{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.998657718120805,
  "global_step": 3720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.27,
      "learning_rate": 6.904632152588556e-06,
      "loss": 0.6417,
      "step": 100
    },
    {
      "epoch": 0.27,
      "eval_accuracy": 0.653323233127594,
      "eval_loss": 0.6283483505249023,
      "eval_runtime": 8.5429,
      "eval_samples_per_second": 154.983,
      "eval_steps_per_second": 9.716,
      "step": 100
    },
    {
      "epoch": 0.54,
      "learning_rate": 6.713896457765667e-06,
      "loss": 0.5105,
      "step": 200
    },
    {
      "epoch": 0.54,
      "eval_accuracy": 0.791540801525116,
      "eval_loss": 0.45882758498191833,
      "eval_runtime": 8.3778,
      "eval_samples_per_second": 158.036,
      "eval_steps_per_second": 9.907,
      "step": 200
    },
    {
      "epoch": 0.81,
      "learning_rate": 6.523160762942779e-06,
      "loss": 0.4554,
      "step": 300
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.7968277931213379,
      "eval_loss": 0.44999730587005615,
      "eval_runtime": 8.4819,
      "eval_samples_per_second": 156.097,
      "eval_steps_per_second": 9.786,
      "step": 300
    },
    {
      "epoch": 1.08,
      "learning_rate": 6.332425068119891e-06,
      "loss": 0.4212,
      "step": 400
    },
    {
      "epoch": 1.08,
      "eval_accuracy": 0.7938066720962524,
      "eval_loss": 0.47725316882133484,
      "eval_runtime": 8.3161,
      "eval_samples_per_second": 159.209,
      "eval_steps_per_second": 9.981,
      "step": 400
    },
    {
      "epoch": 1.34,
      "learning_rate": 6.141689373297002e-06,
      "loss": 0.4054,
      "step": 500
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.7983383536338806,
      "eval_loss": 0.43111804127693176,
      "eval_runtime": 8.5652,
      "eval_samples_per_second": 154.579,
      "eval_steps_per_second": 9.69,
      "step": 500
    },
    {
      "epoch": 1.61,
      "learning_rate": 5.950953678474114e-06,
      "loss": 0.3922,
      "step": 600
    },
    {
      "epoch": 1.61,
      "eval_accuracy": 0.7998489141464233,
      "eval_loss": 0.4587886333465576,
      "eval_runtime": 8.2972,
      "eval_samples_per_second": 159.573,
      "eval_steps_per_second": 10.003,
      "step": 600
    },
    {
      "epoch": 1.88,
      "learning_rate": 5.760217983651226e-06,
      "loss": 0.3776,
      "step": 700
    },
    {
      "epoch": 1.88,
      "eval_accuracy": 0.8066465258598328,
      "eval_loss": 0.43665462732315063,
      "eval_runtime": 8.2982,
      "eval_samples_per_second": 159.553,
      "eval_steps_per_second": 10.002,
      "step": 700
    },
    {
      "epoch": 2.15,
      "learning_rate": 5.569482288828338e-06,
      "loss": 0.3535,
      "step": 800
    },
    {
      "epoch": 2.15,
      "eval_accuracy": 0.8074018359184265,
      "eval_loss": 0.4674627184867859,
      "eval_runtime": 8.7016,
      "eval_samples_per_second": 152.155,
      "eval_steps_per_second": 9.538,
      "step": 800
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.37874659400545e-06,
      "loss": 0.33,
      "step": 900
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.8021147847175598,
      "eval_loss": 0.48743265867233276,
      "eval_runtime": 8.4057,
      "eval_samples_per_second": 157.512,
      "eval_steps_per_second": 9.874,
      "step": 900
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.188010899182561e-06,
      "loss": 0.3113,
      "step": 1000
    },
    {
