{
  "best_metric": 0.11773921549320221,
  "best_model_checkpoint": "model_training/deepseek_prover_base_no_err/checkpoints-random-09-07-09-02/checkpoint-450",
  "epoch": 3.655837563451777,
  "eval_steps": 50,
  "global_step": 450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04060913705583756,
      "grad_norm": 117.39540100097656,
      "learning_rate": 1e-05,
      "loss": 15.7951,
      "step": 5
    },
    {
      "epoch": 0.08121827411167512,
      "grad_norm": 105.14662170410156,
      "learning_rate": 2e-05,
      "loss": 11.8462,
      "step": 10
    },
    {
      "epoch": 0.1218274111675127,
      "grad_norm": 43.610801696777344,
      "learning_rate": 3e-05,
      "loss": 3.3642,
      "step": 15
    },
    {
      "epoch": 0.16243654822335024,
      "grad_norm": 0.4269501864910126,
      "learning_rate": 4e-05,
      "loss": 0.1821,
      "step": 20
    },
    {
      "epoch": 0.20304568527918782,
      "grad_norm": 0.41744017601013184,
      "learning_rate": 5e-05,
      "loss": 0.4639,
      "step": 25
    },
    {
      "epoch": 0.2436548223350254,
      "grad_norm": 0.45930907130241394,
      "learning_rate": 6e-05,
      "loss": 0.1685,
      "step": 30
    },
    {
      "epoch": 0.28426395939086296,
      "grad_norm": 0.32693225145339966,
      "learning_rate": 7e-05,
      "loss": 0.1772,
      "step": 35
    },
    {
      "epoch": 0.3248730964467005,
      "grad_norm": 0.2242601364850998,
      "learning_rate": 8e-05,
      "loss": 0.1677,
      "step": 40
    },
    {
      "epoch": 0.36548223350253806,
      "grad_norm": 2.6285548210144043,
      "learning_rate": 9e-05,
      "loss": 0.1544,
      "step": 45
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 0.1525067687034607,
      "learning_rate": 0.0001,
      "loss": 0.1655,
      "step": 50
    },
    {
      "epoch": 0.40609137055837563,
      "eval_loss": 0.15786193311214447,
      "eval_runtime": 564.1924,
      "eval_samples_per_second": 1.772,
      "eval_steps_per_second": 0.222,
      "step": 50
    },
    {
      "epoch": 0.4467005076142132,
      "grad_norm": 0.10299257934093475,
      "learning_rate": 9.996842891446092e-05,
      "loss": 0.1514,
      "step": 55
    },
    {
      "epoch": 0.4873096446700508,
      "grad_norm": 0.11552461981773376,
      "learning_rate": 9.987375552718133e-05,
      "loss": 0.1505,
      "step": 60
    },
    {
      "epoch": 0.5279187817258884,
      "grad_norm": 0.07015743851661682,
      "learning_rate": 9.971609939582557e-05,
      "loss": 0.1556,
      "step": 65
    },
    {
      "epoch": 0.5685279187817259,
      "grad_norm": 0.0795639380812645,
      "learning_rate": 9.9495659615402e-05,
      "loss": 0.1515,
      "step": 70
    },
    {
      "epoch": 0.6091370558375635,
      "grad_norm": 0.07625389844179153,
      "learning_rate": 9.921271456683715e-05,
      "loss": 0.1394,
      "step": 75
    },
    {
      "epoch": 0.649746192893401,
      "grad_norm": 0.10126403719186783,
      "learning_rate": 9.886762156542428e-05,
      "loss": 0.1657,
      "step": 80
    },
    {
      "epoch": 0.6903553299492385,
      "grad_norm": 0.06794468313455582,
      "learning_rate": 9.846081640959007e-05,
      "loss": 0.1475,
      "step": 85
    },
    {
      "epoch": 0.7309644670050761,
      "grad_norm": 0.043415650725364685,
      "learning_rate": 9.79928128305494e-05,
      "loss": 0.1474,
      "step": 90
    },
    {
      "epoch": 0.7715736040609137,
      "grad_norm": 0.04360484331846237,
      "learning_rate": 9.746420184354334e-05,
