{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3930,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002544529262086514,
      "grad_norm": 2.004002094268799,
      "learning_rate": 3.0508474576271192e-06,
      "loss": 0.1146,
      "step": 10
    },
    {
      "epoch": 0.005089058524173028,
      "grad_norm": 2.2637343406677246,
      "learning_rate": 6.440677966101695e-06,
      "loss": 0.1546,
      "step": 20
    },
    {
      "epoch": 0.007633587786259542,
      "grad_norm": 1.814893364906311,
      "learning_rate": 9.830508474576272e-06,
      "loss": 0.0832,
      "step": 30
    },
    {
      "epoch": 0.010178117048346057,
      "grad_norm": 2.5804829597473145,
      "learning_rate": 1.3220338983050848e-05,
      "loss": 0.0677,
      "step": 40
    },
    {
      "epoch": 0.01272264631043257,
      "grad_norm": 2.116039991378784,
      "learning_rate": 1.6610169491525424e-05,
      "loss": 0.0305,
      "step": 50
    },
    {
      "epoch": 0.015267175572519083,
      "grad_norm": 3.1931254863739014,
      "learning_rate": 2e-05,
      "loss": 0.0375,
      "step": 60
    },
    {
      "epoch": 0.017811704834605598,
      "grad_norm": 0.12293073534965515,
      "learning_rate": 2.338983050847458e-05,
      "loss": 0.0239,
      "step": 70
    },
    {
      "epoch": 0.020356234096692113,
      "grad_norm": 1.320569634437561,
      "learning_rate": 2.6779661016949153e-05,
      "loss": 0.0311,
      "step": 80
    },
    {
      "epoch": 0.022900763358778626,
      "grad_norm": 1.4948492050170898,
      "learning_rate": 3.016949152542373e-05,
      "loss": 0.0258,
      "step": 90
    },
    {
      "epoch": 0.02544529262086514,
      "grad_norm": 1.2656677961349487,
      "learning_rate": 3.355932203389831e-05,
      "loss": 0.0258,
      "step": 100
    },
    {
      "epoch": 0.027989821882951654,
      "grad_norm": 1.5219528675079346,
      "learning_rate": 3.6949152542372886e-05,
      "loss": 0.0109,
      "step": 110
    },
    {
      "epoch": 0.030534351145038167,
      "grad_norm": 0.7582465410232544,
      "learning_rate": 3.9989506820566634e-05,
      "loss": 0.0178,
      "step": 120
    },
    {
      "epoch": 0.03307888040712468,
      "grad_norm": 0.43421927094459534,
      "learning_rate": 3.9884575026232955e-05,
      "loss": 0.018,
      "step": 130
    },
    {
      "epoch": 0.035623409669211195,
      "grad_norm": 0.03073570877313614,
      "learning_rate": 3.977964323189927e-05,
      "loss": 0.0135,
      "step": 140
    },
    {
      "epoch": 0.03816793893129771,
      "grad_norm": 0.28750094771385193,
      "learning_rate": 3.967471143756559e-05,
      "loss": 0.0173,
      "step": 150
    },
    {
      "epoch": 0.04071246819338423,
      "grad_norm": 0.47052061557769775,
      "learning_rate": 3.9569779643231905e-05,
      "loss": 0.0117,
      "step": 160
    },
    {
      "epoch": 0.043256997455470736,
      "grad_norm": 1.7329446077346802,
      "learning_rate": 3.946484784889822e-05,
      "loss": 0.0089,
      "step": 170
    },
    {
      "epoch": 0.04580152671755725,
      "grad_norm": 0.037718575447797775,
      "learning_rate": 3.935991605456454e-05,
      "loss": 0.0202,
      "step": 180
    },
    {
      "epoch": 0.04834605597964377,
      "grad_norm": 0.4059722423553467,
      "learning_rate": 3.9254984260230855e-05,
      "loss": 0.0066,
      "step": 190
    },
    {
      "epoch": 0.05089058524173028,
      "grad_norm": 0.043499384075403214,
      "learning_rate": 3.915005246589717e-05,
      "loss": 0.0111,
      "step": 200
    },
    {
      "epoch": 0.05343511450381679,
      "grad_norm": 0.504442036151886,
      "learning_rate": 3.904512067156349e-05,
      "loss": 0.0188,
      "step": 210
    },
    {
      "epoch": 0.05597964376590331,
      "grad_norm": 1.7431628704071045,
      "learning_rate": 3.8940188877229805e-05,
      "loss": 0.0129,
      "step": 220
    },
    {
      "epoch": 0.058524173027989825,
      "grad_norm": 0.18788424134254456,
      "learning_rate": 3.8835257082896126e-05,
      "loss": 0.0155,
      "step": 230
    },
    {
      "epoch": 0.061068702290076333,
      "grad_norm": 0.03526950255036354,
      "learning_rate": 3.873032528856244e-05,
      "loss": 0.0134,
      "step": 240
    },
    {
      "epoch": 0.06361323155216285,
      "grad_norm": 0.7251449823379517,
      "learning_rate": 3.8625393494228755e-05,
      "loss": 0.0078,
      "step": 250
    },
    {
      "epoch": 0.06615776081424936,
      "grad_norm": 0.05955138057470322,
      "learning_rate": 3.8520461699895076e-05,
      "loss": 0.016,
      "step": 260
    },
    {
      "epoch": 0.06870229007633588,
      "grad_norm": 0.11732513457536697,
      "learning_rate": 3.841552990556139e-05,
      "loss": 0.022,
      "step": 270
    },
    {
      "epoch": 0.07124681933842239,
      "grad_norm": 0.03472783789038658,
      "learning_rate": 3.8310598111227705e-05,
      "loss": 0.013,
      "step": 280
    },
    {
      "epoch": 0.0737913486005089,
      "grad_norm": 0.7197083234786987,
      "learning_rate": 3.8205666316894026e-05,
      "loss": 0.0102,
      "step": 290
    },
    {
      "epoch": 0.07633587786259542,
      "grad_norm": 0.03828183189034462,
      "learning_rate": 3.810073452256034e-05,
      "loss": 0.0078,
      "step": 300
    },
    {
      "epoch": 0.07888040712468193,
      "grad_norm": 0.6017931699752808,
      "learning_rate": 3.7995802728226654e-05,
      "loss": 0.0205,
      "step": 310
    },
    {
      "epoch": 0.08142493638676845,
      "grad_norm": 0.44738417863845825,
      "learning_rate": 3.7890870933892976e-05,
      "loss": 0.0178,
      "step": 320
    },
    {
      "epoch": 0.08396946564885496,
      "grad_norm": 0.03330178186297417,
      "learning_rate": 3.778593913955929e-05,
      "loss": 0.0145,
      "step": 330
    },
    {
      "epoch": 0.08651399491094147,
      "grad_norm": 0.01718255504965782,
      "learning_rate": 3.768100734522561e-05,
      "loss": 0.0147,
      "step": 340
    },
    {
      "epoch": 0.089058524173028,
      "grad_norm": 0.8818934559822083,
      "learning_rate": 3.7576075550891925e-05,
      "loss": 0.0128,
      "step": 350
    },
    {
      "epoch": 0.0916030534351145,
      "grad_norm": 0.6107049584388733,
      "learning_rate": 3.747114375655824e-05,
      "loss": 0.0182,
      "step": 360
    },
    {
      "epoch": 0.09414758269720101,
      "grad_norm": 0.7103002071380615,
      "learning_rate": 3.736621196222456e-05,
      "loss": 0.005,
      "step": 370
    },
    {
      "epoch": 0.09669211195928754,
      "grad_norm": 0.548647403717041,
      "learning_rate": 3.7261280167890875e-05,
      "loss": 0.0201,
      "step": 380
    },
    {
      "epoch": 0.09923664122137404,
      "grad_norm": 0.0742243304848671,
      "learning_rate": 3.715634837355719e-05,
      "loss": 0.0129,
      "step": 390
    },
    {
      "epoch": 0.10178117048346055,
      "grad_norm": 1.2889010906219482,
      "learning_rate": 3.705141657922351e-05,
      "loss": 0.0231,
      "step": 400
    },
    {
      "epoch": 0.10432569974554708,
      "grad_norm": 1.0989108085632324,
      "learning_rate": 3.6946484784889825e-05,
      "loss": 0.0149,
      "step": 410
    },
    {
      "epoch": 0.10687022900763359,
      "grad_norm": 0.6415719985961914,
      "learning_rate": 3.684155299055614e-05,
      "loss": 0.0143,
      "step": 420
    },
    {
      "epoch": 0.10941475826972011,
      "grad_norm": 0.5659549236297607,
      "learning_rate": 3.673662119622246e-05,
      "loss": 0.0149,
      "step": 430
    },
    {
      "epoch": 0.11195928753180662,
      "grad_norm": 0.05708279460668564,
      "learning_rate": 3.6631689401888775e-05,
      "loss": 0.0043,
      "step": 440
    },
    {
      "epoch": 0.11450381679389313,
      "grad_norm": 1.086536169052124,
      "learning_rate": 3.652675760755509e-05,
      "loss": 0.007,
      "step": 450
    },
    {
      "epoch": 0.11704834605597965,
      "grad_norm": 0.17354388535022736,
      "learning_rate": 3.642182581322141e-05,
      "loss": 0.0037,
      "step": 460
    },
    {
      "epoch": 0.11959287531806616,
      "grad_norm": 0.34753769636154175,
      "learning_rate": 3.6316894018887725e-05,
      "loss": 0.0058,
      "step": 470
    },
    {
      "epoch": 0.12213740458015267,
      "grad_norm": 0.010063248686492443,
      "learning_rate": 3.6211962224554046e-05,
      "loss": 0.0097,
      "step": 480
    },
    {
      "epoch": 0.12468193384223919,
      "grad_norm": 0.027567019686102867,
      "learning_rate": 3.610703043022036e-05,
      "loss": 0.0095,
      "step": 490
    },
    {
      "epoch": 0.1272264631043257,
      "grad_norm": 0.16245301067829132,
      "learning_rate": 3.6002098635886675e-05,
      "loss": 0.0169,
      "step": 500
    },
    {
