{
  "best_metric": 58.90729282066525,
  "best_model_checkpoint": "./checkpoint-3500",
  "epoch": 15.337423312883436,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 4.6e-07,
      "loss": 1.0735,
      "step": 25
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.6e-07,
      "loss": 0.9014,
      "step": 50
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.953535353535352e-07,
      "loss": 0.651,
      "step": 75
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.903030303030303e-07,
      "loss": 0.4902,
      "step": 100
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.852525252525253e-07,
      "loss": 0.4484,
      "step": 125
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.802020202020202e-07,
      "loss": 0.4164,
      "step": 150
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.751515151515151e-07,
      "loss": 0.4303,
      "step": 175
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.7010101010101e-07,
      "loss": 0.4073,
      "step": 200
    },
    {
      "epoch": 0.69,
      "learning_rate": 9.65050505050505e-07,
      "loss": 0.3977,
      "step": 225
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.6e-07,
      "loss": 0.3906,
      "step": 250
    },
    {
      "epoch": 0.84,
      "learning_rate": 9.549494949494948e-07,
      "loss": 0.3687,
      "step": 275
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.498989898989899e-07,
      "loss": 0.3817,
      "step": 300
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.448484848484848e-07,
      "loss": 0.3733,
      "step": 325
    },
    {
      "epoch": 1.07,
      "learning_rate": 9.397979797979797e-07,
      "loss": 0.3608,
      "step": 350
    },
    {
      "epoch": 1.15,
      "learning_rate": 9.347474747474747e-07,
      "loss": 0.3519,
      "step": 375
    },
    {
      "epoch": 1.23,
      "learning_rate": 9.296969696969696e-07,
      "loss": 0.3286,
      "step": 400
    },
    {
      "epoch": 1.3,
      "learning_rate": 9.246464646464645e-07,
      "loss": 0.3468,
      "step": 425
    },
    {
      "epoch": 1.38,
      "learning_rate": 9.195959595959596e-07,
      "loss": 0.3386,
      "step": 450
    },
    {
      "epoch": 1.46,
      "learning_rate": 9.145454545454545e-07,
      "loss": 0.3356,
      "step": 475
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.094949494949494e-07,
      "loss": 0.3404,
      "step": 500
    },
    {
      "epoch": 1.53,
      "eval_loss": 0.46061962842941284,
      "eval_runtime": 2718.8588,
      "eval_samples_per_second": 3.849,
      "eval_steps_per_second": 0.121,
      "eval_wer": 66.62163257585992,
      "step": 500
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.044444444444445e-07,
      "loss": 0.3266,
      "step": 525
    },
    {
      "epoch": 1.69,
      "learning_rate": 8.993939393939394e-07,
      "loss": 0.3369,
      "step": 550
    },
    {
      "epoch": 1.76,
      "learning_rate": 8.943434343434342e-07,
      "loss": 0.3154,
      "step": 575
    },
    {
      "epoch": 1.84,
      "learning_rate": 8.892929292929292e-07,
      "loss": 0.3285,
      "step": 600
    },
    {
      "epoch": 1.92,
      "learning_rate": 8.842424242424242e-07,
      "loss": 0.3154,
      "step": 625
    },
    {
      "epoch": 1.99,
      "learning_rate": 8.791919191919191e-07,
      "loss": 0.3243,
      "step": 650
    },
    {
