{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.999690962871396,
  "eval_steps": 500,
  "global_step": 70780,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 9.416195856873823e-06,
      "loss": 0.7169,
      "step": 500
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.8832391713747646e-05,
      "loss": 0.7282,
      "step": 1000
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.8248587570621472e-05,
      "loss": 0.7519,
      "step": 1500
    },
    {
      "epoch": 0.14,
      "learning_rate": 3.766478342749529e-05,
      "loss": 0.7757,
      "step": 2000
    },
    {
      "epoch": 0.18,
      "learning_rate": 3.9997039890834935e-05,
      "loss": 0.8085,
      "step": 2500
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.998393455306432e-05,
      "loss": 0.8176,
      "step": 3000
    },
    {
      "epoch": 0.25,
      "learning_rate": 3.9960368899072094e-05,
      "loss": 0.8158,
      "step": 3500
    },
    {
      "epoch": 0.28,
      "learning_rate": 3.992635526397635e-05,
      "loss": 0.8239,
      "step": 4000
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.9881911451747484e-05,
      "loss": 0.8059,
      "step": 4500
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.9827060725888914e-05,
      "loss": 0.806,
      "step": 5000
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.9761831797260154e-05,
      "loss": 0.8018,
      "step": 5500
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.96862588090485e-05,
      "loss": 0.7967,
      "step": 6000
    },
    {
      "epoch": 0.46,
      "learning_rate": 3.960038131889723e-05,
      "loss": 0.8037,
      "step": 6500
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.9504244278199726e-05,
      "loss": 0.7933,
      "step": 7000
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.9397898008570265e-05,
      "loss": 0.7962,
      "step": 7500
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.9281398175503866e-05,
      "loss": 0.7918,
      "step": 8000
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.915480575923895e-05,
      "loss": 0.7868,
      "step": 8500
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.901818702283807e-05,
      "loss": 0.7887,
      "step": 9000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8871613477503424e-05,
      "loss": 0.7835,
      "step": 9500
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.87151618451453e-05,
      "loss": 0.7867,
      "step": 10000
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.854891401822304e-05,
      "loss": 0.7844,
      "step": 10500
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.837295701687955e-05,
      "loss": 0.7791,
      "step": 11000
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.818738294339182e-05,
      "loss": 0.785,
      "step": 11500
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.799228893396123e-05,
      "loss": 0.7791,
      "step": 12000
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.778777710786896e-05,
      "loss": 0.7804,
      "step": 12500
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.757395451402304e-05,
      "loss": 0.7794,
      "step": 13000
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.735093307492506e-05,
      "loss": 0.7736,
      "step": 13500
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.7118829528085897e-05,
      "loss": 0.7837,
      "step": 14000
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.687776536492105e-05,
      "loss": 0.7527,
      "step": 14500
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.662786676715763e-05,
      "loss": 0.7403,
      "step": 15000
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.636926454078625e-05,
      "loss": 0.7547,
      "step": 15500
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.610209404759251e-05,
      "loss": 0.7476,
      "step": 16000
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.5826495134303565e-05,
      "loss": 0.7474,
      "step": 16500
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.5542612059387445e-05,
      "loss": 0.747,
      "step": 17000
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.5250593417542837e-05,
      "loss": 0.7441,
      "step": 17500
    },
    {
      "epoch": 1.27,
      "learning_rate": 3.495059206191926e-05,
      "loss": 0.7463,
      "step": 18000