      "epoch": 2.69,
      "eval_accuracy": 0.8043806552886963,
      "eval_loss": 0.49486756324768066,
      "eval_runtime": 8.2742,
      "eval_samples_per_second": 160.015,
      "eval_steps_per_second": 10.031,
      "step": 1000
    },
    {
      "epoch": 2.96,
      "learning_rate": 4.997275204359673e-06,
      "loss": 0.3203,
      "step": 1100
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 0.805891215801239,
      "eval_loss": 0.45497822761535645,
      "eval_runtime": 8.7885,
      "eval_samples_per_second": 150.651,
      "eval_steps_per_second": 9.444,
      "step": 1100
    },
    {
      "epoch": 3.23,
      "learning_rate": 4.806539509536785e-06,
      "loss": 0.248,
      "step": 1200
    },
    {
      "epoch": 3.23,
      "eval_accuracy": 0.8036254048347473,
      "eval_loss": 0.4857841432094574,
      "eval_runtime": 8.5367,
      "eval_samples_per_second": 155.095,
      "eval_steps_per_second": 9.723,
      "step": 1200
    },
    {
      "epoch": 3.49,
      "learning_rate": 4.615803814713896e-06,
      "loss": 0.2478,
      "step": 1300
    },
    {
      "epoch": 3.49,
      "eval_accuracy": 0.8028700947761536,
      "eval_loss": 0.5299096703529358,
      "eval_runtime": 8.4072,
      "eval_samples_per_second": 157.485,
      "eval_steps_per_second": 9.873,
      "step": 1300
    },
    {
      "epoch": 3.76,
      "learning_rate": 4.425068119891008e-06,
      "loss": 0.2371,
      "step": 1400
    },
    {
      "epoch": 3.76,
      "eval_accuracy": 0.7990936636924744,
      "eval_loss": 0.5012686252593994,
      "eval_runtime": 8.2774,
      "eval_samples_per_second": 159.953,
      "eval_steps_per_second": 10.027,
      "step": 1400
    },
    {
      "epoch": 4.03,
      "learning_rate": 4.2343324250681206e-06,
      "loss": 0.2388,
      "step": 1500
    },
    {
      "epoch": 4.03,
      "eval_accuracy": 0.8021147847175598,
      "eval_loss": 0.5519642233848572,
      "eval_runtime": 8.5777,
      "eval_samples_per_second": 154.355,
      "eval_steps_per_second": 9.676,
      "step": 1500
    },
    {
      "epoch": 4.3,
      "learning_rate": 4.043596730245232e-06,
      "loss": 0.1744,
      "step": 1600
    },
    {
      "epoch": 4.3,
      "eval_accuracy": 0.791540801525116,
      "eval_loss": 0.6686806082725525,
      "eval_runtime": 8.5418,
      "eval_samples_per_second": 155.002,
      "eval_steps_per_second": 9.717,
      "step": 1600
    },
    {
      "epoch": 4.57,
      "learning_rate": 3.8528610354223435e-06,
      "loss": 0.1788,
      "step": 1700
    },
    {
      "epoch": 4.57,
      "eval_accuracy": 0.768882155418396,
      "eval_loss": 0.756027340888977,
      "eval_runtime": 8.5142,
      "eval_samples_per_second": 155.506,
      "eval_steps_per_second": 9.748,
      "step": 1700
    },
    {
      "epoch": 4.84,
      "learning_rate": 3.6621253405994546e-06,
      "loss": 0.1652,
      "step": 1800
    },
    {
      "epoch": 4.84,
      "eval_accuracy": 0.7832326292991638,
      "eval_loss": 0.6984805464744568,
      "eval_runtime": 8.5496,
      "eval_samples_per_second": 154.861,
      "eval_steps_per_second": 9.708,
      "step": 1800
    },
    {
      "epoch": 5.11,
      "learning_rate": 3.4713896457765665e-06,
      "loss": 0.1596,
      "step": 1900
    },
    {
      "epoch": 5.11,
      "eval_accuracy": 0.791540801525116,
      "eval_loss": 0.7190905213356018,
      "eval_runtime": 8.5367,
      "eval_samples_per_second": 155.094,
      "eval_steps_per_second": 9.723,
