Upload folder using huggingface_hub

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:288be6c8b98f62975ce52470207e00e09a929ae2be8c56589bc000e8bf648de2
+oid sha256:52a9be062605070742cde00600264cee991bb06edc9c792af4d52f4aaefb6256
 size 13648432

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f358539ba7d7cba232d693740717df16c470360c060dcbbfbdbc0b81740fd9f7
+oid sha256:e4559046956d47204fd03216ecec0a6ab372ec6b1898619008c2c64d7a27e94c
 size 27338810

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:adcaf1c854b7f05c382576167fc5384fcf54c029f39da1646c97625ac98a2bba
+oid sha256:89c22016d47a4c8fc4fba2c7162dce77859e0b25d0082d169067b9fa1803b889
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b41241c255ae14db691f28ac1e568666a54cb5937ca5a6e52af17207a998cbc0
+oid sha256:57802a66d4e06aedffd82faf21afe91d5401e799d1bf827c6e333c3d9d5698a6
 size 1064

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 2.1491148471832275,
-  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-1Q-0U-0C-qa_first/checkpoint-891",
-  "epoch": 2.991187578682333,
+  "best_metric": 2.154229164123535,
+  "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-1Q-0U-0C-qa_first/checkpoint-595",
+  "epoch": 1.9974821653378094,
   "eval_steps": 500,
-  "global_step": 891,
+  "global_step": 595,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -4188,2086 +4188,6 @@
       "eval_samples_per_second": 1.114,
       "eval_steps_per_second": 0.28,
       "step": 595
-    },
-    {
-      "epoch": 2.00083927822073,
-      "grad_norm": 0.5490261316299438,
-      "learning_rate": 1.743498817966903e-07,
-      "loss": 2.0814,
-      "step": 596
-    },
-    {
-      "epoch": 2.0041963911036507,
-      "grad_norm": 0.5828990936279297,
-      "learning_rate": 1.7375886524822695e-07,
-      "loss": 2.0934,
-      "step": 597
-    },
-    {
-      "epoch": 2.0075535039865717,
-      "grad_norm": 0.5824359059333801,
-      "learning_rate": 1.7316784869976357e-07,
-      "loss": 2.1668,
-      "step": 598
-    },
-    {
-      "epoch": 2.0109106168694924,
-      "grad_norm": 0.5972450971603394,
-      "learning_rate": 1.7257683215130022e-07,
-      "loss": 2.1254,
-      "step": 599
-    },
-    {
-      "epoch": 2.014267729752413,
-      "grad_norm": 0.5486287474632263,
-      "learning_rate": 1.7198581560283688e-07,
-      "loss": 2.2053,
-      "step": 600
-    },
-    {
-      "epoch": 2.0176248426353336,
-      "grad_norm": 0.5345571637153625,
-      "learning_rate": 1.713947990543735e-07,
-      "loss": 2.0901,
-      "step": 601
-    },
-    {
-      "epoch": 2.020981955518254,
-      "grad_norm": 0.544323742389679,
-      "learning_rate": 1.7080378250591015e-07,
-      "loss": 2.1225,
-      "step": 602
-    },
-    {
-      "epoch": 2.024339068401175,
-      "grad_norm": 0.5551151633262634,
-      "learning_rate": 1.702127659574468e-07,
-      "loss": 2.1934,
-      "step": 603
-    },
-    {
-      "epoch": 2.027696181284096,
-      "grad_norm": 0.5789584517478943,
-      "learning_rate": 1.6962174940898345e-07,
-      "loss": 2.1936,
-      "step": 604
-    },
-    {
-      "epoch": 2.0310532941670165,
-      "grad_norm": 0.5161600112915039,
-      "learning_rate": 1.6903073286052007e-07,
-      "loss": 2.1503,
-      "step": 605
-    },
-    {
-      "epoch": 2.034410407049937,
-      "grad_norm": 0.5447808504104614,
-      "learning_rate": 1.6843971631205673e-07,
-      "loss": 2.1207,
-      "step": 606
-    },
-    {
-      "epoch": 2.0377675199328578,
-      "grad_norm": 0.5768391489982605,
-      "learning_rate": 1.6784869976359338e-07,
-      "loss": 2.2098,
-      "step": 607
-    },
-    {
-      "epoch": 2.0411246328157784,
-      "grad_norm": 0.5963948965072632,
-      "learning_rate": 1.6725768321513e-07,
-      "loss": 2.1794,
-      "step": 608
-    },
-    {
-      "epoch": 2.044481745698699,
-      "grad_norm": 0.5620808601379395,
-      "learning_rate": 1.6666666666666665e-07,
-      "loss": 2.1656,
-      "step": 609
-    },
-    {
-      "epoch": 2.0478388585816196,
-      "grad_norm": 0.49818509817123413,
-      "learning_rate": 1.660756501182033e-07,
-      "loss": 2.0503,
-      "step": 610
-    },
-    {
-      "epoch": 2.0511959714645407,
-      "grad_norm": 0.5394522547721863,
-      "learning_rate": 1.6548463356973992e-07,
-      "loss": 2.1243,
-      "step": 611
-    },
-    {
-      "epoch": 2.0545530843474613,
-      "grad_norm": 0.5886016488075256,
-      "learning_rate": 1.6489361702127658e-07,
-      "loss": 2.167,
-      "step": 612
-    },
-    {
-      "epoch": 2.057910197230382,
-      "grad_norm": 0.5549562573432922,
-      "learning_rate": 1.6430260047281323e-07,
-      "loss": 2.0998,
-      "step": 613
-    },
-    {
-      "epoch": 2.0612673101133026,
-      "grad_norm": 0.5675773620605469,
-      "learning_rate": 1.6371158392434985e-07,
-      "loss": 2.1334,
-      "step": 614
-    },
-    {
-      "epoch": 2.064624422996223,
-      "grad_norm": 0.5877095460891724,
-      "learning_rate": 1.6312056737588653e-07,
-      "loss": 2.1226,
-      "step": 615
-    },
-    {
-      "epoch": 2.067981535879144,
-      "grad_norm": 0.5441724061965942,
-      "learning_rate": 1.6252955082742318e-07,
-      "loss": 2.1736,
-      "step": 616
-    },
-    {
-      "epoch": 2.0713386487620644,
-      "grad_norm": 0.5380129218101501,
-      "learning_rate": 1.6193853427895983e-07,
-      "loss": 2.1552,
-      "step": 617
-    },
-    {
-      "epoch": 2.0746957616449855,
-      "grad_norm": 0.5300249457359314,
-      "learning_rate": 1.6134751773049645e-07,
-      "loss": 2.1198,
-      "step": 618
-    },
-    {
-      "epoch": 2.078052874527906,
-      "grad_norm": 0.5587384104728699,
-      "learning_rate": 1.607565011820331e-07,
-      "loss": 2.1519,
-      "step": 619
-    },
-    {
-      "epoch": 2.0814099874108267,
-      "grad_norm": 0.5410963296890259,
-      "learning_rate": 1.6016548463356975e-07,
-      "loss": 2.1687,
-      "step": 620
-    },
-    {
-      "epoch": 2.0847671002937473,
-      "grad_norm": 0.5827339291572571,
-      "learning_rate": 1.5957446808510638e-07,
-      "loss": 2.1853,
-      "step": 621
-    },
-    {
-      "epoch": 2.088124213176668,
-      "grad_norm": 0.5589419007301331,
-      "learning_rate": 1.5898345153664303e-07,
-      "loss": 2.1093,
-      "step": 622
-    },
-    {
-      "epoch": 2.0914813260595886,
-      "grad_norm": 0.5821492671966553,
-      "learning_rate": 1.5839243498817968e-07,
-      "loss": 2.1167,
-      "step": 623
-    },
-    {
-      "epoch": 2.0948384389425097,
-      "grad_norm": 0.5975298285484314,
-      "learning_rate": 1.578014184397163e-07,
-      "loss": 2.1522,
-      "step": 624
-    },
-    {
-      "epoch": 2.0981955518254303,
-      "grad_norm": 0.5707703232765198,
-      "learning_rate": 1.5721040189125295e-07,
-      "loss": 2.1726,
-      "step": 625
-    },
-    {
