Upload folder using huggingface_hub

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c040bd2c72181ac32d9eca911ad3511e17a00b7f7692f948f0128919daeeca25
+oid sha256:a4a39d07526ad4c47fcf0ed20eb901c709f826a40aea5a2ba5a7eb336482a344
 size 13648432

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43bbe3ae6b244b1777e6da7982cbf7a2e9cb8022fbeddcd423589d85a6101664
+oid sha256:96f4e1fc406e0ef3079999964fd2e8c98f4ab679496de7aff06c9b03585f1ba4
 size 27338810

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e73f434e8535656b41b1d5bfa3067fcbe838f652992a4fe71ac08e37659f183f
+oid sha256:169aa60243c3fe8bcd221891833b2694719c7c5c130939d89d62ad2ab552d909
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c29143e0a84c1bbf5f7d6c4cdd96647b5d99b5c3882bb49ef9e6bdc5f0abca94
+oid sha256:7448b75ce12f86e3d3b951700cc1d23b4d30c97166c1f777930e49ec35d82b42
 size 1064

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 2.072087287902832,
-  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-3Q-0U-0C-qa_first/checkpoint-811",
-  "epoch": 1.9993836671802774,
+  "best_metric": 2.093395709991455,
+  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-3Q-0U-0C-qa_first/checkpoint-405",
+  "epoch": 0.9984591679506933,
   "eval_steps": 500,
-  "global_step": 811,
+  "global_step": 405,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2850,2856 +2850,6 @@
       "eval_samples_per_second": 1.234,
       "eval_steps_per_second": 0.308,
       "step": 405
-    },
-    {
-      "epoch": 1.000924499229584,
-      "grad_norm": 0.5109534859657288,
-      "learning_rate": 3.5051993067590985e-07,
-      "loss": 2.0975,
-      "step": 406
-    },
-    {
-      "epoch": 1.0033898305084745,
-      "grad_norm": 0.5240735411643982,
-      "learning_rate": 3.500866551126516e-07,
-      "loss": 2.1655,
-      "step": 407
-    },
-    {
-      "epoch": 1.0058551617873652,
-      "grad_norm": 0.5256450772285461,
-      "learning_rate": 3.496533795493934e-07,
-      "loss": 2.1388,
-      "step": 408
-    },
-    {
-      "epoch": 1.0083204930662557,
-      "grad_norm": 0.5329455733299255,
-      "learning_rate": 3.4922010398613514e-07,
-      "loss": 2.089,
-      "step": 409
-    },
-    {
-      "epoch": 1.0107858243451464,
-      "grad_norm": 0.5201774835586548,
-      "learning_rate": 3.4878682842287693e-07,
-      "loss": 2.0812,
-      "step": 410
-    },
-    {
-      "epoch": 1.013251155624037,
-      "grad_norm": 0.482138454914093,
-      "learning_rate": 3.483535528596187e-07,
-      "loss": 2.1011,
-      "step": 411
-    },
-    {
-      "epoch": 1.0157164869029276,
-      "grad_norm": 0.5111163258552551,
-      "learning_rate": 3.479202772963605e-07,
-      "loss": 2.1351,
-      "step": 412
-    },
-    {
-      "epoch": 1.018181818181818,
-      "grad_norm": 0.5070258975028992,
-      "learning_rate": 3.474870017331022e-07,
-      "loss": 2.0726,
-      "step": 413
-    },
-    {
-      "epoch": 1.0206471494607088,
-      "grad_norm": 0.5213172435760498,
-      "learning_rate": 3.47053726169844e-07,
-      "loss": 2.0709,
-      "step": 414
-    },
-    {
-      "epoch": 1.0231124807395995,
-      "grad_norm": 0.5116674304008484,
-      "learning_rate": 3.4662045060658576e-07,
-      "loss": 2.1098,
-      "step": 415
-    },
-    {
-      "epoch": 1.02557781201849,
-      "grad_norm": 0.5342626571655273,
-      "learning_rate": 3.4618717504332756e-07,
-      "loss": 2.0572,
-      "step": 416
-    },
-    {
-      "epoch": 1.0280431432973807,
-      "grad_norm": 0.5366671085357666,
-      "learning_rate": 3.457538994800693e-07,
-      "loss": 2.1119,
-      "step": 417
-    },
-    {
-      "epoch": 1.0305084745762711,
-      "grad_norm": 0.5127258896827698,
-      "learning_rate": 3.4532062391681105e-07,
-      "loss": 2.1222,
-      "step": 418
-    },
-    {
-      "epoch": 1.0329738058551619,
-      "grad_norm": 0.5240333080291748,
-      "learning_rate": 3.448873483535528e-07,
-      "loss": 2.0789,
-      "step": 419
-    },
-    {
-      "epoch": 1.0354391371340523,
-      "grad_norm": 0.529655933380127,
-      "learning_rate": 3.444540727902946e-07,
-      "loss": 2.1118,
-      "step": 420
-    },
-    {
-      "epoch": 1.037904468412943,
-      "grad_norm": 0.5909092426300049,
-      "learning_rate": 3.4402079722703634e-07,
-      "loss": 2.0552,
-      "step": 421
-    },
-    {
-      "epoch": 1.0403697996918335,
-      "grad_norm": 0.5189386010169983,
-      "learning_rate": 3.4358752166377814e-07,
-      "loss": 2.0904,
-      "step": 422
-    },
-    {
-      "epoch": 1.0428351309707242,
-      "grad_norm": 0.5365529656410217,
-      "learning_rate": 3.431542461005199e-07,
-      "loss": 2.0861,
-      "step": 423
-    },
-    {
-      "epoch": 1.0453004622496147,
-      "grad_norm": 0.5217022895812988,
-      "learning_rate": 3.427209705372617e-07,
-      "loss": 2.0698,
-      "step": 424
-    },
-    {
-      "epoch": 1.0477657935285054,
-      "grad_norm": 0.5230218172073364,
-      "learning_rate": 3.4228769497400343e-07,
-      "loss": 2.0571,
-      "step": 425
-    },
-    {
-      "epoch": 1.0502311248073959,
-      "grad_norm": 0.5243125557899475,
-      "learning_rate": 3.418544194107452e-07,
-      "loss": 2.1333,
-      "step": 426
-    },
-    {
-      "epoch": 1.0526964560862866,
-      "grad_norm": 0.5268332958221436,
-      "learning_rate": 3.4142114384748697e-07,
-      "loss": 2.0533,
-      "step": 427
-    },
-    {
-      "epoch": 1.0551617873651773,
-      "grad_norm": 0.5369420051574707,
-      "learning_rate": 3.4098786828422877e-07,
-      "loss": 2.0523,
-      "step": 428
-    },
-    {
-      "epoch": 1.0576271186440678,
-      "grad_norm": 0.5415328145027161,
-      "learning_rate": 3.405545927209705e-07,
-      "loss": 2.1297,
-      "step": 429
-    },
-    {
-      "epoch": 1.0600924499229585,
-      "grad_norm": 0.5068016648292542,
-      "learning_rate": 3.401213171577123e-07,
-      "loss": 2.1055,
-      "step": 430
-    },
-    {
-      "epoch": 1.062557781201849,
-      "grad_norm": 0.5332845449447632,
-      "learning_rate": 3.3968804159445406e-07,
-      "loss": 2.0956,
-      "step": 431
-    },
-    {
-      "epoch": 1.0650231124807397,
-      "grad_norm": 0.5464420318603516,
-      "learning_rate": 3.3925476603119585e-07,
-      "loss": 2.1394,
-      "step": 432
-    },
-    {
-      "epoch": 1.0674884437596301,
-      "grad_norm": 0.5103365778923035,
-      "learning_rate": 3.388214904679376e-07,
-      "loss": 2.0665,
-      "step": 433
-    },
-    {
-      "epoch": 1.0699537750385208,
-      "grad_norm": 0.516782820224762,
-      "learning_rate": 3.3838821490467934e-07,
-      "loss": 2.0944,
-      "step": 434
-    },
-    {
-      "epoch": 1.0724191063174113,
-      "grad_norm": 0.518387496471405,
-      "learning_rate": 3.379549393414211e-07,
-      "loss": 2.071,
-      "step": 435
-    },
-    {
-      "epoch": 1.074884437596302,
-      "grad_norm": 0.5234736204147339,
-      "learning_rate": 3.375216637781629e-07,
-      "loss": 2.1083,
-      "step": 436
-    },
-    {
-      "epoch": 1.0773497688751925,
-      "grad_norm": 0.5194358229637146,
-      "learning_rate": 3.3708838821490463e-07,
-      "loss": 2.0615,
-      "step": 437
-    },
-    {
-      "epoch": 1.0798151001540832,
-      "grad_norm": 0.4968649744987488,
-      "learning_rate": 3.3665511265164643e-07,
-      "loss": 2.0345,
-      "step": 438
-    },
-    {
-      "epoch": 1.0822804314329737,
-      "grad_norm": 0.5469282865524292,
-      "learning_rate": 3.362218370883882e-07,
-      "loss": 2.1193,
-      "step": 439
-    },
-    {
-      "epoch": 1.0847457627118644,
-      "grad_norm": 0.509821891784668,
-      "learning_rate": 3.3578856152512997e-07,
-      "loss": 2.092,
-      "step": 440
-    },
-    {
-      "epoch": 1.087211093990755,
-      "grad_norm": 0.5226166248321533,
-      "learning_rate": 3.353552859618717e-07,
-      "loss": 2.0753,
-      "step": 441
-    },
-    {
-      "epoch": 1.0896764252696456,
-      "grad_norm": 0.5321833491325378,
-      "learning_rate": 3.349220103986135e-07,
-      "loss": 2.0423,
-      "step": 442
-    },
-    {
-      "epoch": 1.0921417565485363,
-      "grad_norm": 0.5013507604598999,
-      "learning_rate": 3.3448873483535526e-07,
-      "loss": 2.0561,
-      "step": 443
-    },
-    {
-      "epoch": 1.0946070878274268,
-      "grad_norm": 0.5189869999885559,
-      "learning_rate": 3.3405545927209706e-07,
-      "loss": 2.074,
-      "step": 444
-    },
-    {
-      "epoch": 1.0970724191063175,
-      "grad_norm": 0.5598823428153992,
-      "learning_rate": 3.336221837088388e-07,
-      "loss": 2.082,
-      "step": 445
-    },
-    {
-      "epoch": 1.099537750385208,
-      "grad_norm": 0.553289532661438,
-      "learning_rate": 3.331889081455806e-07,
-      "loss": 2.1035,
-      "step": 446
-    },
-    {
-      "epoch": 1.1020030816640987,
-      "grad_norm": 0.5163353681564331,
-      "learning_rate": 3.3275563258232235e-07,
-      "loss": 2.0894,
-      "step": 447
-    },
-    {
-      "epoch": 1.1044684129429891,
-      "grad_norm": 0.5245928168296814,
-      "learning_rate": 3.3232235701906414e-07,
-      "loss": 2.0705,
-      "step": 448
-    },
-    {
-      "epoch": 1.1069337442218798,
-      "grad_norm": 0.5174081921577454,
-      "learning_rate": 3.3188908145580584e-07,
-      "loss": 2.1132,
-      "step": 449
