{
  "best_global_step": 570,
  "best_metric": 0.2172553390264511,
  "best_model_checkpoint": "/home/longtail/data/outputs/qwen25-coder-7b-verireason-no-reasoning-co-ratio1.0-epoch3-exaone-matched/checkpoint-570",
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 570,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.210302734375,
      "epoch": 0.026420079260237782,
      "grad_norm": 0.2272326648235321,
      "learning_rate": 2.1052631578947366e-06,
      "loss": 0.38262381553649905,
      "mean_token_accuracy": 0.9088647723197937,
      "num_tokens": 15331.0,
      "step": 5
    },
    {
      "entropy": 0.214697265625,
      "epoch": 0.052840158520475564,
      "grad_norm": 0.24048204720020294,
      "learning_rate": 4.736842105263158e-06,
      "loss": 0.36759538650512696,
      "mean_token_accuracy": 0.9038324475288391,
      "num_tokens": 31218.0,
      "step": 10
    },
    {
      "entropy": 0.265625,
      "epoch": 0.07926023778071334,
      "grad_norm": 0.2194337695837021,
      "learning_rate": 7.3684210526315784e-06,
      "loss": 0.44762325286865234,
      "mean_token_accuracy": 0.897509315609932,
      "num_tokens": 46827.0,
      "step": 15
    },
    {
      "entropy": 0.25048828125,
      "epoch": 0.10568031704095113,
      "grad_norm": 0.2488376647233963,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.41580977439880373,
      "mean_token_accuracy": 0.9002776592969894,
      "num_tokens": 61381.0,
      "step": 20
    },
    {
      "entropy": 0.2287353515625,
      "epoch": 0.13210039630118892,
      "grad_norm": 0.28977280855178833,
      "learning_rate": 1.263157894736842e-05,
      "loss": 0.42945280075073244,
      "mean_token_accuracy": 0.8985304534435272,
      "num_tokens": 77061.0,
      "step": 25
    },
    {
      "entropy": 0.2453125,
      "epoch": 0.15852047556142668,
      "grad_norm": 0.2707628011703491,
      "learning_rate": 1.5263157894736842e-05,
      "loss": 0.40612149238586426,
      "mean_token_accuracy": 0.9037269622087478,
      "num_tokens": 91074.0,
      "step": 30
    },
    {
      "entropy": 0.227783203125,
      "epoch": 0.18494055482166447,
      "grad_norm": 0.22182075679302216,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 0.34377403259277345,
      "mean_token_accuracy": 0.908874437212944,
      "num_tokens": 106214.0,
      "step": 35
    },
    {
      "entropy": 0.2180419921875,
      "epoch": 0.21136063408190225,
      "grad_norm": 0.24515125155448914,
      "learning_rate": 2.0526315789473685e-05,
      "loss": 0.27584066390991213,
      "mean_token_accuracy": 0.9267852157354355,
      "num_tokens": 120134.0,
      "step": 40
    },
    {
      "entropy": 0.30302734375,
      "epoch": 0.23778071334214002,
      "grad_norm": 0.36891454458236694,
      "learning_rate": 2.3157894736842103e-05,
      "loss": 0.35630002021789553,
      "mean_token_accuracy": 0.9039895206689834,
      "num_tokens": 134060.0,
      "step": 45
    },
    {
      "entropy": 0.264599609375,
      "epoch": 0.26420079260237783,
      "grad_norm": 0.21175123751163483,
      "learning_rate": 2.578947368421053e-05,
      "loss": 0.2749725341796875,
      "mean_token_accuracy": 0.9225947976112365,
      "num_tokens": 149458.0,
      "step": 50
    },
    {
      "entropy": 0.264453125,
      "epoch": 0.2906208718626156,
      "grad_norm": 0.3016625940799713,
      "learning_rate": 2.8421052631578946e-05,
      "loss": 0.2754631042480469,
      "mean_token_accuracy": 0.9233797520399094,
      "num_tokens": 163845.0,
      "step": 55
    },
    {
      "entropy": 0.2982421875,
      "epoch": 0.31704095112285335,
      "grad_norm": 0.2732570469379425,
      "learning_rate": 2.999887492664158e-05,
      "loss": 0.3260418653488159,
      "mean_token_accuracy": 0.9141247242689132,
      "num_tokens": 179412.0,
      "step": 60
    },
    {
      "entropy": 0.2699951171875,
      "epoch": 0.34346103038309117,
      "grad_norm": 0.3051692545413971,
      "learning_rate": 2.9986219789520436e-05,
      "loss": 0.2863668441772461,
      "mean_token_accuracy": 0.9155579835176468,
      "num_tokens": 193314.0,
      "step": 65
    },
    {
      "entropy": 0.241455078125,
      "epoch": 0.36988110964332893,
      "grad_norm": 0.23183029890060425,
      "learning_rate": 2.9959515077322413e-05,
      "loss": 0.2552649736404419,
      "mean_token_accuracy": 0.9229518651962281,
      "num_tokens": 208409.0,
      "step": 70
    },
    {
      "entropy": 0.2355712890625,
      "epoch": 0.3963011889035667,
      "grad_norm": 0.23595623672008514,
      "learning_rate": 2.9918785825704544e-05,
