{
  "best_global_step": 500,
  "best_metric": 0.2209562063217163,
  "best_model_checkpoint": "/dfs/project/kgrlm/akhatua/CodeConflictBenchmark/finetuned_model/checkpoint-500",
  "epoch": 2.564102564102564,
  "eval_steps": 50,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.0350996673107147,
      "epoch": 0.05128205128205128,
      "grad_norm": 1.3478177785873413,
      "learning_rate": 1.8e-05,
      "loss": 1.5545,
      "mean_token_accuracy": 0.6928696542978287,
      "num_tokens": 45381.0,
      "step": 10
    },
    {
      "entropy": 1.0789292216300965,
      "epoch": 0.10256410256410256,
      "grad_norm": 1.0828419923782349,
      "learning_rate": 3.8e-05,
      "loss": 1.3792,
      "mean_token_accuracy": 0.70970538854599,
      "num_tokens": 90953.0,
      "step": 20
    },
    {
      "entropy": 0.9691586464643478,
      "epoch": 0.15384615384615385,
      "grad_norm": 1.148255467414856,
      "learning_rate": 5.8e-05,
      "loss": 0.9285,
      "mean_token_accuracy": 0.7836786180734634,
      "num_tokens": 140176.0,
      "step": 30
    },
    {
      "entropy": 0.6125603139400482,
      "epoch": 0.20512820512820512,
      "grad_norm": 0.7212587594985962,
      "learning_rate": 7.800000000000001e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8730687618255615,
      "num_tokens": 188322.0,
      "step": 40
    },
    {
      "entropy": 0.37974261194467546,
      "epoch": 0.2564102564102564,
      "grad_norm": 0.555002748966217,
      "learning_rate": 9.8e-05,
      "loss": 0.3664,
      "mean_token_accuracy": 0.9067715287208558,
      "num_tokens": 234634.0,
      "step": 50
    },
    {
      "epoch": 0.2564102564102564,
      "eval_entropy": 0.33798640830950305,
      "eval_loss": 0.33060547709465027,
      "eval_mean_token_accuracy": 0.9126620414582166,
      "eval_num_tokens": 234634.0,
      "eval_runtime": 3.3954,
      "eval_samples_per_second": 51.246,
      "eval_steps_per_second": 12.959,
      "step": 50
    },
    {
      "entropy": 0.3335741743445396,
      "epoch": 0.3076923076923077,
      "grad_norm": 0.40053829550743103,
      "learning_rate": 0.000118,
      "loss": 0.3268,
      "mean_token_accuracy": 0.9144781798124313,
      "num_tokens": 282032.0,
      "step": 60
    },
    {
      "entropy": 0.3074036821722984,
      "epoch": 0.358974358974359,
      "grad_norm": 0.429716557264328,
      "learning_rate": 0.000138,
      "loss": 0.3003,
      "mean_token_accuracy": 0.9197664111852646,
      "num_tokens": 330105.0,
      "step": 70
    },
    {
      "entropy": 0.28681637942790983,
      "epoch": 0.41025641025641024,
      "grad_norm": 0.39301180839538574,
      "learning_rate": 0.00015800000000000002,
      "loss": 0.2728,
      "mean_token_accuracy": 0.9240950107574463,
      "num_tokens": 376760.0,
      "step": 80
    },
    {
      "entropy": 0.2691986866295338,
      "epoch": 0.46153846153846156,
      "grad_norm": 0.4236898124217987,
      "learning_rate": 0.00017800000000000002,
      "loss": 0.2648,
      "mean_token_accuracy": 0.9252654194831849,
      "num_tokens": 423748.0,
      "step": 90
    },
    {
      "entropy": 0.253852054476738,
      "epoch": 0.5128205128205128,
      "grad_norm": 0.39244791865348816,
      "learning_rate": 0.00019800000000000002,
      "loss": 0.2489,
      "mean_token_accuracy": 0.9291503459215165,
      "num_tokens": 470327.0,
      "step": 100
    },
    {
      "epoch": 0.5128205128205128,
      "eval_entropy": 0.249941445548426,
      "eval_loss": 0.2573370337486267,
      "eval_mean_token_accuracy": 0.926737438548695,
      "eval_num_tokens": 470327.0,
      "eval_runtime": 3.3989,