-    },
-    {
-      "epoch": 1.1093990755007703,
-      "grad_norm": 0.5218523740768433,
-      "learning_rate": 3.3145580589254763e-07,
-      "loss": 2.0536,
-      "step": 450
-    },
-    {
-      "epoch": 1.111864406779661,
-      "grad_norm": 0.525867760181427,
-      "learning_rate": 3.310225303292894e-07,
-      "loss": 2.1543,
-      "step": 451
-    },
-    {
-      "epoch": 1.1143297380585517,
-      "grad_norm": 0.5667258501052856,
-      "learning_rate": 3.305892547660312e-07,
-      "loss": 2.0583,
-      "step": 452
-    },
-    {
-      "epoch": 1.1167950693374422,
-      "grad_norm": 0.541770875453949,
-      "learning_rate": 3.301559792027729e-07,
-      "loss": 2.1327,
-      "step": 453
-    },
-    {
-      "epoch": 1.119260400616333,
-      "grad_norm": 0.5466485023498535,
-      "learning_rate": 3.297227036395147e-07,
-      "loss": 2.0528,
-      "step": 454
-    },
-    {
-      "epoch": 1.1217257318952234,
-      "grad_norm": 0.5483329892158508,
-      "learning_rate": 3.2928942807625647e-07,
-      "loss": 2.1536,
-      "step": 455
-    },
-    {
-      "epoch": 1.124191063174114,
-      "grad_norm": 0.5235666632652283,
-      "learning_rate": 3.2885615251299826e-07,
-      "loss": 2.0838,
-      "step": 456
-    },
-    {
-      "epoch": 1.1266563944530046,
-      "grad_norm": 0.5424934029579163,
-      "learning_rate": 3.2842287694974e-07,
-      "loss": 2.113,
-      "step": 457
-    },
-    {
-      "epoch": 1.1291217257318953,
-      "grad_norm": 0.5532853007316589,
-      "learning_rate": 3.279896013864818e-07,
-      "loss": 2.0799,
-      "step": 458
-    },
-    {
-      "epoch": 1.1315870570107858,
-      "grad_norm": 0.5158779621124268,
-      "learning_rate": 3.2755632582322355e-07,
-      "loss": 2.0674,
-      "step": 459
-    },
-    {
-      "epoch": 1.1340523882896765,
-      "grad_norm": 0.5670775175094604,
-      "learning_rate": 3.2712305025996535e-07,
-      "loss": 2.0898,
-      "step": 460
-    },
-    {
-      "epoch": 1.136517719568567,
-      "grad_norm": 0.5700893402099609,
-      "learning_rate": 3.266897746967071e-07,
-      "loss": 2.0833,
-      "step": 461
-    },
-    {
-      "epoch": 1.1389830508474577,
-      "grad_norm": 0.5165070295333862,
-      "learning_rate": 3.262564991334489e-07,
-      "loss": 2.1227,
-      "step": 462
-    },
-    {
-      "epoch": 1.1414483821263481,
-      "grad_norm": 0.5242537260055542,
-      "learning_rate": 3.2582322357019064e-07,
-      "loss": 2.0674,
-      "step": 463
-    },
-    {
-      "epoch": 1.1439137134052388,
-      "grad_norm": 0.5270352363586426,
-      "learning_rate": 3.253899480069324e-07,
-      "loss": 2.0815,
-      "step": 464
-    },
-    {
-      "epoch": 1.1463790446841293,
-      "grad_norm": 0.5376419425010681,
-      "learning_rate": 3.2495667244367413e-07,
-      "loss": 2.0699,
-      "step": 465
-    },
-    {
-      "epoch": 1.14884437596302,
-      "grad_norm": 0.5110929012298584,
-      "learning_rate": 3.245233968804159e-07,
-      "loss": 2.0393,
-      "step": 466
-    },
-    {
-      "epoch": 1.1513097072419107,
-      "grad_norm": 0.5514217019081116,
-      "learning_rate": 3.2409012131715767e-07,
-      "loss": 2.0484,
-      "step": 467
-    },
-    {
-      "epoch": 1.1537750385208012,
-      "grad_norm": 0.5122043490409851,
-      "learning_rate": 3.2365684575389947e-07,
-      "loss": 2.0707,
-      "step": 468
-    },
-    {
-      "epoch": 1.156240369799692,
-      "grad_norm": 0.5329337120056152,
-      "learning_rate": 3.232235701906412e-07,
-      "loss": 2.1176,
-      "step": 469
-    },
-    {
-      "epoch": 1.1587057010785824,
-      "grad_norm": 0.546440064907074,
-      "learning_rate": 3.22790294627383e-07,
-      "loss": 2.0814,
-      "step": 470
-    },
-    {
-      "epoch": 1.161171032357473,
-      "grad_norm": 0.5156552791595459,
-      "learning_rate": 3.2235701906412476e-07,
-      "loss": 2.0764,
-      "step": 471
-    },
-    {
-      "epoch": 1.1636363636363636,
-      "grad_norm": 0.5265877842903137,
-      "learning_rate": 3.2192374350086656e-07,
-      "loss": 2.0558,
-      "step": 472
-    },
-    {
-      "epoch": 1.1661016949152543,
-      "grad_norm": 0.5469325184822083,
-      "learning_rate": 3.214904679376083e-07,
-      "loss": 2.1109,
-      "step": 473
-    },
-    {
-      "epoch": 1.1685670261941448,
-      "grad_norm": 0.5088624358177185,
-      "learning_rate": 3.210571923743501e-07,
-      "loss": 2.1014,
-      "step": 474
-    },
-    {
-      "epoch": 1.1710323574730355,
-      "grad_norm": 0.5592899322509766,
-      "learning_rate": 3.2062391681109184e-07,
-      "loss": 2.0925,
-      "step": 475
-    },
-    {
-      "epoch": 1.173497688751926,
-      "grad_norm": 0.5526501536369324,
-      "learning_rate": 3.2019064124783364e-07,
-      "loss": 2.0594,
-      "step": 476
-    },
-    {
-      "epoch": 1.1759630200308167,
-      "grad_norm": 0.5420464277267456,
-      "learning_rate": 3.197573656845754e-07,
-      "loss": 2.1364,
-      "step": 477
-    },
-    {
-      "epoch": 1.1784283513097074,
-      "grad_norm": 0.53159099817276,
-      "learning_rate": 3.193240901213172e-07,
-      "loss": 2.0358,
-      "step": 478
-    },
-    {
-      "epoch": 1.1808936825885978,
-      "grad_norm": 0.5268188118934631,
-      "learning_rate": 3.188908145580589e-07,
-      "loss": 2.0514,
-      "step": 479
-    },
-    {
-      "epoch": 1.1833590138674885,
-      "grad_norm": 0.5499210953712463,
-      "learning_rate": 3.184575389948007e-07,
-      "loss": 2.0831,
-      "step": 480
-    },
-    {
-      "epoch": 1.185824345146379,
-      "grad_norm": 0.5099960565567017,
-      "learning_rate": 3.180242634315424e-07,
-      "loss": 2.0724,
-      "step": 481
-    },
-    {
-      "epoch": 1.1882896764252697,
-      "grad_norm": 0.5238215327262878,
-      "learning_rate": 3.175909878682842e-07,
-      "loss": 2.0665,
-      "step": 482
-    },
-    {
-      "epoch": 1.1907550077041602,
-      "grad_norm": 0.5423519611358643,
-      "learning_rate": 3.1715771230502596e-07,
-      "loss": 2.1059,
-      "step": 483
-    },
-    {
-      "epoch": 1.193220338983051,
-      "grad_norm": 0.5435311198234558,
-      "learning_rate": 3.1672443674176776e-07,
-      "loss": 2.1405,
-      "step": 484
-    },
-    {
-      "epoch": 1.1956856702619414,
-      "grad_norm": 0.5252048373222351,
-      "learning_rate": 3.162911611785095e-07,
-      "loss": 2.0534,
-      "step": 485
-    },
-    {
-      "epoch": 1.198151001540832,
-      "grad_norm": 0.5369661450386047,
-      "learning_rate": 3.158578856152513e-07,
-      "loss": 2.1207,
-      "step": 486
-    },
-    {
-      "epoch": 1.2006163328197226,
-      "grad_norm": 0.5454829931259155,
-      "learning_rate": 3.1542461005199305e-07,
-      "loss": 2.086,
-      "step": 487
-    },
-    {
-      "epoch": 1.2030816640986133,
-      "grad_norm": 0.5264859199523926,
-      "learning_rate": 3.1499133448873485e-07,
-      "loss": 2.0628,
-      "step": 488
-    },
-    {
-      "epoch": 1.2055469953775038,
-      "grad_norm": 0.5218833684921265,
-      "learning_rate": 3.145580589254766e-07,
-      "loss": 2.0481,
-      "step": 489
-    },
-    {
-      "epoch": 1.2080123266563945,
-      "grad_norm": 0.552793025970459,
-      "learning_rate": 3.141247833622184e-07,
-      "loss": 2.0154,
-      "step": 490
-    },
-    {
-      "epoch": 1.210477657935285,
-      "grad_norm": 0.5695677995681763,
-      "learning_rate": 3.1369150779896013e-07,
-      "loss": 2.0963,
-      "step": 491
-    },
-    {
-      "epoch": 1.2129429892141756,
-      "grad_norm": 0.5433192849159241,
-      "learning_rate": 3.1325823223570193e-07,
-      "loss": 2.1519,
-      "step": 492
-    },
-    {
-      "epoch": 1.2154083204930664,
-      "grad_norm": 0.5215097069740295,
-      "learning_rate": 3.128249566724437e-07,
-      "loss": 2.0963,
-      "step": 493
-    },
-    {
-      "epoch": 1.2178736517719568,
-      "grad_norm": 0.5091007351875305,
-      "learning_rate": 3.123916811091854e-07,
-      "loss": 2.1167,
-      "step": 494
-    },
-    {
-      "epoch": 1.2203389830508475,
-      "grad_norm": 0.500039279460907,
-      "learning_rate": 3.1195840554592717e-07,
-      "loss": 2.0676,
-      "step": 495
-    },
-    {
-      "epoch": 1.222804314329738,
-      "grad_norm": 0.550313413143158,
-      "learning_rate": 3.1152512998266897e-07,
-      "loss": 2.141,
-      "step": 496
-    },
-    {
-      "epoch": 1.2252696456086287,
-      "grad_norm": 0.5415179133415222,
-      "learning_rate": 3.110918544194107e-07,
-      "loss": 2.0414,
-      "step": 497
-    },
-    {
-      "epoch": 1.2277349768875192,
-      "grad_norm": 0.5273166298866272,
-      "learning_rate": 3.106585788561525e-07,
-      "loss": 2.0847,
-      "step": 498
-    },
-    {
-      "epoch": 1.23020030816641,
-      "grad_norm": 0.5227318406105042,
-      "learning_rate": 3.1022530329289425e-07,
-      "loss": 2.074,
-      "step": 499
-    },
-    {
-      "epoch": 1.2326656394453004,
-      "grad_norm": 0.5437159538269043,
-      "learning_rate": 3.0979202772963605e-07,
-      "loss": 2.072,
-      "step": 500
-    },
-    {
-      "epoch": 1.235130970724191,
-      "grad_norm": 0.5285058617591858,
-      "learning_rate": 3.093587521663778e-07,
-      "loss": 2.0242,
-      "step": 501
-    },
-    {