      "loss": 0.2474513530731201,
      "mean_token_accuracy": 0.9276347905397415,
      "num_tokens": 223599.0,
      "step": 75
    },
    {
      "entropy": 0.249169921875,
      "epoch": 0.4227212681638045,
      "grad_norm": 0.3135967254638672,
      "learning_rate": 2.9864070218323984e-05,
      "loss": 0.2542789697647095,
      "mean_token_accuracy": 0.9266749203205109,
      "num_tokens": 238278.0,
      "step": 80
    },
    {
      "entropy": 0.2227783203125,
      "epoch": 0.44914134742404227,
      "grad_norm": 0.27829885482788086,
      "learning_rate": 2.9795419551040836e-05,
      "loss": 0.21763288974761963,
      "mean_token_accuracy": 0.9401752114295959,
      "num_tokens": 252548.0,
      "step": 85
    },
    {
      "entropy": 0.2646728515625,
      "epoch": 0.47556142668428003,
      "grad_norm": 0.2549667954444885,
      "learning_rate": 2.9712898183828297e-05,
      "loss": 0.2652446269989014,
      "mean_token_accuracy": 0.9231121420860291,
      "num_tokens": 267931.0,
      "step": 90
    },
    {
      "entropy": 0.2947021484375,
      "epoch": 0.5019815059445178,
      "grad_norm": 0.2779158651828766,
      "learning_rate": 2.9616583480435235e-05,
      "loss": 0.28800928592681885,
      "mean_token_accuracy": 0.9192559570074081,
      "num_tokens": 282065.0,
      "step": 95
    },
    {
      "entropy": 0.2372802734375,
      "epoch": 0.5284015852047557,
      "grad_norm": 0.25696665048599243,
      "learning_rate": 2.950656573585766e-05,
      "loss": 0.25461769104003906,
      "mean_token_accuracy": 0.9287555605173111,
      "num_tokens": 297844.0,
      "step": 100
    },
    {
      "epoch": 0.5284015852047557,
      "eval_entropy": 0.26014083059210524,
      "eval_loss": 0.25216802954673767,
      "eval_mean_token_accuracy": 0.9233670849549143,
      "eval_num_tokens": 297844.0,
      "eval_runtime": 4.7366,
      "eval_samples_per_second": 39.902,
      "eval_steps_per_second": 20.056,
      "step": 100
    },
    {
      "entropy": 0.212158203125,
      "epoch": 0.5548216644649934,
      "grad_norm": 0.3313332200050354,
      "learning_rate": 2.9382948091687255e-05,
      "loss": 0.21078295707702638,
      "mean_token_accuracy": 0.9359580397605896,
      "num_tokens": 311548.0,
      "step": 105
    },
    {
      "entropy": 0.2112060546875,
      "epoch": 0.5812417437252312,
      "grad_norm": 0.250174880027771,
      "learning_rate": 2.9245846439416132e-05,
      "loss": 0.2149496078491211,
      "mean_token_accuracy": 0.9396522492170334,
      "num_tokens": 326762.0,
      "step": 110
    },
    {
      "entropy": 0.233544921875,
      "epoch": 0.607661822985469,
      "grad_norm": 0.31775665283203125,
      "learning_rate": 2.9095389311788626e-05,
      "loss": 0.258730411529541,
      "mean_token_accuracy": 0.9234512418508529,
      "num_tokens": 342499.0,
      "step": 115
    },
    {
      "entropy": 0.229541015625,
      "epoch": 0.6340819022457067,
      "grad_norm": 0.32010215520858765,
      "learning_rate": 2.8931717762301893e-05,
      "loss": 0.23711071014404297,
      "mean_token_accuracy": 0.9289378106594086,
      "num_tokens": 355651.0,
      "step": 120
    },
    {
      "entropy": 0.1684326171875,
      "epoch": 0.6605019815059445,
      "grad_norm": 0.2775934338569641,
      "learning_rate": 2.8754985232968315e-05,
      "loss": 0.14103322029113768,
      "mean_token_accuracy": 0.9497533321380616,
      "num_tokens": 369599.0,
      "step": 125
    },
    {
      "entropy": 0.2874755859375,
      "epoch": 0.6869220607661823,
      "grad_norm": 0.3574838638305664,
      "learning_rate": 2.8565357410463664e-05,
      "loss": 0.35373797416687014,
      "mean_token_accuracy": 0.910257437825203,
      "num_tokens": 386714.0,
      "step": 130
    },
    {
      "entropy": 0.2102783203125,
      "epoch": 0.71334214002642,
      "grad_norm": 0.3869127929210663,
      "learning_rate": 2.8363012070795927e-05,
      "loss": 0.20922679901123048,
      "mean_token_accuracy": 0.9394314408302307,
      "num_tokens": 399622.0,
      "step": 135
    },
    {
      "entropy": 0.22381591796875,
      "epoch": 0.7397622192866579,
      "grad_norm": 0.3797408640384674,
      "learning_rate": 2.8148138912640355e-05,
      "loss": 0.2222966432571411,
      "mean_token_accuracy": 0.9335697323083878,
      "num_tokens": 415317.0,
      "step": 140
    },
    {
      "entropy": 0.2136474609375,
      "epoch": 0.7661822985468957,
      "grad_norm": 0.331587553024292,
      "learning_rate": 2.792093937949704e-05,
      "loss": 0.23508837223052978,