-      "epoch": 2.101552664708351,
-      "grad_norm": 0.550392746925354,
-      "learning_rate": 1.566193853427896e-07,
-      "loss": 2.1443,
-      "step": 626
-    },
-    {
-      "epoch": 2.1049097775912715,
-      "grad_norm": 0.5648649334907532,
-      "learning_rate": 1.5602836879432623e-07,
-      "loss": 2.1134,
-      "step": 627
-    },
-    {
-      "epoch": 2.108266890474192,
-      "grad_norm": 0.5239277482032776,
-      "learning_rate": 1.5543735224586288e-07,
-      "loss": 2.1372,
-      "step": 628
-    },
-    {
-      "epoch": 2.1116240033571128,
-      "grad_norm": 0.5361467003822327,
-      "learning_rate": 1.5484633569739953e-07,
-      "loss": 2.0948,
-      "step": 629
-    },
-    {
-      "epoch": 2.1149811162400334,
-      "grad_norm": 2.1900720596313477,
-      "learning_rate": 1.5425531914893615e-07,
-      "loss": 2.2063,
-      "step": 630
-    },
-    {
-      "epoch": 2.1183382291229544,
-      "grad_norm": 0.539082944393158,
-      "learning_rate": 1.536643026004728e-07,
-      "loss": 2.1237,
-      "step": 631
-    },
-    {
-      "epoch": 2.121695342005875,
-      "grad_norm": 0.5592411160469055,
-      "learning_rate": 1.5307328605200945e-07,
-      "loss": 2.2125,
-      "step": 632
-    },
-    {
-      "epoch": 2.1250524548887957,
-      "grad_norm": 0.5461694002151489,
-      "learning_rate": 1.524822695035461e-07,
-      "loss": 2.1581,
-      "step": 633
-    },
-    {
-      "epoch": 2.1284095677717163,
-      "grad_norm": 0.5573883652687073,
-      "learning_rate": 1.5189125295508273e-07,
-      "loss": 2.1502,
-      "step": 634
-    },
-    {
-      "epoch": 2.131766680654637,
-      "grad_norm": 0.6131634712219238,
-      "learning_rate": 1.5130023640661938e-07,
-      "loss": 2.1482,
-      "step": 635
-    },
-    {
-      "epoch": 2.1351237935375575,
-      "grad_norm": 0.509318470954895,
-      "learning_rate": 1.5070921985815603e-07,
-      "loss": 2.0815,
-      "step": 636
-    },
-    {
-      "epoch": 2.1384809064204786,
-      "grad_norm": 0.5671557188034058,
-      "learning_rate": 1.5011820330969265e-07,
-      "loss": 2.1446,
-      "step": 637
-    },
-    {
-      "epoch": 2.1418380193033992,
-      "grad_norm": 0.5513312816619873,
-      "learning_rate": 1.495271867612293e-07,
-      "loss": 2.1495,
-      "step": 638
-    },
-    {
-      "epoch": 2.14519513218632,
-      "grad_norm": 0.57234787940979,
-      "learning_rate": 1.4893617021276595e-07,
-      "loss": 2.2077,
-      "step": 639
-    },
-    {
-      "epoch": 2.1485522450692405,
-      "grad_norm": 0.5913609862327576,
-      "learning_rate": 1.4834515366430258e-07,
-      "loss": 2.2186,
-      "step": 640
-    },
-    {
-      "epoch": 2.151909357952161,
-      "grad_norm": 0.5753595232963562,
-      "learning_rate": 1.4775413711583923e-07,
-      "loss": 2.1373,
-      "step": 641
-    },
-    {
-      "epoch": 2.1552664708350817,
-      "grad_norm": 0.5550244450569153,
-      "learning_rate": 1.4716312056737588e-07,
-      "loss": 2.1694,
-      "step": 642
-    },
-    {
-      "epoch": 2.1586235837180023,
-      "grad_norm": 0.5355656743049622,
-      "learning_rate": 1.465721040189125e-07,
-      "loss": 2.2296,
-      "step": 643
-    },
-    {
-      "epoch": 2.1619806966009234,
-      "grad_norm": 0.5573230981826782,
-      "learning_rate": 1.4598108747044915e-07,
-      "loss": 2.1731,
-      "step": 644
-    },
-    {
-      "epoch": 2.165337809483844,
-      "grad_norm": 0.5603845715522766,
-      "learning_rate": 1.453900709219858e-07,
-      "loss": 2.1463,
-      "step": 645
-    },
-    {
-      "epoch": 2.1686949223667646,
-      "grad_norm": 0.5660983920097351,
-      "learning_rate": 1.4479905437352243e-07,
-      "loss": 2.1158,
-      "step": 646
-    },
-    {
-      "epoch": 2.1720520352496853,
-      "grad_norm": 0.5612486004829407,
-      "learning_rate": 1.442080378250591e-07,
-      "loss": 2.1343,
-      "step": 647
-    },
-    {
-      "epoch": 2.175409148132606,
-      "grad_norm": 0.563666045665741,
-      "learning_rate": 1.4361702127659576e-07,
-      "loss": 2.1491,
-      "step": 648
-    },
-    {
-      "epoch": 2.1787662610155265,
-      "grad_norm": 0.5978438258171082,
-      "learning_rate": 1.430260047281324e-07,
-      "loss": 2.213,
-      "step": 649
-    },
-    {
-      "epoch": 2.182123373898447,
-      "grad_norm": 0.5486546158790588,
-      "learning_rate": 1.4243498817966903e-07,
-      "loss": 2.1282,
-      "step": 650
-    },
-    {
-      "epoch": 2.185480486781368,
-      "grad_norm": 0.5446551442146301,
-      "learning_rate": 1.4184397163120568e-07,
-      "loss": 2.1639,
-      "step": 651
-    },
-    {
-      "epoch": 2.188837599664289,
-      "grad_norm": 0.5193803310394287,
-      "learning_rate": 1.4125295508274233e-07,
-      "loss": 2.1069,
-      "step": 652
-    },
-    {
-      "epoch": 2.1921947125472094,
-      "grad_norm": 0.5513821840286255,
-      "learning_rate": 1.4066193853427895e-07,
-      "loss": 2.1852,
-      "step": 653
-    },
-    {
-      "epoch": 2.19555182543013,
-      "grad_norm": 0.5725794434547424,
-      "learning_rate": 1.400709219858156e-07,
-      "loss": 2.1547,
-      "step": 654
-    },
-    {
-      "epoch": 2.1989089383130507,
-      "grad_norm": 0.5426892042160034,
-      "learning_rate": 1.3947990543735226e-07,
-      "loss": 2.1286,
-      "step": 655
-    },
-    {
-      "epoch": 2.2022660511959713,
-      "grad_norm": 0.5511768460273743,
-      "learning_rate": 1.3888888888888888e-07,
-      "loss": 2.0826,
-      "step": 656
-    },
-    {
-      "epoch": 2.205623164078892,
-      "grad_norm": 0.5628213286399841,
-      "learning_rate": 1.3829787234042553e-07,
-      "loss": 2.1114,
-      "step": 657
-    },
-    {
-      "epoch": 2.208980276961813,
-      "grad_norm": 0.5592065453529358,
-      "learning_rate": 1.3770685579196218e-07,
-      "loss": 2.1663,
-      "step": 658
-    },
-    {
-      "epoch": 2.2123373898447336,
-      "grad_norm": 0.584954559803009,
-      "learning_rate": 1.371158392434988e-07,
-      "loss": 2.2195,
-      "step": 659
-    },
-    {
-      "epoch": 2.2156945027276542,
-      "grad_norm": 0.6203845739364624,
-      "learning_rate": 1.3652482269503546e-07,
-      "loss": 2.1448,
-      "step": 660
-    },
-    {
-      "epoch": 2.219051615610575,
-      "grad_norm": 0.5427147746086121,
-      "learning_rate": 1.359338061465721e-07,
-      "loss": 2.1436,
-      "step": 661
-    },
-    {
-      "epoch": 2.2224087284934955,
-      "grad_norm": 0.5791121125221252,
-      "learning_rate": 1.3534278959810873e-07,
-      "loss": 2.1052,
-      "step": 662
-    },
-    {
-      "epoch": 2.225765841376416,
-      "grad_norm": 0.5903706550598145,
-      "learning_rate": 1.3475177304964538e-07,
-      "loss": 2.1175,
-      "step": 663
-    },
-    {
-      "epoch": 2.229122954259337,