      "loss": 0.1495,
      "step": 95
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 0.044484805315732956,
      "learning_rate": 9.687565100147939e-05,
      "loss": 0.1466,
      "step": 100
    },
    {
      "epoch": 0.8121827411167513,
      "eval_loss": 0.14690107107162476,
      "eval_runtime": 564.1055,
      "eval_samples_per_second": 1.773,
      "eval_steps_per_second": 0.222,
      "step": 100
    },
    {
      "epoch": 0.8527918781725888,
      "grad_norm": 0.0379624105989933,
      "learning_rate": 9.622790355191672e-05,
      "loss": 0.1393,
      "step": 105
    },
    {
      "epoch": 0.8934010152284264,
      "grad_norm": 0.046673040837049484,
      "learning_rate": 9.552177749846083e-05,
      "loss": 0.1408,
      "step": 110
    },
    {
      "epoch": 0.934010152284264,
      "grad_norm": 0.04505982622504234,
      "learning_rate": 9.475816456775313e-05,
      "loss": 0.1482,
      "step": 115
    },
    {
      "epoch": 0.9746192893401016,
      "grad_norm": 0.04212146997451782,
      "learning_rate": 9.393802908335977e-05,
      "loss": 0.141,
      "step": 120
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 0.044183164834976196,
      "learning_rate": 9.306240674798203e-05,
      "loss": 0.144,
      "step": 125
    },
    {
      "epoch": 1.0558375634517767,
      "grad_norm": 0.050254255533218384,
      "learning_rate": 9.213240333552589e-05,
      "loss": 0.1376,
      "step": 130
    },
    {
      "epoch": 1.0964467005076142,
      "grad_norm": 0.050791483372449875,
      "learning_rate": 9.114919329468282e-05,
      "loss": 0.1583,
      "step": 135
    },
    {
      "epoch": 1.1370558375634519,
      "grad_norm": 0.05157487466931343,
      "learning_rate": 9.011401826578492e-05,
      "loss": 0.1307,
      "step": 140
    },
    {
      "epoch": 1.1776649746192893,
      "grad_norm": 0.05917412042617798,
      "learning_rate": 8.902818551280758e-05,
      "loss": 0.134,
      "step": 145
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 0.047791987657547,
      "learning_rate": 8.789306627249985e-05,
      "loss": 0.1252,
      "step": 150
    },
    {
      "epoch": 1.218274111675127,
      "eval_loss": 0.1394234001636505,
      "eval_runtime": 564.1635,
      "eval_samples_per_second": 1.773,
      "eval_steps_per_second": 0.222,
      "step": 150
    },
    {
      "epoch": 1.2588832487309645,
      "grad_norm": 0.06823377311229706,
      "learning_rate": 8.6710094022727e-05,
      "loss": 0.1289,
      "step": 155
    },
    {
      "epoch": 1.299492385786802,
      "grad_norm": 0.10027821362018585,
      "learning_rate": 8.548076267221256e-05,
      "loss": 0.1345,
      "step": 160
    },
    {
      "epoch": 1.3401015228426396,
      "grad_norm": 0.08134859055280685,
      "learning_rate": 8.420662467396547e-05,
      "loss": 0.147,
      "step": 165
    },
    {
      "epoch": 1.380710659898477,
      "grad_norm": 0.0758972093462944,
      "learning_rate": 8.288928906477496e-05,
      "loss": 0.1328,
      "step": 170
    },
    {
      "epoch": 1.4213197969543148,
      "grad_norm": 0.07111922651529312,
      "learning_rate": 8.15304194332491e-05,
      "loss": 0.1311,
      "step": 175
    },
    {
      "epoch": 1.4619289340101522,
      "grad_norm": 0.05676901340484619,
      "learning_rate": 8.013173181896283e-05,
      "loss": 0.126,
      "step": 180
    },
    {
      "epoch": 1.50253807106599,
      "grad_norm": 0.058128587901592255,
      "learning_rate": 7.869499254536865e-05,
      "loss": 0.1265,
      "step": 185
