{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 2690,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03717472118959108,
      "grad_norm": 4.877419471740723,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.8539,
      "step": 10
    },
    {
      "epoch": 0.07434944237918216,
      "grad_norm": 5.054717540740967,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.7058,
      "step": 20
    },
    {
      "epoch": 0.11152416356877323,
      "grad_norm": 2.2075462341308594,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.4233,
      "step": 30
    },
    {
      "epoch": 0.14869888475836432,
      "grad_norm": 2.1013667583465576,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.3267,
      "step": 40
    },
    {
      "epoch": 0.18587360594795538,
      "grad_norm": 2.8093650341033936,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.2644,
      "step": 50
    },
    {
      "epoch": 0.22304832713754646,
      "grad_norm": 3.7132174968719482,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.2234,
      "step": 60
    },
    {
      "epoch": 0.26022304832713755,
      "grad_norm": 31.805850982666016,
      "learning_rate": 5.185185185185185e-05,
      "loss": 0.2112,
      "step": 70
    },
    {
      "epoch": 0.29739776951672864,
      "grad_norm": 2.9102275371551514,
      "learning_rate": 5.925925925925926e-05,
      "loss": 0.1907,
      "step": 80
    },
    {
      "epoch": 0.3345724907063197,
      "grad_norm": 2.2478322982788086,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.1921,
      "step": 90
    },
    {
      "epoch": 0.37174721189591076,
      "grad_norm": 1.355301856994629,
      "learning_rate": 7.407407407407407e-05,
      "loss": 0.1638,
      "step": 100
    },
    {
      "epoch": 0.40892193308550184,
      "grad_norm": 1.8597526550292969,
      "learning_rate": 8.148148148148148e-05,
      "loss": 0.1569,
      "step": 110
    },
    {
      "epoch": 0.44609665427509293,
      "grad_norm": 1.6024093627929688,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.1676,
      "step": 120
    },
    {
      "epoch": 0.483271375464684,
      "grad_norm": 1.8739852905273438,
      "learning_rate": 9.62962962962963e-05,
      "loss": 0.1504,
      "step": 130
    },
    {
      "epoch": 0.5204460966542751,
      "grad_norm": 1.0192468166351318,
      "learning_rate": 9.999905507663936e-05,
      "loss": 0.136,
      "step": 140
    },
    {
      "epoch": 0.5576208178438662,
      "grad_norm": 1.430972933769226,
      "learning_rate": 9.9991495904044e-05,
      "loss": 0.1305,
      "step": 150
    },
    {
      "epoch": 0.5947955390334573,
      "grad_norm": 1.3280889987945557,
      "learning_rate": 9.997637870169672e-05,
      "loss": 0.1248,
      "step": 160
    },
    {
      "epoch": 0.6319702602230484,
      "grad_norm": 1.390118956565857,
      "learning_rate": 9.995370575511151e-05,
      "loss": 0.1359,
      "step": 170
    },
    {
      "epoch": 0.6691449814126395,
      "grad_norm": 2.348705530166626,
      "learning_rate": 9.99234804921275e-05,
      "loss": 0.1252,
      "step": 180
    },
    {
      "epoch": 0.7063197026022305,
      "grad_norm": 0.7154766917228699,
      "learning_rate": 9.988570748239062e-05,
      "loss": 0.1143,
      "step": 190
    },
    {
      "epoch": 0.7434944237918215,
      "grad_norm": 0.9727993011474609,
      "learning_rate": 9.984039243666283e-05,
      "loss": 0.1045,
      "step": 200
    },
    {
      "epoch": 0.7806691449814126,
      "grad_norm": 1.0658169984817505,
      "learning_rate": 9.978754220595861e-05,
      "loss": 0.114,
      "step": 210
    },
    {
      "epoch": 0.8178438661710037,
      "grad_norm": 0.9481009840965271,
      "learning_rate": 9.97271647805093e-05,
      "loss": 0.1051,
      "step": 220
    },
    {
      "epoch": 0.8550185873605948,
      "grad_norm": 0.8618531227111816,
      "learning_rate": 9.965926928855499e-05,
      "loss": 0.0996,
      "step": 230
    },
    {
      "epoch": 0.8921933085501859,
      "grad_norm": 0.6774977445602417,
      "learning_rate": 9.95838659949645e-05,
      "loss": 0.102,
      "step": 240
    },
    {
      "epoch": 0.929368029739777,
      "grad_norm": 0.7148956656455994,
      "learning_rate": 9.950096629968352e-05,
      "loss": 0.1072,
      "step": 250
    },
    {
      "epoch": 0.966542750929368,
      "grad_norm": 0.9607800245285034,
      "learning_rate": 9.941058273601096e-05,
      "loss": 0.1025,
      "step": 260
    },
    {
      "epoch": 1.003717472118959,
      "grad_norm": 0.6112203001976013,
      "learning_rate": 9.931272896870426e-05,
      "loss": 0.0927,
      "step": 270
    },
    {
      "epoch": 1.0408921933085502,
      "grad_norm": 0.9129474759101868,
      "learning_rate": 9.920741979191331e-05,
      "loss": 0.0905,
      "step": 280
    },
    {
      "epoch": 1.0780669144981412,
      "grad_norm": 0.7284883856773376,
      "learning_rate": 9.909467112694384e-05,
      "loss": 0.0924,
      "step": 290
    },
    {
      "epoch": 1.1152416356877324,
      "grad_norm": 0.8254210948944092,
      "learning_rate": 9.897450001985039e-05,
      "loss": 0.0901,
      "step": 300
    },
    {
      "epoch": 1.1524163568773234,
      "grad_norm": 0.9527810215950012,
      "learning_rate": 9.88469246388591e-05,
      "loss": 0.0862,
      "step": 310
    },
    {
      "epoch": 1.1895910780669146,
      "grad_norm": 0.9555880427360535,
      "learning_rate": 9.871196427162092e-05,
      "loss": 0.0951,
      "step": 320
    },
    {
      "epoch": 1.2267657992565055,
      "grad_norm": 0.511620044708252,
      "learning_rate": 9.85696393222957e-05,
      "loss": 0.0939,
      "step": 330
    },
    {
      "epoch": 1.2639405204460967,
      "grad_norm": 1.2294408082962036,
      "learning_rate": 9.84199713084672e-05,
      "loss": 0.0862,
      "step": 340