      "epoch": 0.1272264631043257,
      "eval_loss": 0.03691313415765762,
      "eval_runtime": 134.0261,
      "eval_samples_per_second": 59.556,
      "eval_steps_per_second": 0.47,
      "step": 500
    },
    {
      "epoch": 0.1297709923664122,
      "grad_norm": 1.1503355503082275,
      "learning_rate": 3.5897166841552996e-05,
      "loss": 0.0209,
      "step": 510
    },
    {
      "epoch": 0.13231552162849872,
      "grad_norm": 0.19143138825893402,
      "learning_rate": 3.579223504721931e-05,
      "loss": 0.0116,
      "step": 520
    },
    {
      "epoch": 0.13486005089058525,
      "grad_norm": 0.09249290823936462,
      "learning_rate": 3.5687303252885625e-05,
      "loss": 0.0052,
      "step": 530
    },
    {
      "epoch": 0.13740458015267176,
      "grad_norm": 0.4119039475917816,
      "learning_rate": 3.5582371458551946e-05,
      "loss": 0.0092,
      "step": 540
    },
    {
      "epoch": 0.13994910941475827,
      "grad_norm": 0.48269760608673096,
      "learning_rate": 3.547743966421826e-05,
      "loss": 0.0087,
      "step": 550
    },
    {
      "epoch": 0.14249363867684478,
      "grad_norm": 0.7463809251785278,
      "learning_rate": 3.5372507869884575e-05,
      "loss": 0.0122,
      "step": 560
    },
    {
      "epoch": 0.1450381679389313,
      "grad_norm": 0.09356890618801117,
      "learning_rate": 3.5267576075550896e-05,
      "loss": 0.0115,
      "step": 570
    },
    {
      "epoch": 0.1475826972010178,
      "grad_norm": 0.6823534965515137,
      "learning_rate": 3.516264428121721e-05,
      "loss": 0.0038,
      "step": 580
    },
    {
      "epoch": 0.15012722646310434,
      "grad_norm": 0.012240025214850903,
      "learning_rate": 3.505771248688353e-05,
      "loss": 0.0104,
      "step": 590
    },
    {
      "epoch": 0.15267175572519084,
      "grad_norm": 0.22518590092658997,
      "learning_rate": 3.4952780692549846e-05,
      "loss": 0.0027,
      "step": 600
    },
    {
      "epoch": 0.15521628498727735,
      "grad_norm": 0.32648953795433044,
      "learning_rate": 3.484784889821616e-05,
      "loss": 0.0112,
      "step": 610
    },
    {
      "epoch": 0.15776081424936386,
      "grad_norm": 0.07640829682350159,
      "learning_rate": 3.474291710388248e-05,
      "loss": 0.0122,
      "step": 620
    },
    {
      "epoch": 0.16030534351145037,
      "grad_norm": 0.027291180565953255,
      "learning_rate": 3.4637985309548795e-05,
      "loss": 0.0145,
      "step": 630
    },
    {
      "epoch": 0.1628498727735369,
      "grad_norm": 0.9284016489982605,
      "learning_rate": 3.453305351521511e-05,
      "loss": 0.0086,
      "step": 640
    },
    {
      "epoch": 0.16539440203562342,
      "grad_norm": 0.06572818756103516,
      "learning_rate": 3.442812172088143e-05,
      "loss": 0.0046,
      "step": 650
    },
    {
      "epoch": 0.16793893129770993,
      "grad_norm": 0.5768460631370544,
      "learning_rate": 3.4323189926547745e-05,
      "loss": 0.0099,
      "step": 660
    },
    {
      "epoch": 0.17048346055979643,
      "grad_norm": 0.13030396401882172,
      "learning_rate": 3.421825813221406e-05,
      "loss": 0.006,
      "step": 670
    },
    {
      "epoch": 0.17302798982188294,
      "grad_norm": 0.061677299439907074,
      "learning_rate": 3.411332633788038e-05,
      "loss": 0.0172,
      "step": 680
    },
    {
      "epoch": 0.17557251908396945,
      "grad_norm": 0.4030856788158417,
      "learning_rate": 3.4008394543546695e-05,
      "loss": 0.0205,
      "step": 690
    },
    {
      "epoch": 0.178117048346056,
      "grad_norm": 0.7717676758766174,
      "learning_rate": 3.3903462749213016e-05,
      "loss": 0.0112,
      "step": 700
    },
    {
      "epoch": 0.1806615776081425,
      "grad_norm": 0.867887556552887,
      "learning_rate": 3.379853095487933e-05,
      "loss": 0.0065,
      "step": 710
    },
    {
      "epoch": 0.183206106870229,
      "grad_norm": 0.12486184388399124,
      "learning_rate": 3.3693599160545645e-05,
      "loss": 0.0105,
      "step": 720
    },
    {
      "epoch": 0.18575063613231552,
      "grad_norm": 0.6370356678962708,
      "learning_rate": 3.3588667366211966e-05,
      "loss": 0.0058,
      "step": 730
    },
    {
      "epoch": 0.18829516539440203,
      "grad_norm": 0.30326998233795166,
      "learning_rate": 3.348373557187828e-05,
      "loss": 0.0114,
      "step": 740
    },
    {
      "epoch": 0.19083969465648856,
      "grad_norm": 0.03924720734357834,
      "learning_rate": 3.3378803777544595e-05,
      "loss": 0.0103,
      "step": 750
    },
    {
      "epoch": 0.19338422391857507,
      "grad_norm": 0.5556716918945312,
      "learning_rate": 3.3273871983210916e-05,
      "loss": 0.0059,
      "step": 760
    },
    {
      "epoch": 0.19592875318066158,
      "grad_norm": 0.12600132822990417,
      "learning_rate": 3.316894018887723e-05,
      "loss": 0.0032,
      "step": 770
    },
    {
      "epoch": 0.1984732824427481,
      "grad_norm": 0.21691930294036865,
      "learning_rate": 3.3064008394543545e-05,
      "loss": 0.0099,
      "step": 780
    },
    {
      "epoch": 0.2010178117048346,
      "grad_norm": 0.8290544748306274,
      "learning_rate": 3.2959076600209866e-05,
      "loss": 0.006,
      "step": 790
    },
    {
      "epoch": 0.2035623409669211,
      "grad_norm": 1.1044622659683228,
      "learning_rate": 3.285414480587618e-05,
      "loss": 0.0169,
      "step": 800
    },
    {
      "epoch": 0.20610687022900764,
      "grad_norm": 0.24777106940746307,
      "learning_rate": 3.27492130115425e-05,
      "loss": 0.0144,
      "step": 810
    },
    {
      "epoch": 0.20865139949109415,
      "grad_norm": 0.06799839437007904,
      "learning_rate": 3.2644281217208816e-05,
      "loss": 0.0038,
      "step": 820
    },
    {
      "epoch": 0.21119592875318066,
      "grad_norm": 0.5728234052658081,
      "learning_rate": 3.253934942287513e-05,
      "loss": 0.0085,
      "step": 830
    },
    {
      "epoch": 0.21374045801526717,
      "grad_norm": 1.0202137231826782,
      "learning_rate": 3.243441762854145e-05,
      "loss": 0.0163,
      "step": 840
    },
    {
      "epoch": 0.21628498727735368,
      "grad_norm": 0.00811342429369688,
      "learning_rate": 3.2329485834207766e-05,
      "loss": 0.0057,
      "step": 850
    },
    {
      "epoch": 0.21882951653944022,
      "grad_norm": 0.22676686942577362,
      "learning_rate": 3.222455403987408e-05,
      "loss": 0.0153,
      "step": 860
    },
    {
      "epoch": 0.22137404580152673,
      "grad_norm": 0.02477932535111904,
      "learning_rate": 3.21196222455404e-05,
      "loss": 0.0035,
      "step": 870
    },
    {
      "epoch": 0.22391857506361323,
      "grad_norm": 0.08315183967351913,
      "learning_rate": 3.2014690451206715e-05,
      "loss": 0.0068,
      "step": 880
    },
    {
      "epoch": 0.22646310432569974,
      "grad_norm": 1.0003092288970947,
      "learning_rate": 3.190975865687304e-05,
      "loss": 0.0119,
      "step": 890
    },
    {
      "epoch": 0.22900763358778625,
      "grad_norm": 0.034200772643089294,
      "learning_rate": 3.180482686253935e-05,
      "loss": 0.008,
      "step": 900
    },
    {
      "epoch": 0.23155216284987276,
      "grad_norm": 0.0161778274923563,
      "learning_rate": 3.1699895068205665e-05,
      "loss": 0.0105,
      "step": 910
    },
    {
      "epoch": 0.2340966921119593,
      "grad_norm": 0.07032974809408188,
      "learning_rate": 3.1594963273871987e-05,
      "loss": 0.0096,
      "step": 920
    },
    {
      "epoch": 0.2366412213740458,
      "grad_norm": 0.10155557096004486,
      "learning_rate": 3.14900314795383e-05,
      "loss": 0.0066,
      "step": 930
    },
    {
      "epoch": 0.23918575063613232,
      "grad_norm": 0.004129286855459213,
      "learning_rate": 3.1385099685204615e-05,
      "loss": 0.0016,
      "step": 940
    },
    {
      "epoch": 0.24173027989821882,
      "grad_norm": 0.00802706554532051,
      "learning_rate": 3.1280167890870936e-05,
      "loss": 0.0118,
      "step": 950
    },
    {
      "epoch": 0.24427480916030533,
      "grad_norm": 0.09856593608856201,
      "learning_rate": 3.117523609653725e-05,
      "loss": 0.0023,
      "step": 960
    },
    {
      "epoch": 0.24681933842239187,
      "grad_norm": 0.11473847180604935,
      "learning_rate": 3.107030430220357e-05,
      "loss": 0.0056,
      "step": 970
    },
    {
      "epoch": 0.24936386768447838,
      "grad_norm": 0.02289162203669548,
      "learning_rate": 3.0965372507869886e-05,
      "loss": 0.0131,
      "step": 980
    },
    {
      "epoch": 0.25190839694656486,
      "grad_norm": 0.08203335851430893,
      "learning_rate": 3.08604407135362e-05,
      "loss": 0.0049,
      "step": 990
    },
    {
      "epoch": 0.2544529262086514,
