 2.965255310170207,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.5218697801456802e-06,
+      "loss": 1.5855,
+      "step": 10540
+    },
+    {
+      "epoch": 2.9666619777746517,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5180166246258846e-06,
+      "loss": 1.5531,
+      "step": 10545
+    },
+    {
+      "epoch": 2.968068645379097,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.5141672104348708e-06,
+      "loss": 1.6195,
+      "step": 10550
+    },
+    {
+      "epoch": 2.969475312983542,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.5103215433752245e-06,
+      "loss": 1.616,
+      "step": 10555
+    },
+    {
+      "epoch": 2.970881980587987,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.5064796292438868e-06,
+      "loss": 1.5293,
+      "step": 10560
+    },
+    {
+      "epoch": 2.972288648192432,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.502641473832137e-06,
+      "loss": 1.4734,
+      "step": 10565
+    },
+    {
+      "epoch": 2.9736953157968773,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.4988070829255902e-06,
+      "loss": 1.772,
+      "step": 10570
+    },
+    {
+      "epoch": 2.975101983401322,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.4949764623041907e-06,
+      "loss": 1.6811,
+      "step": 10575
+    },
+    {
+      "epoch": 2.9765086510057674,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4911496177421903e-06,
+      "loss": 1.6491,
+      "step": 10580
+    },
+    {
+      "epoch": 2.9779153186102123,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4873265550081593e-06,
+      "loss": 1.5423,
+      "step": 10585
+    },
+    {
+      "epoch": 2.9793219862146576,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.4835072798649607e-06,
+      "loss": 1.8184,
+      "step": 10590
+    },
+    {
+      "epoch": 2.9807286538191025,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.47969179806975e-06,
+      "loss": 1.658,
+      "step": 10595
+    },
+    {
+      "epoch": 2.9821353214235478,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4758801153739632e-06,
+      "loss": 1.7618,
+      "step": 10600
+    },
+    {
+      "epoch": 2.9835419890279926,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.4720722375233154e-06,
+      "loss": 1.693,
+      "step": 10605
+    },
+    {
+      "epoch": 2.984948656632438,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4682681702577756e-06,
+      "loss": 1.5824,
+      "step": 10610
+    },
+    {
+      "epoch": 2.9863553242368828,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.4644679193115793e-06,
+      "loss": 1.3904,
+      "step": 10615
+    },
+    {
+      "epoch": 2.987761991841328,
+      "grad_norm": 4.625,
+      "learning_rate": 1.4606714904132034e-06,
+      "loss": 1.4736,
+      "step": 10620
+    },
+    {
+      "epoch": 2.989168659445773,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.4568788892853653e-06,
+      "loss": 1.4967,
+      "step": 10625
+    },
+    {
+      "epoch": 2.990575327050218,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4530901216450113e-06,
+      "loss": 1.5161,
+      "step": 10630
+    },
+    {
+      "epoch": 2.991981994654663,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4493051932033113e-06,
+      "loss": 1.6695,
+      "step": 10635
+    },
+    {
+      "epoch": 2.9933886622591084,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.4455241096656466e-06,
+      "loss": 1.4315,
+      "step": 10640
+    },
+    {
+      "epoch": 2.9947953298635532,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.4417468767316022e-06,
+      "loss": 1.7368,
+      "step": 10645
+    },
+    {
+      "epoch": 2.996201997467998,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.437973500094959e-06,
+      "loss": 1.5735,
+      "step": 10650
+    },
+    {
+      "epoch": 2.9976086650724434,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.4342039854436849e-06,
+      "loss": 1.4599,
+      "step": 10655
+    },
+    {
+      "epoch": 2.9990153326768887,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.4304383384599281e-06,
+      "loss": 1.7635,
+      "step": 10660
+    },
+    {
+      "epoch": 2.9998593332395553,
+      "eval_loss": 1.5767865180969238,
+      "eval_runtime": 330.5515,
+      "eval_samples_per_second": 9.554,
+      "eval_steps_per_second": 4.777,
+      "step": 10663
+    },
+    {
+      "epoch": 3.0004220002813335,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.4266765648200045e-06,
+      "loss": 1.6251,
+      "step": 10665
+    },
+    {
+      "epoch": 3.0018286678857784,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.4229186701943925e-06,
+      "loss": 1.4799,
+      "step": 10670
+    },
+    {
+      "epoch": 3.0032353354902237,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.4191646602477216e-06,
+      "loss": 1.5905,
+      "step": 10675
+    },
+    {
+      "epoch": 3.0046420030946686,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.4154145406387681e-06,
+      "loss": 1.5103,
+      "step": 10680
+    },
+    {
+      "epoch": 3.006048670699114,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4116683170204407e-06,
+      "loss": 1.5813,
+      "step": 10685
+    },
+    {
+      "epoch": 3.0074553383035587,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.40792599503978e-06,
+      "loss": 1.4847,
+      "step": 10690
+    },
+    {
+      "epoch": 3.008862005908004,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.404187580337941e-06,
+      "loss": 1.6993,
+      "step": 10695
+    },
+    {
+      "epoch": 3.010268673512449,
+      "grad_norm": 2.875,
+      "learning_rate": 1.40045307855019e-06,
+      "loss": 1.7316,
+      "step": 10700
+    },
+    {
+      "epoch": 3.011675341116894,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3967224953058988e-06,
+      "loss": 1.6364,
+      "step": 10705
+    },
+    {
+      "epoch": 3.013082008721339,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3929958362285242e-06,
+      "loss": 1.6369,
+      "step": 10710
+    },
+    {
+      "epoch": 3.0144886763257843,
+      "grad_norm": 5.75,
+      "learning_rate": 1.3892731069356161e-06,
+      "loss": 1.6918,
+      "step": 10715
+    },
+    {
+      "epoch": 3.015895343930229,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.3855543130387965e-06,
+      "loss": 1.6818,
+      "step": 10720
+    },
+    {
+      "epoch": 3.0173020115346745,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.3818394601437557e-06,
+      "loss": 1.6582,
+      "step": 10725
+    },
+    {
+      "epoch": 3.0187086791391193,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.3781285538502418e-06,
+      "loss": 1.522,
+      "step": 10730
+    },
+    {
+      "epoch": 3.0201153467435646,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3744215997520602e-06,
+      "loss": 1.4227,
+      "step": 10735
+    },
+    {
+      "epoch": 3.0215220143480095,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3707186034370484e-06,
+      "loss": 1.6587,
+      "step": 10740
+    },
+    {
+      "epoch": 3.022928681952455,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.3670195704870883e-06,
+      "loss": 1.5284,
+      "step": 10745
+    },
+    {
+      "epoch": 3.0243353495568996,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3633245064780803e-06,
+      "loss": 1.6379,
+      "step": 10750
+    },
+    {
+      "epoch": 3.025742017161345,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.3596334169799457e-06,
+      "loss": 1.4993,
+      "step": 10755
+    },
+    {
+      "epoch": 3.02714868476579,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.355946307556612e-06,
+      "loss": 1.6514,
+      "step": 10760
+    },
+    {
+      "epoch": 3.028555352370235,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.3522631837660123e-06,
+      "loss": 1.5571,
+      "step": 10765
+    },
+    {
+      "epoch": 3.02996201997468,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.3485840511600636e-06,
+      "loss": 1.47,
+      "step": 10770
+    },
+    {
+      "epoch": 3.0313686875791253,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.3449089152846726e-06,
+      "loss": 1.6482,
+      "step": 10775
+    },
+    {
+      "epoch": 3.03277535518357,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.341237781679724e-06,
+      "loss": 1.5908,
+      "step": 10780
+    },
+    {
+      "epoch": 3.0341820227880154,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.337570655879059e-06,
+      "loss": 1.6287,
+      "step": 10785
+    },
+    {
+      "epoch": 3.0355886903924603,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.3339075434104885e-06,
+      "loss": 1.6187,
+      "step": 10790
+    },
+    {
+      "epoch": 3.036995357996905,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3302484497957678e-06,
+      "loss": 1.6408,
+      "step": 10795
+    },
+    {
+      "epoch": 3.0384020256013504,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.3265933805505954e-06,
+      "loss": 1.7244,
+      "step": 10800
+    },
+    {
+      "epoch": 3.0398086932057953,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3229423411846018e-06,
+      "loss": 1.8399,
+      "step": 10805
+    },
+    {
+      "epoch": 3.0412153608102406,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.319295337201349e-06,
+      "loss": 1.4242,
+      "step": 10810
+    },
+    {
+      "epoch": 3.0426220284146854,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.315652374098307e-06,
+      "loss": 1.5995,
+      "step": 10815
+    },
+    {
+      "epoch": 3.0440286960191307,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.3120134573668624e-06,
+      "loss": 1.6642,
+      "step": 10820
+    },
+    {
+      "epoch": 3.0454353636235756,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3083785924922986e-06,
+      "loss": 1.3545,
+      "step": 10825
+    },
+    {
+      "epoch": 3.046842031228021,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.3047477849537916e-06,
+      "loss": 1.3587,
+      "step": 10830
+    },
+    {
+      "epoch": 3.0482486988324657,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.3011210402244008e-06,
+      "loss": 1.6701,
+      "step": 10835
+    },
+    {
+      "epoch": 3.049655366436911,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2974983637710644e-06,
+      "loss": 1.5038,
+      "step": 10840
+    },
+    {
+      "epoch": 3.051062034041356,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.293879761054585e-06,
+      "loss": 1.3233,
+      "step": 10845
+    },
+    {
+      "epoch": 3.052468701645801,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2902652375296255e-06,
+      "loss": 1.3758,
+      "step": 10850
+    },
+    {
+      "epoch": 3.053875369250246,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.2866547986446993e-06,
+      "loss": 1.495,
+      "step": 10855
+    },
+    {
+      "epoch": 3.0552820368546914,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.283048449842162e-06,
+      "loss": 1.4547,
+      "step": 10860
+    },
+    {
+      "epoch": 3.056688704459136,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2794461965582098e-06,
+      "loss": 1.5248,
+      "step": 10865
+    },
+    {
+      "epoch": 3.0580953720635815,
+      "grad_norm": 3.25,
+      "learning_rate": 1.275848044222854e-06,
+      "loss": 1.5124,
+      "step": 10870
+    },
+    {
+      "epoch": 3.0595020396680264,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2722539982599352e-06,
+      "loss": 1.4788,
+      "step": 10875
+    },
+    {
+      "epoch": 3.0609087072724717,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.268664064087098e-06,
+      "loss": 1.5596,
+      "step": 10880
+    },
+    {
+      "epoch": 3.0623153748769165,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.2650782471157904e-06,
+      "loss": 1.6641,
+      "step": 10885
+    },
+    {
+      "epoch": 3.063722042481362,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2614965527512533e-06,
+      "loss": 1.4911,
+      "step": 10890
+    },
+    {
+      "epoch": 3.0651287100858067,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2579189863925175e-06,
+      "loss": 1.5126,
+      "step": 10895
+    },
+    {
+      "epoch": 3.066535377690252,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.2543455534323828e-06,
+      "loss": 1.3572,
+      "step": 10900
+    },
+    {
+      "epoch": 3.067942045294697,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.2507762592574272e-06,
+      "loss": 1.7074,
+      "step": 10905
+    },
+    {
+      "epoch": 3.069348712899142,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.2472111092479853e-06,
+      "loss": 1.5994,
+      "step": 10910
+    },
+    {
+      "epoch": 3.070755380503587,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2436501087781435e-06,
+      "loss": 1.6489,
+      "step": 10915
+    },
+    {
+      "epoch": 3.0721620481080323,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.2400932632157389e-06,
+      "loss": 1.4295,
+      "step": 10920
+    },
+    {
+      "epoch": 3.073568715712477,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.2365405779223395e-06,
+      "loss": 1.6276,
+      "step": 10925
+    },
+    {
+      "epoch": 3.0749753833169224,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2329920582532451e-06,
+      "loss": 1.5665,
+      "step": 10930
+    },
+    {
+      "epoch": 3.0763820509213673,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.229447709557475e-06,
+      "loss": 1.7107,
+      "step": 10935
+    },
+    {
+      "epoch": 3.077788718525812,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.2259075371777648e-06,
+      "loss": 1.6144,
+      "step": 10940
+    },
+    {
+      "epoch": 3.0791953861302575,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.2223715464505473e-06,
+      "loss": 1.7557,
+      "step": 10945
+    },
+    {
+      "epoch": 3.0806020537347023,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.2188397427059607e-06,
+      "loss": 1.5287,
+      "step": 10950
+    },
+    {
+      "epoch": 3.0820087213391476,
+      "grad_norm": 3.125,
+      "learning_rate": 1.215312131267825e-06,
+      "loss": 1.8295,
+      "step": 10955
+    },
+    {
+      "epoch": 3.0834153889435925,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.2117887174536444e-06,
+      "loss": 1.646,
+      "step": 10960
+    },
+    {
+      "epoch": 3.0848220565480378,
+      "grad_norm": 4.375,
+      "learning_rate": 1.2082695065745925e-06,
+      "loss": 1.6581,
+      "step": 10965
+    },
+    {
+      "epoch": 3.0862287241524826,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.2047545039355141e-06,
+      "loss": 1.7364,
+      "step": 10970
+    },
+    {
+      "epoch": 3.087635391756928,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.2012437148348994e-06,
+      "loss": 1.4243,
+      "step": 10975
+    },
+    {
+      "epoch": 3.0890420593613728,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.1977371445648988e-06,
+      "loss": 1.5345,
+      "step": 10980
+    },
+    {
+      "epoch": 3.090448726965818,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.1942347984112959e-06,
+      "loss": 1.6182,
+      "step": 10985
+    },
+    {
+      "epoch": 3.091855394570263,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1907366816535076e-06,
+      "loss": 1.686,
+      "step": 10990
+    },
+    {