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.464276502410819e-05,
      "loss": 0.7452,
      "step": 18500
    },
    {
      "epoch": 1.34,
      "learning_rate": 3.432727343194701e-05,
      "loss": 0.7439,
      "step": 19000
    },
    {
      "epoch": 1.38,
      "learning_rate": 3.400428242517889e-05,
      "loss": 0.7438,
      "step": 19500
    },
    {
      "epoch": 1.41,
      "learning_rate": 3.367396106901259e-05,
      "loss": 0.7427,
      "step": 20000
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.3336482265627675e-05,
      "loss": 0.7471,
      "step": 20500
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.299202266367119e-05,
      "loss": 0.745,
      "step": 21000
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.2640762565793374e-05,
      "loss": 0.7418,
      "step": 21500
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.2282885834270696e-05,
      "loss": 0.7465,
      "step": 22000
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.191857979476569e-05,
      "loss": 0.7416,
      "step": 22500
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.154803513827388e-05,
      "loss": 0.7421,
      "step": 23000
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.117144582130925e-05,
      "loss": 0.7396,
      "step": 23500
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.078900896438028e-05,
      "loss": 0.7371,
      "step": 24000
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.040092474881003e-05,
      "loss": 0.7393,
      "step": 24500
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.0007396311953882e-05,
      "loss": 0.7366,
      "step": 25000
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.9608629640870138e-05,
      "loss": 0.7455,
      "step": 25500
    },
    {
      "epoch": 1.84,
      "learning_rate": 2.9204833464498878e-05,
      "loss": 0.7387,
      "step": 26000
    },
    {
      "epoch": 1.87,
      "learning_rate": 2.8796219144405713e-05,
      "loss": 0.7363,
      "step": 26500
    },
    {
      "epoch": 1.91,
      "learning_rate": 2.838300056414743e-05,
      "loss": 0.7387,
      "step": 27000
    },
    {
      "epoch": 1.94,
      "learning_rate": 2.7965394017317587e-05,
      "loss": 0.7317,
      "step": 27500
    },
    {
      "epoch": 1.98,
      "learning_rate": 2.754361809433063e-05,
      "loss": 0.7331,
      "step": 28000
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.711789356800372e-05,
      "loss": 0.7049,
      "step": 28500
    },
    {
      "epoch": 2.05,
      "learning_rate": 2.6688443277996225e-05,
      "loss": 0.6657,
      "step": 29000
    },
    {
      "epoch": 2.08,
      "learning_rate": 2.6255492014167356e-05,
      "loss": 0.6651,
      "step": 29500
    },
    {
      "epoch": 2.12,
      "learning_rate": 2.581926639891304e-05,
      "loss": 0.6717,
      "step": 30000
    },
    {
      "epoch": 2.15,
      "learning_rate": 2.537999476854349e-05,
      "loss": 0.6688,
      "step": 30500
    },
    {
      "epoch": 2.19,
      "learning_rate": 2.4937907053763732e-05,
      "loss": 0.6726,
      "step": 31000
    },
    {
      "epoch": 2.23,
      "learning_rate": 2.4493234659319507e-05,
      "loss": 0.6681,
      "step": 31500
    },
    {
      "epoch": 2.26,
      "learning_rate": 2.404621034287166e-05,
      "loss": 0.6656,
      "step": 32000
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.359706809316231e-05,
      "loss": 0.6633,
      "step": 32500
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.314604300753667e-05,
      "loss": 0.6647,
      "step": 33000
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.2693371168884593e-05,
      "loss": 0.6632,
      "step": 33500
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.2239289522066157e-05,
      "loss": 0.6642,
      "step": 34000
    },
    {
      "epoch": 2.44,
      "learning_rate": 2.178403574988621e-05,
      "loss": 0.6626,
      "step": 34500
    },
    {
      "epoch": 2.47,
      "learning_rate": 2.1327848148682503e-05,
      "loss": 0.659,
      "step": 35000
    },
    {
      "epoch": 2.51,
      "learning_rate": 2.0870965503592795e-05,
      "loss": 0.6626,
      "step": 35500
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.0413626963566004e-05,
      "loss": 0.6605,
      "step": 36000
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.9956071916183e-05,