      "grad_norm": 0.8752655982971191,
      "learning_rate": 3.075550891920252e-05,
      "loss": 0.0102,
      "step": 1000
    },
    {
      "epoch": 0.2544529262086514,
      "eval_loss": 0.03172941133379936,
      "eval_runtime": 133.9654,
      "eval_samples_per_second": 59.583,
      "eval_steps_per_second": 0.47,
      "step": 1000
    },
    {
      "epoch": 0.25699745547073793,
      "grad_norm": 0.04296811670064926,
      "learning_rate": 3.0650577124868836e-05,
      "loss": 0.0045,
      "step": 1010
    },
    {
      "epoch": 0.2595419847328244,
      "grad_norm": 0.14025937020778656,
      "learning_rate": 3.054564533053516e-05,
      "loss": 0.0069,
      "step": 1020
    },
    {
      "epoch": 0.26208651399491095,
      "grad_norm": 0.5874730348587036,
      "learning_rate": 3.0440713536201468e-05,
      "loss": 0.0057,
      "step": 1030
    },
    {
      "epoch": 0.26463104325699743,
      "grad_norm": 0.45463526248931885,
      "learning_rate": 3.033578174186779e-05,
      "loss": 0.0119,
      "step": 1040
    },
    {
      "epoch": 0.26717557251908397,
      "grad_norm": 0.4644736051559448,
      "learning_rate": 3.0230849947534104e-05,
      "loss": 0.004,
      "step": 1050
    },
    {
      "epoch": 0.2697201017811705,
      "grad_norm": 0.23533551394939423,
      "learning_rate": 3.0125918153200425e-05,
      "loss": 0.002,
      "step": 1060
    },
    {
      "epoch": 0.272264631043257,
      "grad_norm": 0.744405210018158,
      "learning_rate": 3.002098635886674e-05,
      "loss": 0.0088,
      "step": 1070
    },
    {
      "epoch": 0.2748091603053435,
      "grad_norm": 0.4194672703742981,
      "learning_rate": 2.9916054564533054e-05,
      "loss": 0.0108,
      "step": 1080
    },
    {
      "epoch": 0.27735368956743,
      "grad_norm": 0.5709512829780579,
      "learning_rate": 2.9811122770199375e-05,
      "loss": 0.0068,
      "step": 1090
    },
    {
      "epoch": 0.27989821882951654,
      "grad_norm": 0.029444962739944458,
      "learning_rate": 2.970619097586569e-05,
      "loss": 0.0103,
      "step": 1100
    },
    {
      "epoch": 0.2824427480916031,
      "grad_norm": 0.024460772052407265,
      "learning_rate": 2.9601259181532003e-05,
      "loss": 0.0069,
      "step": 1110
    },
    {
      "epoch": 0.28498727735368956,
      "grad_norm": 0.9423603415489197,
      "learning_rate": 2.9496327387198325e-05,
      "loss": 0.018,
      "step": 1120
    },
    {
      "epoch": 0.2875318066157761,
      "grad_norm": 0.1192903071641922,
      "learning_rate": 2.939139559286464e-05,
      "loss": 0.007,
      "step": 1130
    },
    {
      "epoch": 0.2900763358778626,
      "grad_norm": 0.12459497898817062,
      "learning_rate": 2.9286463798530957e-05,
      "loss": 0.004,
      "step": 1140
    },
    {
      "epoch": 0.2926208651399491,
      "grad_norm": 0.02133631706237793,
      "learning_rate": 2.9181532004197274e-05,
      "loss": 0.0097,
      "step": 1150
    },
    {
      "epoch": 0.2951653944020356,
      "grad_norm": 0.20045505464076996,
      "learning_rate": 2.907660020986359e-05,
      "loss": 0.0096,
      "step": 1160
    },
    {
      "epoch": 0.29770992366412213,
      "grad_norm": 0.8285555243492126,
      "learning_rate": 2.897166841552991e-05,
      "loss": 0.0153,
      "step": 1170
    },
    {
      "epoch": 0.30025445292620867,
      "grad_norm": 0.11353142559528351,
      "learning_rate": 2.8866736621196224e-05,
      "loss": 0.004,
      "step": 1180
    },
    {
      "epoch": 0.30279898218829515,
      "grad_norm": 0.316145658493042,
      "learning_rate": 2.8761804826862542e-05,
      "loss": 0.0118,
      "step": 1190
    },
    {
      "epoch": 0.3053435114503817,
      "grad_norm": 0.20705455541610718,
      "learning_rate": 2.865687303252886e-05,
      "loss": 0.0048,
      "step": 1200
    },
    {
      "epoch": 0.30788804071246817,
      "grad_norm": 0.004804595839232206,
      "learning_rate": 2.8551941238195174e-05,
      "loss": 0.0077,
      "step": 1210
    },
    {
      "epoch": 0.3104325699745547,
      "grad_norm": 0.003488596761599183,
      "learning_rate": 2.8447009443861492e-05,
      "loss": 0.0115,
      "step": 1220
    },
    {
      "epoch": 0.31297709923664124,
      "grad_norm": 0.10154404491186142,
      "learning_rate": 2.834207764952781e-05,
      "loss": 0.0146,
      "step": 1230
    },
    {
      "epoch": 0.3155216284987277,
      "grad_norm": 0.008709642104804516,
      "learning_rate": 2.8237145855194124e-05,
      "loss": 0.0061,
      "step": 1240
    },
    {
      "epoch": 0.31806615776081426,
      "grad_norm": 0.018860768526792526,
      "learning_rate": 2.8132214060860442e-05,
      "loss": 0.0091,
      "step": 1250
    },
    {
      "epoch": 0.32061068702290074,
      "grad_norm": 0.10052315890789032,
      "learning_rate": 2.802728226652676e-05,
      "loss": 0.0189,
      "step": 1260
    },
    {
      "epoch": 0.3231552162849873,
      "grad_norm": 0.012599390931427479,
      "learning_rate": 2.7922350472193077e-05,
      "loss": 0.0037,
      "step": 1270
    },
    {
      "epoch": 0.3256997455470738,
      "grad_norm": 0.08410943299531937,
      "learning_rate": 2.7817418677859395e-05,
      "loss": 0.0047,
      "step": 1280
    },
    {
      "epoch": 0.3282442748091603,
      "grad_norm": 0.11278515309095383,
      "learning_rate": 2.771248688352571e-05,
      "loss": 0.0033,
      "step": 1290
    },
    {
      "epoch": 0.33078880407124683,
      "grad_norm": 0.508403480052948,
      "learning_rate": 2.7607555089192027e-05,
      "loss": 0.008,
      "step": 1300
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.2811458706855774,
      "learning_rate": 2.7502623294858345e-05,
      "loss": 0.0116,
      "step": 1310
    },
    {
      "epoch": 0.33587786259541985,
      "grad_norm": 0.012120573781430721,
      "learning_rate": 2.7397691500524663e-05,
      "loss": 0.0094,
      "step": 1320
    },
    {
      "epoch": 0.3384223918575064,
      "grad_norm": 0.24956151843070984,
      "learning_rate": 2.7292759706190977e-05,
      "loss": 0.0031,
      "step": 1330
    },
    {
      "epoch": 0.34096692111959287,
      "grad_norm": 0.7218794226646423,
      "learning_rate": 2.7187827911857295e-05,
      "loss": 0.0105,
      "step": 1340
    },
    {
      "epoch": 0.3435114503816794,
      "grad_norm": 0.24596373736858368,
      "learning_rate": 2.7082896117523613e-05,
      "loss": 0.0123,
      "step": 1350
    },
    {
      "epoch": 0.3460559796437659,
      "grad_norm": 1.128461480140686,
      "learning_rate": 2.6977964323189927e-05,
      "loss": 0.0129,
      "step": 1360
    },
    {
      "epoch": 0.3486005089058524,
      "grad_norm": 0.029505103826522827,
      "learning_rate": 2.6873032528856248e-05,
      "loss": 0.0066,
      "step": 1370
    },
    {
      "epoch": 0.3511450381679389,
      "grad_norm": 0.2863447964191437,
      "learning_rate": 2.6768100734522562e-05,
      "loss": 0.0084,
      "step": 1380
    },
    {
      "epoch": 0.35368956743002544,
      "grad_norm": 0.6585884690284729,
      "learning_rate": 2.666316894018888e-05,
      "loss": 0.0121,
      "step": 1390
    },
    {
      "epoch": 0.356234096692112,
      "grad_norm": 0.12626522779464722,
      "learning_rate": 2.6558237145855198e-05,
      "loss": 0.0061,
      "step": 1400
    },
    {
      "epoch": 0.35877862595419846,
      "grad_norm": 0.042731188237667084,
      "learning_rate": 2.6453305351521512e-05,
      "loss": 0.0079,
      "step": 1410
    },
    {
      "epoch": 0.361323155216285,
      "grad_norm": 0.49549469351768494,
      "learning_rate": 2.6348373557187833e-05,
      "loss": 0.0107,
      "step": 1420
    },
    {
      "epoch": 0.3638676844783715,
      "grad_norm": 0.05457068234682083,
      "learning_rate": 2.6243441762854148e-05,
      "loss": 0.0051,
      "step": 1430
    },
    {
      "epoch": 0.366412213740458,
      "grad_norm": 0.03797609731554985,
      "learning_rate": 2.6138509968520462e-05,
      "loss": 0.0146,
      "step": 1440
    },
    {
      "epoch": 0.36895674300254455,
      "grad_norm": 0.025207195430994034,
      "learning_rate": 2.6033578174186783e-05,
      "loss": 0.0081,
      "step": 1450
    },
    {
      "epoch": 0.37150127226463103,
      "grad_norm": 0.03196549415588379,
      "learning_rate": 2.5928646379853098e-05,
      "loss": 0.01,
      "step": 1460
    },
    {
      "epoch": 0.37404580152671757,
      "grad_norm": 0.322093665599823,
      "learning_rate": 2.5823714585519412e-05,
      "loss": 0.0031,
      "step": 1470
    },
    {
      "epoch": 0.37659033078880405,
      "grad_norm": 0.29793310165405273,
      "learning_rate": 2.5718782791185733e-05,
      "loss": 0.0138,
      "step": 1480
    },
    {
      "epoch": 0.3791348600508906,
      "grad_norm": 0.1609342247247696,