      "epoch": 2.07,
      "learning_rate": 8.741414141414141e-07,
      "loss": 0.3045,
      "step": 675
    },
    {
      "epoch": 2.15,
      "learning_rate": 8.690909090909091e-07,
      "loss": 0.2976,
      "step": 700
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.64040404040404e-07,
      "loss": 0.2947,
      "step": 725
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.58989898989899e-07,
      "loss": 0.2951,
      "step": 750
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.539393939393939e-07,
      "loss": 0.2865,
      "step": 775
    },
    {
      "epoch": 2.45,
      "learning_rate": 8.488888888888888e-07,
      "loss": 0.3027,
      "step": 800
    },
    {
      "epoch": 2.53,
      "learning_rate": 8.438383838383838e-07,
      "loss": 0.2963,
      "step": 825
    },
    {
      "epoch": 2.61,
      "learning_rate": 8.387878787878787e-07,
      "loss": 0.2903,
      "step": 850
    },
    {
      "epoch": 2.68,
      "learning_rate": 8.337373737373737e-07,
      "loss": 0.2887,
      "step": 875
    },
    {
      "epoch": 2.76,
      "learning_rate": 8.286868686868687e-07,
      "loss": 0.2903,
      "step": 900
    },
    {
      "epoch": 2.84,
      "learning_rate": 8.236363636363636e-07,
      "loss": 0.2836,
      "step": 925
    },
    {
      "epoch": 2.91,
      "learning_rate": 8.185858585858586e-07,
      "loss": 0.2832,
      "step": 950
    },
    {
      "epoch": 2.99,
      "learning_rate": 8.135353535353536e-07,
      "loss": 0.2984,
      "step": 975
    },
    {
      "epoch": 3.07,
      "learning_rate": 8.084848484848484e-07,
      "loss": 0.2707,
      "step": 1000
    },
    {
      "epoch": 3.07,
      "eval_loss": 0.42951369285583496,
      "eval_runtime": 2739.5863,
      "eval_samples_per_second": 3.82,
      "eval_steps_per_second": 0.12,
      "eval_wer": 66.84995541625011,
      "step": 1000
    },
    {
      "epoch": 3.14,
      "learning_rate": 8.034343434343433e-07,
      "loss": 0.2778,
      "step": 1025
    },
    {
      "epoch": 3.22,
      "learning_rate": 7.983838383838384e-07,
      "loss": 0.2742,
      "step": 1050
    },
    {
      "epoch": 3.3,
      "learning_rate": 7.933333333333333e-07,
      "loss": 0.2626,
      "step": 1075
    },
    {
      "epoch": 3.37,
      "learning_rate": 7.882828282828282e-07,
      "loss": 0.2658,
      "step": 1100
    },
    {
      "epoch": 3.45,
      "learning_rate": 7.832323232323233e-07,
      "loss": 0.2654,
      "step": 1125
    },
    {
      "epoch": 3.53,
      "learning_rate": 7.781818181818182e-07,
      "loss": 0.2634,
      "step": 1150
    },
    {
      "epoch": 3.6,
      "learning_rate": 7.731313131313131e-07,
      "loss": 0.2609,
      "step": 1175
    },
    {
      "epoch": 3.68,
      "learning_rate": 7.68080808080808e-07,
      "loss": 0.262,
      "step": 1200
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.63030303030303e-07,
      "loss": 0.2701,
      "step": 1225
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.579797979797979e-07,
      "loss": 0.2646,
      "step": 1250
    },
    {
      "epoch": 3.91,
      "learning_rate": 7.529292929292928e-07,
      "loss": 0.2692,
      "step": 1275
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.478787878787879e-07,
      "loss": 0.2588,
      "step": 1300
    },
    {
      "epoch": 4.06,
      "learning_rate": 7.428282828282828e-07,
