{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "global_step": 49420,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "learning_rate": 4.896398219344395e-05,
      "loss": 0.2574,
      "step": 1024
    },
    {
      "epoch": 0.1,
      "eval_loss": 0.7888669967651367,
      "eval_runtime": 25.6574,
      "eval_samples_per_second": 81.536,
      "eval_steps_per_second": 3.43,
      "step": 1024
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.79279643868879e-05,
      "loss": 0.2633,
      "step": 2048
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.780219316482544,
      "eval_runtime": 25.6023,
      "eval_samples_per_second": 81.711,
      "eval_steps_per_second": 3.437,
      "step": 2048
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.6891946580331854e-05,
      "loss": 0.2635,
      "step": 3072
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.7877150774002075,
      "eval_runtime": 25.6162,
      "eval_samples_per_second": 81.667,
      "eval_steps_per_second": 3.435,
      "step": 3072
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.5855928773775804e-05,
      "loss": 0.2639,
      "step": 4096
    },
    {
      "epoch": 0.41,
      "eval_loss": 0.7751328945159912,
      "eval_runtime": 25.6157,
      "eval_samples_per_second": 81.669,
      "eval_steps_per_second": 3.435,
      "step": 4096
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.4819910967219755e-05,
      "loss": 0.2625,
      "step": 5120
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.7836451530456543,
      "eval_runtime": 25.6812,
      "eval_samples_per_second": 81.46,
      "eval_steps_per_second": 3.427,
      "step": 5120
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.3783893160663705e-05,
      "loss": 0.2609,
      "step": 6144
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.7758111357688904,
      "eval_runtime": 25.6349,
      "eval_samples_per_second": 81.607,
      "eval_steps_per_second": 3.433,
      "step": 6144
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.2747875354107656e-05,
      "loss": 0.2597,
      "step": 7168
    },
    {
      "epoch": 0.73,
      "eval_loss": 0.7923067212104797,
      "eval_runtime": 25.6218,
      "eval_samples_per_second": 81.649,
      "eval_steps_per_second": 3.435,
      "step": 7168
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.17118575475516e-05,
      "loss": 0.2612,
      "step": 8192
    },
    {
      "epoch": 0.83,
      "eval_loss": 0.8000349402427673,
      "eval_runtime": 25.646,
      "eval_samples_per_second": 81.572,
      "eval_steps_per_second": 3.431,
      "step": 8192
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.0676851477134767e-05,
      "loss": 0.2612,
      "step": 9216
    },
    {
      "epoch": 0.93,
      "eval_loss": 0.7934667468070984,
      "eval_runtime": 25.6816,
      "eval_samples_per_second": 81.459,
      "eval_steps_per_second": 3.427,
      "step": 9216
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.964083367057872e-05,
      "loss": 0.2554,
      "step": 10240
    },
    {
      "epoch": 1.04,
      "eval_loss": 0.7942602038383484,
      "eval_runtime": 25.7479,
      "eval_samples_per_second": 81.249,
      "eval_steps_per_second": 3.418,
      "step": 10240
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.860481586402267e-05,
      "loss": 0.2524,
      "step": 11264
    },
    {
      "epoch": 1.14,
      "eval_loss": 0.8014633655548096,
      "eval_runtime": 25.6355,
      "eval_samples_per_second": 81.606,
      "eval_steps_per_second": 3.433,
      "step": 11264
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.7569809793605834e-05,
      "loss": 0.2504,
      "step": 12288
    },
    {
      "epoch": 1.24,
      "eval_loss": 0.7962351441383362,
      "eval_runtime": 25.6019,
      "eval_samples_per_second": 81.713,
      "eval_steps_per_second": 3.437,
      "step": 12288
    },
    {
      "epoch": 1.35,
      "learning_rate": 3.653379198704978e-05,