-      "epoch": 1.2375963020030816,
-      "grad_norm": 0.5310484766960144,
-      "learning_rate": 3.089254766031196e-07,
-      "loss": 2.0797,
-      "step": 502
-    },
-    {
-      "epoch": 1.2400616332819723,
-      "grad_norm": 0.5042904019355774,
-      "learning_rate": 3.0849220103986134e-07,
-      "loss": 2.0703,
-      "step": 503
-    },
-    {
-      "epoch": 1.242526964560863,
-      "grad_norm": 0.5337966084480286,
-      "learning_rate": 3.0805892547660314e-07,
-      "loss": 2.0958,
-      "step": 504
-    },
-    {
-      "epoch": 1.2449922958397535,
-      "grad_norm": 0.5376964211463928,
-      "learning_rate": 3.076256499133449e-07,
-      "loss": 2.0587,
-      "step": 505
-    },
-    {
-      "epoch": 1.2474576271186442,
-      "grad_norm": 0.5448833703994751,
-      "learning_rate": 3.071923743500867e-07,
-      "loss": 2.0371,
-      "step": 506
-    },
-    {
-      "epoch": 1.2499229583975346,
-      "grad_norm": 0.5545293688774109,
-      "learning_rate": 3.067590987868284e-07,
-      "loss": 2.1334,
-      "step": 507
-    },
-    {
-      "epoch": 1.2523882896764253,
-      "grad_norm": 0.5596605539321899,
-      "learning_rate": 3.063258232235702e-07,
-      "loss": 2.1101,
-      "step": 508
-    },
-    {
-      "epoch": 1.2548536209553158,
-      "grad_norm": 0.5680539011955261,
-      "learning_rate": 3.058925476603119e-07,
-      "loss": 2.1224,
-      "step": 509
-    },
-    {
-      "epoch": 1.2573189522342065,
-      "grad_norm": 0.5574333071708679,
-      "learning_rate": 3.054592720970537e-07,
-      "loss": 2.0774,
-      "step": 510
-    },
-    {
-      "epoch": 1.259784283513097,
-      "grad_norm": 0.5248346924781799,
-      "learning_rate": 3.0502599653379546e-07,
-      "loss": 2.0657,
-      "step": 511
-    },
-    {
-      "epoch": 1.2622496147919877,
-      "grad_norm": 0.5491170883178711,
-      "learning_rate": 3.0459272097053726e-07,
-      "loss": 2.0696,
-      "step": 512
-    },
-    {
-      "epoch": 1.2647149460708782,
-      "grad_norm": 0.5188306570053101,
-      "learning_rate": 3.04159445407279e-07,
-      "loss": 2.0528,
-      "step": 513
-    },
-    {
-      "epoch": 1.267180277349769,
-      "grad_norm": 0.5485676527023315,
-      "learning_rate": 3.037261698440208e-07,
-      "loss": 2.0722,
-      "step": 514
-    },
-    {
-      "epoch": 1.2696456086286596,
-      "grad_norm": 0.5163741707801819,
-      "learning_rate": 3.0329289428076254e-07,
-      "loss": 2.0316,
-      "step": 515
-    },
-    {
-      "epoch": 1.27211093990755,
-      "grad_norm": 0.5422912836074829,
-      "learning_rate": 3.0285961871750434e-07,
-      "loss": 2.131,
-      "step": 516
-    },
-    {
-      "epoch": 1.2745762711864406,
-      "grad_norm": 0.5590441823005676,
-      "learning_rate": 3.024263431542461e-07,
-      "loss": 2.136,
-      "step": 517
-    },
-    {
-      "epoch": 1.2770416024653313,
-      "grad_norm": 0.5339657068252563,
-      "learning_rate": 3.019930675909879e-07,
-      "loss": 2.0866,
-      "step": 518
-    },
-    {
-      "epoch": 1.279506933744222,
-      "grad_norm": 0.5249587893486023,
-      "learning_rate": 3.0155979202772963e-07,
-      "loss": 2.0505,
-      "step": 519
-    },
-    {
-      "epoch": 1.2819722650231125,
-      "grad_norm": 0.5352839827537537,
-      "learning_rate": 3.0112651646447143e-07,
-      "loss": 2.0719,
-      "step": 520
-    },
-    {
-      "epoch": 1.2844375963020032,
-      "grad_norm": 0.521338939666748,
-      "learning_rate": 3.0069324090121317e-07,
-      "loss": 2.1264,
-      "step": 521
-    },
-    {
-      "epoch": 1.2869029275808936,
-      "grad_norm": 0.5433731079101562,
-      "learning_rate": 3.0025996533795497e-07,
-      "loss": 2.1272,
-      "step": 522
-    },
-    {
-      "epoch": 1.2893682588597843,
-      "grad_norm": 0.5479230880737305,
-      "learning_rate": 2.998266897746967e-07,
-      "loss": 2.0996,
-      "step": 523
-    },
-    {
-      "epoch": 1.2918335901386748,
-      "grad_norm": 0.5389401912689209,
-      "learning_rate": 2.993934142114385e-07,
-      "loss": 2.1212,
-      "step": 524
-    },
-    {
-      "epoch": 1.2942989214175655,
-      "grad_norm": 0.5670454502105713,
-      "learning_rate": 2.989601386481802e-07,
-      "loss": 2.0662,
-      "step": 525
-    },
-    {
-      "epoch": 1.296764252696456,
-      "grad_norm": 0.5215103030204773,
-      "learning_rate": 2.9852686308492195e-07,
-      "loss": 2.0437,
-      "step": 526
-    },
-    {
-      "epoch": 1.2992295839753467,
-      "grad_norm": 0.5562092065811157,
-      "learning_rate": 2.9809358752166375e-07,
-      "loss": 2.0364,
-      "step": 527
-    },
-    {
-      "epoch": 1.3016949152542372,
-      "grad_norm": 0.5172645449638367,
-      "learning_rate": 2.976603119584055e-07,
-      "loss": 2.0302,
-      "step": 528
-    },
-    {
-      "epoch": 1.304160246533128,
-      "grad_norm": 0.5652945041656494,
-      "learning_rate": 2.972270363951473e-07,
-      "loss": 2.1963,
-      "step": 529
-    },
-    {
-      "epoch": 1.3066255778120186,
-      "grad_norm": 0.5477254986763,
-      "learning_rate": 2.9679376083188904e-07,
-      "loss": 2.0815,
-      "step": 530
-    },
-    {
-      "epoch": 1.309090909090909,
-      "grad_norm": 0.5283427834510803,
-      "learning_rate": 2.9636048526863084e-07,
-      "loss": 2.0962,
-      "step": 531
-    },
-    {
-      "epoch": 1.3115562403697996,
-      "grad_norm": 0.5537866950035095,
-      "learning_rate": 2.959272097053726e-07,
-      "loss": 2.0885,
-      "step": 532
-    },
-    {
-      "epoch": 1.3140215716486903,
-      "grad_norm": 0.5142833590507507,
-      "learning_rate": 2.954939341421144e-07,
-      "loss": 2.0664,
-      "step": 533
-    },
-    {
-      "epoch": 1.316486902927581,
-      "grad_norm": 0.5292583107948303,
-      "learning_rate": 2.950606585788561e-07,
-      "loss": 2.0065,
-      "step": 534
-    },
-    {
-      "epoch": 1.3189522342064715,
-      "grad_norm": 0.5204624533653259,
-      "learning_rate": 2.946273830155979e-07,
-      "loss": 2.0522,
-      "step": 535
-    },
-    {
-      "epoch": 1.3214175654853622,
-      "grad_norm": 0.534831702709198,
-      "learning_rate": 2.9419410745233967e-07,
-      "loss": 2.1378,
-      "step": 536
-    },
-    {
-      "epoch": 1.3238828967642526,
-      "grad_norm": 0.5467716455459595,
-      "learning_rate": 2.9376083188908146e-07,
-      "loss": 2.0894,
-      "step": 537
-    },
-    {
-      "epoch": 1.3263482280431433,
-      "grad_norm": 0.536545991897583,
-      "learning_rate": 2.933275563258232e-07,
-      "loss": 2.0737,
-      "step": 538
-    },
-    {
-      "epoch": 1.3288135593220338,
-      "grad_norm": 0.5458997488021851,
-      "learning_rate": 2.92894280762565e-07,
-      "loss": 2.0601,
-      "step": 539
-    },
-    {
-      "epoch": 1.3312788906009245,
-      "grad_norm": 0.5358441472053528,
-      "learning_rate": 2.924610051993067e-07,
-      "loss": 2.0644,
-      "step": 540
-    },
-    {
-      "epoch": 1.3337442218798152,
-      "grad_norm": 0.5691521763801575,
-      "learning_rate": 2.920277296360485e-07,
-      "loss": 2.1359,
-      "step": 541
-    },
-    {
-      "epoch": 1.3362095531587057,
-      "grad_norm": 0.5218629837036133,
-      "learning_rate": 2.9159445407279024e-07,
-      "loss": 2.058,
-      "step": 542
-    },
-    {
-      "epoch": 1.3386748844375962,
-      "grad_norm": 0.5232541561126709,
-      "learning_rate": 2.9116117850953204e-07,
-      "loss": 2.0103,
-      "step": 543
-    },
-    {
-      "epoch": 1.341140215716487,
-      "grad_norm": 0.4909007251262665,
-      "learning_rate": 2.907279029462738e-07,
-      "loss": 2.0264,
-      "step": 544
-    },
-    {
-      "epoch": 1.3436055469953776,
-      "grad_norm": 0.5056614875793457,
-      "learning_rate": 2.902946273830156e-07,
-      "loss": 2.0628,
-      "step": 545
-    },
-    {
-      "epoch": 1.346070878274268,
-      "grad_norm": 0.5265542268753052,
-      "learning_rate": 2.8986135181975733e-07,
-      "loss": 2.098,
-      "step": 546
-    },
-    {
-      "epoch": 1.3485362095531588,
-      "grad_norm": 0.5243731141090393,
-      "learning_rate": 2.894280762564991e-07,
-      "loss": 2.0943,
-      "step": 547
-    },
-    {
-      "epoch": 1.3510015408320493,
-      "grad_norm": 0.5294694304466248,
-      "learning_rate": 2.8899480069324087e-07,
-      "loss": 2.0928,
-      "step": 548
-    },
-    {
-      "epoch": 1.35346687211094,
-      "grad_norm": 0.5630142688751221,
-      "learning_rate": 2.8856152512998267e-07,
-      "loss": 2.0451,
-      "step": 549
-    },
-    {
-      "epoch": 1.3559322033898304,
-      "grad_norm": 0.517575740814209,
-      "learning_rate": 2.881282495667244e-07,
-      "loss": 2.0556,
-      "step": 550
-    },
-    {
-      "epoch": 1.3583975346687212,
-      "grad_norm": 0.5253962874412537,
-      "learning_rate": 2.876949740034662e-07,
-      "loss": 2.1088,
-      "step": 551
-    },
-    {
-      "epoch": 1.3608628659476116,
-      "grad_norm": 0.5190132260322571,
-      "learning_rate": 2.8726169844020796e-07,
-      "loss": 2.0459,
-      "step": 552
-    },
-    {
-      "epoch": 1.3633281972265023,
-      "grad_norm": 0.5542247891426086,
-      "learning_rate": 2.8682842287694976e-07,
-      "loss": 2.0736,
-      "step": 553
-    },
-    {
