{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1032,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05830903790087463,
      "grad_norm": 14.626969535556903,
      "learning_rate": 9.000000000000001e-07,
      "loss": 1.5047,
      "step": 10
    },
    {
      "epoch": 0.11661807580174927,
      "grad_norm": 7.637318846998983,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.1973,
      "step": 20
    },
    {
      "epoch": 0.1749271137026239,
      "grad_norm": 3.691260729275177,
      "learning_rate": 2.9e-06,
      "loss": 0.8,
      "step": 30
    },
    {
      "epoch": 0.23323615160349853,
      "grad_norm": 2.1291989636182427,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.6175,
      "step": 40
    },
    {
      "epoch": 0.2915451895043732,
      "grad_norm": 2.14798091897299,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.5677,
      "step": 50
    },
    {
      "epoch": 0.3498542274052478,
      "grad_norm": 2.0456305815246005,
      "learning_rate": 5.9e-06,
      "loss": 0.5051,
      "step": 60
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 1.988581124794002,
      "learning_rate": 6.9e-06,
      "loss": 0.4964,
      "step": 70
    },
    {
      "epoch": 0.46647230320699706,
      "grad_norm": 2.303466906806559,
      "learning_rate": 7.9e-06,
      "loss": 0.4929,
      "step": 80
    },
    {
      "epoch": 0.5247813411078717,
      "grad_norm": 1.9377203696272898,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.4703,
      "step": 90
    },
    {
      "epoch": 0.5830903790087464,
      "grad_norm": 2.190981800812453,
      "learning_rate": 9.9e-06,
      "loss": 0.4443,
      "step": 100
    },
    {
      "epoch": 0.641399416909621,
      "grad_norm": 2.0744832345963493,
      "learning_rate": 9.997699301870489e-06,
      "loss": 0.4364,
      "step": 110
    },
    {
      "epoch": 0.6997084548104956,
      "grad_norm": 1.7704017643532202,
      "learning_rate": 9.98974898953517e-06,
      "loss": 0.4147,
      "step": 120
    },
    {
      "epoch": 0.7580174927113703,
      "grad_norm": 1.747873159096446,
      "learning_rate": 9.976129689724575e-06,
      "loss": 0.4294,
      "step": 130
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 1.7960844701869685,
      "learning_rate": 9.956856875690006e-06,
      "loss": 0.4168,
      "step": 140
    },
    {
      "epoch": 0.8746355685131195,
      "grad_norm": 1.7736875630417643,
      "learning_rate": 9.931952443791704e-06,
      "loss": 0.4085,
      "step": 150
    },
    {
      "epoch": 0.9329446064139941,
      "grad_norm": 2.2132067152288752,
      "learning_rate": 9.901444688621801e-06,
      "loss": 0.4063,
      "step": 160
    },
    {
      "epoch": 0.9912536443148688,
      "grad_norm": 1.8315327950009368,
      "learning_rate": 9.865368270858082e-06,
      "loss": 0.3978,
      "step": 170
    },
    {
      "epoch": 1.0466472303206997,
      "grad_norm": 1.6782697102135526,
      "learning_rate": 9.82376417788506e-06,
      "loss": 0.343,
      "step": 180
    },
    {
      "epoch": 1.1049562682215743,
      "grad_norm": 1.5675824197763997,
      "learning_rate": 9.77667967722711e-06,
      "loss": 0.297,
      "step": 190
    },
    {
      "epoch": 1.163265306122449,
      "grad_norm": 1.9145136006604275,
      "learning_rate": 9.724168262846567e-06,
      "loss": 0.3153,
      "step": 200
    },
    {
      "epoch": 1.2215743440233235,
      "grad_norm": 1.7203060932188563,
      "learning_rate": 9.666289594367804e-06,
      "loss": 0.3036,
      "step": 210
    },
    {
      "epoch": 1.2798833819241984,
