{
  "best_metric": 1.0948013067245483,
  "best_model_checkpoint": "output_dir/checkpoint-8000",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 8420,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12,
      "learning_rate": 4.9406175771971496e-05,
      "loss": 2.9665,
      "step": 100
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.8812351543942996e-05,
      "loss": 2.3245,
      "step": 200
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.821852731591449e-05,
      "loss": 2.1171,
      "step": 300
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.7624703087885984e-05,
      "loss": 2.0118,
      "step": 400
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.7030878859857484e-05,
      "loss": 1.9308,
      "step": 500
    },
    {
      "epoch": 0.59,
      "eval_loss": 1.7528491020202637,
      "eval_runtime": 55.8537,
      "eval_samples_per_second": 80.335,
      "eval_steps_per_second": 5.031,
      "step": 500
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.643705463182898e-05,
      "loss": 1.8934,
      "step": 600
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.584323040380048e-05,
      "loss": 1.7962,
      "step": 700
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.524940617577197e-05,
      "loss": 1.7785,
      "step": 800
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.465558194774347e-05,
      "loss": 1.7253,
      "step": 900
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.406175771971497e-05,
      "loss": 1.7079,
      "step": 1000
    },
    {
      "epoch": 1.19,
      "eval_loss": 1.5516979694366455,
      "eval_runtime": 56.1057,
      "eval_samples_per_second": 79.974,
      "eval_steps_per_second": 5.008,
      "step": 1000
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.3467933491686466e-05,
      "loss": 1.629,
      "step": 1100
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.287410926365796e-05,
      "loss": 1.6273,
      "step": 1200
    },
    {
      "epoch": 1.54,
      "learning_rate": 4.228028503562946e-05,
      "loss": 1.5842,
      "step": 1300
    },
    {
      "epoch": 1.66,
      "learning_rate": 4.168646080760095e-05,
      "loss": 1.6089,
      "step": 1400
    },
    {
      "epoch": 1.78,
      "learning_rate": 4.109263657957245e-05,
      "loss": 1.5614,
      "step": 1500
    },
    {
      "epoch": 1.78,
      "eval_loss": 1.4558075666427612,
      "eval_runtime": 56.2297,
      "eval_samples_per_second": 79.798,
      "eval_steps_per_second": 4.997,
      "step": 1500
    },
    {
      "epoch": 1.9,
      "learning_rate": 4.049881235154395e-05,
      "loss": 1.5462,
      "step": 1600
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.990498812351544e-05,
      "loss": 1.5253,
      "step": 1700
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.9311163895486934e-05,
      "loss": 1.4815,
      "step": 1800
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.8717339667458435e-05,
      "loss": 1.5012,
      "step": 1900
    },
    {
      "epoch": 2.38,
      "learning_rate": 3.812351543942993e-05,
      "loss": 1.4384,
      "step": 2000
    },
    {
      "epoch": 2.38,
      "eval_loss": 1.3825178146362305,
      "eval_runtime": 56.3145,
      "eval_samples_per_second": 79.678,
      "eval_steps_per_second": 4.99,
      "step": 2000
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.752969121140142e-05,
      "loss": 1.4401,
      "step": 2100
    },
    {
      "epoch": 2.61,
      "learning_rate": 3.693586698337292e-05,
      "loss": 1.4834,
      "step": 2200
    },
    {