-      "epoch": 1.3657935285053928,
-      "grad_norm": 0.5446171164512634,
-      "learning_rate": 2.863951473136915e-07,
-      "loss": 2.0366,
-      "step": 554
-    },
-    {
-      "epoch": 1.3682588597842835,
-      "grad_norm": 0.5620147585868835,
-      "learning_rate": 2.8596187175043325e-07,
-      "loss": 2.0593,
-      "step": 555
-    },
-    {
-      "epoch": 1.3707241910631742,
-      "grad_norm": 0.5434351563453674,
-      "learning_rate": 2.85528596187175e-07,
-      "loss": 2.0697,
-      "step": 556
-    },
-    {
-      "epoch": 1.3731895223420647,
-      "grad_norm": 0.5291680693626404,
-      "learning_rate": 2.850953206239168e-07,
-      "loss": 2.117,
-      "step": 557
-    },
-    {
-      "epoch": 1.3756548536209552,
-      "grad_norm": 0.5382573008537292,
-      "learning_rate": 2.8466204506065853e-07,
-      "loss": 2.0755,
-      "step": 558
-    },
-    {
-      "epoch": 1.378120184899846,
-      "grad_norm": 0.5260967016220093,
-      "learning_rate": 2.8422876949740033e-07,
-      "loss": 2.0305,
-      "step": 559
-    },
-    {
-      "epoch": 1.3805855161787366,
-      "grad_norm": 0.5255964994430542,
-      "learning_rate": 2.837954939341421e-07,
-      "loss": 2.0419,
-      "step": 560
-    },
-    {
-      "epoch": 1.383050847457627,
-      "grad_norm": 0.541347086429596,
-      "learning_rate": 2.833622183708839e-07,
-      "loss": 2.0988,
-      "step": 561
-    },
-    {
-      "epoch": 1.3855161787365178,
-      "grad_norm": 0.5098375082015991,
-      "learning_rate": 2.829289428076256e-07,
-      "loss": 2.0677,
-      "step": 562
-    },
-    {
-      "epoch": 1.3879815100154083,
-      "grad_norm": 0.5496364235877991,
-      "learning_rate": 2.824956672443674e-07,
-      "loss": 2.0748,
-      "step": 563
-    },
-    {
-      "epoch": 1.390446841294299,
-      "grad_norm": 0.5202908515930176,
-      "learning_rate": 2.8206239168110916e-07,
-      "loss": 2.1031,
-      "step": 564
-    },
-    {
-      "epoch": 1.3929121725731894,
-      "grad_norm": 0.5658861398696899,
-      "learning_rate": 2.8162911611785096e-07,
-      "loss": 2.1063,
-      "step": 565
-    },
-    {
-      "epoch": 1.3953775038520801,
-      "grad_norm": 0.5813626646995544,
-      "learning_rate": 2.811958405545927e-07,
-      "loss": 2.0727,
-      "step": 566
-    },
-    {
-      "epoch": 1.3978428351309709,
-      "grad_norm": 0.5407326817512512,
-      "learning_rate": 2.807625649913345e-07,
-      "loss": 2.1281,
-      "step": 567
-    },
-    {
-      "epoch": 1.4003081664098613,
-      "grad_norm": 0.5366054177284241,
-      "learning_rate": 2.8032928942807625e-07,
-      "loss": 2.0793,
-      "step": 568
-    },
-    {
-      "epoch": 1.4027734976887518,
-      "grad_norm": 0.5454252362251282,
-      "learning_rate": 2.7989601386481805e-07,
-      "loss": 2.0573,
-      "step": 569
-    },
-    {
-      "epoch": 1.4052388289676425,
-      "grad_norm": 0.5705124735832214,
-      "learning_rate": 2.7946273830155974e-07,
-      "loss": 2.0594,
-      "step": 570
-    },
-    {
-      "epoch": 1.4077041602465332,
-      "grad_norm": 0.5513876676559448,
-      "learning_rate": 2.7902946273830154e-07,
-      "loss": 2.0789,
-      "step": 571
-    },
-    {
-      "epoch": 1.4101694915254237,
-      "grad_norm": 0.5202826261520386,
-      "learning_rate": 2.785961871750433e-07,
-      "loss": 2.0809,
-      "step": 572
-    },
-    {
-      "epoch": 1.4126348228043144,
-      "grad_norm": 0.5438618659973145,
-      "learning_rate": 2.781629116117851e-07,
-      "loss": 2.1247,
-      "step": 573
-    },
-    {
-      "epoch": 1.4151001540832049,
-      "grad_norm": 0.5352728962898254,
-      "learning_rate": 2.777296360485268e-07,
-      "loss": 2.0723,
-      "step": 574
-    },
-    {
-      "epoch": 1.4175654853620956,
-      "grad_norm": 0.5342223048210144,
-      "learning_rate": 2.772963604852686e-07,
-      "loss": 2.0088,
-      "step": 575
-    },
-    {
-      "epoch": 1.420030816640986,
-      "grad_norm": 0.556289553642273,
-      "learning_rate": 2.7686308492201037e-07,
-      "loss": 2.088,
-      "step": 576
-    },
-    {
-      "epoch": 1.4224961479198768,
-      "grad_norm": 0.5240088105201721,
-      "learning_rate": 2.7642980935875217e-07,
-      "loss": 2.05,
-      "step": 577
-    },
-    {
-      "epoch": 1.4249614791987673,
-      "grad_norm": 0.5449855327606201,
-      "learning_rate": 2.759965337954939e-07,
-      "loss": 2.0436,
-      "step": 578
-    },
-    {
-      "epoch": 1.427426810477658,
-      "grad_norm": 0.5392140746116638,
-      "learning_rate": 2.755632582322357e-07,
-      "loss": 2.0478,
-      "step": 579
-    },
-    {
-      "epoch": 1.4298921417565484,
-      "grad_norm": 0.5651275515556335,
-      "learning_rate": 2.7512998266897745e-07,
-      "loss": 2.0421,
-      "step": 580
-    },
-    {
-      "epoch": 1.4323574730354391,
-      "grad_norm": 0.5078974366188049,
-      "learning_rate": 2.7469670710571925e-07,
-      "loss": 2.0267,
-      "step": 581
-    },
-    {
-      "epoch": 1.4348228043143298,
-      "grad_norm": 0.5486698150634766,
-      "learning_rate": 2.74263431542461e-07,
-      "loss": 2.0553,
-      "step": 582
-    },
-    {
-      "epoch": 1.4372881355932203,
-      "grad_norm": 0.5565054416656494,
-      "learning_rate": 2.738301559792028e-07,
-      "loss": 2.1163,
-      "step": 583
-    },
-    {
-      "epoch": 1.4397534668721108,
-      "grad_norm": 0.5276550650596619,
-      "learning_rate": 2.7339688041594454e-07,
-      "loss": 2.0427,
-      "step": 584
-    },
-    {
-      "epoch": 1.4422187981510015,
-      "grad_norm": 0.5744592547416687,
-      "learning_rate": 2.729636048526863e-07,
-      "loss": 2.1292,
-      "step": 585
-    },
-    {
-      "epoch": 1.4446841294298922,
-      "grad_norm": 0.5316488146781921,
-      "learning_rate": 2.7253032928942803e-07,
-      "loss": 2.0663,
-      "step": 586
-    },
-    {
-      "epoch": 1.4471494607087827,
-      "grad_norm": 0.562899649143219,
-      "learning_rate": 2.7209705372616983e-07,
-      "loss": 2.1515,
-      "step": 587
-    },
-    {
-      "epoch": 1.4496147919876734,
-      "grad_norm": 0.5735856294631958,
-      "learning_rate": 2.7166377816291157e-07,
-      "loss": 2.0883,
-      "step": 588
-    },
-    {
-      "epoch": 1.4520801232665639,
-      "grad_norm": 0.5496561527252197,
-      "learning_rate": 2.7123050259965337e-07,
-      "loss": 2.069,
-      "step": 589
-    },
-    {
-      "epoch": 1.4545454545454546,
-      "grad_norm": 0.5434293150901794,
-      "learning_rate": 2.707972270363951e-07,
-      "loss": 2.0468,
-      "step": 590
-    },
-    {
-      "epoch": 1.457010785824345,
-      "grad_norm": 0.5591951012611389,
-      "learning_rate": 2.703639514731369e-07,
-      "loss": 2.0979,
-      "step": 591
-    },
-    {
-      "epoch": 1.4594761171032358,
-      "grad_norm": 0.556535542011261,
-      "learning_rate": 2.6993067590987866e-07,
-      "loss": 2.104,
-      "step": 592
-    },
-    {
-      "epoch": 1.4619414483821265,
-      "grad_norm": 0.5265719890594482,
-      "learning_rate": 2.6949740034662046e-07,
-      "loss": 2.0565,
-      "step": 593
-    },
-    {
-      "epoch": 1.464406779661017,
-      "grad_norm": 0.5520715117454529,
-      "learning_rate": 2.690641247833622e-07,
-      "loss": 2.1049,
-      "step": 594
-    },
-    {
-      "epoch": 1.4668721109399074,
-      "grad_norm": 0.5084496736526489,
-      "learning_rate": 2.68630849220104e-07,
-      "loss": 2.0027,
-      "step": 595
-    },
-    {
-      "epoch": 1.4693374422187981,
-      "grad_norm": 0.5238492488861084,
-      "learning_rate": 2.6819757365684574e-07,
-      "loss": 2.019,
-      "step": 596
-    },
-    {
-      "epoch": 1.4718027734976888,
-      "grad_norm": 0.5607398748397827,
-      "learning_rate": 2.6776429809358754e-07,
-      "loss": 2.097,
-      "step": 597
-    },
-    {
-      "epoch": 1.4742681047765793,
-      "grad_norm": 0.5568619966506958,
-      "learning_rate": 2.673310225303293e-07,
-      "loss": 2.0772,
-      "step": 598
-    },
-    {
-      "epoch": 1.47673343605547,
-      "grad_norm": 0.553952157497406,
-      "learning_rate": 2.668977469670711e-07,
-      "loss": 2.0929,
-      "step": 599
-    },
-    {
-      "epoch": 1.4791987673343605,
-      "grad_norm": 0.5402290225028992,
-      "learning_rate": 2.6646447140381283e-07,
-      "loss": 2.062,
-      "step": 600
-    },
-    {
-      "epoch": 1.4816640986132512,
-      "grad_norm": 0.5399553179740906,
-      "learning_rate": 2.660311958405546e-07,
-      "loss": 2.0673,
-      "step": 601
-    },
-    {
-      "epoch": 1.4841294298921417,
-      "grad_norm": 0.5668056011199951,
-      "learning_rate": 2.655979202772963e-07,
-      "loss": 2.0596,
-      "step": 602
-    },
-    {
-      "epoch": 1.4865947611710324,
-      "grad_norm": 0.5426621437072754,
-      "learning_rate": 2.651646447140381e-07,
-      "loss": 2.0718,
-      "step": 603
-    },
-    {
-      "epoch": 1.4890600924499229,
-      "grad_norm": 0.5283502340316772,
-      "learning_rate": 2.6473136915077986e-07,
-      "loss": 2.0544,
-      "step": 604
-    },
-    {
-      "epoch": 1.4915254237288136,
-      "grad_norm": 0.5753063559532166,
-      "learning_rate": 2.6429809358752166e-07,
-      "loss": 2.1082,
-      "step": 605
-    },
-    {
-      "epoch": 1.493990755007704,
