{
  "best_metric": 0.2909594774246216,
  "best_model_checkpoint": "/data2/ckpts/GenRM/qwen-2.5-math-instruct/full/GenPRM-78k-train-5:5-decontamination/checkpoint-1200",
  "epoch": 0.9995883079456567,
  "eval_steps": 100,
  "global_step": 1214,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016467682173734045,
      "grad_norm": 2.2536449432373047,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.9706,
      "step": 20
    },
    {
      "epoch": 0.03293536434746809,
      "grad_norm": 0.7532587051391602,
      "learning_rate": 4.999919851200522e-06,
      "loss": 0.5433,
      "step": 40
    },
    {
      "epoch": 0.04940304652120214,
      "grad_norm": 0.6202924847602844,
      "learning_rate": 4.995290485881111e-06,
      "loss": 0.4319,
      "step": 60
    },
    {
      "epoch": 0.06587072869493618,
      "grad_norm": 0.5724588632583618,
      "learning_rate": 4.983551854245604e-06,
      "loss": 0.3985,
      "step": 80
    },
    {
      "epoch": 0.08233841086867023,
      "grad_norm": 0.6023426651954651,
      "learning_rate": 4.9647374005198125e-06,
      "loss": 0.3951,
      "step": 100
    },
    {
      "epoch": 0.08233841086867023,
      "eval_loss": 0.37706664204597473,
      "eval_runtime": 38.4153,
      "eval_samples_per_second": 52.063,
      "eval_steps_per_second": 6.508,
      "step": 100
    },
    {
      "epoch": 0.09880609304240429,
      "grad_norm": 0.6422224640846252,
      "learning_rate": 4.938900728467664e-06,
      "loss": 0.381,
      "step": 120
    },
    {
      "epoch": 0.11527377521613832,
      "grad_norm": 0.5755812525749207,
      "learning_rate": 4.9061154486701204e-06,
      "loss": 0.3649,
      "step": 140
    },
    {
      "epoch": 0.13174145738987236,
      "grad_norm": 0.5953928232192993,
      "learning_rate": 4.866474968803222e-06,
      "loss": 0.3605,
      "step": 160
    },
    {
      "epoch": 0.14820913956360643,
      "grad_norm": 0.608504593372345,
      "learning_rate": 4.820092227512736e-06,
      "loss": 0.3514,
      "step": 180
    },
    {
      "epoch": 0.16467682173734047,
      "grad_norm": 0.6068744659423828,
      "learning_rate": 4.767099372643641e-06,
      "loss": 0.3471,
      "step": 200
    },
    {
      "epoch": 0.16467682173734047,
      "eval_loss": 0.3430534303188324,
      "eval_runtime": 38.4427,
      "eval_samples_per_second": 52.026,
      "eval_steps_per_second": 6.503,
      "step": 200
    },
    {
      "epoch": 0.1811445039110745,
      "grad_norm": 0.6543622612953186,
      "learning_rate": 4.707647384741187e-06,
      "loss": 0.3494,
      "step": 220
    },
    {
      "epoch": 0.19761218608480857,
      "grad_norm": 0.6283196210861206,
      "learning_rate": 4.641905646896205e-06,
      "loss": 0.3382,
      "step": 240
    },
    {
      "epoch": 0.2140798682585426,
      "grad_norm": 0.6050875186920166,
      "learning_rate": 4.570061462160209e-06,
      "loss": 0.3382,
      "step": 260
    },
    {
      "epoch": 0.23054755043227665,
      "grad_norm": 0.6354455947875977,
      "learning_rate": 4.492319519905217e-06,
      "loss": 0.3319,
      "step": 280
    },
    {
      "epoch": 0.24701523260601072,
      "grad_norm": 0.5852451920509338,
      "learning_rate": 4.408901312648652e-06,
      "loss": 0.3295,
      "step": 300
    },
    {
      "epoch": 0.24701523260601072,
      "eval_loss": 0.3265990614891052,
      "eval_runtime": 38.4536,