+      "epoch": 3.0932620621747082,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1872427995645803e-06,
+      "loss": 1.6519,
+      "step": 10995
+    },
+    {
+      "epoch": 3.094668729779153,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.183753157411171e-06,
+      "loss": 1.5121,
+      "step": 11000
+    },
+    {
+      "epoch": 3.0960753973835984,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1802677604535496e-06,
+      "loss": 1.5297,
+      "step": 11005
+    },
+    {
+      "epoch": 3.0974820649880432,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.176786613945584e-06,
+      "loss": 1.3636,
+      "step": 11010
+    },
+    {
+      "epoch": 3.0988887325924885,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1733097231347372e-06,
+      "loss": 1.6959,
+      "step": 11015
+    },
+    {
+      "epoch": 3.1002954001969334,
+      "grad_norm": 2.875,
+      "learning_rate": 1.1698370932620538e-06,
+      "loss": 1.5115,
+      "step": 11020
+    },
+    {
+      "epoch": 3.1017020678013787,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.1663687295621621e-06,
+      "loss": 1.5369,
+      "step": 11025
+    },
+    {
+      "epoch": 3.1031087354058235,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1629046372632524e-06,
+      "loss": 1.5513,
+      "step": 11030
+    },
+    {
+      "epoch": 3.104515403010269,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1594448215870812e-06,
+      "loss": 1.4403,
+      "step": 11035
+    },
+    {
+      "epoch": 3.1059220706147137,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.1559892877489548e-06,
+      "loss": 1.6092,
+      "step": 11040
+    },
+    {
+      "epoch": 3.107328738219159,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1525380409577282e-06,
+      "loss": 1.6913,
+      "step": 11045
+    },
+    {
+      "epoch": 3.108735405823604,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.1490910864157907e-06,
+      "loss": 1.4355,
+      "step": 11050
+    },
+    {
+      "epoch": 3.110142073428049,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.1456484293190668e-06,
+      "loss": 1.624,
+      "step": 11055
+    },
+    {
+      "epoch": 3.111548741032494,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.1422100748569982e-06,
+      "loss": 1.2994,
+      "step": 11060
+    },
+    {
+      "epoch": 3.112955408636939,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.1387760282125412e-06,
+      "loss": 1.52,
+      "step": 11065
+    },
+    {
+      "epoch": 3.114362076241384,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.1353462945621632e-06,
+      "loss": 1.5788,
+      "step": 11070
+    },
+    {
+      "epoch": 3.115768743845829,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.1319208790758223e-06,
+      "loss": 1.542,
+      "step": 11075
+    },
+    {
+      "epoch": 3.1171754114502743,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1284997869169756e-06,
+      "loss": 1.5894,
+      "step": 11080
+    },
+    {
+      "epoch": 3.118582079054719,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.125083023242558e-06,
+      "loss": 1.6021,
+      "step": 11085
+    },
+    {
+      "epoch": 3.1199887466591645,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.1216705932029816e-06,
+      "loss": 1.7714,
+      "step": 11090
+    },
+    {
+      "epoch": 3.1213954142636093,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.1182625019421244e-06,
+      "loss": 1.5162,
+      "step": 11095
+    },
+    {
+      "epoch": 3.1228020818680546,
+      "grad_norm": 4.25,
+      "learning_rate": 1.114858754597329e-06,
+      "loss": 1.7289,
+      "step": 11100
+    },
+    {
+      "epoch": 3.1242087494724995,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.111459356299381e-06,
+      "loss": 1.5885,
+      "step": 11105
+    },
+    {
+      "epoch": 3.125615417076945,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.10806431217252e-06,
+      "loss": 1.5521,
+      "step": 11110
+    },
+    {
+      "epoch": 3.1270220846813896,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.104673627334416e-06,
+      "loss": 1.578,
+      "step": 11115
+    },
+    {
+      "epoch": 3.128428752285835,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1012873068961702e-06,
+      "loss": 1.6254,
+      "step": 11120
+    },
+    {
+      "epoch": 3.12983541989028,
+      "grad_norm": 4.0,
+      "learning_rate": 1.0979053559623026e-06,
+      "loss": 1.4843,
+      "step": 11125
+    },
+    {
+      "epoch": 3.131242087494725,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0945277796307513e-06,
+      "loss": 1.5013,
+      "step": 11130
+    },
+    {
+      "epoch": 3.13264875509917,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.0911545829928552e-06,
+      "loss": 1.5731,
+      "step": 11135
+    },
+    {
+      "epoch": 3.1340554227036153,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.0877857711333534e-06,
+      "loss": 1.6624,
+      "step": 11140
+    },
+    {
+      "epoch": 3.13546209030806,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0844213491303772e-06,
+      "loss": 1.766,
+      "step": 11145
+    },
+    {
+      "epoch": 3.1368687579125054,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.0810613220554356e-06,
+      "loss": 1.6699,
+      "step": 11150
+    },
+    {
+      "epoch": 3.1382754255169503,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0777056949734187e-06,
+      "loss": 1.7605,
+      "step": 11155
+    },
+    {
+      "epoch": 3.1396820931213956,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0743544729425802e-06,
+      "loss": 1.6822,
+      "step": 11160
+    },
+    {
+      "epoch": 3.1410887607258404,
+      "grad_norm": 4.25,
+      "learning_rate": 1.0710076610145344e-06,
+      "loss": 1.6808,
+      "step": 11165
+    },
+    {
+      "epoch": 3.1424954283302857,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0676652642342471e-06,
+      "loss": 1.5603,
+      "step": 11170
+    },
+    {
+      "epoch": 3.1439020959347306,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.064327287640034e-06,
+      "loss": 1.6544,
+      "step": 11175
+    },
+    {
+      "epoch": 3.145308763539176,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.0609937362635376e-06,
+      "loss": 1.6709,
+      "step": 11180
+    },
+    {
+      "epoch": 3.1467154311436207,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0576646151297404e-06,
+      "loss": 1.6512,
+      "step": 11185
+    },
+    {
+      "epoch": 3.148122098748066,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.0543399292569404e-06,
+      "loss": 1.7634,
+      "step": 11190
+    },
+    {
+      "epoch": 3.149528766352511,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.0510196836567522e-06,
+      "loss": 1.4378,
+      "step": 11195
+    },
+    {
+      "epoch": 3.150935433956956,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.0477038833340964e-06,
+      "loss": 1.5599,
+      "step": 11200
+    },
+    {
+      "epoch": 3.152342101561401,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.0443925332871914e-06,
+      "loss": 1.6039,
+      "step": 11205
+    },
+    {
+      "epoch": 3.153748769165846,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.0410856385075528e-06,
+      "loss": 1.5945,
+      "step": 11210
+    },
+    {
+      "epoch": 3.155155436770291,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.037783203979974e-06,
+      "loss": 1.514,
+      "step": 11215
+    },
+    {
+      "epoch": 3.156562104374736,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0344852346825282e-06,
+      "loss": 1.3017,
+      "step": 11220
+    },
+    {
+      "epoch": 3.1579687719791814,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.0311917355865554e-06,
+      "loss": 1.4708,
+      "step": 11225
+    },
+    {
+      "epoch": 3.159375439583626,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.027902711656663e-06,
+      "loss": 1.676,
+      "step": 11230
+    },
+    {
+      "epoch": 3.1607821071880715,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.0246181678507038e-06,
+      "loss": 1.5518,
+      "step": 11235
+    },
+    {
+      "epoch": 3.1621887747925164,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.0213381091197852e-06,
+      "loss": 1.561,
+      "step": 11240
+    },
+    {
+      "epoch": 3.1635954423969617,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0180625404082497e-06,
+      "loss": 1.7335,
+      "step": 11245
+    },
+    {
+      "epoch": 3.1650021100014065,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0147914666536718e-06,
+      "loss": 1.5096,
+      "step": 11250
+    },
+    {
+      "epoch": 3.166408777605852,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.011524892786851e-06,
+      "loss": 1.5547,
+      "step": 11255
+    },
+    {
+      "epoch": 3.1678154452102967,
+      "grad_norm": 4.125,
+      "learning_rate": 1.0082628237318065e-06,
+      "loss": 1.6803,
+      "step": 11260
+    },
+    {
+      "epoch": 3.169222112814742,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.0050052644057592e-06,
+      "loss": 1.6371,
+      "step": 11265
+    },
+    {
+      "epoch": 3.170628780419187,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.00175221971914e-06,
+      "loss": 1.6701,
+      "step": 11270
+    },
+    {
+      "epoch": 3.172035448023632,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.985036945755734e-07,
+      "loss": 1.7211,
+      "step": 11275
+    },
+    {
+      "epoch": 3.173442115628077,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.952596938718648e-07,
+      "loss": 1.5572,
+      "step": 11280
+    },
+    {
+      "epoch": 3.1748487832325223,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.920202224980072e-07,
+      "loss": 1.5249,
+      "step": 11285
+    },
+    {
+      "epoch": 3.176255450836967,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.887852853371615e-07,
+      "loss": 1.4818,
+      "step": 11290
+    },
+    {
+      "epoch": 3.1776621184414124,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.855548872656557e-07,
+      "loss": 1.5171,
+      "step": 11295
+    },
+    {
+      "epoch": 3.1790687860458573,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.823290331529736e-07,
+      "loss": 1.7529,
+      "step": 11300
+    },
+    {
+      "epoch": 3.1804754536503026,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.791077278617538e-07,
+      "loss": 1.546,
+      "step": 11305
+    },
+    {
+      "epoch": 3.1818821212547475,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.758909762477717e-07,
+      "loss": 1.6178,
+      "step": 11310
+    },
+    {
+      "epoch": 3.1832887888591928,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.726787831599436e-07,
+      "loss": 1.4941,
+      "step": 11315
+    },
+    {
+      "epoch": 3.1846954564636376,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.694711534403128e-07,
+      "loss": 1.6382,
+      "step": 11320
+    },
+    {
+      "epoch": 3.186102124068083,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.662680919240434e-07,
+      "loss": 1.4399,
+      "step": 11325
+    },
+    {
+      "epoch": 3.1875087916725278,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.630696034394118e-07,
+      "loss": 1.2156,
+      "step": 11330
+    },
+    {
+      "epoch": 3.1889154592769726,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.598756928078069e-07,
+      "loss": 1.7228,
+      "step": 11335
+    },
+    {
+      "epoch": 3.190322126881418,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.56686364843708e-07,
+      "loss": 1.6953,
+      "step": 11340
+    },
+    {
+      "epoch": 3.191728794485863,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.535016243546952e-07,
+      "loss": 1.586,
+      "step": 11345
+    },
+    {
+      "epoch": 3.193135462090308,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.503214761414277e-07,
+      "loss": 1.7237,
+      "step": 11350
+    },
+    {
+      "epoch": 3.194542129694753,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.471459249976446e-07,
+      "loss": 1.4558,
+      "step": 11355
+    },
+    {
+      "epoch": 3.1959487972991982,
+      "grad_norm": 4.5,
+      "learning_rate": 9.439749757101561e-07,
+      "loss": 1.5914,
+      "step": 11360
+    },
+    {
+      "epoch": 3.197355464903643,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.408086330588343e-07,
+      "loss": 1.6958,
+      "step": 11365
+    },
+    {
+      "epoch": 3.1987621325080884,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.376469018166071e-07,
+      "loss": 1.6927,
+      "step": 11370
+    },
+    {
+      "epoch": 3.2001688001125332,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.344897867494515e-07,
+      "loss": 1.6201,
+      "step": 11375
+    },
+    {
+      "epoch": 3.2015754677169785,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.313372926163867e-07,
+      "loss": 1.5303,
+      "step": 11380
+    },
+    {
+      "epoch": 3.2029821353214234,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.28189424169465e-07,
+      "loss": 1.7088,
+      "step": 11385
+    },
+    {
+      "epoch": 3.2043888029258687,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.250461861537684e-07,
+      "loss": 1.5421,
+      "step": 11390
+    },
+    {
+      "epoch": 3.2057954705303136,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.21907583307397e-07,
+      "loss": 1.6744,
+      "step": 11395
+    },
+    {
+      "epoch": 3.207202138134759,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.187736203614638e-07,
+      "loss": 1.5669,
+      "step": 11400
+    },
+    {
+      "epoch": 3.2086088057392037,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.156443020400883e-07,
+      "loss": 1.6098,
+      "step": 11405
+    },
+    {
+      "epoch": 3.210015473343649,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.125196330603877e-07,
+      "loss": 1.504,
+      "step": 11410
+    },
+    {
+      "epoch": 3.211422140948094,
+      "grad_norm": 9.0,
+      "learning_rate": 9.093996181324742e-07,
+      "loss": 1.4803,
+      "step": 11415
+    },
+    {
+      "epoch": 3.212828808552539,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.062842619594402e-07,
+      "loss": 1.5536,
+      "step": 11420
+    },
+    {
+      "epoch": 3.214235476156984,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.031735692373578e-07,
+      "loss": 1.8039,
+      "step": 11425
+    },
+    {
+      "epoch": 3.2156421437614293,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.000675446552679e-07,
+      "loss": 1.7332,
+      "step": 11430
+    },
+    {
+      "epoch": 3.217048811365874,
+      "grad_norm": 3.0,
+      "learning_rate": 8.969661928951789e-07,
+      "loss": 1.3582,
+      "step": 11435
+    },
+    {
+      "epoch": 3.2184554789703195,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.938695186320475e-07,
+      "loss": 1.5736,
+      "step": 11440
+    },
+    {
+      "epoch": 3.2198621465747643,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.90777526533788e-07,
+      "loss": 1.5083,
+      "step": 11445
+    },
+    {
+      "epoch": 3.2212688141792096,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.87690221261252e-07,
+      "loss": 1.5167,
+      "step": 11450
+    },
+    {
+      "epoch": 3.2226754817836545,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.846076074682276e-07,
+      "loss": 1.7006,
+      "step": 11455
