{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.844374470146542,
  "global_step": 80000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 4.96875e-05,
      "loss": 0.5806,
      "step": 500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.937500000000001e-05,
      "loss": 0.3929,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.90625e-05,
      "loss": 0.3707,
      "step": 1500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.875e-05,
      "loss": 0.3568,
      "step": 2000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.8437500000000005e-05,
      "loss": 0.3479,
      "step": 2500
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.8125000000000004e-05,
      "loss": 0.3437,
      "step": 3000
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.7812500000000003e-05,
      "loss": 0.3398,
      "step": 3500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.75e-05,
      "loss": 0.3362,
      "step": 4000
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.71875e-05,
      "loss": 0.3326,
      "step": 4500
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.6875e-05,
      "loss": 0.3302,
      "step": 5000
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.65625e-05,
      "loss": 0.3311,
      "step": 5500
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.6250000000000006e-05,
      "loss": 0.3276,
      "step": 6000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.59375e-05,
      "loss": 0.3241,
      "step": 6500
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.5625e-05,
      "loss": 0.3214,
      "step": 7000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.5312500000000004e-05,
      "loss": 0.3217,
      "step": 7500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.5e-05,
      "loss": 0.3211,
      "step": 8000
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.46875e-05,
      "loss": 0.3204,
      "step": 8500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.4375e-05,
      "loss": 0.3177,
      "step": 9000
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.40625e-05,
      "loss": 0.3166,
      "step": 9500
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.375e-05,
      "loss": 0.3158,
      "step": 10000
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.3437500000000006e-05,
      "loss": 0.3147,
      "step": 10500
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.3125000000000005e-05,
      "loss": 0.3143,
      "step": 11000
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.28125e-05,
      "loss": 0.313,
      "step": 11500
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.25e-05,
      "loss": 0.3136,
      "step": 12000
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.21875e-05,
      "loss": 0.3118,
      "step": 12500
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.1875e-05,
      "loss": 0.3129,
      "step": 13000
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.156250000000001e-05,
      "loss": 0.3113,
      "step": 13500
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.125e-05,
      "loss": 0.3095,
      "step": 14000
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.09375e-05,
      "loss": 0.3108,
      "step": 14500
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 0.3096,
      "step": 15000
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.0312500000000004e-05,
      "loss": 0.3107,
      "step": 15500
    },
    {
      "epoch": 0.97,
      "learning_rate": 4e-05,
      "loss": 0.309,
      "step": 16000
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.96875e-05,
      "loss": 0.3075,
      "step": 16500
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.9375e-05,
      "loss": 0.3069,
      "step": 17000
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.90625e-05,
      "loss": 0.3067,
      "step": 17500
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.875e-05,
      "loss": 0.3066,
      "step": 18000
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.8437500000000006e-05,
      "loss": 0.3069,
      "step": 18500
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.8125e-05,
      "loss": 0.3067,
      "step": 19000
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.78125e-05,
      "loss": 0.3044,
      "step": 19500
    },
    {
      "epoch": 1.21,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.3071,
      "step": 20000
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.71875e-05,