      "learning_rate": 2.5613850996852048e-05,
      "loss": 0.0057,
      "step": 1490
    },
    {
      "epoch": 0.3816793893129771,
      "grad_norm": 0.014492412097752094,
      "learning_rate": 2.5508919202518362e-05,
      "loss": 0.0124,
      "step": 1500
    },
    {
      "epoch": 0.3816793893129771,
      "eval_loss": 0.03146525099873543,
      "eval_runtime": 133.9818,
      "eval_samples_per_second": 59.575,
      "eval_steps_per_second": 0.47,
      "step": 1500
    },
    {
      "epoch": 0.3842239185750636,
      "grad_norm": 0.6495288610458374,
      "learning_rate": 2.5403987408184683e-05,
      "loss": 0.0137,
      "step": 1510
    },
    {
      "epoch": 0.38676844783715014,
      "grad_norm": 0.1830611675977707,
      "learning_rate": 2.5299055613850997e-05,
      "loss": 0.0029,
      "step": 1520
    },
    {
      "epoch": 0.3893129770992366,
      "grad_norm": 0.1593904048204422,
      "learning_rate": 2.519412381951732e-05,
      "loss": 0.006,
      "step": 1530
    },
    {
      "epoch": 0.39185750636132316,
      "grad_norm": 0.012680516578257084,
      "learning_rate": 2.5089192025183633e-05,
      "loss": 0.0032,
      "step": 1540
    },
    {
      "epoch": 0.3944020356234097,
      "grad_norm": 0.6925283074378967,
      "learning_rate": 2.4984260230849947e-05,
      "loss": 0.0074,
      "step": 1550
    },
    {
      "epoch": 0.3969465648854962,
      "grad_norm": 0.3493160605430603,
      "learning_rate": 2.487932843651627e-05,
      "loss": 0.012,
      "step": 1560
    },
    {
      "epoch": 0.3994910941475827,
      "grad_norm": 0.14689674973487854,
      "learning_rate": 2.4774396642182583e-05,
      "loss": 0.0038,
      "step": 1570
    },
    {
      "epoch": 0.4020356234096692,
      "grad_norm": 0.015744954347610474,
      "learning_rate": 2.4669464847848897e-05,
      "loss": 0.003,
      "step": 1580
    },
    {
      "epoch": 0.40458015267175573,
      "grad_norm": 0.8252820372581482,
      "learning_rate": 2.456453305351522e-05,
      "loss": 0.0084,
      "step": 1590
    },
    {
      "epoch": 0.4071246819338422,
      "grad_norm": 0.07908014953136444,
      "learning_rate": 2.4459601259181533e-05,
      "loss": 0.0031,
      "step": 1600
    },
    {
      "epoch": 0.40966921119592875,
      "grad_norm": 0.019779745489358902,
      "learning_rate": 2.435466946484785e-05,
      "loss": 0.0025,
      "step": 1610
    },
    {
      "epoch": 0.4122137404580153,
      "grad_norm": 0.19202356040477753,
      "learning_rate": 2.4249737670514168e-05,
      "loss": 0.0018,
      "step": 1620
    },
    {
      "epoch": 0.41475826972010177,
      "grad_norm": 0.9459726214408875,
      "learning_rate": 2.4144805876180483e-05,
      "loss": 0.005,
      "step": 1630
    },
    {
      "epoch": 0.4173027989821883,
      "grad_norm": 0.7893335819244385,
      "learning_rate": 2.4039874081846804e-05,
      "loss": 0.0081,
      "step": 1640
    },
    {
      "epoch": 0.4198473282442748,
      "grad_norm": 0.0010036466410383582,
      "learning_rate": 2.3934942287513118e-05,
      "loss": 0.0021,
      "step": 1650
    },
    {
      "epoch": 0.4223918575063613,
      "grad_norm": 0.015236877836287022,
      "learning_rate": 2.3830010493179432e-05,
      "loss": 0.008,
      "step": 1660
    },
    {
      "epoch": 0.42493638676844786,
      "grad_norm": 0.00819339882582426,
      "learning_rate": 2.3725078698845754e-05,
      "loss": 0.005,
      "step": 1670
    },
    {
      "epoch": 0.42748091603053434,
      "grad_norm": 0.05410633608698845,
      "learning_rate": 2.3620146904512068e-05,
      "loss": 0.0053,
      "step": 1680
    },
    {
      "epoch": 0.4300254452926209,
      "grad_norm": 0.10227911174297333,
      "learning_rate": 2.3515215110178386e-05,
      "loss": 0.0101,
      "step": 1690
    },
    {
      "epoch": 0.43256997455470736,
      "grad_norm": 0.08396944403648376,
      "learning_rate": 2.3410283315844703e-05,
      "loss": 0.0074,
      "step": 1700
    },
    {
      "epoch": 0.4351145038167939,
      "grad_norm": 0.8800942897796631,
      "learning_rate": 2.3305351521511018e-05,
      "loss": 0.0108,
      "step": 1710
    },
    {
      "epoch": 0.43765903307888043,
      "grad_norm": 0.0064217569306492805,
      "learning_rate": 2.3200419727177336e-05,
      "loss": 0.0053,
      "step": 1720
    },
    {
      "epoch": 0.4402035623409669,
      "grad_norm": 0.9264568090438843,
      "learning_rate": 2.3095487932843653e-05,
      "loss": 0.0084,
      "step": 1730
    },
    {
      "epoch": 0.44274809160305345,
      "grad_norm": 0.03843522444367409,
      "learning_rate": 2.299055613850997e-05,
      "loss": 0.0025,
      "step": 1740
    },
    {
      "epoch": 0.44529262086513993,
      "grad_norm": 0.7688780426979065,
      "learning_rate": 2.288562434417629e-05,
      "loss": 0.0094,
      "step": 1750
    },
    {
      "epoch": 0.44783715012722647,
      "grad_norm": 0.007313898764550686,
      "learning_rate": 2.2780692549842603e-05,
      "loss": 0.0074,
      "step": 1760
    },
    {
      "epoch": 0.45038167938931295,
      "grad_norm": 0.10244712233543396,
      "learning_rate": 2.267576075550892e-05,
      "loss": 0.0075,
      "step": 1770
    },
    {
      "epoch": 0.4529262086513995,
      "grad_norm": 0.6030476093292236,
      "learning_rate": 2.257082896117524e-05,
      "loss": 0.0172,
      "step": 1780
    },
    {
      "epoch": 0.455470737913486,
      "grad_norm": 0.08637584000825882,
      "learning_rate": 2.2465897166841556e-05,
      "loss": 0.0068,
      "step": 1790
    },
    {
      "epoch": 0.4580152671755725,
      "grad_norm": 0.5561503767967224,
      "learning_rate": 2.236096537250787e-05,
      "loss": 0.0104,
      "step": 1800
    },
    {
      "epoch": 0.46055979643765904,
      "grad_norm": 0.08013448119163513,
      "learning_rate": 2.225603357817419e-05,
      "loss": 0.0078,
      "step": 1810
    },
    {
      "epoch": 0.4631043256997455,
      "grad_norm": 0.08120064437389374,
      "learning_rate": 2.2151101783840506e-05,
      "loss": 0.0047,
      "step": 1820
    },
    {
      "epoch": 0.46564885496183206,
      "grad_norm": 0.021734222769737244,
      "learning_rate": 2.204616998950682e-05,
      "loss": 0.0067,
      "step": 1830
    },
    {
      "epoch": 0.4681933842239186,
      "grad_norm": 0.014017133973538876,
      "learning_rate": 2.194123819517314e-05,
      "loss": 0.0047,
      "step": 1840
    },
    {
      "epoch": 0.4707379134860051,
      "grad_norm": 0.10124704986810684,
      "learning_rate": 2.1836306400839456e-05,
      "loss": 0.0041,
      "step": 1850
    },
    {
      "epoch": 0.4732824427480916,
      "grad_norm": 0.2559342086315155,
      "learning_rate": 2.1731374606505774e-05,
      "loss": 0.0171,
      "step": 1860
    },
    {
      "epoch": 0.4758269720101781,
      "grad_norm": 0.003074784530326724,
      "learning_rate": 2.1626442812172092e-05,
      "loss": 0.0087,
      "step": 1870
    },
    {
      "epoch": 0.47837150127226463,
      "grad_norm": 0.2251855880022049,
      "learning_rate": 2.1521511017838406e-05,
      "loss": 0.0026,
      "step": 1880
    },
    {
      "epoch": 0.48091603053435117,
      "grad_norm": 0.7686161994934082,
      "learning_rate": 2.1416579223504724e-05,
      "loss": 0.011,
      "step": 1890
    },
    {
      "epoch": 0.48346055979643765,
      "grad_norm": 0.06315213441848755,
      "learning_rate": 2.131164742917104e-05,
      "loss": 0.0098,
      "step": 1900
    },
    {
      "epoch": 0.4860050890585242,
      "grad_norm": 0.1871805340051651,
      "learning_rate": 2.1206715634837356e-05,
      "loss": 0.0065,
      "step": 1910
    },
    {
      "epoch": 0.48854961832061067,
      "grad_norm": 0.1880640983581543,
      "learning_rate": 2.1101783840503677e-05,
      "loss": 0.0021,
      "step": 1920
    },
    {
      "epoch": 0.4910941475826972,
      "grad_norm": 1.1694693565368652,
      "learning_rate": 2.099685204616999e-05,
      "loss": 0.0083,
      "step": 1930
    },
    {
      "epoch": 0.49363867684478374,
      "grad_norm": 0.2776516079902649,
      "learning_rate": 2.0891920251836306e-05,
      "loss": 0.0053,
      "step": 1940
    },
    {
      "epoch": 0.4961832061068702,
      "grad_norm": 0.023003289476037025,
      "learning_rate": 2.0786988457502627e-05,
      "loss": 0.01,
      "step": 1950
    },
    {
      "epoch": 0.49872773536895676,
      "grad_norm": 0.07819438725709915,
      "learning_rate": 2.068205666316894e-05,
      "loss": 0.0038,
      "step": 1960
    },
    {
      "epoch": 0.5012722646310432,
      "grad_norm": 0.6131492853164673,
      "learning_rate": 2.0577124868835262e-05,
      "loss": 0.0053,
      "step": 1970
    },
    {
      "epoch": 0.5038167938931297,
