{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 30.0,
  "eval_steps": 500,
  "global_step": 19710,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.76103500761035,
      "grad_norm": 1.1888039112091064,
      "learning_rate": 0.0009746321664129883,
      "loss": 5.3071,
      "max_memory_allocated (GB)": 5.75,
      "memory_allocated (GB)": 3.2,
      "step": 500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.08006198116451355,
      "eval_loss": 6.240572929382324,
      "eval_runtime": 1138.1454,
      "eval_samples_per_second": 590.273,
      "eval_steps_per_second": 0.577,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 657,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.5220700152207,
      "grad_norm": 1.0501078367233276,
      "learning_rate": 0.0009492643328259766,
      "loss": 3.1366,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 1000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.10465945339281382,
      "eval_loss": 5.848066329956055,
      "eval_runtime": 1134.9424,
      "eval_samples_per_second": 591.939,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 1314,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.2831050228310503,
      "grad_norm": 0.972637414932251,
      "learning_rate": 0.0009238964992389651,
      "loss": 2.6048,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 1500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.12382538697294054,
      "eval_loss": 5.552162170410156,
      "eval_runtime": 1132.6347,
      "eval_samples_per_second": 593.145,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 1971,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.0441400304414,
      "grad_norm": 0.8711762428283691,
      "learning_rate": 0.0008985286656519534,
      "loss": 2.3103,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 2000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.8051750380517504,
      "grad_norm": 0.859586775302887,
      "learning_rate": 0.0008731608320649417,
      "loss": 1.9918,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 2500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.1300875089496098,
      "eval_loss": 5.555095672607422,
      "eval_runtime": 1133.2064,
      "eval_samples_per_second": 592.846,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 2628,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.566210045662101,
      "grad_norm": 0.8556590676307678,
      "learning_rate": 0.00084779299847793,
      "loss": 1.8353,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 3000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.141504308464954,
      "eval_loss": 5.414204120635986,
      "eval_runtime": 1132.3487,
      "eval_samples_per_second": 593.295,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 3285,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 5.327245053272451,
      "grad_norm": 0.8324838280677795,
      "learning_rate": 0.0008224251648909183,
      "loss": 1.7262,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 3500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.14951988413511416,
      "eval_loss": 5.40610933303833,
      "eval_runtime": 1133.7065,
      "eval_samples_per_second": 592.585,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 3942,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 6.0882800608828,
      "grad_norm": 0.7686742544174194,
      "learning_rate": 0.0007970573313039067,
      "loss": 1.6381,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 4000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 6.8493150684931505,
      "grad_norm": 0.7999989986419678,
      "learning_rate": 0.000771689497716895,
      "loss": 1.5135,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 4500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.14675722704248328,
      "eval_loss": 5.426120758056641,
      "eval_runtime": 1131.9083,
      "eval_samples_per_second": 593.526,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 4599,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 7.610350076103501,
      "grad_norm": 0.8491269946098328,
      "learning_rate": 0.0007463216641298833,
      "loss": 1.4225,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 5000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.15733153522462218,
      "eval_loss": 5.333346843719482,
      "eval_runtime": 1130.7533,
      "eval_samples_per_second": 594.132,
      "eval_steps_per_second": 0.581,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 5256,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 8.37138508371385,
      "grad_norm": 0.7948514819145203,
      "learning_rate": 0.0007209538305428717,
      "loss": 1.354,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 5500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.16383479429666115,
      "eval_loss": 5.220494747161865,
      "eval_runtime": 1131.3928,
      "eval_samples_per_second": 593.796,
      "eval_steps_per_second": 0.581,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 5913,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 9.132420091324201,
      "grad_norm": 0.7878388166427612,
      "learning_rate": 0.00069558599695586,
      "loss": 1.3172,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 6000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 9.89345509893455,
      "grad_norm": 0.7506768703460693,
      "learning_rate": 0.0006702181633688484,
      "loss": 1.2511,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 6500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.17084414356885877,
      "eval_loss": 5.212928295135498,
      "eval_runtime": 1133.5622,
      "eval_samples_per_second": 592.66,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 6570,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 10.654490106544902,
      "grad_norm": 0.7633622288703918,
      "learning_rate": 0.0006448503297818367,
