{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.981488477521722,
  "eval_steps": 500,
  "global_step": 2640,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03022289384208538,
      "grad_norm": 2.8042919635772705,
      "learning_rate": 3.7878787878787882e-06,
      "loss": 2.0238,
      "step": 10
    },
    {
      "epoch": 0.06044578768417076,
      "grad_norm": 1.7337448596954346,
      "learning_rate": 7.5757575757575764e-06,
      "loss": 1.7313,
      "step": 20
    },
    {
      "epoch": 0.09066868152625614,
      "grad_norm": 1.5929458141326904,
      "learning_rate": 1.1363636363636365e-05,
      "loss": 1.2745,
      "step": 30
    },
    {
      "epoch": 0.12089157536834153,
      "grad_norm": 0.9085432887077332,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.8377,
      "step": 40
    },
    {
      "epoch": 0.1511144692104269,
      "grad_norm": 1.534300684928894,
      "learning_rate": 1.893939393939394e-05,
      "loss": 0.6819,
      "step": 50
    },
    {
      "epoch": 0.18133736305251227,
      "grad_norm": 1.3361095190048218,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.6215,
      "step": 60
    },
    {
      "epoch": 0.21156025689459765,
      "grad_norm": 1.5394957065582275,
      "learning_rate": 2.6515151515151516e-05,
      "loss": 0.583,
      "step": 70
    },
    {
      "epoch": 0.24178315073668305,
      "grad_norm": 2.153825521469116,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.5443,
      "step": 80
    },
    {
      "epoch": 0.2720060445787684,
      "grad_norm": 1.6891471147537231,
      "learning_rate": 3.409090909090909e-05,
      "loss": 0.5418,
      "step": 90
    },
    {
      "epoch": 0.3022289384208538,
      "grad_norm": 1.127630591392517,
      "learning_rate": 3.787878787878788e-05,
      "loss": 0.527,
      "step": 100
    },
    {
      "epoch": 0.3324518322629392,
      "grad_norm": 3.1586360931396484,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.5007,
      "step": 110
    },
    {
      "epoch": 0.36267472610502455,
      "grad_norm": 1.173498272895813,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.5222,
      "step": 120
    },
    {
      "epoch": 0.3928976199471099,
      "grad_norm": 1.0684623718261719,
      "learning_rate": 4.9242424242424245e-05,
      "loss": 0.4919,
      "step": 130
    },
    {
      "epoch": 0.4231205137891953,
      "grad_norm": 0.9527800679206848,
      "learning_rate": 5.303030303030303e-05,
      "loss": 0.4892,
      "step": 140
    },
    {
      "epoch": 0.45334340763128067,
      "grad_norm": 0.7964572906494141,
      "learning_rate": 5.6818181818181825e-05,
      "loss": 0.49,
      "step": 150
    },
    {
      "epoch": 0.4835663014733661,
      "grad_norm": 0.8295568823814392,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.4796,
      "step": 160
    },
    {
      "epoch": 0.5137891953154514,
      "grad_norm": 0.9569409489631653,
      "learning_rate": 6.439393939393939e-05,
      "loss": 0.4702,
      "step": 170
    },
    {
      "epoch": 0.5440120891575368,
      "grad_norm": 0.7955500483512878,
      "learning_rate": 6.818181818181818e-05,
      "loss": 0.4818,
      "step": 180
    },
    {
      "epoch": 0.5742349829996222,
      "grad_norm": 1.0039526224136353,
      "learning_rate": 7.196969696969698e-05,
      "loss": 0.471,
      "step": 190
    },
    {
      "epoch": 0.6044578768417076,
      "grad_norm": 1.1145097017288208,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.4874,
      "step": 200
    },
    {
      "epoch": 0.634680770683793,
      "grad_norm": 0.8231551647186279,
      "learning_rate": 7.954545454545455e-05,
      "loss": 0.4641,
      "step": 210
    },
    {
      "epoch": 0.6649036645258783,
      "grad_norm": 0.7706289887428284,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.4703,
      "step": 220
    },
    {
      "epoch": 0.6951265583679638,
      "grad_norm": 0.9305810332298279,
      "learning_rate": 8.712121212121212e-05,
      "loss": 0.4589,
      "step": 230
    },
    {
      "epoch": 0.7253494522100491,
      "grad_norm": 0.895759105682373,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.465,
      "step": 240
    },
    {
      "epoch": 0.7555723460521345,
      "grad_norm": 0.8712740540504456,
      "learning_rate": 9.469696969696971e-05,
      "loss": 0.4665,
      "step": 250
    },
    {
      "epoch": 0.7857952398942198,
      "grad_norm": 0.9580489993095398,
      "learning_rate": 9.848484848484849e-05,
      "loss": 0.4478,
      "step": 260
    },
    {
      "epoch": 0.8160181337363053,
      "grad_norm": 1.2908588647842407,
      "learning_rate": 9.999842657116665e-05,
      "loss": 0.4545,
      "step": 270
    },
    {
      "epoch": 0.8462410275783906,
      "grad_norm": 0.865259051322937,
      "learning_rate": 9.99888115313551e-05,
      "loss": 0.4572,
      "step": 280
    },
    {
      "epoch": 0.876463921420476,
      "grad_norm": 0.9201974272727966,
      "learning_rate": 9.997045725776174e-05,
      "loss": 0.4283,
      "step": 290
    },
    {
      "epoch": 0.9066868152625613,
      "grad_norm": 0.7685046792030334,
      "learning_rate": 9.99433669591504e-05,
      "loss": 0.4374,
      "step": 300
    },
    {
      "epoch": 0.9369097091046468,
      "grad_norm": 3.0810554027557373,
      "learning_rate": 9.99075453715499e-05,
      "loss": 0.4333,
      "step": 310
    },
    {
      "epoch": 0.9671326029467322,
      "grad_norm": 1.222629189491272,
      "learning_rate": 9.986299875742613e-05,
      "loss": 0.4485,
      "step": 320
    },
    {
      "epoch": 0.9973554967888175,
      "grad_norm": 1.5258204936981201,
      "learning_rate": 9.980973490458728e-05,
      "loss": 0.4491,