      "grad_norm": 0.2298940122127533,
      "learning_rate": 2.0472193074501577e-05,
      "loss": 0.0075,
      "step": 1980
    },
    {
      "epoch": 0.5063613231552163,
      "grad_norm": 0.09217698127031326,
      "learning_rate": 2.036726128016789e-05,
      "loss": 0.0024,
      "step": 1990
    },
    {
      "epoch": 0.5089058524173028,
      "grad_norm": 0.11719799041748047,
      "learning_rate": 2.0262329485834212e-05,
      "loss": 0.0017,
      "step": 2000
    },
    {
      "epoch": 0.5089058524173028,
      "eval_loss": 0.028622597455978394,
      "eval_runtime": 133.9853,
      "eval_samples_per_second": 59.574,
      "eval_steps_per_second": 0.47,
      "step": 2000
    },
    {
      "epoch": 0.5114503816793893,
      "grad_norm": 0.0031085775699466467,
      "learning_rate": 2.0157397691500527e-05,
      "loss": 0.0109,
      "step": 2010
    },
    {
      "epoch": 0.5139949109414759,
      "grad_norm": 0.45600050687789917,
      "learning_rate": 2.005246589716684e-05,
      "loss": 0.0052,
      "step": 2020
    },
    {
      "epoch": 0.5165394402035624,
      "grad_norm": 0.15607476234436035,
      "learning_rate": 1.994753410283316e-05,
      "loss": 0.0072,
      "step": 2030
    },
    {
      "epoch": 0.5190839694656488,
      "grad_norm": 0.4832943081855774,
      "learning_rate": 1.9842602308499477e-05,
      "loss": 0.0088,
      "step": 2040
    },
    {
      "epoch": 0.5216284987277354,
      "grad_norm": 0.13958820700645447,
      "learning_rate": 1.9737670514165794e-05,
      "loss": 0.0075,
      "step": 2050
    },
    {
      "epoch": 0.5241730279898219,
      "grad_norm": 0.9070898294448853,
      "learning_rate": 1.9632738719832112e-05,
      "loss": 0.0062,
      "step": 2060
    },
    {
      "epoch": 0.5267175572519084,
      "grad_norm": 0.3955434262752533,
      "learning_rate": 1.9527806925498426e-05,
      "loss": 0.0112,
      "step": 2070
    },
    {
      "epoch": 0.5292620865139949,
      "grad_norm": 0.616914689540863,
      "learning_rate": 1.9422875131164744e-05,
      "loss": 0.0133,
      "step": 2080
    },
    {
      "epoch": 0.5318066157760815,
      "grad_norm": 0.006997487973421812,
      "learning_rate": 1.9317943336831062e-05,
      "loss": 0.0103,
      "step": 2090
    },
    {
      "epoch": 0.5343511450381679,
      "grad_norm": 0.7367590665817261,
      "learning_rate": 1.9213011542497376e-05,
      "loss": 0.0117,
      "step": 2100
    },
    {
      "epoch": 0.5368956743002544,
      "grad_norm": 0.10802368819713593,
      "learning_rate": 1.9108079748163694e-05,
      "loss": 0.0092,
      "step": 2110
    },
    {
      "epoch": 0.539440203562341,
      "grad_norm": 0.1059407964348793,
      "learning_rate": 1.9003147953830012e-05,
      "loss": 0.0065,
      "step": 2120
    },
    {
      "epoch": 0.5419847328244275,
      "grad_norm": 0.03868965432047844,
      "learning_rate": 1.889821615949633e-05,
      "loss": 0.0052,
      "step": 2130
    },
    {
      "epoch": 0.544529262086514,
      "grad_norm": 0.004114597570151091,
      "learning_rate": 1.8793284365162644e-05,
      "loss": 0.011,
      "step": 2140
    },
    {
      "epoch": 0.5470737913486005,
      "grad_norm": 0.15624436736106873,
      "learning_rate": 1.868835257082896e-05,
      "loss": 0.0098,
      "step": 2150
    },
    {
      "epoch": 0.549618320610687,
      "grad_norm": 0.46892330050468445,
      "learning_rate": 1.858342077649528e-05,
      "loss": 0.0059,
      "step": 2160
    },
    {
      "epoch": 0.5521628498727735,
      "grad_norm": 0.046456072479486465,
      "learning_rate": 1.8478488982161594e-05,
      "loss": 0.0041,
      "step": 2170
    },
    {
      "epoch": 0.55470737913486,
      "grad_norm": 0.6324033737182617,
      "learning_rate": 1.837355718782791e-05,
      "loss": 0.0079,
      "step": 2180
    },
    {
      "epoch": 0.5572519083969466,
      "grad_norm": 0.08168061077594757,
      "learning_rate": 1.826862539349423e-05,
      "loss": 0.0081,
      "step": 2190
    },
    {
      "epoch": 0.5597964376590331,
      "grad_norm": 0.0022136031184345484,
      "learning_rate": 1.8163693599160547e-05,
      "loss": 0.0087,
      "step": 2200
    },
    {
      "epoch": 0.5623409669211196,
      "grad_norm": 0.8067908883094788,
      "learning_rate": 1.8058761804826865e-05,
      "loss": 0.0054,
      "step": 2210
    },
    {
      "epoch": 0.5648854961832062,
      "grad_norm": 0.438232958316803,
      "learning_rate": 1.795383001049318e-05,
      "loss": 0.0091,
      "step": 2220
    },
    {
      "epoch": 0.5674300254452926,
      "grad_norm": 0.8731472492218018,
      "learning_rate": 1.7848898216159497e-05,
      "loss": 0.0153,
      "step": 2230
    },
    {
      "epoch": 0.5699745547073791,
      "grad_norm": 0.09311806410551071,
      "learning_rate": 1.7743966421825815e-05,
      "loss": 0.0025,
      "step": 2240
    },
    {
      "epoch": 0.5725190839694656,
      "grad_norm": 0.27608296275138855,
      "learning_rate": 1.7639034627492132e-05,
      "loss": 0.0097,
      "step": 2250
    },
    {
      "epoch": 0.5750636132315522,
      "grad_norm": 0.023715125396847725,
      "learning_rate": 1.7534102833158447e-05,
      "loss": 0.0068,
      "step": 2260
    },
    {
      "epoch": 0.5776081424936387,
      "grad_norm": 0.390055388212204,
      "learning_rate": 1.7429171038824765e-05,
      "loss": 0.0063,
      "step": 2270
    },
    {
      "epoch": 0.5801526717557252,
      "grad_norm": 0.5786376595497131,
      "learning_rate": 1.7324239244491082e-05,
      "loss": 0.0086,
      "step": 2280
    },
    {
      "epoch": 0.5826972010178118,
      "grad_norm": 0.7800132632255554,
      "learning_rate": 1.72193074501574e-05,
      "loss": 0.0076,
      "step": 2290
    },
    {
      "epoch": 0.5852417302798982,
      "grad_norm": 0.022780494764447212,
      "learning_rate": 1.7114375655823718e-05,
      "loss": 0.0007,
      "step": 2300
    },
    {
      "epoch": 0.5877862595419847,
      "grad_norm": 0.4210648238658905,
      "learning_rate": 1.7009443861490032e-05,
      "loss": 0.0026,
      "step": 2310
    },
    {
      "epoch": 0.5903307888040712,
      "grad_norm": 0.046605486422777176,
      "learning_rate": 1.690451206715635e-05,
      "loss": 0.0096,
      "step": 2320
    },
    {
      "epoch": 0.5928753180661578,
      "grad_norm": 0.09751195460557938,
      "learning_rate": 1.6799580272822668e-05,
      "loss": 0.0049,
      "step": 2330
    },
    {
      "epoch": 0.5954198473282443,
      "grad_norm": 0.08370577543973923,
      "learning_rate": 1.6694648478488985e-05,
      "loss": 0.0057,
      "step": 2340
    },
    {
      "epoch": 0.5979643765903307,
      "grad_norm": 0.003502418752759695,
      "learning_rate": 1.65897166841553e-05,
      "loss": 0.0084,
      "step": 2350
    },
    {
      "epoch": 0.6005089058524173,
      "grad_norm": 0.27206945419311523,
      "learning_rate": 1.6484784889821618e-05,
      "loss": 0.0022,
      "step": 2360
    },
    {
      "epoch": 0.6030534351145038,
      "grad_norm": 0.11041576415300369,
      "learning_rate": 1.6379853095487935e-05,
      "loss": 0.0045,
      "step": 2370
    },
    {
      "epoch": 0.6055979643765903,
      "grad_norm": 0.0404311828315258,
      "learning_rate": 1.6274921301154253e-05,
      "loss": 0.0101,
      "step": 2380
    },
    {
      "epoch": 0.6081424936386769,
      "grad_norm": 0.47096529603004456,
      "learning_rate": 1.6169989506820567e-05,
      "loss": 0.0073,
      "step": 2390
    },
    {
      "epoch": 0.6106870229007634,
      "grad_norm": 0.0981617122888565,
      "learning_rate": 1.6065057712486885e-05,
      "loss": 0.011,
      "step": 2400
    },
    {
      "epoch": 0.6132315521628499,
      "grad_norm": 0.029454654082655907,
      "learning_rate": 1.5960125918153203e-05,
      "loss": 0.0083,
      "step": 2410
    },
    {
      "epoch": 0.6157760814249363,
      "grad_norm": 0.6878058910369873,
      "learning_rate": 1.585519412381952e-05,
      "loss": 0.0134,
      "step": 2420
    },
    {
      "epoch": 0.6183206106870229,
      "grad_norm": 0.06291262060403824,
      "learning_rate": 1.5750262329485835e-05,
      "loss": 0.0043,
      "step": 2430
    },
    {
      "epoch": 0.6208651399491094,
      "grad_norm": 0.01707449182868004,
      "learning_rate": 1.5645330535152153e-05,
      "loss": 0.0087,
      "step": 2440
    },
    {
      "epoch": 0.6234096692111959,
      "grad_norm": 0.18086305260658264,
      "learning_rate": 1.554039874081847e-05,
      "loss": 0.004,
      "step": 2450
    },
    {
      "epoch": 0.6259541984732825,
      "grad_norm": 0.632637619972229,
      "learning_rate": 1.5435466946484785e-05,
      "loss": 0.0116,
      "step": 2460
    },
    {
      "epoch": 0.628498727735369,
      "grad_norm": 0.2877883017063141,