      "epoch": 2.73,
      "learning_rate": 3.6342042755344416e-05,
      "loss": 1.4534,
      "step": 2300
    },
    {
      "epoch": 2.85,
      "learning_rate": 3.5748218527315916e-05,
      "loss": 1.4139,
      "step": 2400
    },
    {
      "epoch": 2.97,
      "learning_rate": 3.515439429928741e-05,
      "loss": 1.3972,
      "step": 2500
    },
    {
      "epoch": 2.97,
      "eval_loss": 1.3105206489562988,
      "eval_runtime": 56.1399,
      "eval_samples_per_second": 79.925,
      "eval_steps_per_second": 5.005,
      "step": 2500
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.456057007125891e-05,
      "loss": 1.4154,
      "step": 2600
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.396674584323041e-05,
      "loss": 1.3644,
      "step": 2700
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3372921615201904e-05,
      "loss": 1.3769,
      "step": 2800
    },
    {
      "epoch": 3.44,
      "learning_rate": 3.27790973871734e-05,
      "loss": 1.3772,
      "step": 2900
    },
    {
      "epoch": 3.56,
      "learning_rate": 3.21852731591449e-05,
      "loss": 1.3532,
      "step": 3000
    },
    {
      "epoch": 3.56,
      "eval_loss": 1.2871623039245605,
      "eval_runtime": 56.4013,
      "eval_samples_per_second": 79.555,
      "eval_steps_per_second": 4.982,
      "step": 3000
    },
    {
      "epoch": 3.68,
      "learning_rate": 3.159144893111639e-05,
      "loss": 1.3614,
      "step": 3100
    },
    {
      "epoch": 3.8,
      "learning_rate": 3.0997624703087885e-05,
      "loss": 1.3531,
      "step": 3200
    },
    {
      "epoch": 3.92,
      "learning_rate": 3.0403800475059386e-05,
      "loss": 1.3365,
      "step": 3300
    },
    {
      "epoch": 4.04,
      "learning_rate": 2.980997624703088e-05,
      "loss": 1.3075,
      "step": 3400
    },
    {
      "epoch": 4.16,
      "learning_rate": 2.9216152019002373e-05,
      "loss": 1.3114,
      "step": 3500
    },
    {
      "epoch": 4.16,
      "eval_loss": 1.2454777956008911,
      "eval_runtime": 56.1952,
      "eval_samples_per_second": 79.847,
      "eval_steps_per_second": 5.0,
      "step": 3500
    },
    {
      "epoch": 4.28,
      "learning_rate": 2.8622327790973873e-05,
      "loss": 1.2952,
      "step": 3600
    },
    {
      "epoch": 4.39,
      "learning_rate": 2.8028503562945367e-05,
      "loss": 1.3133,
      "step": 3700
    },
    {
      "epoch": 4.51,
      "learning_rate": 2.7434679334916867e-05,
      "loss": 1.3225,
      "step": 3800
    },
    {
      "epoch": 4.63,
      "learning_rate": 2.6840855106888364e-05,
      "loss": 1.2958,
      "step": 3900
    },
    {
      "epoch": 4.75,
      "learning_rate": 2.6247030878859858e-05,
      "loss": 1.269,
      "step": 4000
    },
    {
      "epoch": 4.75,
      "eval_loss": 1.2113255262374878,
      "eval_runtime": 57.2608,
      "eval_samples_per_second": 78.361,
      "eval_steps_per_second": 4.907,
      "step": 4000
    },
    {
      "epoch": 4.87,
      "learning_rate": 2.5653206650831358e-05,
      "loss": 1.2693,
      "step": 4100
    },
    {
      "epoch": 4.99,
      "learning_rate": 2.5059382422802852e-05,
      "loss": 1.2646,
      "step": 4200
    },
    {
      "epoch": 5.11,
      "learning_rate": 2.446555819477435e-05,
      "loss": 1.2702,
      "step": 4300
    },
    {
      "epoch": 5.23,
      "learning_rate": 2.3871733966745842e-05,
      "loss": 1.2568,
      "step": 4400
    },
    {
      "epoch": 5.34,
      "learning_rate": 2.327790973871734e-05,
      "loss": 1.2525,
      "step": 4500
    },
    {