      "step": 330
    },
    {
      "epoch": 1.027956176803929,
      "grad_norm": 3.25398325920105,
      "learning_rate": 9.97477631248223e-05,
      "loss": 0.4296,
      "step": 340
    },
    {
      "epoch": 1.0581790706460144,
      "grad_norm": 0.6229041814804077,
      "learning_rate": 9.967709425227294e-05,
      "loss": 0.4209,
      "step": 350
    },
    {
      "epoch": 1.0884019644880998,
      "grad_norm": 0.7541171312332153,
      "learning_rate": 9.959774064153977e-05,
      "loss": 0.3989,
      "step": 360
    },
    {
      "epoch": 1.1186248583301852,
      "grad_norm": 1.236759066581726,
      "learning_rate": 9.950971616552222e-05,
      "loss": 0.4226,
      "step": 370
    },
    {
      "epoch": 1.1488477521722704,
      "grad_norm": 1.4303499460220337,
      "learning_rate": 9.941303621299332e-05,
      "loss": 0.3918,
      "step": 380
    },
    {
      "epoch": 1.1790706460143558,
      "grad_norm": 1.2374080419540405,
      "learning_rate": 9.930771768590933e-05,
      "loss": 0.4012,
      "step": 390
    },
    {
      "epoch": 1.2092935398564413,
      "grad_norm": 0.7818596959114075,
      "learning_rate": 9.919377899645497e-05,
      "loss": 0.4226,
      "step": 400
    },
    {
      "epoch": 1.2395164336985267,
      "grad_norm": 0.8049737811088562,
      "learning_rate": 9.907124006382438e-05,
      "loss": 0.3948,
      "step": 410
    },
    {
      "epoch": 1.269739327540612,
      "grad_norm": 2.329979181289673,
      "learning_rate": 9.894012231073894e-05,
      "loss": 0.4052,
      "step": 420
    },
    {
      "epoch": 1.2999622213826973,
      "grad_norm": 0.8067731261253357,
      "learning_rate": 9.880044865970192e-05,
      "loss": 0.4049,
      "step": 430
    },
    {
      "epoch": 1.3301851152247828,
      "grad_norm": 0.8137128949165344,
      "learning_rate": 9.865224352899119e-05,
      "loss": 0.4083,
      "step": 440
    },
    {
      "epoch": 1.3604080090668682,
      "grad_norm": 1.8560458421707153,
      "learning_rate": 9.849553282839025e-05,
      "loss": 0.3963,
      "step": 450
    },
    {
      "epoch": 1.3906309029089536,
      "grad_norm": 0.6536078453063965,
      "learning_rate": 9.833034395465866e-05,
      "loss": 0.3951,
      "step": 460
    },
    {
      "epoch": 1.4208537967510388,
      "grad_norm": 0.57303386926651,
      "learning_rate": 9.815670578674232e-05,
      "loss": 0.3823,
      "step": 470
    },
    {
      "epoch": 1.4510766905931243,
      "grad_norm": 0.713214635848999,
      "learning_rate": 9.797464868072488e-05,
      "loss": 0.4022,
      "step": 480
    },
    {
      "epoch": 1.4812995844352097,
      "grad_norm": 0.6472973227500916,
      "learning_rate": 9.778420446452063e-05,
      "loss": 0.4026,
      "step": 490
    },
    {
      "epoch": 1.511522478277295,
      "grad_norm": 0.7143449783325195,
      "learning_rate": 9.75854064323104e-05,
      "loss": 0.4079,
      "step": 500
    },
    {
      "epoch": 1.5417453721193803,
      "grad_norm": 0.5345566272735596,
      "learning_rate": 9.737828933872075e-05,
      "loss": 0.3996,
      "step": 510
    },
    {
      "epoch": 1.5719682659614658,
      "grad_norm": 1.006264090538025,
      "learning_rate": 9.716288939274819e-05,
      "loss": 0.4077,
      "step": 520
    },
    {
      "epoch": 1.6021911598035512,
      "grad_norm": 0.7178054451942444,
      "learning_rate": 9.693924425142886e-05,
      "loss": 0.4167,
      "step": 530
    },
    {
      "epoch": 1.6324140536456366,
      "grad_norm": 0.8414012789726257,
      "learning_rate": 9.670739301325534e-05,
      "loss": 0.4104,
      "step": 540
    },
    {
      "epoch": 1.662636947487722,
      "grad_norm": 0.6398180723190308,
      "learning_rate": 9.646737621134112e-05,
      "loss": 0.3756,
      "step": 550
    },
    {
      "epoch": 1.6928598413298075,
      "grad_norm": 0.7549942135810852,
      "learning_rate": 9.62192358063346e-05,
      "loss": 0.3879,
      "step": 560
    },
    {
      "epoch": 1.7230827351718927,
      "grad_norm": 0.6766812205314636,
      "learning_rate": 9.596301517908328e-05,
      "loss": 0.3841,
      "step": 570
    },
    {
      "epoch": 1.7533056290139781,
      "grad_norm": 0.8327857255935669,
      "learning_rate": 9.56987591230498e-05,
      "loss": 0.3958,
      "step": 580
    },
    {
      "epoch": 1.7835285228560633,
      "grad_norm": 0.6985905766487122,
      "learning_rate": 9.542651383648091e-05,
      "loss": 0.4113,
      "step": 590
    },
    {
      "epoch": 1.8137514166981488,
      "grad_norm": 0.6444863677024841,
      "learning_rate": 9.514632691433107e-05,
      "loss": 0.392,
      "step": 600
    },
    {
      "epoch": 1.8439743105402342,
      "grad_norm": 0.8134333491325378,
      "learning_rate": 9.48582473399415e-05,
      "loss": 0.3996,
      "step": 610
    },
    {
      "epoch": 1.8741972043823196,
      "grad_norm": 0.6443466544151306,
      "learning_rate": 9.456232547647694e-05,
      "loss": 0.3596,
      "step": 620
    },
    {
      "epoch": 1.904420098224405,
      "grad_norm": 0.7170634269714355,
      "learning_rate": 9.425861305812082e-05,
      "loss": 0.3732,
      "step": 630
    },
    {
      "epoch": 1.9346429920664905,
      "grad_norm": 1.3517498970031738,
      "learning_rate": 9.394716318103098e-05,
      "loss": 0.3791,
      "step": 640
    },
    {
      "epoch": 1.9648658859085757,
      "grad_norm": 2.1151044368743896,
      "learning_rate": 9.362803029405712e-05,
      "loss": 0.3634,
      "step": 650
    },
    {
      "epoch": 1.9950887797506611,
      "grad_norm": 0.6365861296653748,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.393,
      "step": 660
    },
    {