-      "grad_norm": 0.5240095257759094,
-      "learning_rate": 2.638648180242634e-07,
-      "loss": 2.054,
-      "step": 606
-    },
-    {
-      "epoch": 1.4964560862865948,
-      "grad_norm": 0.5336340069770813,
-      "learning_rate": 2.634315424610052e-07,
-      "loss": 2.0585,
-      "step": 607
-    },
-    {
-      "epoch": 1.4989214175654855,
-      "grad_norm": 0.5449371337890625,
-      "learning_rate": 2.6299826689774695e-07,
-      "loss": 2.0645,
-      "step": 608
-    },
-    {
-      "epoch": 1.501386748844376,
-      "grad_norm": 0.553382396697998,
-      "learning_rate": 2.6256499133448875e-07,
-      "loss": 2.0974,
-      "step": 609
-    },
-    {
-      "epoch": 1.5038520801232664,
-      "grad_norm": 0.5372682809829712,
-      "learning_rate": 2.621317157712305e-07,
-      "loss": 2.1287,
-      "step": 610
-    },
-    {
-      "epoch": 1.5063174114021571,
-      "grad_norm": 0.5454490780830383,
-      "learning_rate": 2.616984402079723e-07,
-      "loss": 2.0699,
-      "step": 611
-    },
-    {
-      "epoch": 1.5087827426810478,
-      "grad_norm": 0.5598608255386353,
-      "learning_rate": 2.6126516464471404e-07,
-      "loss": 2.1242,
-      "step": 612
-    },
-    {
-      "epoch": 1.5112480739599383,
-      "grad_norm": 0.5377222895622253,
-      "learning_rate": 2.6083188908145583e-07,
-      "loss": 2.078,
-      "step": 613
-    },
-    {
-      "epoch": 1.5137134052388288,
-      "grad_norm": 0.549659013748169,
-      "learning_rate": 2.603986135181976e-07,
-      "loss": 2.077,
-      "step": 614
-    },
-    {
-      "epoch": 1.5161787365177197,
-      "grad_norm": 0.5235285758972168,
-      "learning_rate": 2.599653379549394e-07,
-      "loss": 2.0384,
-      "step": 615
-    },
-    {
-      "epoch": 1.5186440677966102,
-      "grad_norm": 0.5353394746780396,
-      "learning_rate": 2.5953206239168107e-07,
-      "loss": 2.084,
-      "step": 616
-    },
-    {
-      "epoch": 1.5211093990755007,
-      "grad_norm": 0.5501331686973572,
-      "learning_rate": 2.5909878682842287e-07,
-      "loss": 2.0653,
-      "step": 617
-    },
-    {
-      "epoch": 1.5235747303543914,
-      "grad_norm": 0.5552558898925781,
-      "learning_rate": 2.586655112651646e-07,
-      "loss": 2.0984,
-      "step": 618
-    },
-    {
-      "epoch": 1.526040061633282,
-      "grad_norm": 0.5290137529373169,
-      "learning_rate": 2.582322357019064e-07,
-      "loss": 2.101,
-      "step": 619
-    },
-    {
-      "epoch": 1.5285053929121726,
-      "grad_norm": 0.5513224005699158,
-      "learning_rate": 2.5779896013864815e-07,
-      "loss": 2.0511,
-      "step": 620
-    },
-    {
-      "epoch": 1.530970724191063,
-      "grad_norm": 0.525702714920044,
-      "learning_rate": 2.5736568457538995e-07,
-      "loss": 2.1566,
-      "step": 621
-    },
-    {
-      "epoch": 1.5334360554699538,
-      "grad_norm": 0.5237429738044739,
-      "learning_rate": 2.569324090121317e-07,
-      "loss": 2.0632,
-      "step": 622
-    },
-    {
-      "epoch": 1.5359013867488445,
-      "grad_norm": 0.5647119879722595,
-      "learning_rate": 2.564991334488735e-07,
-      "loss": 2.0885,
-      "step": 623
-    },
-    {
-      "epoch": 1.538366718027735,
-      "grad_norm": 0.5435053706169128,
-      "learning_rate": 2.5606585788561524e-07,
-      "loss": 2.063,
-      "step": 624
-    },
-    {
-      "epoch": 1.5408320493066254,
-      "grad_norm": 0.5810684561729431,
-      "learning_rate": 2.5563258232235704e-07,
-      "loss": 2.1319,
-      "step": 625
-    },
-    {
-      "epoch": 1.5432973805855161,
-      "grad_norm": 0.5526551008224487,
-      "learning_rate": 2.551993067590988e-07,
-      "loss": 2.1514,
-      "step": 626
-    },
-    {
-      "epoch": 1.5457627118644068,
-      "grad_norm": 0.5331485271453857,
-      "learning_rate": 2.547660311958406e-07,
-      "loss": 1.982,
-      "step": 627
-    },
-    {
-      "epoch": 1.5482280431432973,
-      "grad_norm": 0.5447163581848145,
-      "learning_rate": 2.5433275563258233e-07,
-      "loss": 2.0662,
-      "step": 628
-    },
-    {
-      "epoch": 1.550693374422188,
-      "grad_norm": 0.5696806311607361,
-      "learning_rate": 2.538994800693241e-07,
-      "loss": 2.0573,
-      "step": 629
-    },
-    {
-      "epoch": 1.5531587057010787,
-      "grad_norm": 0.5400473475456238,
-      "learning_rate": 2.5346620450606587e-07,
-      "loss": 2.0772,
-      "step": 630
-    },
-    {
-      "epoch": 1.5556240369799692,
-      "grad_norm": 0.543637752532959,
-      "learning_rate": 2.530329289428076e-07,
-      "loss": 2.0869,
-      "step": 631
-    },
-    {
-      "epoch": 1.5580893682588597,
-      "grad_norm": 0.5363517999649048,
-      "learning_rate": 2.5259965337954936e-07,
-      "loss": 2.106,
-      "step": 632
-    },
-    {
-      "epoch": 1.5605546995377504,
-      "grad_norm": 0.5055550932884216,
-      "learning_rate": 2.5216637781629116e-07,
-      "loss": 2.0274,
-      "step": 633
-    },
-    {
-      "epoch": 1.563020030816641,
-      "grad_norm": 0.5467242002487183,
-      "learning_rate": 2.517331022530329e-07,
-      "loss": 2.0927,
-      "step": 634
-    },
-    {
-      "epoch": 1.5654853620955316,
-      "grad_norm": 0.5361837148666382,
-      "learning_rate": 2.512998266897747e-07,
-      "loss": 2.1079,
-      "step": 635
-    },
-    {
-      "epoch": 1.567950693374422,
-      "grad_norm": 0.5386531949043274,
-      "learning_rate": 2.5086655112651645e-07,
-      "loss": 2.0499,
-      "step": 636
-    },
-    {
-      "epoch": 1.5704160246533128,
-      "grad_norm": 0.5682772994041443,
-      "learning_rate": 2.5043327556325824e-07,
-      "loss": 2.1246,
-      "step": 637
-    },
-    {
-      "epoch": 1.5728813559322035,
-      "grad_norm": 0.5458372831344604,
-      "learning_rate": 2.5e-07,
-      "loss": 2.1237,
-      "step": 638
-    },
-    {
-      "epoch": 1.575346687211094,
-      "grad_norm": 0.5346107482910156,
-      "learning_rate": 2.4956672443674173e-07,
-      "loss": 2.0898,
-      "step": 639
-    },
-    {
-      "epoch": 1.5778120184899846,
-      "grad_norm": 0.5748855471611023,
-      "learning_rate": 2.4913344887348353e-07,
-      "loss": 2.1254,
-      "step": 640
-    },
-    {
-      "epoch": 1.5802773497688754,
-      "grad_norm": 0.5122947692871094,
-      "learning_rate": 2.487001733102253e-07,
-      "loss": 2.057,
-      "step": 641
-    },
-    {
-      "epoch": 1.5827426810477658,
-      "grad_norm": 0.5241718888282776,
-      "learning_rate": 2.48266897746967e-07,
-      "loss": 2.0911,
-      "step": 642
-    },
-    {
-      "epoch": 1.5852080123266563,
-      "grad_norm": 0.5585358142852783,
-      "learning_rate": 2.478336221837088e-07,
-      "loss": 2.0615,
-      "step": 643
-    },
-    {
-      "epoch": 1.587673343605547,
-      "grad_norm": 0.6165497899055481,
-      "learning_rate": 2.4740034662045056e-07,
-      "loss": 2.179,
-      "step": 644
-    },
-    {
-      "epoch": 1.5901386748844377,
-      "grad_norm": 0.5242385864257812,
-      "learning_rate": 2.4696707105719236e-07,
-      "loss": 2.0643,
-      "step": 645
-    },
-    {
-      "epoch": 1.5926040061633282,
-      "grad_norm": 0.5357149243354797,
-      "learning_rate": 2.465337954939341e-07,
-      "loss": 2.0407,
-      "step": 646
-    },
-    {
-      "epoch": 1.5950693374422187,
-      "grad_norm": 0.525974452495575,
-      "learning_rate": 2.461005199306759e-07,
-      "loss": 2.0345,
-      "step": 647
-    },
-    {
-      "epoch": 1.5975346687211094,
-      "grad_norm": 0.5614073872566223,
-      "learning_rate": 2.4566724436741765e-07,
-      "loss": 2.0699,
-      "step": 648
-    },
-    {
-      "epoch": 1.6,
-      "grad_norm": 0.526637613773346,
-      "learning_rate": 2.4523396880415945e-07,
-      "loss": 2.0783,
-      "step": 649
-    },
-    {
-      "epoch": 1.6024653312788906,
-      "grad_norm": 0.5309383869171143,
-      "learning_rate": 2.448006932409012e-07,
-      "loss": 2.0257,
-      "step": 650
-    },
-    {
-      "epoch": 1.604930662557781,
-      "grad_norm": 0.5437653660774231,
-      "learning_rate": 2.4436741767764294e-07,
-      "loss": 2.073,
-      "step": 651
-    },
-    {
-      "epoch": 1.6073959938366718,
-      "grad_norm": 0.5697052478790283,
-      "learning_rate": 2.4393414211438474e-07,
-      "loss": 2.0774,
-      "step": 652
-    },
-    {
-      "epoch": 1.6098613251155625,
-      "grad_norm": 0.5350382328033447,
-      "learning_rate": 2.435008665511265e-07,
-      "loss": 2.0817,
-      "step": 653
-    },
-    {
-      "epoch": 1.612326656394453,
-      "grad_norm": 0.5554140210151672,
-      "learning_rate": 2.430675909878683e-07,
-      "loss": 2.0842,
-      "step": 654
-    },
-    {
-      "epoch": 1.6147919876733436,
-      "grad_norm": 0.5539381504058838,
-      "learning_rate": 2.4263431542461e-07,
-      "loss": 2.0684,
-      "step": 655
-    },
-    {
-      "epoch": 1.6172573189522343,
-      "grad_norm": 0.5573647618293762,
-      "learning_rate": 2.422010398613518e-07,
-      "loss": 2.0848,
-      "step": 656
-    },
-    {
-      "epoch": 1.6197226502311248,
-      "grad_norm": 0.5368602871894836,
-      "learning_rate": 2.4176776429809357e-07,
-      "loss": 2.0463,
-      "step": 657
-    },
-    {
-      "epoch": 1.6221879815100153,
-      "grad_norm": 0.5693660974502563,