      "loss": 1.1742,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 7000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.17239664968287494,
      "eval_loss": 5.200212001800537,
      "eval_runtime": 1129.7037,
      "eval_samples_per_second": 594.684,
      "eval_steps_per_second": 0.582,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 7227,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 11.415525114155251,
      "grad_norm": 0.7618717551231384,
      "learning_rate": 0.000619482496194825,
      "loss": 1.1342,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 7500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.17819584797645788,
      "eval_loss": 5.163547039031982,
      "eval_runtime": 1128.6949,
      "eval_samples_per_second": 595.216,
      "eval_steps_per_second": 0.582,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 7884,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 12.1765601217656,
      "grad_norm": 0.7090550661087036,
      "learning_rate": 0.0005941146626078133,
      "loss": 1.1111,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 8000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 12.937595129375952,
      "grad_norm": 0.7710525393486023,
      "learning_rate": 0.0005687468290208016,
      "loss": 1.0711,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 8500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.17787879735106435,
      "eval_loss": 5.143550872802734,
      "eval_runtime": 1133.3031,
      "eval_samples_per_second": 592.796,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 8541,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 13.698630136986301,
      "grad_norm": 0.7961007952690125,
      "learning_rate": 0.00054337899543379,
      "loss": 0.9971,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 9000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.18167893935402052,
      "eval_loss": 5.107571125030518,
      "eval_runtime": 1132.5925,
      "eval_samples_per_second": 593.167,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 9198,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 14.459665144596652,
      "grad_norm": 0.7081454992294312,
      "learning_rate": 0.0005180111618467784,
      "loss": 0.9774,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 9500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.19349168002595946,
      "eval_loss": 4.9076433181762695,
      "eval_runtime": 1135.4783,
      "eval_samples_per_second": 591.66,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 9855,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 15.220700152207002,
      "grad_norm": 0.706643283367157,
      "learning_rate": 0.0004926433282597666,
      "loss": 0.9457,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 10000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 15.981735159817351,
      "grad_norm": 0.7469919323921204,
      "learning_rate": 0.0004672754946727549,
      "loss": 0.9174,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 10500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.18904255176632923,
      "eval_loss": 5.03179407119751,
      "eval_runtime": 1127.5619,
      "eval_samples_per_second": 595.814,
      "eval_steps_per_second": 0.583,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 10512,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 16.7427701674277,
      "grad_norm": 0.7092038989067078,
      "learning_rate": 0.0004419076610857433,
      "loss": 0.8675,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 11000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.19512456517176552,
      "eval_loss": 5.039154052734375,
      "eval_runtime": 1129.1255,
      "eval_samples_per_second": 594.989,
      "eval_steps_per_second": 0.582,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 11169,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 17.503805175038053,
      "grad_norm": 0.722985029220581,
      "learning_rate": 0.0004165398274987316,
      "loss": 0.8499,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 11500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.19776069971435672,
      "eval_loss": 5.024279594421387,
      "eval_runtime": 1134.2545,
      "eval_samples_per_second": 592.298,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 11826,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 18.264840182648403,
      "grad_norm": 0.703146755695343,
      "learning_rate": 0.0003911719939117199,
      "loss": 0.8262,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 12000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.19716678797946466,
      "eval_loss": 5.084349632263184,
      "eval_runtime": 1134.1402,
      "eval_samples_per_second": 592.358,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 12483,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 19.025875190258752,
      "grad_norm": 0.6394225358963013,
      "learning_rate": 0.00036580416032470827,
      "loss": 0.8039,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 12500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 19.7869101978691,
      "grad_norm": 0.6850036978721619,
      "learning_rate": 0.0003404363267376966,
      "loss": 0.7623,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 13000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.20482214650715894,
      "eval_loss": 5.0004353523254395,
      "eval_runtime": 1132.7333,
      "eval_samples_per_second": 593.094,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 13140,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 20.54794520547945,
      "grad_norm": 0.7249587178230286,
      "learning_rate": 0.00031506849315068495,
      "loss": 0.7481,
      "max_memory_allocated (GB)": 60.21,
      "memory_allocated (GB)": 3.2,
      "step": 13500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.21318007731272057,
      "eval_loss": 4.842759609222412,
      "eval_runtime": 1132.229,