      "epoch": 2.0256894597657724,
      "grad_norm": 1.6099416017532349,
      "learning_rate": 9.296693999196728e-05,
      "loss": 0.3329,
      "step": 670
    },
    {
      "epoch": 2.055912353607858,
      "grad_norm": 1.0266637802124023,
      "learning_rate": 9.262509815116732e-05,
      "loss": 0.3534,
      "step": 680
    },
    {
      "epoch": 2.0861352474499433,
      "grad_norm": 0.803153395652771,
      "learning_rate": 9.227580442891022e-05,
      "loss": 0.3328,
      "step": 690
    },
    {
      "epoch": 2.1163581412920287,
      "grad_norm": 0.6713061928749084,
      "learning_rate": 9.191911989005037e-05,
      "loss": 0.3524,
      "step": 700
    },
    {
      "epoch": 2.146581035134114,
      "grad_norm": 0.5472012162208557,
      "learning_rate": 9.155510689153282e-05,
      "loss": 0.3368,
      "step": 710
    },
    {
      "epoch": 2.1768039289761996,
      "grad_norm": 1.0651007890701294,
      "learning_rate": 9.118382907149165e-05,
      "loss": 0.2971,
      "step": 720
    },
    {
      "epoch": 2.207026822818285,
      "grad_norm": 0.7630440592765808,
      "learning_rate": 9.080535133812469e-05,
      "loss": 0.3375,
      "step": 730
    },
    {
      "epoch": 2.2372497166603704,
      "grad_norm": 0.6513319611549377,
      "learning_rate": 9.041973985834595e-05,
      "loss": 0.3234,
      "step": 740
    },
    {
      "epoch": 2.2674726105024554,
      "grad_norm": 0.5795676112174988,
      "learning_rate": 9.002706204621803e-05,
      "loss": 0.3288,
      "step": 750
    },
    {
      "epoch": 2.297695504344541,
      "grad_norm": 0.6500043272972107,
      "learning_rate": 8.962738655116658e-05,
      "loss": 0.3356,
      "step": 760
    },
    {
      "epoch": 2.3279183981866263,
      "grad_norm": 0.665492594242096,
      "learning_rate": 8.922078324597879e-05,
      "loss": 0.3359,
      "step": 770
    },
    {
      "epoch": 2.3581412920287117,
      "grad_norm": 0.7372997403144836,
      "learning_rate": 8.880732321458784e-05,
      "loss": 0.3347,
      "step": 780
    },
    {
      "epoch": 2.388364185870797,
      "grad_norm": 0.617290735244751,
      "learning_rate": 8.838707873964587e-05,
      "loss": 0.3365,
      "step": 790
    },
    {
      "epoch": 2.4185870797128826,
      "grad_norm": 0.7033390402793884,
      "learning_rate": 8.796012328988716e-05,
      "loss": 0.3386,
      "step": 800
    },
    {
      "epoch": 2.448809973554968,
      "grad_norm": 1.019696831703186,
      "learning_rate": 8.752653150728411e-05,
      "loss": 0.3244,
      "step": 810
    },
    {
      "epoch": 2.4790328673970534,
      "grad_norm": 0.8520340323448181,
      "learning_rate": 8.708637919399798e-05,
      "loss": 0.3387,
      "step": 820
    },
    {
      "epoch": 2.509255761239139,
      "grad_norm": 1.2463282346725464,
      "learning_rate": 8.663974329912696e-05,
      "loss": 0.3334,
      "step": 830
    },
    {
      "epoch": 2.539478655081224,
      "grad_norm": 0.6218689680099487,
      "learning_rate": 8.618670190525352e-05,
      "loss": 0.3412,
      "step": 840
    },
    {
      "epoch": 2.5697015489233093,
      "grad_norm": 0.635360836982727,
      "learning_rate": 8.572733421479382e-05,
      "loss": 0.3292,
      "step": 850
    },
    {
      "epoch": 2.5999244427653947,
      "grad_norm": 0.5387209057807922,
      "learning_rate": 8.526172053615121e-05,
      "loss": 0.3359,
      "step": 860
    },
    {
      "epoch": 2.63014733660748,
      "grad_norm": 0.5592532157897949,
      "learning_rate": 8.478994226967638e-05,
      "loss": 0.3466,
      "step": 870
    },
    {
      "epoch": 2.6603702304495656,
      "grad_norm": 0.7285254597663879,
      "learning_rate": 8.43120818934367e-05,
      "loss": 0.3345,
      "step": 880
    },
    {
      "epoch": 2.690593124291651,
      "grad_norm": 0.6214547753334045,
      "learning_rate": 8.382822294879698e-05,
      "loss": 0.3464,
      "step": 890
    },
    {
      "epoch": 2.7208160181337364,
      "grad_norm": 0.4706718325614929,
      "learning_rate": 8.333845002581458e-05,
      "loss": 0.3267,
      "step": 900
    },
    {
      "epoch": 2.751038911975822,
      "grad_norm": 1.0929789543151855,
      "learning_rate": 8.284284874845104e-05,
      "loss": 0.3327,
      "step": 910
    },
    {
      "epoch": 2.7812618058179073,
      "grad_norm": 0.6653083562850952,
      "learning_rate": 8.234150575960288e-05,
      "loss": 0.3146,
      "step": 920
    },
    {
      "epoch": 2.8114846996599923,
      "grad_norm": 0.4961538314819336,
      "learning_rate": 8.183450870595441e-05,
      "loss": 0.3382,
      "step": 930
    },
    {
      "epoch": 2.8417075935020777,
      "grad_norm": 1.2516436576843262,
      "learning_rate": 8.132194622265507e-05,
      "loss": 0.3293,
      "step": 940
    },
    {
      "epoch": 2.871930487344163,
      "grad_norm": 1.3802598714828491,
      "learning_rate": 8.080390791782375e-05,
      "loss": 0.3344,
      "step": 950
    },
    {
      "epoch": 2.9021533811862485,
      "grad_norm": 0.619377851486206,
      "learning_rate": 8.028048435688333e-05,
      "loss": 0.3396,
      "step": 960
    },
    {
      "epoch": 2.932376275028334,
      "grad_norm": 0.5353909134864807,
      "learning_rate": 7.975176704672758e-05,
      "loss": 0.3375,
      "step": 970
    },
    {
      "epoch": 2.9625991688704194,
      "grad_norm": 0.6385555267333984,
      "learning_rate": 7.921784841972355e-05,
      "loss": 0.3308,
      "step": 980
    },
    {
      "epoch": 2.992822062712505,
      "grad_norm": 1.1463056802749634,
      "learning_rate": 7.86788218175523e-05,
      "loss": 0.3283,
      "step": 990
    },
    {
      "epoch": 3.0234227427276164,