      "step": 1900
    },
    {
      "epoch": 5.38,
      "learning_rate": 3.2806539509536784e-06,
      "loss": 0.1214,
      "step": 2000
    },
    {
      "epoch": 5.38,
      "eval_accuracy": 0.7892749309539795,
      "eval_loss": 0.909650981426239,
      "eval_runtime": 8.5312,
      "eval_samples_per_second": 155.195,
      "eval_steps_per_second": 9.729,
      "step": 2000
    },
    {
      "epoch": 5.64,
      "learning_rate": 3.0899182561307904e-06,
      "loss": 0.1432,
      "step": 2100
    },
    {
      "epoch": 5.64,
      "eval_accuracy": 0.7787008881568909,
      "eval_loss": 0.9183990359306335,
      "eval_runtime": 8.556,
      "eval_samples_per_second": 154.745,
      "eval_steps_per_second": 9.701,
      "step": 2100
    },
    {
      "epoch": 5.91,
      "learning_rate": 2.899182561307902e-06,
      "loss": 0.1145,
      "step": 2200
    },
    {
      "epoch": 5.91,
      "eval_accuracy": 0.7877643704414368,
      "eval_loss": 0.962020993232727,
      "eval_runtime": 8.5378,
      "eval_samples_per_second": 155.075,
      "eval_steps_per_second": 9.721,
      "step": 2200
    },
    {
      "epoch": 6.18,
      "learning_rate": 2.7084468664850138e-06,
      "loss": 0.1069,
      "step": 2300
    },
    {
      "epoch": 6.18,
      "eval_accuracy": 0.7892749309539795,
      "eval_loss": 0.9488775134086609,
      "eval_runtime": 8.3193,
      "eval_samples_per_second": 159.148,
      "eval_steps_per_second": 9.977,
      "step": 2300
    },
    {
      "epoch": 6.45,
      "learning_rate": 2.5177111716621253e-06,
      "loss": 0.1012,
      "step": 2400
    },
    {
      "epoch": 6.45,
      "eval_accuracy": 0.7817220687866211,
      "eval_loss": 1.0107486248016357,
      "eval_runtime": 8.2713,
      "eval_samples_per_second": 160.071,
      "eval_steps_per_second": 10.035,
      "step": 2400
    },
    {
      "epoch": 6.72,
      "learning_rate": 2.326975476839237e-06,
      "loss": 0.0942,
      "step": 2500
    },
    {
      "epoch": 6.72,
      "eval_accuracy": 0.7885196208953857,
      "eval_loss": 1.002103567123413,
      "eval_runtime": 8.5971,
      "eval_samples_per_second": 154.005,
      "eval_steps_per_second": 9.654,
      "step": 2500
    },
    {
      "epoch": 6.99,
      "learning_rate": 2.1362397820163487e-06,
      "loss": 0.087,
      "step": 2600
    },
    {
      "epoch": 6.99,
      "eval_accuracy": 0.791540801525116,
      "eval_loss": 1.1089540719985962,
      "eval_runtime": 8.5285,
      "eval_samples_per_second": 155.244,
      "eval_steps_per_second": 9.732,
      "step": 2600
    },
    {
      "epoch": 7.26,
      "learning_rate": 1.9455040871934606e-06,
      "loss": 0.0598,
      "step": 2700
    },
    {
      "epoch": 7.26,
      "eval_accuracy": 0.7794561982154846,
      "eval_loss": 1.1735244989395142,
      "eval_runtime": 8.6408,
      "eval_samples_per_second": 153.227,
      "eval_steps_per_second": 9.606,
      "step": 2700
    },
    {
      "epoch": 7.53,
      "learning_rate": 1.754768392370572e-06,
      "loss": 0.0742,
      "step": 2800
    },
    {
      "epoch": 7.53,
      "eval_accuracy": 0.7817220687866211,
      "eval_loss": 1.1433196067810059,
      "eval_runtime": 8.3802,
      "eval_samples_per_second": 157.992,
      "eval_steps_per_second": 9.904,
      "step": 2800
    },
    {
      "epoch": 7.79,
      "learning_rate": 1.5640326975476838e-06,
      "loss": 0.073,
      "step": 2900
    },
    {
      "epoch": 7.79,
      "eval_accuracy": 0.7953172326087952,