      "learning_rate": 1.5330535152151103e-05,
      "loss": 0.0042,
      "step": 2470
    },
    {
      "epoch": 0.6310432569974554,
      "grad_norm": 0.42460644245147705,
      "learning_rate": 1.522560335781742e-05,
      "loss": 0.0197,
      "step": 2480
    },
    {
      "epoch": 0.6335877862595419,
      "grad_norm": 0.013305261731147766,
      "learning_rate": 1.5120671563483738e-05,
      "loss": 0.0101,
      "step": 2490
    },
    {
      "epoch": 0.6361323155216285,
      "grad_norm": 0.5830298066139221,
      "learning_rate": 1.5015739769150053e-05,
      "loss": 0.0051,
      "step": 2500
    },
    {
      "epoch": 0.6361323155216285,
      "eval_loss": 0.026752423495054245,
      "eval_runtime": 133.9291,
      "eval_samples_per_second": 59.599,
      "eval_steps_per_second": 0.47,
      "step": 2500
    },
    {
      "epoch": 0.638676844783715,
      "grad_norm": 0.48519253730773926,
      "learning_rate": 1.491080797481637e-05,
      "loss": 0.0136,
      "step": 2510
    },
    {
      "epoch": 0.6412213740458015,
      "grad_norm": 0.4439609944820404,
      "learning_rate": 1.4805876180482688e-05,
      "loss": 0.0098,
      "step": 2520
    },
    {
      "epoch": 0.6437659033078881,
      "grad_norm": 0.27948108315467834,
      "learning_rate": 1.4700944386149006e-05,
      "loss": 0.0086,
      "step": 2530
    },
    {
      "epoch": 0.6463104325699746,
      "grad_norm": 0.01617351733148098,
      "learning_rate": 1.459601259181532e-05,
      "loss": 0.0079,
      "step": 2540
    },
    {
      "epoch": 0.648854961832061,
      "grad_norm": 0.8637906312942505,
      "learning_rate": 1.4491080797481638e-05,
      "loss": 0.0083,
      "step": 2550
    },
    {
      "epoch": 0.6513994910941476,
      "grad_norm": 0.47218918800354004,
      "learning_rate": 1.4386149003147956e-05,
      "loss": 0.0243,
      "step": 2560
    },
    {
      "epoch": 0.6539440203562341,
      "grad_norm": 0.5402618646621704,
      "learning_rate": 1.428121720881427e-05,
      "loss": 0.0082,
      "step": 2570
    },
    {
      "epoch": 0.6564885496183206,
      "grad_norm": 0.03326863422989845,
      "learning_rate": 1.4176285414480588e-05,
      "loss": 0.0112,
      "step": 2580
    },
    {
      "epoch": 0.6590330788804071,
      "grad_norm": 1.7870920896530151,
      "learning_rate": 1.4071353620146906e-05,
      "loss": 0.0132,
      "step": 2590
    },
    {
      "epoch": 0.6615776081424937,
      "grad_norm": 0.06419245153665543,
      "learning_rate": 1.3966421825813223e-05,
      "loss": 0.0037,
      "step": 2600
    },
    {
      "epoch": 0.6641221374045801,
      "grad_norm": 0.020061425864696503,
      "learning_rate": 1.386149003147954e-05,
      "loss": 0.0104,
      "step": 2610
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.018068986013531685,
      "learning_rate": 1.3756558237145855e-05,
      "loss": 0.0038,
      "step": 2620
    },
    {
      "epoch": 0.6692111959287532,
      "grad_norm": 0.038132619112730026,
      "learning_rate": 1.3651626442812173e-05,
      "loss": 0.0029,
      "step": 2630
    },
    {
      "epoch": 0.6717557251908397,
      "grad_norm": 0.6635996103286743,
      "learning_rate": 1.3546694648478491e-05,
      "loss": 0.0064,
      "step": 2640
    },
    {
      "epoch": 0.6743002544529262,
      "grad_norm": 0.01742667704820633,
      "learning_rate": 1.3441762854144807e-05,
      "loss": 0.0045,
      "step": 2650
    },
    {
      "epoch": 0.6768447837150128,
      "grad_norm": 0.5041590332984924,
      "learning_rate": 1.3336831059811125e-05,
      "loss": 0.0028,
      "step": 2660
    },
    {
      "epoch": 0.6793893129770993,
      "grad_norm": 0.2859777808189392,
      "learning_rate": 1.323189926547744e-05,
      "loss": 0.0085,
      "step": 2670
    },
    {
      "epoch": 0.6819338422391857,
      "grad_norm": 0.8118346929550171,
      "learning_rate": 1.3126967471143757e-05,
      "loss": 0.0261,
      "step": 2680
    },
    {
      "epoch": 0.6844783715012722,
      "grad_norm": 0.898858904838562,
      "learning_rate": 1.3022035676810075e-05,
      "loss": 0.0151,
      "step": 2690
    },
    {
      "epoch": 0.6870229007633588,
      "grad_norm": 1.0029804706573486,
      "learning_rate": 1.2917103882476392e-05,
      "loss": 0.0079,
      "step": 2700
    },
    {
      "epoch": 0.6895674300254453,
      "grad_norm": 1.1035183668136597,
      "learning_rate": 1.2812172088142708e-05,
      "loss": 0.0082,
      "step": 2710
    },
    {
      "epoch": 0.6921119592875318,
      "grad_norm": 0.05390460416674614,
      "learning_rate": 1.2707240293809024e-05,
      "loss": 0.0085,
      "step": 2720
    },
    {
      "epoch": 0.6946564885496184,
      "grad_norm": 0.12546318769454956,
      "learning_rate": 1.2602308499475342e-05,
      "loss": 0.0084,
      "step": 2730
    },
    {
      "epoch": 0.6972010178117048,
      "grad_norm": 0.0319611094892025,
      "learning_rate": 1.249737670514166e-05,
      "loss": 0.0112,
      "step": 2740
    },
    {
      "epoch": 0.6997455470737913,
      "grad_norm": 0.46342381834983826,
      "learning_rate": 1.2392444910807974e-05,
      "loss": 0.0042,
      "step": 2750
    },
    {
      "epoch": 0.7022900763358778,
      "grad_norm": 0.17658625543117523,
      "learning_rate": 1.2287513116474292e-05,
      "loss": 0.0049,
      "step": 2760
    },
    {
      "epoch": 0.7048346055979644,
      "grad_norm": 0.18953083455562592,
      "learning_rate": 1.218258132214061e-05,
      "loss": 0.0061,
      "step": 2770
    },
    {
      "epoch": 0.7073791348600509,
      "grad_norm": 0.010515202768146992,
      "learning_rate": 1.2077649527806928e-05,
      "loss": 0.0008,
      "step": 2780
    },
    {
      "epoch": 0.7099236641221374,
      "grad_norm": 0.004728379193693399,
      "learning_rate": 1.1972717733473242e-05,
      "loss": 0.0026,
      "step": 2790
    },
    {
      "epoch": 0.712468193384224,
      "grad_norm": 0.29097694158554077,
      "learning_rate": 1.186778593913956e-05,
      "loss": 0.0054,
      "step": 2800
    },
    {
      "epoch": 0.7150127226463104,
      "grad_norm": 0.010688016191124916,
      "learning_rate": 1.1762854144805877e-05,
      "loss": 0.0048,
      "step": 2810
    },
    {
      "epoch": 0.7175572519083969,
      "grad_norm": 0.5915406346321106,
      "learning_rate": 1.1657922350472195e-05,
      "loss": 0.0046,
      "step": 2820
    },
    {
      "epoch": 0.7201017811704835,
      "grad_norm": 0.009668947197496891,
      "learning_rate": 1.155299055613851e-05,
      "loss": 0.0018,
      "step": 2830
    },
    {
      "epoch": 0.72264631043257,
      "grad_norm": 0.010339943692088127,
      "learning_rate": 1.1448058761804827e-05,
      "loss": 0.0137,
      "step": 2840
    },
    {
      "epoch": 0.7251908396946565,
      "grad_norm": 0.00322342268191278,
      "learning_rate": 1.1343126967471145e-05,
      "loss": 0.0059,
      "step": 2850
    },
    {
      "epoch": 0.727735368956743,
      "grad_norm": 0.5668567419052124,
      "learning_rate": 1.1238195173137461e-05,
      "loss": 0.0128,
      "step": 2860
    },
    {
      "epoch": 0.7302798982188295,
      "grad_norm": 0.30744919180870056,
      "learning_rate": 1.1133263378803777e-05,
      "loss": 0.0009,
      "step": 2870
    },
    {
      "epoch": 0.732824427480916,
      "grad_norm": 1.329842448234558,
      "learning_rate": 1.1028331584470095e-05,
      "loss": 0.0056,
      "step": 2880
    },
    {
      "epoch": 0.7353689567430025,
      "grad_norm": 0.3226444125175476,
      "learning_rate": 1.0923399790136413e-05,
      "loss": 0.0084,
      "step": 2890
    },
    {
      "epoch": 0.7379134860050891,
      "grad_norm": 0.26656076312065125,
      "learning_rate": 1.0818467995802729e-05,
      "loss": 0.01,
      "step": 2900
    },
    {
      "epoch": 0.7404580152671756,
      "grad_norm": 0.029421096667647362,
      "learning_rate": 1.0713536201469047e-05,
      "loss": 0.0122,
      "step": 2910
    },
    {
      "epoch": 0.7430025445292621,
      "grad_norm": 0.036649610847234726,
      "learning_rate": 1.0608604407135363e-05,
      "loss": 0.0041,
      "step": 2920
    },
    {
      "epoch": 0.7455470737913485,
      "grad_norm": 0.14049018919467926,
      "learning_rate": 1.050367261280168e-05,
      "loss": 0.0021,
      "step": 2930
    },
    {
      "epoch": 0.7480916030534351,
      "grad_norm": 0.8101888298988342,
      "learning_rate": 1.0398740818467996e-05,
      "loss": 0.0061,
      "step": 2940
    },
    {
      "epoch": 0.7506361323155216,
      "grad_norm": 0.3166317045688629,
      "learning_rate": 1.0293809024134314e-05,
      "loss": 0.0103,
      "step": 2950
    },
    {
      "epoch": 0.7531806615776081,
      "grad_norm": 0.4369036853313446,