      "grad_norm": 1.7012502607051327,
      "learning_rate": 9.603109429296333e-06,
      "loss": 0.3246,
      "step": 220
    },
    {
      "epoch": 1.3381924198250728,
      "grad_norm": 1.6624551186289933,
      "learning_rate": 9.534699548309949e-06,
      "loss": 0.302,
      "step": 230
    },
    {
      "epoch": 1.3965014577259476,
      "grad_norm": 1.442216855717199,
      "learning_rate": 9.461137673706768e-06,
      "loss": 0.3068,
      "step": 240
    },
    {
      "epoch": 1.4548104956268222,
      "grad_norm": 1.6576009785614365,
      "learning_rate": 9.382507381102848e-06,
      "loss": 0.298,
      "step": 250
    },
    {
      "epoch": 1.5131195335276968,
      "grad_norm": 1.8058128149355792,
      "learning_rate": 9.298898004479698e-06,
      "loss": 0.3075,
      "step": 260
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 1.4784242670602321,
      "learning_rate": 9.210404534689536e-06,
      "loss": 0.3057,
      "step": 270
    },
    {
      "epoch": 1.629737609329446,
      "grad_norm": 1.546258930542768,
      "learning_rate": 9.117127511533654e-06,
      "loss": 0.2988,
      "step": 280
    },
    {
      "epoch": 1.6880466472303208,
      "grad_norm": 1.695476568663106,
      "learning_rate": 9.019172909536442e-06,
      "loss": 0.3033,
      "step": 290
    },
    {
      "epoch": 1.7463556851311952,
      "grad_norm": 1.6544653059743208,
      "learning_rate": 8.9166520175449e-06,
      "loss": 0.3035,
      "step": 300
    },
    {
      "epoch": 1.80466472303207,
      "grad_norm": 1.6077779186418766,
      "learning_rate": 8.809681312290398e-06,
      "loss": 0.3037,
      "step": 310
    },
    {
      "epoch": 1.8629737609329446,
      "grad_norm": 1.6177268312586421,
      "learning_rate": 8.698382326056341e-06,
      "loss": 0.2921,
      "step": 320
    },
    {
      "epoch": 1.9212827988338192,
      "grad_norm": 1.603497878343536,
      "learning_rate": 8.582881508602082e-06,
      "loss": 0.2948,
      "step": 330
    },
    {
      "epoch": 1.9795918367346939,
      "grad_norm": 1.3886954878888524,
      "learning_rate": 8.46331008349997e-06,
      "loss": 0.2946,
      "step": 340
    },
    {
      "epoch": 2.0349854227405246,
      "grad_norm": 1.499985665960878,
      "learning_rate": 8.339803899048737e-06,
      "loss": 0.2193,
      "step": 350
    },
    {
      "epoch": 2.0932944606413995,
      "grad_norm": 1.435584840559424,
      "learning_rate": 8.212503273932594e-06,
      "loss": 0.1747,
      "step": 360
    },
    {
      "epoch": 2.151603498542274,
      "grad_norm": 1.4871062012555394,
      "learning_rate": 8.081552837801428e-06,
      "loss": 0.1778,
      "step": 370
    },
    {
      "epoch": 2.2099125364431487,
      "grad_norm": 1.4218330629504066,
      "learning_rate": 7.947101366953177e-06,
      "loss": 0.176,
      "step": 380
    },
    {
      "epoch": 2.2682215743440235,
      "grad_norm": 1.2427066721460804,
      "learning_rate": 7.809301615305098e-06,
      "loss": 0.169,
      "step": 390
    },
    {
      "epoch": 2.326530612244898,
      "grad_norm": 1.491373179347201,
      "learning_rate": 7.668310140845944e-06,
      "loss": 0.1828,
      "step": 400
    },
    {
      "epoch": 2.3848396501457727,
      "grad_norm": 1.1382514278247085,
      "learning_rate": 7.524287127766245e-06,
      "loss": 0.1691,
      "step": 410
    },
    {
      "epoch": 2.443148688046647,
      "grad_norm": 1.4250482419303996,
      "learning_rate": 7.377396204468754e-06,
      "loss": 0.1813,
      "step": 420
    },
    {
      "epoch": 2.501457725947522,
      "grad_norm": 1.3115974493545381,