    },
    {
      "epoch": 1.5431472081218274,
      "grad_norm": 0.07095002382993698,
      "learning_rate": 7.722201598920673e-05,
      "loss": 0.1194,
      "step": 190
    },
    {
      "epoch": 1.5837563451776648,
      "grad_norm": 0.06738153845071793,
      "learning_rate": 7.571466228923115e-05,
      "loss": 0.1274,
      "step": 195
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 0.08010146021842957,
      "learning_rate": 7.417483499714589e-05,
      "loss": 0.1396,
      "step": 200
    },
    {
      "epoch": 1.6243654822335025,
      "eval_loss": 0.1322861909866333,
      "eval_runtime": 564.0964,
      "eval_samples_per_second": 1.773,
      "eval_steps_per_second": 0.222,
      "step": 200
    },
    {
      "epoch": 1.665989847715736,
      "grad_norm": 0.1655184030532837,
      "learning_rate": 7.260447867371709e-05,
      "loss": 0.1362,
      "step": 205
    },
    {
      "epoch": 1.7065989847715737,
      "grad_norm": 0.07665544748306274,
      "learning_rate": 7.100557643309732e-05,
      "loss": 0.1194,
      "step": 210
    },
    {
      "epoch": 1.7472081218274111,
      "grad_norm": 0.07500068098306656,
      "learning_rate": 6.938014743846285e-05,
      "loss": 0.1319,
      "step": 215
    },
    {
      "epoch": 1.7878172588832486,
      "grad_norm": 0.0936647355556488,
      "learning_rate": 6.773024435212678e-05,
      "loss": 0.1164,
      "step": 220
    },
    {
      "epoch": 1.8284263959390863,
      "grad_norm": 0.08089397102594376,
      "learning_rate": 6.605795074334794e-05,
      "loss": 0.1242,
      "step": 225
    },
    {
      "epoch": 1.869035532994924,
      "grad_norm": 0.06260140240192413,
      "learning_rate": 6.436537845710903e-05,
      "loss": 0.1223,
      "step": 230
    },
    {
      "epoch": 1.9096446700507614,
      "grad_norm": 0.08003013581037521,
      "learning_rate": 6.265466494718732e-05,
      "loss": 0.1251,
      "step": 235
    },
    {
      "epoch": 1.950253807106599,
      "grad_norm": 0.09465450048446655,
      "learning_rate": 6.092797057688495e-05,
      "loss": 0.1254,
      "step": 240
    },
    {
      "epoch": 1.9908629441624366,
      "grad_norm": 0.1107378676533699,
      "learning_rate": 5.918747589082853e-05,
      "loss": 0.1198,
      "step": 245
    },
    {
      "epoch": 2.0314720812182743,
      "grad_norm": 0.08334667980670929,
      "learning_rate": 5.7435378861282585e-05,
      "loss": 0.1283,
      "step": 250
    },
    {
      "epoch": 2.0314720812182743,
      "eval_loss": 0.1270112842321396,
      "eval_runtime": 566.0408,
      "eval_samples_per_second": 1.767,
      "eval_steps_per_second": 0.221,
      "step": 250
    },
    {
      "epoch": 2.0720812182741115,
      "grad_norm": 0.07709399610757828,
      "learning_rate": 5.567389211245485e-05,
      "loss": 0.1204,
      "step": 255
    },
    {
      "epoch": 2.112690355329949,
      "grad_norm": 0.11390144377946854,
      "learning_rate": 5.390524012629824e-05,
      "loss": 0.1165,
      "step": 260
    },
    {
      "epoch": 2.153299492385787,
      "grad_norm": 0.09440046548843384,
      "learning_rate": 5.2131656433338506e-05,
      "loss": 0.1209,
      "step": 265
    },
    {
      "epoch": 2.1939086294416246,
      "grad_norm": 0.2260516881942749,
      "learning_rate": 5.035538079207488e-05,
      "loss": 0.1265,
      "step": 270
    },
    {
      "epoch": 2.234517766497462,
      "grad_norm": 0.08039233833551407,
      "learning_rate": 4.857865636051585e-05,
      "loss": 0.1231,
      "step": 275