-      "grad_norm": 0.5627418160438538,
-      "learning_rate": 1.3416075650118203e-07,
-      "loss": 2.1011,
-      "step": 664
-    },
-    {
-      "epoch": 2.2324800671422578,
-      "grad_norm": 0.5707918405532837,
-      "learning_rate": 1.3356973995271868e-07,
-      "loss": 2.0925,
-      "step": 665
-    },
-    {
-      "epoch": 2.2358371800251784,
-      "grad_norm": 0.5552944540977478,
-      "learning_rate": 1.329787234042553e-07,
-      "loss": 2.132,
-      "step": 666
-    },
-    {
-      "epoch": 2.239194292908099,
-      "grad_norm": 0.5708072185516357,
-      "learning_rate": 1.3238770685579196e-07,
-      "loss": 2.148,
-      "step": 667
-    },
-    {
-      "epoch": 2.2425514057910196,
-      "grad_norm": 0.5760245323181152,
-      "learning_rate": 1.317966903073286e-07,
-      "loss": 2.173,
-      "step": 668
-    },
-    {
-      "epoch": 2.2459085186739403,
-      "grad_norm": 0.5791294574737549,
-      "learning_rate": 1.3120567375886523e-07,
-      "loss": 2.1913,
-      "step": 669
-    },
-    {
-      "epoch": 2.2492656315568613,
-      "grad_norm": 0.5469384789466858,
-      "learning_rate": 1.3061465721040188e-07,
-      "loss": 2.0626,
-      "step": 670
-    },
-    {
-      "epoch": 2.252622744439782,
-      "grad_norm": 0.5542003512382507,
-      "learning_rate": 1.3002364066193853e-07,
-      "loss": 2.1759,
-      "step": 671
-    },
-    {
-      "epoch": 2.2559798573227026,
-      "grad_norm": 0.5164523720741272,
-      "learning_rate": 1.2943262411347516e-07,
-      "loss": 2.0917,
-      "step": 672
-    },
-    {
-      "epoch": 2.259336970205623,
-      "grad_norm": 0.5994356274604797,
-      "learning_rate": 1.288416075650118e-07,
-      "loss": 2.1832,
-      "step": 673
-    },
-    {
-      "epoch": 2.262694083088544,
-      "grad_norm": 0.5621914267539978,
-      "learning_rate": 1.2825059101654846e-07,
-      "loss": 2.1212,
-      "step": 674
-    },
-    {
-      "epoch": 2.2660511959714644,
-      "grad_norm": 0.576374351978302,
-      "learning_rate": 1.2765957446808508e-07,
-      "loss": 2.1408,
-      "step": 675
-    },
-    {
-      "epoch": 2.269408308854385,
-      "grad_norm": 0.5409926176071167,
-      "learning_rate": 1.2706855791962173e-07,
-      "loss": 2.1872,
-      "step": 676
-    },
-    {
-      "epoch": 2.272765421737306,
-      "grad_norm": 0.5715805292129517,
-      "learning_rate": 1.2647754137115838e-07,
-      "loss": 2.1805,
-      "step": 677
-    },
-    {
-      "epoch": 2.2761225346202267,
-      "grad_norm": 0.5578653812408447,
-      "learning_rate": 1.25886524822695e-07,
-      "loss": 2.1738,
-      "step": 678
-    },
-    {
-      "epoch": 2.2794796475031474,
-      "grad_norm": 0.574573278427124,
-      "learning_rate": 1.2529550827423166e-07,
-      "loss": 2.1258,
-      "step": 679
-    },
-    {
-      "epoch": 2.282836760386068,
-      "grad_norm": 0.5796377658843994,
-      "learning_rate": 1.247044917257683e-07,
-      "loss": 2.1559,
-      "step": 680
-    },
-    {
-      "epoch": 2.2861938732689886,
-      "grad_norm": 0.5414871573448181,
-      "learning_rate": 1.2411347517730496e-07,
-      "loss": 2.1258,
-      "step": 681
-    },
-    {
-      "epoch": 2.289550986151909,
-      "grad_norm": 0.5621658563613892,
-      "learning_rate": 1.235224586288416e-07,
-      "loss": 2.0942,
-      "step": 682
-    },
-    {
-      "epoch": 2.29290809903483,
-      "grad_norm": 0.5886093974113464,
-      "learning_rate": 1.2293144208037823e-07,
-      "loss": 2.1605,
-      "step": 683
-    },
-    {
-      "epoch": 2.296265211917751,
-      "grad_norm": 0.528712809085846,
-      "learning_rate": 1.2234042553191488e-07,
-      "loss": 2.1128,
-      "step": 684
-    },
-    {
-      "epoch": 2.2996223248006715,
-      "grad_norm": 0.5604879260063171,
-      "learning_rate": 1.2174940898345153e-07,
-      "loss": 2.1542,
-      "step": 685
-    },
-    {
-      "epoch": 2.302979437683592,
-      "grad_norm": 0.5524810552597046,
-      "learning_rate": 1.2115839243498816e-07,
-      "loss": 2.1249,
-      "step": 686
-    },
-    {
-      "epoch": 2.3063365505665128,
-      "grad_norm": 0.5685278177261353,
-      "learning_rate": 1.205673758865248e-07,
-      "loss": 2.1189,
-      "step": 687
-    },
-    {
-      "epoch": 2.3096936634494334,
-      "grad_norm": 0.5520109534263611,
-      "learning_rate": 1.1997635933806146e-07,
-      "loss": 2.1371,
-      "step": 688
-    },
-    {
-      "epoch": 2.313050776332354,
-      "grad_norm": 0.5782994627952576,
-      "learning_rate": 1.193853427895981e-07,
-      "loss": 2.1783,
-      "step": 689
-    },
-    {
-      "epoch": 2.3164078892152746,
-      "grad_norm": 0.5418346524238586,
-      "learning_rate": 1.1879432624113475e-07,
-      "loss": 2.1516,
-      "step": 690
-    },
-    {
-      "epoch": 2.3197650020981957,
-      "grad_norm": 0.5413779616355896,
-      "learning_rate": 1.182033096926714e-07,
-      "loss": 2.104,
-      "step": 691
-    },
-    {
-      "epoch": 2.3231221149811163,
-      "grad_norm": 0.5890994668006897,
-      "learning_rate": 1.1761229314420803e-07,
-      "loss": 2.1653,
-      "step": 692
-    },
-    {
-      "epoch": 2.326479227864037,
-      "grad_norm": 0.5684483051300049,
-      "learning_rate": 1.1702127659574468e-07,
-      "loss": 2.1468,
-      "step": 693
-    },
-    {
-      "epoch": 2.3298363407469576,
-      "grad_norm": 0.5710468888282776,
-      "learning_rate": 1.1643026004728132e-07,
-      "loss": 2.1695,
-      "step": 694
-    },
-    {
-      "epoch": 2.333193453629878,
-      "grad_norm": 0.5511908531188965,
-      "learning_rate": 1.1583924349881796e-07,
-      "loss": 2.153,
-      "step": 695
-    },
-    {
-      "epoch": 2.336550566512799,
-      "grad_norm": 0.5747969150543213,
-      "learning_rate": 1.1524822695035461e-07,
-      "loss": 2.1164,
-      "step": 696
-    },
-    {
-      "epoch": 2.33990767939572,
-      "grad_norm": 0.5769465565681458,
-      "learning_rate": 1.1465721040189125e-07,
-      "loss": 2.0983,
-      "step": 697
-    },
-    {
-      "epoch": 2.3432647922786405,
-      "grad_norm": 0.5540648698806763,
-      "learning_rate": 1.1406619385342788e-07,
-      "loss": 2.123,
-      "step": 698
-    },
-    {
-      "epoch": 2.346621905161561,
-      "grad_norm": 0.5682700276374817,
-      "learning_rate": 1.1347517730496453e-07,
-      "loss": 2.1387,
-      "step": 699
-    },
-    {
-      "epoch": 2.3499790180444817,
-      "grad_norm": 0.5571325421333313,
-      "learning_rate": 1.1288416075650117e-07,
-      "loss": 2.1504,
-      "step": 700
-    },
-    {
-      "epoch": 2.3533361309274023,
-      "grad_norm": 0.592702329158783,
-      "learning_rate": 1.1229314420803782e-07,
-      "loss": 2.1569,
-      "step": 701
-    },
-    {
-      "epoch": 2.356693243810323,
-      "grad_norm": 0.5651736259460449,
-      "learning_rate": 1.1170212765957446e-07,