      "eval_loss": 1.1342711448669434,
      "eval_runtime": 8.549,
      "eval_samples_per_second": 154.873,
      "eval_steps_per_second": 9.709,
      "step": 2900
    },
    {
      "epoch": 8.06,
      "learning_rate": 1.3732970027247957e-06,
      "loss": 0.0553,
      "step": 3000
    },
    {
      "epoch": 8.06,
      "eval_accuracy": 0.7839879393577576,
      "eval_loss": 1.2258013486862183,
      "eval_runtime": 8.4899,
      "eval_samples_per_second": 155.95,
      "eval_steps_per_second": 9.776,
      "step": 3000
    },
    {
      "epoch": 8.33,
      "learning_rate": 1.1825613079019074e-06,
      "loss": 0.0474,
      "step": 3100
    },
    {
      "epoch": 8.33,
      "eval_accuracy": 0.7817220687866211,
      "eval_loss": 1.2460782527923584,
      "eval_runtime": 8.5462,
      "eval_samples_per_second": 154.923,
      "eval_steps_per_second": 9.712,
      "step": 3100
    },
    {
      "epoch": 8.6,
      "learning_rate": 9.91825613079019e-07,
      "loss": 0.0515,
      "step": 3200
    },
    {
      "epoch": 8.6,
      "eval_accuracy": 0.7824773192405701,
      "eval_loss": 1.2996242046356201,
      "eval_runtime": 8.5173,
      "eval_samples_per_second": 155.448,
      "eval_steps_per_second": 9.745,
      "step": 3200
    },
    {
      "epoch": 8.87,
      "learning_rate": 8.010899182561308e-07,
      "loss": 0.0551,
      "step": 3300
    },
    {
      "epoch": 8.87,
      "eval_accuracy": 0.7854984998703003,
      "eval_loss": 1.281937837600708,
      "eval_runtime": 8.2769,
      "eval_samples_per_second": 159.964,
      "eval_steps_per_second": 10.028,
      "step": 3300
    },
    {
      "epoch": 9.14,
      "learning_rate": 6.103542234332425e-07,
      "loss": 0.0541,
      "step": 3400
    },
    {
      "epoch": 9.14,
      "eval_accuracy": 0.7854984998703003,
      "eval_loss": 1.2807722091674805,
      "eval_runtime": 8.5739,
      "eval_samples_per_second": 154.423,
      "eval_steps_per_second": 9.681,
      "step": 3400
    },
    {
      "epoch": 9.41,
      "learning_rate": 4.196185286103542e-07,
      "loss": 0.0465,
      "step": 3500
    },
    {
      "epoch": 9.41,
      "eval_accuracy": 0.7817220687866211,
      "eval_loss": 1.3397763967514038,
      "eval_runtime": 8.503,
      "eval_samples_per_second": 155.71,
      "eval_steps_per_second": 9.761,
      "step": 3500
    },
    {
      "epoch": 9.68,
      "learning_rate": 2.288828337874659e-07,
      "loss": 0.0407,
      "step": 3600
    },
    {
      "epoch": 9.68,
      "eval_accuracy": 0.7824773192405701,
      "eval_loss": 1.3231298923492432,
      "eval_runtime": 8.2453,
      "eval_samples_per_second": 160.577,
      "eval_steps_per_second": 10.066,
      "step": 3600
    },
    {
      "epoch": 9.94,
      "learning_rate": 3.8147138964577657e-08,
      "loss": 0.0343,
      "step": 3700
    },
    {
      "epoch": 9.94,
      "eval_accuracy": 0.7824773192405701,
      "eval_loss": 1.3330293893814087,
      "eval_runtime": 8.5211,
      "eval_samples_per_second": 155.378,
      "eval_steps_per_second": 9.74,
      "step": 3700
    },
    {
      "epoch": 10.0,
      "step": 3720,
      "total_flos": 5.551925384610202e+16,
      "train_loss": 0.20273424893297176,
      "train_runtime": 2723.3449,
      "train_samples_per_second": 43.755,
      "train_steps_per_second": 1.366
    }
  ],
  "max_steps": 3720,
  "num_train_epochs": 10,
  "total_flos": 5.551925384610202e+16,
  "trial_name": null,
  "trial_params": null
}