      "loss": 0.2549,
      "step": 1325
    },
    {
      "epoch": 4.14,
      "learning_rate": 7.377777777777777e-07,
      "loss": 0.2376,
      "step": 1350
    },
    {
      "epoch": 4.22,
      "learning_rate": 7.327272727272728e-07,
      "loss": 0.2542,
      "step": 1375
    },
    {
      "epoch": 4.29,
      "learning_rate": 7.276767676767677e-07,
      "loss": 0.2634,
      "step": 1400
    },
    {
      "epoch": 4.37,
      "learning_rate": 7.226262626262625e-07,
      "loss": 0.2453,
      "step": 1425
    },
    {
      "epoch": 4.45,
      "learning_rate": 7.175757575757575e-07,
      "loss": 0.2393,
      "step": 1450
    },
    {
      "epoch": 4.52,
      "learning_rate": 7.125252525252525e-07,
      "loss": 0.2466,
      "step": 1475
    },
    {
      "epoch": 4.6,
      "learning_rate": 7.074747474747474e-07,
      "loss": 0.2427,
      "step": 1500
    },
    {
      "epoch": 4.6,
      "eval_loss": 0.4123845398426056,
      "eval_runtime": 2708.1735,
      "eval_samples_per_second": 3.865,
      "eval_steps_per_second": 0.121,
      "eval_wer": 61.166202815531356,
      "step": 1500
    },
    {
      "epoch": 4.68,
      "learning_rate": 7.024242424242424e-07,
      "loss": 0.247,
      "step": 1525
    },
    {
      "epoch": 4.75,
      "learning_rate": 6.973737373737374e-07,
      "loss": 0.2531,
      "step": 1550
    },
    {
      "epoch": 4.83,
      "learning_rate": 6.923232323232323e-07,
      "loss": 0.2496,
      "step": 1575
    },
    {
      "epoch": 4.91,
      "learning_rate": 6.872727272727273e-07,
      "loss": 0.2366,
      "step": 1600
    },
    {
      "epoch": 4.98,
      "learning_rate": 6.822222222222221e-07,
      "loss": 0.2287,
      "step": 1625
    },
    {
      "epoch": 5.06,
      "learning_rate": 6.771717171717171e-07,
      "loss": 0.2352,
      "step": 1650
    },
    {
      "epoch": 5.14,
      "learning_rate": 6.721212121212121e-07,
      "loss": 0.2242,
      "step": 1675
    },
    {
      "epoch": 5.21,
      "learning_rate": 6.67070707070707e-07,
      "loss": 0.2288,
      "step": 1700
    },
    {
      "epoch": 5.29,
      "learning_rate": 6.62020202020202e-07,
      "loss": 0.2374,
      "step": 1725
    },
    {
      "epoch": 5.37,
      "learning_rate": 6.56969696969697e-07,
      "loss": 0.2312,
      "step": 1750
    },
    {
      "epoch": 5.44,
      "learning_rate": 6.519191919191919e-07,
      "loss": 0.2279,
      "step": 1775
    },
    {
      "epoch": 5.52,
      "learning_rate": 6.468686868686868e-07,
      "loss": 0.2263,
      "step": 1800
    },
    {
      "epoch": 5.6,
      "learning_rate": 6.418181818181818e-07,
      "loss": 0.2305,
      "step": 1825
    },
    {
      "epoch": 5.67,
      "learning_rate": 6.367676767676767e-07,
      "loss": 0.2251,
      "step": 1850
    },
    {
      "epoch": 5.75,
      "learning_rate": 6.317171717171716e-07,
      "loss": 0.2285,
      "step": 1875
    },
    {
      "epoch": 5.83,
      "learning_rate": 6.266666666666667e-07,
      "loss": 0.2308,
      "step": 1900
    },
    {
      "epoch": 5.9,
      "learning_rate": 6.216161616161616e-07,
      "loss": 0.2343,
      "step": 1925
    },
    {
      "epoch": 5.98,
      "learning_rate": 6.165656565656565e-07,
      "loss": 0.2335,
      "step": 1950
    },
    {
      "epoch": 6.06,
      "learning_rate": 6.115151515151516e-07,