      "grad_norm": 0.7241145968437195,
      "learning_rate": 7.813478147489052e-05,
      "loss": 0.2587,
      "step": 1000
    },
    {
      "epoch": 3.0536456365697013,
      "grad_norm": 0.8696322441101074,
      "learning_rate": 7.758582250293596e-05,
      "loss": 0.2624,
      "step": 1010
    },
    {
      "epoch": 3.0838685304117868,
      "grad_norm": 0.7954265475273132,
      "learning_rate": 7.703204087277988e-05,
      "loss": 0.2414,
      "step": 1020
    },
    {
      "epoch": 3.114091424253872,
      "grad_norm": 0.611754298210144,
      "learning_rate": 7.647353339862895e-05,
      "loss": 0.2506,
      "step": 1030
    },
    {
      "epoch": 3.1443143180959576,
      "grad_norm": 0.7143287658691406,
      "learning_rate": 7.591039772087977e-05,
      "loss": 0.2516,
      "step": 1040
    },
    {
      "epoch": 3.174537211938043,
      "grad_norm": 0.597282350063324,
      "learning_rate": 7.534273228904915e-05,
      "loss": 0.2628,
      "step": 1050
    },
    {
      "epoch": 3.2047601057801285,
      "grad_norm": 0.7054301500320435,
      "learning_rate": 7.477063634456263e-05,
      "loss": 0.2611,
      "step": 1060
    },
    {
      "epoch": 3.234982999622214,
      "grad_norm": 0.5925849080085754,
      "learning_rate": 7.41942099034048e-05,
      "loss": 0.2782,
      "step": 1070
    },
    {
      "epoch": 3.2652058934642993,
      "grad_norm": 0.6402950286865234,
      "learning_rate": 7.361355373863414e-05,
      "loss": 0.2506,
      "step": 1080
    },
    {
      "epoch": 3.2954287873063848,
      "grad_norm": 0.6893470287322998,
      "learning_rate": 7.302876936276546e-05,
      "loss": 0.2593,
      "step": 1090
    },
    {
      "epoch": 3.3256516811484698,
      "grad_norm": 0.7923823595046997,
      "learning_rate": 7.243995901002312e-05,
      "loss": 0.2677,
      "step": 1100
    },
    {
      "epoch": 3.355874574990555,
      "grad_norm": 0.6485563516616821,
      "learning_rate": 7.184722561846798e-05,
      "loss": 0.2596,
      "step": 1110
    },
    {
      "epoch": 3.3860974688326406,
      "grad_norm": 1.6831997632980347,
      "learning_rate": 7.12506728120015e-05,
      "loss": 0.2507,
      "step": 1120
    },
    {
      "epoch": 3.416320362674726,
      "grad_norm": 0.6467958092689514,
      "learning_rate": 7.065040488224974e-05,
      "loss": 0.2626,
      "step": 1130
    },
    {
      "epoch": 3.4465432565168115,
      "grad_norm": 0.7198959589004517,
      "learning_rate": 7.004652677033068e-05,
      "loss": 0.2601,
      "step": 1140
    },
    {
      "epoch": 3.476766150358897,
      "grad_norm": 0.6917047500610352,
      "learning_rate": 6.94391440485081e-05,
      "loss": 0.2692,
      "step": 1150
    },
    {
      "epoch": 3.5069890442009823,
      "grad_norm": 0.9341140985488892,
      "learning_rate": 6.882836290173493e-05,
      "loss": 0.2609,
      "step": 1160
    },
    {
      "epoch": 3.5372119380430678,
      "grad_norm": 0.7227875590324402,
      "learning_rate": 6.821429010908971e-05,
      "loss": 0.2561,
      "step": 1170
    },
    {
      "epoch": 3.567434831885153,
      "grad_norm": 0.7613323926925659,
      "learning_rate": 6.759703302510898e-05,
      "loss": 0.2461,
      "step": 1180
    },
    {
      "epoch": 3.597657725727238,
      "grad_norm": 0.5406620502471924,
      "learning_rate": 6.697669956101914e-05,
      "loss": 0.2669,
      "step": 1190
    },
    {
      "epoch": 3.6278806195693236,
      "grad_norm": 0.6179462671279907,
      "learning_rate": 6.635339816587109e-05,
      "loss": 0.2655,
      "step": 1200
    },
    {
      "epoch": 3.658103513411409,
      "grad_norm": 0.7479238510131836,
      "learning_rate": 6.572723780758069e-05,
      "loss": 0.2567,
      "step": 1210
    },
    {
      "epoch": 3.6883264072534945,
      "grad_norm": 0.6450394988059998,
      "learning_rate": 6.509832795387858e-05,
      "loss": 0.2721,
      "step": 1220
    },
    {
      "epoch": 3.71854930109558,
      "grad_norm": 0.7902218103408813,
      "learning_rate": 6.446677855317264e-05,
      "loss": 0.2745,
      "step": 1230
    },
    {
      "epoch": 3.7487721949376653,
      "grad_norm": 0.5468536019325256,
      "learning_rate": 6.383270001532635e-05,
      "loss": 0.2662,
      "step": 1240
    },
    {
      "epoch": 3.7789950887797508,
      "grad_norm": 0.5629965662956238,
      "learning_rate": 6.319620319235659e-05,
      "loss": 0.2583,
      "step": 1250
    },
    {
      "epoch": 3.809217982621836,
      "grad_norm": 0.601555585861206,
      "learning_rate": 6.255739935905396e-05,
      "loss": 0.2681,
      "step": 1260
    },
    {
      "epoch": 3.8394408764639216,
      "grad_norm": 0.7035398483276367,
      "learning_rate": 6.191640019352942e-05,
      "loss": 0.2723,
      "step": 1270
    },
    {
      "epoch": 3.8696637703060066,
      "grad_norm": 0.6691277623176575,
      "learning_rate": 6.127331775769023e-05,
      "loss": 0.2582,
      "step": 1280
    },
    {
      "epoch": 3.899886664148092,
      "grad_norm": 0.5650741457939148,
      "learning_rate": 6.062826447764883e-05,
      "loss": 0.2667,
      "step": 1290
    },
    {
      "epoch": 3.9301095579901775,
      "grad_norm": 0.8159789443016052,
      "learning_rate": 5.998135312406821e-05,
      "loss": 0.2657,
      "step": 1300
    },
    {
      "epoch": 3.960332451832263,
      "grad_norm": 0.6696954965591431,
      "learning_rate": 5.9332696792446727e-05,
      "loss": 0.2697,
      "step": 1310
    },
    {
      "epoch": 3.9905553456743483,
      "grad_norm": 0.584435760974884,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.246,
      "step": 1320
    },
    {
      "epoch": 4.02115602568946,
      "grad_norm": 0.5735254883766174,