      "eval_samples_per_second": 52.011,
      "eval_steps_per_second": 6.501,
      "step": 300
    },
    {
      "epoch": 0.2634829147797447,
      "grad_norm": 0.6466448307037354,
      "learning_rate": 4.3200445050048545e-06,
      "loss": 0.3322,
      "step": 320
    },
    {
      "epoch": 0.2799505969534788,
      "grad_norm": 0.6878635287284851,
      "learning_rate": 4.226002256561101e-06,
      "loss": 0.3336,
      "step": 340
    },
    {
      "epoch": 0.29641827912721286,
      "grad_norm": 0.6180429458618164,
      "learning_rate": 4.127042500607298e-06,
      "loss": 0.3287,
      "step": 360
    },
    {
      "epoch": 0.31288596130094687,
      "grad_norm": 0.6696975231170654,
      "learning_rate": 4.023447180774308e-06,
      "loss": 0.3208,
      "step": 380
    },
    {
      "epoch": 0.32935364347468093,
      "grad_norm": 0.6044342517852783,
      "learning_rate": 3.915511447755793e-06,
      "loss": 0.3162,
      "step": 400
    },
    {
      "epoch": 0.32935364347468093,
      "eval_loss": 0.3160727918148041,
      "eval_runtime": 38.4863,
      "eval_samples_per_second": 51.967,
      "eval_steps_per_second": 6.496,
      "step": 400
    },
    {
      "epoch": 0.345821325648415,
      "grad_norm": 0.6135950088500977,
      "learning_rate": 3.803542818402154e-06,
      "loss": 0.3167,
      "step": 420
    },
    {
      "epoch": 0.362289007822149,
      "grad_norm": 0.6029672026634216,
      "learning_rate": 3.687860299582381e-06,
      "loss": 0.3221,
      "step": 440
    },
    {
      "epoch": 0.3787566899958831,
      "grad_norm": 0.6770069003105164,
      "learning_rate": 3.568793479309998e-06,
      "loss": 0.319,
      "step": 460
    },
    {
      "epoch": 0.39522437216961714,
      "grad_norm": 0.6067743897438049,
      "learning_rate": 3.4466815877225456e-06,
      "loss": 0.3122,
      "step": 480
    },
    {
      "epoch": 0.41169205434335115,
      "grad_norm": 0.5844372510910034,
      "learning_rate": 3.3218725305899402e-06,
      "loss": 0.3143,
      "step": 500
    },
    {
      "epoch": 0.41169205434335115,
      "eval_loss": 0.3084495961666107,
      "eval_runtime": 38.5709,
      "eval_samples_per_second": 51.853,
      "eval_steps_per_second": 6.482,
      "step": 500
    },
    {
      "epoch": 0.4281597365170852,
      "grad_norm": 0.6176360249519348,
      "learning_rate": 3.194721898105323e-06,
      "loss": 0.3128,
      "step": 520
    },
    {
      "epoch": 0.4446274186908193,
      "grad_norm": 0.6112409830093384,
      "learning_rate": 3.0655919517824167e-06,
      "loss": 0.315,
      "step": 540
    },
    {
      "epoch": 0.4610951008645533,
      "grad_norm": 0.5909674763679504,
      "learning_rate": 2.9348505923458008e-06,
      "loss": 0.3179,
      "step": 560
    },
    {
      "epoch": 0.47756278303828736,
      "grad_norm": 0.5776228904724121,
      "learning_rate": 2.802870311554661e-06,
      "loss": 0.3109,
      "step": 580
    },
    {
      "epoch": 0.49403046521202143,
      "grad_norm": 0.5729939937591553,
      "learning_rate": 2.6700271309463397e-06,
      "loss": 0.3054,
      "step": 600
    },
    {
      "epoch": 0.49403046521202143,
      "eval_loss": 0.3028527796268463,
      "eval_runtime": 38.5353,
      "eval_samples_per_second": 51.9,
      "eval_steps_per_second": 6.488,
      "step": 600
    },
    {
      "epoch": 0.5104981473857555,
      "grad_norm": 0.6428977847099304,
