{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7659607062157712,
  "global_step": 50000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.8817,
      "step": 500
    },
    {
      "epoch": 0.02,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.7274,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "learning_rate": 9e-07,
      "loss": 0.5446,
      "step": 1500
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.3974,
      "step": 2000
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.5e-06,
      "loss": 0.2731,
      "step": 2500
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.8e-06,
      "loss": 0.1759,
      "step": 3000
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.112,
      "step": 3500
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0785,
      "step": 4000
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.7e-06,
      "loss": 0.0652,
      "step": 4500
    },
    {
      "epoch": 0.08,
      "learning_rate": 3e-06,
      "loss": 0.0614,
      "step": 5000
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.3e-06,
      "loss": 0.0605,
      "step": 5500
    },
    {
      "epoch": 0.09,
      "learning_rate": 3.6e-06,
      "loss": 0.0602,
      "step": 6000
    },
    {
      "epoch": 0.1,
      "learning_rate": 3.9e-06,
      "loss": 0.0601,
      "step": 6500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0599,
      "step": 7000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.5e-06,
      "loss": 0.0598,
      "step": 7500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0597,
      "step": 8000
    },
    {
      "epoch": 0.13,
      "learning_rate": 5.1e-06,
      "loss": 0.0596,
      "step": 8500
    },
    {
      "epoch": 0.14,
      "learning_rate": 5.4e-06,
      "loss": 0.0595,
      "step": 9000
    },
    {
      "epoch": 0.15,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 0.0594,
      "step": 9500
    },
    {
      "epoch": 0.15,
      "learning_rate": 6e-06,
      "loss": 0.0594,
      "step": 10000
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.3e-06,
      "loss": 0.0593,
      "step": 10500
    },
    {
      "epoch": 0.17,
      "learning_rate": 6.6e-06,
      "loss": 0.0593,
      "step": 11000
    },
    {
      "epoch": 0.18,
      "learning_rate": 6.900000000000001e-06,
      "loss": 0.0592,
      "step": 11500
    },
    {
      "epoch": 0.18,
      "learning_rate": 7.2e-06,
      "loss": 0.0591,
      "step": 12000
    },
    {
      "epoch": 0.19,
      "learning_rate": 7.5e-06,
      "loss": 0.059,
      "step": 12500
    },
    {
      "epoch": 0.2,
      "learning_rate": 7.8e-06,
      "loss": 0.0588,
      "step": 13000
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.1e-06,
      "loss": 0.0586,
      "step": 13500
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.0584,
      "step": 14000
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.7e-06,
      "loss": 0.0583,
      "step": 14500
    },
    {
      "epoch": 0.23,
      "learning_rate": 9e-06,
      "loss": 0.0582,
      "step": 15000
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.3e-06,
      "loss": 0.0581,
      "step": 15500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0581,
      "step": 16000
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.9e-06,
      "loss": 0.058,
      "step": 16500
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.02e-05,
      "loss": 0.058,
      "step": 17000
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.05e-05,
      "loss": 0.058,
      "step": 17500
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.08e-05,
      "loss": 0.058,
      "step": 18000
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.11e-05,
      "loss": 0.058,
      "step": 18500
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.058,
      "step": 19000
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 0.058,
      "step": 19500
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.2e-05,
      "loss": 0.058,
      "step": 20000
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.2299999999999999e-05,
      "loss": 0.0579,
      "step": 20500
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.26e-05,
      "loss": 0.0579,
      "step": 21000
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.29e-05,
      "loss": 0.0579,
      "step": 21500
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.32e-05,
      "loss": 0.0579,
      "step": 22000
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.0579,
      "step": 22500
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.0579,
      "step": 23000
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 0.0579,
      "step": 23500
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.44e-05,
      "loss": 0.0579,
      "step": 24000
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.47e-05,
      "loss": 0.0579,
      "step": 24500
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 0.0579,
      "step": 25000
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.4999972660400536e-05,
      "loss": 0.0579,
      "step": 25500
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.4999890641901121e-05,
      "loss": 0.0578,
      "step": 26000
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.4999753945398704e-05,
      "loss": 0.0578,
      "step": 26500
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.499956257238817e-05,
      "loss": 0.0576,
      "step": 27000
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.4999316524962345e-05,