      "eval_samples_per_second": 51.192,
      "eval_steps_per_second": 12.945,
      "step": 100
    },
    {
      "entropy": 0.24774221405386926,
      "epoch": 0.5641025641025641,
      "grad_norm": 0.3382515013217926,
      "learning_rate": 0.00019628865979381442,
      "loss": 0.2523,
      "mean_token_accuracy": 0.9275001347064972,
      "num_tokens": 517684.0,
      "step": 110
    },
    {
      "entropy": 0.26822882816195487,
      "epoch": 0.6153846153846154,
      "grad_norm": 0.3561689555644989,
      "learning_rate": 0.00019216494845360825,
      "loss": 0.2609,
      "mean_token_accuracy": 0.925196173787117,
      "num_tokens": 563536.0,
      "step": 120
    },
    {
      "entropy": 0.2486445650458336,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.40442419052124023,
      "learning_rate": 0.00018804123711340206,
      "loss": 0.2474,
      "mean_token_accuracy": 0.9291001617908478,
      "num_tokens": 608035.0,
      "step": 130
    },
    {
      "entropy": 0.25348711013793945,
      "epoch": 0.717948717948718,
      "grad_norm": 0.33539682626724243,
      "learning_rate": 0.00018391752577319586,
      "loss": 0.2456,
      "mean_token_accuracy": 0.9287373900413514,
      "num_tokens": 655355.0,
      "step": 140
    },
    {
      "entropy": 0.2404498890042305,
      "epoch": 0.7692307692307693,
      "grad_norm": 0.3772445619106293,
      "learning_rate": 0.0001797938144329897,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9304307967424392,
      "num_tokens": 700330.0,
      "step": 150
    },
    {
      "epoch": 0.7692307692307693,
      "eval_entropy": 0.2422055324370211,
      "eval_loss": 0.24134406447410583,
      "eval_mean_token_accuracy": 0.9292049421505495,
      "eval_num_tokens": 700330.0,
      "eval_runtime": 3.3903,
      "eval_samples_per_second": 51.323,
      "eval_steps_per_second": 12.978,
      "step": 150
    },
    {
      "entropy": 0.23829926177859306,
      "epoch": 0.8205128205128205,
      "grad_norm": 0.3463786542415619,
      "learning_rate": 0.0001756701030927835,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9308693438768387,
      "num_tokens": 745583.0,
      "step": 160
    },
    {
      "entropy": 0.23628849536180496,
      "epoch": 0.8717948717948718,
      "grad_norm": 0.3511769771575928,
      "learning_rate": 0.0001715463917525773,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9328561395406723,
      "num_tokens": 793508.0,
      "step": 170
    },
    {
      "entropy": 0.2514991283416748,
      "epoch": 0.9230769230769231,
      "grad_norm": 0.3281315565109253,
      "learning_rate": 0.00016742268041237114,
      "loss": 0.245,
      "mean_token_accuracy": 0.9293459743261338,
      "num_tokens": 838666.0,
      "step": 180
    },
    {
      "entropy": 0.23273598849773408,
      "epoch": 0.9743589743589743,
      "grad_norm": 0.3585456311702728,
      "learning_rate": 0.00016329896907216494,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9312855184078217,
      "num_tokens": 884925.0,
      "step": 190
    },
    {
      "entropy": 0.23954743817448615,
      "epoch": 1.0256410256410255,
      "grad_norm": 0.30551019310951233,
      "learning_rate": 0.00015917525773195875,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9325341433286667,
      "num_tokens": 929033.0,
      "step": 200
    },
    {
      "epoch": 1.0256410256410255,
      "eval_entropy": 0.2335269803350622,
      "eval_loss": 0.2343294471502304,
      "eval_mean_token_accuracy": 0.9310184608806263,
      "eval_num_tokens": 929033.0,
      "eval_runtime": 3.3954,
      "eval_samples_per_second": 51.246,
      "eval_steps_per_second": 12.959,