      "learning_rate": 2.536699530523292e-06,
      "loss": 0.3051,
      "step": 620
    },
    {
      "epoch": 0.5269658295594895,
      "grad_norm": 0.652348518371582,
      "learning_rate": 2.4032673704356912e-06,
      "loss": 0.3006,
      "step": 640
    },
    {
      "epoch": 0.5434335117332235,
      "grad_norm": 0.665675699710846,
      "learning_rate": 2.2701108087318935e-06,
      "loss": 0.3068,
      "step": 660
    },
    {
      "epoch": 0.5599011939069576,
      "grad_norm": 0.568688154220581,
      "learning_rate": 2.1376092182601694e-06,
      "loss": 0.3079,
      "step": 680
    },
    {
      "epoch": 0.5763688760806917,
      "grad_norm": 0.6417534351348877,
      "learning_rate": 2.006140105807523e-06,
      "loss": 0.3031,
      "step": 700
    },
    {
      "epoch": 0.5763688760806917,
      "eval_loss": 0.29845306277275085,
      "eval_runtime": 38.6545,
      "eval_samples_per_second": 51.74,
      "eval_steps_per_second": 6.468,
      "step": 700
    },
    {
      "epoch": 0.5928365582544257,
      "grad_norm": 0.6650199294090271,
      "learning_rate": 1.8760780365550654e-06,
      "loss": 0.304,
      "step": 720
    },
    {
      "epoch": 0.6093042404281598,
      "grad_norm": 0.6189924478530884,
      "learning_rate": 1.7477935669142319e-06,
      "loss": 0.3034,
      "step": 740
    },
    {
      "epoch": 0.6257719226018937,
      "grad_norm": 0.64354008436203,
      "learning_rate": 1.6216521887842863e-06,
      "loss": 0.306,
      "step": 760
    },
    {
      "epoch": 0.6422396047756278,
      "grad_norm": 0.5726205110549927,
      "learning_rate": 1.4980132882389836e-06,
      "loss": 0.2985,
      "step": 780
    },
    {
      "epoch": 0.6587072869493619,
      "grad_norm": 0.6365756988525391,
      "learning_rate": 1.3772291216091954e-06,
      "loss": 0.2988,
      "step": 800
    },
    {
      "epoch": 0.6587072869493619,
      "eval_loss": 0.2953297197818756,
      "eval_runtime": 38.4039,
      "eval_samples_per_second": 52.078,
      "eval_steps_per_second": 6.51,
      "step": 800
    },
    {
      "epoch": 0.6751749691230959,
      "grad_norm": 0.6121465563774109,
      "learning_rate": 1.2596438118786732e-06,
      "loss": 0.3006,
      "step": 820
    },
    {
      "epoch": 0.69164265129683,
      "grad_norm": 0.601615309715271,
      "learning_rate": 1.1455923682523476e-06,
      "loss": 0.298,
      "step": 840
    },
    {
      "epoch": 0.7081103334705641,
      "grad_norm": 0.5637004971504211,
      "learning_rate": 1.0353997316904204e-06,
      "loss": 0.3021,
      "step": 860
    },
    {
      "epoch": 0.724578015644298,
      "grad_norm": 0.6115686893463135,
      "learning_rate": 9.293798491276612e-07,
      "loss": 0.3007,
      "step": 880
    },
    {
      "epoch": 0.7410456978180321,
      "grad_norm": 0.56075119972229,
      "learning_rate": 8.278347790154595e-07,
      "loss": 0.2965,
      "step": 900
    },
    {
      "epoch": 0.7410456978180321,
      "eval_loss": 0.2931554615497589,
      "eval_runtime": 38.421,
      "eval_samples_per_second": 52.055,
      "eval_steps_per_second": 6.507,
      "step": 900
    },
    {
      "epoch": 0.7575133799917662,
      "grad_norm": 0.5739229917526245,
      "learning_rate": 7.310538307350684e-07,
      "loss": 0.2949,
      "step": 920
    },
    {
      "epoch": 0.7739810621655002,
      "grad_norm": 0.6038621664047241,