      "loss": 0.6583,
      "step": 36500
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.9498539862352476e-05,
      "loss": 0.6554,
      "step": 37000
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.904127029094744e-05,
      "loss": 0.6588,
      "step": 37500
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.8584502553448085e-05,
      "loss": 0.6561,
      "step": 38000
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.812847573865655e-05,
      "loss": 0.6529,
      "step": 38500
    },
    {
      "epoch": 2.75,
      "learning_rate": 1.7673428547549134e-05,
      "loss": 0.6499,
      "step": 39000
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.721959916833157e-05,
      "loss": 0.648,
      "step": 39500
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.6767225151762676e-05,
      "loss": 0.6479,
      "step": 40000
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.631654328681168e-05,
      "loss": 0.6482,
      "step": 40500
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.586778947671426e-05,
      "loss": 0.6446,
      "step": 41000
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.5421198615492244e-05,
      "loss": 0.6432,
      "step": 41500
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.4977004465001586e-05,
      "loss": 0.6425,
      "step": 42000
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.4535439532572877e-05,
      "loss": 0.6361,
      "step": 42500
    },
    {
      "epoch": 3.04,
      "learning_rate": 1.4096734949308623e-05,
      "loss": 0.5411,
      "step": 43000
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.3661120349100823e-05,
      "loss": 0.5375,
      "step": 43500
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.3228823748432258e-05,
      "loss": 0.5394,
      "step": 44000
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.2800071427024391e-05,
      "loss": 0.5369,
      "step": 44500
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.2375087809394368e-05,
      "loss": 0.5378,
      "step": 45000
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.1954095347383076e-05,
      "loss": 0.5365,
      "step": 45500
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.153731440371577e-05,
      "loss": 0.5358,
      "step": 46000
    },
    {
      "epoch": 3.28,
      "learning_rate": 1.1124963136656253e-05,
      "loss": 0.5375,
      "step": 46500
    },
    {
      "epoch": 3.32,
      "learning_rate": 1.0717257385814897e-05,
      "loss": 0.537,
      "step": 47000
    },
    {
      "epoch": 3.36,
      "learning_rate": 1.0314410559170397e-05,
      "loss": 0.5348,
      "step": 47500
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.916633521364266e-06,
      "loss": 0.5342,
      "step": 48000
    },
    {
      "epoch": 3.43,
      "learning_rate": 9.524134483326633e-06,
      "loss": 0.5321,
      "step": 48500
    },
    {
      "epoch": 3.46,
      "learning_rate": 9.137118893291118e-06,
      "loss": 0.5311,
      "step": 49000
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.755789329255755e-06,
      "loss": 0.5303,
      "step": 49500
    },
    {
      "epoch": 3.53,
      "learning_rate": 8.3803453929463e-06,
      "loss": 0.5296,
      "step": 50000
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.01098360533749e-06,
      "loss": 0.5288,
      "step": 50500
    },
    {
      "epoch": 3.6,
      "learning_rate": 7.647897303786813e-06,
      "loss": 0.5266,
      "step": 51000
    },
    {
      "epoch": 3.64,
      "learning_rate": 7.291276540834699e-06,
      "loss": 0.5263,
      "step": 51500
    },
    {
      "epoch": 3.67,
      "learning_rate": 6.941307984724182e-06,
      "loss": 0.5259,
      "step": 52000
    },
    {
      "epoch": 3.71,
      "learning_rate": 6.598174821691929e-06,
      "loss": 0.5231,
      "step": 52500
    },
    {
      "epoch": 3.74,
      "learning_rate": 6.262056660081919e-06,
      "loss": 0.5255,
      "step": 53000
    },
    {
      "epoch": 3.78,
      "learning_rate": 5.933129436331942e-06,
      "loss": 0.5237,
      "step": 53500
    },
    {
      "epoch": 3.81,
      "learning_rate": 5.611565322882084e-06,
      "loss": 0.5213,
      "step": 54000
    },
    {
      "epoch": 3.85,
      "learning_rate": 5.297532638053395e-06,
      "loss": 0.5203,