      "learning_rate": 1.0188877229800632e-05,
      "loss": 0.006,
      "step": 2960
    },
    {
      "epoch": 0.7557251908396947,
      "grad_norm": 0.6003983616828918,
      "learning_rate": 1.0083945435466946e-05,
      "loss": 0.0063,
      "step": 2970
    },
    {
      "epoch": 0.7582697201017812,
      "grad_norm": 0.3092733323574066,
      "learning_rate": 9.979013641133264e-06,
      "loss": 0.0061,
      "step": 2980
    },
    {
      "epoch": 0.7608142493638677,
      "grad_norm": 0.17131660878658295,
      "learning_rate": 9.874081846799582e-06,
      "loss": 0.0026,
      "step": 2990
    },
    {
      "epoch": 0.7633587786259542,
      "grad_norm": 0.6556366682052612,
      "learning_rate": 9.769150052465898e-06,
      "loss": 0.0071,
      "step": 3000
    },
    {
      "epoch": 0.7633587786259542,
      "eval_loss": 0.02788231521844864,
      "eval_runtime": 134.0287,
      "eval_samples_per_second": 59.554,
      "eval_steps_per_second": 0.47,
      "step": 3000
    },
    {
      "epoch": 0.7659033078880407,
      "grad_norm": 0.02394355833530426,
      "learning_rate": 9.664218258132216e-06,
      "loss": 0.0062,
      "step": 3010
    },
    {
      "epoch": 0.7684478371501272,
      "grad_norm": 0.047644879668951035,
      "learning_rate": 9.559286463798532e-06,
      "loss": 0.0061,
      "step": 3020
    },
    {
      "epoch": 0.7709923664122137,
      "grad_norm": 1.3568006753921509,
      "learning_rate": 9.45435466946485e-06,
      "loss": 0.0058,
      "step": 3030
    },
    {
      "epoch": 0.7735368956743003,
      "grad_norm": 0.316701740026474,
      "learning_rate": 9.349422875131165e-06,
      "loss": 0.0034,
      "step": 3040
    },
    {
      "epoch": 0.7760814249363868,
      "grad_norm": 0.02915433794260025,
      "learning_rate": 9.244491080797482e-06,
      "loss": 0.0024,
      "step": 3050
    },
    {
      "epoch": 0.7786259541984732,
      "grad_norm": 0.9025008678436279,
      "learning_rate": 9.1395592864638e-06,
      "loss": 0.01,
      "step": 3060
    },
    {
      "epoch": 0.7811704834605598,
      "grad_norm": 0.05889623612165451,
      "learning_rate": 9.034627492130115e-06,
      "loss": 0.0175,
      "step": 3070
    },
    {
      "epoch": 0.7837150127226463,
      "grad_norm": 0.2562934160232544,
      "learning_rate": 8.929695697796433e-06,
      "loss": 0.0037,
      "step": 3080
    },
    {
      "epoch": 0.7862595419847328,
      "grad_norm": 0.019730977714061737,
      "learning_rate": 8.824763903462749e-06,
      "loss": 0.0072,
      "step": 3090
    },
    {
      "epoch": 0.7888040712468194,
      "grad_norm": 0.24575015902519226,
      "learning_rate": 8.719832109129067e-06,
      "loss": 0.0066,
      "step": 3100
    },
    {
      "epoch": 0.7913486005089059,
      "grad_norm": 0.13040441274642944,
      "learning_rate": 8.614900314795383e-06,
      "loss": 0.0091,
      "step": 3110
    },
    {
      "epoch": 0.7938931297709924,
      "grad_norm": 0.10326778143644333,
      "learning_rate": 8.5099685204617e-06,
      "loss": 0.0053,
      "step": 3120
    },
    {
      "epoch": 0.7964376590330788,
      "grad_norm": 0.011873223818838596,
      "learning_rate": 8.405036726128017e-06,
      "loss": 0.0114,
      "step": 3130
    },
    {
      "epoch": 0.7989821882951654,
      "grad_norm": 0.1212642639875412,
      "learning_rate": 8.300104931794335e-06,
      "loss": 0.0059,
      "step": 3140
    },
    {
      "epoch": 0.8015267175572519,
      "grad_norm": 0.04367879405617714,
      "learning_rate": 8.19517313746065e-06,
      "loss": 0.0052,
      "step": 3150
    },
    {
      "epoch": 0.8040712468193384,
      "grad_norm": 0.5733098387718201,
      "learning_rate": 8.090241343126968e-06,
      "loss": 0.0158,
      "step": 3160
    },
    {
      "epoch": 0.806615776081425,
      "grad_norm": 0.03896990418434143,
      "learning_rate": 7.985309548793284e-06,
      "loss": 0.0092,
      "step": 3170
    },
    {
      "epoch": 0.8091603053435115,
      "grad_norm": 0.39722007513046265,
      "learning_rate": 7.880377754459602e-06,
      "loss": 0.0069,
      "step": 3180
    },
    {
      "epoch": 0.811704834605598,
      "grad_norm": 0.06284640729427338,
      "learning_rate": 7.77544596012592e-06,
      "loss": 0.0124,
      "step": 3190
    },
    {
      "epoch": 0.8142493638676844,
      "grad_norm": 0.8230289816856384,
      "learning_rate": 7.670514165792236e-06,
      "loss": 0.0253,
      "step": 3200
    },
    {
      "epoch": 0.816793893129771,
      "grad_norm": 0.045468445867300034,
      "learning_rate": 7.565582371458553e-06,
      "loss": 0.0012,
      "step": 3210
    },
    {
      "epoch": 0.8193384223918575,
      "grad_norm": 0.3366369307041168,
      "learning_rate": 7.46065057712487e-06,
      "loss": 0.0021,
      "step": 3220
    },
    {
      "epoch": 0.821882951653944,
      "grad_norm": 0.47629743814468384,
      "learning_rate": 7.355718782791186e-06,
      "loss": 0.0134,
      "step": 3230
    },
    {
      "epoch": 0.8244274809160306,
      "grad_norm": 0.18790696561336517,
      "learning_rate": 7.250786988457504e-06,
      "loss": 0.0045,
      "step": 3240
    },
    {
      "epoch": 0.8269720101781171,
      "grad_norm": 0.17202238738536835,
      "learning_rate": 7.14585519412382e-06,
      "loss": 0.0042,
      "step": 3250
    },
    {
      "epoch": 0.8295165394402035,
      "grad_norm": 0.4404882788658142,
      "learning_rate": 7.040923399790137e-06,
      "loss": 0.0042,
      "step": 3260
    },
    {
      "epoch": 0.8320610687022901,
      "grad_norm": 0.5554683208465576,
      "learning_rate": 6.9359916054564535e-06,
      "loss": 0.0049,
      "step": 3270
    },
    {
      "epoch": 0.8346055979643766,
      "grad_norm": 0.6356526613235474,
      "learning_rate": 6.831059811122771e-06,
      "loss": 0.0142,
      "step": 3280
    },
    {
      "epoch": 0.8371501272264631,
      "grad_norm": 0.2484181523323059,
      "learning_rate": 6.726128016789087e-06,
      "loss": 0.0065,
      "step": 3290
    },
    {
      "epoch": 0.8396946564885496,
      "grad_norm": 0.1756751537322998,
      "learning_rate": 6.621196222455405e-06,
      "loss": 0.0102,
      "step": 3300
    },
    {
      "epoch": 0.8422391857506362,
      "grad_norm": 0.5005545616149902,
      "learning_rate": 6.516264428121721e-06,
      "loss": 0.0018,
      "step": 3310
    },
    {
      "epoch": 0.8447837150127226,
      "grad_norm": 0.06506593525409698,
      "learning_rate": 6.411332633788039e-06,
      "loss": 0.0058,
      "step": 3320
    },
    {
      "epoch": 0.8473282442748091,
      "grad_norm": 0.07923749834299088,
      "learning_rate": 6.306400839454355e-06,
      "loss": 0.0055,
      "step": 3330
    },
    {
      "epoch": 0.8498727735368957,
      "grad_norm": 0.03359365090727806,
      "learning_rate": 6.201469045120672e-06,
      "loss": 0.01,
      "step": 3340
    },
    {
      "epoch": 0.8524173027989822,
      "grad_norm": 0.5358174443244934,
      "learning_rate": 6.096537250786989e-06,
      "loss": 0.0041,
      "step": 3350
    },
    {
      "epoch": 0.8549618320610687,
      "grad_norm": 0.26434558629989624,
      "learning_rate": 5.991605456453306e-06,
      "loss": 0.0117,
      "step": 3360
    },
    {
      "epoch": 0.8575063613231552,
      "grad_norm": 0.1969001740217209,
      "learning_rate": 5.886673662119623e-06,
      "loss": 0.0079,
      "step": 3370
    },
    {
      "epoch": 0.8600508905852418,
      "grad_norm": 0.3824959993362427,
      "learning_rate": 5.781741867785939e-06,
      "loss": 0.008,
      "step": 3380
    },
    {
      "epoch": 0.8625954198473282,
      "grad_norm": 0.0074090552516281605,
      "learning_rate": 5.676810073452257e-06,
      "loss": 0.0031,
      "step": 3390
    },
    {
      "epoch": 0.8651399491094147,
      "grad_norm": 0.22043420374393463,
      "learning_rate": 5.571878279118573e-06,
      "loss": 0.0048,
      "step": 3400
    },
    {
      "epoch": 0.8676844783715013,
      "grad_norm": 0.6048237681388855,
      "learning_rate": 5.466946484784891e-06,
      "loss": 0.0127,
      "step": 3410
    },
    {
      "epoch": 0.8702290076335878,
      "grad_norm": 0.004376487340778112,
      "learning_rate": 5.362014690451207e-06,
      "loss": 0.0044,
      "step": 3420
    },
    {
      "epoch": 0.8727735368956743,
      "grad_norm": 0.3400680422782898,
      "learning_rate": 5.257082896117524e-06,
      "loss": 0.0025,
      "step": 3430
    },
    {
      "epoch": 0.8753180661577609,
      "grad_norm": 0.004490234889090061,
      "learning_rate": 5.152151101783841e-06,
      "loss": 0.0024,
      "step": 3440
    },
    {
      "epoch": 0.8778625954198473,
      "grad_norm": 0.005273368675261736,
      "learning_rate": 5.047219307450158e-06,
      "loss": 0.0089,