      "loss": 0.2524,
      "step": 13312
    },
    {
      "epoch": 1.35,
      "eval_loss": 0.8086119294166565,
      "eval_runtime": 25.7264,
      "eval_samples_per_second": 81.317,
      "eval_steps_per_second": 3.421,
      "step": 13312
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.549777418049373e-05,
      "loss": 0.2521,
      "step": 14336
    },
    {
      "epoch": 1.45,
      "eval_loss": 0.8062009215354919,
      "eval_runtime": 25.6245,
      "eval_samples_per_second": 81.641,
      "eval_steps_per_second": 3.434,
      "step": 14336
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.446175637393768e-05,
      "loss": 0.2503,
      "step": 15360
    },
    {
      "epoch": 1.55,
      "eval_loss": 0.7997719049453735,
      "eval_runtime": 25.6588,
      "eval_samples_per_second": 81.532,
      "eval_steps_per_second": 3.43,
      "step": 15360
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.3426750303520846e-05,
      "loss": 0.2523,
      "step": 16384
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.8097875118255615,
      "eval_runtime": 25.6111,
      "eval_samples_per_second": 81.683,
      "eval_steps_per_second": 3.436,
      "step": 16384
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.239073249696479e-05,
      "loss": 0.251,
      "step": 17408
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.8213343620300293,
      "eval_runtime": 25.6868,
      "eval_samples_per_second": 81.443,
      "eval_steps_per_second": 3.426,
      "step": 17408
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.135572642654796e-05,
      "loss": 0.2509,
      "step": 18432
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.8137653470039368,
      "eval_runtime": 25.6223,
      "eval_samples_per_second": 81.648,
      "eval_steps_per_second": 3.435,
      "step": 18432
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.031970861999191e-05,
      "loss": 0.2533,
      "step": 19456
    },
    {
      "epoch": 1.97,
      "eval_loss": 0.8182128071784973,
      "eval_runtime": 25.6272,
      "eval_samples_per_second": 81.632,
      "eval_steps_per_second": 3.434,
      "step": 19456
    },
    {
      "epoch": 2.07,
      "learning_rate": 2.928369081343586e-05,
      "loss": 0.245,
      "step": 20480
    },
    {
      "epoch": 2.07,
      "eval_loss": 0.8290057182312012,
      "eval_runtime": 25.6288,
      "eval_samples_per_second": 81.627,
      "eval_steps_per_second": 3.434,
      "step": 20480
    },
    {
      "epoch": 2.18,
      "learning_rate": 2.824868474301902e-05,
      "loss": 0.2432,
      "step": 21504
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.8328016400337219,
      "eval_runtime": 25.7111,
      "eval_samples_per_second": 81.366,
      "eval_steps_per_second": 3.423,
      "step": 21504
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.7212666936462972e-05,
      "loss": 0.2435,
      "step": 22528
    },
    {
      "epoch": 2.28,
      "eval_loss": 0.8186773657798767,
      "eval_runtime": 25.61,
      "eval_samples_per_second": 81.687,
      "eval_steps_per_second": 3.436,
      "step": 22528
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.617766086604614e-05,
      "loss": 0.2423,
      "step": 23552
    },
    {
      "epoch": 2.38,
      "eval_loss": 0.8238145112991333,
      "eval_runtime": 25.6287,
      "eval_samples_per_second": 81.627,
      "eval_steps_per_second": 3.434,
      "step": 23552
    },
    {
      "epoch": 2.49,
      "learning_rate": 2.51426547956293e-05,
      "loss": 0.2443,
      "step": 24576
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.8248941898345947,
      "eval_runtime": 25.6771,
      "eval_samples_per_second": 81.473,
      "eval_steps_per_second": 3.427,
      "step": 24576
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.410663698907325e-05,
      "loss": 0.2431,
      "step": 25600
    },
    {
      "epoch": 2.59,
      "eval_loss": 0.8253324031829834,