-      "learning_rate": 2.413344887348353e-07,
-      "loss": 2.1469,
-      "step": 658
-    },
-    {
-      "epoch": 1.624653312788906,
-      "grad_norm": 0.5438942909240723,
-      "learning_rate": 2.409012131715771e-07,
-      "loss": 2.0725,
-      "step": 659
-    },
-    {
-      "epoch": 1.6271186440677967,
-      "grad_norm": 0.5400587916374207,
-      "learning_rate": 2.4046793760831886e-07,
-      "loss": 2.1452,
-      "step": 660
-    },
-    {
-      "epoch": 1.6295839753466872,
-      "grad_norm": 0.5040330290794373,
-      "learning_rate": 2.4003466204506065e-07,
-      "loss": 2.0522,
-      "step": 661
-    },
-    {
-      "epoch": 1.6320493066255777,
-      "grad_norm": 0.5188309550285339,
-      "learning_rate": 2.396013864818024e-07,
-      "loss": 2.0538,
-      "step": 662
-    },
-    {
-      "epoch": 1.6345146379044684,
-      "grad_norm": 0.5574179887771606,
-      "learning_rate": 2.391681109185442e-07,
-      "loss": 2.1391,
-      "step": 663
-    },
-    {
-      "epoch": 1.636979969183359,
-      "grad_norm": 0.5268549919128418,
-      "learning_rate": 2.3873483535528594e-07,
-      "loss": 2.0742,
-      "step": 664
-    },
-    {
-      "epoch": 1.6394453004622496,
-      "grad_norm": 0.5314316749572754,
-      "learning_rate": 2.3830155979202771e-07,
-      "loss": 2.0721,
-      "step": 665
-    },
-    {
-      "epoch": 1.6419106317411403,
-      "grad_norm": 0.5841025710105896,
-      "learning_rate": 2.3786828422876948e-07,
-      "loss": 2.124,
-      "step": 666
-    },
-    {
-      "epoch": 1.644375963020031,
-      "grad_norm": 0.5572120547294617,
-      "learning_rate": 2.3743500866551126e-07,
-      "loss": 2.0926,
-      "step": 667
-    },
-    {
-      "epoch": 1.6468412942989215,
-      "grad_norm": 0.5612335801124573,
-      "learning_rate": 2.3700173310225303e-07,
-      "loss": 2.07,
-      "step": 668
-    },
-    {
-      "epoch": 1.649306625577812,
-      "grad_norm": 0.545322835445404,
-      "learning_rate": 2.365684575389948e-07,
-      "loss": 2.0675,
-      "step": 669
-    },
-    {
-      "epoch": 1.6517719568567026,
-      "grad_norm": 0.5538606643676758,
-      "learning_rate": 2.3613518197573657e-07,
-      "loss": 2.0667,
-      "step": 670
-    },
-    {
-      "epoch": 1.6542372881355933,
-      "grad_norm": 0.5166656374931335,
-      "learning_rate": 2.3570190641247834e-07,
-      "loss": 2.0332,
-      "step": 671
-    },
-    {
-      "epoch": 1.6567026194144838,
-      "grad_norm": 0.5593602061271667,
-      "learning_rate": 2.3526863084922011e-07,
-      "loss": 2.1419,
-      "step": 672
-    },
-    {
-      "epoch": 1.6591679506933743,
-      "grad_norm": 0.5563687086105347,
-      "learning_rate": 2.3483535528596186e-07,
-      "loss": 2.0222,
-      "step": 673
-    },
-    {
-      "epoch": 1.661633281972265,
-      "grad_norm": 0.5505043864250183,
-      "learning_rate": 2.3440207972270363e-07,
-      "loss": 2.0949,
-      "step": 674
-    },
-    {
-      "epoch": 1.6640986132511557,
-      "grad_norm": 0.49905431270599365,
-      "learning_rate": 2.339688041594454e-07,
-      "loss": 2.0702,
-      "step": 675
-    },
-    {
-      "epoch": 1.6665639445300462,
-      "grad_norm": 0.5658527612686157,
-      "learning_rate": 2.3353552859618717e-07,
-      "loss": 2.0925,
-      "step": 676
-    },
-    {
-      "epoch": 1.6690292758089367,
-      "grad_norm": 0.5240568518638611,
-      "learning_rate": 2.3310225303292894e-07,
-      "loss": 2.0719,
-      "step": 677
-    },
-    {
-      "epoch": 1.6714946070878274,
-      "grad_norm": 0.5115277767181396,
-      "learning_rate": 2.3266897746967072e-07,
-      "loss": 2.0976,
-      "step": 678
-    },
-    {
-      "epoch": 1.673959938366718,
-      "grad_norm": 0.5566647052764893,
-      "learning_rate": 2.322357019064125e-07,
-      "loss": 2.0161,
-      "step": 679
-    },
-    {
-      "epoch": 1.6764252696456086,
-      "grad_norm": 0.5448424816131592,
-      "learning_rate": 2.3180242634315423e-07,
-      "loss": 2.1174,
-      "step": 680
-    },
-    {
-      "epoch": 1.6788906009244993,
-      "grad_norm": 0.5539455413818359,
-      "learning_rate": 2.31369150779896e-07,
-      "loss": 2.0338,
-      "step": 681
-    },
-    {
-      "epoch": 1.68135593220339,
-      "grad_norm": 0.5492231845855713,
-      "learning_rate": 2.3093587521663778e-07,
-      "loss": 2.1217,
-      "step": 682
-    },
-    {
-      "epoch": 1.6838212634822805,
-      "grad_norm": 0.5186160206794739,
-      "learning_rate": 2.3050259965337955e-07,
-      "loss": 2.0538,
-      "step": 683
-    },
-    {
-      "epoch": 1.686286594761171,
-      "grad_norm": 0.5424460768699646,
-      "learning_rate": 2.3006932409012132e-07,
-      "loss": 2.0764,
-      "step": 684
-    },
-    {
-      "epoch": 1.6887519260400616,
-      "grad_norm": 0.5673890709877014,
-      "learning_rate": 2.296360485268631e-07,
-      "loss": 2.1029,
-      "step": 685
-    },
-    {
-      "epoch": 1.6912172573189523,
-      "grad_norm": 0.5408136248588562,
-      "learning_rate": 2.2920277296360486e-07,
-      "loss": 2.1321,
-      "step": 686
-    },
-    {
-      "epoch": 1.6936825885978428,
-      "grad_norm": 0.5323325991630554,
-      "learning_rate": 2.2876949740034663e-07,
-      "loss": 2.1003,
-      "step": 687
-    },
-    {
-      "epoch": 1.6961479198767333,
-      "grad_norm": 0.5330997109413147,
-      "learning_rate": 2.2833622183708838e-07,
-      "loss": 2.0747,
-      "step": 688
-    },
-    {
-      "epoch": 1.698613251155624,
-      "grad_norm": 0.5504052042961121,
-      "learning_rate": 2.2790294627383015e-07,
-      "loss": 2.1067,
-      "step": 689
-    },
-    {
-      "epoch": 1.7010785824345147,
-      "grad_norm": 0.5399577617645264,
-      "learning_rate": 2.2746967071057192e-07,
-      "loss": 2.0032,
-      "step": 690
-    },
-    {
-      "epoch": 1.7035439137134052,
-      "grad_norm": 0.5435107350349426,
-      "learning_rate": 2.270363951473137e-07,
-      "loss": 2.0389,
-      "step": 691
-    },
-    {
-      "epoch": 1.706009244992296,
-      "grad_norm": 0.5398039221763611,
-      "learning_rate": 2.2660311958405546e-07,
-      "loss": 2.0745,
-      "step": 692
-    },
-    {
-      "epoch": 1.7084745762711866,
-      "grad_norm": 0.5262869000434875,
-      "learning_rate": 2.2616984402079724e-07,
-      "loss": 2.0557,
-      "step": 693
-    },
-    {
-      "epoch": 1.710939907550077,
-      "grad_norm": 0.5872994661331177,
-      "learning_rate": 2.25736568457539e-07,
-      "loss": 2.1679,
-      "step": 694
-    },
-    {
-      "epoch": 1.7134052388289676,
-      "grad_norm": 0.526508092880249,
-      "learning_rate": 2.2530329289428073e-07,
-      "loss": 2.088,
-      "step": 695
-    },
-    {
-      "epoch": 1.7158705701078583,
-      "grad_norm": 0.5324400663375854,
-      "learning_rate": 2.248700173310225e-07,
-      "loss": 2.0343,
-      "step": 696
-    },
-    {
-      "epoch": 1.718335901386749,
-      "grad_norm": 0.5571790933609009,
-      "learning_rate": 2.2443674176776427e-07,
-      "loss": 2.1619,
-      "step": 697
-    },
-    {
-      "epoch": 1.7208012326656394,
-      "grad_norm": 0.5518919825553894,
-      "learning_rate": 2.2400346620450604e-07,
-      "loss": 2.1053,
-      "step": 698
-    },
-    {
-      "epoch": 1.72326656394453,
-      "grad_norm": 0.5263517498970032,
-      "learning_rate": 2.235701906412478e-07,
-      "loss": 2.1048,
-      "step": 699
-    },
-    {
-      "epoch": 1.7257318952234206,
-      "grad_norm": 0.5402522087097168,
-      "learning_rate": 2.2313691507798958e-07,
-      "loss": 2.1042,
-      "step": 700
-    },
-    {
-      "epoch": 1.7281972265023113,
-      "grad_norm": 0.527553141117096,
-      "learning_rate": 2.2270363951473135e-07,
-      "loss": 2.0133,
-      "step": 701
-    },
-    {
-      "epoch": 1.7306625577812018,
-      "grad_norm": 0.5525938868522644,
-      "learning_rate": 2.2227036395147313e-07,
-      "loss": 2.0717,
-      "step": 702
-    },
-    {
-      "epoch": 1.7331278890600923,
-      "grad_norm": 0.49240729212760925,
-      "learning_rate": 2.2183708838821487e-07,
-      "loss": 2.0334,
-      "step": 703
-    },
-    {
-      "epoch": 1.735593220338983,
-      "grad_norm": 0.5105845928192139,
-      "learning_rate": 2.2140381282495664e-07,
-      "loss": 2.0768,
-      "step": 704
-    },
-    {
-      "epoch": 1.7380585516178737,
-      "grad_norm": 0.5725539326667786,
-      "learning_rate": 2.2097053726169841e-07,
-      "loss": 2.0955,
-      "step": 705
-    },
-    {
-      "epoch": 1.7405238828967642,
-      "grad_norm": 0.5486234426498413,
-      "learning_rate": 2.2053726169844019e-07,
-      "loss": 2.0214,
-      "step": 706
-    },
-    {
-      "epoch": 1.7429892141756549,
-      "grad_norm": 0.5501362681388855,
-      "learning_rate": 2.2010398613518196e-07,
-      "loss": 2.1141,
-      "step": 707
-    },
-    {
-      "epoch": 1.7454545454545456,
-      "grad_norm": 0.543329656124115,
-      "learning_rate": 2.1967071057192373e-07,
-      "loss": 2.0873,
-      "step": 708
-    },
-    {
-      "epoch": 1.747919876733436,
-      "grad_norm": 0.5453404188156128,
-      "learning_rate": 2.192374350086655e-07,
-      "loss": 2.1165,
-      "step": 709
-    },
-    {
-      "epoch": 1.7503852080123266,
-      "grad_norm": 0.5702621936798096,