    },
    {
      "epoch": 1.3011152416356877,
      "grad_norm": 1.0503334999084473,
      "learning_rate": 9.826298285789002e-05,
      "loss": 0.0914,
      "step": 350
    },
    {
      "epoch": 1.3382899628252787,
      "grad_norm": 0.6006482839584351,
      "learning_rate": 9.809869770506856e-05,
      "loss": 0.0897,
      "step": 360
    },
    {
      "epoch": 1.3754646840148699,
      "grad_norm": 1.2222704887390137,
      "learning_rate": 9.792714068766872e-05,
      "loss": 0.0919,
      "step": 370
    },
    {
      "epoch": 1.412639405204461,
      "grad_norm": 0.9911707043647766,
      "learning_rate": 9.774833774276278e-05,
      "loss": 0.0993,
      "step": 380
    },
    {
      "epoch": 1.449814126394052,
      "grad_norm": 0.6083410978317261,
      "learning_rate": 9.7562315902908e-05,
      "loss": 0.0854,
      "step": 390
    },
    {
      "epoch": 1.486988847583643,
      "grad_norm": 0.9849829077720642,
      "learning_rate": 9.736910329205978e-05,
      "loss": 0.0826,
      "step": 400
    },
    {
      "epoch": 1.5241635687732342,
      "grad_norm": 0.4323747158050537,
      "learning_rate": 9.716872912131964e-05,
      "loss": 0.0866,
      "step": 410
    },
    {
      "epoch": 1.5613382899628254,
      "grad_norm": 0.8285312056541443,
      "learning_rate": 9.696122368451886e-05,
      "loss": 0.082,
      "step": 420
    },
    {
      "epoch": 1.5985130111524164,
      "grad_norm": 1.4286469221115112,
      "learning_rate": 9.674661835363858e-05,
      "loss": 0.0818,
      "step": 430
    },
    {
      "epoch": 1.6356877323420074,
      "grad_norm": 0.5847893953323364,
      "learning_rate": 9.652494557406666e-05,
      "loss": 0.0735,
      "step": 440
    },
    {
      "epoch": 1.6728624535315983,
      "grad_norm": 0.7277955412864685,
      "learning_rate": 9.62962388596925e-05,
      "loss": 0.0752,
      "step": 450
    },
    {
      "epoch": 1.7100371747211895,
      "grad_norm": 0.5724025368690491,
      "learning_rate": 9.606053278784009e-05,
      "loss": 0.0689,
      "step": 460
    },
    {
      "epoch": 1.7472118959107807,
      "grad_norm": 0.49029338359832764,
      "learning_rate": 9.581786299404045e-05,
      "loss": 0.0726,
      "step": 470
    },
    {
      "epoch": 1.7843866171003717,
      "grad_norm": 0.5488936305046082,
      "learning_rate": 9.556826616664407e-05,
      "loss": 0.0748,
      "step": 480
    },
    {
      "epoch": 1.8215613382899627,
      "grad_norm": 0.7589528560638428,
      "learning_rate": 9.531178004127403e-05,
      "loss": 0.0838,
      "step": 490
    },
    {
      "epoch": 1.858736059479554,
      "grad_norm": 0.4553124010562897,
      "learning_rate": 9.504844339512095e-05,
      "loss": 0.0714,
      "step": 500
    },
    {
      "epoch": 1.895910780669145,
      "grad_norm": 0.4341762959957123,
      "learning_rate": 9.477829604108044e-05,
      "loss": 0.0759,
      "step": 510
    },
    {
      "epoch": 1.933085501858736,
      "grad_norm": 0.4884495139122009,
      "learning_rate": 9.450137882173384e-05,
      "loss": 0.0708,
      "step": 520
    },
    {
      "epoch": 1.970260223048327,
      "grad_norm": 0.8140550851821899,
      "learning_rate": 9.421773360317347e-05,
      "loss": 0.0705,
      "step": 530
    },
    {
      "epoch": 2.007434944237918,
      "grad_norm": 0.5736023187637329,
      "learning_rate": 9.392740326867304e-05,
      "loss": 0.0658,
      "step": 540
    },
    {
      "epoch": 2.0446096654275094,
      "grad_norm": 0.6355158686637878,
      "learning_rate": 9.363043171220423e-05,
      "loss": 0.082,
      "step": 550
    },
    {
      "epoch": 2.0817843866171004,
      "grad_norm": 0.6205140948295593,
      "learning_rate": 9.332686383180055e-05,
      "loss": 0.0695,
      "step": 560
    },
    {
      "epoch": 2.1189591078066914,
      "grad_norm": 0.8051413297653198,
      "learning_rate": 9.301674552276942e-05,
      "loss": 0.0683,
      "step": 570
    },
    {
      "epoch": 2.1561338289962824,
      "grad_norm": 0.5303524732589722,
      "learning_rate": 9.270012367075336e-05,
      "loss": 0.0656,
      "step": 580
    },
    {
      "epoch": 2.193308550185874,
      "grad_norm": 0.6141722798347473,
      "learning_rate": 9.237704614464156e-05,
      "loss": 0.0717,
      "step": 590
    },
    {
      "epoch": 2.2304832713754648,
      "grad_norm": 0.645879328250885,
      "learning_rate": 9.204756178933274e-05,
      "loss": 0.0742,
      "step": 600
    },
    {
      "epoch": 2.2676579925650557,
      "grad_norm": 0.49008268117904663,
      "learning_rate": 9.17117204183505e-05,
      "loss": 0.0681,
      "step": 610
    },
    {
      "epoch": 2.3048327137546467,
      "grad_norm": 0.6824775338172913,
      "learning_rate": 9.136957280631212e-05,
      "loss": 0.0651,
      "step": 620
    },
    {
      "epoch": 2.3420074349442377,
      "grad_norm": 0.47334712743759155,
      "learning_rate": 9.102117068125226e-05,
      "loss": 0.0778,
      "step": 630
    },
    {
      "epoch": 2.379182156133829,
      "grad_norm": 0.6936899423599243,
      "learning_rate": 9.06665667168023e-05,
      "loss": 0.0682,
      "step": 640
    },
    {
      "epoch": 2.41635687732342,
      "grad_norm": 0.5970759391784668,
      "learning_rate": 9.030581452422679e-05,
      "loss": 0.0632,
      "step": 650
    },
    {
      "epoch": 2.453531598513011,
      "grad_norm": 0.7608362436294556,
      "learning_rate": 8.993896864431826e-05,
      "loss": 0.0624,
      "step": 660
    },
    {
      "epoch": 2.4907063197026025,
      "grad_norm": 0.5167849659919739,
      "learning_rate": 8.956608453915125e-05,
      "loss": 0.0624,
      "step": 670
    },
    {
      "epoch": 2.5278810408921935,
      "grad_norm": 0.48926475644111633,
      "learning_rate": 8.918721858369738e-05,
      "loss": 0.0606,
      "step": 680
    },