      "loss": 0.305,
      "step": 20500
    },
    {
      "epoch": 1.27,
      "learning_rate": 3.6875e-05,
      "loss": 0.3035,
      "step": 21000
    },
    {
      "epoch": 1.3,
      "learning_rate": 3.65625e-05,
      "loss": 0.3032,
      "step": 21500
    },
    {
      "epoch": 1.33,
      "learning_rate": 3.625e-05,
      "loss": 0.3042,
      "step": 22000
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.59375e-05,
      "loss": 0.3045,
      "step": 22500
    },
    {
      "epoch": 1.39,
      "learning_rate": 3.5625000000000005e-05,
      "loss": 0.3041,
      "step": 23000
    },
    {
      "epoch": 1.42,
      "learning_rate": 3.5312500000000005e-05,
      "loss": 0.302,
      "step": 23500
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.5e-05,
      "loss": 0.303,
      "step": 24000
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.46875e-05,
      "loss": 0.3019,
      "step": 24500
    },
    {
      "epoch": 1.51,
      "learning_rate": 3.4375e-05,
      "loss": 0.301,
      "step": 25000
    },
    {
      "epoch": 1.54,
      "learning_rate": 3.40625e-05,
      "loss": 0.3024,
      "step": 25500
    },
    {
      "epoch": 1.57,
      "learning_rate": 3.375000000000001e-05,
      "loss": 0.3031,
      "step": 26000
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.34375e-05,
      "loss": 0.3012,
      "step": 26500
    },
    {
      "epoch": 1.63,
      "learning_rate": 3.3125e-05,
      "loss": 0.3009,
      "step": 27000
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.2812500000000005e-05,
      "loss": 0.3006,
      "step": 27500
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.301,
      "step": 28000
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.21875e-05,
      "loss": 0.3001,
      "step": 28500
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.1875e-05,
      "loss": 0.3006,
      "step": 29000
    },
    {
      "epoch": 1.79,
      "learning_rate": 3.15625e-05,
      "loss": 0.2997,
      "step": 29500
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.125e-05,
      "loss": 0.2996,
      "step": 30000
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.09375e-05,
      "loss": 0.2994,
      "step": 30500
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.0625000000000006e-05,
      "loss": 0.2994,
      "step": 31000
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.0312499999999998e-05,
      "loss": 0.2981,
      "step": 31500
    },
    {
      "epoch": 1.94,
      "learning_rate": 3e-05,
      "loss": 0.2994,
      "step": 32000
    },
    {
      "epoch": 1.97,
      "learning_rate": 2.96875e-05,
      "loss": 0.2998,
      "step": 32500
    },
    {
      "epoch": 2.0,
      "learning_rate": 2.9375000000000003e-05,
      "loss": 0.299,
      "step": 33000
    },
    {
      "epoch": 2.03,
      "learning_rate": 2.9062500000000005e-05,
      "loss": 0.2984,
      "step": 33500
    },
    {
      "epoch": 2.06,
      "learning_rate": 2.8749999999999997e-05,
      "loss": 0.298,
      "step": 34000
    },
    {
      "epoch": 2.09,
      "learning_rate": 2.84375e-05,
      "loss": 0.2984,
      "step": 34500
    },
    {
      "epoch": 2.12,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 0.2961,
      "step": 35000
    },
    {
      "epoch": 2.15,
      "learning_rate": 2.7812500000000002e-05,
      "loss": 0.2983,
      "step": 35500
    },
    {
      "epoch": 2.18,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2963,
      "step": 36000
    },
    {
      "epoch": 2.21,
      "learning_rate": 2.71875e-05,
      "loss": 0.2974,
      "step": 36500
    },
    {
      "epoch": 2.24,
      "learning_rate": 2.6875e-05,
      "loss": 0.2964,
      "step": 37000
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.6562500000000002e-05,
      "loss": 0.2959,
      "step": 37500
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.625e-05,
      "loss": 0.2965,
      "step": 38000
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.5937500000000004e-05,
      "loss": 0.2956,
      "step": 38500
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.5625e-05,
      "loss": 0.2967,
      "step": 39000
    },
    {
      "epoch": 2.39,
      "learning_rate": 2.53125e-05,
      "loss": 0.2966,
      "step": 39500
    },
    {
      "epoch": 2.42,
      "learning_rate": 2.5e-05,
      "loss": 0.2949,
      "step": 40000
    },
    {
      "epoch": 2.45,
      "learning_rate": 2.4687500000000004e-05,