      "mean_token_accuracy": 0.9313608318567276,
      "num_tokens": 429274.0,
      "step": 145
    },
    {
      "entropy": 0.19384765625,
      "epoch": 0.7926023778071334,
      "grad_norm": 0.2861741781234741,
      "learning_rate": 2.7681626470837708e-05,
      "loss": 0.2089618682861328,
      "mean_token_accuracy": 0.93760344684124,
      "num_tokens": 445079.0,
      "step": 150
    },
    {
      "entropy": 0.2195068359375,
      "epoch": 0.8190224570673712,
      "grad_norm": 0.31503307819366455,
      "learning_rate": 2.7430424542418798e-05,
      "loss": 0.26315999031066895,
      "mean_token_accuracy": 0.9285398453474045,
      "num_tokens": 459855.0,
      "step": 155
    },
    {
      "entropy": 0.211865234375,
      "epoch": 0.845442536327609,
      "grad_norm": 0.3392992615699768,
      "learning_rate": 2.7167569095948032e-05,
      "loss": 0.22536911964416503,
      "mean_token_accuracy": 0.9318837016820908,
      "num_tokens": 473254.0,
      "step": 160
    },
    {
      "entropy": 0.199951171875,
      "epoch": 0.8718626155878467,
      "grad_norm": 0.34515535831451416,
      "learning_rate": 2.6893306558301674e-05,
      "loss": 0.22186641693115233,
      "mean_token_accuracy": 0.9428424179553986,
      "num_tokens": 487870.0,
      "step": 165
    },
    {
      "entropy": 0.275048828125,
      "epoch": 0.8982826948480845,
      "grad_norm": 0.3238794505596161,
      "learning_rate": 2.660789405049939e-05,
      "loss": 0.2929158449172974,
      "mean_token_accuracy": 0.917644801735878,
      "num_tokens": 502746.0,
      "step": 170
    },
    {
      "entropy": 0.318017578125,
      "epoch": 0.9247027741083224,
      "grad_norm": 0.3478052318096161,
      "learning_rate": 2.6311599146653446e-05,
      "loss": 0.33602657318115237,
      "mean_token_accuracy": 0.9080141842365265,
      "num_tokens": 517990.0,
      "step": 175
    },
    {
      "entropy": 0.2086669921875,
      "epoch": 0.9511228533685601,
      "grad_norm": 0.25240081548690796,
      "learning_rate": 2.600469962311804e-05,
      "loss": 0.23073394298553468,
      "mean_token_accuracy": 0.9393861293792725,
      "num_tokens": 532382.0,
      "step": 180
    },
    {
      "entropy": 0.2622802734375,
      "epoch": 0.9775429326287979,
      "grad_norm": 0.44267305731773376,
      "learning_rate": 2.5687483198074123e-05,
      "loss": 0.26755404472351074,
      "mean_token_accuracy": 0.9224933117628098,
      "num_tokens": 547708.0,
      "step": 185
    },
    {
      "entropy": 0.20502068014705882,
      "epoch": 1.0,
      "grad_norm": 0.8547039031982422,
      "learning_rate": 2.5360247261793666e-05,
      "loss": 0.21282377243041992,
      "mean_token_accuracy": 0.935039313400493,
      "num_tokens": 560030.0,
      "step": 190
    },
    {
      "entropy": 0.19251708984375,
      "epoch": 1.0264200792602378,
      "grad_norm": 0.31648746132850647,
      "learning_rate": 2.502329859783646e-05,
      "loss": 0.194023060798645,
      "mean_token_accuracy": 0.9428818017244339,
      "num_tokens": 575211.0,
      "step": 195
    },
    {
      "entropy": 0.24560546875,
      "epoch": 1.0528401585204756,
      "grad_norm": 0.40644362568855286,
      "learning_rate": 2.46769530954406e-05,
      "loss": 0.2631294012069702,
      "mean_token_accuracy": 0.9278771877288818,
      "num_tokens": 590045.0,
      "step": 200
    },
    {
      "epoch": 1.0528401585204756,
      "eval_entropy": 0.2268117804276316,
      "eval_loss": 0.23400349915027618,
      "eval_mean_token_accuracy": 0.9259416059443825,
      "eval_num_tokens": 590045.0,
      "eval_runtime": 4.7111,
      "eval_samples_per_second": 40.118,
      "eval_steps_per_second": 20.165,
      "step": 200
    },
    {
      "entropy": 0.22978515625,
      "epoch": 1.0792602377807134,
      "grad_norm": 0.4463113248348236,
      "learning_rate": 2.4321535453376472e-05,
      "loss": 0.24853811264038086,
      "mean_token_accuracy": 0.9308895230293274,
      "num_tokens": 605344.0,
      "step": 205
    },
    {
      "entropy": 0.221337890625,
      "epoch": 1.105680317040951,
      "grad_norm": 0.4098075032234192,
      "learning_rate": 2.3957378875541795e-05,
      "loss": 0.24046854972839354,
      "mean_token_accuracy": 0.9350315988063812,
      "num_tokens": 620661.0,
      "step": 210
    },
    {
      "entropy": 0.2391845703125,
      "epoch": 1.1321003963011889,
      "grad_norm": 0.3997504711151123,
      "learning_rate": 2.3584824758583077e-05,
      "loss": 0.24855997562408447,
      "mean_token_accuracy": 0.9297777205705643,
      "num_tokens": 636157.0,
      "step": 215