+    },
+    {
+      "epoch": 3.2240821493881,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.815296898014293e-07,
+      "loss": 1.5955,
+      "step": 11460
+    },
+    {
+      "epoch": 3.2254888169925446,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.784564729004978e-07,
+      "loss": 1.6101,
+      "step": 11465
+    },
+    {
+      "epoch": 3.22689548459699,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.753879613979789e-07,
+      "loss": 1.6842,
+      "step": 11470
+    },
+    {
+      "epoch": 3.228302152201435,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.723241599193349e-07,
+      "loss": 1.2444,
+      "step": 11475
+    },
+    {
+      "epoch": 3.2297088198058796,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.692650730829232e-07,
+      "loss": 1.2935,
+      "step": 11480
+    },
+    {
+      "epoch": 3.231115487410325,
+      "grad_norm": 4.375,
+      "learning_rate": 8.662107054999936e-07,
+      "loss": 1.6327,
+      "step": 11485
+    },
+    {
+      "epoch": 3.23252215501477,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.631610617746865e-07,
+      "loss": 1.7187,
+      "step": 11490
+    },
+    {
+      "epoch": 3.233928822619215,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.601161465040179e-07,
+      "loss": 1.5457,
+      "step": 11495
+    },
+    {
+      "epoch": 3.23533549022366,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.570759642778766e-07,
+      "loss": 1.6327,
+      "step": 11500
+    },
+    {
+      "epoch": 3.2367421578281053,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.54040519679017e-07,
+      "loss": 1.6916,
+      "step": 11505
+    },
+    {
+      "epoch": 3.23814882543255,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.510098172830553e-07,
+      "loss": 1.7347,
+      "step": 11510
+    },
+    {
+      "epoch": 3.2395554930369954,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.479838616584523e-07,
+      "loss": 1.763,
+      "step": 11515
+    },
+    {
+      "epoch": 3.2409621606414403,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.449626573665209e-07,
+      "loss": 1.791,
+      "step": 11520
+    },
+    {
+      "epoch": 3.2423688282458856,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.419462089614073e-07,
+      "loss": 1.5492,
+      "step": 11525
+    },
+    {
+      "epoch": 3.2437754958503304,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.389345209900907e-07,
+      "loss": 1.3123,
+      "step": 11530
+    },
+    {
+      "epoch": 3.2451821634547757,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.359275979923723e-07,
+      "loss": 1.6477,
+      "step": 11535
+    },
+    {
+      "epoch": 3.2465888310592206,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.329254445008755e-07,
+      "loss": 1.663,
+      "step": 11540
+    },
+    {
+      "epoch": 3.247995498663666,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.299280650410265e-07,
+      "loss": 1.3256,
+      "step": 11545
+    },
+    {
+      "epoch": 3.2494021662681107,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.269354641310627e-07,
+      "loss": 1.5578,
+      "step": 11550
+    },
+    {
+      "epoch": 3.250808833872556,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.239476462820136e-07,
+      "loss": 1.6555,
+      "step": 11555
+    },
+    {
+      "epoch": 3.252215501477001,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.209646159976999e-07,
+      "loss": 1.6347,
+      "step": 11560
+    },
+    {
+      "epoch": 3.253622169081446,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.179863777747287e-07,
+      "loss": 1.4629,
+      "step": 11565
+    },
+    {
+      "epoch": 3.255028836685891,
+      "grad_norm": 3.5,
+      "learning_rate": 8.150129361024762e-07,
+      "loss": 1.6009,
+      "step": 11570
+    },
+    {
+      "epoch": 3.2564355042903363,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.120442954630964e-07,
+      "loss": 1.3342,
+      "step": 11575
+    },
+    {
+      "epoch": 3.257842171894781,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.090804603315016e-07,
+      "loss": 1.6218,
+      "step": 11580
+    },
+    {
+      "epoch": 3.2592488394992265,
+      "grad_norm": 4.75,
+      "learning_rate": 8.061214351753616e-07,
+      "loss": 1.6242,
+      "step": 11585
+    },
+    {
+      "epoch": 3.2606555071036714,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.031672244550938e-07,
+      "loss": 1.5137,
+      "step": 11590
+    },
+    {
+      "epoch": 3.2620621747081167,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.002178326238636e-07,
+      "loss": 1.6331,
+      "step": 11595
+    },
+    {
+      "epoch": 3.2634688423125615,
+      "grad_norm": 4.21875,
+      "learning_rate": 7.972732641275648e-07,
+      "loss": 1.2351,
+      "step": 11600
+    },
+    {
+      "epoch": 3.2648755099170064,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.943335234048274e-07,
+      "loss": 1.539,
+      "step": 11605
+    },
+    {
+      "epoch": 3.2662821775214517,
+      "grad_norm": 5.375,
+      "learning_rate": 7.91398614887e-07,
+      "loss": 1.53,
+      "step": 11610
+    },
+    {
+      "epoch": 3.267688845125897,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.88468542998149e-07,
+      "loss": 1.5524,
+      "step": 11615
+    },
+    {
+      "epoch": 3.269095512730342,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.855433121550481e-07,
+      "loss": 1.4712,
+      "step": 11620
+    },
+    {
+      "epoch": 3.2705021803347867,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.826229267671771e-07,
+      "loss": 1.4857,
+      "step": 11625
+    },
+    {
+      "epoch": 3.271908847939232,
+      "grad_norm": 4.4375,
+      "learning_rate": 7.797073912367085e-07,
+      "loss": 1.6786,
+      "step": 11630
+    },
+    {
+      "epoch": 3.2733155155436773,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.767967099585044e-07,
+      "loss": 1.5137,
+      "step": 11635
+    },
+    {
+      "epoch": 3.274722183148122,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.73890887320114e-07,
+      "loss": 1.7585,
+      "step": 11640
+    },
+    {
+      "epoch": 3.276128850752567,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.709899277017546e-07,
+      "loss": 1.6304,
+      "step": 11645
+    },
+    {
+      "epoch": 3.2775355183570123,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.680938354763205e-07,
+      "loss": 1.5074,
+      "step": 11650
+    },
+    {
+      "epoch": 3.278942185961457,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.652026150093656e-07,
+      "loss": 1.6919,
+      "step": 11655
+    },
+    {
+      "epoch": 3.2803488535659024,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.623162706591002e-07,
+      "loss": 1.5654,
+      "step": 11660
+    },
+    {
+      "epoch": 3.2817555211703473,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.594348067763837e-07,
+      "loss": 1.5234,
+      "step": 11665
+    },
+    {
+      "epoch": 3.2831621887747926,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.565582277047227e-07,
+      "loss": 1.4964,
+      "step": 11670
+    },
+    {
+      "epoch": 3.2845688563792375,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.536865377802532e-07,
+      "loss": 1.5708,
+      "step": 11675
+    },
+    {
+      "epoch": 3.2859755239836828,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.508197413317491e-07,
+      "loss": 1.4245,
+      "step": 11680
+    },
+    {
+      "epoch": 3.2873821915881276,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.479578426806035e-07,
+      "loss": 1.4793,
+      "step": 11685
+    },
+    {
+      "epoch": 3.288788859192573,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.45100846140827e-07,
+      "loss": 1.6957,
+      "step": 11690
+    },
+    {
+      "epoch": 3.2901955267970178,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.422487560190407e-07,
+      "loss": 1.4461,
+      "step": 11695
+    },
+    {
+      "epoch": 3.291602194401463,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.394015766144717e-07,
+      "loss": 1.7007,
+      "step": 11700
+    },
+    {
+      "epoch": 3.293008862005908,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.365593122189428e-07,
+      "loss": 1.4112,
+      "step": 11705
+    },
+    {
+      "epoch": 3.294415529610353,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.337219671168689e-07,
+      "loss": 1.5714,
+      "step": 11710
+    },
+    {
+      "epoch": 3.295822197214798,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.308895455852484e-07,
+      "loss": 1.5193,
+      "step": 11715
+    },
+    {
+      "epoch": 3.2972288648192434,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.280620518936582e-07,
+      "loss": 1.4024,
+      "step": 11720
+    },
+    {
+      "epoch": 3.2986355324236882,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.252394903042498e-07,
+      "loss": 1.6127,
+      "step": 11725
+    },
+    {
+      "epoch": 3.3000422000281335,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.224218650717361e-07,
+      "loss": 1.302,
+      "step": 11730
+    },
+    {
+      "epoch": 3.3014488676325784,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.196091804433911e-07,
+      "loss": 1.7009,
+      "step": 11735
+    },
+    {
+      "epoch": 3.3028555352370237,
+      "grad_norm": 3.375,
+      "learning_rate": 7.168014406590405e-07,
+      "loss": 1.6338,
+      "step": 11740
+    },
+    {
+      "epoch": 3.3042622028414685,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.139986499510575e-07,
+      "loss": 1.6938,
+      "step": 11745
+    },
+    {
+      "epoch": 3.3056688704459134,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.112008125443524e-07,
+      "loss": 1.4994,
+      "step": 11750
+    },
+    {
+      "epoch": 3.3070755380503587,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.084079326563728e-07,
+      "loss": 1.5465,
+      "step": 11755
+    },
+    {
+      "epoch": 3.308482205654804,
+      "grad_norm": 4.96875,
+      "learning_rate": 7.056200144970907e-07,
+      "loss": 1.2758,
+      "step": 11760
+    },
+    {
+      "epoch": 3.309888873259249,
+      "grad_norm": 4.25,
+      "learning_rate": 7.02837062268999e-07,
+      "loss": 1.6104,
+      "step": 11765
+    },
+    {
+      "epoch": 3.3112955408636937,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.000590801671049e-07,
+      "loss": 1.3466,
+      "step": 11770
+    },
+    {
+      "epoch": 3.312702208468139,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.972860723789243e-07,
+      "loss": 1.7502,
+      "step": 11775
+    },
+    {
+      "epoch": 3.314108876072584,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.945180430844754e-07,
+      "loss": 1.7875,
+      "step": 11780
+    },
+    {
+      "epoch": 3.315515543677029,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.917549964562712e-07,
+      "loss": 1.503,
+      "step": 11785
+    },
+    {
+      "epoch": 3.316922211281474,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.889969366593127e-07,
+      "loss": 1.4491,
+      "step": 11790
+    },
+    {
+      "epoch": 3.3183288788859193,
+      "grad_norm": 4.78125,
+      "learning_rate": 6.862438678510849e-07,
+      "loss": 1.2961,
+      "step": 11795
+    },
+    {
+      "epoch": 3.319735546490364,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.834957941815518e-07,
+      "loss": 1.6573,
+      "step": 11800
+    },
+    {
+      "epoch": 3.3211422140948095,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.807527197931411e-07,
+      "loss": 1.3254,
+      "step": 11805
+    },
+    {
+      "epoch": 3.3225488816992543,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.780146488207524e-07,
+      "loss": 1.5744,
+      "step": 11810
+    },
+    {
+      "epoch": 3.3239555493036996,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.752815853917377e-07,
+      "loss": 1.5929,
+      "step": 11815
+    },
+    {
+      "epoch": 3.3253622169081445,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.725535336259036e-07,
+      "loss": 1.687,
+      "step": 11820
+    },
+    {
+      "epoch": 3.32676888451259,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.698304976354992e-07,
+      "loss": 1.5552,
+      "step": 11825
+    },
+    {
+      "epoch": 3.3281755521170346,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.671124815252182e-07,
+      "loss": 1.4686,
+      "step": 11830
+    },
+    {
+      "epoch": 3.32958221972148,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.643994893921801e-07,
+      "loss": 1.5944,
+      "step": 11835
+    },
+    {
+      "epoch": 3.330988887325925,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.616915253259367e-07,
+      "loss": 1.3498,
+      "step": 11840
+    },
+    {
+      "epoch": 3.33239555493037,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.589885934084609e-07,
+      "loss": 1.8226,
+      "step": 11845
+    },
+    {
+      "epoch": 3.333802222534815,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.562906977141342e-07,
+      "loss": 1.5916,
+      "step": 11850
+    },
+    {
+      "epoch": 3.3352088901392603,
+      "grad_norm": 3.75,
+      "learning_rate": 6.535978423097535e-07,
+      "loss": 1.2771,
+      "step": 11855
+    },
+    {
+      "epoch": 3.336615557743705,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.509100312545142e-07,
+      "loss": 1.7791,
+      "step": 11860
+    },
+    {
+      "epoch": 3.3380222253481504,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.482272686000083e-07,
+      "loss": 1.5699,
+      "step": 11865
+    },
+    {
+      "epoch": 3.3394288929525953,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.455495583902175e-07,
+      "loss": 1.5746,
+      "step": 11870
+    },
+    {
+      "epoch": 3.3408355605570406,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.428769046615108e-07,
+      "loss": 1.7,
+      "step": 11875
+    },
+    {
+      "epoch": 3.3422422281614854,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.402093114426291e-07,
+      "loss": 1.6023,
+      "step": 11880
+    },
+    {
+      "epoch": 3.3436488957659307,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.375467827546908e-07,
+      "loss": 1.6378,
+      "step": 11885
+    },
+    {
+      "epoch": 3.3450555633703756,
+      "grad_norm": 4.71875,
+      "learning_rate": 6.348893226111775e-07,
+      "loss": 1.561,
+      "step": 11890
+    },
+    {
+      "epoch": 3.3464622309748204,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.3223693501793e-07,
+      "loss": 1.6388,
+      "step": 11895
+    },
+    {
+      "epoch": 3.3478688985792657,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.29589623973143e-07,
+      "loss": 1.6767,
+      "step": 11900
+    },
+    {
+      "epoch": 3.349275566183711,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.269473934673617e-07,
+      "loss": 1.6953,
+      "step": 11905
+    },
+    {
+      "epoch": 3.350682233788156,
+      "grad_norm": 4.125,
+      "learning_rate": 6.243102474834679e-07,
+      "loss": 1.6556,
+      "step": 11910
+    },
+    {
+      "epoch": 3.3520889013926007,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.21678189996683e-07,
+      "loss": 1.5383,
+      "step": 11915
+    },
+    {
+      "epoch": 3.353495568997046,
+      "grad_norm": 4.125,
+      "learning_rate": 6.19051224974557e-07,
+      "loss": 1.2658,