      "eval_runtime": 25.6574,
      "eval_samples_per_second": 81.536,
      "eval_steps_per_second": 3.43,
      "step": 25600
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.30706191825172e-05,
      "loss": 0.2432,
      "step": 26624
    },
    {
      "epoch": 2.69,
      "eval_loss": 0.8268725275993347,
      "eval_runtime": 25.6215,
      "eval_samples_per_second": 81.65,
      "eval_steps_per_second": 3.435,
      "step": 26624
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.203460137596115e-05,
      "loss": 0.2421,
      "step": 27648
    },
    {
      "epoch": 2.8,
      "eval_loss": 0.8282283544540405,
      "eval_runtime": 25.6027,
      "eval_samples_per_second": 81.71,
      "eval_steps_per_second": 3.437,
      "step": 27648
    },
    {
      "epoch": 2.9,
      "learning_rate": 2.09985835694051e-05,
      "loss": 0.2421,
      "step": 28672
    },
    {
      "epoch": 2.9,
      "eval_loss": 0.8267728686332703,
      "eval_runtime": 25.6314,
      "eval_samples_per_second": 81.619,
      "eval_steps_per_second": 3.433,
      "step": 28672
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.9963577498988265e-05,
      "loss": 0.243,
      "step": 29696
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.8345196843147278,
      "eval_runtime": 25.6765,
      "eval_samples_per_second": 81.475,
      "eval_steps_per_second": 3.427,
      "step": 29696
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.8927559692432215e-05,
      "loss": 0.2367,
      "step": 30720
    },
    {
      "epoch": 3.11,
      "eval_loss": 0.8424174189567566,
      "eval_runtime": 25.634,
      "eval_samples_per_second": 81.61,
      "eval_steps_per_second": 3.433,
      "step": 30720
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.789255362201538e-05,
      "loss": 0.237,
      "step": 31744
    },
    {
      "epoch": 3.21,
      "eval_loss": 0.8373659253120422,
      "eval_runtime": 25.5978,
      "eval_samples_per_second": 81.726,
      "eval_steps_per_second": 3.438,
      "step": 31744
    },
    {
      "epoch": 3.32,
      "learning_rate": 1.6857547551598542e-05,
      "loss": 0.2351,
      "step": 32768
    },
    {
      "epoch": 3.32,
      "eval_loss": 0.8430664539337158,
      "eval_runtime": 25.6675,
      "eval_samples_per_second": 81.504,
      "eval_steps_per_second": 3.428,
      "step": 32768
    },
    {
      "epoch": 3.42,
      "learning_rate": 1.5821529745042493e-05,
      "loss": 0.2374,
      "step": 33792
    },
    {
      "epoch": 3.42,
      "eval_loss": 0.8425076603889465,
      "eval_runtime": 25.7071,
      "eval_samples_per_second": 81.378,
      "eval_steps_per_second": 3.423,
      "step": 33792
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.4786523674625658e-05,
      "loss": 0.2355,
      "step": 34816
    },
    {
      "epoch": 3.52,
      "eval_loss": 0.8351919054985046,
      "eval_runtime": 25.6245,
      "eval_samples_per_second": 81.641,
      "eval_steps_per_second": 3.434,
      "step": 34816
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.3750505868069607e-05,
      "loss": 0.2373,
      "step": 35840
    },
    {
      "epoch": 3.63,
      "eval_loss": 0.845241904258728,
      "eval_runtime": 25.6392,
      "eval_samples_per_second": 81.594,
      "eval_steps_per_second": 3.432,
      "step": 35840
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.2714488061513558e-05,
      "loss": 0.2356,
      "step": 36864
    },
    {
      "epoch": 3.73,
      "eval_loss": 0.8382672071456909,
      "eval_runtime": 25.6391,
      "eval_samples_per_second": 81.594,
      "eval_steps_per_second": 3.432,
      "step": 36864
    },
    {
      "epoch": 3.83,
      "learning_rate": 1.1678470254957508e-05,
      "loss": 0.2343,
      "step": 37888
    },
    {
      "epoch": 3.83,
      "eval_loss": 0.8443974852561951,
      "eval_runtime": 25.7018,
      "eval_samples_per_second": 81.395,
      "eval_steps_per_second": 3.424,