    },
    {
      "entropy": 0.2037841796875,
      "epoch": 1.1585204755614267,
      "grad_norm": 0.3310227692127228,
      "learning_rate": 2.320422237183641e-05,
      "loss": 0.1977330446243286,
      "mean_token_accuracy": 0.9382714837789535,
      "num_tokens": 651029.0,
      "step": 220
    },
    {
      "entropy": 0.182568359375,
      "epoch": 1.1849405548216645,
      "grad_norm": 0.31502988934516907,
      "learning_rate": 2.28159285298876e-05,
      "loss": 0.17885322570800782,
      "mean_token_accuracy": 0.9421977251768112,
      "num_tokens": 666993.0,
      "step": 225
    },
    {
      "entropy": 0.228564453125,
      "epoch": 1.2113606340819023,
      "grad_norm": 0.4962048828601837,
      "learning_rate": 2.242030725805864e-05,
      "loss": 0.2582087516784668,
      "mean_token_accuracy": 0.9375840485095978,
      "num_tokens": 683273.0,
      "step": 230
    },
    {
      "entropy": 0.21556396484375,
      "epoch": 1.23778071334214,
      "grad_norm": 0.34862184524536133,
      "learning_rate": 2.201772945113412e-05,
      "loss": 0.2517971992492676,
      "mean_token_accuracy": 0.9303504943847656,
      "num_tokens": 698334.0,
      "step": 235
    },
    {
      "entropy": 0.211962890625,
      "epoch": 1.2642007926023777,
      "grad_norm": 0.33244001865386963,
      "learning_rate": 2.16085725256475e-05,
      "loss": 0.22715094089508056,
      "mean_token_accuracy": 0.936177110671997,
      "num_tokens": 714090.0,
      "step": 240
    },
    {
      "entropy": 0.1701171875,
      "epoch": 1.2906208718626155,
      "grad_norm": 0.4048219919204712,
      "learning_rate": 2.1193220066053276e-05,
      "loss": 0.16675148010253907,
      "mean_token_accuracy": 0.9470200836658478,
      "num_tokens": 728351.0,
      "step": 245
    },
    {
      "entropy": 0.14979248046875,
      "epoch": 1.3170409511228534,
      "grad_norm": 0.369083434343338,
      "learning_rate": 2.0772061465116688e-05,
      "loss": 0.14205996990203856,
      "mean_token_accuracy": 0.9517278134822845,
      "num_tokens": 741598.0,
      "step": 250
    },
    {
      "entropy": 0.205712890625,
      "epoch": 1.3434610303830912,
      "grad_norm": 0.39064091444015503,
      "learning_rate": 2.0345491558858175e-05,
      "loss": 0.19915249347686767,
      "mean_token_accuracy": 0.9415161877870559,
      "num_tokens": 756707.0,
      "step": 255
    },
    {
      "entropy": 0.20458984375,
      "epoch": 1.369881109643329,
      "grad_norm": 0.41757169365882874,
      "learning_rate": 1.9913910256394766e-05,
      "loss": 0.21049420833587645,
      "mean_token_accuracy": 0.9382226914167404,
      "num_tokens": 772153.0,
      "step": 260
    },
    {
      "entropy": 0.1894775390625,
      "epoch": 1.3963011889035668,
      "grad_norm": 0.5959511995315552,
      "learning_rate": 1.9477722165025422e-05,
      "loss": 0.20970487594604492,
      "mean_token_accuracy": 0.9384240061044693,
      "num_tokens": 785641.0,
      "step": 265
    },
    {
      "entropy": 0.2056396484375,
      "epoch": 1.4227212681638046,
      "grad_norm": 0.4974619448184967,
      "learning_rate": 1.9037336210911868e-05,
      "loss": 0.21267409324645997,
      "mean_token_accuracy": 0.9356545269489288,
      "num_tokens": 801277.0,
      "step": 270
    },
    {
      "entropy": 0.1760009765625,
      "epoch": 1.4491413474240422,
      "grad_norm": 0.5783056616783142,
      "learning_rate": 1.8593165255710483e-05,
      "loss": 0.18950119018554687,
      "mean_token_accuracy": 0.9396244764328003,
      "num_tokens": 815204.0,
      "step": 275
    },
    {
      "entropy": 0.2078857421875,
      "epoch": 1.47556142668428,
      "grad_norm": 0.4639170169830322,
      "learning_rate": 1.8145625709514654e-05,
      "loss": 0.23868813514709472,
      "mean_token_accuracy": 0.9355578929185867,
      "num_tokens": 830311.0,
      "step": 280
    },
    {
      "entropy": 0.205615234375,
      "epoch": 1.5019815059445178,
      "grad_norm": 0.5348495244979858,
      "learning_rate": 1.7695137140470484e-05,
      "loss": 0.21911923885345458,
      "mean_token_accuracy": 0.9332752346992492,
      "num_tokens": 845164.0,
      "step": 285
    },
    {
      "entropy": 0.20986328125,
      "epoch": 1.5284015852047557,
      "grad_norm": 0.38183921575546265,
      "learning_rate": 1.724212188143182e-05,
      "loss": 0.21561951637268068,
      "mean_token_accuracy": 0.9330598264932632,
      "num_tokens": 859545.0,
      "step": 290
    },
    {
      "entropy": 0.1962890625,
      "epoch": 1.5548216644649933,
      "grad_norm": 0.45823004841804504,