    {
      "epoch": 2.5650557620817844,
      "grad_norm": 0.7994228005409241,
      "learning_rate": 8.880242805730208e-05,
      "loss": 0.0627,
      "step": 690
    },
    {
      "epoch": 2.6022304832713754,
      "grad_norm": 0.5655274391174316,
      "learning_rate": 8.841177113502482e-05,
      "loss": 0.0756,
      "step": 700
    },
    {
      "epoch": 2.6394052044609664,
      "grad_norm": 0.5700337290763855,
      "learning_rate": 8.801530687884378e-05,
      "loss": 0.0664,
      "step": 710
    },
    {
      "epoch": 2.6765799256505574,
      "grad_norm": 0.920259952545166,
      "learning_rate": 8.761309522872657e-05,
      "loss": 0.0756,
      "step": 720
    },
    {
      "epoch": 2.7137546468401488,
      "grad_norm": 0.6414476633071899,
      "learning_rate": 8.720519699356804e-05,
      "loss": 0.0636,
      "step": 730
    },
    {
      "epoch": 2.7509293680297398,
      "grad_norm": 0.48423320055007935,
      "learning_rate": 8.679167384199684e-05,
      "loss": 0.0596,
      "step": 740
    },
    {
      "epoch": 2.7881040892193307,
      "grad_norm": 0.8158369660377502,
      "learning_rate": 8.637258829305199e-05,
      "loss": 0.0568,
      "step": 750
    },
    {
      "epoch": 2.825278810408922,
      "grad_norm": 0.41953688859939575,
      "learning_rate": 8.594800370673083e-05,
      "loss": 0.0615,
      "step": 760
    },
    {
      "epoch": 2.862453531598513,
      "grad_norm": 0.5318211317062378,
      "learning_rate": 8.551798427440986e-05,
      "loss": 0.0633,
      "step": 770
    },
    {
      "epoch": 2.899628252788104,
      "grad_norm": 0.49588924646377563,
      "learning_rate": 8.50825950091399e-05,
      "loss": 0.0563,
      "step": 780
    },
    {
      "epoch": 2.936802973977695,
      "grad_norm": 0.5286570191383362,
      "learning_rate": 8.464190173581699e-05,
      "loss": 0.0663,
      "step": 790
    },
    {
      "epoch": 2.973977695167286,
      "grad_norm": 0.5693170428276062,
      "learning_rate": 8.419597108123054e-05,
      "loss": 0.0727,
      "step": 800
    },
    {
      "epoch": 3.0111524163568775,
      "grad_norm": 0.4871514141559601,
      "learning_rate": 8.374487046399036e-05,
      "loss": 0.0678,
      "step": 810
    },
    {
      "epoch": 3.0483271375464684,
      "grad_norm": 0.4753323197364807,
      "learning_rate": 8.328866808433378e-05,
      "loss": 0.0554,
      "step": 820
    },
    {
      "epoch": 3.0855018587360594,
      "grad_norm": 0.6938982009887695,
      "learning_rate": 8.282743291381481e-05,
      "loss": 0.0597,
      "step": 830
    },
    {
      "epoch": 3.1226765799256504,
      "grad_norm": 0.4666888415813446,
      "learning_rate": 8.236123468487648e-05,
      "loss": 0.0625,
      "step": 840
    },
    {
      "epoch": 3.159851301115242,
      "grad_norm": 0.6246622800827026,
      "learning_rate": 8.189014388030833e-05,
      "loss": 0.0687,
      "step": 850
    },
    {
      "epoch": 3.197026022304833,
      "grad_norm": 0.46097248792648315,
      "learning_rate": 8.141423172259038e-05,
      "loss": 0.0601,
      "step": 860
    },
    {
      "epoch": 3.2342007434944238,
      "grad_norm": 0.698952317237854,
      "learning_rate": 8.093357016312517e-05,
      "loss": 0.0572,
      "step": 870
    },
    {
      "epoch": 3.2713754646840147,
      "grad_norm": 0.39901071786880493,
      "learning_rate": 8.044823187135984e-05,
      "loss": 0.0577,
      "step": 880
    },
    {
      "epoch": 3.3085501858736057,
      "grad_norm": 0.4128256142139435,
      "learning_rate": 7.995829022379936e-05,
      "loss": 0.0669,
      "step": 890
    },
    {
      "epoch": 3.345724907063197,
      "grad_norm": 0.3928859829902649,
      "learning_rate": 7.94638192929131e-05,
      "loss": 0.0514,
      "step": 900
    },
    {
      "epoch": 3.382899628252788,
      "grad_norm": 0.42870765924453735,
      "learning_rate": 7.896489383593606e-05,
      "loss": 0.0591,
      "step": 910
    },
    {
      "epoch": 3.420074349442379,
      "grad_norm": 0.5093951225280762,
      "learning_rate": 7.84615892835666e-05,
      "loss": 0.0599,
      "step": 920
    },
    {
      "epoch": 3.45724907063197,
      "grad_norm": 0.9360804557800293,
      "learning_rate": 7.795398172856233e-05,
      "loss": 0.0541,
      "step": 930
    },
    {
      "epoch": 3.4944237918215615,
      "grad_norm": 0.6116214394569397,
      "learning_rate": 7.744214791423596e-05,
      "loss": 0.0572,
      "step": 940
    },
    {
      "epoch": 3.5315985130111525,
      "grad_norm": 0.6456897854804993,
      "learning_rate": 7.692616522285278e-05,
      "loss": 0.0653,
      "step": 950
    },
    {
      "epoch": 3.5687732342007434,
      "grad_norm": 0.5947780013084412,
      "learning_rate": 7.640611166393141e-05,
      "loss": 0.0627,
      "step": 960
    },
    {
      "epoch": 3.6059479553903344,
      "grad_norm": 1.3824375867843628,
      "learning_rate": 7.588206586245002e-05,
      "loss": 0.061,
      "step": 970
    },
    {
      "epoch": 3.6431226765799254,
      "grad_norm": 0.5360904932022095,
      "learning_rate": 7.535410704695906e-05,
      "loss": 0.049,
      "step": 980
    },
    {
      "epoch": 3.680297397769517,
      "grad_norm": 0.47667422890663147,
      "learning_rate": 7.482231503760325e-05,
      "loss": 0.0674,
      "step": 990
    },
    {
      "epoch": 3.717472118959108,
      "grad_norm": 0.9161400198936462,
      "learning_rate": 7.428677023405366e-05,
      "loss": 0.0567,
      "step": 1000
    },
    {
      "epoch": 3.7546468401486988,
      "grad_norm": 0.5789083242416382,
      "learning_rate": 7.374755360335253e-05,
      "loss": 0.0676,
      "step": 1010
    },
    {
      "epoch": 3.79182156133829,
      "grad_norm": 1.225460171699524,
      "learning_rate": 7.320474666767201e-05,
      "loss": 0.0568,
      "step": 1020
    },
    {
