{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.984974958263773,
  "eval_steps": 10,
  "global_step": 447,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 10
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.3519736842105263,
      "eval_loss": 0.6503910422325134,
      "eval_runtime": 16.3619,
      "eval_samples_per_second": 18.58,
      "eval_steps_per_second": 4.645,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 3e-05,
      "loss": 0.6652,
      "step": 20
    },
    {
      "epoch": 0.13,
      "eval_accuracy": 0.3782894736842105,
      "eval_loss": 0.6469289660453796,
      "eval_runtime": 16.3596,
      "eval_samples_per_second": 18.582,
      "eval_steps_per_second": 4.646,
      "step": 20
    },
    {
      "epoch": 0.2,
      "learning_rate": 3e-05,
      "loss": 0.6523,
      "step": 30
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.3651315789473684,
      "eval_loss": 0.6429744362831116,
      "eval_runtime": 16.3534,
      "eval_samples_per_second": 18.589,
      "eval_steps_per_second": 4.647,
      "step": 30
    },
    {
      "epoch": 0.27,
      "learning_rate": 3e-05,
      "loss": 0.613,
      "step": 40
    },
    {
      "epoch": 0.27,
      "eval_accuracy": 0.40789473684210525,
      "eval_loss": 0.6341487765312195,
      "eval_runtime": 16.4446,
      "eval_samples_per_second": 18.486,
      "eval_steps_per_second": 4.622,
      "step": 40
    },
    {
      "epoch": 0.33,
      "learning_rate": 3e-05,
      "loss": 0.6586,
      "step": 50
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.3881578947368421,
      "eval_loss": 0.6206462979316711,
      "eval_runtime": 16.366,
      "eval_samples_per_second": 18.575,
      "eval_steps_per_second": 4.644,
      "step": 50
    },
    {
      "epoch": 0.4,
      "learning_rate": 3e-05,
      "loss": 0.586,
      "step": 60
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.41776315789473684,
      "eval_loss": 0.6268599033355713,
      "eval_runtime": 16.3556,
      "eval_samples_per_second": 18.587,
      "eval_steps_per_second": 4.647,
      "step": 60
    },
    {
      "epoch": 0.47,
      "learning_rate": 3e-05,
      "loss": 0.594,
      "step": 70
    },
    {
      "epoch": 0.47,
      "eval_accuracy": 0.4276315789473684,
      "eval_loss": 0.6045505404472351,
      "eval_runtime": 16.3658,
      "eval_samples_per_second": 18.575,
      "eval_steps_per_second": 4.644,
      "step": 70
    },
    {
      "epoch": 0.53,
      "learning_rate": 3e-05,
      "loss": 0.6063,
      "step": 80
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.41776315789473684,
      "eval_loss": 0.6135305762290955,
      "eval_runtime": 16.3648,
      "eval_samples_per_second": 18.576,
      "eval_steps_per_second": 4.644,
      "step": 80
    },
    {
      "epoch": 0.6,
      "learning_rate": 3e-05,
      "loss": 0.5988,
      "step": 90
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.4276315789473684,
      "eval_loss": 0.6097424626350403,
      "eval_runtime": 16.3755,
      "eval_samples_per_second": 18.564,
      "eval_steps_per_second": 4.641,
      "step": 90
    },
    {
      "epoch": 0.67,
      "learning_rate": 3e-05,
      "loss": 0.6217,
      "step": 100
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.45394736842105265,
      "eval_loss": 0.6098220348358154,
      "eval_runtime": 16.3674,
      "eval_samples_per_second": 18.574,
      "eval_steps_per_second": 4.643,
      "step": 100
    },
    {
      "epoch": 0.73,
      "learning_rate": 3e-05,
      "loss": 0.5817,
      "step": 110
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.45394736842105265,
      "eval_loss": 0.6021943092346191,
      "eval_runtime": 16.3678,
      "eval_samples_per_second": 18.573,
      "eval_steps_per_second": 4.643,
      "step": 110
    },
    {
      "epoch": 0.8,
      "learning_rate": 3e-05,
      "loss": 0.6219,