      "learning_rate": 1.6787004634023356e-05,
      "loss": 0.20297410488128662,
      "mean_token_accuracy": 0.9378657519817353,
      "num_tokens": 873695.0,
      "step": 295
    },
    {
      "entropy": 0.1971923828125,
      "epoch": 1.581241743725231,
      "grad_norm": 0.48585113883018494,
      "learning_rate": 1.6330212070483076e-05,
      "loss": 0.19309521913528443,
      "mean_token_accuracy": 0.9413891911506653,
      "num_tokens": 889312.0,
      "step": 300
    },
    {
      "epoch": 1.581241743725231,
      "eval_entropy": 0.2150930304276316,
      "eval_loss": 0.22393172979354858,
      "eval_mean_token_accuracy": 0.9294123360985204,
      "eval_num_tokens": 889312.0,
      "eval_runtime": 4.6991,
      "eval_samples_per_second": 40.221,
      "eval_steps_per_second": 20.217,
      "step": 300
    },
    {
      "entropy": 0.1765625,
      "epoch": 1.607661822985469,
      "grad_norm": 0.5171866416931152,
      "learning_rate": 1.587217243365714e-05,
      "loss": 0.18057866096496583,
      "mean_token_accuracy": 0.9421164453029632,
      "num_tokens": 903271.0,
      "step": 305
    },
    {
      "entropy": 0.15986328125,
      "epoch": 1.6340819022457067,
      "grad_norm": 0.5997929573059082,
      "learning_rate": 1.5413315135522434e-05,
      "loss": 0.15566400289535523,
      "mean_token_accuracy": 0.9444721043109894,
      "num_tokens": 915512.0,
      "step": 310
    },
    {
      "entropy": 0.1656494140625,
      "epoch": 1.6605019815059445,
      "grad_norm": 0.6050997972488403,
      "learning_rate": 1.4954070354612982e-05,
      "loss": 0.1810092568397522,
      "mean_token_accuracy": 0.9434872210025788,
      "num_tokens": 929316.0,
      "step": 315
    },
    {
      "entropy": 0.2291015625,
      "epoch": 1.6869220607661823,
      "grad_norm": 0.47083452343940735,
      "learning_rate": 1.4494868632727769e-05,
      "loss": 0.26879181861877444,
      "mean_token_accuracy": 0.9259828954935074,
      "num_tokens": 944991.0,
      "step": 320
    },
    {
      "entropy": 0.25474853515625,
      "epoch": 1.7133421400264202,
      "grad_norm": 0.578498899936676,
      "learning_rate": 1.403614047129795e-05,
      "loss": 0.2804116725921631,
      "mean_token_accuracy": 0.9253529846668244,
      "num_tokens": 961086.0,
      "step": 325
    },
    {
      "entropy": 0.188720703125,
      "epoch": 1.739762219286658,
      "grad_norm": 0.5830754041671753,
      "learning_rate": 1.3578315927791948e-05,
      "loss": 0.2004934310913086,
      "mean_token_accuracy": 0.9403171420097352,
      "num_tokens": 974874.0,
      "step": 330
    },
    {
      "entropy": 0.2357177734375,
      "epoch": 1.7661822985468958,
      "grad_norm": 0.4751978814601898,
      "learning_rate": 1.3121824212536742e-05,
      "loss": 0.24819355010986327,
      "mean_token_accuracy": 0.9310738503932953,
      "num_tokens": 989482.0,
      "step": 335
    },
    {
      "entropy": 0.2209228515625,
      "epoch": 1.7926023778071334,
      "grad_norm": 0.5777033567428589,
      "learning_rate": 1.2667093286333357e-05,
      "loss": 0.23880820274353026,
      "mean_token_accuracy": 0.9315720736980438,
      "num_tokens": 1005284.0,
      "step": 340
    },
    {
      "entropy": 0.22216796875,
      "epoch": 1.8190224570673712,
      "grad_norm": 0.5043455958366394,
      "learning_rate": 1.2214549459243775e-05,
      "loss": 0.2275393486022949,
      "mean_token_accuracy": 0.9350559681653976,
      "num_tokens": 1020967.0,
      "step": 345
    },
    {
      "entropy": 0.1806640625,
      "epoch": 1.845442536327609,
      "grad_norm": 0.4179781675338745,
      "learning_rate": 1.1764616990925452e-05,
      "loss": 0.1624812126159668,
      "mean_token_accuracy": 0.9455937176942826,
      "num_tokens": 1035449.0,
      "step": 350
    },
    {
      "entropy": 0.21103515625,
      "epoch": 1.8718626155878466,
      "grad_norm": 0.4346652925014496,
      "learning_rate": 1.1317717692888014e-05,
      "loss": 0.22333147525787353,
      "mean_token_accuracy": 0.9315652191638947,
      "num_tokens": 1051018.0,
      "step": 355
    },
    {
      "entropy": 0.1916748046875,
      "epoch": 1.8982826948480844,
      "grad_norm": 0.5184327960014343,
      "learning_rate": 1.0874270533045177e-05,
      "loss": 0.19307322502136232,
      "mean_token_accuracy": 0.9433722555637359,
      "num_tokens": 1064913.0,
      "step": 360
    },
    {
      "entropy": 0.18074951171875,
      "epoch": 1.9247027741083222,
      "grad_norm": 0.6973323225975037,
      "learning_rate": 1.043469124293246e-05,
      "loss": 0.18255283832550048,