      "step": 200
    },
    {
      "entropy": 0.2196498766541481,
      "epoch": 1.0769230769230769,
      "grad_norm": 0.3471911549568176,
      "learning_rate": 0.00015505154639175258,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9367556869983673,
      "num_tokens": 975806.0,
      "step": 210
    },
    {
      "entropy": 0.21442087069153787,
      "epoch": 1.1282051282051282,
      "grad_norm": 0.3200732469558716,
      "learning_rate": 0.00015092783505154638,
      "loss": 0.2072,
      "mean_token_accuracy": 0.937502408027649,
      "num_tokens": 1022737.0,
      "step": 220
    },
    {
      "entropy": 0.21722501143813133,
      "epoch": 1.1794871794871795,
      "grad_norm": 0.3441697657108307,
      "learning_rate": 0.0001468041237113402,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9341749101877213,
      "num_tokens": 1069470.0,
      "step": 230
    },
    {
      "entropy": 0.21757920682430268,
      "epoch": 1.2307692307692308,
      "grad_norm": 0.33167675137519836,
      "learning_rate": 0.00014268041237113402,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9353529185056686,
      "num_tokens": 1115922.0,
      "step": 240
    },
    {
      "entropy": 0.2290858395397663,
      "epoch": 1.282051282051282,
      "grad_norm": 0.3299613893032074,
      "learning_rate": 0.00013855670103092783,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9326711863279342,
      "num_tokens": 1163335.0,
      "step": 250
    },
    {
      "epoch": 1.282051282051282,
      "eval_entropy": 0.22082097286527808,
      "eval_loss": 0.2295188456773758,
      "eval_mean_token_accuracy": 0.9324770271778107,
      "eval_num_tokens": 1163335.0,
      "eval_runtime": 3.3967,
      "eval_samples_per_second": 51.227,
      "eval_steps_per_second": 12.954,
      "step": 250
    },
    {
      "entropy": 0.20567527562379836,
      "epoch": 1.3333333333333333,
      "grad_norm": 0.34776246547698975,
      "learning_rate": 0.00013443298969072166,
      "loss": 0.2102,
      "mean_token_accuracy": 0.9354015827178955,
      "num_tokens": 1210637.0,
      "step": 260
    },
    {
      "entropy": 0.2168171338737011,
      "epoch": 1.3846153846153846,
      "grad_norm": 0.34032920002937317,
      "learning_rate": 0.00013030927835051546,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9366940945386887,
      "num_tokens": 1258602.0,
      "step": 270
    },
    {
      "entropy": 0.2166948698461056,
      "epoch": 1.435897435897436,
      "grad_norm": 0.34751859307289124,
      "learning_rate": 0.00012618556701030927,
      "loss": 0.213,
      "mean_token_accuracy": 0.9365024119615555,
      "num_tokens": 1305357.0,
      "step": 280
    },
    {
      "entropy": 0.2134072668850422,
      "epoch": 1.4871794871794872,
      "grad_norm": 0.33452409505844116,
      "learning_rate": 0.00012206185567010311,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9365908294916153,
      "num_tokens": 1351788.0,
      "step": 290
    },
    {
      "entropy": 0.2126854807138443,
      "epoch": 1.5384615384615383,
      "grad_norm": 0.3085568845272064,
      "learning_rate": 0.00011793814432989692,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9375448822975159,
      "num_tokens": 1397353.0,
      "step": 300
    },
    {
      "epoch": 1.5384615384615383,
      "eval_entropy": 0.21527046744119038,
      "eval_loss": 0.22627197206020355,
      "eval_mean_token_accuracy": 0.9322315928610888,
      "eval_num_tokens": 1397353.0,
      "eval_runtime": 3.3989,
      "eval_samples_per_second": 51.193,
      "eval_steps_per_second": 12.945,
      "step": 300
    },
    {
      "entropy": 0.21225061118602753,