      "learning_rate": 5.803060308256824e-05,
      "loss": 0.1935,
      "step": 1330
    },
    {
      "epoch": 4.051378919531545,
      "grad_norm": 0.6788007616996765,
      "learning_rate": 5.737739334127611e-05,
      "loss": 0.1943,
      "step": 1340
    },
    {
      "epoch": 4.081601813373631,
      "grad_norm": 0.6538478136062622,
      "learning_rate": 5.6722893856076596e-05,
      "loss": 0.1877,
      "step": 1350
    },
    {
      "epoch": 4.111824707215716,
      "grad_norm": 0.6879804134368896,
      "learning_rate": 5.60672190490541e-05,
      "loss": 0.189,
      "step": 1360
    },
    {
      "epoch": 4.142047601057802,
      "grad_norm": 0.642549991607666,
      "learning_rate": 5.541048354776721e-05,
      "loss": 0.1848,
      "step": 1370
    },
    {
      "epoch": 4.1722704948998866,
      "grad_norm": 0.6130525469779968,
      "learning_rate": 5.475280216520913e-05,
      "loss": 0.1901,
      "step": 1380
    },
    {
      "epoch": 4.202493388741972,
      "grad_norm": 0.880362331867218,
      "learning_rate": 5.409428987973564e-05,
      "loss": 0.1804,
      "step": 1390
    },
    {
      "epoch": 4.232716282584057,
      "grad_norm": 0.6228994727134705,
      "learning_rate": 5.343506181496405e-05,
      "loss": 0.1915,
      "step": 1400
    },
    {
      "epoch": 4.262939176426142,
      "grad_norm": 0.6418446898460388,
      "learning_rate": 5.277523321964701e-05,
      "loss": 0.1877,
      "step": 1410
    },
    {
      "epoch": 4.293162070268228,
      "grad_norm": 0.7304417490959167,
      "learning_rate": 5.2114919447524155e-05,
      "loss": 0.1905,
      "step": 1420
    },
    {
      "epoch": 4.323384964110313,
      "grad_norm": 1.017180323600769,
      "learning_rate": 5.145423593715557e-05,
      "loss": 0.1736,
      "step": 1430
    },
    {
      "epoch": 4.353607857952399,
      "grad_norm": 0.7358663082122803,
      "learning_rate": 5.0793298191740404e-05,
      "loss": 0.19,
      "step": 1440
    },
    {
      "epoch": 4.383830751794484,
      "grad_norm": 0.674080491065979,
      "learning_rate": 5.013222175892411e-05,
      "loss": 0.1907,
      "step": 1450
    },
    {
      "epoch": 4.41405364563657,
      "grad_norm": 0.645470142364502,
      "learning_rate": 4.9471122210598034e-05,
      "loss": 0.196,
      "step": 1460
    },
    {
      "epoch": 4.444276539478655,
      "grad_norm": 0.5910624861717224,
      "learning_rate": 4.881011512269463e-05,
      "loss": 0.187,
      "step": 1470
    },
    {
      "epoch": 4.474499433320741,
      "grad_norm": 0.7850512862205505,
      "learning_rate": 4.8149316054982095e-05,
      "loss": 0.186,
      "step": 1480
    },
    {
      "epoch": 4.504722327162826,
      "grad_norm": 1.1707135438919067,
      "learning_rate": 4.748884053086175e-05,
      "loss": 0.1843,
      "step": 1490
    },
    {
      "epoch": 4.534945221004911,
      "grad_norm": 0.7980189323425293,
      "learning_rate": 4.6828804017171776e-05,
      "loss": 0.1985,
      "step": 1500
    },
    {
      "epoch": 4.565168114846997,
      "grad_norm": 0.5942202806472778,
      "learning_rate": 4.616932190400089e-05,
      "loss": 0.1949,
      "step": 1510
    },
    {
      "epoch": 4.595391008689082,
      "grad_norm": 0.6443480253219604,
      "learning_rate": 4.551050948451542e-05,
      "loss": 0.1847,
      "step": 1520
    },
    {
      "epoch": 4.6256139025311676,
      "grad_norm": 0.6935557126998901,
      "learning_rate": 4.485248193480328e-05,
      "loss": 0.1819,
      "step": 1530
    },
    {
      "epoch": 4.6558367963732525,
      "grad_norm": 0.8106910586357117,
      "learning_rate": 4.4195354293738484e-05,
      "loss": 0.1839,
      "step": 1540
    },
    {
      "epoch": 4.686059690215338,
      "grad_norm": 0.8624558448791504,
      "learning_rate": 4.353924144286963e-05,
      "loss": 0.1848,
      "step": 1550
    },
    {
      "epoch": 4.716282584057423,
      "grad_norm": 0.7457184791564941,
      "learning_rate": 4.288425808633575e-05,
      "loss": 0.1843,
      "step": 1560
    },
    {
      "epoch": 4.746505477899509,
      "grad_norm": 0.7222162485122681,
      "learning_rate": 4.223051873081349e-05,
      "loss": 0.1962,
      "step": 1570
    },
    {
      "epoch": 4.776728371741594,
      "grad_norm": 0.6403977274894714,
      "learning_rate": 4.157813766549848e-05,
      "loss": 0.1872,
      "step": 1580
    },
    {
      "epoch": 4.806951265583679,
      "grad_norm": 1.0512725114822388,
      "learning_rate": 4.092722894212487e-05,
      "loss": 0.186,
      "step": 1590
    },
    {
      "epoch": 4.837174159425765,
      "grad_norm": 0.7039276361465454,
      "learning_rate": 4.027790635502646e-05,
      "loss": 0.1854,
      "step": 1600
    },
    {
      "epoch": 4.86739705326785,
      "grad_norm": 1.100012183189392,
      "learning_rate": 3.963028342124265e-05,
      "loss": 0.1904,
      "step": 1610
    },
    {
      "epoch": 4.897619947109936,
      "grad_norm": 0.9361763000488281,
      "learning_rate": 3.898447336067297e-05,
      "loss": 0.1905,
      "step": 1620
    },
    {
      "epoch": 4.927842840952021,
      "grad_norm": 0.7108214497566223,
      "learning_rate": 3.8340589076283645e-05,
      "loss": 0.1807,
      "step": 1630
    },
    {
      "epoch": 4.958065734794107,
      "grad_norm": 0.5470288991928101,
      "learning_rate": 3.769874313436933e-05,
      "loss": 0.1765,
      "step": 1640
    },
    {
      "epoch": 4.988288628636192,
      "grad_norm": 1.0886564254760742,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.178,
      "step": 1650
    },
    {
      "epoch": 5.018889308651303,
      "grad_norm": 1.7127904891967773,
      "learning_rate": 3.64216147417737e-05,