      "mean_token_accuracy": 0.9395459204912185,
      "num_tokens": 1080655.0,
      "step": 365
    },
    {
      "entropy": 0.1902587890625,
      "epoch": 1.95112285336856,
      "grad_norm": 0.5347604155540466,
      "learning_rate": 9.999391927959071e-06,
      "loss": 0.2131338119506836,
      "mean_token_accuracy": 0.9324799060821534,
      "num_tokens": 1093614.0,
      "step": 370
    },
    {
      "entropy": 0.203662109375,
      "epoch": 1.9775429326287979,
      "grad_norm": 0.5682111978530884,
      "learning_rate": 9.568780681059207e-06,
      "loss": 0.21848342418670655,
      "mean_token_accuracy": 0.932270273566246,
      "num_tokens": 1107999.0,
      "step": 375
    },
    {
      "entropy": 0.17408662683823528,
      "epoch": 2.0,
      "grad_norm": 1.124240517616272,
      "learning_rate": 9.143261200105145e-06,
      "loss": 0.2077638626098633,
      "mean_token_accuracy": 0.9436776988646564,
      "num_tokens": 1120060.0,
      "step": 380
    },
    {
      "entropy": 0.2104248046875,
      "epoch": 2.026420079260238,
      "grad_norm": 0.5043058395385742,
      "learning_rate": 8.72323240944058e-06,
      "loss": 0.20539186000823975,
      "mean_token_accuracy": 0.9394132256507873,
      "num_tokens": 1134276.0,
      "step": 385
    },
    {
      "entropy": 0.178759765625,
      "epoch": 2.0528401585204756,
      "grad_norm": 0.5097795128822327,
      "learning_rate": 8.309088085889304e-06,
      "loss": 0.16731362342834472,
      "mean_token_accuracy": 0.9490650534629822,
      "num_tokens": 1148618.0,
      "step": 390
    },
    {
      "entropy": 0.24921875,
      "epoch": 2.0792602377807134,
      "grad_norm": 0.5066947340965271,
      "learning_rate": 7.901216489589477e-06,
      "loss": 0.23654639720916748,
      "mean_token_accuracy": 0.9348665356636048,
      "num_tokens": 1163557.0,
      "step": 395
    },
    {
      "entropy": 0.18203125,
      "epoch": 2.1056803170409513,
      "grad_norm": 0.5836538672447205,
      "learning_rate": 7.500000000000004e-06,
      "loss": 0.13899633884429932,
      "mean_token_accuracy": 0.9545765340328216,
      "num_tokens": 1176678.0,
      "step": 400
    },
    {
      "epoch": 2.1056803170409513,
      "eval_entropy": 0.20543791118421054,
      "eval_loss": 0.21899865567684174,
      "eval_mean_token_accuracy": 0.931007537088896,
      "eval_num_tokens": 1176678.0,
      "eval_runtime": 4.7063,
      "eval_samples_per_second": 40.159,
      "eval_steps_per_second": 20.186,
      "step": 400
    },
    {
      "entropy": 0.17667236328125,
      "epoch": 2.132100396301189,
      "grad_norm": 0.5883168578147888,
      "learning_rate": 7.105814757419883e-06,
      "loss": 0.16830549240112305,
      "mean_token_accuracy": 0.9500027000904083,
      "num_tokens": 1192300.0,
      "step": 405
    },
    {
      "entropy": 0.1467529296875,
      "epoch": 2.158520475561427,
      "grad_norm": 0.36971086263656616,
      "learning_rate": 6.719030310356854e-06,
      "loss": 0.14136677980422974,
      "mean_token_accuracy": 0.9549454599618912,
      "num_tokens": 1206433.0,
      "step": 410
    },
    {
      "entropy": 0.195703125,
      "epoch": 2.1849405548216643,
      "grad_norm": 0.5005773901939392,
      "learning_rate": 6.340009269075807e-06,
      "loss": 0.19400241374969482,
      "mean_token_accuracy": 0.9394445419311523,
      "num_tokens": 1221815.0,
      "step": 415
    },
    {
      "entropy": 0.1435791015625,
      "epoch": 2.211360634081902,
      "grad_norm": 0.44894537329673767,
      "learning_rate": 5.969106965651805e-06,
      "loss": 0.13873835802078247,
      "mean_token_accuracy": 0.9550708144903183,
      "num_tokens": 1237748.0,
      "step": 420
    },
    {
      "entropy": 0.2187744140625,
      "epoch": 2.23778071334214,
      "grad_norm": 0.5867881178855896,
      "learning_rate": 5.6066711208464115e-06,
      "loss": 0.22510697841644287,
      "mean_token_accuracy": 0.9315517574548722,
      "num_tokens": 1252360.0,
      "step": 425
    },
    {
      "entropy": 0.235498046875,
      "epoch": 2.2642007926023777,
      "grad_norm": 0.605404257774353,
      "learning_rate": 5.253041518119608e-06,
      "loss": 0.24080452919006348,
      "mean_token_accuracy": 0.9283825039863587,
      "num_tokens": 1268082.0,
      "step": 430
    },
    {
      "entropy": 0.15987548828125,
      "epoch": 2.2906208718626155,
      "grad_norm": 0.5086158514022827,
      "learning_rate": 4.908549685082943e-06,
      "loss": 0.1456709623336792,
      "mean_token_accuracy": 0.9460086256265641,
      "num_tokens": 1283428.0,