-      "loss": 2.1349,
-      "step": 702
-    },
-    {
-      "epoch": 2.360050356693244,
-      "grad_norm": 0.5222681164741516,
-      "learning_rate": 1.111111111111111e-07,
-      "loss": 2.1089,
-      "step": 703
-    },
-    {
-      "epoch": 2.3634074695761647,
-      "grad_norm": 0.5406708717346191,
-      "learning_rate": 1.1052009456264776e-07,
-      "loss": 2.1394,
-      "step": 704
-    },
-    {
-      "epoch": 2.3667645824590853,
-      "grad_norm": 0.5576866269111633,
-      "learning_rate": 1.099290780141844e-07,
-      "loss": 2.1586,
-      "step": 705
-    },
-    {
-      "epoch": 2.370121695342006,
-      "grad_norm": 0.5735974311828613,
-      "learning_rate": 1.0933806146572103e-07,
-      "loss": 2.175,
-      "step": 706
-    },
-    {
-      "epoch": 2.3734788082249265,
-      "grad_norm": 0.5451417565345764,
-      "learning_rate": 1.0874704491725768e-07,
-      "loss": 2.1319,
-      "step": 707
-    },
-    {
-      "epoch": 2.376835921107847,
-      "grad_norm": 0.5943117737770081,
-      "learning_rate": 1.0815602836879432e-07,
-      "loss": 2.1199,
-      "step": 708
-    },
-    {
-      "epoch": 2.3801930339907678,
-      "grad_norm": 0.6147474050521851,
-      "learning_rate": 1.0756501182033097e-07,
-      "loss": 2.0827,
-      "step": 709
-    },
-    {
-      "epoch": 2.383550146873689,
-      "grad_norm": 0.5546717047691345,
-      "learning_rate": 1.0697399527186761e-07,
-      "loss": 2.1564,
-      "step": 710
-    },
-    {
-      "epoch": 2.3869072597566094,
-      "grad_norm": 0.5820872783660889,
-      "learning_rate": 1.0638297872340425e-07,
-      "loss": 2.1466,
-      "step": 711
-    },
-    {
-      "epoch": 2.39026437263953,
-      "grad_norm": 0.5819923877716064,
-      "learning_rate": 1.057919621749409e-07,
-      "loss": 2.1618,
-      "step": 712
-    },
-    {
-      "epoch": 2.3936214855224507,
-      "grad_norm": 0.6026069521903992,
-      "learning_rate": 1.0520094562647753e-07,
-      "loss": 2.118,
-      "step": 713
-    },
-    {
-      "epoch": 2.3969785984053713,
-      "grad_norm": 0.5684935450553894,
-      "learning_rate": 1.0460992907801417e-07,
-      "loss": 2.1309,
-      "step": 714
-    },
-    {
-      "epoch": 2.400335711288292,
-      "grad_norm": 0.5980827212333679,
-      "learning_rate": 1.0401891252955082e-07,
-      "loss": 2.1596,
-      "step": 715
-    },
-    {
-      "epoch": 2.4036928241712125,
-      "grad_norm": 0.4979473054409027,
-      "learning_rate": 1.0342789598108746e-07,
-      "loss": 2.0867,
-      "step": 716
-    },
-    {
-      "epoch": 2.4070499370541336,
-      "grad_norm": 0.5678048729896545,
-      "learning_rate": 1.0283687943262411e-07,
-      "loss": 2.1685,
-      "step": 717
-    },
-    {
-      "epoch": 2.4104070499370542,
-      "grad_norm": 0.5005086660385132,
-      "learning_rate": 1.0224586288416075e-07,
-      "loss": 2.0723,
-      "step": 718
-    },
-    {
-      "epoch": 2.413764162819975,
-      "grad_norm": 0.5812935829162598,
-      "learning_rate": 1.0165484633569738e-07,
-      "loss": 2.1636,
-      "step": 719
-    },
-    {
-      "epoch": 2.4171212757028955,
-      "grad_norm": 0.5400813817977905,
-      "learning_rate": 1.0106382978723404e-07,
-      "loss": 2.083,
-      "step": 720
-    },
-    {
-      "epoch": 2.420478388585816,
-      "grad_norm": 0.5791892409324646,
-      "learning_rate": 1.0047281323877069e-07,
-      "loss": 2.1372,
-      "step": 721
-    },
-    {
-      "epoch": 2.4238355014687367,
-      "grad_norm": 0.5753834843635559,
-      "learning_rate": 9.988179669030732e-08,
-      "loss": 2.1322,
-      "step": 722
-    },
-    {
-      "epoch": 2.4271926143516573,
-      "grad_norm": 0.5643473267555237,
-      "learning_rate": 9.929078014184397e-08,
-      "loss": 2.1252,
-      "step": 723
-    },
-    {
-      "epoch": 2.4305497272345784,
-      "grad_norm": 0.5979934930801392,
-      "learning_rate": 9.869976359338061e-08,
-      "loss": 2.1401,
-      "step": 724
-    },
-    {
-      "epoch": 2.433906840117499,
-      "grad_norm": 0.5440395474433899,
-      "learning_rate": 9.810874704491726e-08,
-      "loss": 2.165,
-      "step": 725
-    },
-    {
-      "epoch": 2.4372639530004196,
-      "grad_norm": 0.5564695000648499,
-      "learning_rate": 9.75177304964539e-08,
-      "loss": 2.0954,
-      "step": 726
-    },
-    {
-      "epoch": 2.4406210658833403,
-      "grad_norm": 0.6030122637748718,
-      "learning_rate": 9.692671394799054e-08,
-      "loss": 2.1144,
-      "step": 727
-    },
-    {
-      "epoch": 2.443978178766261,
-      "grad_norm": 0.5595952272415161,
-      "learning_rate": 9.633569739952719e-08,
-      "loss": 2.1009,
-      "step": 728
-    },
-    {
-      "epoch": 2.4473352916491815,
-      "grad_norm": 0.6225353479385376,
-      "learning_rate": 9.574468085106382e-08,
-      "loss": 2.1491,
-      "step": 729
-    },
-    {
-      "epoch": 2.4506924045321026,
-      "grad_norm": 0.5642361044883728,
-      "learning_rate": 9.515366430260046e-08,
-      "loss": 2.1863,
-      "step": 730
-    },
-    {
-      "epoch": 2.454049517415023,
-      "grad_norm": 0.5429561138153076,
-      "learning_rate": 9.456264775413711e-08,
-      "loss": 2.1346,
-      "step": 731
-    },
-    {
-      "epoch": 2.457406630297944,
-      "grad_norm": 0.603590726852417,
-      "learning_rate": 9.397163120567375e-08,
-      "loss": 2.1782,
-      "step": 732
-    },
-    {
-      "epoch": 2.4607637431808644,
-      "grad_norm": 0.5963548421859741,
-      "learning_rate": 9.33806146572104e-08,
-      "loss": 2.1252,
-      "step": 733
-    },
-    {
-      "epoch": 2.464120856063785,
-      "grad_norm": 0.5711669325828552,
-      "learning_rate": 9.278959810874704e-08,
-      "loss": 2.1494,
-      "step": 734
-    },
-    {
-      "epoch": 2.4674779689467057,
-      "grad_norm": 0.5727424025535583,
-      "learning_rate": 9.219858156028367e-08,
-      "loss": 2.1455,
-      "step": 735
-    },
-    {
-      "epoch": 2.4708350818296267,
-      "grad_norm": 0.5406821370124817,
-      "learning_rate": 9.160756501182032e-08,
-      "loss": 2.1044,
-      "step": 736
-    },
-    {
-      "epoch": 2.4741921947125474,
-      "grad_norm": 0.5895271301269531,
-      "learning_rate": 9.101654846335697e-08,
-      "loss": 2.1494,
-      "step": 737
-    },
-    {
-      "epoch": 2.477549307595468,
-      "grad_norm": 0.5411199331283569,
-      "learning_rate": 9.042553191489363e-08,
-      "loss": 2.1043,
-      "step": 738
-    },
-    {
-      "epoch": 2.4809064204783886,
-      "grad_norm": 0.6222611665725708,
-      "learning_rate": 8.983451536643026e-08,
-      "loss": 2.1783,
-      "step": 739
-    },
-    {
-      "epoch": 2.4842635333613092,
-      "grad_norm": 0.6279255747795105,
-      "learning_rate": 8.92434988179669e-08,
-      "loss": 2.1208,
-      "step": 740