      "epoch": 3.828996282527881,
      "grad_norm": 0.6838318705558777,
      "learning_rate": 7.265843149198931e-05,
      "loss": 0.0577,
      "step": 1030
    },
    {
      "epoch": 3.866171003717472,
      "grad_norm": 0.6001195311546326,
      "learning_rate": 7.210869067167942e-05,
      "loss": 0.0575,
      "step": 1040
    },
    {
      "epoch": 3.903345724907063,
      "grad_norm": 0.3830925524234772,
      "learning_rate": 7.155560732002791e-05,
      "loss": 0.0536,
      "step": 1050
    },
    {
      "epoch": 3.940520446096654,
      "grad_norm": 0.41186103224754333,
      "learning_rate": 7.099926505566537e-05,
      "loss": 0.0584,
      "step": 1060
    },
    {
      "epoch": 3.9776951672862455,
      "grad_norm": 0.42513903975486755,
      "learning_rate": 7.043974798992532e-05,
      "loss": 0.0611,
      "step": 1070
    },
    {
      "epoch": 4.014869888475836,
      "grad_norm": 0.3477434515953064,
      "learning_rate": 6.98771407141278e-05,
      "loss": 0.0481,
      "step": 1080
    },
    {
      "epoch": 4.052044609665428,
      "grad_norm": 0.5160051584243774,
      "learning_rate": 6.931152828679033e-05,
      "loss": 0.0535,
      "step": 1090
    },
    {
      "epoch": 4.089219330855019,
      "grad_norm": 0.6074197888374329,
      "learning_rate": 6.874299622076816e-05,
      "loss": 0.0505,
      "step": 1100
    },
    {
      "epoch": 4.12639405204461,
      "grad_norm": 0.6713132262229919,
      "learning_rate": 6.817163047032598e-05,
      "loss": 0.0533,
      "step": 1110
    },
    {
      "epoch": 4.163568773234201,
      "grad_norm": 0.37607887387275696,
      "learning_rate": 6.759751741814271e-05,
      "loss": 0.0537,
      "step": 1120
    },
    {
      "epoch": 4.200743494423792,
      "grad_norm": 0.6844078302383423,
      "learning_rate": 6.702074386225175e-05,
      "loss": 0.061,
      "step": 1130
    },
    {
      "epoch": 4.237918215613383,
      "grad_norm": 0.46288058161735535,
      "learning_rate": 6.644139700291817e-05,
      "loss": 0.0432,
      "step": 1140
    },
    {
      "epoch": 4.275092936802974,
      "grad_norm": 0.35319188237190247,
      "learning_rate": 6.585956442945532e-05,
      "loss": 0.0499,
      "step": 1150
    },
    {
      "epoch": 4.312267657992565,
      "grad_norm": 0.47214651107788086,
      "learning_rate": 6.52753341069825e-05,
      "loss": 0.0537,
      "step": 1160
    },
    {
      "epoch": 4.349442379182156,
      "grad_norm": 0.3913021385669708,
      "learning_rate": 6.468879436312584e-05,
      "loss": 0.0549,
      "step": 1170
    },
    {
      "epoch": 4.386617100371748,
      "grad_norm": 0.5337435603141785,
      "learning_rate": 6.410003387466433e-05,
      "loss": 0.0497,
      "step": 1180
    },
    {
      "epoch": 4.4237918215613385,
      "grad_norm": 0.5121909976005554,
      "learning_rate": 6.35091416541232e-05,
      "loss": 0.0497,
      "step": 1190
    },
    {
      "epoch": 4.4609665427509295,
      "grad_norm": 0.5067112445831299,
      "learning_rate": 6.29162070363163e-05,
      "loss": 0.0504,
      "step": 1200
    },
    {
      "epoch": 4.4981412639405205,
      "grad_norm": 0.48435378074645996,
      "learning_rate": 6.232131966484006e-05,
      "loss": 0.0504,
      "step": 1210
    },
    {
      "epoch": 4.5353159851301115,
      "grad_norm": 0.4863722324371338,
      "learning_rate": 6.172456947852049e-05,
      "loss": 0.0494,
      "step": 1220
    },
    {
      "epoch": 4.5724907063197024,
      "grad_norm": 0.4769513010978699,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.0616,
      "step": 1230
    },
    {
      "epoch": 4.609665427509293,
      "grad_norm": 0.4153217077255249,
      "learning_rate": 6.052584181117589e-05,
      "loss": 0.0514,
      "step": 1240
    },
    {
      "epoch": 4.646840148698884,
      "grad_norm": 0.35905030369758606,
      "learning_rate": 5.9924045561362474e-05,
      "loss": 0.0482,
      "step": 1250
    },
    {
      "epoch": 4.684014869888475,
      "grad_norm": 0.35207000374794006,
      "learning_rate": 5.9320748931729344e-05,
      "loss": 0.0652,
      "step": 1260
    },
    {
      "epoch": 4.721189591078067,
      "grad_norm": 0.461542546749115,
      "learning_rate": 5.87160431324672e-05,
      "loss": 0.0461,
      "step": 1270
    },
    {
      "epoch": 4.758364312267658,
      "grad_norm": 0.47476083040237427,
      "learning_rate": 5.8110019586813946e-05,
      "loss": 0.0526,
      "step": 1280
    },
    {
      "epoch": 4.795539033457249,
      "grad_norm": 0.5205811262130737,
      "learning_rate": 5.7502769917232635e-05,
      "loss": 0.0474,
      "step": 1290
    },
    {
      "epoch": 4.83271375464684,
      "grad_norm": 0.36620450019836426,
      "learning_rate": 5.689438593155956e-05,
      "loss": 0.0405,
      "step": 1300
    },
    {
      "epoch": 4.869888475836431,
      "grad_norm": 0.4866809546947479,
      "learning_rate": 5.6284959609124e-05,
      "loss": 0.0407,
      "step": 1310
    },
    {
      "epoch": 4.907063197026022,
      "grad_norm": 0.4499811828136444,
      "learning_rate": 5.567458308684232e-05,
      "loss": 0.0555,
      "step": 1320
    },
    {
      "epoch": 4.944237918215613,
      "grad_norm": 0.49062612652778625,
      "learning_rate": 5.506334864528808e-05,
      "loss": 0.0502,
      "step": 1330
    },
    {
      "epoch": 4.981412639405205,
      "grad_norm": 0.28027549386024475,
      "learning_rate": 5.445134869474049e-05,
      "loss": 0.0538,
      "step": 1340
    },
    {
      "epoch": 5.018587360594796,
      "grad_norm": 0.4135720431804657,
      "learning_rate": 5.3838675761213244e-05,
      "loss": 0.0541,
      "step": 1350
    },
    {
      "epoch": 5.055762081784387,
      "grad_norm": 0.9072251319885254,
      "learning_rate": 5.3225422472465824e-05,
      "loss": 0.0553,
      "step": 1360