+      "step": 11920
+    },
+    {
+      "epoch": 3.354902236601491,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.164293563769618e-07,
+      "loss": 1.6482,
+      "step": 11925
+    },
+    {
+      "epoch": 3.356308904205936,
+      "grad_norm": 3.875,
+      "learning_rate": 6.138125881560912e-07,
+      "loss": 1.7493,
+      "step": 11930
+    },
+    {
+      "epoch": 3.357715571810381,
+      "grad_norm": 4.5,
+      "learning_rate": 6.112009242564444e-07,
+      "loss": 1.4005,
+      "step": 11935
+    },
+    {
+      "epoch": 3.3591222394148263,
+      "grad_norm": 3.0,
+      "learning_rate": 6.085943686148329e-07,
+      "loss": 1.6748,
+      "step": 11940
+    },
+    {
+      "epoch": 3.360528907019271,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.059929251603635e-07,
+      "loss": 1.4217,
+      "step": 11945
+    },
+    {
+      "epoch": 3.3619355746237165,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.033965978144393e-07,
+      "loss": 1.7179,
+      "step": 11950
+    },
+    {
+      "epoch": 3.3633422422281614,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.008053904907489e-07,
+      "loss": 1.5697,
+      "step": 11955
+    },
+    {
+      "epoch": 3.3647489098326067,
+      "grad_norm": 3.0,
+      "learning_rate": 5.982193070952677e-07,
+      "loss": 1.4438,
+      "step": 11960
+    },
+    {
+      "epoch": 3.3661555774370515,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.956383515262411e-07,
+      "loss": 1.6678,
+      "step": 11965
+    },
+    {
+      "epoch": 3.367562245041497,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.930625276741903e-07,
+      "loss": 1.66,
+      "step": 11970
+    },
+    {
+      "epoch": 3.3689689126459417,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.904918394218978e-07,
+      "loss": 1.4462,
+      "step": 11975
+    },
+    {
+      "epoch": 3.370375580250387,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.879262906444049e-07,
+      "loss": 1.6758,
+      "step": 11980
+    },
+    {
+      "epoch": 3.371782247854832,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.853658852090082e-07,
+      "loss": 1.4858,
+      "step": 11985
+    },
+    {
+      "epoch": 3.373188915459277,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.828106269752488e-07,
+      "loss": 1.3861,
+      "step": 11990
+    },
+    {
+      "epoch": 3.374595583063722,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.802605197949093e-07,
+      "loss": 1.5823,
+      "step": 11995
+    },
+    {
+      "epoch": 3.3760022506681673,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.777155675120071e-07,
+      "loss": 1.5136,
+      "step": 12000
+    },
+    {
+      "epoch": 3.377408918272612,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.751757739627931e-07,
+      "loss": 1.4484,
+      "step": 12005
+    },
+    {
+      "epoch": 3.3788155858770574,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.726411429757347e-07,
+      "loss": 1.6736,
+      "step": 12010
+    },
+    {
+      "epoch": 3.3802222534815023,
+      "grad_norm": 15.5625,
+      "learning_rate": 5.701116783715241e-07,
+      "loss": 1.399,
+      "step": 12015
+    },
+    {
+      "epoch": 3.381628921085947,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.675873839630627e-07,
+      "loss": 1.7158,
+      "step": 12020
+    },
+    {
+      "epoch": 3.3830355886903924,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.65068263555458e-07,
+      "loss": 1.6117,
+      "step": 12025
+    },
+    {
+      "epoch": 3.3844422562948377,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.625543209460186e-07,
+      "loss": 1.592,
+      "step": 12030
+    },
+    {
+      "epoch": 3.3858489238992826,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.60045559924251e-07,
+      "loss": 1.6539,
+      "step": 12035
+    },
+    {
+      "epoch": 3.3872555915037275,
+      "grad_norm": 2.75,
+      "learning_rate": 5.57541984271845e-07,
+      "loss": 1.5038,
+      "step": 12040
+    },
+    {
+      "epoch": 3.3886622591081728,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.550435977626797e-07,
+      "loss": 1.4934,
+      "step": 12045
+    },
+    {
+      "epoch": 3.390068926712618,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.525504041628095e-07,
+      "loss": 1.4347,
+      "step": 12050
+    },
+    {
+      "epoch": 3.391475594317063,
+      "grad_norm": 2.875,
+      "learning_rate": 5.5006240723046e-07,
+      "loss": 1.698,
+      "step": 12055
+    },
+    {
+      "epoch": 3.3928822619215078,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.475796107160273e-07,
+      "loss": 1.5576,
+      "step": 12060
+    },
+    {
+      "epoch": 3.394288929525953,
+      "grad_norm": 4.25,
+      "learning_rate": 5.451020183620642e-07,
+      "loss": 1.5214,
+      "step": 12065
+    },
+    {
+      "epoch": 3.395695597130398,
+      "grad_norm": 4.875,
+      "learning_rate": 5.426296339032812e-07,
+      "loss": 1.5787,
+      "step": 12070
+    },
+    {
+      "epoch": 3.3971022647348432,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.401624610665374e-07,
+      "loss": 1.5412,
+      "step": 12075
+    },
+    {
+      "epoch": 3.398508932339288,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.377005035708362e-07,
+      "loss": 1.5295,
+      "step": 12080
+    },
+    {
+      "epoch": 3.3999155999437334,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.352437651273183e-07,
+      "loss": 1.5275,
+      "step": 12085
+    },
+    {
+      "epoch": 3.4013222675481782,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.32792249439261e-07,
+      "loss": 1.542,
+      "step": 12090
+    },
+    {
+      "epoch": 3.4027289351526235,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.303459602020646e-07,
+      "loss": 1.3692,
+      "step": 12095
+    },
+    {
+      "epoch": 3.4041356027570684,
+      "grad_norm": 1.9296875,
+      "learning_rate": 5.279049011032533e-07,
+      "loss": 1.631,
+      "step": 12100
+    },
+    {
+      "epoch": 3.4055422703615137,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.254690758224663e-07,
+      "loss": 1.7059,
+      "step": 12105
+    },
+    {
+      "epoch": 3.4069489379659585,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.23038488031454e-07,
+      "loss": 1.3645,
+      "step": 12110
+    },
+    {
+      "epoch": 3.408355605570404,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.206131413940711e-07,
+      "loss": 1.4428,
+      "step": 12115
+    },
+    {
+      "epoch": 3.4097622731748487,
+      "grad_norm": 2.875,
+      "learning_rate": 5.181930395662744e-07,
+      "loss": 1.6279,
+      "step": 12120
+    },
+    {
+      "epoch": 3.411168940779294,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.157781861961115e-07,
+      "loss": 1.7865,
+      "step": 12125
+    },
+    {
+      "epoch": 3.412575608383739,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.133685849237191e-07,
+      "loss": 1.6429,
+      "step": 12130
+    },
+    {
+      "epoch": 3.413982275988184,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.109642393813201e-07,
+      "loss": 1.6585,
+      "step": 12135
+    },
+    {
+      "epoch": 3.415388943592629,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.085651531932087e-07,
+      "loss": 1.6252,
+      "step": 12140
+    },
+    {
+      "epoch": 3.4167956111970743,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.061713299757579e-07,
+      "loss": 1.647,
+      "step": 12145
+    },
+    {
+      "epoch": 3.418202278801519,
+      "grad_norm": 3.125,
+      "learning_rate": 5.037827733374031e-07,
+      "loss": 1.5837,
+      "step": 12150
+    },
+    {
+      "epoch": 3.4196089464059645,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.013994868786429e-07,
+      "loss": 1.3133,
+      "step": 12155
+    },
+    {
+      "epoch": 3.4210156140104093,
+      "grad_norm": 3.375,
+      "learning_rate": 4.990214741920287e-07,
+      "loss": 1.81,
+      "step": 12160
+    },
+    {
+      "epoch": 3.422422281614854,
+      "grad_norm": 2.875,
+      "learning_rate": 4.966487388621679e-07,
+      "loss": 1.3786,
+      "step": 12165
+    },
+    {
+      "epoch": 3.4238289492192995,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.942812844657061e-07,
+      "loss": 1.3051,
+      "step": 12170
+    },
+    {
+      "epoch": 3.4252356168237448,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.919191145713335e-07,
+      "loss": 1.7033,
+      "step": 12175
+    },
+    {
+      "epoch": 3.4266422844281896,
+      "grad_norm": 2.875,
+      "learning_rate": 4.895622327397722e-07,
+      "loss": 1.54,
+      "step": 12180
+    },
+    {
+      "epoch": 3.4280489520326345,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.872106425237734e-07,
+      "loss": 1.4154,
+      "step": 12185
+    },
+    {
+      "epoch": 3.42945561963708,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.848643474681115e-07,
+      "loss": 1.3742,
+      "step": 12190
+    },
+    {
+      "epoch": 3.4308622872415246,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.82523351109581e-07,
+      "loss": 1.5347,
+      "step": 12195
+    },
+    {
+      "epoch": 3.43226895484597,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.801876569769865e-07,
+      "loss": 1.6441,
+      "step": 12200
+    },
+    {
+      "epoch": 3.433675622450415,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.778572685911402e-07,
+      "loss": 1.4506,
+      "step": 12205
+    },
+    {
+      "epoch": 3.43508229005486,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7553218946486007e-07,
+      "loss": 1.4641,
+      "step": 12210
+    },
+    {
+      "epoch": 3.436488957659305,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.732124231029546e-07,
+      "loss": 1.6343,
+      "step": 12215
+    },
+    {
+      "epoch": 3.4378956252637503,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.708979730022307e-07,
+      "loss": 1.3519,
+      "step": 12220
+    },
+    {
+      "epoch": 3.439302292868195,
+      "grad_norm": 4.375,
+      "learning_rate": 4.6858884265147705e-07,
+      "loss": 1.6333,
+      "step": 12225
+    },
+    {
+      "epoch": 3.4407089604726404,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.662850355314649e-07,
+      "loss": 1.4561,
+      "step": 12230
+    },
+    {
+      "epoch": 3.4421156280770853,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.6398655511494e-07,
+      "loss": 1.569,
+      "step": 12235
+    },
+    {
+      "epoch": 3.4435222956815306,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.6169340486662234e-07,
+      "loss": 1.684,
+      "step": 12240
+    },
+    {
+      "epoch": 3.4449289632859754,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.594055882431913e-07,
+      "loss": 1.3547,
+      "step": 12245
+    },
+    {
+      "epoch": 3.4463356308904207,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.571231086932923e-07,
+      "loss": 1.6273,
+      "step": 12250
+    },
+    {
+      "epoch": 3.4477422984948656,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.54845969657522e-07,
+      "loss": 1.7027,
+      "step": 12255
+    },
+    {
+      "epoch": 3.449148966099311,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.525741745684284e-07,
+      "loss": 1.5842,
+      "step": 12260
+    },
+    {
+      "epoch": 3.4505556337037557,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.50307726850502e-07,
+      "loss": 1.6175,
+      "step": 12265
+    },
+    {
+      "epoch": 3.451962301308201,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.480466299201766e-07,
+      "loss": 1.4567,
+      "step": 12270
+    },
+    {
+      "epoch": 3.453368968912646,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.457908871858169e-07,
+      "loss": 1.4125,
+      "step": 12275
+    },
+    {
+      "epoch": 3.454775636517091,
+      "grad_norm": 3.25,
+      "learning_rate": 4.435405020477172e-07,
+      "loss": 1.5601,
+      "step": 12280
+    },
+    {
+      "epoch": 3.456182304121536,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.412954778980968e-07,
+      "loss": 1.6602,
+      "step": 12285
+    },
+    {
+      "epoch": 3.4575889717259813,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.390558181210928e-07,
+      "loss": 1.8053,
+      "step": 12290
+    },
+    {
+      "epoch": 3.458995639330426,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.368215260927588e-07,
+      "loss": 1.5461,
+      "step": 12295
+    },
+    {
+      "epoch": 3.4604023069348715,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.3459260518105134e-07,
+      "loss": 1.5155,
+      "step": 12300
+    },
+    {
+      "epoch": 3.4618089745393164,
+      "grad_norm": 4.5,
+      "learning_rate": 4.3236905874583704e-07,
+      "loss": 1.3204,
+      "step": 12305
+    },
+    {
+      "epoch": 3.463215642143761,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.3015089013887753e-07,
+      "loss": 1.4738,
+      "step": 12310
+    },
+    {
+      "epoch": 3.4646223097482065,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.279381027038278e-07,
+      "loss": 1.5011,
+      "step": 12315
+    },
+    {
+      "epoch": 3.466028977352652,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.257306997762322e-07,
+      "loss": 1.3831,
+      "step": 12320
+    },
+    {
+      "epoch": 3.4674356449570967,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.235286846835202e-07,
+      "loss": 1.5323,
+      "step": 12325
+    },
+    {
+      "epoch": 3.4688423125615415,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.2133206074499527e-07,
+      "loss": 1.5963,
+      "step": 12330
+    },
+    {
+      "epoch": 3.470248980165987,
+      "grad_norm": 3.875,
+      "learning_rate": 4.191408312718385e-07,
+      "loss": 1.6031,
+      "step": 12335
+    },
+    {
+      "epoch": 3.4716556477704317,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.169549995670971e-07,
+      "loss": 1.6792,
+      "step": 12340
+    },
+    {
+      "epoch": 3.473062315374877,
+      "grad_norm": 4.375,
+      "learning_rate": 4.147745689256821e-07,
+      "loss": 1.526,
+      "step": 12345
+    },
+    {
+      "epoch": 3.474468982979322,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.1259954263436426e-07,
+      "loss": 1.6437,
+      "step": 12350
+    },
+    {
+      "epoch": 3.475875650583767,
+      "grad_norm": 4.125,
+      "learning_rate": 4.104299239717668e-07,
+      "loss": 1.5831,
+      "step": 12355
+    },
+    {
+      "epoch": 3.477282318188212,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.082657162083607e-07,
+      "loss": 1.6606,
+      "step": 12360
+    },
+    {
+      "epoch": 3.4786889857926573,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0610692260646085e-07,
+      "loss": 1.6935,
+      "step": 12365
+    },
+    {
+      "epoch": 3.480095653397102,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.039535464202242e-07,
+      "loss": 1.6452,
+      "step": 12370
+    },
+    {
+      "epoch": 3.4815023210015474,
+      "grad_norm": 2.75,
+      "learning_rate": 4.018055908956355e-07,
+      "loss": 1.5527,
+      "step": 12375
+    },
+    {
+      "epoch": 3.4829089886059923,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.9966305927051416e-07,
+      "loss": 1.5381,
+      "step": 12380
+    },
+    {
+      "epoch": 3.4843156562104376,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.975259547744998e-07,