-    },
-    {
-      "epoch": 2.48762064624423,
-      "grad_norm": 0.5328918099403381,
-      "learning_rate": 8.865248226950355e-08,
-      "loss": 2.0602,
-      "step": 741
-    },
-    {
-      "epoch": 2.4909777591271505,
-      "grad_norm": 0.5722616314888,
-      "learning_rate": 8.806146572104019e-08,
-      "loss": 2.1542,
-      "step": 742
-    },
-    {
-      "epoch": 2.4943348720100715,
-      "grad_norm": 0.6053423881530762,
-      "learning_rate": 8.747044917257682e-08,
-      "loss": 2.1443,
-      "step": 743
-    },
-    {
-      "epoch": 2.497691984892992,
-      "grad_norm": 0.5643149614334106,
-      "learning_rate": 8.687943262411348e-08,
-      "loss": 2.1571,
-      "step": 744
-    },
-    {
-      "epoch": 2.5010490977759128,
-      "grad_norm": 0.5647066831588745,
-      "learning_rate": 8.628841607565011e-08,
-      "loss": 2.0443,
-      "step": 745
-    },
-    {
-      "epoch": 2.5044062106588334,
-      "grad_norm": 0.5630561113357544,
-      "learning_rate": 8.569739952718675e-08,
-      "loss": 2.0837,
-      "step": 746
-    },
-    {
-      "epoch": 2.507763323541754,
-      "grad_norm": 0.5802566409111023,
-      "learning_rate": 8.51063829787234e-08,
-      "loss": 2.2105,
-      "step": 747
-    },
-    {
-      "epoch": 2.5111204364246746,
-      "grad_norm": 0.5902342796325684,
-      "learning_rate": 8.451536643026004e-08,
-      "loss": 2.1874,
-      "step": 748
-    },
-    {
-      "epoch": 2.5144775493075953,
-      "grad_norm": 0.5961369872093201,
-      "learning_rate": 8.392434988179669e-08,
-      "loss": 2.1717,
-      "step": 749
-    },
-    {
-      "epoch": 2.5178346621905163,
-      "grad_norm": 0.5433673858642578,
-      "learning_rate": 8.333333333333333e-08,
-      "loss": 2.1003,
-      "step": 750
-    },
-    {
-      "epoch": 2.521191775073437,
-      "grad_norm": 0.5389529466629028,
-      "learning_rate": 8.274231678486996e-08,
-      "loss": 2.1282,
-      "step": 751
-    },
-    {
-      "epoch": 2.5245488879563576,
-      "grad_norm": 0.5602272748947144,
-      "learning_rate": 8.215130023640661e-08,
-      "loss": 2.1452,
-      "step": 752
-    },
-    {
-      "epoch": 2.527906000839278,
-      "grad_norm": 0.5761686563491821,
-      "learning_rate": 8.156028368794326e-08,
-      "loss": 2.1399,
-      "step": 753
-    },
-    {
-      "epoch": 2.531263113722199,
-      "grad_norm": 0.5529290437698364,
-      "learning_rate": 8.096926713947991e-08,
-      "loss": 2.0855,
-      "step": 754
-    },
-    {
-      "epoch": 2.5346202266051194,
-      "grad_norm": 0.5970497727394104,
-      "learning_rate": 8.037825059101655e-08,
-      "loss": 2.1162,
-      "step": 755
-    },
-    {
-      "epoch": 2.53797733948804,
-      "grad_norm": 0.5574461221694946,
-      "learning_rate": 7.978723404255319e-08,
-      "loss": 2.169,
-      "step": 756
-    },
-    {
-      "epoch": 2.541334452370961,
-      "grad_norm": 0.5463849306106567,
-      "learning_rate": 7.919621749408984e-08,
-      "loss": 2.1589,
-      "step": 757
-    },
-    {
-      "epoch": 2.5446915652538817,
-      "grad_norm": 0.5842500925064087,
-      "learning_rate": 7.860520094562648e-08,
-      "loss": 2.1216,
-      "step": 758
-    },
-    {
-      "epoch": 2.5480486781368024,
-      "grad_norm": 0.5609316229820251,
-      "learning_rate": 7.801418439716311e-08,
-      "loss": 2.1259,
-      "step": 759
-    },
-    {
-      "epoch": 2.551405791019723,
-      "grad_norm": 0.563157856464386,
-      "learning_rate": 7.742316784869976e-08,
-      "loss": 2.1739,
-      "step": 760
-    },
-    {
-      "epoch": 2.5547629039026436,
-      "grad_norm": 0.5769089460372925,
-      "learning_rate": 7.68321513002364e-08,
-      "loss": 2.1196,
-      "step": 761
-    },
-    {
-      "epoch": 2.5581200167855647,
-      "grad_norm": 0.5630622506141663,
-      "learning_rate": 7.624113475177305e-08,
-      "loss": 2.1243,
-      "step": 762
-    },
-    {
-      "epoch": 2.561477129668485,
-      "grad_norm": 0.5799427032470703,
-      "learning_rate": 7.565011820330969e-08,
-      "loss": 2.0763,
-      "step": 763
-    },
-    {
-      "epoch": 2.564834242551406,
-      "grad_norm": 0.5502346754074097,
-      "learning_rate": 7.505910165484633e-08,
-      "loss": 2.1204,
-      "step": 764
-    },
-    {
-      "epoch": 2.5681913554343265,
-      "grad_norm": 0.5592603087425232,
-      "learning_rate": 7.446808510638298e-08,
-      "loss": 2.1178,
-      "step": 765
-    },
-    {
-      "epoch": 2.571548468317247,
-      "grad_norm": 0.5861775875091553,
-      "learning_rate": 7.387706855791961e-08,
-      "loss": 2.2152,
-      "step": 766
-    },
-    {
-      "epoch": 2.5749055812001678,
-      "grad_norm": 0.5786147713661194,
-      "learning_rate": 7.328605200945625e-08,
-      "loss": 2.1161,
-      "step": 767
-    },
-    {
-      "epoch": 2.5782626940830884,
-      "grad_norm": 0.5833945274353027,
-      "learning_rate": 7.26950354609929e-08,
-      "loss": 2.1747,
-      "step": 768
-    },
-    {
-      "epoch": 2.5816198069660095,
-      "grad_norm": 0.5661745667457581,
-      "learning_rate": 7.210401891252955e-08,
-      "loss": 2.1337,
-      "step": 769
-    },
-    {
-      "epoch": 2.58497691984893,
-      "grad_norm": 0.57346510887146,
-      "learning_rate": 7.15130023640662e-08,
-      "loss": 2.1262,
-      "step": 770
-    },
-    {
-      "epoch": 2.5883340327318507,
-      "grad_norm": 0.5714176893234253,
-      "learning_rate": 7.092198581560284e-08,
-      "loss": 2.1607,
-      "step": 771
-    },
-    {
-      "epoch": 2.5916911456147713,
-      "grad_norm": 0.5676454901695251,
-      "learning_rate": 7.033096926713948e-08,
-      "loss": 2.1609,
-      "step": 772
-    },
-    {
-      "epoch": 2.595048258497692,
-      "grad_norm": 0.6049889326095581,
-      "learning_rate": 6.973995271867613e-08,
-      "loss": 2.1581,
-      "step": 773
-    },
-    {
-      "epoch": 2.5984053713806126,
-      "grad_norm": 0.5774138569831848,
-      "learning_rate": 6.914893617021277e-08,
-      "loss": 2.111,
-      "step": 774
-    },
-    {
-      "epoch": 2.601762484263533,
-      "grad_norm": 0.5626241564750671,
-      "learning_rate": 6.85579196217494e-08,
-      "loss": 2.1279,
-      "step": 775
-    },
-    {
-      "epoch": 2.6051195971464542,
-      "grad_norm": 0.5682496428489685,
-      "learning_rate": 6.796690307328605e-08,
-      "loss": 2.1819,
-      "step": 776
-    },
-    {
-      "epoch": 2.608476710029375,
-      "grad_norm": 0.5871005058288574,
-      "learning_rate": 6.737588652482269e-08,
-      "loss": 2.1623,
-      "step": 777
-    },
-    {
-      "epoch": 2.6118338229122955,
-      "grad_norm": 0.6054531335830688,
-      "learning_rate": 6.678486997635934e-08,
-      "loss": 2.1136,
-      "step": 778
-    },
-    {
-      "epoch": 2.615190935795216,
-      "grad_norm": 0.5541722774505615,