      "loss": 0.2246,
      "step": 1975
    },
    {
      "epoch": 6.13,
      "learning_rate": 6.064646464646465e-07,
      "loss": 0.2131,
      "step": 2000
    },
    {
      "epoch": 6.13,
      "eval_loss": 0.4056357443332672,
      "eval_runtime": 2709.0159,
      "eval_samples_per_second": 3.863,
      "eval_steps_per_second": 0.121,
      "eval_wer": 62.303763949309634,
      "step": 2000
    },
    {
      "epoch": 6.21,
      "learning_rate": 6.014141414141413e-07,
      "loss": 0.2113,
      "step": 2025
    },
    {
      "epoch": 6.29,
      "learning_rate": 5.963636363636363e-07,
      "loss": 0.2207,
      "step": 2050
    },
    {
      "epoch": 6.37,
      "learning_rate": 5.913131313131313e-07,
      "loss": 0.2171,
      "step": 2075
    },
    {
      "epoch": 6.44,
      "learning_rate": 5.862626262626262e-07,
      "loss": 0.2112,
      "step": 2100
    },
    {
      "epoch": 6.52,
      "learning_rate": 5.812121212121212e-07,
      "loss": 0.2098,
      "step": 2125
    },
    {
      "epoch": 6.6,
      "learning_rate": 5.761616161616162e-07,
      "loss": 0.2094,
      "step": 2150
    },
    {
      "epoch": 6.67,
      "learning_rate": 5.711111111111111e-07,
      "loss": 0.2149,
      "step": 2175
    },
    {
      "epoch": 6.75,
      "learning_rate": 5.660606060606061e-07,
      "loss": 0.2162,
      "step": 2200
    },
    {
      "epoch": 6.83,
      "learning_rate": 5.610101010101009e-07,
      "loss": 0.2216,
      "step": 2225
    },
    {
      "epoch": 6.9,
      "learning_rate": 5.559595959595959e-07,
      "loss": 0.2169,
      "step": 2250
    },
    {
      "epoch": 6.98,
      "learning_rate": 5.509090909090909e-07,
      "loss": 0.2209,
      "step": 2275
    },
    {
      "epoch": 7.06,
      "learning_rate": 5.458585858585858e-07,
      "loss": 0.2108,
      "step": 2300
    },
    {
      "epoch": 7.13,
      "learning_rate": 5.408080808080808e-07,
      "loss": 0.198,
      "step": 2325
    },
    {
      "epoch": 7.21,
      "learning_rate": 5.357575757575758e-07,
      "loss": 0.2075,
      "step": 2350
    },
    {
      "epoch": 7.29,
      "learning_rate": 5.307070707070707e-07,
      "loss": 0.2067,
      "step": 2375
    },
    {
      "epoch": 7.36,
      "learning_rate": 5.256565656565657e-07,
      "loss": 0.2193,
      "step": 2400
    },
    {
      "epoch": 7.44,
      "learning_rate": 5.206060606060607e-07,
      "loss": 0.1996,
      "step": 2425
    },
    {
      "epoch": 7.52,
      "learning_rate": 5.155555555555555e-07,
      "loss": 0.2118,
      "step": 2450
    },
    {
      "epoch": 7.59,
      "learning_rate": 5.105050505050504e-07,
      "loss": 0.1938,
      "step": 2475
    },
    {
      "epoch": 7.67,
      "learning_rate": 5.054545454545454e-07,
      "loss": 0.2085,
      "step": 2500
    },
    {
      "epoch": 7.67,
      "eval_loss": 0.401197612285614,
      "eval_runtime": 2721.3063,
      "eval_samples_per_second": 3.846,
      "eval_steps_per_second": 0.121,
      "eval_wer": 62.27539247210137,
      "step": 2500
    },
    {
      "epoch": 7.75,
      "learning_rate": 5.004040404040404e-07,
      "loss": 0.1983,
      "step": 2525
    },
    {
      "epoch": 7.82,
      "learning_rate": 4.953535353535353e-07,
      "loss": 0.2098,
      "step": 2550
    },
    {
      "epoch": 7.9,
      "learning_rate": 4.903030303030302e-07,