    },
    {
      "epoch": 5.092936802973978,
      "grad_norm": 0.3564969599246979,
      "learning_rate": 5.261168154399952e-05,
      "loss": 0.0507,
      "step": 1370
    },
    {
      "epoch": 5.130111524163569,
      "grad_norm": 0.47934380173683167,
      "learning_rate": 5.199754576504006e-05,
      "loss": 0.0529,
      "step": 1380
    },
    {
      "epoch": 5.16728624535316,
      "grad_norm": 0.5730791687965393,
      "learning_rate": 5.138310798450912e-05,
      "loss": 0.0508,
      "step": 1390
    },
    {
      "epoch": 5.204460966542751,
      "grad_norm": 0.4801822006702423,
      "learning_rate": 5.076846109698693e-05,
      "loss": 0.0484,
      "step": 1400
    },
    {
      "epoch": 5.241635687732342,
      "grad_norm": 0.5102419853210449,
      "learning_rate": 5.01536980286678e-05,
      "loss": 0.0479,
      "step": 1410
    },
    {
      "epoch": 5.278810408921933,
      "grad_norm": 0.4119831323623657,
      "learning_rate": 4.9538911723310976e-05,
      "loss": 0.0513,
      "step": 1420
    },
    {
      "epoch": 5.315985130111525,
      "grad_norm": 0.5217913389205933,
      "learning_rate": 4.892419512818889e-05,
      "loss": 0.046,
      "step": 1430
    },
    {
      "epoch": 5.353159851301116,
      "grad_norm": 0.5952945351600647,
      "learning_rate": 4.830964118003468e-05,
      "loss": 0.0461,
      "step": 1440
    },
    {
      "epoch": 5.390334572490707,
      "grad_norm": 0.412113755941391,
      "learning_rate": 4.769534279099152e-05,
      "loss": 0.0492,
      "step": 1450
    },
    {
      "epoch": 5.4275092936802976,
      "grad_norm": 0.4845413565635681,
      "learning_rate": 4.708139283456551e-05,
      "loss": 0.049,
      "step": 1460
    },
    {
      "epoch": 5.4646840148698885,
      "grad_norm": 0.33267199993133545,
      "learning_rate": 4.646788413158455e-05,
      "loss": 0.0466,
      "step": 1470
    },
    {
      "epoch": 5.5018587360594795,
      "grad_norm": 0.35786089301109314,
      "learning_rate": 4.585490943616504e-05,
      "loss": 0.0439,
      "step": 1480
    },
    {
      "epoch": 5.5390334572490705,
      "grad_norm": 0.6732335686683655,
      "learning_rate": 4.524256142168874e-05,
      "loss": 0.0529,
      "step": 1490
    },
    {
      "epoch": 5.5762081784386615,
      "grad_norm": 0.5137932896614075,
      "learning_rate": 4.463093266679185e-05,
      "loss": 0.0415,
      "step": 1500
    },
    {
      "epoch": 5.613382899628252,
      "grad_norm": 0.3910720646381378,
      "learning_rate": 4.402011564136836e-05,
      "loss": 0.0523,
      "step": 1510
    },
    {
      "epoch": 5.650557620817844,
      "grad_norm": 0.5063146948814392,
      "learning_rate": 4.341020269258987e-05,
      "loss": 0.0564,
      "step": 1520
    },
    {
      "epoch": 5.687732342007435,
      "grad_norm": 0.5695778131484985,
      "learning_rate": 4.2801286030943985e-05,
      "loss": 0.0551,
      "step": 1530
    },
    {
      "epoch": 5.724907063197026,
      "grad_norm": 0.5160698294639587,
      "learning_rate": 4.219345771629333e-05,
      "loss": 0.0468,
      "step": 1540
    },
    {
      "epoch": 5.762081784386617,
      "grad_norm": 0.38890841603279114,
      "learning_rate": 4.1586809643957345e-05,
      "loss": 0.042,
      "step": 1550
    },
    {
      "epoch": 5.799256505576208,
      "grad_norm": 0.6440842747688293,
      "learning_rate": 4.098143353081902e-05,
      "loss": 0.0448,
      "step": 1560
    },
    {
      "epoch": 5.836431226765799,
      "grad_norm": 0.37773260474205017,
      "learning_rate": 4.0377420901458506e-05,
      "loss": 0.0487,
      "step": 1570
    },
    {
      "epoch": 5.87360594795539,
      "grad_norm": 0.44040870666503906,
      "learning_rate": 3.977486307431589e-05,
      "loss": 0.0433,
      "step": 1580
    },
    {
      "epoch": 5.910780669144981,
      "grad_norm": 0.36806249618530273,
      "learning_rate": 3.9173851147885075e-05,
      "loss": 0.0398,
      "step": 1590
    },
    {
      "epoch": 5.947955390334572,
      "grad_norm": 0.49731066823005676,
      "learning_rate": 3.857447598694094e-05,
      "loss": 0.0396,
      "step": 1600
    },
    {
      "epoch": 5.985130111524164,
      "grad_norm": 0.5180796384811401,
      "learning_rate": 3.797682820880184e-05,
      "loss": 0.0461,
      "step": 1610
    },
    {
      "epoch": 6.022304832713755,
      "grad_norm": 0.34899625182151794,
      "learning_rate": 3.7380998169629476e-05,
      "loss": 0.0433,
      "step": 1620
    },
    {
      "epoch": 6.059479553903346,
      "grad_norm": 0.32863014936447144,
      "learning_rate": 3.678707595076834e-05,
      "loss": 0.0396,
      "step": 1630
    },
    {
      "epoch": 6.096654275092937,
      "grad_norm": 0.5975285768508911,
      "learning_rate": 3.619515134512656e-05,
      "loss": 0.0418,
      "step": 1640
    },
    {
      "epoch": 6.133828996282528,
      "grad_norm": 0.6041769981384277,
      "learning_rate": 3.560531384360055e-05,
      "loss": 0.0373,
      "step": 1650
    },
    {
      "epoch": 6.171003717472119,
      "grad_norm": 0.5462443828582764,
      "learning_rate": 3.501765262154513e-05,
      "loss": 0.0455,
      "step": 1660
    },
    {
      "epoch": 6.20817843866171,
      "grad_norm": 0.4584652781486511,
      "learning_rate": 3.4432256525291465e-05,
      "loss": 0.0424,
      "step": 1670
    },
    {
      "epoch": 6.245353159851301,
      "grad_norm": 1.1305296421051025,
      "learning_rate": 3.3849214058714704e-05,
      "loss": 0.0463,
      "step": 1680
    },
    {
      "epoch": 6.282527881040892,
      "grad_norm": 0.4728774428367615,
      "learning_rate": 3.326861336985341e-05,
      "loss": 0.0422,
      "step": 1690
    },
    {
      "epoch": 6.319702602230484,
      "grad_norm": 0.347769558429718,
      "learning_rate": 3.269054223758279e-05,
      "loss": 0.042,