      "loss": 0.0574,
      "step": 27500
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.4999015805811964e-05,
      "loss": 0.057,
      "step": 28000
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.4998660418225645e-05,
      "loss": 0.0565,
      "step": 28500
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.4998250366089847e-05,
      "loss": 0.0561,
      "step": 29000
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.4997785653888835e-05,
      "loss": 0.0557,
      "step": 29500
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.499726628670463e-05,
      "loss": 0.0554,
      "step": 30000
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.4996692270216944e-05,
      "loss": 0.0551,
      "step": 30500
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.4996063610703138e-05,
      "loss": 0.0548,
      "step": 31000
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.4995380315038117e-05,
      "loss": 0.0545,
      "step": 31500
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.4994642390694308e-05,
      "loss": 0.0543,
      "step": 32000
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.4993849845741523e-05,
      "loss": 0.054,
      "step": 32500
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.499300268884691e-05,
      "loss": 0.0538,
      "step": 33000
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.4992100929274847e-05,
      "loss": 0.0537,
      "step": 33500
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.4991144576886822e-05,
      "loss": 0.0535,
      "step": 34000
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.4990133642141357e-05,
      "loss": 0.0533,
      "step": 34500
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.4989068136093872e-05,
      "loss": 0.0531,
      "step": 35000
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.498794807039657e-05,
      "loss": 0.053,
      "step": 35500
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.4986773457298309e-05,
      "loss": 0.0528,
      "step": 36000
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.4985544309644472e-05,
      "loss": 0.0527,
      "step": 36500
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.4984260640876819e-05,
      "loss": 0.0525,
      "step": 37000
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.4982922465033348e-05,
      "loss": 0.0524,
      "step": 37500
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.4981529796748133e-05,
      "loss": 0.0523,
      "step": 38000
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.4980082651251175e-05,
      "loss": 0.0522,
      "step": 38500
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.497858104436822e-05,
      "loss": 0.0521,
      "step": 39000
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.4977024992520602e-05,
      "loss": 0.052,
      "step": 39500
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.4975414512725056e-05,
      "loss": 0.0519,
      "step": 40000
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.4973749622593532e-05,
      "loss": 0.0518,
      "step": 40500
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.4972030340333e-05,
      "loss": 0.0517,
      "step": 41000
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.4970256684745257e-05,
      "loss": 0.0516,
      "step": 41500
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.4968428675226714e-05,
      "loss": 0.0516,
      "step": 42000
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.4966546331768191e-05,
      "loss": 0.0515,
      "step": 42500
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.4964609674954694e-05,
      "loss": 0.0514,
      "step": 43000
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.4962618725965194e-05,
      "loss": 0.0512,
      "step": 43500
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.496057350657239e-05,
      "loss": 0.051,
      "step": 44000
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4958474039142468e-05,
      "loss": 0.0509,
      "step": 44500
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.4956320346634877e-05,
      "loss": 0.0507,
      "step": 45000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.4954112452602043e-05,
      "loss": 0.0505,
      "step": 45500
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.4951850381189151e-05,
      "loss": 0.0503,
      "step": 46000
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.4949534157133844e-05,
      "loss": 0.0502,
      "step": 46500
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.494716380576598e-05,
      "loss": 0.05,
      "step": 47000
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.4944739353007343e-05,
      "loss": 0.0498,
      "step": 47500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.4942260825371358e-05,
      "loss": 0.0496,
      "step": 48000
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.4939728249962806e-05,
      "loss": 0.0495,
      "step": 48500
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.4937141654477529e-05,
      "loss": 0.0493,
      "step": 49000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.4934501067202117e-05,
      "loss": 0.0491,
      "step": 49500
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.4931806517013612e-05,
      "loss": 0.0489,
      "step": 50000
    }
  ],
  "max_steps": 500000,
  "num_train_epochs": 8,
  "total_flos": 7.9871784693625e+20,
  "trial_name": null,
  "trial_params": null
}