      "step": 120
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.45723684210526316,
      "eval_loss": 0.5926041007041931,
      "eval_runtime": 16.3627,
      "eval_samples_per_second": 18.579,
      "eval_steps_per_second": 4.645,
      "step": 120
    },
    {
      "epoch": 0.87,
      "learning_rate": 3e-05,
      "loss": 0.559,
      "step": 130
    },
    {
      "epoch": 0.87,
      "eval_accuracy": 0.4605263157894737,
      "eval_loss": 0.5816267728805542,
      "eval_runtime": 16.3673,
      "eval_samples_per_second": 18.574,
      "eval_steps_per_second": 4.643,
      "step": 130
    },
    {
      "epoch": 0.93,
      "learning_rate": 3e-05,
      "loss": 0.5514,
      "step": 140
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.47368421052631576,
      "eval_loss": 0.5783373713493347,
      "eval_runtime": 16.368,
      "eval_samples_per_second": 18.573,
      "eval_steps_per_second": 4.643,
      "step": 140
    },
    {
      "epoch": 1.0,
      "learning_rate": 3e-05,
      "loss": 0.59,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4868421052631579,
      "eval_loss": 0.5621668100357056,
      "eval_runtime": 16.3648,
      "eval_samples_per_second": 18.576,
      "eval_steps_per_second": 4.644,
      "step": 150
    },
    {
      "epoch": 1.07,
      "learning_rate": 3e-05,
      "loss": 0.46,
      "step": 160
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.48026315789473684,
      "eval_loss": 0.5867581367492676,
      "eval_runtime": 16.3579,
      "eval_samples_per_second": 18.584,
      "eval_steps_per_second": 4.646,
      "step": 160
    },
    {
      "epoch": 1.14,
      "learning_rate": 3e-05,
      "loss": 0.4484,
      "step": 170
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.4868421052631579,
      "eval_loss": 0.5666611194610596,
      "eval_runtime": 16.3613,
      "eval_samples_per_second": 18.58,
      "eval_steps_per_second": 4.645,
      "step": 170
    },
    {
      "epoch": 1.2,
      "learning_rate": 3e-05,
      "loss": 0.4162,
      "step": 180
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.48026315789473684,
      "eval_loss": 0.5819750428199768,
      "eval_runtime": 16.3725,
      "eval_samples_per_second": 18.568,
      "eval_steps_per_second": 4.642,
      "step": 180
    },
    {
      "epoch": 1.27,
      "learning_rate": 3e-05,
      "loss": 0.4716,
      "step": 190
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 0.46381578947368424,
      "eval_loss": 0.590432345867157,
      "eval_runtime": 16.361,
      "eval_samples_per_second": 18.581,
      "eval_steps_per_second": 4.645,
      "step": 190
    },
    {
      "epoch": 1.34,
      "learning_rate": 3e-05,
      "loss": 0.4486,
      "step": 200
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.5098684210526315,
      "eval_loss": 0.5777420997619629,
      "eval_runtime": 16.354,
      "eval_samples_per_second": 18.589,
      "eval_steps_per_second": 4.647,
      "step": 200
    },
    {
      "epoch": 1.4,
      "learning_rate": 3e-05,
      "loss": 0.4264,
      "step": 210
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.4967105263157895,
      "eval_loss": 0.6482496857643127,
      "eval_runtime": 16.3696,
      "eval_samples_per_second": 18.571,
      "eval_steps_per_second": 4.643,
      "step": 210
    },
    {
      "epoch": 1.47,
      "learning_rate": 3e-05,
      "loss": 0.4236,
      "step": 220
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.5032894736842105,
      "eval_loss": 0.5741015076637268,
      "eval_runtime": 16.3648,
      "eval_samples_per_second": 18.576,
      "eval_steps_per_second": 4.644,
      "step": 220
    },
    {
      "epoch": 1.54,
      "learning_rate": 3e-05,
      "loss": 0.4141,
      "step": 230
    },
    {
      "epoch": 1.54,
      "eval_accuracy": 0.5164473684210527,
      "eval_loss": 0.5607666373252869,