+      "loss": 1.6809,
+      "step": 12385
+    },
+    {
+      "epoch": 3.4857223238148825,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.953942806290533e-07,
+      "loss": 1.3621,
+      "step": 12390
+    },
+    {
+      "epoch": 3.4871289914193278,
+      "grad_norm": 4.4375,
+      "learning_rate": 3.9326804004744794e-07,
+      "loss": 1.4407,
+      "step": 12395
+    },
+    {
+      "epoch": 3.4885356590237726,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.911472362347701e-07,
+      "loss": 1.7222,
+      "step": 12400
+    },
+    {
+      "epoch": 3.489942326628218,
+      "grad_norm": 4.375,
+      "learning_rate": 3.8903187238790514e-07,
+      "loss": 1.7223,
+      "step": 12405
+    },
+    {
+      "epoch": 3.4913489942326628,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.869219516955442e-07,
+      "loss": 1.7887,
+      "step": 12410
+    },
+    {
+      "epoch": 3.492755661837108,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.8481747733816984e-07,
+      "loss": 1.5084,
+      "step": 12415
+    },
+    {
+      "epoch": 3.494162329441553,
+      "grad_norm": 3.25,
+      "learning_rate": 3.827184524880542e-07,
+      "loss": 1.6298,
+      "step": 12420
+    },
+    {
+      "epoch": 3.495568997045998,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.8062488030925887e-07,
+      "loss": 1.5018,
+      "step": 12425
+    },
+    {
+      "epoch": 3.496975664650443,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.785367639576225e-07,
+      "loss": 1.8452,
+      "step": 12430
+    },
+    {
+      "epoch": 3.498382332254888,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.764541065807609e-07,
+      "loss": 1.562,
+      "step": 12435
+    },
+    {
+      "epoch": 3.4997889998593332,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.7437691131806083e-07,
+      "loss": 1.6698,
+      "step": 12440
+    },
+    {
+      "epoch": 3.5011956674637785,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.723051813006752e-07,
+      "loss": 1.2999,
+      "step": 12445
+    },
+    {
+      "epoch": 3.5026023350682234,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.7023891965151853e-07,
+      "loss": 1.5353,
+      "step": 12450
+    },
+    {
+      "epoch": 3.5040090026726682,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.6817812948526506e-07,
+      "loss": 1.4993,
+      "step": 12455
+    },
+    {
+      "epoch": 3.5054156702771135,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.66122813908337e-07,
+      "loss": 1.4794,
+      "step": 12460
+    },
+    {
+      "epoch": 3.506822337881559,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.6407297601890763e-07,
+      "loss": 1.6521,
+      "step": 12465
+    },
+    {
+      "epoch": 3.5082290054860037,
+      "grad_norm": 4.625,
+      "learning_rate": 3.6202861890689105e-07,
+      "loss": 1.4697,
+      "step": 12470
+    },
+    {
+      "epoch": 3.5096356730904485,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.599897456539409e-07,
+      "loss": 1.6966,
+      "step": 12475
+    },
+    {
+      "epoch": 3.511042340694894,
+      "grad_norm": 2.75,
+      "learning_rate": 3.5795635933344313e-07,
+      "loss": 1.6049,
+      "step": 12480
+    },
+    {
+      "epoch": 3.5124490082993387,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.5592846301051525e-07,
+      "loss": 1.5986,
+      "step": 12485
+    },
+    {
+      "epoch": 3.513855675903784,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.5390605974199697e-07,
+      "loss": 1.4704,
+      "step": 12490
+    },
+    {
+      "epoch": 3.515262343508229,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.518891525764474e-07,
+      "loss": 1.5354,
+      "step": 12495
+    },
+    {
+      "epoch": 3.516669011112674,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.4987774455414434e-07,
+      "loss": 1.3862,
+      "step": 12500
+    },
+    {
+      "epoch": 3.518075678717119,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.478718387070705e-07,
+      "loss": 1.7165,
+      "step": 12505
+    },
+    {
+      "epoch": 3.5194823463215643,
+      "grad_norm": 5.125,
+      "learning_rate": 3.458714380589205e-07,
+      "loss": 1.5756,
+      "step": 12510
+    },
+    {
+      "epoch": 3.520889013926009,
+      "grad_norm": 4.6875,
+      "learning_rate": 3.438765456250867e-07,
+      "loss": 1.5997,
+      "step": 12515
+    },
+    {
+      "epoch": 3.5222956815304545,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.418871644126593e-07,
+      "loss": 1.5862,
+      "step": 12520
+    },
+    {
+      "epoch": 3.5237023491348993,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.399032974204212e-07,
+      "loss": 1.647,
+      "step": 12525
+    },
+    {
+      "epoch": 3.5251090167393446,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.3792494763884527e-07,
+      "loss": 1.6057,
+      "step": 12530
+    },
+    {
+      "epoch": 3.5265156843437895,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.3595211805008193e-07,
+      "loss": 1.5112,
+      "step": 12535
+    },
+    {
+      "epoch": 3.527922351948235,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.339848116279671e-07,
+      "loss": 1.3215,
+      "step": 12540
+    },
+    {
+      "epoch": 3.5293290195526796,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.3202303133800724e-07,
+      "loss": 1.5533,
+      "step": 12545
+    },
+    {
+      "epoch": 3.530735687157125,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.300667801373791e-07,
+      "loss": 1.6153,
+      "step": 12550
+    },
+    {
+      "epoch": 3.53214235476157,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.281160609749265e-07,
+      "loss": 1.5961,
+      "step": 12555
+    },
+    {
+      "epoch": 3.5335490223660146,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.261708767911533e-07,
+      "loss": 1.4118,
+      "step": 12560
+    },
+    {
+      "epoch": 3.53495568997046,
+      "grad_norm": 2.5,
+      "learning_rate": 3.242312305182193e-07,
+      "loss": 1.4728,
+      "step": 12565
+    },
+    {
+      "epoch": 3.5363623575749052,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.222971250799373e-07,
+      "loss": 1.6069,
+      "step": 12570
+    },
+    {
+      "epoch": 3.53776902517935,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.2036856339176897e-07,
+      "loss": 1.6743,
+      "step": 12575
+    },
+    {
+      "epoch": 3.539175692783795,
+      "grad_norm": 2.75,
+      "learning_rate": 3.1844554836081596e-07,
+      "loss": 1.7418,
+      "step": 12580
+    },
+    {
+      "epoch": 3.5405823603882403,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.16528082885823e-07,
+      "loss": 1.7357,
+      "step": 12585
+    },
+    {
+      "epoch": 3.5419890279926856,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.1461616985716655e-07,
+      "loss": 1.7516,
+      "step": 12590
+    },
+    {
+      "epoch": 3.5433956955971304,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.12709812156855e-07,
+      "loss": 1.7343,
+      "step": 12595
+    },
+    {
+      "epoch": 3.5448023632015753,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.1080901265852034e-07,
+      "loss": 1.5757,
+      "step": 12600
+    },
+    {
+      "epoch": 3.5462090308060206,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.0891377422742084e-07,
+      "loss": 1.693,
+      "step": 12605
+    },
+    {
+      "epoch": 3.547615698410466,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.070240997204254e-07,
+      "loss": 1.5975,
+      "step": 12610
+    },
+    {
+      "epoch": 3.5490223660149107,
+      "grad_norm": 4.625,
+      "learning_rate": 3.051399919860222e-07,
+      "loss": 1.6121,
+      "step": 12615
+    },
+    {
+      "epoch": 3.5504290336193556,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.0326145386430433e-07,
+      "loss": 1.4582,
+      "step": 12620
+    },
+    {
+      "epoch": 3.551835701223801,
+      "grad_norm": 4.25,
+      "learning_rate": 3.013884881869702e-07,
+      "loss": 1.5848,
+      "step": 12625
+    },
+    {
+      "epoch": 3.5532423688282457,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9952109777731947e-07,
+      "loss": 1.6932,
+      "step": 12630
+    },
+    {
+      "epoch": 3.554649036432691,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.97659285450246e-07,
+      "loss": 1.7294,
+      "step": 12635
+    },
+    {
+      "epoch": 3.556055704037136,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.958030540122358e-07,
+      "loss": 1.7316,
+      "step": 12640
+    },
+    {
+      "epoch": 3.557462371641581,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.9395240626136276e-07,
+      "loss": 1.7024,
+      "step": 12645
+    },
+    {
+      "epoch": 3.558869039246026,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9210734498728375e-07,
+      "loss": 1.6179,
+      "step": 12650
+    },
+    {
+      "epoch": 3.5602757068504713,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.902678729712336e-07,
+      "loss": 1.6973,
+      "step": 12655
+    },
+    {
+      "epoch": 3.561682374454916,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8843399298602446e-07,
+      "loss": 1.3544,
+      "step": 12660
+    },
+    {
+      "epoch": 3.5630890420593615,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.866057077960353e-07,
+      "loss": 1.7027,
+      "step": 12665
+    },
+    {
+      "epoch": 3.5644957096638064,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.847830201572159e-07,
+      "loss": 1.9461,
+      "step": 12670
+    },
+    {
+      "epoch": 3.5659023772682517,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.8296593281707457e-07,
+      "loss": 1.4954,
+      "step": 12675
+    },
+    {
+      "epoch": 3.5673090448726965,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.811544485146804e-07,
+      "loss": 1.6632,
+      "step": 12680
+    },
+    {
+      "epoch": 3.568715712477142,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.793485699806535e-07,
+      "loss": 1.5274,
+      "step": 12685
+    },
+    {
+      "epoch": 3.5701223800815867,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.7754829993716876e-07,
+      "loss": 1.5179,
+      "step": 12690
+    },
+    {
+      "epoch": 3.571529047686032,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.757536410979404e-07,
+      "loss": 1.404,
+      "step": 12695
+    },
+    {
+      "epoch": 3.572935715290477,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.7396459616822974e-07,
+      "loss": 1.4433,
+      "step": 12700
+    },
+    {
+      "epoch": 3.5743423828949217,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.721811678448347e-07,
+      "loss": 1.7107,
+      "step": 12705
+    },
+    {
+      "epoch": 3.575749050499367,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.70403358816083e-07,
+      "loss": 1.3703,
+      "step": 12710
+    },
+    {
+      "epoch": 3.5771557181038123,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.6863117176183727e-07,
+      "loss": 1.4843,
+      "step": 12715
+    },
+    {
+      "epoch": 3.578562385708257,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.6686460935348187e-07,
+      "loss": 1.2866,
+      "step": 12720
+    },
+    {
+      "epoch": 3.579969053312702,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.651036742539241e-07,
+      "loss": 1.511,
+      "step": 12725
+    },
+    {
+      "epoch": 3.5813757209171473,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.633483691175877e-07,
+      "loss": 1.5051,
+      "step": 12730
+    },
+    {
+      "epoch": 3.5827823885215926,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6159869659041176e-07,
+      "loss": 1.5787,
+      "step": 12735
+    },
+    {
+      "epoch": 3.5841890561260374,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.5985465930984163e-07,
+      "loss": 1.4837,
+      "step": 12740
+    },
+    {
+      "epoch": 3.5855957237304823,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.5811625990483164e-07,
+      "loss": 1.4975,
+      "step": 12745
+    },
+    {
+      "epoch": 3.5870023913349276,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.563835009958355e-07,
+      "loss": 1.624,
+      "step": 12750
+    },
+    {
+      "epoch": 3.588409058939373,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.546563851948047e-07,
+      "loss": 1.5078,
+      "step": 12755
+    },
+    {
+      "epoch": 3.5898157265438178,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.5293491510518425e-07,
+      "loss": 1.5151,
+      "step": 12760
+    },
+    {
+      "epoch": 3.5912223941482626,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.5121909332191047e-07,
+      "loss": 1.4745,
+      "step": 12765
+    },
+    {
+      "epoch": 3.592629061752708,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.4950892243140285e-07,
+      "loss": 1.6577,
+      "step": 12770
+    },
+    {
+      "epoch": 3.5940357293571528,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.478044050115646e-07,
+      "loss": 1.5741,
+      "step": 12775
+    },
+    {
+      "epoch": 3.595442396961598,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.4610554363177647e-07,
+      "loss": 1.4498,
+      "step": 12780
+    },
+    {
+      "epoch": 3.596849064566043,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.4441234085289264e-07,
+      "loss": 1.7303,
+      "step": 12785
+    },
+    {
+      "epoch": 3.598255732170488,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4272479922723897e-07,
+      "loss": 1.6942,
+      "step": 12790
+    },
+    {
+      "epoch": 3.599662399774933,
+      "grad_norm": 3.0,
+      "learning_rate": 2.410429212986065e-07,
+      "loss": 1.4784,
+      "step": 12795
+    },
+    {
+      "epoch": 3.6010690673793784,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.3936670960224935e-07,
+      "loss": 1.6887,
+      "step": 12800
+    },
+    {
+      "epoch": 3.6024757349838232,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.3769616666488024e-07,
+      "loss": 1.6325,
+      "step": 12805
+    },
+    {
+      "epoch": 3.6038824025882685,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.36031295004667e-07,
+      "loss": 1.6284,
+      "step": 12810
+    },
+    {
+      "epoch": 3.6052890701927134,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.3437209713122707e-07,
+      "loss": 1.458,
+      "step": 12815
+    },
+    {
+      "epoch": 3.6066957377971587,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.3271857554562914e-07,
+      "loss": 1.4742,
+      "step": 12820
+    },
+    {
+      "epoch": 3.6081024054016035,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.3107073274038157e-07,
+      "loss": 1.5134,
+      "step": 12825
+    },
+    {
+      "epoch": 3.6095090730060484,
+      "grad_norm": 3.375,
+      "learning_rate": 2.2942857119943392e-07,
+      "loss": 1.6525,
+      "step": 12830
+    },
+    {
+      "epoch": 3.6109157406104937,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.277920933981723e-07,
+      "loss": 1.8801,
+      "step": 12835
+    },
+    {
+      "epoch": 3.612322408214939,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.2616130180341408e-07,
+      "loss": 1.4974,
+      "step": 12840
+    },
+    {
+      "epoch": 3.613729075819384,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.245361988734076e-07,
+      "loss": 1.6507,
+      "step": 12845
+    },
+    {