      "learning_rate": 7.227804257665838e-06,
      "loss": 0.1739,
      "step": 430
    },
    {
      "epoch": 2.5597667638483967,
      "grad_norm": 1.460585330094477,
      "learning_rate": 7.075681242775017e-06,
      "loss": 0.1857,
      "step": 440
    },
    {
      "epoch": 2.618075801749271,
      "grad_norm": 1.2364257305075577,
      "learning_rate": 6.921199990828056e-06,
      "loss": 0.1864,
      "step": 450
    },
    {
      "epoch": 2.6763848396501455,
      "grad_norm": 1.460927926445741,
      "learning_rate": 6.764536012113005e-06,
      "loss": 0.1927,
      "step": 460
    },
    {
      "epoch": 2.7346938775510203,
      "grad_norm": 1.1802915819618254,
      "learning_rate": 6.605867296772262e-06,
      "loss": 0.1716,
      "step": 470
    },
    {
      "epoch": 2.793002915451895,
      "grad_norm": 1.2215463060704888,
      "learning_rate": 6.445374112583196e-06,
      "loss": 0.1843,
      "step": 480
    },
    {
      "epoch": 2.8513119533527695,
      "grad_norm": 1.1217299502317768,
      "learning_rate": 6.2832388001511034e-06,
      "loss": 0.1711,
      "step": 490
    },
    {
      "epoch": 2.9096209912536444,
      "grad_norm": 1.359355058632944,
      "learning_rate": 6.119645565747165e-06,
      "loss": 0.1739,
      "step": 500
    },
    {
      "epoch": 2.9679300291545188,
      "grad_norm": 1.2912101573020145,
      "learning_rate": 5.954780272026761e-06,
      "loss": 0.1757,
      "step": 510
    },
    {
      "epoch": 3.0233236151603498,
      "grad_norm": 1.0681809205213206,
      "learning_rate": 5.788830226865929e-06,
      "loss": 0.1391,
      "step": 520
    },
    {
      "epoch": 3.0816326530612246,
      "grad_norm": 1.2260857222791004,
      "learning_rate": 5.621983970555881e-06,
      "loss": 0.0892,
      "step": 530
    },
    {
      "epoch": 3.139941690962099,
      "grad_norm": 1.0189023598379496,
      "learning_rate": 5.454431061597312e-06,
      "loss": 0.0872,
      "step": 540
    },
    {
      "epoch": 3.198250728862974,
      "grad_norm": 1.1572966564929497,
      "learning_rate": 5.286361861337924e-06,
      "loss": 0.0849,
      "step": 550
    },
    {
      "epoch": 3.256559766763848,
      "grad_norm": 1.3012849394193384,
      "learning_rate": 5.1179673176977915e-06,
      "loss": 0.0871,
      "step": 560
    },
    {
      "epoch": 3.314868804664723,
      "grad_norm": 1.0416650556204414,
      "learning_rate": 4.9494387482283185e-06,
      "loss": 0.0905,
      "step": 570
    },
    {
      "epoch": 3.373177842565598,
      "grad_norm": 1.210951177453276,
      "learning_rate": 4.7809676227512455e-06,
      "loss": 0.0895,
      "step": 580
    },
    {
      "epoch": 3.431486880466472,
      "grad_norm": 1.21811830246229,
      "learning_rate": 4.612745345824652e-06,
      "loss": 0.0953,
      "step": 590
    },
    {
      "epoch": 3.489795918367347,
      "grad_norm": 1.0389113088365345,
      "learning_rate": 4.444963039283114e-06,
      "loss": 0.0924,
      "step": 600
    },
    {
      "epoch": 3.5481049562682214,
      "grad_norm": 1.0509213042877452,
      "learning_rate": 4.277811325099072e-06,
      "loss": 0.0933,
      "step": 610
    },
    {
      "epoch": 3.6064139941690962,
      "grad_norm": 1.031466931202058,
      "learning_rate": 4.111480108812096e-06,
      "loss": 0.0921,
      "step": 620
    },
    {
      "epoch": 3.664723032069971,
      "grad_norm": 0.9674006736435579,
      "learning_rate": 3.946158363772118e-06,
      "loss": 0.0884,
      "step": 630
    },
    {
      "epoch": 3.7230320699708455,
      "grad_norm": 1.052555764127678,