      "step": 1700
    },
    {
      "epoch": 6.356877323420075,
      "grad_norm": 0.4270273447036743,
      "learning_rate": 3.2115088058343725e-05,
      "loss": 0.046,
      "step": 1710
    },
    {
      "epoch": 6.394052044609666,
      "grad_norm": 0.46025773882865906,
      "learning_rate": 3.1542337832929644e-05,
      "loss": 0.0452,
      "step": 1720
    },
    {
      "epoch": 6.431226765799257,
      "grad_norm": 0.5107239484786987,
      "learning_rate": 3.0972378153333145e-05,
      "loss": 0.0416,
      "step": 1730
    },
    {
      "epoch": 6.4684014869888475,
      "grad_norm": 0.309860497713089,
      "learning_rate": 3.0405295189654537e-05,
      "loss": 0.0394,
      "step": 1740
    },
    {
      "epoch": 6.5055762081784385,
      "grad_norm": 0.33350470662117004,
      "learning_rate": 2.9841174677074035e-05,
      "loss": 0.043,
      "step": 1750
    },
    {
      "epoch": 6.5427509293680295,
      "grad_norm": 0.42869654297828674,
      "learning_rate": 2.9280101902889824e-05,
      "loss": 0.0446,
      "step": 1760
    },
    {
      "epoch": 6.5799256505576205,
      "grad_norm": 0.2736971974372864,
      "learning_rate": 2.8722161693623772e-05,
      "loss": 0.0393,
      "step": 1770
    },
    {
      "epoch": 6.617100371747211,
      "grad_norm": 0.4107191562652588,
      "learning_rate": 2.8167438402196805e-05,
      "loss": 0.0412,
      "step": 1780
    },
    {
      "epoch": 6.654275092936803,
      "grad_norm": 0.5272426009178162,
      "learning_rate": 2.761601589517595e-05,
      "loss": 0.0502,
      "step": 1790
    },
    {
      "epoch": 6.691449814126394,
      "grad_norm": 0.48449358344078064,
      "learning_rate": 2.706797754009476e-05,
      "loss": 0.0416,
      "step": 1800
    },
    {
      "epoch": 6.728624535315985,
      "grad_norm": 0.46288979053497314,
      "learning_rate": 2.6523406192849386e-05,
      "loss": 0.0431,
      "step": 1810
    },
    {
      "epoch": 6.765799256505576,
      "grad_norm": 0.37995341420173645,
      "learning_rate": 2.5982384185171906e-05,
      "loss": 0.0365,
      "step": 1820
    },
    {
      "epoch": 6.802973977695167,
      "grad_norm": 0.3601038157939911,
      "learning_rate": 2.544499331218274e-05,
      "loss": 0.0523,
      "step": 1830
    },
    {
      "epoch": 6.840148698884758,
      "grad_norm": 0.3422638177871704,
      "learning_rate": 2.4911314820024568e-05,
      "loss": 0.0398,
      "step": 1840
    },
    {
      "epoch": 6.877323420074349,
      "grad_norm": 0.4652217626571655,
      "learning_rate": 2.438142939357882e-05,
      "loss": 0.0384,
      "step": 1850
    },
    {
      "epoch": 6.91449814126394,
      "grad_norm": 0.4344288408756256,
      "learning_rate": 2.38554171442674e-05,
      "loss": 0.0358,
      "step": 1860
    },
    {
      "epoch": 6.951672862453531,
      "grad_norm": 0.550430953502655,
      "learning_rate": 2.3333357597940793e-05,
      "loss": 0.045,
      "step": 1870
    },
    {
      "epoch": 6.988847583643123,
      "grad_norm": 0.5860105156898499,
      "learning_rate": 2.281532968285494e-05,
      "loss": 0.0425,
      "step": 1880
    },
    {
      "epoch": 7.026022304832714,
      "grad_norm": 0.3104412257671356,
      "learning_rate": 2.230141171773836e-05,
      "loss": 0.0377,
      "step": 1890
    },
    {
      "epoch": 7.063197026022305,
      "grad_norm": 0.629639744758606,
      "learning_rate": 2.179168139995134e-05,
      "loss": 0.0398,
      "step": 1900
    },
    {
      "epoch": 7.100371747211896,
      "grad_norm": 0.280269980430603,
      "learning_rate": 2.1286215793739302e-05,
      "loss": 0.0337,
      "step": 1910
    },
    {
      "epoch": 7.137546468401487,
      "grad_norm": 1.1272166967391968,
      "learning_rate": 2.0785091318581577e-05,
      "loss": 0.0409,
      "step": 1920
    },
    {
      "epoch": 7.174721189591078,
      "grad_norm": 0.33485621213912964,
      "learning_rate": 2.0288383737638006e-05,
      "loss": 0.0403,
      "step": 1930
    },
    {
      "epoch": 7.211895910780669,
      "grad_norm": 0.6893181204795837,
      "learning_rate": 1.9796168146294412e-05,
      "loss": 0.0428,
      "step": 1940
    },
    {
      "epoch": 7.24907063197026,
      "grad_norm": 1.7540059089660645,
      "learning_rate": 1.9308518960809353e-05,
      "loss": 0.0365,
      "step": 1950
    },
    {
      "epoch": 7.286245353159852,
      "grad_norm": 0.6964354515075684,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.0413,
      "step": 1960
    },
    {
      "epoch": 7.323420074349443,
      "grad_norm": 0.5227733850479126,
      "learning_rate": 1.83472140094124e-05,
      "loss": 0.0356,
      "step": 1970
    },
    {
      "epoch": 7.360594795539034,
      "grad_norm": 0.3736005127429962,
      "learning_rate": 1.7873703579648034e-05,
      "loss": 0.0371,
      "step": 1980
    },
    {
      "epoch": 7.397769516728625,
      "grad_norm": 0.6643156409263611,
      "learning_rate": 1.7405050206064373e-05,
      "loss": 0.0412,
      "step": 1990
    },
    {
      "epoch": 7.434944237918216,
      "grad_norm": 0.34871557354927063,
      "learning_rate": 1.694132474263526e-05,
      "loss": 0.0469,
      "step": 2000
    },
    {
      "epoch": 7.4721189591078065,
      "grad_norm": 0.44656774401664734,
      "learning_rate": 1.6482597298301916e-05,
      "loss": 0.0352,
      "step": 2010
    },
    {
      "epoch": 7.5092936802973975,
      "grad_norm": 0.28243181109428406,
      "learning_rate": 1.6028937226373536e-05,
      "loss": 0.0387,
      "step": 2020
    },
    {
      "epoch": 7.5464684014869885,
      "grad_norm": 0.2621857225894928,
      "learning_rate": 1.5580413114042e-05,
      "loss": 0.0403,
      "step": 2030
    },
    {
      "epoch": 7.58364312267658,
      "grad_norm": 0.3578774929046631,
      "learning_rate": 1.5137092772012368e-05,