+      "epoch": 3.6151357434238287,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.2291678705782303e-07,
+      "loss": 1.811,
+      "step": 12850
+    },
+    {
+      "epoch": 3.616542411028274,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.2130306879775396e-07,
+      "loss": 1.6014,
+      "step": 12855
+    },
+    {
+      "epoch": 3.6179490786327193,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.1969504652571014e-07,
+      "loss": 1.7031,
+      "step": 12860
+    },
+    {
+      "epoch": 3.619355746237164,
+      "grad_norm": 2.625,
+      "learning_rate": 2.1809272266561796e-07,
+      "loss": 1.7339,
+      "step": 12865
+    },
+    {
+      "epoch": 3.620762413841609,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.1649609963280892e-07,
+      "loss": 1.6556,
+      "step": 12870
+    },
+    {
+      "epoch": 3.6221690814460543,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.1490517983402667e-07,
+      "loss": 1.6822,
+      "step": 12875
+    },
+    {
+      "epoch": 3.6235757490504996,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.1331996566741473e-07,
+      "loss": 1.4878,
+      "step": 12880
+    },
+    {
+      "epoch": 3.6249824166549445,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.1174045952251674e-07,
+      "loss": 1.7963,
+      "step": 12885
+    },
+    {
+      "epoch": 3.6263890842593893,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.1016666378027127e-07,
+      "loss": 1.6533,
+      "step": 12890
+    },
+    {
+      "epoch": 3.6277957518638346,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.085985808130113e-07,
+      "loss": 1.6496,
+      "step": 12895
+    },
+    {
+      "epoch": 3.6292024194682795,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.070362129844554e-07,
+      "loss": 1.4314,
+      "step": 12900
+    },
+    {
+      "epoch": 3.630609087072725,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.0547956264970946e-07,
+      "loss": 1.3795,
+      "step": 12905
+    },
+    {
+      "epoch": 3.6320157546771696,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0392863215525957e-07,
+      "loss": 1.4565,
+      "step": 12910
+    },
+    {
+      "epoch": 3.633422422281615,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0238342383897032e-07,
+      "loss": 1.4031,
+      "step": 12915
+    },
+    {
+      "epoch": 3.63482908988606,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.0084394003008165e-07,
+      "loss": 1.5884,
+      "step": 12920
+    },
+    {
+      "epoch": 3.636235757490505,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.9931018304920256e-07,
+      "loss": 1.7368,
+      "step": 12925
+    },
+    {
+      "epoch": 3.63764242509495,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.9778215520831076e-07,
+      "loss": 1.5598,
+      "step": 12930
+    },
+    {
+      "epoch": 3.6390490926993952,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9625985881074603e-07,
+      "loss": 1.4654,
+      "step": 12935
+    },
+    {
+      "epoch": 3.64045576030384,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.9474329615121232e-07,
+      "loss": 1.6227,
+      "step": 12940
+    },
+    {
+      "epoch": 3.6418624279082854,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.9323246951576633e-07,
+      "loss": 1.4179,
+      "step": 12945
+    },
+    {
+      "epoch": 3.6432690955127303,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.9172738118182098e-07,
+      "loss": 1.417,
+      "step": 12950
+    },
+    {
+      "epoch": 3.6446757631171756,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.902280334181392e-07,
+      "loss": 1.65,
+      "step": 12955
+    },
+    {
+      "epoch": 3.6460824307216204,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.8873442848482868e-07,
+      "loss": 1.5986,
+      "step": 12960
+    },
+    {
+      "epoch": 3.6474890983260657,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.872465686333422e-07,
+      "loss": 1.6104,
+      "step": 12965
+    },
+    {
+      "epoch": 3.6488957659305106,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.857644561064733e-07,
+      "loss": 1.5107,
+      "step": 12970
+    },
+    {
+      "epoch": 3.6503024335349554,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.842880931383477e-07,
+      "loss": 1.5454,
+      "step": 12975
+    },
+    {
+      "epoch": 3.6517091011394007,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8281748195443015e-07,
+      "loss": 1.416,
+      "step": 12980
+    },
+    {
+      "epoch": 3.653115768743846,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.8135262477151092e-07,
+      "loss": 1.6158,
+      "step": 12985
+    },
+    {
+      "epoch": 3.654522436348291,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7989352379770773e-07,
+      "loss": 1.5509,
+      "step": 12990
+    },
+    {
+      "epoch": 3.6559291039527357,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7844018123246295e-07,
+      "loss": 1.493,
+      "step": 12995
+    },
+    {
+      "epoch": 3.657335771557181,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.7699259926653665e-07,
+      "loss": 1.7625,
+      "step": 13000
+    },
+    {
+      "epoch": 3.6587424391616263,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.7555078008200685e-07,
+      "loss": 1.3462,
+      "step": 13005
+    },
+    {
+      "epoch": 3.660149106766071,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.741147258522635e-07,
+      "loss": 1.5288,
+      "step": 13010
+    },
+    {
+      "epoch": 3.661555774370516,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.7268443874200834e-07,
+      "loss": 1.7067,
+      "step": 13015
+    },
+    {
+      "epoch": 3.6629624419749613,
+      "grad_norm": 2.75,
+      "learning_rate": 1.712599209072474e-07,
+      "loss": 1.7385,
+      "step": 13020
+    },
+    {
+      "epoch": 3.6643691095794066,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.6984117449529324e-07,
+      "loss": 1.5687,
+      "step": 13025
+    },
+    {
+      "epoch": 3.6657757771838515,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.684282016447547e-07,
+      "loss": 1.7137,
+      "step": 13030
+    },
+    {
+      "epoch": 3.6671824447882964,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.670210044855409e-07,
+      "loss": 1.5094,
+      "step": 13035
+    },
+    {
+      "epoch": 3.6685891123927417,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.6561958513885332e-07,
+      "loss": 1.7145,
+      "step": 13040
+    },
+    {
+      "epoch": 3.6699957799971865,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.6422394571718435e-07,
+      "loss": 1.4869,
+      "step": 13045
+    },
+    {
+      "epoch": 3.671402447601632,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.628340883243129e-07,
+      "loss": 1.6588,
+      "step": 13050
+    },
+    {
+      "epoch": 3.6728091152060767,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.6145001505530353e-07,
+      "loss": 1.6234,
+      "step": 13055
+    },
+    {
+      "epoch": 3.674215782810522,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.6007172799650027e-07,
+      "loss": 1.4941,
+      "step": 13060
+    },
+    {
+      "epoch": 3.675622450414967,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.5869922922552649e-07,
+      "loss": 1.6677,
+      "step": 13065
+    },
+    {
+      "epoch": 3.677029118019412,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.573325208112801e-07,
+      "loss": 1.5113,
+      "step": 13070
+    },
+    {
+      "epoch": 3.678435785623857,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.5597160481392834e-07,
+      "loss": 1.4624,
+      "step": 13075
+    },
+    {
+      "epoch": 3.6798424532283023,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.5461648328491106e-07,
+      "loss": 1.333,
+      "step": 13080
+    },
+    {
+      "epoch": 3.681249120832747,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.5326715826693027e-07,
+      "loss": 1.5943,
+      "step": 13085
+    },
+    {
+      "epoch": 3.6826557884371924,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.519236317939514e-07,
+      "loss": 1.5913,
+      "step": 13090
+    },
+    {
+      "epoch": 3.6840624560416373,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.5058590589119936e-07,
+      "loss": 1.59,
+      "step": 13095
+    },
+    {
+      "epoch": 3.6854691236460826,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.492539825751562e-07,
+      "loss": 1.6069,
+      "step": 13100
+    },
+    {
+      "epoch": 3.6868757912505274,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4792786385355415e-07,
+      "loss": 1.8708,
+      "step": 13105
+    },
+    {
+      "epoch": 3.6882824588549727,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.4660755172537953e-07,
+      "loss": 1.4281,
+      "step": 13110
+    },
+    {
+      "epoch": 3.6896891264594176,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4529304818086297e-07,
+      "loss": 1.8313,
+      "step": 13115
+    },
+    {
+      "epoch": 3.6910957940638625,
+      "grad_norm": 3.375,
+      "learning_rate": 1.4398435520147988e-07,
+      "loss": 1.4211,
+      "step": 13120
+    },
+    {
+      "epoch": 3.6925024616683078,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.426814747599483e-07,
+      "loss": 1.5624,
+      "step": 13125
+    },
+    {
+      "epoch": 3.693909129272753,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4138440882022297e-07,
+      "loss": 1.5928,
+      "step": 13130
+    },
+    {
+      "epoch": 3.695315796877198,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.4009315933749411e-07,
+      "loss": 1.6827,
+      "step": 13135
+    },
+    {
+      "epoch": 3.6967224644816428,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.388077282581852e-07,
+      "loss": 1.6348,
+      "step": 13140
+    },
+    {
+      "epoch": 3.698129132086088,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.375281175199472e-07,
+      "loss": 1.5618,
+      "step": 13145
+    },
+    {
+      "epoch": 3.6995357996905334,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.362543290516589e-07,
+      "loss": 1.621,
+      "step": 13150
+    },
+    {
+      "epoch": 3.700942467294978,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.3498636477342307e-07,
+      "loss": 1.6355,
+      "step": 13155
+    },
+    {
+      "epoch": 3.702349134899423,
+      "grad_norm": 3.375,
+      "learning_rate": 1.337242265965619e-07,
+      "loss": 1.584,
+      "step": 13160
+    },
+    {
+      "epoch": 3.7037558025038684,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3246791642361622e-07,
+      "loss": 1.5461,
+      "step": 13165
+    },
+    {
+      "epoch": 3.7051624701083137,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.3121743614834135e-07,
+      "loss": 1.7194,
+      "step": 13170
+    },
+    {
+      "epoch": 3.7065691377127585,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.2997278765570463e-07,
+      "loss": 1.1595,
+      "step": 13175
+    },
+    {
+      "epoch": 3.7079758053172034,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.2873397282188215e-07,
+      "loss": 1.7544,
+      "step": 13180
+    },
+    {
+      "epoch": 3.7093824729216487,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2750099351425792e-07,
+      "loss": 1.6543,
+      "step": 13185
+    },
+    {
+      "epoch": 3.7107891405260935,
+      "grad_norm": 6.375,
+      "learning_rate": 1.2627385159141812e-07,
+      "loss": 1.4372,
+      "step": 13190
+    },
+    {
+      "epoch": 3.712195808130539,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.250525489031493e-07,
+      "loss": 1.525,
+      "step": 13195
+    },
+    {
+      "epoch": 3.7136024757349837,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.2383708729043886e-07,
+      "loss": 1.7147,
+      "step": 13200
+    },
+    {
+      "epoch": 3.715009143339429,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2262746858546468e-07,
+      "loss": 1.4541,
+      "step": 13205
+    },
+    {
+      "epoch": 3.716415810943874,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.214236946116012e-07,
+      "loss": 1.696,
+      "step": 13210
+    },
+    {
+      "epoch": 3.717822478548319,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2022576718341104e-07,
+      "loss": 1.614,
+      "step": 13215
+    },
+    {
+      "epoch": 3.719229146152764,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.1903368810664315e-07,
+      "loss": 1.7279,
+      "step": 13220
+    },
+    {
+      "epoch": 3.7206358137572093,
+      "grad_norm": 6.03125,
+      "learning_rate": 1.1784745917823169e-07,
+      "loss": 1.4992,
+      "step": 13225
+    },
+    {
+      "epoch": 3.722042481361654,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.1666708218629206e-07,
+      "loss": 1.4643,
+      "step": 13230
+    },
+    {
+      "epoch": 3.7234491489660995,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1549255891011788e-07,
+      "loss": 1.3531,
+      "step": 13235
+    },
+    {
+      "epoch": 3.7248558165705443,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1432389112017959e-07,
+      "loss": 1.5313,
+      "step": 13240
+    },
+    {
+      "epoch": 3.726262484174989,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.1316108057812135e-07,
+      "loss": 1.5671,
+      "step": 13245
+    },
+    {
+      "epoch": 3.7276691517794345,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1200412903675749e-07,
+      "loss": 1.5695,
+      "step": 13250
+    },
+    {
+      "epoch": 3.7290758193838798,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.1085303824006986e-07,
+      "loss": 1.505,
+      "step": 13255
+    },
+    {
+      "epoch": 3.7304824869883246,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.0970780992320871e-07,
+      "loss": 1.4891,
+      "step": 13260
+    },
+    {
+      "epoch": 3.7318891545927695,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0856844581248292e-07,
+      "loss": 1.6268,
+      "step": 13265
+    },
+    {
+      "epoch": 3.733295822197215,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0743494762536486e-07,
+      "loss": 1.4452,
+      "step": 13270
+    },
+    {
+      "epoch": 3.73470248980166,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0630731707048513e-07,
+      "loss": 1.5661,
+      "step": 13275
+    },
+    {
+      "epoch": 3.736109157406105,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0518555584762578e-07,
+      "loss": 1.5987,
+      "step": 13280
+    },
+    {
+      "epoch": 3.73751582501055,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0406966564772578e-07,
+      "loss": 1.6404,
+      "step": 13285
+    },
+    {
+      "epoch": 3.738922492614995,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.0295964815287117e-07,
+      "loss": 1.5056,
+      "step": 13290
+    },
+    {
+      "epoch": 3.7403291602194404,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.0185550503629725e-07,
+      "loss": 1.595,
+      "step": 13295
+    },
+    {
+      "epoch": 3.7417358278238853,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.0075723796238244e-07,
+      "loss": 1.3048,
+      "step": 13300
+    },
+    {
+      "epoch": 3.74314249542833,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.966484858665003e-08,
+      "loss": 1.5691,
+      "step": 13305
+    },
+    {
+      "epoch": 3.7445491630327754,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.857833855576103e-08,