-      "learning_rate": 2.1880415944540727e-07,
-      "loss": 2.1072,
-      "step": 710
-    },
-    {
-      "epoch": 1.7528505392912173,
-      "grad_norm": 0.5359793305397034,
-      "learning_rate": 2.1837088388214902e-07,
-      "loss": 2.0051,
-      "step": 711
-    },
-    {
-      "epoch": 1.755315870570108,
-      "grad_norm": 0.5496872663497925,
-      "learning_rate": 2.179376083188908e-07,
-      "loss": 2.0871,
-      "step": 712
-    },
-    {
-      "epoch": 1.7577812018489984,
-      "grad_norm": 0.5678308606147766,
-      "learning_rate": 2.1750433275563256e-07,
-      "loss": 2.0548,
-      "step": 713
-    },
-    {
-      "epoch": 1.760246533127889,
-      "grad_norm": 0.546625554561615,
-      "learning_rate": 2.1707105719237433e-07,
-      "loss": 2.1175,
-      "step": 714
-    },
-    {
-      "epoch": 1.7627118644067796,
-      "grad_norm": 0.5344644784927368,
-      "learning_rate": 2.166377816291161e-07,
-      "loss": 2.0468,
-      "step": 715
-    },
-    {
-      "epoch": 1.7651771956856703,
-      "grad_norm": 0.5369846820831299,
-      "learning_rate": 2.1620450606585787e-07,
-      "loss": 2.0596,
-      "step": 716
-    },
-    {
-      "epoch": 1.7676425269645608,
-      "grad_norm": 0.5267930626869202,
-      "learning_rate": 2.1577123050259965e-07,
-      "loss": 2.0495,
-      "step": 717
-    },
-    {
-      "epoch": 1.7701078582434515,
-      "grad_norm": 0.5174152255058289,
-      "learning_rate": 2.153379549393414e-07,
-      "loss": 2.0451,
-      "step": 718
-    },
-    {
-      "epoch": 1.7725731895223422,
-      "grad_norm": 0.5533501505851746,
-      "learning_rate": 2.1490467937608316e-07,
-      "loss": 2.0884,
-      "step": 719
-    },
-    {
-      "epoch": 1.7750385208012327,
-      "grad_norm": 0.5463818907737732,
-      "learning_rate": 2.1447140381282493e-07,
-      "loss": 2.1242,
-      "step": 720
-    },
-    {
-      "epoch": 1.7775038520801232,
-      "grad_norm": 0.5650577545166016,
-      "learning_rate": 2.140381282495667e-07,
-      "loss": 2.0692,
-      "step": 721
-    },
-    {
-      "epoch": 1.7799691833590139,
-      "grad_norm": 0.55766361951828,
-      "learning_rate": 2.1360485268630848e-07,
-      "loss": 2.0454,
-      "step": 722
-    },
-    {
-      "epoch": 1.7824345146379046,
-      "grad_norm": 0.5192793607711792,
-      "learning_rate": 2.1317157712305025e-07,
-      "loss": 2.0684,
-      "step": 723
-    },
-    {
-      "epoch": 1.784899845916795,
-      "grad_norm": 0.5581333041191101,
-      "learning_rate": 2.1273830155979202e-07,
-      "loss": 2.0862,
-      "step": 724
-    },
-    {
-      "epoch": 1.7873651771956856,
-      "grad_norm": 0.5046393275260925,
-      "learning_rate": 2.123050259965338e-07,
-      "loss": 2.0813,
-      "step": 725
-    },
-    {
-      "epoch": 1.7898305084745763,
-      "grad_norm": 0.537695586681366,
-      "learning_rate": 2.1187175043327554e-07,
-      "loss": 2.0809,
-      "step": 726
-    },
-    {
-      "epoch": 1.792295839753467,
-      "grad_norm": 0.5130316615104675,
-      "learning_rate": 2.114384748700173e-07,
-      "loss": 2.0937,
-      "step": 727
-    },
-    {
-      "epoch": 1.7947611710323574,
-      "grad_norm": 0.535754382610321,
-      "learning_rate": 2.1100519930675908e-07,
-      "loss": 2.0515,
-      "step": 728
-    },
-    {
-      "epoch": 1.797226502311248,
-      "grad_norm": 0.5362329483032227,
-      "learning_rate": 2.1057192374350085e-07,
-      "loss": 2.0982,
-      "step": 729
-    },
-    {
-      "epoch": 1.7996918335901386,
-      "grad_norm": 0.4987218379974365,
-      "learning_rate": 2.1013864818024262e-07,
-      "loss": 1.9768,
-      "step": 730
-    },
-    {
-      "epoch": 1.8021571648690293,
-      "grad_norm": 0.5529645085334778,
-      "learning_rate": 2.097053726169844e-07,
-      "loss": 2.0765,
-      "step": 731
-    },
-    {
-      "epoch": 1.8046224961479198,
-      "grad_norm": 0.5982069373130798,
-      "learning_rate": 2.0927209705372617e-07,
-      "loss": 2.1271,
-      "step": 732
-    },
-    {
-      "epoch": 1.8070878274268105,
-      "grad_norm": 0.5242738723754883,
-      "learning_rate": 2.088388214904679e-07,
-      "loss": 2.0285,
-      "step": 733
-    },
-    {
-      "epoch": 1.8095531587057012,
-      "grad_norm": 0.522243857383728,
-      "learning_rate": 2.0840554592720968e-07,
-      "loss": 2.0701,
-      "step": 734
-    },
-    {
-      "epoch": 1.8120184899845917,
-      "grad_norm": 0.5549618601799011,
-      "learning_rate": 2.0797227036395145e-07,
-      "loss": 2.0853,
-      "step": 735
-    },
-    {
-      "epoch": 1.8144838212634822,
-      "grad_norm": 0.5378853678703308,
-      "learning_rate": 2.0753899480069323e-07,
-      "loss": 2.129,
-      "step": 736
-    },
-    {
-      "epoch": 1.8169491525423729,
-      "grad_norm": 0.5620313286781311,
-      "learning_rate": 2.07105719237435e-07,
-      "loss": 2.0955,
-      "step": 737
-    },
-    {
-      "epoch": 1.8194144838212636,
-      "grad_norm": 0.5175241231918335,
-      "learning_rate": 2.0667244367417677e-07,
-      "loss": 2.0136,
-      "step": 738
-    },
-    {
-      "epoch": 1.821879815100154,
-      "grad_norm": 0.5567770600318909,
-      "learning_rate": 2.0623916811091854e-07,
-      "loss": 2.0785,
-      "step": 739
-    },
-    {
-      "epoch": 1.8243451463790445,
-      "grad_norm": 0.5436593294143677,
-      "learning_rate": 2.058058925476603e-07,
-      "loss": 2.0805,
-      "step": 740
-    },
-    {
-      "epoch": 1.8268104776579352,
-      "grad_norm": 0.5625580549240112,
-      "learning_rate": 2.0537261698440206e-07,
-      "loss": 2.0666,
-      "step": 741
-    },
-    {
-      "epoch": 1.829275808936826,
-      "grad_norm": 0.5570465922355652,
-      "learning_rate": 2.0493934142114383e-07,
-      "loss": 2.0822,
-      "step": 742
-    },
-    {
-      "epoch": 1.8317411402157164,
-      "grad_norm": 0.5406922698020935,
-      "learning_rate": 2.045060658578856e-07,
-      "loss": 2.08,
-      "step": 743
-    },
-    {
-      "epoch": 1.8342064714946071,
-      "grad_norm": 0.5050183534622192,
-      "learning_rate": 2.0407279029462737e-07,
-      "loss": 2.0316,
-      "step": 744
-    },
-    {
-      "epoch": 1.8366718027734978,
-      "grad_norm": 0.5195122361183167,
-      "learning_rate": 2.0363951473136914e-07,
-      "loss": 2.0036,
-      "step": 745
-    },
-    {
-      "epoch": 1.8391371340523883,
-      "grad_norm": 0.5114386677742004,
-      "learning_rate": 2.0320623916811091e-07,
-      "loss": 2.0347,
-      "step": 746
-    },
-    {
-      "epoch": 1.8416024653312788,
-      "grad_norm": 0.556669294834137,
-      "learning_rate": 2.0277296360485269e-07,
-      "loss": 2.0139,
-      "step": 747
-    },
-    {
-      "epoch": 1.8440677966101695,
-      "grad_norm": 0.5159513354301453,
-      "learning_rate": 2.0233968804159446e-07,
-      "loss": 2.0301,
-      "step": 748
-    },
-    {
-      "epoch": 1.8465331278890602,
-      "grad_norm": 0.5236509442329407,
-      "learning_rate": 2.019064124783362e-07,
-      "loss": 1.995,
-      "step": 749
-    },
-    {
-      "epoch": 1.8489984591679507,
-      "grad_norm": 0.5232506990432739,
-      "learning_rate": 2.0147313691507797e-07,
-      "loss": 1.9718,
-      "step": 750
-    },
-    {
-      "epoch": 1.8514637904468412,
-      "grad_norm": 0.5529718399047852,
-      "learning_rate": 2.0103986135181974e-07,
-      "loss": 2.0509,
-      "step": 751
-    },
-    {
-      "epoch": 1.8539291217257319,
-      "grad_norm": 0.5564405918121338,
-      "learning_rate": 2.0060658578856152e-07,
-      "loss": 2.1508,
-      "step": 752
-    },
-    {
-      "epoch": 1.8563944530046226,
-      "grad_norm": 0.5209116339683533,
-      "learning_rate": 2.001733102253033e-07,
-      "loss": 2.0108,
-      "step": 753
-    },
-    {
-      "epoch": 1.858859784283513,
-      "grad_norm": 0.5505017638206482,
-      "learning_rate": 1.9974003466204506e-07,
-      "loss": 2.0589,
-      "step": 754
-    },
-    {
-      "epoch": 1.8613251155624035,
-      "grad_norm": 0.5664639472961426,
-      "learning_rate": 1.9930675909878683e-07,
-      "loss": 2.1018,
-      "step": 755
-    },
-    {
-      "epoch": 1.8637904468412942,
-      "grad_norm": 0.5276853442192078,
-      "learning_rate": 1.9887348353552858e-07,
-      "loss": 2.095,
-      "step": 756
-    },
-    {
-      "epoch": 1.866255778120185,
-      "grad_norm": 0.5520254373550415,
-      "learning_rate": 1.9844020797227035e-07,
-      "loss": 2.1249,
-      "step": 757
-    },
-    {
-      "epoch": 1.8687211093990754,
-      "grad_norm": 0.5317251086235046,
-      "learning_rate": 1.9800693240901212e-07,
-      "loss": 2.0812,
-      "step": 758
-    },
-    {
-      "epoch": 1.8711864406779661,
-      "grad_norm": 0.5425392985343933,
-      "learning_rate": 1.975736568457539e-07,
-      "loss": 2.0785,
-      "step": 759
-    },
-    {
-      "epoch": 1.8736517719568568,
-      "grad_norm": 0.5492266416549683,
-      "learning_rate": 1.9714038128249566e-07,
-      "loss": 2.0745,
-      "step": 760
-    },
-    {
-      "epoch": 1.8761171032357473,
-      "grad_norm": 0.5728384256362915,
-      "learning_rate": 1.9670710571923743e-07,
-      "loss": 2.0847,
-      "step": 761
-    },
-    {
-      "epoch": 1.8785824345146378,
-      "grad_norm": 0.5570703744888306,