      "loss": 0.2036,
      "step": 2575
    },
    {
      "epoch": 7.98,
      "learning_rate": 4.852525252525253e-07,
      "loss": 0.2024,
      "step": 2600
    },
    {
      "epoch": 8.05,
      "learning_rate": 4.802020202020202e-07,
      "loss": 0.2074,
      "step": 2625
    },
    {
      "epoch": 8.13,
      "learning_rate": 4.7515151515151514e-07,
      "loss": 0.1989,
      "step": 2650
    },
    {
      "epoch": 8.21,
      "learning_rate": 4.701010101010101e-07,
      "loss": 0.1904,
      "step": 2675
    },
    {
      "epoch": 8.28,
      "learning_rate": 4.65050505050505e-07,
      "loss": 0.1921,
      "step": 2700
    },
    {
      "epoch": 8.36,
      "learning_rate": 4.6e-07,
      "loss": 0.1993,
      "step": 2725
    },
    {
      "epoch": 8.44,
      "learning_rate": 4.549494949494949e-07,
      "loss": 0.1966,
      "step": 2750
    },
    {
      "epoch": 8.51,
      "learning_rate": 4.498989898989899e-07,
      "loss": 0.1912,
      "step": 2775
    },
    {
      "epoch": 8.59,
      "learning_rate": 4.448484848484848e-07,
      "loss": 0.2001,
      "step": 2800
    },
    {
      "epoch": 8.67,
      "learning_rate": 4.3979797979797977e-07,
      "loss": 0.2025,
      "step": 2825
    },
    {
      "epoch": 8.74,
      "learning_rate": 4.3474747474747475e-07,
      "loss": 0.192,
      "step": 2850
    },
    {
      "epoch": 8.82,
      "learning_rate": 4.296969696969697e-07,
      "loss": 0.1979,
      "step": 2875
    },
    {
      "epoch": 8.9,
      "learning_rate": 4.246464646464646e-07,
      "loss": 0.1928,
      "step": 2900
    },
    {
      "epoch": 8.97,
      "learning_rate": 4.1959595959595955e-07,
      "loss": 0.1845,
      "step": 2925
    },
    {
      "epoch": 9.05,
      "learning_rate": 4.1454545454545453e-07,
      "loss": 0.1846,
      "step": 2950
    },
    {
      "epoch": 9.13,
      "learning_rate": 4.0949494949494946e-07,
      "loss": 0.1926,
      "step": 2975
    },
    {
      "epoch": 9.2,
      "learning_rate": 4.044444444444444e-07,
      "loss": 0.1904,
      "step": 3000
    },
    {
      "epoch": 9.2,
      "eval_loss": 0.3976161777973175,
      "eval_runtime": 2684.5933,
      "eval_samples_per_second": 3.899,
      "eval_steps_per_second": 0.122,
      "eval_wer": 59.734118727877004,
      "step": 3000
    },
    {
      "epoch": 9.28,
      "learning_rate": 3.993939393939394e-07,
      "loss": 0.1972,
      "step": 3025
    },
    {
      "epoch": 9.36,
      "learning_rate": 3.943434343434343e-07,
      "loss": 0.1881,
      "step": 3050
    },
    {
      "epoch": 9.43,
      "learning_rate": 3.892929292929293e-07,
      "loss": 0.1935,
      "step": 3075
    },
    {
      "epoch": 9.51,
      "learning_rate": 3.8424242424242423e-07,
      "loss": 0.1911,
      "step": 3100
    },
    {
      "epoch": 9.59,
      "learning_rate": 3.7919191919191916e-07,
      "loss": 0.1852,
      "step": 3125
    },
    {
      "epoch": 9.66,
      "learning_rate": 3.7414141414141414e-07,
      "loss": 0.1939,
      "step": 3150
    },
    {
      "epoch": 9.74,
      "learning_rate": 3.690909090909091e-07,
      "loss": 0.1834,
      "step": 3175
    },
    {
      "epoch": 9.82,
      "learning_rate": 3.6404040404040406e-07,
      "loss": 0.1794,
      "step": 3200
    },
    {
      "epoch": 9.89,
      "learning_rate": 3.5898989898989894e-07,