    },
    {
      "epoch": 2.2751269035532995,
      "grad_norm": 0.13909466564655304,
      "learning_rate": 4.6803726863421725e-05,
      "loss": 0.1135,
      "step": 280
    },
    {
      "epoch": 2.315736040609137,
      "grad_norm": 0.12441807985305786,
      "learning_rate": 4.503283375883165e-05,
      "loss": 0.103,
      "step": 285
    },
    {
      "epoch": 2.3563451776649744,
      "grad_norm": 0.09395081549882889,
      "learning_rate": 4.326821340745304e-05,
      "loss": 0.1136,
      "step": 290
    },
    {
      "epoch": 2.396954314720812,
      "grad_norm": 0.09730109572410583,
      "learning_rate": 4.151209424848819e-05,
      "loss": 0.1133,
      "step": 295
    },
    {
      "epoch": 2.43756345177665,
      "grad_norm": 0.09149850159883499,
      "learning_rate": 3.976669398546451e-05,
      "loss": 0.1177,
      "step": 300
    },
    {
      "epoch": 2.43756345177665,
      "eval_loss": 0.12281998991966248,
      "eval_runtime": 566.7313,
      "eval_samples_per_second": 1.765,
      "eval_steps_per_second": 0.221,
      "step": 300
    },
    {
      "epoch": 2.4781725888324875,
      "grad_norm": 0.09771734476089478,
      "learning_rate": 3.803421678562213e-05,
      "loss": 0.117,
      "step": 305
    },
    {
      "epoch": 2.5187817258883247,
      "grad_norm": 0.11150355637073517,
      "learning_rate": 3.631685049639586e-05,
      "loss": 0.1121,
      "step": 310
    },
    {
      "epoch": 2.5593908629441624,
      "grad_norm": 0.09728321433067322,
      "learning_rate": 3.461676388250651e-05,
      "loss": 0.1123,
      "step": 315
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.1387196183204651,
      "learning_rate": 3.293610388715048e-05,
      "loss": 0.1091,
      "step": 320
    },
    {
      "epoch": 2.6406091370558373,
      "grad_norm": 0.08297745138406754,
      "learning_rate": 3.127699292074683e-05,
      "loss": 0.1123,
      "step": 325
    },
    {
      "epoch": 2.681218274111675,
      "grad_norm": 0.07975372672080994,
      "learning_rate": 2.964152618066508e-05,
      "loss": 0.117,
      "step": 330
    },
    {
      "epoch": 2.7218274111675127,
      "grad_norm": 0.09079435467720032,
      "learning_rate": 2.8031769005319147e-05,
      "loss": 0.1167,
      "step": 335
    },
    {
      "epoch": 2.7624365482233504,
      "grad_norm": 0.11697285622358322,
      "learning_rate": 2.6449754265968264e-05,
      "loss": 0.1149,
      "step": 340
    },
    {
      "epoch": 2.803045685279188,
      "grad_norm": 0.14380836486816406,
      "learning_rate": 2.4897479799518796e-05,
      "loss": 0.1078,
      "step": 345
    },
    {
      "epoch": 2.8436548223350253,
      "grad_norm": 0.08891697227954865,
      "learning_rate": 2.3376905885569182e-05,
      "loss": 0.1041,
      "step": 350
    },
    {
      "epoch": 2.8436548223350253,
      "eval_loss": 0.11983277648687363,
      "eval_runtime": 566.6646,
      "eval_samples_per_second": 1.765,
      "eval_steps_per_second": 0.221,
      "step": 350
    },
    {
      "epoch": 2.884263959390863,
      "grad_norm": 0.09689132869243622,
      "learning_rate": 2.1889952770883643e-05,
      "loss": 0.1212,
      "step": 355
    },
    {
      "epoch": 2.9248730964467002,
      "grad_norm": 0.08049995452165604,
      "learning_rate": 2.043849824442124e-05,
      "loss": 0.105,
      "step": 360
    },
    {
      "epoch": 2.965482233502538,
      "grad_norm": 0.0926508679986,
      "learning_rate": 1.9024375265982384e-05,
      "loss": 0.11,
      "step": 365
    },
    {