-      "learning_rate": 1.962738301559792e-07,
-      "loss": 2.0432,
-      "step": 762
-    },
-    {
-      "epoch": 1.8810477657935285,
-      "grad_norm": 0.5224375128746033,
-      "learning_rate": 1.9584055459272098e-07,
-      "loss": 2.0734,
-      "step": 763
-    },
-    {
-      "epoch": 1.8835130970724192,
-      "grad_norm": 0.5276843905448914,
-      "learning_rate": 1.9540727902946272e-07,
-      "loss": 2.0405,
-      "step": 764
-    },
-    {
-      "epoch": 1.8859784283513097,
-      "grad_norm": 0.521706223487854,
-      "learning_rate": 1.949740034662045e-07,
-      "loss": 2.0807,
-      "step": 765
-    },
-    {
-      "epoch": 1.8884437596302002,
-      "grad_norm": 0.5321924090385437,
-      "learning_rate": 1.9454072790294626e-07,
-      "loss": 1.9975,
-      "step": 766
-    },
-    {
-      "epoch": 1.8909090909090909,
-      "grad_norm": 0.5504716038703918,
-      "learning_rate": 1.9410745233968804e-07,
-      "loss": 2.1148,
-      "step": 767
-    },
-    {
-      "epoch": 1.8933744221879816,
-      "grad_norm": 0.5735941529273987,
-      "learning_rate": 1.936741767764298e-07,
-      "loss": 2.0855,
-      "step": 768
-    },
-    {
-      "epoch": 1.895839753466872,
-      "grad_norm": 0.5594768524169922,
-      "learning_rate": 1.9324090121317158e-07,
-      "loss": 2.0675,
-      "step": 769
-    },
-    {
-      "epoch": 1.8983050847457628,
-      "grad_norm": 0.5534265041351318,
-      "learning_rate": 1.9280762564991335e-07,
-      "loss": 2.059,
-      "step": 770
-    },
-    {
-      "epoch": 1.9007704160246535,
-      "grad_norm": 0.5681132674217224,
-      "learning_rate": 1.923743500866551e-07,
-      "loss": 2.1262,
-      "step": 771
-    },
-    {
-      "epoch": 1.903235747303544,
-      "grad_norm": 0.52858567237854,
-      "learning_rate": 1.9194107452339687e-07,
-      "loss": 2.1065,
-      "step": 772
-    },
-    {
-      "epoch": 1.9057010785824344,
-      "grad_norm": 0.5450841784477234,
-      "learning_rate": 1.9150779896013864e-07,
-      "loss": 2.0728,
-      "step": 773
-    },
-    {
-      "epoch": 1.9081664098613251,
-      "grad_norm": 0.558519184589386,
-      "learning_rate": 1.910745233968804e-07,
-      "loss": 2.0356,
-      "step": 774
-    },
-    {
-      "epoch": 1.9106317411402158,
-      "grad_norm": 0.5373818874359131,
-      "learning_rate": 1.9064124783362218e-07,
-      "loss": 2.1255,
-      "step": 775
-    },
-    {
-      "epoch": 1.9130970724191063,
-      "grad_norm": 0.5443577170372009,
-      "learning_rate": 1.9020797227036395e-07,
-      "loss": 2.0232,
-      "step": 776
-    },
-    {
-      "epoch": 1.9155624036979968,
-      "grad_norm": 0.5216531157493591,
-      "learning_rate": 1.8977469670710572e-07,
-      "loss": 2.091,
-      "step": 777
-    },
-    {
-      "epoch": 1.9180277349768875,
-      "grad_norm": 0.5133727788925171,
-      "learning_rate": 1.893414211438475e-07,
-      "loss": 2.0208,
-      "step": 778
-    },
-    {
-      "epoch": 1.9204930662557782,
-      "grad_norm": 0.5339122414588928,
-      "learning_rate": 1.8890814558058924e-07,
-      "loss": 2.085,
-      "step": 779
-    },
-    {
-      "epoch": 1.9229583975346687,
-      "grad_norm": 0.529255211353302,
-      "learning_rate": 1.88474870017331e-07,
-      "loss": 2.0794,
-      "step": 780
-    },
-    {
-      "epoch": 1.9254237288135592,
-      "grad_norm": 0.54831463098526,
-      "learning_rate": 1.8804159445407278e-07,
-      "loss": 2.1017,
-      "step": 781
-    },
-    {
-      "epoch": 1.9278890600924499,
-      "grad_norm": 0.5309552550315857,
-      "learning_rate": 1.8760831889081456e-07,
-      "loss": 2.1126,
-      "step": 782
-    },
-    {
-      "epoch": 1.9303543913713406,
-      "grad_norm": 0.5368801355361938,
-      "learning_rate": 1.8717504332755633e-07,
-      "loss": 2.0153,
-      "step": 783
-    },
-    {
-      "epoch": 1.932819722650231,
-      "grad_norm": 0.5303763747215271,
-      "learning_rate": 1.867417677642981e-07,
-      "loss": 2.0798,
-      "step": 784
-    },
-    {
-      "epoch": 1.9352850539291218,
-      "grad_norm": 0.5320292115211487,
-      "learning_rate": 1.8630849220103987e-07,
-      "loss": 2.0749,
-      "step": 785
-    },
-    {
-      "epoch": 1.9377503852080125,
-      "grad_norm": 0.5434033274650574,
-      "learning_rate": 1.8587521663778164e-07,
-      "loss": 1.9961,
-      "step": 786
-    },
-    {
-      "epoch": 1.940215716486903,
-      "grad_norm": 0.5612481236457825,
-      "learning_rate": 1.8544194107452339e-07,
-      "loss": 2.0702,
-      "step": 787
-    },
-    {
-      "epoch": 1.9426810477657934,
-      "grad_norm": 0.5725658535957336,
-      "learning_rate": 1.8500866551126516e-07,
-      "loss": 2.106,
-      "step": 788
-    },
-    {
-      "epoch": 1.9451463790446841,
-      "grad_norm": 0.5459932088851929,
-      "learning_rate": 1.8457538994800693e-07,
-      "loss": 2.1029,
-      "step": 789
-    },
-    {
-      "epoch": 1.9476117103235748,
-      "grad_norm": 0.5336212515830994,
-      "learning_rate": 1.841421143847487e-07,
-      "loss": 2.1044,
-      "step": 790
-    },
-    {
-      "epoch": 1.9500770416024653,
-      "grad_norm": 0.5421571135520935,
-      "learning_rate": 1.8370883882149047e-07,
-      "loss": 2.111,
-      "step": 791
-    },
-    {
-      "epoch": 1.9525423728813558,
-      "grad_norm": 0.5272592306137085,
-      "learning_rate": 1.8327556325823224e-07,
-      "loss": 2.0597,
-      "step": 792
-    },
-    {
-      "epoch": 1.9550077041602465,
-      "grad_norm": 0.5724270939826965,
-      "learning_rate": 1.8284228769497402e-07,
-      "loss": 2.1377,
-      "step": 793
-    },
-    {
-      "epoch": 1.9574730354391372,
-      "grad_norm": 0.544668972492218,
-      "learning_rate": 1.8240901213171576e-07,
-      "loss": 2.0321,
-      "step": 794
-    },
-    {
-      "epoch": 1.9599383667180277,
-      "grad_norm": 0.5739486217498779,
-      "learning_rate": 1.8197573656845753e-07,
-      "loss": 2.1109,
-      "step": 795
-    },
-    {
-      "epoch": 1.9624036979969184,
-      "grad_norm": 0.5566874742507935,
-      "learning_rate": 1.815424610051993e-07,
-      "loss": 2.0582,
-      "step": 796
-    },
-    {
-      "epoch": 1.964869029275809,
-      "grad_norm": 0.5381664633750916,
-      "learning_rate": 1.8110918544194107e-07,
-      "loss": 2.0734,
-      "step": 797
-    },
-    {
-      "epoch": 1.9673343605546996,
-      "grad_norm": 0.5530892014503479,
-      "learning_rate": 1.8067590987868285e-07,
-      "loss": 2.0903,
-      "step": 798
-    },
-    {
-      "epoch": 1.96979969183359,
-      "grad_norm": 0.5485751628875732,
-      "learning_rate": 1.8024263431542462e-07,
-      "loss": 2.0376,
-      "step": 799
-    },
-    {
-      "epoch": 1.9722650231124808,
-      "grad_norm": 0.5459527373313904,
-      "learning_rate": 1.798093587521664e-07,
-      "loss": 2.0328,
-      "step": 800
-    },
-    {
-      "epoch": 1.9747303543913715,
-      "grad_norm": 0.5666062235832214,
-      "learning_rate": 1.7937608318890816e-07,
-      "loss": 2.0966,
-      "step": 801
-    },
-    {
-      "epoch": 1.977195685670262,
-      "grad_norm": 0.5256234407424927,
-      "learning_rate": 1.789428076256499e-07,
-      "loss": 2.0458,
-      "step": 802
-    },
-    {
-      "epoch": 1.9796610169491524,
-      "grad_norm": 0.567815899848938,
-      "learning_rate": 1.7850953206239168e-07,
-      "loss": 2.0793,
-      "step": 803
-    },
-    {
-      "epoch": 1.9821263482280431,
-      "grad_norm": 0.5290260910987854,
-      "learning_rate": 1.7807625649913345e-07,
-      "loss": 2.0759,
-      "step": 804
-    },
-    {
-      "epoch": 1.9845916795069338,
-      "grad_norm": 0.5459809303283691,
-      "learning_rate": 1.7764298093587522e-07,
-      "loss": 2.0634,
-      "step": 805
-    },
-    {
-      "epoch": 1.9870570107858243,
-      "grad_norm": 0.5483869910240173,
-      "learning_rate": 1.77209705372617e-07,
-      "loss": 2.1077,
-      "step": 806
-    },
-    {
-      "epoch": 1.9895223420647148,
-      "grad_norm": 0.5340948700904846,
-      "learning_rate": 1.7677642980935876e-07,
-      "loss": 2.0584,
-      "step": 807
-    },
-    {
-      "epoch": 1.9919876733436055,
-      "grad_norm": 0.5514506101608276,
-      "learning_rate": 1.7634315424610053e-07,
-      "loss": 2.1299,
-      "step": 808
-    },
-    {
-      "epoch": 1.9944530046224962,
-      "grad_norm": 0.5255659222602844,
-      "learning_rate": 1.7590987868284225e-07,
-      "loss": 2.0081,
-      "step": 809
-    },
-    {
-      "epoch": 1.9969183359013867,
-      "grad_norm": 0.5237001776695251,
-      "learning_rate": 1.7547660311958402e-07,
-      "loss": 1.9692,
-      "step": 810
-    },
-    {
-      "epoch": 1.9993836671802774,
-      "grad_norm": 0.51847243309021,
-      "learning_rate": 1.750433275563258e-07,
-      "loss": 2.0306,
-      "step": 811
-    },
-    {
-      "epoch": 1.9993836671802774,
-      "eval_loss": 2.072087287902832,
-      "eval_runtime": 399.1878,
-      "eval_samples_per_second": 1.233,
-      "eval_steps_per_second": 0.308,
-      "step": 811
     }
   ],
   "logging_steps": 1,
@@ -5719,7 +2869,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.3947626026442424e+18,
+  "total_flos": 1.1973813013221212e+18,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null