      "loss": 0.1878,
      "step": 3225
    },
    {
      "epoch": 9.97,
      "learning_rate": 3.539393939393939e-07,
      "loss": 0.1884,
      "step": 3250
    },
    {
      "epoch": 10.05,
      "learning_rate": 3.488888888888889e-07,
      "loss": 0.1844,
      "step": 3275
    },
    {
      "epoch": 10.12,
      "learning_rate": 3.4383838383838384e-07,
      "loss": 0.1744,
      "step": 3300
    },
    {
      "epoch": 10.2,
      "learning_rate": 3.3878787878787877e-07,
      "loss": 0.1848,
      "step": 3325
    },
    {
      "epoch": 10.28,
      "learning_rate": 3.337373737373737e-07,
      "loss": 0.1792,
      "step": 3350
    },
    {
      "epoch": 10.35,
      "learning_rate": 3.286868686868687e-07,
      "loss": 0.1757,
      "step": 3375
    },
    {
      "epoch": 10.43,
      "learning_rate": 3.236363636363636e-07,
      "loss": 0.1815,
      "step": 3400
    },
    {
      "epoch": 10.51,
      "learning_rate": 3.1858585858585855e-07,
      "loss": 0.1758,
      "step": 3425
    },
    {
      "epoch": 10.58,
      "learning_rate": 3.1353535353535353e-07,
      "loss": 0.1895,
      "step": 3450
    },
    {
      "epoch": 10.66,
      "learning_rate": 3.0848484848484847e-07,
      "loss": 0.177,
      "step": 3475
    },
    {
      "epoch": 10.74,
      "learning_rate": 3.0343434343434345e-07,
      "loss": 0.1836,
      "step": 3500
    },
    {
      "epoch": 10.74,
      "eval_loss": 0.40053287148475647,
      "eval_runtime": 2681.3744,
      "eval_samples_per_second": 3.903,
      "eval_steps_per_second": 0.122,
      "eval_wer": 58.90729282066525,
      "step": 3500
    },
    {
      "epoch": 10.81,
      "learning_rate": 2.9838383838383833e-07,
      "loss": 0.1861,
      "step": 3525
    },
    {
      "epoch": 10.89,
      "learning_rate": 2.933333333333333e-07,
      "loss": 0.1817,
      "step": 3550
    },
    {
      "epoch": 10.97,
      "learning_rate": 2.882828282828283e-07,
      "loss": 0.1854,
      "step": 3575
    },
    {
      "epoch": 11.04,
      "learning_rate": 2.8323232323232323e-07,
      "loss": 0.1796,
      "step": 3600
    },
    {
      "epoch": 11.12,
      "learning_rate": 2.7818181818181816e-07,
      "loss": 0.1719,
      "step": 3625
    },
    {
      "epoch": 11.2,
      "learning_rate": 2.731313131313131e-07,
      "loss": 0.1718,
      "step": 3650
    },
    {
      "epoch": 11.27,
      "learning_rate": 2.680808080808081e-07,
      "loss": 0.1774,
      "step": 3675
    },
    {
      "epoch": 11.35,
      "learning_rate": 2.63030303030303e-07,
      "loss": 0.1795,
      "step": 3700
    },
    {
      "epoch": 11.43,
      "learning_rate": 2.5797979797979794e-07,
      "loss": 0.1784,
      "step": 3725
    },
    {
      "epoch": 11.5,
      "learning_rate": 2.529292929292929e-07,
      "loss": 0.1689,
      "step": 3750
    },
    {
      "epoch": 11.58,
      "learning_rate": 2.4787878787878786e-07,
      "loss": 0.1792,
      "step": 3775
    },
    {
      "epoch": 11.66,
      "learning_rate": 2.4282828282828284e-07,
      "loss": 0.1749,
      "step": 3800
    },
    {
      "epoch": 11.73,
      "learning_rate": 2.3777777777777777e-07,
      "loss": 0.1787,
      "step": 3825
    },
    {
      "epoch": 11.81,
      "learning_rate": 2.3272727272727273e-07,
      "loss": 0.1816,
      "step": 3850
    },
    {
      "epoch": 11.89,