      "step": 435
    },
    {
      "entropy": 0.1646484375,
      "epoch": 2.3170409511228534,
      "grad_norm": 0.5739629864692688,
      "learning_rate": 4.573518582692519e-06,
      "loss": 0.1556909441947937,
      "mean_token_accuracy": 0.9473109751939773,
      "num_tokens": 1299298.0,
      "step": 440
    },
    {
      "entropy": 0.16197509765625,
      "epoch": 2.343461030383091,
      "grad_norm": 0.5335649251937866,
      "learning_rate": 4.248262302473233e-06,
      "loss": 0.16208385229110717,
      "mean_token_accuracy": 0.9521175414323807,
      "num_tokens": 1314206.0,
      "step": 445
    },
    {
      "entropy": 0.177294921875,
      "epoch": 2.369881109643329,
      "grad_norm": 0.6850025057792664,
      "learning_rate": 3.933085772058094e-06,
      "loss": 0.19847280979156495,
      "mean_token_accuracy": 0.9465788096189499,
      "num_tokens": 1328448.0,
      "step": 450
    },
    {
      "entropy": 0.2244384765625,
      "epoch": 2.396301188903567,
      "grad_norm": 0.6370927691459656,
      "learning_rate": 3.628284469318688e-06,
      "loss": 0.23741717338562013,
      "mean_token_accuracy": 0.9319147199392319,
      "num_tokens": 1343514.0,
      "step": 455
    },
    {
      "entropy": 0.15863037109375,
      "epoch": 2.4227212681638046,
      "grad_norm": 0.6529363989830017,
      "learning_rate": 3.3341441453547932e-06,
      "loss": 0.16858468055725098,
      "mean_token_accuracy": 0.9459336966276168,
      "num_tokens": 1357222.0,
      "step": 460
    },
    {
      "entropy": 0.20355224609375,
      "epoch": 2.4491413474240424,
      "grad_norm": 0.4986702501773834,
      "learning_rate": 3.050940556602836e-06,
      "loss": 0.22168829441070556,
      "mean_token_accuracy": 0.9359017014503479,
      "num_tokens": 1372978.0,
      "step": 465
    },
    {
      "entropy": 0.217138671875,
      "epoch": 2.47556142668428,
      "grad_norm": 0.7848091125488281,
      "learning_rate": 2.778939206314343e-06,
      "loss": 0.23623857498168946,
      "mean_token_accuracy": 0.9339287459850312,
      "num_tokens": 1388009.0,
      "step": 470
    },
    {
      "entropy": 0.2067138671875,
      "epoch": 2.501981505944518,
      "grad_norm": 0.6232095956802368,
      "learning_rate": 2.51839509564675e-06,
      "loss": 0.20749149322509766,
      "mean_token_accuracy": 0.9373960226774216,
      "num_tokens": 1403597.0,
      "step": 475
    },
    {
      "entropy": 0.1603271484375,
      "epoch": 2.5284015852047554,
      "grad_norm": 0.5246099829673767,
      "learning_rate": 2.2695524845999143e-06,
      "loss": 0.18473849296569825,
      "mean_token_accuracy": 0.9495263069868087,
      "num_tokens": 1418083.0,
      "step": 480
    },
    {
      "entropy": 0.1822998046875,
      "epoch": 2.5548216644649933,
      "grad_norm": 0.5171620845794678,
      "learning_rate": 2.0326446630224656e-06,
      "loss": 0.19002891778945924,
      "mean_token_accuracy": 0.9404330343008042,
      "num_tokens": 1433006.0,
      "step": 485
    },
    {
      "entropy": 0.198486328125,
      "epoch": 2.581241743725231,
      "grad_norm": 0.6369736790657043,
      "learning_rate": 1.8078937319026655e-06,
      "loss": 0.20044972896575927,
      "mean_token_accuracy": 0.9355429112911224,
      "num_tokens": 1447394.0,
      "step": 490
    },
    {
      "entropy": 0.16258544921875,
      "epoch": 2.607661822985469,
      "grad_norm": 0.48347020149230957,
      "learning_rate": 1.5955103951488177e-06,
      "loss": 0.1729520797729492,
      "mean_token_accuracy": 0.9477747470140457,
      "num_tokens": 1462124.0,
      "step": 495
    },
    {
      "entropy": 0.144384765625,
      "epoch": 2.6340819022457067,
      "grad_norm": 0.5637415051460266,
      "learning_rate": 1.3956937620544408e-06,
      "loss": 0.15064214468002318,
      "mean_token_accuracy": 0.9508807033300399,
      "num_tokens": 1476343.0,
      "step": 500
    },
    {
      "epoch": 2.6340819022457067,
      "eval_entropy": 0.18974609375,
      "eval_loss": 0.21745847165584564,
      "eval_mean_token_accuracy": 0.9320169153966402,
      "eval_num_tokens": 1476343.0,
      "eval_runtime": 4.7158,
      "eval_samples_per_second": 40.078,
      "eval_steps_per_second": 20.145,
      "step": 500
    },
    {
      "entropy": 0.1888671875,
      "epoch": 2.6605019815059445,
      "grad_norm": 0.6523160934448242,
      "learning_rate": 1.2086311606333845e-06,
      "loss": 0.18436098098754883,
      "mean_token_accuracy": 0.9459800988435745,
      "num_tokens": 1490632.0,
      "step": 505
    },
    {
      "entropy": 0.17581787109375,