      "epoch": 1.5897435897435899,
      "grad_norm": 0.32770541310310364,
      "learning_rate": 0.00011381443298969072,
      "loss": 0.2127,
      "mean_token_accuracy": 0.935514223575592,
      "num_tokens": 1444470.0,
      "step": 310
    },
    {
      "entropy": 0.2156527377665043,
      "epoch": 1.641025641025641,
      "grad_norm": 0.3466593623161316,
      "learning_rate": 0.00010969072164948456,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9328227639198303,
      "num_tokens": 1488716.0,
      "step": 320
    },
    {
      "entropy": 0.2182770349085331,
      "epoch": 1.6923076923076923,
      "grad_norm": 0.31441858410835266,
      "learning_rate": 0.00010556701030927836,
      "loss": 0.2141,
      "mean_token_accuracy": 0.9356919556856156,
      "num_tokens": 1535315.0,
      "step": 330
    },
    {
      "entropy": 0.20888199508190156,
      "epoch": 1.7435897435897436,
      "grad_norm": 0.35723230242729187,
      "learning_rate": 0.00010144329896907217,
      "loss": 0.2059,
      "mean_token_accuracy": 0.9381601482629776,
      "num_tokens": 1580453.0,
      "step": 340
    },
    {
      "entropy": 0.19662269279360772,
      "epoch": 1.7948717948717947,
      "grad_norm": 0.32357287406921387,
      "learning_rate": 9.731958762886598e-05,
      "loss": 0.1946,
      "mean_token_accuracy": 0.9412949442863464,
      "num_tokens": 1626061.0,
      "step": 350
    },
    {
      "epoch": 1.7948717948717947,
      "eval_entropy": 0.20560899884863335,
      "eval_loss": 0.22450383007526398,
      "eval_mean_token_accuracy": 0.9330938959663565,
      "eval_num_tokens": 1626061.0,
      "eval_runtime": 3.3963,
      "eval_samples_per_second": 51.232,
      "eval_steps_per_second": 12.955,
      "step": 350
    },
    {
      "entropy": 0.21225450485944747,
      "epoch": 1.8461538461538463,
      "grad_norm": 0.32507598400115967,
      "learning_rate": 9.31958762886598e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9344300121068955,
      "num_tokens": 1672735.0,
      "step": 360
    },
    {
      "entropy": 0.2201077848672867,
      "epoch": 1.8974358974358974,
      "grad_norm": 0.3187689483165741,
      "learning_rate": 8.907216494845362e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9343109726905823,
      "num_tokens": 1721643.0,
      "step": 370
    },
    {
      "entropy": 0.19808203503489494,
      "epoch": 1.9487179487179487,
      "grad_norm": 0.31342393159866333,
      "learning_rate": 8.494845360824743e-05,
      "loss": 0.2018,
      "mean_token_accuracy": 0.9375469982624054,
      "num_tokens": 1767639.0,
      "step": 380
    },
    {
      "entropy": 0.21093383952975273,
      "epoch": 2.0,
      "grad_norm": 0.3826819062232971,
      "learning_rate": 8.082474226804125e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9374133050441742,
      "num_tokens": 1813280.0,
      "step": 390
    },
    {
      "entropy": 0.2017211489379406,
      "epoch": 2.051282051282051,
      "grad_norm": 0.33081358671188354,
      "learning_rate": 7.670103092783506e-05,
      "loss": 0.1857,
      "mean_token_accuracy": 0.9430070728063583,
      "num_tokens": 1860011.0,
      "step": 400
    },
    {
      "epoch": 2.051282051282051,
      "eval_entropy": 0.19915587759830736,
      "eval_loss": 0.2229449599981308,
      "eval_mean_token_accuracy": 0.9337983199141242,
      "eval_num_tokens": 1860011.0,
      "eval_runtime": 3.3893,
      "eval_samples_per_second": 51.339,
      "eval_steps_per_second": 12.982,
      "step": 400
    },
    {
      "entropy": 0.1856864832341671,
      "epoch": 2.1025641025641026,
      "grad_norm": 0.33279746770858765,