      "eval_runtime": 16.3604,
      "eval_samples_per_second": 18.581,
      "eval_steps_per_second": 4.645,
      "step": 230
    },
    {
      "epoch": 1.6,
      "learning_rate": 3e-05,
      "loss": 0.4308,
      "step": 240
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.5098684210526315,
      "eval_loss": 0.5539161562919617,
      "eval_runtime": 16.3597,
      "eval_samples_per_second": 18.582,
      "eval_steps_per_second": 4.646,
      "step": 240
    },
    {
      "epoch": 1.67,
      "learning_rate": 3e-05,
      "loss": 0.4505,
      "step": 250
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.5032894736842105,
      "eval_loss": 0.5494562387466431,
      "eval_runtime": 16.3671,
      "eval_samples_per_second": 18.574,
      "eval_steps_per_second": 4.643,
      "step": 250
    },
    {
      "epoch": 1.74,
      "learning_rate": 3e-05,
      "loss": 0.3958,
      "step": 260
    },
    {
      "epoch": 1.74,
      "eval_accuracy": 0.5098684210526315,
      "eval_loss": 0.5593812465667725,
      "eval_runtime": 16.3598,
      "eval_samples_per_second": 18.582,
      "eval_steps_per_second": 4.646,
      "step": 260
    },
    {
      "epoch": 1.8,
      "learning_rate": 3e-05,
      "loss": 0.4432,
      "step": 270
    },
    {
      "epoch": 1.8,
      "eval_accuracy": 0.5164473684210527,
      "eval_loss": 0.5492013096809387,
      "eval_runtime": 16.3635,
      "eval_samples_per_second": 18.578,
      "eval_steps_per_second": 4.644,
      "step": 270
    },
    {
      "epoch": 1.87,
      "learning_rate": 3e-05,
      "loss": 0.4067,
      "step": 280
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.506578947368421,
      "eval_loss": 0.6023809313774109,
      "eval_runtime": 16.3658,
      "eval_samples_per_second": 18.575,
      "eval_steps_per_second": 4.644,
      "step": 280
    },
    {
      "epoch": 1.94,
      "learning_rate": 3e-05,
      "loss": 0.3988,
      "step": 290
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.5098684210526315,
      "eval_loss": 0.5606762766838074,
      "eval_runtime": 16.3642,
      "eval_samples_per_second": 18.577,
      "eval_steps_per_second": 4.644,
      "step": 290
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 0.3992,
      "step": 300
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5164473684210527,
      "eval_loss": 0.5669550895690918,
      "eval_runtime": 16.3688,
      "eval_samples_per_second": 18.572,
      "eval_steps_per_second": 4.643,
      "step": 300
    },
    {
      "epoch": 2.07,
      "learning_rate": 3e-05,
      "loss": 0.2304,
      "step": 310
    },
    {
      "epoch": 2.07,
      "eval_accuracy": 0.5361842105263158,
      "eval_loss": 0.8199814558029175,
      "eval_runtime": 16.3644,
      "eval_samples_per_second": 18.577,
      "eval_steps_per_second": 4.644,
      "step": 310
    },
    {
      "epoch": 2.14,
      "learning_rate": 3e-05,
      "loss": 0.1696,
      "step": 320
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.5296052631578947,
      "eval_loss": 0.9087279438972473,
      "eval_runtime": 16.362,
      "eval_samples_per_second": 18.58,
      "eval_steps_per_second": 4.645,
      "step": 320
    },
    {
      "epoch": 2.2,
      "learning_rate": 3e-05,
      "loss": 0.2255,
      "step": 330
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.5361842105263158,
      "eval_loss": 0.7565640211105347,
      "eval_runtime": 16.3681,
      "eval_samples_per_second": 18.573,
      "eval_steps_per_second": 4.643,
      "step": 330
    },
    {
      "epoch": 2.27,
      "learning_rate": 3e-05,
      "loss": 0.1923,
      "step": 340
    },
    {
      "epoch": 2.27,
      "eval_accuracy": 0.5197368421052632,
      "eval_loss": 0.701989471912384,
      "eval_runtime": 16.3692,
      "eval_samples_per_second": 18.571,
      "eval_steps_per_second": 4.643,
      "step": 340