-      "learning_rate": 6.619385342789598e-08,
-      "loss": 2.1334,
-      "step": 779
-    },
-    {
-      "epoch": 2.6185480486781367,
-      "grad_norm": 0.556919515132904,
-      "learning_rate": 6.560283687943262e-08,
-      "loss": 2.1148,
-      "step": 780
-    },
-    {
-      "epoch": 2.6219051615610574,
-      "grad_norm": 0.5363909602165222,
-      "learning_rate": 6.501182033096927e-08,
-      "loss": 2.0795,
-      "step": 781
-    },
-    {
-      "epoch": 2.625262274443978,
-      "grad_norm": 0.5697099566459656,
-      "learning_rate": 6.44208037825059e-08,
-      "loss": 2.1522,
-      "step": 782
-    },
-    {
-      "epoch": 2.628619387326899,
-      "grad_norm": 0.571094810962677,
-      "learning_rate": 6.382978723404254e-08,
-      "loss": 2.1545,
-      "step": 783
-    },
-    {
-      "epoch": 2.6319765002098197,
-      "grad_norm": 0.5669984817504883,
-      "learning_rate": 6.323877068557919e-08,
-      "loss": 2.1427,
-      "step": 784
-    },
-    {
-      "epoch": 2.6353336130927403,
-      "grad_norm": 0.5831683874130249,
-      "learning_rate": 6.264775413711583e-08,
-      "loss": 2.188,
-      "step": 785
-    },
-    {
-      "epoch": 2.638690725975661,
-      "grad_norm": 0.5660560131072998,
-      "learning_rate": 6.205673758865248e-08,
-      "loss": 2.1572,
-      "step": 786
-    },
-    {
-      "epoch": 2.6420478388585815,
-      "grad_norm": 0.5728089809417725,
-      "learning_rate": 6.146572104018912e-08,
-      "loss": 2.0789,
-      "step": 787
-    },
-    {
-      "epoch": 2.645404951741502,
-      "grad_norm": 0.5673388242721558,
-      "learning_rate": 6.087470449172577e-08,
-      "loss": 2.1508,
-      "step": 788
-    },
-    {
-      "epoch": 2.6487620646244228,
-      "grad_norm": 0.568686842918396,
-      "learning_rate": 6.02836879432624e-08,
-      "loss": 2.1811,
-      "step": 789
-    },
-    {
-      "epoch": 2.652119177507344,
-      "grad_norm": 0.5477533936500549,
-      "learning_rate": 5.969267139479905e-08,
-      "loss": 2.1047,
-      "step": 790
-    },
-    {
-      "epoch": 2.6554762903902644,
-      "grad_norm": 0.627791702747345,
-      "learning_rate": 5.91016548463357e-08,
-      "loss": 2.1543,
-      "step": 791
-    },
-    {
-      "epoch": 2.658833403273185,
-      "grad_norm": 0.5684884786605835,
-      "learning_rate": 5.851063829787234e-08,
-      "loss": 2.1225,
-      "step": 792
-    },
-    {
-      "epoch": 2.6621905161561057,
-      "grad_norm": 0.6162928938865662,
-      "learning_rate": 5.791962174940898e-08,
-      "loss": 2.1343,
-      "step": 793
-    },
-    {
-      "epoch": 2.6655476290390263,
-      "grad_norm": 0.5560017228126526,
-      "learning_rate": 5.732860520094562e-08,
-      "loss": 2.1797,
-      "step": 794
-    },
-    {
-      "epoch": 2.6689047419219474,
-      "grad_norm": 0.5683034658432007,
-      "learning_rate": 5.673758865248227e-08,
-      "loss": 2.1552,
-      "step": 795
-    },
-    {
-      "epoch": 2.6722618548048676,
-      "grad_norm": 0.6046074628829956,
-      "learning_rate": 5.614657210401891e-08,
-      "loss": 2.1549,
-      "step": 796
-    },
-    {
-      "epoch": 2.6756189676877886,
-      "grad_norm": 0.5512291789054871,
-      "learning_rate": 5.555555555555555e-08,
-      "loss": 2.1443,
-      "step": 797
-    },
-    {
-      "epoch": 2.6789760805707092,
-      "grad_norm": 0.5601515173912048,
-      "learning_rate": 5.49645390070922e-08,
-      "loss": 2.135,
-      "step": 798
-    },
-    {
-      "epoch": 2.68233319345363,
-      "grad_norm": 0.5837010741233826,
-      "learning_rate": 5.437352245862884e-08,
-      "loss": 2.0827,
-      "step": 799
-    },
-    {
-      "epoch": 2.6856903063365505,
-      "grad_norm": 0.5594925284385681,
-      "learning_rate": 5.3782505910165486e-08,
-      "loss": 2.1214,
-      "step": 800
-    },
-    {
-      "epoch": 2.689047419219471,
-      "grad_norm": 0.5061306953430176,
-      "learning_rate": 5.3191489361702123e-08,
-      "loss": 2.1342,
-      "step": 801
-    },
-    {
-      "epoch": 2.692404532102392,
-      "grad_norm": 0.5876765251159668,
-      "learning_rate": 5.260047281323877e-08,
-      "loss": 2.1541,
-      "step": 802
-    },
-    {
-      "epoch": 2.695761644985313,
-      "grad_norm": 0.5789502263069153,
-      "learning_rate": 5.200945626477541e-08,
-      "loss": 2.1463,
-      "step": 803
-    },
-    {
-      "epoch": 2.6991187578682334,
-      "grad_norm": 0.579647421836853,
-      "learning_rate": 5.1418439716312055e-08,
-      "loss": 2.1836,
-      "step": 804
-    },
-    {
-      "epoch": 2.702475870751154,
-      "grad_norm": 0.5416697859764099,
-      "learning_rate": 5.082742316784869e-08,
-      "loss": 2.0999,
-      "step": 805
-    },
-    {
-      "epoch": 2.7058329836340747,
-      "grad_norm": 0.5642104148864746,
-      "learning_rate": 5.023640661938534e-08,
-      "loss": 2.1421,
-      "step": 806
-    },
-    {
-      "epoch": 2.7091900965169953,
-      "grad_norm": 0.554743766784668,
-      "learning_rate": 4.964539007092199e-08,
-      "loss": 2.1678,
-      "step": 807
-    },
-    {
-      "epoch": 2.712547209399916,
-      "grad_norm": 0.5825777053833008,
-      "learning_rate": 4.905437352245863e-08,
-      "loss": 2.1865,
-      "step": 808
-    },
-    {
-      "epoch": 2.715904322282837,
-      "grad_norm": 0.5957197546958923,
-      "learning_rate": 4.846335697399527e-08,
-      "loss": 2.1653,
-      "step": 809
-    },
-    {
-      "epoch": 2.7192614351657576,
-      "grad_norm": 0.5674082040786743,
-      "learning_rate": 4.787234042553191e-08,
-      "loss": 2.1615,
-      "step": 810
-    },
-    {
-      "epoch": 2.722618548048678,
-      "grad_norm": 0.5784925818443298,
-      "learning_rate": 4.7281323877068556e-08,
-      "loss": 2.1737,
-      "step": 811
-    },
-    {
-      "epoch": 2.725975660931599,
-      "grad_norm": 0.5676373243331909,
-      "learning_rate": 4.66903073286052e-08,
-      "loss": 2.0867,
-      "step": 812
-    },
-    {
-      "epoch": 2.7293327738145194,
-      "grad_norm": 0.5750331878662109,
-      "learning_rate": 4.609929078014184e-08,
-      "loss": 2.1381,
-      "step": 813
-    },
-    {
-      "epoch": 2.73268988669744,
-      "grad_norm": 0.572400689125061,
-      "learning_rate": 4.550827423167849e-08,
-      "loss": 2.128,
-      "step": 814
-    },
-    {
-      "epoch": 2.7360469995803607,
-      "grad_norm": 0.5287570953369141,
-      "learning_rate": 4.491725768321513e-08,
-      "loss": 2.041,
-      "step": 815
-    },
-    {
-      "epoch": 2.7394041124632817,
-      "grad_norm": 0.5819547176361084,
-      "learning_rate": 4.4326241134751775e-08,
-      "loss": 2.1262,
-      "step": 816
-    },
-    {
-      "epoch": 2.7427612253462024,
-      "grad_norm": 0.5426435470581055,
-      "learning_rate": 4.373522458628841e-08,