      "step": 37888
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.0642452448401457e-05,
      "loss": 0.2348,
      "step": 38912
    },
    {
      "epoch": 3.94,
      "eval_loss": 0.8427897095680237,
      "eval_runtime": 25.6146,
      "eval_samples_per_second": 81.672,
      "eval_steps_per_second": 3.436,
      "step": 38912
    },
    {
      "epoch": 4.04,
      "learning_rate": 9.606434641845407e-06,
      "loss": 0.2349,
      "step": 39936
    },
    {
      "epoch": 4.04,
      "eval_loss": 0.8479835987091064,
      "eval_runtime": 25.6232,
      "eval_samples_per_second": 81.645,
      "eval_steps_per_second": 3.434,
      "step": 39936
    },
    {
      "epoch": 4.14,
      "learning_rate": 8.571428571428573e-06,
      "loss": 0.2327,
      "step": 40960
    },
    {
      "epoch": 4.14,
      "eval_loss": 0.8481870889663696,
      "eval_runtime": 25.6261,
      "eval_samples_per_second": 81.636,
      "eval_steps_per_second": 3.434,
      "step": 40960
    },
    {
      "epoch": 4.25,
      "learning_rate": 7.536422501011736e-06,
      "loss": 0.2337,
      "step": 41984
    },
    {
      "epoch": 4.25,
      "eval_loss": 0.850956916809082,
      "eval_runtime": 25.6867,
      "eval_samples_per_second": 81.443,
      "eval_steps_per_second": 3.426,
      "step": 41984
    },
    {
      "epoch": 4.35,
      "learning_rate": 6.500404694455686e-06,
      "loss": 0.2288,
      "step": 43008
    },
    {
      "epoch": 4.35,
      "eval_loss": 0.8499058485031128,
      "eval_runtime": 25.6219,
      "eval_samples_per_second": 81.649,
      "eval_steps_per_second": 3.435,
      "step": 43008
    },
    {
      "epoch": 4.45,
      "learning_rate": 5.464386887899636e-06,
      "loss": 0.2299,
      "step": 44032
    },
    {
      "epoch": 4.45,
      "eval_loss": 0.8522343039512634,
      "eval_runtime": 25.6037,
      "eval_samples_per_second": 81.707,
      "eval_steps_per_second": 3.437,
      "step": 44032
    },
    {
      "epoch": 4.56,
      "learning_rate": 4.429380817482801e-06,
      "loss": 0.2277,
      "step": 45056
    },
    {
      "epoch": 4.56,
      "eval_loss": 0.85256427526474,
      "eval_runtime": 25.6116,
      "eval_samples_per_second": 81.682,
      "eval_steps_per_second": 3.436,
      "step": 45056
    },
    {
      "epoch": 4.66,
      "learning_rate": 3.3933630109267505e-06,
      "loss": 0.2301,
      "step": 46080
    },
    {
      "epoch": 4.66,
      "eval_loss": 0.851773202419281,
      "eval_runtime": 25.6644,
      "eval_samples_per_second": 81.514,
      "eval_steps_per_second": 3.429,
      "step": 46080
    },
    {
      "epoch": 4.77,
      "learning_rate": 2.3573452043707e-06,
      "loss": 0.2312,
      "step": 47104
    },
    {
      "epoch": 4.77,
      "eval_loss": 0.8511470556259155,
      "eval_runtime": 25.6121,
      "eval_samples_per_second": 81.68,
      "eval_steps_per_second": 3.436,
      "step": 47104
    },
    {
      "epoch": 4.87,
      "learning_rate": 1.32132739781465e-06,
      "loss": 0.2284,
      "step": 48128
    },
    {
      "epoch": 4.87,
      "eval_loss": 0.8506836295127869,
      "eval_runtime": 25.6413,
      "eval_samples_per_second": 81.587,
      "eval_steps_per_second": 3.432,
      "step": 48128
    },
    {
      "epoch": 4.97,
      "learning_rate": 2.8632132739781467e-07,
      "loss": 0.2294,
      "step": 49152
    },
    {
      "epoch": 4.97,
      "eval_loss": 0.8519254922866821,
      "eval_runtime": 25.6136,
      "eval_samples_per_second": 81.675,
      "eval_steps_per_second": 3.436,
      "step": 49152
    },
    {
      "epoch": 5.0,
      "step": 49420,
      "total_flos": 3.1008084491562394e+17,
      "train_loss": 0.244437825115332,
      "train_runtime": 40418.7139,
      "train_samples_per_second": 25.676,
      "train_steps_per_second": 1.223
    }
  ],
  "max_steps": 49420,
  "num_train_epochs": 5,
  "total_flos": 3.1008084491562394e+17,
  "trial_name": null,
  "trial_params": null
}