      "epoch": 2.6869220607661823,
      "grad_norm": 0.5741255879402161,
      "learning_rate": 1.0344979619998874e-06,
      "loss": 0.17965226173400878,
      "mean_token_accuracy": 0.9434189975261689,
      "num_tokens": 1505385.0,
      "step": 510
    },
    {
      "entropy": 0.1491943359375,
      "epoch": 2.71334214002642,
      "grad_norm": 0.6013323068618774,
      "learning_rate": 8.734574159582343e-07,
      "loss": 0.14173706769943237,
      "mean_token_accuracy": 0.9502143740653992,
      "num_tokens": 1520285.0,
      "step": 515
    },
    {
      "entropy": 0.14251708984375,
      "epoch": 2.739762219286658,
      "grad_norm": 0.4806106686592102,
      "learning_rate": 7.256604979561238e-07,
      "loss": 0.15166788101196288,
      "mean_token_accuracy": 0.9553648352622985,
      "num_tokens": 1535153.0,
      "step": 520
    },
    {
      "entropy": 0.1333984375,
      "epoch": 2.766182298546896,
      "grad_norm": 0.49979710578918457,
      "learning_rate": 5.912457675452465e-07,
      "loss": 0.12939850091934205,
      "mean_token_accuracy": 0.9568995028734207,
      "num_tokens": 1549952.0,
      "step": 525
    },
    {
      "entropy": 0.16025390625,
      "epoch": 2.7926023778071336,
      "grad_norm": 0.6719100475311279,
      "learning_rate": 4.703392384817656e-07,
      "loss": 0.16313141584396362,
      "mean_token_accuracy": 0.9468693345785141,
      "num_tokens": 1563426.0,
      "step": 530
    },
    {
      "entropy": 0.22427978515625,
      "epoch": 2.819022457067371,
      "grad_norm": 0.5740999579429626,
      "learning_rate": 3.630542605884657e-07,
      "loss": 0.23523077964782715,
      "mean_token_accuracy": 0.9297642588615418,
      "num_tokens": 1580594.0,
      "step": 535
    },
    {
      "entropy": 0.15897216796875,
      "epoch": 2.8454425363276092,
      "grad_norm": 0.6617240905761719,
      "learning_rate": 2.694914134893428e-07,
      "loss": 0.17223806381225587,
      "mean_token_accuracy": 0.9480368942022324,
      "num_tokens": 1594450.0,
      "step": 540
    },
    {
      "entropy": 0.17318115234375,
      "epoch": 2.8718626155878466,
      "grad_norm": 0.5243069529533386,
      "learning_rate": 1.8973841231624323e-07,
      "loss": 0.20006287097930908,
      "mean_token_accuracy": 0.9483350962400436,
      "num_tokens": 1609571.0,
      "step": 545
    },
    {
      "entropy": 0.1625244140625,
      "epoch": 2.8982826948480844,
      "grad_norm": 0.7395583987236023,
      "learning_rate": 1.2387002547596615e-07,
      "loss": 0.18141778707504272,
      "mean_token_accuracy": 0.941899985074997,
      "num_tokens": 1623621.0,
      "step": 550
    },
    {
      "entropy": 0.2183837890625,
      "epoch": 2.9247027741083222,
      "grad_norm": 0.5586023926734924,
      "learning_rate": 7.194800455490591e-08,
      "loss": 0.2549168825149536,
      "mean_token_accuracy": 0.928812712430954,
      "num_tokens": 1638889.0,
      "step": 555
    },
    {
      "entropy": 0.1492431640625,
      "epoch": 2.95112285336856,
      "grad_norm": 0.5419667959213257,
      "learning_rate": 3.40210264269758e-08,
      "loss": 0.14560272693634033,
      "mean_token_accuracy": 0.9526803165674209,
      "num_tokens": 1654140.0,
      "step": 560
    },
    {
      "entropy": 0.13837890625,
      "epoch": 2.977542932628798,
      "grad_norm": 0.5562270283699036,
      "learning_rate": 1.0124647619045302e-08,
      "loss": 0.13998043537139893,
      "mean_token_accuracy": 0.958392658829689,
      "num_tokens": 1668045.0,
      "step": 565
    },
    {
      "entropy": 0.19031479779411764,
      "epoch": 3.0,
      "grad_norm": 1.064183235168457,
      "learning_rate": 2.812709767163746e-10,
      "loss": 0.17594786882400512,
      "mean_token_accuracy": 0.9378625014249016,
      "num_tokens": 1680090.0,
      "step": 570
    },
    {
      "epoch": 3.0,
      "eval_entropy": 0.188671875,
      "eval_loss": 0.2172553390264511,
      "eval_mean_token_accuracy": 0.9322408381261323,
      "eval_num_tokens": 1680090.0,
      "eval_runtime": 4.7022,
      "eval_samples_per_second": 40.194,
      "eval_steps_per_second": 20.203,
      "step": 570
    },
    {
      "epoch": 3.0,
      "step": 570,
      "total_flos": 7.171819552165069e+16,
      "train_loss": 0.22585551466858178,
      "train_runtime": 751.1949,
      "train_samples_per_second": 6.042,
      "train_steps_per_second": 0.759
    }
  ],
  "logging_steps": 5,
  "max_steps": 570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.171819552165069e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}