      "eval_samples_per_second": 593.358,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 13797,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 21.308980213089804,
      "grad_norm": 0.651644766330719,
      "learning_rate": 0.00028970065956367326,
      "loss": 0.7284,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 14000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.2148576174761877,
      "eval_loss": 4.846081733703613,
      "eval_runtime": 1134.6589,
      "eval_samples_per_second": 592.087,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 14454,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 22.070015220700153,
      "grad_norm": 0.6403504610061646,
      "learning_rate": 0.00026433282597666157,
      "loss": 0.706,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 14500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 22.831050228310502,
      "grad_norm": 0.6770262718200684,
      "learning_rate": 0.0002389649923896499,
      "loss": 0.6834,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 15000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.2159159413947548,
      "eval_loss": 4.8741374015808105,
      "eval_runtime": 1130.2975,
      "eval_samples_per_second": 594.372,
      "eval_steps_per_second": 0.581,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 15111,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 23.59208523592085,
      "grad_norm": 0.6229885816574097,
      "learning_rate": 0.00021359715880263824,
      "loss": 0.6591,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 15500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.2186681789832648,
      "eval_loss": 4.89931058883667,
      "eval_runtime": 1133.9373,
      "eval_samples_per_second": 592.464,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 15768,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 24.3531202435312,
      "grad_norm": 0.6464186310768127,
      "learning_rate": 0.00018822932521562658,
      "loss": 0.6447,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 16000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.21962528486180016,
      "eval_loss": 4.8415398597717285,
      "eval_runtime": 1126.1392,
      "eval_samples_per_second": 596.567,
      "eval_steps_per_second": 0.583,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 16425,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 25.114155251141554,
      "grad_norm": 0.695124626159668,
      "learning_rate": 0.00016286149162861492,
      "loss": 0.6323,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 16500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 25.875190258751903,
      "grad_norm": 0.7219062447547913,
      "learning_rate": 0.00013749365804160323,
      "loss": 0.6107,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 17000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.22164369166008005,
      "eval_loss": 4.859982967376709,
      "eval_runtime": 1131.8158,
      "eval_samples_per_second": 593.574,
      "eval_steps_per_second": 0.58,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 17082,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 26.636225266362253,
      "grad_norm": 0.6680580377578735,
      "learning_rate": 0.00011212582445459158,
      "loss": 0.5958,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 17500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.22447184277861382,
      "eval_loss": 4.839137554168701,
      "eval_runtime": 1135.3766,
      "eval_samples_per_second": 591.713,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 17739,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 27.397260273972602,
      "grad_norm": 0.6511676907539368,
      "learning_rate": 8.67579908675799e-05,
      "loss": 0.5836,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 18000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.22654234709749826,
      "eval_loss": 4.856111526489258,
      "eval_runtime": 1131.6209,
      "eval_samples_per_second": 593.677,
      "eval_steps_per_second": 0.581,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 18396,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 28.15829528158295,
      "grad_norm": 0.6694862842559814,
      "learning_rate": 6.139015728056824e-05,
      "loss": 0.5713,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 18500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 28.919330289193304,
      "grad_norm": 0.6698545813560486,
      "learning_rate": 3.6022323693556566e-05,
      "loss": 0.5547,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 19000,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.2294940437648943,
      "eval_loss": 4.793288230895996,
      "eval_runtime": 1134.2517,
      "eval_samples_per_second": 592.3,
      "eval_steps_per_second": 0.579,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 19053,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 29.680365296803654,
      "grad_norm": 0.5940834879875183,
      "learning_rate": 1.06544901065449e-05,
      "loss": 0.547,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 19500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.22931691219483877,
      "eval_loss": 4.809002876281738,
      "eval_runtime": 1130.6604,
      "eval_samples_per_second": 594.181,
      "eval_steps_per_second": 0.581,
      "max_memory_allocated (GB)": 60.24,
      "memory_allocated (GB)": 3.2,
      "step": 19710,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 30.0,
      "max_memory_allocated (GB)": 1.42,
      "memory_allocated (GB)": 1.42,
      "step": 19710,
      "total_flos": 2.9333313524800244e+21,
      "total_memory_available (GB)": 94.62,
      "train_loss": 0.0,
      "train_runtime": 0.2168,
      "train_samples_per_second": 92964947.027,
      "train_steps_per_second": 90914.595
    }
  ],
  "logging_steps": 500,
  "max_steps": 19710,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.9333313524800244e+21,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}