    },
    {
      "epoch": 2.34,
      "learning_rate": 3e-05,
      "loss": 0.281,
      "step": 350
    },
    {
      "epoch": 2.34,
      "eval_accuracy": 0.5032894736842105,
      "eval_loss": 0.6653422117233276,
      "eval_runtime": 16.3703,
      "eval_samples_per_second": 18.57,
      "eval_steps_per_second": 4.643,
      "step": 350
    },
    {
      "epoch": 2.4,
      "learning_rate": 3e-05,
      "loss": 0.2311,
      "step": 360
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.5131578947368421,
      "eval_loss": 0.6411683559417725,
      "eval_runtime": 16.3702,
      "eval_samples_per_second": 18.57,
      "eval_steps_per_second": 4.643,
      "step": 360
    },
    {
      "epoch": 2.47,
      "learning_rate": 3e-05,
      "loss": 0.1523,
      "step": 370
    },
    {
      "epoch": 2.47,
      "eval_accuracy": 0.5230263157894737,
      "eval_loss": 0.8846220970153809,
      "eval_runtime": 16.3728,
      "eval_samples_per_second": 18.567,
      "eval_steps_per_second": 4.642,
      "step": 370
    },
    {
      "epoch": 2.54,
      "learning_rate": 3e-05,
      "loss": 0.2451,
      "step": 380
    },
    {
      "epoch": 2.54,
      "eval_accuracy": 0.5164473684210527,
      "eval_loss": 0.9251819252967834,
      "eval_runtime": 16.379,
      "eval_samples_per_second": 18.56,
      "eval_steps_per_second": 4.64,
      "step": 380
    },
    {
      "epoch": 2.6,
      "learning_rate": 3e-05,
      "loss": 0.2022,
      "step": 390
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.5197368421052632,
      "eval_loss": 0.7422206401824951,
      "eval_runtime": 16.3727,
      "eval_samples_per_second": 18.567,
      "eval_steps_per_second": 4.642,
      "step": 390
    },
    {
      "epoch": 2.67,
      "learning_rate": 3e-05,
      "loss": 0.217,
      "step": 400
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.5328947368421053,
      "eval_loss": 0.7557851076126099,
      "eval_runtime": 16.3638,
      "eval_samples_per_second": 18.578,
      "eval_steps_per_second": 4.644,
      "step": 400
    },
    {
      "epoch": 2.74,
      "learning_rate": 3e-05,
      "loss": 0.165,
      "step": 410
    },
    {
      "epoch": 2.74,
      "eval_accuracy": 0.5427631578947368,
      "eval_loss": 0.7846018075942993,
      "eval_runtime": 16.3674,
      "eval_samples_per_second": 18.574,
      "eval_steps_per_second": 4.643,
      "step": 410
    },
    {
      "epoch": 2.8,
      "learning_rate": 3e-05,
      "loss": 0.2025,
      "step": 420
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 0.5230263157894737,
      "eval_loss": 0.725389301776886,
      "eval_runtime": 16.3775,
      "eval_samples_per_second": 18.562,
      "eval_steps_per_second": 4.641,
      "step": 420
    },
    {
      "epoch": 2.87,
      "learning_rate": 3e-05,
      "loss": 0.2201,
      "step": 430
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.5296052631578947,
      "eval_loss": 0.6530900001525879,
      "eval_runtime": 16.3694,
      "eval_samples_per_second": 18.571,
      "eval_steps_per_second": 4.643,
      "step": 430
    },
    {
      "epoch": 2.94,
      "learning_rate": 3e-05,
      "loss": 0.2037,
      "step": 440
    },
    {
      "epoch": 2.94,
      "eval_accuracy": 0.5493421052631579,
      "eval_loss": 0.7827179431915283,
      "eval_runtime": 16.3682,
      "eval_samples_per_second": 18.573,
      "eval_steps_per_second": 4.643,
      "step": 440
    },
    {
      "epoch": 2.98,
      "step": 447,
      "total_flos": 1.09848374069035e+17,
      "train_loss": 0.43748926956381573,
      "train_runtime": 1886.9121,
      "train_samples_per_second": 3.809,
      "train_steps_per_second": 0.237
    }
  ],
  "logging_steps": 10,
  "max_steps": 447,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 250,
  "total_flos": 1.09848374069035e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}