      "epoch": 5.34,
      "eval_loss": 1.1863404512405396,
      "eval_runtime": 56.357,
      "eval_samples_per_second": 79.617,
      "eval_steps_per_second": 4.986,
      "step": 4500
    },
    {
      "epoch": 5.46,
      "learning_rate": 2.2684085510688836e-05,
      "loss": 1.2343,
      "step": 4600
    },
    {
      "epoch": 5.58,
      "learning_rate": 2.2090261282660333e-05,
      "loss": 1.2289,
      "step": 4700
    },
    {
      "epoch": 5.7,
      "learning_rate": 2.149643705463183e-05,
      "loss": 1.2091,
      "step": 4800
    },
    {
      "epoch": 5.82,
      "learning_rate": 2.0902612826603327e-05,
      "loss": 1.2231,
      "step": 4900
    },
    {
      "epoch": 5.94,
      "learning_rate": 2.0308788598574824e-05,
      "loss": 1.2171,
      "step": 5000
    },
    {
      "epoch": 5.94,
      "eval_loss": 1.1596688032150269,
      "eval_runtime": 56.2736,
      "eval_samples_per_second": 79.735,
      "eval_steps_per_second": 4.993,
      "step": 5000
    },
    {
      "epoch": 6.06,
      "learning_rate": 1.9714964370546318e-05,
      "loss": 1.1989,
      "step": 5100
    },
    {
      "epoch": 6.18,
      "learning_rate": 1.9121140142517815e-05,
      "loss": 1.2272,
      "step": 5200
    },
    {
      "epoch": 6.29,
      "learning_rate": 1.8527315914489312e-05,
      "loss": 1.2426,
      "step": 5300
    },
    {
      "epoch": 6.41,
      "learning_rate": 1.793349168646081e-05,
      "loss": 1.1808,
      "step": 5400
    },
    {
      "epoch": 6.53,
      "learning_rate": 1.7339667458432306e-05,
      "loss": 1.1669,
      "step": 5500
    },
    {
      "epoch": 6.53,
      "eval_loss": 1.141239047050476,
      "eval_runtime": 56.1754,
      "eval_samples_per_second": 79.875,
      "eval_steps_per_second": 5.002,
      "step": 5500
    },
    {
      "epoch": 6.65,
      "learning_rate": 1.6745843230403803e-05,
      "loss": 1.2044,
      "step": 5600
    },
    {
      "epoch": 6.77,
      "learning_rate": 1.61520190023753e-05,
      "loss": 1.1781,
      "step": 5700
    },
    {
      "epoch": 6.89,
      "learning_rate": 1.5558194774346793e-05,
      "loss": 1.1757,
      "step": 5800
    },
    {
      "epoch": 7.01,
      "learning_rate": 1.496437054631829e-05,
      "loss": 1.1777,
      "step": 5900
    },
    {
      "epoch": 7.13,
      "learning_rate": 1.4370546318289787e-05,
      "loss": 1.1529,
      "step": 6000
    },
    {
      "epoch": 7.13,
      "eval_loss": 1.1352198123931885,
      "eval_runtime": 56.8461,
      "eval_samples_per_second": 78.932,
      "eval_steps_per_second": 4.943,
      "step": 6000
    },
    {
      "epoch": 7.24,
      "learning_rate": 1.3776722090261283e-05,
      "loss": 1.1854,
      "step": 6100
    },
    {
      "epoch": 7.36,
      "learning_rate": 1.318289786223278e-05,
      "loss": 1.1736,
      "step": 6200
    },
    {
      "epoch": 7.48,
      "learning_rate": 1.2589073634204277e-05,
      "loss": 1.1932,
      "step": 6300
    },
    {
      "epoch": 7.6,
      "learning_rate": 1.1995249406175772e-05,
      "loss": 1.1431,
      "step": 6400
    },
    {
      "epoch": 7.72,
      "learning_rate": 1.1401425178147269e-05,
      "loss": 1.1532,
      "step": 6500
    },
    {
      "epoch": 7.72,
      "eval_loss": 1.1115353107452393,
      "eval_runtime": 56.4012,
      "eval_samples_per_second": 79.555,
      "eval_steps_per_second": 4.982,
      "step": 6500
    },
    {
      "epoch": 7.84,
      "learning_rate": 1.0807600950118766e-05,
      "loss": 1.1762,
      "step": 6600
    },
    {
      "epoch": 7.96,