      "learning_rate": 3.78203391644175e-06,
      "loss": 0.0881,
      "step": 640
    },
    {
      "epoch": 3.78134110787172,
      "grad_norm": 1.041731792136989,
      "learning_rate": 3.6192932330016074e-06,
      "loss": 0.0904,
      "step": 650
    },
    {
      "epoch": 3.8396501457725947,
      "grad_norm": 1.1688175176919982,
      "learning_rate": 3.4581212075010834e-06,
      "loss": 0.0901,
      "step": 660
    },
    {
      "epoch": 3.8979591836734695,
      "grad_norm": 1.102720481784069,
      "learning_rate": 3.2987009517952716e-06,
      "loss": 0.0915,
      "step": 670
    },
    {
      "epoch": 3.956268221574344,
      "grad_norm": 1.0261487225837105,
      "learning_rate": 3.1412135875066853e-06,
      "loss": 0.0886,
      "step": 680
    },
    {
      "epoch": 4.011661807580175,
      "grad_norm": 0.6607620413640864,
      "learning_rate": 2.9858380402481203e-06,
      "loss": 0.0711,
      "step": 690
    },
    {
      "epoch": 4.069970845481049,
      "grad_norm": 0.8936608342352569,
      "learning_rate": 2.8327508363404816e-06,
      "loss": 0.0389,
      "step": 700
    },
    {
      "epoch": 4.128279883381924,
      "grad_norm": 0.7891996422928471,
      "learning_rate": 2.6821259022565106e-06,
      "loss": 0.0367,
      "step": 710
    },
    {
      "epoch": 4.186588921282799,
      "grad_norm": 0.9396857177599008,
      "learning_rate": 2.5341343670182373e-06,
      "loss": 0.0371,
      "step": 720
    },
    {
      "epoch": 4.244897959183674,
      "grad_norm": 0.7935821446040255,
      "learning_rate": 2.3889443677727386e-06,
      "loss": 0.035,
      "step": 730
    },
    {
      "epoch": 4.303206997084548,
      "grad_norm": 0.7947020670633791,
      "learning_rate": 2.2467208587670054e-06,
      "loss": 0.0384,
      "step": 740
    },
    {
      "epoch": 4.3615160349854225,
      "grad_norm": 0.8946315335018554,
      "learning_rate": 2.107625423939015e-06,
      "loss": 0.0368,
      "step": 750
    },
    {
      "epoch": 4.419825072886297,
      "grad_norm": 0.7944653560307688,
      "learning_rate": 1.9718160933378905e-06,
      "loss": 0.0368,
      "step": 760
    },
    {
      "epoch": 4.478134110787172,
      "grad_norm": 0.7774636550435158,
      "learning_rate": 1.8394471635817158e-06,
      "loss": 0.0349,
      "step": 770
    },
    {
      "epoch": 4.536443148688047,
      "grad_norm": 0.8756408218531322,
      "learning_rate": 1.7106690225570343e-06,
      "loss": 0.0371,
      "step": 780
    },
    {
      "epoch": 4.594752186588921,
      "grad_norm": 0.8687861976493633,
      "learning_rate": 1.5856279785591321e-06,
      "loss": 0.0339,
      "step": 790
    },
    {
      "epoch": 4.653061224489796,
      "grad_norm": 0.8189000224644959,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.0323,
      "step": 800
    },
    {
      "epoch": 4.711370262390671,
      "grad_norm": 0.8369884028081186,
      "learning_rate": 1.3473210243436673e-06,
      "loss": 0.0334,
      "step": 810
    },
    {
      "epoch": 4.769679300291545,
      "grad_norm": 0.7785311317469292,
      "learning_rate": 1.2343258610397397e-06,
      "loss": 0.0326,
      "step": 820
    },
    {
      "epoch": 4.827988338192419,
      "grad_norm": 0.8404324074760597,
      "learning_rate": 1.1256089809870336e-06,
      "loss": 0.0334,
      "step": 830
    },
    {
      "epoch": 4.886297376093294,
      "grad_norm": 0.8675304198425063,
      "learning_rate": 1.0212939003449128e-06,
      "loss": 0.0345,
      "step": 840
    },
    {
      "epoch": 4.944606413994169,
      "grad_norm": 0.8431136894720528,