      "loss": 0.1444,
      "step": 1660
    },
    {
      "epoch": 5.049112202493388,
      "grad_norm": 0.5874884128570557,
      "learning_rate": 3.5786555563525745e-05,
      "loss": 0.1326,
      "step": 1670
    },
    {
      "epoch": 5.079335096335474,
      "grad_norm": 0.6397553086280823,
      "learning_rate": 3.515398123358627e-05,
      "loss": 0.1269,
      "step": 1680
    },
    {
      "epoch": 5.109557990177559,
      "grad_norm": 0.8377091288566589,
      "learning_rate": 3.4524002341000715e-05,
      "loss": 0.1311,
      "step": 1690
    },
    {
      "epoch": 5.139780884019645,
      "grad_norm": 0.6393041014671326,
      "learning_rate": 3.389672902107044e-05,
      "loss": 0.132,
      "step": 1700
    },
    {
      "epoch": 5.17000377786173,
      "grad_norm": 0.8685671091079712,
      "learning_rate": 3.3272270936098246e-05,
      "loss": 0.1195,
      "step": 1710
    },
    {
      "epoch": 5.200226671703816,
      "grad_norm": 1.0302200317382812,
      "learning_rate": 3.2650737256216886e-05,
      "loss": 0.1309,
      "step": 1720
    },
    {
      "epoch": 5.230449565545901,
      "grad_norm": 0.6609297394752502,
      "learning_rate": 3.2032236640303545e-05,
      "loss": 0.127,
      "step": 1730
    },
    {
      "epoch": 5.260672459387987,
      "grad_norm": 0.717876136302948,
      "learning_rate": 3.141687721698363e-05,
      "loss": 0.1288,
      "step": 1740
    },
    {
      "epoch": 5.290895353230072,
      "grad_norm": 0.7257722616195679,
      "learning_rate": 3.0804766565727316e-05,
      "loss": 0.1289,
      "step": 1750
    },
    {
      "epoch": 5.321118247072157,
      "grad_norm": 0.5968477129936218,
      "learning_rate": 3.019601169804216e-05,
      "loss": 0.1246,
      "step": 1760
    },
    {
      "epoch": 5.351341140914243,
      "grad_norm": 0.5718531012535095,
      "learning_rate": 2.959071903876486e-05,
      "loss": 0.1221,
      "step": 1770
    },
    {
      "epoch": 5.381564034756328,
      "grad_norm": 0.7471643090248108,
      "learning_rate": 2.898899440745569e-05,
      "loss": 0.1247,
      "step": 1780
    },
    {
      "epoch": 5.4117869285984135,
      "grad_norm": 0.8166024684906006,
      "learning_rate": 2.8390942999898766e-05,
      "loss": 0.125,
      "step": 1790
    },
    {
      "epoch": 5.4420098224404985,
      "grad_norm": 0.6300913691520691,
      "learning_rate": 2.7796669369711294e-05,
      "loss": 0.1252,
      "step": 1800
    },
    {
      "epoch": 5.472232716282584,
      "grad_norm": 0.6817363500595093,
      "learning_rate": 2.720627741006505e-05,
      "loss": 0.1272,
      "step": 1810
    },
    {
      "epoch": 5.502455610124669,
      "grad_norm": 0.6244085431098938,
      "learning_rate": 2.6619870335523432e-05,
      "loss": 0.1218,
      "step": 1820
    },
    {
      "epoch": 5.532678503966755,
      "grad_norm": 0.7697808742523193,
      "learning_rate": 2.603755066399718e-05,
      "loss": 0.1221,
      "step": 1830
    },
    {
      "epoch": 5.56290139780884,
      "grad_norm": 0.728665292263031,
      "learning_rate": 2.5459420198821605e-05,
      "loss": 0.1267,
      "step": 1840
    },
    {
      "epoch": 5.593124291650925,
      "grad_norm": 0.8812174797058105,
      "learning_rate": 2.4885580010959153e-05,
      "loss": 0.1271,
      "step": 1850
    },
    {
      "epoch": 5.623347185493011,
      "grad_norm": 0.6068354845046997,
      "learning_rate": 2.4316130421329697e-05,
      "loss": 0.1257,
      "step": 1860
    },
    {
      "epoch": 5.653570079335096,
      "grad_norm": 0.6595832109451294,
      "learning_rate": 2.3751170983272e-05,
      "loss": 0.1233,
      "step": 1870
    },
    {
      "epoch": 5.683792973177182,
      "grad_norm": 0.5363273024559021,
      "learning_rate": 2.319080046513954e-05,
      "loss": 0.1256,
      "step": 1880
    },
    {
      "epoch": 5.714015867019267,
      "grad_norm": 0.5247385501861572,
      "learning_rate": 2.2635116833033393e-05,
      "loss": 0.1221,
      "step": 1890
    },
    {
      "epoch": 5.744238760861353,
      "grad_norm": 0.7972837686538696,
      "learning_rate": 2.2084217233675385e-05,
      "loss": 0.1191,
      "step": 1900
    },
    {
      "epoch": 5.774461654703438,
      "grad_norm": 0.766128659248352,
      "learning_rate": 2.1538197977424618e-05,
      "loss": 0.1207,
      "step": 1910
    },
    {
      "epoch": 5.804684548545524,
      "grad_norm": 0.6039882302284241,
      "learning_rate": 2.09971545214401e-05,
      "loss": 0.1187,
      "step": 1920
    },
    {
      "epoch": 5.834907442387609,
      "grad_norm": 0.6861318349838257,
      "learning_rate": 2.0461181452992496e-05,
      "loss": 0.1214,
      "step": 1930
    },
    {
      "epoch": 5.865130336229694,
      "grad_norm": 0.530323326587677,
      "learning_rate": 1.9930372472928095e-05,
      "loss": 0.1295,
      "step": 1940
    },
    {
      "epoch": 5.8953532300717795,
      "grad_norm": 0.6027620434761047,
      "learning_rate": 1.9404820379287675e-05,
      "loss": 0.1316,
      "step": 1950
    },
    {
      "epoch": 5.9255761239138645,
      "grad_norm": 0.5568642020225525,
      "learning_rate": 1.888461705108318e-05,
      "loss": 0.1185,
      "step": 1960
    },
    {
      "epoch": 5.95579901775595,
      "grad_norm": 0.677137553691864,
      "learning_rate": 1.8369853432235024e-05,
      "loss": 0.1213,
      "step": 1970
    },
    {
      "epoch": 5.986021911598035,
      "grad_norm": 0.7476499676704407,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.1258,
      "step": 1980
    },
    {
      "epoch": 6.016622591613147,
      "grad_norm": 0.3778318464756012,
      "learning_rate": 1.73570043276035e-05,