      "loss": 0.2954,
      "step": 40500
    },
    {
      "epoch": 2.48,
      "learning_rate": 2.4375e-05,
      "loss": 0.296,
      "step": 41000
    },
    {
      "epoch": 2.51,
      "learning_rate": 2.4062500000000002e-05,
      "loss": 0.2947,
      "step": 41500
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.375e-05,
      "loss": 0.2944,
      "step": 42000
    },
    {
      "epoch": 2.57,
      "learning_rate": 2.34375e-05,
      "loss": 0.296,
      "step": 42500
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.3125000000000003e-05,
      "loss": 0.2948,
      "step": 43000
    },
    {
      "epoch": 2.63,
      "learning_rate": 2.28125e-05,
      "loss": 0.2947,
      "step": 43500
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.25e-05,
      "loss": 0.2959,
      "step": 44000
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.21875e-05,
      "loss": 0.2953,
      "step": 44500
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.1875e-05,
      "loss": 0.2941,
      "step": 45000
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.1562500000000002e-05,
      "loss": 0.2945,
      "step": 45500
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.125e-05,
      "loss": 0.295,
      "step": 46000
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.09375e-05,
      "loss": 0.296,
      "step": 46500
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.0625e-05,
      "loss": 0.2949,
      "step": 47000
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.0312500000000002e-05,
      "loss": 0.2954,
      "step": 47500
    },
    {
      "epoch": 2.91,
      "learning_rate": 2e-05,
      "loss": 0.293,
      "step": 48000
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.96875e-05,
      "loss": 0.2937,
      "step": 48500
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.9375e-05,
      "loss": 0.2938,
      "step": 49000
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.90625e-05,
      "loss": 0.2939,
      "step": 49500
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.2952,
      "step": 50000
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.84375e-05,
      "loss": 0.2925,
      "step": 50500
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.8125e-05,
      "loss": 0.2932,
      "step": 51000
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.7812500000000003e-05,
      "loss": 0.2926,
      "step": 51500
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.75e-05,
      "loss": 0.2941,
      "step": 52000
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.71875e-05,
      "loss": 0.2929,
      "step": 52500
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.6875000000000004e-05,
      "loss": 0.2925,
      "step": 53000
    },
    {
      "epoch": 3.24,
      "learning_rate": 1.65625e-05,
      "loss": 0.293,
      "step": 53500
    },
    {
      "epoch": 3.27,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.2932,
      "step": 54000
    },
    {
      "epoch": 3.3,
      "learning_rate": 1.59375e-05,
      "loss": 0.2932,
      "step": 54500
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.5625e-05,
      "loss": 0.2939,
      "step": 55000
    },
    {
      "epoch": 3.36,
      "learning_rate": 1.5312500000000003e-05,
      "loss": 0.2906,
      "step": 55500
    },
    {
      "epoch": 3.39,
      "learning_rate": 1.5e-05,
      "loss": 0.2918,
      "step": 56000
    },
    {
      "epoch": 3.42,
      "learning_rate": 1.4687500000000001e-05,
      "loss": 0.2927,
      "step": 56500
    },
    {
      "epoch": 3.45,
      "learning_rate": 1.4374999999999999e-05,
      "loss": 0.2906,
      "step": 57000
    },
    {
      "epoch": 3.48,
      "learning_rate": 1.4062500000000001e-05,
      "loss": 0.2922,
      "step": 57500
    },
    {
      "epoch": 3.51,
      "learning_rate": 1.3750000000000002e-05,
      "loss": 0.2917,
      "step": 58000
    },
    {
      "epoch": 3.54,
      "learning_rate": 1.34375e-05,
      "loss": 0.2926,
      "step": 58500
    },
    {
      "epoch": 3.57,
      "learning_rate": 1.3125e-05,
      "loss": 0.2933,
      "step": 59000
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.28125e-05,
      "loss": 0.2915,
      "step": 59500
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.25e-05,
      "loss": 0.292,
      "step": 60000
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.21875e-05,
      "loss": 0.292,