-      "loss": 2.1539,
-      "step": 817
-    },
-    {
-      "epoch": 2.746118338229123,
-      "grad_norm": 0.5462058782577515,
-      "learning_rate": 4.3144208037825056e-08,
-      "loss": 2.1055,
-      "step": 818
-    },
-    {
-      "epoch": 2.7494754511120436,
-      "grad_norm": 0.5718535780906677,
-      "learning_rate": 4.25531914893617e-08,
-      "loss": 2.1508,
-      "step": 819
-    },
-    {
-      "epoch": 2.7528325639949642,
-      "grad_norm": 0.5446327328681946,
-      "learning_rate": 4.1962174940898344e-08,
-      "loss": 2.2267,
-      "step": 820
-    },
-    {
-      "epoch": 2.756189676877885,
-      "grad_norm": 0.5311114192008972,
-      "learning_rate": 4.137115839243498e-08,
-      "loss": 2.0866,
-      "step": 821
-    },
-    {
-      "epoch": 2.7595467897608055,
-      "grad_norm": 0.6080516576766968,
-      "learning_rate": 4.078014184397163e-08,
-      "loss": 2.1363,
-      "step": 822
-    },
-    {
-      "epoch": 2.7629039026437265,
-      "grad_norm": 0.5661442875862122,
-      "learning_rate": 4.0189125295508276e-08,
-      "loss": 2.1195,
-      "step": 823
-    },
-    {
-      "epoch": 2.766261015526647,
-      "grad_norm": 0.5298861265182495,
-      "learning_rate": 3.959810874704492e-08,
-      "loss": 2.1442,
-      "step": 824
-    },
-    {
-      "epoch": 2.769618128409568,
-      "grad_norm": 0.5596882700920105,
-      "learning_rate": 3.900709219858156e-08,
-      "loss": 2.0943,
-      "step": 825
-    },
-    {
-      "epoch": 2.7729752412924884,
-      "grad_norm": 0.5581501126289368,
-      "learning_rate": 3.84160756501182e-08,
-      "loss": 2.1547,
-      "step": 826
-    },
-    {
-      "epoch": 2.776332354175409,
-      "grad_norm": 0.5514313578605652,
-      "learning_rate": 3.7825059101654844e-08,
-      "loss": 2.0646,
-      "step": 827
-    },
-    {
-      "epoch": 2.77968946705833,
-      "grad_norm": 0.5894129872322083,
-      "learning_rate": 3.723404255319149e-08,
-      "loss": 2.1742,
-      "step": 828
-    },
-    {
-      "epoch": 2.7830465799412503,
-      "grad_norm": 0.5762084126472473,
-      "learning_rate": 3.6643026004728126e-08,
-      "loss": 2.1449,
-      "step": 829
-    },
-    {
-      "epoch": 2.7864036928241713,
-      "grad_norm": 0.5430789589881897,
-      "learning_rate": 3.6052009456264776e-08,
-      "loss": 2.1861,
-      "step": 830
-    },
-    {
-      "epoch": 2.789760805707092,
-      "grad_norm": 0.5707167983055115,
-      "learning_rate": 3.546099290780142e-08,
-      "loss": 2.1671,
-      "step": 831
-    },
-    {
-      "epoch": 2.7931179185900126,
-      "grad_norm": 0.5500497221946716,
-      "learning_rate": 3.4869976359338064e-08,
-      "loss": 2.135,
-      "step": 832
-    },
-    {
-      "epoch": 2.796475031472933,
-      "grad_norm": 0.5838862061500549,
-      "learning_rate": 3.42789598108747e-08,
-      "loss": 2.099,
-      "step": 833
-    },
-    {
-      "epoch": 2.799832144355854,
-      "grad_norm": 0.5668932199478149,
-      "learning_rate": 3.3687943262411345e-08,
-      "loss": 2.156,
-      "step": 834
-    },
-    {
-      "epoch": 2.803189257238775,
-      "grad_norm": 0.5672734975814819,
-      "learning_rate": 3.309692671394799e-08,
-      "loss": 2.1534,
-      "step": 835
-    },
-    {
-      "epoch": 2.8065463701216955,
-      "grad_norm": 0.593976616859436,
-      "learning_rate": 3.250591016548463e-08,
-      "loss": 2.1516,
-      "step": 836
-    },
-    {
-      "epoch": 2.809903483004616,
-      "grad_norm": 0.5572679042816162,
-      "learning_rate": 3.191489361702127e-08,
-      "loss": 2.1574,
-      "step": 837
-    },
-    {
-      "epoch": 2.8132605958875367,
-      "grad_norm": 0.570029079914093,
-      "learning_rate": 3.1323877068557914e-08,
-      "loss": 2.1257,
-      "step": 838
-    },
-    {
-      "epoch": 2.8166177087704574,
-      "grad_norm": 0.5792723894119263,
-      "learning_rate": 3.073286052009456e-08,
-      "loss": 2.181,
-      "step": 839
-    },
-    {
-      "epoch": 2.819974821653378,
-      "grad_norm": 0.5601706504821777,
-      "learning_rate": 3.01418439716312e-08,
-      "loss": 2.1163,
-      "step": 840
-    },
-    {
-      "epoch": 2.8233319345362986,
-      "grad_norm": 0.6072015166282654,
-      "learning_rate": 2.955082742316785e-08,
-      "loss": 2.2281,
-      "step": 841
-    },
-    {
-      "epoch": 2.8266890474192197,
-      "grad_norm": 0.5666745901107788,
-      "learning_rate": 2.895981087470449e-08,
-      "loss": 2.1562,
-      "step": 842
-    },
-    {
-      "epoch": 2.8300461603021403,
-      "grad_norm": 0.555409848690033,
-      "learning_rate": 2.8368794326241133e-08,
-      "loss": 2.1212,
-      "step": 843
-    },
-    {
-      "epoch": 2.833403273185061,
-      "grad_norm": 0.5660236477851868,
-      "learning_rate": 2.7777777777777774e-08,
-      "loss": 2.1268,
-      "step": 844
-    },
-    {
-      "epoch": 2.8367603860679815,
-      "grad_norm": 0.5438356399536133,
-      "learning_rate": 2.718676122931442e-08,
-      "loss": 2.1104,
-      "step": 845
-    },
-    {
-      "epoch": 2.840117498950902,
-      "grad_norm": 0.5927281379699707,
-      "learning_rate": 2.6595744680851062e-08,
-      "loss": 2.1474,
-      "step": 846
-    },
-    {
-      "epoch": 2.8434746118338228,
-      "grad_norm": 0.551418125629425,
-      "learning_rate": 2.6004728132387706e-08,
-      "loss": 2.148,
-      "step": 847
-    },
-    {
-      "epoch": 2.8468317247167434,
-      "grad_norm": 0.5623355507850647,
-      "learning_rate": 2.5413711583924346e-08,
-      "loss": 2.1714,
-      "step": 848
-    },
-    {
-      "epoch": 2.8501888375996645,
-      "grad_norm": 0.5470379590988159,
-      "learning_rate": 2.4822695035460993e-08,
-      "loss": 2.1373,
-      "step": 849
-    },
-    {
-      "epoch": 2.853545950482585,
-      "grad_norm": 0.5807493329048157,
-      "learning_rate": 2.4231678486997634e-08,
-      "loss": 2.0945,
-      "step": 850
-    },
-    {
-      "epoch": 2.8569030633655057,
-      "grad_norm": 0.5901828408241272,
-      "learning_rate": 2.3640661938534278e-08,
-      "loss": 2.1589,
-      "step": 851
-    },
-    {
-      "epoch": 2.8602601762484263,
-      "grad_norm": 0.5616815090179443,
-      "learning_rate": 2.304964539007092e-08,
-      "loss": 2.019,
-      "step": 852
-    },
-    {
-      "epoch": 2.863617289131347,
-      "grad_norm": 0.5549256801605225,
-      "learning_rate": 2.2458628841607566e-08,
-      "loss": 2.1154,
-      "step": 853
-    },
-    {
-      "epoch": 2.8669744020142676,
-      "grad_norm": 0.58626788854599,
-      "learning_rate": 2.1867612293144206e-08,
-      "loss": 2.1385,
-      "step": 854
-    },
-    {
-      "epoch": 2.870331514897188,
-      "grad_norm": 0.5496333241462708,
-      "learning_rate": 2.127659574468085e-08,
-      "loss": 2.1115,
-      "step": 855
-    },
-    {