      "loss": 0.1076,
      "step": 1990
    },
    {
      "epoch": 6.046845485455233,
      "grad_norm": 0.5993187427520752,
      "learning_rate": 1.68590959119452e-05,
      "loss": 0.0849,
      "step": 2000
    },
    {
      "epoch": 6.077068379297318,
      "grad_norm": 0.5854635834693909,
      "learning_rate": 1.6366981314937376e-05,
      "loss": 0.0911,
      "step": 2010
    },
    {
      "epoch": 6.107291273139403,
      "grad_norm": 0.48482272028923035,
      "learning_rate": 1.5880746569921868e-05,
      "loss": 0.1051,
      "step": 2020
    },
    {
      "epoch": 6.1375141669814886,
      "grad_norm": 0.42831626534461975,
      "learning_rate": 1.5400476682302454e-05,
      "loss": 0.087,
      "step": 2030
    },
    {
      "epoch": 6.1677370608235735,
      "grad_norm": 0.7245472073554993,
      "learning_rate": 1.4926255614683932e-05,
      "loss": 0.0981,
      "step": 2040
    },
    {
      "epoch": 6.197959954665659,
      "grad_norm": 0.4705452024936676,
      "learning_rate": 1.4458166272193425e-05,
      "loss": 0.0895,
      "step": 2050
    },
    {
      "epoch": 6.228182848507744,
      "grad_norm": 0.438855916261673,
      "learning_rate": 1.3996290487986568e-05,
      "loss": 0.088,
      "step": 2060
    },
    {
      "epoch": 6.25840574234983,
      "grad_norm": 0.641183614730835,
      "learning_rate": 1.354070900894115e-05,
      "loss": 0.0989,
      "step": 2070
    },
    {
      "epoch": 6.288628636191915,
      "grad_norm": 0.4393808841705322,
      "learning_rate": 1.3091501481540674e-05,
      "loss": 0.086,
      "step": 2080
    },
    {
      "epoch": 6.318851530034001,
      "grad_norm": 0.5123514533042908,
      "learning_rate": 1.264874643795021e-05,
      "loss": 0.0874,
      "step": 2090
    },
    {
      "epoch": 6.349074423876086,
      "grad_norm": 0.7096104025840759,
      "learning_rate": 1.2212521282287092e-05,
      "loss": 0.0858,
      "step": 2100
    },
    {
      "epoch": 6.379297317718171,
      "grad_norm": 0.7978008389472961,
      "learning_rate": 1.178290227708888e-05,
      "loss": 0.1001,
      "step": 2110
    },
    {
      "epoch": 6.409520211560257,
      "grad_norm": 0.8465205430984497,
      "learning_rate": 1.1359964529980849e-05,
      "loss": 0.0864,
      "step": 2120
    },
    {
      "epoch": 6.439743105402342,
      "grad_norm": 0.4713567793369293,
      "learning_rate": 1.0943781980545331e-05,
      "loss": 0.0871,
      "step": 2130
    },
    {
      "epoch": 6.469965999244428,
      "grad_norm": 0.5560480356216431,
      "learning_rate": 1.053442738739539e-05,
      "loss": 0.0801,
      "step": 2140
    },
    {
      "epoch": 6.500188893086513,
      "grad_norm": 0.4392213821411133,
      "learning_rate": 1.0131972315454869e-05,
      "loss": 0.0824,
      "step": 2150
    },
    {
      "epoch": 6.530411786928599,
      "grad_norm": 0.7270606160163879,
      "learning_rate": 9.73648712344707e-06,
      "loss": 0.0895,
      "step": 2160
    },
    {
      "epoch": 6.560634680770684,
      "grad_norm": 0.5097915530204773,
      "learning_rate": 9.348040951594477e-06,
      "loss": 0.0889,
      "step": 2170
    },
    {
      "epoch": 6.5908575746127696,
      "grad_norm": 0.7505512833595276,
      "learning_rate": 8.966701709531344e-06,
      "loss": 0.0889,
      "step": 2180
    },
    {
      "epoch": 6.6210804684548545,
      "grad_norm": 0.5032856464385986,
      "learning_rate": 8.592536064431467e-06,
      "loss": 0.0837,
      "step": 2190
    },
    {
      "epoch": 6.6513033622969395,
      "grad_norm": 0.632739245891571,
      "learning_rate": 8.225609429353187e-06,
      "loss": 0.0909,
      "step": 2200
    },
    {
      "epoch": 6.681526256139025,
      "grad_norm": 0.6564577221870422,
      "learning_rate": 7.865985951803645e-06,
      "loss": 0.0985,
      "step": 2210
    },
    {
      "epoch": 6.71174914998111,
      "grad_norm": 0.5407485961914062,
      "learning_rate": 7.513728502524286e-06,
      "loss": 0.0856,
      "step": 2220
    },
    {
      "epoch": 6.741972043823196,
      "grad_norm": 0.5400388836860657,
      "learning_rate": 7.168898664499485e-06,
      "loss": 0.0848,
      "step": 2230
    },
    {
      "epoch": 6.772194937665281,
      "grad_norm": 0.5073435306549072,
      "learning_rate": 6.831556722190452e-06,
      "loss": 0.088,
      "step": 2240
    },
    {
      "epoch": 6.802417831507367,
      "grad_norm": 0.5684376955032349,
      "learning_rate": 6.501761650996052e-06,
      "loss": 0.0937,
      "step": 2250
    },
    {
      "epoch": 6.832640725349452,
      "grad_norm": 0.6096483469009399,
      "learning_rate": 6.1795711069424666e-06,
      "loss": 0.0823,
      "step": 2260
    },
    {
      "epoch": 6.862863619191538,
      "grad_norm": 0.5201295018196106,
      "learning_rate": 5.865041416603601e-06,
      "loss": 0.092,
      "step": 2270
    },
    {
      "epoch": 6.893086513033623,
      "grad_norm": 0.6869983077049255,
      "learning_rate": 5.558227567253832e-06,
      "loss": 0.093,
      "step": 2280
    },
    {
      "epoch": 6.923309406875708,
      "grad_norm": 0.561122715473175,
      "learning_rate": 5.259183197254902e-06,
      "loss": 0.0802,
      "step": 2290
    },
    {
      "epoch": 6.953532300717794,
      "grad_norm": 0.5263656377792358,
      "learning_rate": 4.967960586678721e-06,
      "loss": 0.0839,
      "step": 2300
    },
    {
      "epoch": 6.983755194559879,
      "grad_norm": 0.575607419013977,
      "learning_rate": 4.684610648167503e-06,
      "loss": 0.0896,
      "step": 2310
    },
    {
      "epoch": 7.01435587457499,
      "grad_norm": 0.3270145654678345,
      "learning_rate": 4.4091829180330505e-06,
      "loss": 0.0778,