+      "loss": 1.504,
+      "step": 13310
+    },
+    {
+      "epoch": 3.7459558306372203,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.749770950751601e-08,
+      "loss": 1.6968,
+      "step": 13315
+    },
+    {
+      "epoch": 3.7473624982416656,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.642296307084885e-08,
+      "loss": 1.7236,
+      "step": 13320
+    },
+    {
+      "epoch": 3.7487691658461104,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.535410086582718e-08,
+      "loss": 1.3874,
+      "step": 13325
+    },
+    {
+      "epoch": 3.7501758334505557,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.429112450364707e-08,
+      "loss": 1.319,
+      "step": 13330
+    },
+    {
+      "epoch": 3.7515825010550006,
+      "grad_norm": 3.0,
+      "learning_rate": 9.323403558663523e-08,
+      "loss": 1.6897,
+      "step": 13335
+    },
+    {
+      "epoch": 3.752989168659446,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.218283570824149e-08,
+      "loss": 1.5377,
+      "step": 13340
+    },
+    {
+      "epoch": 3.7543958362638907,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.113752645303829e-08,
+      "loss": 1.4597,
+      "step": 13345
+    },
+    {
+      "epoch": 3.755802503868336,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.009810939671991e-08,
+      "loss": 1.3243,
+      "step": 13350
+    },
+    {
+      "epoch": 3.757209171472781,
+      "grad_norm": 3.75,
+      "learning_rate": 8.906458610609791e-08,
+      "loss": 1.6396,
+      "step": 13355
+    },
+    {
+      "epoch": 3.758615839077226,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.803695813910072e-08,
+      "loss": 1.7091,
+      "step": 13360
+    },
+    {
+      "epoch": 3.760022506681671,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.701522704476838e-08,
+      "loss": 1.6138,
+      "step": 13365
+    },
+    {
+      "epoch": 3.7614291742861163,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.599939436325376e-08,
+      "loss": 1.5434,
+      "step": 13370
+    },
+    {
+      "epoch": 3.762835841890561,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.498946162581732e-08,
+      "loss": 1.6085,
+      "step": 13375
+    },
+    {
+      "epoch": 3.7642425094950065,
+      "grad_norm": 3.75,
+      "learning_rate": 8.39854303548262e-08,
+      "loss": 1.4269,
+      "step": 13380
+    },
+    {
+      "epoch": 3.7656491770994514,
+      "grad_norm": 4.40625,
+      "learning_rate": 8.298730206375237e-08,
+      "loss": 1.65,
+      "step": 13385
+    },
+    {
+      "epoch": 3.767055844703896,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.199507825716923e-08,
+      "loss": 1.6611,
+      "step": 13390
+    },
+    {
+      "epoch": 3.7684625123083415,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.100876043074878e-08,
+      "loss": 1.6301,
+      "step": 13395
+    },
+    {
+      "epoch": 3.769869179912787,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.002835007126263e-08,
+      "loss": 1.7159,
+      "step": 13400
+    },
+    {
+      "epoch": 3.7712758475172317,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.905384865657572e-08,
+      "loss": 1.6362,
+      "step": 13405
+    },
+    {
+      "epoch": 3.7726825151216765,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.808525765564634e-08,
+      "loss": 1.5149,
+      "step": 13410
+    },
+    {
+      "epoch": 3.774089182726122,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.712257852852344e-08,
+      "loss": 1.7875,
+      "step": 13415
+    },
+    {
+      "epoch": 3.775495850330567,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.616581272634493e-08,
+      "loss": 1.5831,
+      "step": 13420
+    },
+    {
+      "epoch": 3.776902517935012,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.521496169133445e-08,
+      "loss": 1.75,
+      "step": 13425
+    },
+    {
+      "epoch": 3.778309185539457,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.427002685679884e-08,
+      "loss": 1.7126,
+      "step": 13430
+    },
+    {
+      "epoch": 3.779715853143902,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.33310096471298e-08,
+      "loss": 1.453,
+      "step": 13435
+    },
+    {
+      "epoch": 3.7811225207483474,
+      "grad_norm": 5.15625,
+      "learning_rate": 7.23979114777955e-08,
+      "loss": 1.6371,
+      "step": 13440
+    },
+    {
+      "epoch": 3.7825291883527923,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.147073375534374e-08,
+      "loss": 1.7193,
+      "step": 13445
+    },
+    {
+      "epoch": 3.783935855957237,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.054947787739785e-08,
+      "loss": 1.4624,
+      "step": 13450
+    },
+    {
+      "epoch": 3.7853425235616824,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.963414523265321e-08,
+      "loss": 1.6463,
+      "step": 13455
+    },
+    {
+      "epoch": 3.7867491911661273,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.872473720087768e-08,
+      "loss": 1.6806,
+      "step": 13460
+    },
+    {
+      "epoch": 3.7881558587705726,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.782125515290937e-08,
+      "loss": 1.5029,
+      "step": 13465
+    },
+    {
+      "epoch": 3.7895625263750174,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.692370045065043e-08,
+      "loss": 1.4976,
+      "step": 13470
+    },
+    {
+      "epoch": 3.7909691939794627,
+      "grad_norm": 3.0,
+      "learning_rate": 6.603207444707149e-08,
+      "loss": 1.4385,
+      "step": 13475
+    },
+    {
+      "epoch": 3.7923758615839076,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.514637848620497e-08,
+      "loss": 1.4465,
+      "step": 13480
+    },
+    {
+      "epoch": 3.793782529188353,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.426661390314336e-08,
+      "loss": 1.6002,
+      "step": 13485
+    },
+    {
+      "epoch": 3.7951891967927978,
+      "grad_norm": 5.125,
+      "learning_rate": 6.339278202404009e-08,
+      "loss": 1.43,
+      "step": 13490
+    },
+    {
+      "epoch": 3.796595864397243,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.252488416610458e-08,
+      "loss": 1.6215,
+      "step": 13495
+    },
+    {
+      "epoch": 3.798002532001688,
+      "grad_norm": 3.25,
+      "learning_rate": 6.166292163760145e-08,
+      "loss": 1.7022,
+      "step": 13500
+    },
+    {
+      "epoch": 3.799409199606133,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.080689573784826e-08,
+      "loss": 1.4709,
+      "step": 13505
+    },
+    {
+      "epoch": 3.800815867210578,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.995680775721457e-08,
+      "loss": 1.5759,
+      "step": 13510
+    },
+    {
+      "epoch": 3.8022225348150234,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.911265897711759e-08,
+      "loss": 1.6665,
+      "step": 13515
+    },
+    {
+      "epoch": 3.8036292024194682,
+      "grad_norm": 4.5,
+      "learning_rate": 5.82744506700239e-08,
+      "loss": 1.7009,
+      "step": 13520
+    },
+    {
+      "epoch": 3.8050358700239135,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.744218409944412e-08,
+      "loss": 1.5255,
+      "step": 13525
+    },
+    {
+      "epoch": 3.8064425376283584,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.6615860519932054e-08,
+      "loss": 1.6446,
+      "step": 13530
+    },
+    {
+      "epoch": 3.8078492052328032,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.5795481177083324e-08,
+      "loss": 1.5484,
+      "step": 13535
+    },
+    {
+      "epoch": 3.8092558728372485,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.49810473075345e-08,
+      "loss": 1.6404,
+      "step": 13540
+    },
+    {
+      "epoch": 3.810662540441694,
+      "grad_norm": 5.59375,
+      "learning_rate": 5.417256013895777e-08,
+      "loss": 1.5426,
+      "step": 13545
+    },
+    {
+      "epoch": 3.8120692080461387,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.337002089006315e-08,
+      "loss": 1.681,
+      "step": 13550
+    },
+    {
+      "epoch": 3.8134758756505835,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.2573430770594505e-08,
+      "loss": 1.7098,
+      "step": 13555
+    },
+    {
+      "epoch": 3.814882543255029,
+      "grad_norm": 7.40625,
+      "learning_rate": 5.178279098132643e-08,
+      "loss": 1.7041,
+      "step": 13560
+    },
+    {
+      "epoch": 3.816289210859474,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.099810271406646e-08,
+      "loss": 1.6355,
+      "step": 13565
+    },
+    {
+      "epoch": 3.817695878463919,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.021936715164843e-08,
+      "loss": 1.6715,
+      "step": 13570
+    },
+    {
+      "epoch": 3.819102546068364,
+      "grad_norm": 6.0,
+      "learning_rate": 4.9446585467935566e-08,
+      "loss": 1.4218,
+      "step": 13575
+    },
+    {
+      "epoch": 3.820509213672809,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.8679758827813835e-08,
+      "loss": 1.7012,
+      "step": 13580
+    },
+    {
+      "epoch": 3.8219158812772545,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.791888838719416e-08,
+      "loss": 1.4577,
+      "step": 13585
+    },
+    {
+      "epoch": 3.8233225488816993,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.7163975293008416e-08,
+      "loss": 1.6121,
+      "step": 13590
+    },
+    {
+      "epoch": 3.824729216486144,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.641502068320946e-08,
+      "loss": 1.3883,
+      "step": 13595
+    },
+    {
+      "epoch": 3.8261358840905895,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.567202568676665e-08,
+      "loss": 1.6198,
+      "step": 13600
+    },
+    {
+      "epoch": 3.8275425516950343,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.493499142366719e-08,
+      "loss": 1.8679,
+      "step": 13605
+    },
+    {
+      "epoch": 3.8289492192994796,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.4203919004912606e-08,
+      "loss": 1.5141,
+      "step": 13610
+    },
+    {
+      "epoch": 3.8303558869039245,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.347880953251737e-08,
+      "loss": 1.5901,
+      "step": 13615
+    },
+    {
+      "epoch": 3.83176255450837,
+      "grad_norm": 3.125,
+      "learning_rate": 4.275966409950804e-08,
+      "loss": 1.5317,
+      "step": 13620
+    },
+    {
+      "epoch": 3.8331692221128146,
+      "grad_norm": 3.625,
+      "learning_rate": 4.2046483789920596e-08,
+      "loss": 1.4834,
+      "step": 13625
+    },
+    {
+      "epoch": 3.83457588971726,
+      "grad_norm": 5.25,
+      "learning_rate": 4.1339269678799525e-08,
+      "loss": 1.6823,
+      "step": 13630
+    },
+    {
+      "epoch": 3.835982557321705,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.0638022832195197e-08,
+      "loss": 1.7195,
+      "step": 13635
+    },
+    {
+      "epoch": 3.83738922492615,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.994274430716427e-08,
+      "loss": 1.8747,
+      "step": 13640
+    },
+    {
+      "epoch": 3.838795892530595,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.925343515176482e-08,
+      "loss": 1.5168,
+      "step": 13645
+    },
+    {
+      "epoch": 3.8402025601350402,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.857009640505859e-08,
+      "loss": 1.7263,
+      "step": 13650
+    },
+    {
+      "epoch": 3.841609227739485,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.7892729097106944e-08,
+      "loss": 1.3747,
+      "step": 13655
+    },
+    {
+      "epoch": 3.84301589534393,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.722133424896956e-08,
+      "loss": 1.4349,
+      "step": 13660
+    },
+    {
+      "epoch": 3.8444225629483753,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.655591287270354e-08,
+      "loss": 1.7337,
+      "step": 13665
+    },
+    {
+      "epoch": 3.8458292305528206,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.589646597136209e-08,
+      "loss": 1.647,
+      "step": 13670
+    },
+    {
+      "epoch": 3.8472358981572654,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.524299453899093e-08,
+      "loss": 1.7049,
+      "step": 13675
+    },
+    {
+      "epoch": 3.8486425657617103,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.459549956063013e-08,
+      "loss": 1.4334,
+      "step": 13680
+    },
+    {
+      "epoch": 3.8500492333661556,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.395398201231048e-08,
+      "loss": 1.685,
+      "step": 13685
+    },
+    {
+      "epoch": 3.851455900970601,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.331844286105179e-08,
+      "loss": 1.6998,
+      "step": 13690
+    },
+    {
+      "epoch": 3.8528625685750457,
+      "grad_norm": 3.0,
+      "learning_rate": 3.268888306486284e-08,
+      "loss": 1.5654,
+      "step": 13695
+    },
+    {
+      "epoch": 3.8542692361794906,
+      "grad_norm": 4.46875,
+      "learning_rate": 3.206530357273829e-08,
+      "loss": 1.623,
+      "step": 13700
+    },
+    {
+      "epoch": 3.855675903783936,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.1447705324659126e-08,
+      "loss": 1.7226,
+      "step": 13705
+    },
+    {
+      "epoch": 3.857082571388381,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.0836089251589535e-08,
+      "loss": 1.5045,
+      "step": 13710
+    },
+    {
+      "epoch": 3.858489238992826,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.0230456275476045e-08,
+      "loss": 1.6649,
+      "step": 13715
+    },
+    {
+      "epoch": 3.859895906597271,
+      "grad_norm": 3.5,
+      "learning_rate": 2.963080730924705e-08,
+      "loss": 1.6491,
+      "step": 13720
+    },
+    {
+      "epoch": 3.861302574201716,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.903714325681017e-08,
+      "loss": 1.5911,
+      "step": 13725
+    },
+    {
+      "epoch": 3.862709241806161,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8449465013051343e-08,
+      "loss": 1.5251,
+      "step": 13730
+    },
+    {
+      "epoch": 3.8641159094106063,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7867773463833954e-08,
+      "loss": 1.6151,
+      "step": 13735
+    },
+    {
+      "epoch": 3.865522577015051,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.7292069485996604e-08,
+      "loss": 1.7493,
+      "step": 13740
+    },
+    {
+      "epoch": 3.8669292446194965,
+      "grad_norm": 4.75,
+      "learning_rate": 2.6722353947352227e-08,
+      "loss": 1.5258,
+      "step": 13745
+    },
+    {
+      "epoch": 3.8683359122239414,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.615862770668764e-08,
+      "loss": 1.4538,
+      "step": 13750
+    },
+    {
+      "epoch": 3.8697425798283867,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.5600891613760445e-08,
+      "loss": 1.5134,
+      "step": 13755
+    },
+    {
+      "epoch": 3.8711492474328315,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.5049146509299012e-08,
+      "loss": 1.4722,
+      "step": 13760
+    },
+    {
+      "epoch": 3.872555915037277,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.450339322500161e-08,
+      "loss": 1.608,
+      "step": 13765
+    },
+    {
+      "epoch": 3.8739625826417217,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.3963632583533733e-08,
+      "loss": 1.6941,
+      "step": 13770
+    },
+    {
+      "epoch": 3.875369250246167,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.342986539852676e-08,