-      "epoch": 2.8736886277801093,
-      "grad_norm": 0.5691447257995605,
-      "learning_rate": 2.068557919621749e-08,
-      "loss": 2.1418,
-      "step": 856
-    },
-    {
-      "epoch": 2.87704574066303,
-      "grad_norm": 0.5428998470306396,
-      "learning_rate": 2.0094562647754138e-08,
-      "loss": 2.0914,
-      "step": 857
-    },
-    {
-      "epoch": 2.8804028535459505,
-      "grad_norm": 0.5734657049179077,
-      "learning_rate": 1.950354609929078e-08,
-      "loss": 2.2244,
-      "step": 858
-    },
-    {
-      "epoch": 2.883759966428871,
-      "grad_norm": 0.5396918058395386,
-      "learning_rate": 1.8912529550827422e-08,
-      "loss": 2.0997,
-      "step": 859
-    },
-    {
-      "epoch": 2.8871170793117917,
-      "grad_norm": 0.5532637238502502,
-      "learning_rate": 1.8321513002364063e-08,
-      "loss": 2.1957,
-      "step": 860
-    },
-    {
-      "epoch": 2.890474192194713,
-      "grad_norm": 0.5601110458374023,
-      "learning_rate": 1.773049645390071e-08,
-      "loss": 2.1385,
-      "step": 861
-    },
-    {
-      "epoch": 2.893831305077633,
-      "grad_norm": 0.5704315304756165,
-      "learning_rate": 1.713947990543735e-08,
-      "loss": 2.1373,
-      "step": 862
-    },
-    {
-      "epoch": 2.897188417960554,
-      "grad_norm": 0.5758122205734253,
-      "learning_rate": 1.6548463356973994e-08,
-      "loss": 2.124,
-      "step": 863
-    },
-    {
-      "epoch": 2.9005455308434747,
-      "grad_norm": 0.5552284121513367,
-      "learning_rate": 1.5957446808510635e-08,
-      "loss": 2.114,
-      "step": 864
-    },
-    {
-      "epoch": 2.9039026437263953,
-      "grad_norm": 0.600894033908844,
-      "learning_rate": 1.536643026004728e-08,
-      "loss": 2.1908,
-      "step": 865
-    },
-    {
-      "epoch": 2.907259756609316,
-      "grad_norm": 0.5430445075035095,
-      "learning_rate": 1.4775413711583924e-08,
-      "loss": 2.0722,
-      "step": 866
-    },
-    {
-      "epoch": 2.9106168694922365,
-      "grad_norm": 0.564946711063385,
-      "learning_rate": 1.4184397163120567e-08,
-      "loss": 2.1676,
-      "step": 867
-    },
-    {
-      "epoch": 2.9139739823751576,
-      "grad_norm": 0.568100094795227,
-      "learning_rate": 1.359338061465721e-08,
-      "loss": 2.0945,
-      "step": 868
-    },
-    {
-      "epoch": 2.917331095258078,
-      "grad_norm": 0.553946316242218,
-      "learning_rate": 1.3002364066193853e-08,
-      "loss": 2.1065,
-      "step": 869
-    },
-    {
-      "epoch": 2.920688208140999,
-      "grad_norm": 0.5421509146690369,
-      "learning_rate": 1.2411347517730497e-08,
-      "loss": 2.0381,
-      "step": 870
-    },
-    {
-      "epoch": 2.9240453210239195,
-      "grad_norm": 0.543483316898346,
-      "learning_rate": 1.1820330969267139e-08,
-      "loss": 2.1513,
-      "step": 871
-    },
-    {
-      "epoch": 2.92740243390684,
-      "grad_norm": 0.5386929512023926,
-      "learning_rate": 1.1229314420803783e-08,
-      "loss": 2.164,
-      "step": 872
-    },
-    {
-      "epoch": 2.9307595467897607,
-      "grad_norm": 0.5322523713111877,
-      "learning_rate": 1.0638297872340425e-08,
-      "loss": 2.12,
-      "step": 873
-    },
-    {
-      "epoch": 2.9341166596726813,
-      "grad_norm": 0.5454387068748474,
-      "learning_rate": 1.0047281323877069e-08,
-      "loss": 2.1669,
-      "step": 874
-    },
-    {
-      "epoch": 2.9374737725556024,
-      "grad_norm": 0.5460171103477478,
-      "learning_rate": 9.456264775413711e-09,
-      "loss": 2.0947,
-      "step": 875
-    },
-    {
-      "epoch": 2.940830885438523,
-      "grad_norm": 0.5782340168952942,
-      "learning_rate": 8.865248226950355e-09,
-      "loss": 2.1667,
-      "step": 876
-    },
-    {
-      "epoch": 2.9441879983214436,
-      "grad_norm": 0.5890206694602966,
-      "learning_rate": 8.274231678486997e-09,
-      "loss": 2.2306,
-      "step": 877
-    },
-    {
-      "epoch": 2.9475451112043642,
-      "grad_norm": 0.5283526182174683,
-      "learning_rate": 7.68321513002364e-09,
-      "loss": 2.163,
-      "step": 878
-    },
-    {
-      "epoch": 2.950902224087285,
-      "grad_norm": 0.578064501285553,
-      "learning_rate": 7.092198581560283e-09,
-      "loss": 2.099,
-      "step": 879
-    },
-    {
-      "epoch": 2.9542593369702055,
-      "grad_norm": 0.5312917232513428,
-      "learning_rate": 6.501182033096926e-09,
-      "loss": 2.1152,
-      "step": 880
-    },
-    {
-      "epoch": 2.957616449853126,
-      "grad_norm": 0.5620092749595642,
-      "learning_rate": 5.9101654846335695e-09,
-      "loss": 2.0636,
-      "step": 881
-    },
-    {
-      "epoch": 2.960973562736047,
-      "grad_norm": 0.572350025177002,
-      "learning_rate": 5.3191489361702125e-09,
-      "loss": 2.1626,
-      "step": 882
-    },
-    {
-      "epoch": 2.964330675618968,
-      "grad_norm": 0.6169728636741638,
-      "learning_rate": 4.7281323877068556e-09,
-      "loss": 2.1022,
-      "step": 883
-    },
-    {
-      "epoch": 2.9676877885018884,
-      "grad_norm": 0.5962982773780823,
-      "learning_rate": 4.137115839243499e-09,
-      "loss": 2.1747,
-      "step": 884
-    },
-    {
-      "epoch": 2.971044901384809,
-      "grad_norm": 0.5752913355827332,
-      "learning_rate": 3.5460992907801417e-09,
-      "loss": 2.1499,
-      "step": 885
-    },
-    {
-      "epoch": 2.9744020142677297,
-      "grad_norm": 0.5664993524551392,
-      "learning_rate": 2.9550827423167847e-09,
-      "loss": 2.1381,
-      "step": 886
-    },
-    {
-      "epoch": 2.9777591271506503,
-      "grad_norm": 0.5439576506614685,
-      "learning_rate": 2.3640661938534278e-09,
-      "loss": 2.1562,
-      "step": 887
-    },
-    {
-      "epoch": 2.981116240033571,
-      "grad_norm": 0.5637330412864685,
-      "learning_rate": 1.7730496453900708e-09,
-      "loss": 2.1475,
-      "step": 888
-    },
-    {
-      "epoch": 2.984473352916492,
-      "grad_norm": 0.6112098693847656,
-      "learning_rate": 1.1820330969267139e-09,
-      "loss": 2.1624,
-      "step": 889
-    },
-    {
-      "epoch": 2.9878304657994126,
-      "grad_norm": 0.5705456733703613,
-      "learning_rate": 5.910165484633569e-10,
-      "loss": 2.1383,
-      "step": 890
-    },
-    {
-      "epoch": 2.991187578682333,
-      "grad_norm": 0.5267236232757568,
-      "learning_rate": 0.0,
-      "loss": 2.1183,
-      "step": 891
-    },
-    {
-      "epoch": 2.991187578682333,
-      "eval_loss": 2.1491148471832275,
-      "eval_runtime": 325.8656,
-      "eval_samples_per_second": 1.111,
-      "eval_steps_per_second": 0.279,
-      "step": 891
     }
   ],
   "logging_steps": 1,
@@ -6282,12 +4202,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }
   },
-  "total_flos": 2.63058527225276e+18,
+  "total_flos": 1.758890208692011e+18,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null