      "learning_rate": 7.257731958762887e-05,
      "loss": 0.1794,
      "mean_token_accuracy": 0.9443148285150528,
      "num_tokens": 1905712.0,
      "step": 410
    },
    {
      "entropy": 0.18336084187030793,
      "epoch": 2.1538461538461537,
      "grad_norm": 0.3445015549659729,
      "learning_rate": 6.845360824742269e-05,
      "loss": 0.1867,
      "mean_token_accuracy": 0.9425998359918595,
      "num_tokens": 1953094.0,
      "step": 420
    },
    {
      "entropy": 0.18557853624224663,
      "epoch": 2.2051282051282053,
      "grad_norm": 0.3739573061466217,
      "learning_rate": 6.43298969072165e-05,
      "loss": 0.1793,
      "mean_token_accuracy": 0.9447199493646622,
      "num_tokens": 1998254.0,
      "step": 430
    },
    {
      "entropy": 0.18786199390888214,
      "epoch": 2.2564102564102564,
      "grad_norm": 0.38025277853012085,
      "learning_rate": 6.020618556701031e-05,
      "loss": 0.1858,
      "mean_token_accuracy": 0.9422474384307862,
      "num_tokens": 2044070.0,
      "step": 440
    },
    {
      "entropy": 0.19071117043495178,
      "epoch": 2.3076923076923075,
      "grad_norm": 0.32729047536849976,
      "learning_rate": 5.608247422680413e-05,
      "loss": 0.1909,
      "mean_token_accuracy": 0.9416741698980331,
      "num_tokens": 2092654.0,
      "step": 450
    },
    {
      "epoch": 2.3076923076923075,
      "eval_entropy": 0.1941434680060907,
      "eval_loss": 0.22272132337093353,
      "eval_mean_token_accuracy": 0.9333756877617403,
      "eval_num_tokens": 2092654.0,
      "eval_runtime": 3.3838,
      "eval_samples_per_second": 51.421,
      "eval_steps_per_second": 13.003,
      "step": 450
    },
    {
      "entropy": 0.1821839489042759,
      "epoch": 2.358974358974359,
      "grad_norm": 0.3813530504703522,
      "learning_rate": 5.195876288659794e-05,
      "loss": 0.1771,
      "mean_token_accuracy": 0.9454336315393448,
      "num_tokens": 2141205.0,
      "step": 460
    },
    {
      "entropy": 0.1962843805551529,
      "epoch": 2.41025641025641,
      "grad_norm": 0.36779674887657166,
      "learning_rate": 4.783505154639176e-05,
      "loss": 0.1957,
      "mean_token_accuracy": 0.938510924577713,
      "num_tokens": 2189080.0,
      "step": 470
    },
    {
      "entropy": 0.19173899814486503,
      "epoch": 2.4615384615384617,
      "grad_norm": 0.3630339503288269,
      "learning_rate": 4.371134020618557e-05,
      "loss": 0.1924,
      "mean_token_accuracy": 0.939791002869606,
      "num_tokens": 2235899.0,
      "step": 480
    },
    {
      "entropy": 0.1869501829147339,
      "epoch": 2.5128205128205128,
      "grad_norm": 0.3583838939666748,
      "learning_rate": 3.9587628865979384e-05,
      "loss": 0.1806,
      "mean_token_accuracy": 0.943246477842331,
      "num_tokens": 2281375.0,
      "step": 490
    },
    {
      "entropy": 0.1777109794318676,
      "epoch": 2.564102564102564,
      "grad_norm": 0.3346199691295624,
      "learning_rate": 3.54639175257732e-05,
      "loss": 0.1728,
      "mean_token_accuracy": 0.9454073548316956,
      "num_tokens": 2327070.0,
      "step": 500
    },
    {
      "epoch": 2.564102564102564,
      "eval_entropy": 0.18993931737813083,
      "eval_loss": 0.2209562063217163,
      "eval_mean_token_accuracy": 0.9337840676307678,
      "eval_num_tokens": 2327070.0,
      "eval_runtime": 3.3769,
      "eval_samples_per_second": 51.527,
      "eval_steps_per_second": 13.03,
      "step": 500
    }
  ],
  "logging_steps": 10,
  "max_steps": 585,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 6286031527403520.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}