      "step": 3450
    },
    {
      "epoch": 0.8804071246819338,
      "grad_norm": 0.0007822876796126366,
      "learning_rate": 4.942287513116475e-06,
      "loss": 0.0039,
      "step": 3460
    },
    {
      "epoch": 0.8829516539440203,
      "grad_norm": 0.088668093085289,
      "learning_rate": 4.8373557187827916e-06,
      "loss": 0.0022,
      "step": 3470
    },
    {
      "epoch": 0.8854961832061069,
      "grad_norm": 0.08840341866016388,
      "learning_rate": 4.7324239244491085e-06,
      "loss": 0.0047,
      "step": 3480
    },
    {
      "epoch": 0.8880407124681934,
      "grad_norm": 0.4743495285511017,
      "learning_rate": 4.627492130115425e-06,
      "loss": 0.0075,
      "step": 3490
    },
    {
      "epoch": 0.8905852417302799,
      "grad_norm": 0.08260899037122726,
      "learning_rate": 4.522560335781742e-06,
      "loss": 0.0188,
      "step": 3500
    },
    {
      "epoch": 0.8905852417302799,
      "eval_loss": 0.026213819161057472,
      "eval_runtime": 133.9096,
      "eval_samples_per_second": 59.607,
      "eval_steps_per_second": 0.47,
      "step": 3500
    },
    {
      "epoch": 0.8931297709923665,
      "grad_norm": 0.002863268367946148,
      "learning_rate": 4.417628541448059e-06,
      "loss": 0.0016,
      "step": 3510
    },
    {
      "epoch": 0.8956743002544529,
      "grad_norm": 0.4324285686016083,
      "learning_rate": 4.312696747114376e-06,
      "loss": 0.0029,
      "step": 3520
    },
    {
      "epoch": 0.8982188295165394,
      "grad_norm": 0.21640141308307648,
      "learning_rate": 4.207764952780693e-06,
      "loss": 0.0045,
      "step": 3530
    },
    {
      "epoch": 0.9007633587786259,
      "grad_norm": 1.0017634630203247,
      "learning_rate": 4.10283315844701e-06,
      "loss": 0.0092,
      "step": 3540
    },
    {
      "epoch": 0.9033078880407125,
      "grad_norm": 0.01565241441130638,
      "learning_rate": 3.997901364113327e-06,
      "loss": 0.0059,
      "step": 3550
    },
    {
      "epoch": 0.905852417302799,
      "grad_norm": 0.25544649362564087,
      "learning_rate": 3.892969569779644e-06,
      "loss": 0.0046,
      "step": 3560
    },
    {
      "epoch": 0.9083969465648855,
      "grad_norm": 0.007674202788621187,
      "learning_rate": 3.78803777544596e-06,
      "loss": 0.0085,
      "step": 3570
    },
    {
      "epoch": 0.910941475826972,
      "grad_norm": 0.004819818306714296,
      "learning_rate": 3.683105981112277e-06,
      "loss": 0.0054,
      "step": 3580
    },
    {
      "epoch": 0.9134860050890585,
      "grad_norm": 0.39050644636154175,
      "learning_rate": 3.578174186778594e-06,
      "loss": 0.0075,
      "step": 3590
    },
    {
      "epoch": 0.916030534351145,
      "grad_norm": 0.5367786884307861,
      "learning_rate": 3.473242392444911e-06,
      "loss": 0.0036,
      "step": 3600
    },
    {
      "epoch": 0.9185750636132316,
      "grad_norm": 0.002182970056310296,
      "learning_rate": 3.368310598111228e-06,
      "loss": 0.0005,
      "step": 3610
    },
    {
      "epoch": 0.9211195928753181,
      "grad_norm": 0.010968153364956379,
      "learning_rate": 3.2633788037775447e-06,
      "loss": 0.006,
      "step": 3620
    },
    {
      "epoch": 0.9236641221374046,
      "grad_norm": 0.04364249110221863,
      "learning_rate": 3.1584470094438616e-06,
      "loss": 0.0101,
      "step": 3630
    },
    {
      "epoch": 0.926208651399491,
      "grad_norm": 0.009574225172400475,
      "learning_rate": 3.053515215110179e-06,
      "loss": 0.0095,
      "step": 3640
    },
    {
      "epoch": 0.9287531806615776,
      "grad_norm": 0.023343412205576897,
      "learning_rate": 2.948583420776496e-06,
      "loss": 0.0057,
      "step": 3650
    },
    {
      "epoch": 0.9312977099236641,
      "grad_norm": 0.013017148710787296,
      "learning_rate": 2.8436516264428128e-06,
      "loss": 0.0039,
      "step": 3660
    },
    {
      "epoch": 0.9338422391857506,
      "grad_norm": 0.07142795622348785,
      "learning_rate": 2.7387198321091293e-06,
      "loss": 0.001,
      "step": 3670
    },
    {
      "epoch": 0.9363867684478372,
      "grad_norm": 0.03349682688713074,
      "learning_rate": 2.633788037775446e-06,
      "loss": 0.0018,
      "step": 3680
    },
    {
      "epoch": 0.9389312977099237,
      "grad_norm": 0.09319507330656052,
      "learning_rate": 2.528856243441763e-06,
      "loss": 0.0049,
      "step": 3690
    },
    {
      "epoch": 0.9414758269720102,
      "grad_norm": 0.08466564863920212,
      "learning_rate": 2.42392444910808e-06,
      "loss": 0.0101,
      "step": 3700
    },
    {
      "epoch": 0.9440203562340967,
      "grad_norm": 0.03436915576457977,
      "learning_rate": 2.318992654774397e-06,
      "loss": 0.0016,
      "step": 3710
    },
    {
      "epoch": 0.9465648854961832,
      "grad_norm": 0.05093689262866974,
      "learning_rate": 2.214060860440714e-06,
      "loss": 0.0055,
      "step": 3720
    },
    {
      "epoch": 0.9491094147582697,
      "grad_norm": 0.0019584556575864553,
      "learning_rate": 2.1091290661070307e-06,
      "loss": 0.0089,
      "step": 3730
    },
    {
      "epoch": 0.9516539440203562,
      "grad_norm": 0.6479087471961975,
      "learning_rate": 2.004197271773347e-06,
      "loss": 0.005,
      "step": 3740
    },
    {
      "epoch": 0.9541984732824428,
      "grad_norm": 0.22916148602962494,
      "learning_rate": 1.8992654774396643e-06,
      "loss": 0.0018,
      "step": 3750
    },
    {
      "epoch": 0.9567430025445293,
      "grad_norm": 0.021227942779660225,
      "learning_rate": 1.7943336831059812e-06,
      "loss": 0.0053,
      "step": 3760
    },
    {
      "epoch": 0.9592875318066157,
      "grad_norm": 0.028795473277568817,
      "learning_rate": 1.6894018887722981e-06,
      "loss": 0.0098,
      "step": 3770
    },
    {
      "epoch": 0.9618320610687023,
      "grad_norm": 0.23935425281524658,
      "learning_rate": 1.5844700944386152e-06,
      "loss": 0.0069,
      "step": 3780
    },
    {
      "epoch": 0.9643765903307888,
      "grad_norm": 0.2875511348247528,
      "learning_rate": 1.479538300104932e-06,
      "loss": 0.0028,
      "step": 3790
    },
    {
      "epoch": 0.9669211195928753,
      "grad_norm": 0.02270675264298916,
      "learning_rate": 1.3746065057712488e-06,
      "loss": 0.0049,
      "step": 3800
    },
    {
      "epoch": 0.9694656488549618,
      "grad_norm": 0.42798078060150146,
      "learning_rate": 1.2696747114375657e-06,
      "loss": 0.0097,
      "step": 3810
    },
    {
      "epoch": 0.9720101781170484,
      "grad_norm": 0.051214683800935745,
      "learning_rate": 1.1647429171038824e-06,
      "loss": 0.0025,
      "step": 3820
    },
    {
      "epoch": 0.9745547073791349,
      "grad_norm": 0.34169483184814453,
      "learning_rate": 1.0598111227701995e-06,
      "loss": 0.0186,
      "step": 3830
    },
    {
      "epoch": 0.9770992366412213,
      "grad_norm": 0.013889187946915627,
      "learning_rate": 9.548793284365165e-07,
      "loss": 0.0134,
      "step": 3840
    },
    {
      "epoch": 0.9796437659033079,
      "grad_norm": 0.08551329374313354,
      "learning_rate": 8.499475341028333e-07,
      "loss": 0.0094,
      "step": 3850
    },
    {
      "epoch": 0.9821882951653944,
      "grad_norm": 0.008397153578698635,
      "learning_rate": 7.450157397691502e-07,
      "loss": 0.0061,
      "step": 3860
    },
    {
      "epoch": 0.9847328244274809,
      "grad_norm": 0.14278727769851685,
      "learning_rate": 6.40083945435467e-07,
      "loss": 0.0015,
      "step": 3870
    },
    {
      "epoch": 0.9872773536895675,
      "grad_norm": 0.07914524525403976,
      "learning_rate": 5.351521511017839e-07,
      "loss": 0.0045,
      "step": 3880
    },
    {
      "epoch": 0.989821882951654,
      "grad_norm": 0.024170175194740295,
      "learning_rate": 4.3022035676810077e-07,
      "loss": 0.0052,
      "step": 3890
    },
    {
      "epoch": 0.9923664122137404,
      "grad_norm": 0.004945623688399792,
      "learning_rate": 3.252885624344176e-07,
      "loss": 0.0048,
      "step": 3900
    },
    {
      "epoch": 0.9949109414758269,
      "grad_norm": 0.21786199510097504,
      "learning_rate": 2.2035676810073456e-07,
      "loss": 0.0049,
      "step": 3910
    },
    {
      "epoch": 0.9974554707379135,
      "grad_norm": 0.11055775731801987,
      "learning_rate": 1.1542497376705142e-07,
      "loss": 0.0009,
      "step": 3920
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.363585296436213e-05,
      "learning_rate": 1.0493179433368311e-08,
      "loss": 0.0035,
      "step": 3930
    }
  ],
  "logging_steps": 10,
  "max_steps": 3930,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.273606668827065e+18,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}