      "learning_rate": 9.214991342705304e-07,
      "loss": 0.0346,
      "step": 850
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1768071446682524,
      "learning_rate": 8.263380622705796e-07,
      "loss": 0.0337,
      "step": 860
    },
    {
      "epoch": 5.058309037900875,
      "grad_norm": 0.40110032583503596,
      "learning_rate": 7.359187993878109e-07,
      "loss": 0.0158,
      "step": 870
    },
    {
      "epoch": 5.11661807580175,
      "grad_norm": 0.49375416237980635,
      "learning_rate": 6.503440733686251e-07,
      "loss": 0.013,
      "step": 880
    },
    {
      "epoch": 5.174927113702624,
      "grad_norm": 0.5363565719022582,
      "learning_rate": 5.69711107951334e-07,
      "loss": 0.013,
      "step": 890
    },
    {
      "epoch": 5.233236151603498,
      "grad_norm": 0.5236558514075167,
      "learning_rate": 4.941115124076679e-07,
      "loss": 0.0123,
      "step": 900
    },
    {
      "epoch": 5.291545189504373,
      "grad_norm": 0.49489565231551663,
      "learning_rate": 4.2363117746302476e-07,
      "loss": 0.012,
      "step": 910
    },
    {
      "epoch": 5.349854227405248,
      "grad_norm": 0.5305713447576351,
      "learning_rate": 3.583501777137227e-07,
      "loss": 0.0125,
      "step": 920
    },
    {
      "epoch": 5.408163265306122,
      "grad_norm": 0.8039219674866297,
      "learning_rate": 2.9834268065210006e-07,
      "loss": 0.0125,
      "step": 930
    },
    {
      "epoch": 5.466472303206997,
      "grad_norm": 0.3862288158846098,
      "learning_rate": 2.4367686240284127e-07,
      "loss": 0.0124,
      "step": 940
    },
    {
      "epoch": 5.524781341107872,
      "grad_norm": 0.47054782769610126,
      "learning_rate": 1.9441483026626185e-07,
      "loss": 0.0113,
      "step": 950
    },
    {
      "epoch": 5.5830903790087465,
      "grad_norm": 0.4671541625745262,
      "learning_rate": 1.506125521565327e-07,
      "loss": 0.0116,
      "step": 960
    },
    {
      "epoch": 5.641399416909621,
      "grad_norm": 0.48837436287072683,
      "learning_rate": 1.1231979301504048e-07,
      "loss": 0.0117,
      "step": 970
    },
    {
      "epoch": 5.699708454810495,
      "grad_norm": 0.4538198562291079,
      "learning_rate": 7.958005827110039e-08,
      "loss": 0.0113,
      "step": 980
    },
    {
      "epoch": 5.75801749271137,
      "grad_norm": 0.5730409751947306,
      "learning_rate": 5.2430544414273645e-08,
      "loss": 0.0126,
      "step": 990
    },
    {
      "epoch": 5.816326530612245,
      "grad_norm": 0.6123539079634027,
      "learning_rate": 3.0902096734442554e-08,
      "loss": 0.0121,
      "step": 1000
    },
    {
      "epoch": 5.87463556851312,
      "grad_norm": 0.5748592685210242,
      "learning_rate": 1.5019174277645098e-08,
      "loss": 0.0114,
      "step": 1010
    },
    {
      "epoch": 5.932944606413994,
      "grad_norm": 0.4633724477797094,
      "learning_rate": 4.799822057502401e-09,
      "loss": 0.0123,
      "step": 1020
    },
    {
      "epoch": 5.9912536443148685,
      "grad_norm": 0.4377554362173321,
      "learning_rate": 2.5565055379261904e-10,
      "loss": 0.0116,
      "step": 1030
    },
    {
      "epoch": 6.0,
      "step": 1032,
      "total_flos": 884370134269952.0,
      "train_loss": 0.20082982304657615,
      "train_runtime": 10237.3944,
      "train_samples_per_second": 12.861,
      "train_steps_per_second": 0.101
    }
  ],
  "logging_steps": 10,
  "max_steps": 1032,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 884370134269952.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}