      "learning_rate": 6.393127403338714e-07,
      "loss": 0.2999,
      "step": 940
    },
    {
      "epoch": 0.7904487443392343,
      "grad_norm": 0.6664382815361023,
      "learning_rate": 5.528728849330817e-07,
      "loss": 0.3014,
      "step": 960
    },
    {
      "epoch": 0.8069164265129684,
      "grad_norm": 0.5887951850891113,
      "learning_rate": 4.719805380450804e-07,
      "loss": 0.2979,
      "step": 980
    },
    {
      "epoch": 0.8233841086867023,
      "grad_norm": 0.584530234336853,
      "learning_rate": 3.9686616792204677e-07,
      "loss": 0.2935,
      "step": 1000
    },
    {
      "epoch": 0.8233841086867023,
      "eval_loss": 0.29177162051200867,
      "eval_runtime": 38.3967,
      "eval_samples_per_second": 52.088,
      "eval_steps_per_second": 6.511,
      "step": 1000
    },
    {
      "epoch": 0.8398517908604364,
      "grad_norm": 0.520577609539032,
      "learning_rate": 3.2774378093494067e-07,
      "loss": 0.2974,
      "step": 1020
    },
    {
      "epoch": 0.8563194730341704,
      "grad_norm": 0.60713130235672,
      "learning_rate": 2.6481031185358995e-07,
      "loss": 0.3022,
      "step": 1040
    },
    {
      "epoch": 0.8727871552079045,
      "grad_norm": 0.655536413192749,
      "learning_rate": 2.0824506276503898e-07,
      "loss": 0.2961,
      "step": 1060
    },
    {
      "epoch": 0.8892548373816386,
      "grad_norm": 0.5735917091369629,
      "learning_rate": 1.5820919222869325e-07,
      "loss": 0.2983,
      "step": 1080
    },
    {
      "epoch": 0.9057225195553725,
      "grad_norm": 0.5977383852005005,
      "learning_rate": 1.1484525612372372e-07,
      "loss": 0.2975,
      "step": 1100
    },
    {
      "epoch": 0.9057225195553725,
      "eval_loss": 0.2911008894443512,
      "eval_runtime": 38.7911,
      "eval_samples_per_second": 51.558,
      "eval_steps_per_second": 6.445,
      "step": 1100
    },
    {
      "epoch": 0.9221902017291066,
      "grad_norm": 0.602390468120575,
      "learning_rate": 7.827680149686879e-08,
      "loss": 0.299,
      "step": 1120
    },
    {
      "epoch": 0.9386578839028407,
      "grad_norm": 0.5963403582572937,
      "learning_rate": 4.86080145678014e-08,
      "loss": 0.2954,
      "step": 1140
    },
    {
      "epoch": 0.9551255660765747,
      "grad_norm": 0.5622375011444092,
      "learning_rate": 2.5923423894919365e-08,
      "loss": 0.2968,
      "step": 1160
    },
    {
      "epoch": 0.9715932482503088,
      "grad_norm": 0.5845054984092712,
      "learning_rate": 1.02876595472573e-08,
      "loss": 0.292,
      "step": 1180
    },
    {
      "epoch": 0.9880609304240429,
      "grad_norm": 0.6160269975662231,
      "learning_rate": 1.7452689686631164e-09,
      "loss": 0.304,
      "step": 1200
    },
    {
      "epoch": 0.9880609304240429,
      "eval_loss": 0.2909594774246216,
      "eval_runtime": 38.9593,
      "eval_samples_per_second": 51.336,
      "eval_steps_per_second": 6.417,
      "step": 1200
    },
    {
      "epoch": 0.9995883079456567,
      "step": 1214,
      "total_flos": 5.587165226987422e+18,
      "train_loss": 0.332974739875982,
      "train_runtime": 7444.239,
      "train_samples_per_second": 10.439,
      "train_steps_per_second": 0.163
    }
  ],
  "logging_steps": 20,
  "max_steps": 1214,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 5.587165226987422e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}