      "step": 60500
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.1875e-05,
      "loss": 0.2916,
      "step": 61000
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.1562500000000002e-05,
      "loss": 0.2925,
      "step": 61500
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.125e-05,
      "loss": 0.2911,
      "step": 62000
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.09375e-05,
      "loss": 0.2925,
      "step": 62500
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.0625e-05,
      "loss": 0.2922,
      "step": 63000
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.03125e-05,
      "loss": 0.2906,
      "step": 63500
    },
    {
      "epoch": 3.88,
      "learning_rate": 1e-05,
      "loss": 0.2927,
      "step": 64000
    },
    {
      "epoch": 3.91,
      "learning_rate": 9.6875e-06,
      "loss": 0.2909,
      "step": 64500
    },
    {
      "epoch": 3.94,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.2909,
      "step": 65000
    },
    {
      "epoch": 3.97,
      "learning_rate": 9.0625e-06,
      "loss": 0.2916,
      "step": 65500
    },
    {
      "epoch": 4.0,
      "learning_rate": 8.75e-06,
      "loss": 0.2906,
      "step": 66000
    },
    {
      "epoch": 4.03,
      "learning_rate": 8.437500000000002e-06,
      "loss": 0.2899,
      "step": 66500
    },
    {
      "epoch": 4.06,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.2929,
      "step": 67000
    },
    {
      "epoch": 4.09,
      "learning_rate": 7.8125e-06,
      "loss": 0.2908,
      "step": 67500
    },
    {
      "epoch": 4.12,
      "learning_rate": 7.5e-06,
      "loss": 0.29,
      "step": 68000
    },
    {
      "epoch": 4.15,
      "learning_rate": 7.187499999999999e-06,
      "loss": 0.2898,
      "step": 68500
    },
    {
      "epoch": 4.18,
      "learning_rate": 6.875000000000001e-06,
      "loss": 0.2893,
      "step": 69000
    },
    {
      "epoch": 4.21,
      "learning_rate": 6.5625e-06,
      "loss": 0.2909,
      "step": 69500
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.25e-06,
      "loss": 0.2918,
      "step": 70000
    },
    {
      "epoch": 4.27,
      "learning_rate": 5.9375e-06,
      "loss": 0.2901,
      "step": 70500
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.625e-06,
      "loss": 0.2901,
      "step": 71000
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.3125e-06,
      "loss": 0.2897,
      "step": 71500
    },
    {
      "epoch": 4.36,
      "learning_rate": 5e-06,
      "loss": 0.2897,
      "step": 72000
    },
    {
      "epoch": 4.39,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.2904,
      "step": 72500
    },
    {
      "epoch": 4.42,
      "learning_rate": 4.375e-06,
      "loss": 0.2901,
      "step": 73000
    },
    {
      "epoch": 4.45,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 0.2896,
      "step": 73500
    },
    {
      "epoch": 4.48,
      "learning_rate": 3.75e-06,
      "loss": 0.2893,
      "step": 74000
    },
    {
      "epoch": 4.51,
      "learning_rate": 3.4375000000000005e-06,
      "loss": 0.2911,
      "step": 74500
    },
    {
      "epoch": 4.54,
      "learning_rate": 3.125e-06,
      "loss": 0.2876,
      "step": 75000
    },
    {
      "epoch": 4.57,
      "learning_rate": 2.8125e-06,
      "loss": 0.2913,
      "step": 75500
    },
    {
      "epoch": 4.6,
      "learning_rate": 2.5e-06,
      "loss": 0.2897,
      "step": 76000
    },
    {
      "epoch": 4.63,
      "learning_rate": 2.1875e-06,
      "loss": 0.2899,
      "step": 76500
    },
    {
      "epoch": 4.66,
      "learning_rate": 1.875e-06,
      "loss": 0.2892,
      "step": 77000
    },
    {
      "epoch": 4.69,
      "learning_rate": 1.5625e-06,
      "loss": 0.2901,
      "step": 77500
    },
    {
      "epoch": 4.72,
      "learning_rate": 1.25e-06,
      "loss": 0.2891,
      "step": 78000
    },
    {
      "epoch": 4.75,
      "learning_rate": 9.375e-07,
      "loss": 0.2904,
      "step": 78500
    },
    {
      "epoch": 4.78,
      "learning_rate": 6.25e-07,
      "loss": 0.2905,
      "step": 79000
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.125e-07,
      "loss": 0.2895,
      "step": 79500
    },
    {
      "epoch": 4.84,
      "learning_rate": 0.0,
      "loss": 0.2909,
      "step": 80000
    }
  ],
  "max_steps": 80000,
  "num_train_epochs": 5,
  "total_flos": 6.973930336308691e+18,
  "trial_name": null,
  "trial_params": null
}