      "learning_rate": 1.0213776722090261e-05,
      "loss": 1.1327,
      "step": 6700
    },
    {
      "epoch": 8.08,
      "learning_rate": 9.619952494061758e-06,
      "loss": 1.1492,
      "step": 6800
    },
    {
      "epoch": 8.19,
      "learning_rate": 9.026128266033253e-06,
      "loss": 1.1214,
      "step": 6900
    },
    {
      "epoch": 8.31,
      "learning_rate": 8.432304038004752e-06,
      "loss": 1.1684,
      "step": 7000
    },
    {
      "epoch": 8.31,
      "eval_loss": 1.1077940464019775,
      "eval_runtime": 56.2764,
      "eval_samples_per_second": 79.731,
      "eval_steps_per_second": 4.993,
      "step": 7000
    },
    {
      "epoch": 8.43,
      "learning_rate": 7.838479809976247e-06,
      "loss": 1.154,
      "step": 7100
    },
    {
      "epoch": 8.55,
      "learning_rate": 7.244655581947744e-06,
      "loss": 1.1439,
      "step": 7200
    },
    {
      "epoch": 8.67,
      "learning_rate": 6.6508313539192404e-06,
      "loss": 1.1398,
      "step": 7300
    },
    {
      "epoch": 8.79,
      "learning_rate": 6.0570071258907366e-06,
      "loss": 1.117,
      "step": 7400
    },
    {
      "epoch": 8.91,
      "learning_rate": 5.4631828978622335e-06,
      "loss": 1.1286,
      "step": 7500
    },
    {
      "epoch": 8.91,
      "eval_loss": 1.0950287580490112,
      "eval_runtime": 56.0134,
      "eval_samples_per_second": 80.106,
      "eval_steps_per_second": 5.017,
      "step": 7500
    },
    {
      "epoch": 9.03,
      "learning_rate": 4.86935866983373e-06,
      "loss": 1.1485,
      "step": 7600
    },
    {
      "epoch": 9.14,
      "learning_rate": 4.275534441805226e-06,
      "loss": 1.1273,
      "step": 7700
    },
    {
      "epoch": 9.26,
      "learning_rate": 3.681710213776722e-06,
      "loss": 1.1412,
      "step": 7800
    },
    {
      "epoch": 9.38,
      "learning_rate": 3.0878859857482185e-06,
      "loss": 1.1206,
      "step": 7900
    },
    {
      "epoch": 9.5,
      "learning_rate": 2.494061757719715e-06,
      "loss": 1.1416,
      "step": 8000
    },
    {
      "epoch": 9.5,
      "eval_loss": 1.0948013067245483,
      "eval_runtime": 56.75,
      "eval_samples_per_second": 79.066,
      "eval_steps_per_second": 4.952,
      "step": 8000
    },
    {
      "epoch": 9.62,
      "learning_rate": 1.9002375296912116e-06,
      "loss": 1.1101,
      "step": 8100
    },
    {
      "epoch": 9.74,
      "learning_rate": 1.306413301662708e-06,
      "loss": 1.1181,
      "step": 8200
    },
    {
      "epoch": 9.86,
      "learning_rate": 7.125890736342043e-07,
      "loss": 1.0831,
      "step": 8300
    },
    {
      "epoch": 9.98,
      "learning_rate": 1.1876484560570073e-07,
      "loss": 1.1164,
      "step": 8400
    },
    {
      "epoch": 10.0,
      "step": 8420,
      "total_flos": 3.5427566278656e+16,
      "train_loss": 1.365048002347244,
      "train_runtime": 5770.015,
      "train_samples_per_second": 23.327,
      "train_steps_per_second": 1.459
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.0815315246582031,
      "eval_runtime": 57.2783,
      "eval_samples_per_second": 78.337,
      "eval_steps_per_second": 4.906,
      "step": 8420
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.0882647037506104,
      "eval_runtime": 56.9424,
      "eval_samples_per_second": 78.799,
      "eval_steps_per_second": 4.935,
      "step": 8420
    }
  ],
  "logging_steps": 100,
  "max_steps": 8420,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 3.5427566278656e+16,
  "trial_name": null,
  "trial_params": null
}