+      "loss": 1.4776,
+      "step": 13775
+    },
+    {
+      "epoch": 3.876775917850612,
+      "grad_norm": 4.0,
+      "learning_rate": 2.2902092474579747e-08,
+      "loss": 1.5991,
+      "step": 13780
+    },
+    {
+      "epoch": 3.878182585455057,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2380314607254536e-08,
+      "loss": 1.542,
+      "step": 13785
+    },
+    {
+      "epoch": 3.879589253059502,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.186453258307619e-08,
+      "loss": 1.6427,
+      "step": 13790
+    },
+    {
+      "epoch": 3.8809959206639473,
+      "grad_norm": 3.125,
+      "learning_rate": 2.1354747179531674e-08,
+      "loss": 1.5292,
+      "step": 13795
+    },
+    {
+      "epoch": 3.882402588268392,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.0850959165069403e-08,
+      "loss": 1.4787,
+      "step": 13800
+    },
+    {
+      "epoch": 3.883809255872837,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.035316929909614e-08,
+      "loss": 1.6944,
+      "step": 13805
+    },
+    {
+      "epoch": 3.8852159234772823,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.9861378331978318e-08,
+      "loss": 1.5436,
+      "step": 13810
+    },
+    {
+      "epoch": 3.8866225910817276,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.937558700503894e-08,
+      "loss": 1.8179,
+      "step": 13815
+    },
+    {
+      "epoch": 3.8880292586861724,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8895796050557134e-08,
+      "loss": 1.3398,
+      "step": 13820
+    },
+    {
+      "epoch": 3.8894359262906173,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.8422006191766813e-08,
+      "loss": 1.6199,
+      "step": 13825
+    },
+    {
+      "epoch": 3.8908425938950626,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.795421814285758e-08,
+      "loss": 1.4014,
+      "step": 13830
+    },
+    {
+      "epoch": 3.892249261499508,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.7492432608969375e-08,
+      "loss": 1.7273,
+      "step": 13835
+    },
+    {
+      "epoch": 3.8936559291039528,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.7036650286196498e-08,
+      "loss": 1.3763,
+      "step": 13840
+    },
+    {
+      "epoch": 3.8950625967083976,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.6586871861581807e-08,
+      "loss": 1.6194,
+      "step": 13845
+    },
+    {
+      "epoch": 3.896469264312843,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.6143098013119415e-08,
+      "loss": 1.4738,
+      "step": 13850
+    },
+    {
+      "epoch": 3.897875931917288,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.5705329409751556e-08,
+      "loss": 1.6051,
+      "step": 13855
+    },
+    {
+      "epoch": 3.899282599521733,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.5273566711369036e-08,
+      "loss": 1.6372,
+      "step": 13860
+    },
+    {
+      "epoch": 3.900689267126178,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4847810568807683e-08,
+      "loss": 1.5585,
+      "step": 13865
+    },
+    {
+      "epoch": 3.902095934730623,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.4428061623850573e-08,
+      "loss": 1.4838,
+      "step": 13870
+    },
+    {
+      "epoch": 3.903502602335068,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4014320509224909e-08,
+      "loss": 1.6295,
+      "step": 13875
+    },
+    {
+      "epoch": 3.9049092699395134,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.3606587848602024e-08,
+      "loss": 1.6314,
+      "step": 13880
+    },
+    {
+      "epoch": 3.9063159375439582,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3204864256596504e-08,
+      "loss": 1.5786,
+      "step": 13885
+    },
+    {
+      "epoch": 3.9077226051484035,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2809150338763064e-08,
+      "loss": 1.345,
+      "step": 13890
+    },
+    {
+      "epoch": 3.9091292727528484,
+      "grad_norm": 3.25,
+      "learning_rate": 1.241944669160011e-08,
+      "loss": 1.7872,
+      "step": 13895
+    },
+    {
+      "epoch": 3.9105359403572937,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.203575390254441e-08,
+      "loss": 1.6342,
+      "step": 13900
+    },
+    {
+      "epoch": 3.9119426079617385,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.1658072549971975e-08,
+      "loss": 1.3083,
+      "step": 13905
+    },
+    {
+      "epoch": 3.913349275566184,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1286403203198513e-08,
+      "loss": 1.6906,
+      "step": 13910
+    },
+    {
+      "epoch": 3.9147559431706287,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.0920746422476313e-08,
+      "loss": 1.3366,
+      "step": 13915
+    },
+    {
+      "epoch": 3.916162610775074,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.05611027589938e-08,
+      "loss": 1.4893,
+      "step": 13920
+    },
+    {
+      "epoch": 3.917569278379519,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.0207472754876878e-08,
+      "loss": 1.844,
+      "step": 13925
+    },
+    {
+      "epoch": 3.918975945983964,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.859856943184919e-09,
+      "loss": 1.7505,
+      "step": 13930
+    },
+    {
+      "epoch": 3.920382613588409,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.518255847912548e-09,
+      "loss": 1.6995,
+      "step": 13935
+    },
+    {
+      "epoch": 3.9217892811928543,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.182669983986979e-09,
+      "loss": 1.4661,
+      "step": 13940
+    },
+    {
+      "epoch": 3.923195948797299,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.853099857269342e-09,
+      "loss": 1.2838,
+      "step": 13945
+    },
+    {
+      "epoch": 3.924602616401744,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.529545964551577e-09,
+      "loss": 1.5647,
+      "step": 13950
+    },
+    {
+      "epoch": 3.9260092840061893,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.212008793556436e-09,
+      "loss": 1.657,
+      "step": 13955
+    },
+    {
+      "epoch": 3.9274159516106346,
+      "grad_norm": 4.0,
+      "learning_rate": 7.900488822939255e-09,
+      "loss": 1.4927,
+      "step": 13960
+    },
+    {
+      "epoch": 3.9288226192150795,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.594986522282188e-09,
+      "loss": 1.604,
+      "step": 13965
+    },
+    {
+      "epoch": 3.9302292868195243,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.295502352098637e-09,
+      "loss": 1.5973,
+      "step": 13970
+    },
+    {
+      "epoch": 3.9316359544239696,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.002036763829267e-09,
+      "loss": 1.6541,
+      "step": 13975
+    },
+    {
+      "epoch": 3.933042622028415,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.7145901998424404e-09,
+      "loss": 1.6558,
+      "step": 13980
+    },
+    {
+      "epoch": 3.93444928963286,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.43316309343378e-09,
+      "loss": 1.7133,
+      "step": 13985
+    },
+    {
+      "epoch": 3.9358559572373046,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.157755868824832e-09,
+      "loss": 1.5935,
+      "step": 13990
+    },
+    {
+      "epoch": 3.93726262484175,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.888368941163513e-09,
+      "loss": 1.4415,
+      "step": 13995
+    },
+    {
+      "epoch": 3.9386692924461952,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.625002716521887e-09,
+      "loss": 1.6804,
+      "step": 14000
+    },
+    {
+      "epoch": 3.94007596005064,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.3676575918966125e-09,
+      "loss": 1.4467,
+      "step": 14005
+    },
+    {
+      "epoch": 3.941482627655085,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.1163339552084964e-09,
+      "loss": 1.4877,
+      "step": 14010
+    },
+    {
+      "epoch": 3.9428892952595302,
+      "grad_norm": 3.75,
+      "learning_rate": 4.871032185302048e-09,
+      "loss": 1.4424,
+      "step": 14015
+    },
+    {
+      "epoch": 3.944295962863975,
+      "grad_norm": 3.625,
+      "learning_rate": 4.631752651943266e-09,
+      "loss": 1.6061,
+      "step": 14020
+    },
+    {
+      "epoch": 3.9457026304684204,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.39849571582096e-09,
+      "loss": 1.4244,
+      "step": 14025
+    },
+    {
+      "epoch": 3.9471092980728653,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.171261728545428e-09,
+      "loss": 1.4968,
+      "step": 14030
+    },
+    {
+      "epoch": 3.9485159656773106,
+      "grad_norm": 2.375,
+      "learning_rate": 3.950051032648449e-09,
+      "loss": 1.5433,
+      "step": 14035
+    },
+    {
+      "epoch": 3.9499226332817554,
+      "grad_norm": 4.4375,
+      "learning_rate": 3.734863961581069e-09,
+      "loss": 1.6523,
+      "step": 14040
+    },
+    {
+      "epoch": 3.9513293008862007,
+      "grad_norm": 4.6875,
+      "learning_rate": 3.525700839715817e-09,
+      "loss": 1.3554,
+      "step": 14045
+    },
+    {
+      "epoch": 3.9527359684906456,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.322561982343597e-09,
+      "loss": 1.5458,
+      "step": 14050
+    },
+    {
+      "epoch": 3.954142636095091,
+      "grad_norm": 2.625,
+      "learning_rate": 3.1254476956750207e-09,
+      "loss": 1.5498,
+      "step": 14055
+    },
+    {
+      "epoch": 3.9555493036995357,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9343582768395215e-09,
+      "loss": 1.5235,
+      "step": 14060
+    },
+    {
+      "epoch": 3.956955971303981,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7492940138840183e-09,
+      "loss": 1.4832,
+      "step": 14065
+    },
+    {
+      "epoch": 3.958362638908426,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.5702551857733623e-09,
+      "loss": 1.6262,
+      "step": 14070
+    },
+    {
+      "epoch": 3.9597693065128707,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3972420623898927e-09,
+      "loss": 1.6804,
+      "step": 14075
+    },
+    {
+      "epoch": 3.961175974117316,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.230254904532547e-09,
+      "loss": 1.3863,
+      "step": 14080
+    },
+    {
+      "epoch": 3.9625826417217613,
+      "grad_norm": 5.03125,
+      "learning_rate": 2.069293963916419e-09,
+      "loss": 1.4973,
+      "step": 14085
+    },
+    {
+      "epoch": 3.963989309326206,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.9143594831740882e-09,
+      "loss": 1.6832,
+      "step": 14090
+    },
+    {
+      "epoch": 3.965395976930651,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.7654516958525156e-09,
+      "loss": 1.3712,
+      "step": 14095
+    },
+    {
+      "epoch": 3.9668026445350963,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.6225708264148153e-09,
+      "loss": 1.4175,
+      "step": 14100
+    },
+    {
+      "epoch": 3.9682093121395416,
+      "grad_norm": 4.125,
+      "learning_rate": 1.4857170902384807e-09,
+      "loss": 1.5806,
+      "step": 14105
+    },
+    {
+      "epoch": 3.9696159797439865,
+      "grad_norm": 2.875,
+      "learning_rate": 1.354890693616273e-09,
+      "loss": 1.7819,
+      "step": 14110
+    },
+    {
+      "epoch": 3.9710226473484314,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.2300918337553312e-09,
+      "loss": 1.413,
+      "step": 14115
+    },
+    {
+      "epoch": 3.9724293149528767,
+      "grad_norm": 2.875,
+      "learning_rate": 1.1113206987767298e-09,
+      "loss": 1.66,
+      "step": 14120
+    },
+    {
+      "epoch": 3.973835982557322,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.985774677150339e-10,
+      "loss": 1.5321,
+      "step": 14125
+    },
+    {
+      "epoch": 3.975242650161767,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.91862310519631e-10,
+      "loss": 1.4343,
+      "step": 14130
+    },
+    {
+      "epoch": 3.9766493177662117,
+      "grad_norm": 3.0,
+      "learning_rate": 7.911753880516236e-10,
+      "loss": 1.6013,
+      "step": 14135
+    },
+    {
+      "epoch": 3.978055985370657,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.965168520864928e-10,
+      "loss": 1.517,
+      "step": 14140
+    },
+    {
+      "epoch": 3.979462652975102,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.07886845311878e-10,
+      "loss": 1.6839,
+      "step": 14145
+    },
+    {
+      "epoch": 3.980869320579547,
+      "grad_norm": 3.625,
+      "learning_rate": 5.252855013280211e-10,
+      "loss": 1.5155,
+      "step": 14150
+    },
+    {
+      "epoch": 3.982275988183992,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.487129446477667e-10,
+      "loss": 1.6638,
+      "step": 14155
+    },
+    {
+      "epoch": 3.9836826557884373,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.7816929069656165e-10,
+      "loss": 1.6047,
+      "step": 14160
+    },
+    {
+      "epoch": 3.985089323392882,
+      "grad_norm": 3.25,
+      "learning_rate": 3.1365464581112334e-10,
+      "loss": 1.5561,
+      "step": 14165
+    },
+    {
+      "epoch": 3.9864959909973274,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.5516910724077133e-10,
+      "loss": 1.7709,
+      "step": 14170
+    },
+    {
+      "epoch": 3.9879026586017723,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0271276314565155e-10,
+      "loss": 1.612,
+      "step": 14175
+    },
+    {
+      "epoch": 3.9893093262062176,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.5628569259940049e-10,
+      "loss": 1.4502,
+      "step": 14180
+    },
+    {
+      "epoch": 3.9907159938106624,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1588796558470449e-10,
+      "loss": 1.6444,
+      "step": 14185
+    },
+    {
+      "epoch": 3.9921226614151077,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.15196429977405e-11,
+      "loss": 1.2351,
+      "step": 14190
+    },
+    {
+      "epoch": 3.9935293290195526,
+      "grad_norm": 4.125,
+      "learning_rate": 5.3180776644623505e-11,
+      "loss": 1.6336,
+      "step": 14195
+    },
+    {
+      "epoch": 3.994935996623998,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.087140924318277e-11,
+      "loss": 1.638,
+      "step": 14200
+    },
+    {
+      "epoch": 3.9963426642284428,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.4591574422961883e-11,
+      "loss": 1.7758,
+      "step": 14205
+    },
+    {
+      "epoch": 3.997749331832888,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.3412967238865004e-12,
+      "loss": 1.8328,
+      "step": 14210
+    },
+    {
+      "epoch": 3.999155999437333,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2059159670840813e-13,
+      "loss": 1.8058,
+      "step": 14215
+    },
+    {
+      "epoch": 3.9994373329582222,
+      "eval_loss": 1.576697826385498,
+      "eval_runtime": 330.563,
+      "eval_samples_per_second": 9.553,
+      "eval_steps_per_second": 4.777,
+      "step": 14216
+    },
+    {
+      "epoch": 3.9994373329582222,
+      "step": 14216,
+      "total_flos": 1.684239053169361e+18,
+      "train_loss": 1.6145593146796164,
+      "train_runtime": 58184.933,
+      "train_samples_per_second": 1.955,
+      "train_steps_per_second": 0.244
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 14216,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.684239053169361e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}