      "step": 2320
    },
    {
      "epoch": 7.044578768417076,
      "grad_norm": 0.5017428994178772,
      "learning_rate": 4.1417255475966145e-06,
      "loss": 0.0818,
      "step": 2330
    },
    {
      "epoch": 7.074801662259161,
      "grad_norm": 0.5966259837150574,
      "learning_rate": 3.8822852947709375e-06,
      "loss": 0.0747,
      "step": 2340
    },
    {
      "epoch": 7.105024556101247,
      "grad_norm": 0.758906900882721,
      "learning_rate": 3.630907515885801e-06,
      "loss": 0.0743,
      "step": 2350
    },
    {
      "epoch": 7.135247449943332,
      "grad_norm": 0.564600944519043,
      "learning_rate": 3.3876361577587113e-06,
      "loss": 0.0831,
      "step": 2360
    },
    {
      "epoch": 7.165470343785417,
      "grad_norm": 0.548526406288147,
      "learning_rate": 3.1525137500119207e-06,
      "loss": 0.0753,
      "step": 2370
    },
    {
      "epoch": 7.195693237627503,
      "grad_norm": 0.619285523891449,
      "learning_rate": 2.9255813976372226e-06,
      "loss": 0.0777,
      "step": 2380
    },
    {
      "epoch": 7.225916131469588,
      "grad_norm": 0.5830234289169312,
      "learning_rate": 2.7068787738098735e-06,
      "loss": 0.0803,
      "step": 2390
    },
    {
      "epoch": 7.256139025311674,
      "grad_norm": 0.40243130922317505,
      "learning_rate": 2.496444112952734e-06,
      "loss": 0.063,
      "step": 2400
    },
    {
      "epoch": 7.286361919153759,
      "grad_norm": 0.42115914821624756,
      "learning_rate": 2.2943142040519837e-06,
      "loss": 0.068,
      "step": 2410
    },
    {
      "epoch": 7.316584812995845,
      "grad_norm": 0.7173041105270386,
      "learning_rate": 2.100524384225555e-06,
      "loss": 0.0808,
      "step": 2420
    },
    {
      "epoch": 7.34680770683793,
      "grad_norm": 0.48172470927238464,
      "learning_rate": 1.915108532545351e-06,
      "loss": 0.0756,
      "step": 2430
    },
    {
      "epoch": 7.3770306006800155,
      "grad_norm": 0.541785717010498,
      "learning_rate": 1.738099064114368e-06,
      "loss": 0.0717,
      "step": 2440
    },
    {
      "epoch": 7.4072534945221005,
      "grad_norm": 0.6277313828468323,
      "learning_rate": 1.569526924399778e-06,
      "loss": 0.067,
      "step": 2450
    },
    {
      "epoch": 7.4374763883641855,
      "grad_norm": 1.088949203491211,
      "learning_rate": 1.4094215838229176e-06,
      "loss": 0.0773,
      "step": 2460
    },
    {
      "epoch": 7.467699282206271,
      "grad_norm": 0.7406247854232788,
      "learning_rate": 1.2578110326071702e-06,
      "loss": 0.0695,
      "step": 2470
    },
    {
      "epoch": 7.497922176048356,
      "grad_norm": 0.5518158078193665,
      "learning_rate": 1.1147217758845751e-06,
      "loss": 0.0711,
      "step": 2480
    },
    {
      "epoch": 7.528145069890442,
      "grad_norm": 1.5025277137756348,
      "learning_rate": 9.801788290621505e-07,
      "loss": 0.0749,
      "step": 2490
    },
    {
      "epoch": 7.558367963732527,
      "grad_norm": 0.43806299567222595,
      "learning_rate": 8.542057134485638e-07,
      "loss": 0.0679,
      "step": 2500
    },
    {
      "epoch": 7.588590857574613,
      "grad_norm": 0.604317843914032,
      "learning_rate": 7.368244521420486e-07,
      "loss": 0.0674,
      "step": 2510
    },
    {
      "epoch": 7.618813751416698,
      "grad_norm": 0.5833208560943604,
      "learning_rate": 6.280555661802856e-07,
      "loss": 0.0771,
      "step": 2520
    },
    {
      "epoch": 7.649036645258784,
      "grad_norm": 0.6899517178535461,
      "learning_rate": 5.279180709527765e-07,
      "loss": 0.072,
      "step": 2530
    },
    {
      "epoch": 7.679259539100869,
      "grad_norm": 0.5004780888557434,
      "learning_rate": 4.3642947287654277e-07,
      "loss": 0.0702,
      "step": 2540
    },
    {
      "epoch": 7.709482432942954,
      "grad_norm": 1.1061525344848633,
      "learning_rate": 3.536057663355852e-07,
      "loss": 0.0726,
      "step": 2550
    },
    {
      "epoch": 7.73970532678504,
      "grad_norm": 1.5561116933822632,
      "learning_rate": 2.794614308846644e-07,
      "loss": 0.0755,
      "step": 2560
    },
    {
      "epoch": 7.769928220627125,
      "grad_norm": 0.4418126940727234,
      "learning_rate": 2.1400942871794283e-07,
      "loss": 0.0783,
      "step": 2570
    },
    {
      "epoch": 7.800151114469211,
      "grad_norm": 0.6038776636123657,
      "learning_rate": 1.5726120240288634e-07,
      "loss": 0.0769,
      "step": 2580
    },
    {
      "epoch": 7.830374008311296,
      "grad_norm": 0.6262177228927612,
      "learning_rate": 1.0922667287981969e-07,
      "loss": 0.0748,
      "step": 2590
    },
    {
      "epoch": 7.8605969021533815,
      "grad_norm": 0.8348723649978638,
      "learning_rate": 6.991423772753636e-08,
      "loss": 0.0749,
      "step": 2600
    },
    {
      "epoch": 7.8908197959954665,
      "grad_norm": 0.4923211634159088,
      "learning_rate": 3.9330769695167245e-08,
      "loss": 0.0744,
      "step": 2610
    },
    {
      "epoch": 7.921042689837552,
      "grad_norm": 0.5318350791931152,
      "learning_rate": 1.748161550069183e-08,
      "loss": 0.0731,
      "step": 2620
    },
    {
      "epoch": 7.951265583679637,
      "grad_norm": 0.756984293460846,
      "learning_rate": 4.3705948961692e-09,
      "loss": 0.0716,
      "step": 2630
    },
    {
      "epoch": 7.981488477521722,
      "grad_norm": 0.797049880027771,
      "learning_rate": 0.0,
      "loss": 0.0746,
      "step": 2640
    }
  ],
  "logging_steps": 10,
  "max_steps": 2640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 4.018451794664358e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}