      "epoch": 3.0060913705583756,
      "grad_norm": 0.09762662649154663,
      "learning_rate": 1.764936965146773e-05,
      "loss": 0.1067,
      "step": 370
    },
    {
      "epoch": 3.0467005076142133,
      "grad_norm": 0.11719583719968796,
      "learning_rate": 1.631521781767214e-05,
      "loss": 0.0963,
      "step": 375
    },
    {
      "epoch": 3.087309644670051,
      "grad_norm": 0.09774573147296906,
      "learning_rate": 1.502360458946232e-05,
      "loss": 0.1081,
      "step": 380
    },
    {
      "epoch": 3.127918781725888,
      "grad_norm": 0.10405320674180984,
      "learning_rate": 1.3776161072106702e-05,
      "loss": 0.107,
      "step": 385
    },
    {
      "epoch": 3.168527918781726,
      "grad_norm": 0.10721355676651001,
      "learning_rate": 1.257446259144494e-05,
      "loss": 0.1101,
      "step": 390
    },
    {
      "epoch": 3.2091370558375636,
      "grad_norm": 0.12284258753061295,
      "learning_rate": 1.1420026704498077e-05,
      "loss": 0.1096,
      "step": 395
    },
    {
      "epoch": 3.249746192893401,
      "grad_norm": 0.09421534836292267,
      "learning_rate": 1.031431128303153e-05,
      "loss": 0.1059,
      "step": 400
    },
    {
      "epoch": 3.249746192893401,
      "eval_loss": 0.11832693219184875,
      "eval_runtime": 566.7299,
      "eval_samples_per_second": 1.765,
      "eval_steps_per_second": 0.221,
      "step": 400
    },
    {
      "epoch": 3.2903553299492385,
      "grad_norm": 0.09948204457759857,
      "learning_rate": 9.258712672491415e-06,
      "loss": 0.1129,
      "step": 405
    },
    {
      "epoch": 3.330964467005076,
      "grad_norm": 0.10511817038059235,
      "learning_rate": 8.254563928638893e-06,
      "loss": 0.1101,
      "step": 410
    },
    {
      "epoch": 3.371573604060914,
      "grad_norm": 0.11750485748052597,
      "learning_rate": 7.3031331341093915e-06,
      "loss": 0.1121,
      "step": 415
    },
    {
      "epoch": 3.412182741116751,
      "grad_norm": 0.09111586213111877,
      "learning_rate": 6.405621797022848e-06,
      "loss": 0.1102,
      "step": 420
    },
    {
      "epoch": 3.452791878172589,
      "grad_norm": 0.08644920587539673,
      "learning_rate": 5.563163333667099e-06,
      "loss": 0.0995,
      "step": 425
    },
    {
      "epoch": 3.4934010152284265,
      "grad_norm": 0.10021346062421799,
      "learning_rate": 4.776821637170526e-06,
      "loss": 0.0984,
      "step": 430
    },
    {
      "epoch": 3.5340101522842637,
      "grad_norm": 0.09591208398342133,
      "learning_rate": 4.047589733971646e-06,
      "loss": 0.1102,
      "step": 435
    },
    {
      "epoch": 3.5746192893401014,
      "grad_norm": 0.11921706050634384,
      "learning_rate": 3.376388529782215e-06,
      "loss": 0.1085,
      "step": 440
    },
    {
      "epoch": 3.615228426395939,
      "grad_norm": 0.08839091658592224,
      "learning_rate": 2.7640656466274782e-06,
      "loss": 0.1082,
      "step": 445
    },
    {
      "epoch": 3.655837563451777,
      "grad_norm": 0.09160158783197403,
      "learning_rate": 2.2113943524323167e-06,
      "loss": 0.1041,
      "step": 450
    },
    {
      "epoch": 3.655837563451777,
      "eval_loss": 0.11773921549320221,
      "eval_runtime": 566.7284,
      "eval_samples_per_second": 1.765,
      "eval_steps_per_second": 0.221,
      "step": 450
    }
  ],
  "logging_steps": 5,
  "max_steps": 492,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8705510622704435e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}