      "loss": 0.039,
      "step": 2040
    },
    {
      "epoch": 7.620817843866171,
      "grad_norm": 0.41617679595947266,
      "learning_rate": 1.469904322425092e-05,
      "loss": 0.0414,
      "step": 2050
    },
    {
      "epoch": 7.657992565055762,
      "grad_norm": 0.44881588220596313,
      "learning_rate": 1.4266330697851954e-05,
      "loss": 0.0386,
      "step": 2060
    },
    {
      "epoch": 7.695167286245353,
      "grad_norm": 0.4774656295776367,
      "learning_rate": 1.3839020613025266e-05,
      "loss": 0.0394,
      "step": 2070
    },
    {
      "epoch": 7.732342007434944,
      "grad_norm": 0.3506743609905243,
      "learning_rate": 1.3417177573205398e-05,
      "loss": 0.044,
      "step": 2080
    },
    {
      "epoch": 7.769516728624535,
      "grad_norm": 0.35965225100517273,
      "learning_rate": 1.3000865355284563e-05,
      "loss": 0.0377,
      "step": 2090
    },
    {
      "epoch": 7.806691449814126,
      "grad_norm": 0.3799026906490326,
      "learning_rate": 1.2590146899970446e-05,
      "loss": 0.0359,
      "step": 2100
    },
    {
      "epoch": 7.843866171003717,
      "grad_norm": 0.38570570945739746,
      "learning_rate": 1.218508430227035e-05,
      "loss": 0.038,
      "step": 2110
    },
    {
      "epoch": 7.881040892193308,
      "grad_norm": 0.4911622107028961,
      "learning_rate": 1.1785738802103396e-05,
      "loss": 0.0373,
      "step": 2120
    },
    {
      "epoch": 7.9182156133829,
      "grad_norm": 0.8531225919723511,
      "learning_rate": 1.1392170775041788e-05,
      "loss": 0.0386,
      "step": 2130
    },
    {
      "epoch": 7.955390334572491,
      "grad_norm": 0.41892483830451965,
      "learning_rate": 1.1004439723182942e-05,
      "loss": 0.0421,
      "step": 2140
    },
    {
      "epoch": 7.992565055762082,
      "grad_norm": 0.45428428053855896,
      "learning_rate": 1.0622604266153485e-05,
      "loss": 0.041,
      "step": 2150
    },
    {
      "epoch": 8.029739776951672,
      "grad_norm": 0.30198222398757935,
      "learning_rate": 1.0246722132246856e-05,
      "loss": 0.0366,
      "step": 2160
    },
    {
      "epoch": 8.066914498141264,
      "grad_norm": 0.3173958659172058,
      "learning_rate": 9.876850149695555e-06,
      "loss": 0.0341,
      "step": 2170
    },
    {
      "epoch": 8.104089219330856,
      "grad_norm": 0.4124772548675537,
      "learning_rate": 9.513044238079427e-06,
      "loss": 0.034,
      "step": 2180
    },
    {
      "epoch": 8.141263940520446,
      "grad_norm": 0.2519373893737793,
      "learning_rate": 9.15535939987151e-06,
      "loss": 0.0313,
      "step": 2190
    },
    {
      "epoch": 8.178438661710038,
      "grad_norm": 0.41054391860961914,
      "learning_rate": 8.803849712122292e-06,
      "loss": 0.0371,
      "step": 2200
    },
    {
      "epoch": 8.215613382899628,
      "grad_norm": 0.4737650454044342,
      "learning_rate": 8.458568318284088e-06,
      "loss": 0.0427,
      "step": 2210
    },
    {
      "epoch": 8.25278810408922,
      "grad_norm": 0.2717382609844208,
      "learning_rate": 8.11956742017641e-06,
      "loss": 0.0354,
      "step": 2220
    },
    {
      "epoch": 8.28996282527881,
      "grad_norm": 0.3701151907444,
      "learning_rate": 7.786898270093846e-06,
      "loss": 0.0347,
      "step": 2230
    },
    {
      "epoch": 8.327137546468402,
      "grad_norm": 0.3946413993835449,
      "learning_rate": 7.460611163057346e-06,
      "loss": 0.0314,
      "step": 2240
    },
    {
      "epoch": 8.364312267657992,
      "grad_norm": 0.4067074656486511,
      "learning_rate": 7.140755429210316e-06,
      "loss": 0.0365,
      "step": 2250
    },
    {
      "epoch": 8.401486988847584,
      "grad_norm": 0.2950913906097412,
      "learning_rate": 6.827379426360614e-06,
      "loss": 0.0407,
      "step": 2260
    },
    {
      "epoch": 8.438661710037175,
      "grad_norm": 0.22738321125507355,
      "learning_rate": 6.52053053266945e-06,
      "loss": 0.0359,
      "step": 2270
    },
    {
      "epoch": 8.475836431226766,
      "grad_norm": 0.4955553114414215,
      "learning_rate": 6.22025513948854e-06,
      "loss": 0.033,
      "step": 2280
    },
    {
      "epoch": 8.513011152416357,
      "grad_norm": 0.31211531162261963,
      "learning_rate": 5.926598644346259e-06,
      "loss": 0.0378,
      "step": 2290
    },
    {
      "epoch": 8.550185873605948,
      "grad_norm": 0.28172406554222107,
      "learning_rate": 5.639605444084273e-06,
      "loss": 0.0358,
      "step": 2300
    },
    {
      "epoch": 8.58736059479554,
      "grad_norm": 0.5845323204994202,
      "learning_rate": 5.3593189281452625e-06,
      "loss": 0.0382,
      "step": 2310
    },
    {
      "epoch": 8.62453531598513,
      "grad_norm": 0.27726075053215027,
      "learning_rate": 5.085781472013051e-06,
      "loss": 0.0326,
      "step": 2320
    },
    {
      "epoch": 8.661710037174721,
      "grad_norm": 0.504346489906311,
      "learning_rate": 4.819034430806046e-06,
      "loss": 0.0297,
      "step": 2330
    },
    {
      "epoch": 8.698884758364311,
      "grad_norm": 0.23243720829486847,
      "learning_rate": 4.559118133024853e-06,
      "loss": 0.0405,
      "step": 2340
    },
    {
      "epoch": 8.736059479553903,
      "grad_norm": 0.5230311155319214,
      "learning_rate": 4.3060718744552256e-06,
      "loss": 0.0443,
      "step": 2350
    },
    {
      "epoch": 8.773234200743495,
      "grad_norm": 0.38711124658584595,
      "learning_rate": 4.05993391222701e-06,
      "loss": 0.0382,
      "step": 2360
    },
    {
      "epoch": 8.810408921933085,
      "grad_norm": 0.49542826414108276,
      "learning_rate": 3.820741459030253e-06,
      "loss": 0.0324,
      "step": 2370
    },
    {
      "epoch": 8.847583643122677,
      "grad_norm": 0.41099298000335693,