      "step": 54500
    },
    {
      "epoch": 3.89,
      "learning_rate": 4.991195757944023e-06,
      "loss": 0.5191,
      "step": 55000
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.6927150303887505e-06,
      "loss": 0.5177,
      "step": 55500
    },
    {
      "epoch": 3.96,
      "learning_rate": 4.402246691027168e-06,
      "loss": 0.5161,
      "step": 56000
    },
    {
      "epoch": 3.99,
      "learning_rate": 4.119942781524248e-06,
      "loss": 0.516,
      "step": 56500
    },
    {
      "epoch": 4.03,
      "learning_rate": 3.845951069986216e-06,
      "loss": 0.4345,
      "step": 57000
    },
    {
      "epoch": 4.06,
      "learning_rate": 3.5804149736133887e-06,
      "loss": 0.4057,
      "step": 57500
    },
    {
      "epoch": 4.1,
      "learning_rate": 3.3234734836303883e-06,
      "loss": 0.4048,
      "step": 58000
    },
    {
      "epoch": 4.13,
      "learning_rate": 3.075261092533097e-06,
      "loss": 0.4048,
      "step": 58500
    },
    {
      "epoch": 4.17,
      "learning_rate": 2.8359077236904165e-06,
      "loss": 0.4041,
      "step": 59000
    },
    {
      "epoch": 4.2,
      "learning_rate": 2.6055386633376613e-06,
      "loss": 0.4026,
      "step": 59500
    },
    {
      "epoch": 4.24,
      "learning_rate": 2.3842744949971765e-06,
      "loss": 0.4039,
      "step": 60000
    },
    {
      "epoch": 4.27,
      "learning_rate": 2.172231036360588e-06,
      "loss": 0.403,
      "step": 60500
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.9695192786655902e-06,
      "loss": 0.4038,
      "step": 61000
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.776245328599111e-06,
      "loss": 0.4023,
      "step": 61500
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.5925103527572395e-06,
      "loss": 0.4018,
      "step": 62000
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.4184105246909429e-06,
      "loss": 0.4017,
      "step": 62500
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.2540369745653446e-06,
      "loss": 0.4022,
      "step": 63000
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.099475741458904e-06,
      "loss": 0.4016,
      "step": 63500
    },
    {
      "epoch": 4.52,
      "learning_rate": 9.548077283274115e-07,
      "loss": 0.4021,
      "step": 64000
    },
    {
      "epoch": 4.56,
      "learning_rate": 8.201086596564867e-07,
      "loss": 0.4003,
      "step": 64500
    },
    {
      "epoch": 4.59,
      "learning_rate": 6.954490418246052e-07,
      "loss": 0.4001,
      "step": 65000
    },
    {
      "epoch": 4.63,
      "learning_rate": 5.808941261975087e-07,
      "loss": 0.4004,
      "step": 65500
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.765038749732864e-07,
      "loss": 0.4003,
      "step": 66000
    },
    {
      "epoch": 4.7,
      "learning_rate": 3.823329297959766e-07,
      "loss": 0.4002,
      "step": 66500
    },
    {
      "epoch": 4.73,
      "learning_rate": 2.98430583154139e-07,
      "loss": 0.4002,
      "step": 67000
    },
    {
      "epoch": 4.77,
      "learning_rate": 2.2484075257939165e-07,
      "loss": 0.4006,
      "step": 67500
    },
    {
      "epoch": 4.8,
      "learning_rate": 1.6160195765838605e-07,
      "loss": 0.3993,
      "step": 68000
    },
    {
      "epoch": 4.84,
      "learning_rate": 1.0874729987023547e-07,
      "loss": 0.3989,
      "step": 68500
    },
    {
      "epoch": 4.87,
      "learning_rate": 6.630444526002367e-08,
      "loss": 0.3991,
      "step": 69000
    },
    {
      "epoch": 4.91,
      "learning_rate": 3.4295609957382126e-08,
      "loss": 0.3996,
      "step": 69500
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.2737548547760991e-08,
      "loss": 0.3992,
      "step": 70000
    },
    {
      "epoch": 4.98,
      "learning_rate": 1.6415453024820617e-09,
      "loss": 0.3995,
      "step": 70500
    },
    {
      "epoch": 5.0,
      "step": 70780,
      "total_flos": 1.007604716180962e+24,
      "train_loss": 0.6235090117226821,
      "train_runtime": 442317.6902,
      "train_samples_per_second": 40.968,
      "train_steps_per_second": 0.16
    }
  ],
  "logging_steps": 500,
  "max_steps": 70780,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 3000,
  "total_flos": 1.007604716180962e+24,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}