      "learning_rate": 3.5885306774891215e-06,
      "loss": 0.0302,
      "step": 2380
    },
    {
      "epoch": 8.884758364312267,
      "grad_norm": 0.2901303768157959,
      "learning_rate": 3.3633366746946004e-06,
      "loss": 0.0328,
      "step": 2390
    },
    {
      "epoch": 8.921933085501859,
      "grad_norm": 0.3307854235172272,
      "learning_rate": 3.145193496896809e-06,
      "loss": 0.0385,
      "step": 2400
    },
    {
      "epoch": 8.95910780669145,
      "grad_norm": 0.5515095591545105,
      "learning_rate": 2.934134124357646e-06,
      "loss": 0.0386,
      "step": 2410
    },
    {
      "epoch": 8.996282527881041,
      "grad_norm": 0.3325430154800415,
      "learning_rate": 2.7301904663646516e-06,
      "loss": 0.0418,
      "step": 2420
    },
    {
      "epoch": 9.033457249070633,
      "grad_norm": 0.38487276434898376,
      "learning_rate": 2.5333933564067492e-06,
      "loss": 0.0466,
      "step": 2430
    },
    {
      "epoch": 9.070631970260223,
      "grad_norm": 0.4031588137149811,
      "learning_rate": 2.343772547512613e-06,
      "loss": 0.0299,
      "step": 2440
    },
    {
      "epoch": 9.107806691449815,
      "grad_norm": 0.4202631115913391,
      "learning_rate": 2.1613567077524875e-06,
      "loss": 0.0289,
      "step": 2450
    },
    {
      "epoch": 9.144981412639405,
      "grad_norm": 0.3162038028240204,
      "learning_rate": 1.9861734159038968e-06,
      "loss": 0.0297,
      "step": 2460
    },
    {
      "epoch": 9.182156133828997,
      "grad_norm": 0.37053269147872925,
      "learning_rate": 1.8182491572821536e-06,
      "loss": 0.03,
      "step": 2470
    },
    {
      "epoch": 9.219330855018587,
      "grad_norm": 0.36643552780151367,
      "learning_rate": 1.6576093197361253e-06,
      "loss": 0.0356,
      "step": 2480
    },
    {
      "epoch": 9.256505576208179,
      "grad_norm": 0.4086417257785797,
      "learning_rate": 1.5042781898099434e-06,
      "loss": 0.0296,
      "step": 2490
    },
    {
      "epoch": 9.293680297397769,
      "grad_norm": 0.9764392971992493,
      "learning_rate": 1.3582789490712179e-06,
      "loss": 0.03,
      "step": 2500
    },
    {
      "epoch": 9.33085501858736,
      "grad_norm": 0.24946707487106323,
      "learning_rate": 1.2196336706062739e-06,
      "loss": 0.0342,
      "step": 2510
    },
    {
      "epoch": 9.368029739776953,
      "grad_norm": 0.3657890856266022,
      "learning_rate": 1.0883633156830553e-06,
      "loss": 0.0294,
      "step": 2520
    },
    {
      "epoch": 9.405204460966543,
      "grad_norm": 0.34119731187820435,
      "learning_rate": 9.644877305819976e-07,
      "loss": 0.034,
      "step": 2530
    },
    {
      "epoch": 9.442379182156134,
      "grad_norm": 0.5186340808868408,
      "learning_rate": 8.480256435956124e-07,
      "loss": 0.0355,
      "step": 2540
    },
    {
      "epoch": 9.479553903345725,
      "grad_norm": 0.790932834148407,
      "learning_rate": 7.389946621969679e-07,
      "loss": 0.0409,
      "step": 2550
    },
    {
      "epoch": 9.516728624535316,
      "grad_norm": 0.3314131796360016,
      "learning_rate": 6.374112703777301e-07,
      "loss": 0.0321,
      "step": 2560
    },
    {
      "epoch": 9.553903345724907,
      "grad_norm": 0.6170405149459839,
      "learning_rate": 5.432908261559733e-07,
      "loss": 0.0352,
      "step": 2570
    },
    {
      "epoch": 9.591078066914498,
      "grad_norm": 0.30839791893959045,
      "learning_rate": 4.5664755925426403e-07,
      "loss": 0.0283,
      "step": 2580
    },
    {
      "epoch": 9.628252788104088,
      "grad_norm": 0.2792086899280548,
      "learning_rate": 3.7749456894834447e-07,
      "loss": 0.035,
      "step": 2590
    },
    {
      "epoch": 9.66542750929368,
      "grad_norm": 0.3965803384780884,
      "learning_rate": 3.058438220866544e-07,
      "loss": 0.0324,
      "step": 2600
    },
    {
      "epoch": 9.702602230483272,
      "grad_norm": 0.3901006877422333,
      "learning_rate": 2.4170615128117357e-07,
      "loss": 0.0323,
      "step": 2610
    },
    {
      "epoch": 9.739776951672862,
      "grad_norm": 0.32952436804771423,
      "learning_rate": 1.850912532696092e-07,
      "loss": 0.0281,
      "step": 2620
    },
    {
      "epoch": 9.776951672862454,
      "grad_norm": 0.2635156214237213,
      "learning_rate": 1.3600768744944647e-07,
      "loss": 0.0282,
      "step": 2630
    },
    {
      "epoch": 9.814126394052044,
      "grad_norm": 0.1897970736026764,
      "learning_rate": 9.446287458383385e-08,
      "loss": 0.0388,
      "step": 2640
    },
    {
      "epoch": 9.851301115241636,
      "grad_norm": 0.3042493462562561,
      "learning_rate": 6.046309567968588e-08,
      "loss": 0.0316,
      "step": 2650
    },
    {
      "epoch": 9.888475836431226,
      "grad_norm": 0.42293256521224976,
      "learning_rate": 3.4013491038087375e-08,
      "loss": 0.0299,
      "step": 2660
    },
    {
      "epoch": 9.925650557620818,
      "grad_norm": 0.7281969785690308,
      "learning_rate": 1.511805947714273e-08,
      "loss": 0.0412,
      "step": 2670
    },
    {
      "epoch": 9.962825278810408,
      "grad_norm": 0.3139171004295349,
      "learning_rate": 3.7796577274096245e-09,
      "loss": 0.0341,
      "step": 2680
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.9909800887107849,
      "learning_rate": 0.0,
      "loss": 0.0364,
      "step": 2690
    },
    {
      "epoch": 10.0,
      "step": 2690,
      "total_flos": 2.2056377044963104e+17,
      "train_loss": 0.06680747533154753,
      "train_runtime": 2132.8764,
      "train_samples_per_second": 33.978,
      "train_steps_per_second": 1.261
    }
  ],
  "logging_steps": 10,
  "max_steps": 2690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2056377044963104e+17,
  "train_batch_size": 27,
  "trial_name": null,
  "trial_params": null
}