      "learning_rate": 2.2767676767676766e-07,
      "loss": 0.171,
      "step": 3875
    },
    {
      "epoch": 11.96,
      "learning_rate": 2.2262626262626262e-07,
      "loss": 0.1794,
      "step": 3900
    },
    {
      "epoch": 12.04,
      "learning_rate": 2.1757575757575758e-07,
      "loss": 0.1651,
      "step": 3925
    },
    {
      "epoch": 12.12,
      "learning_rate": 2.1252525252525254e-07,
      "loss": 0.1689,
      "step": 3950
    },
    {
      "epoch": 12.19,
      "learning_rate": 2.0747474747474747e-07,
      "loss": 0.1747,
      "step": 3975
    },
    {
      "epoch": 12.27,
      "learning_rate": 2.0242424242424242e-07,
      "loss": 0.1653,
      "step": 4000
    },
    {
      "epoch": 12.27,
      "eval_loss": 0.3988991677761078,
      "eval_runtime": 2674.7246,
      "eval_samples_per_second": 3.913,
      "eval_steps_per_second": 0.123,
      "eval_wer": 59.77735145505147,
      "step": 4000
    },
    {
      "epoch": 12.35,
      "learning_rate": 1.9737373737373736e-07,
      "loss": 0.1748,
      "step": 4025
    },
    {
      "epoch": 12.42,
      "learning_rate": 1.9232323232323231e-07,
      "loss": 0.1722,
      "step": 4050
    },
    {
      "epoch": 12.5,
      "learning_rate": 1.8727272727272727e-07,
      "loss": 0.1678,
      "step": 4075
    },
    {
      "epoch": 12.58,
      "learning_rate": 1.8222222222222223e-07,
      "loss": 0.1784,
      "step": 4100
    },
    {
      "epoch": 12.65,
      "learning_rate": 1.7717171717171716e-07,
      "loss": 0.1795,
      "step": 4125
    },
    {
      "epoch": 12.73,
      "learning_rate": 1.7212121212121212e-07,
      "loss": 0.1752,
      "step": 4150
    },
    {
      "epoch": 12.81,
      "learning_rate": 1.6707070707070705e-07,
      "loss": 0.1725,
      "step": 4175
    },
    {
      "epoch": 12.88,
      "learning_rate": 1.62020202020202e-07,
      "loss": 0.1792,
      "step": 4200
    },
    {
      "epoch": 12.96,
      "learning_rate": 1.5696969696969697e-07,
      "loss": 0.169,
      "step": 4225
    },
    {
      "epoch": 13.04,
      "learning_rate": 1.5191919191919193e-07,
      "loss": 0.1643,
      "step": 4250
    },
    {
      "epoch": 13.11,
      "learning_rate": 1.4686868686868686e-07,
      "loss": 0.1653,
      "step": 4275
    },
    {
      "epoch": 13.19,
      "learning_rate": 1.4181818181818182e-07,
      "loss": 0.1743,
      "step": 4300
    },
    {
      "epoch": 13.27,
      "learning_rate": 1.3676767676767675e-07,
      "loss": 0.1677,
      "step": 4325
    },
    {
      "epoch": 13.34,
      "learning_rate": 1.317171717171717e-07,
      "loss": 0.1724,
      "step": 4350
    },
    {
      "epoch": 13.42,
      "learning_rate": 1.2666666666666666e-07,
      "loss": 0.17,
      "step": 4375
    },
    {
      "epoch": 13.5,
      "learning_rate": 1.2161616161616162e-07,
      "loss": 0.1665,
      "step": 4400
    },
    {
      "epoch": 13.57,
      "learning_rate": 1.1656565656565657e-07,
      "loss": 0.1733,
      "step": 4425
    },
    {
      "epoch": 13.65,
      "learning_rate": 1.1151515151515151e-07,
      "loss": 0.1655,
      "step": 4450
    },
    {
      "epoch": 13.73,
      "learning_rate": 1.0646464646464647e-07,
      "loss": 0.1642,
      "step": 4475
    },
    {
      "epoch": 13.8,
      "learning_rate": 1.0141414141414141e-07,
      "loss": 0.1693,
      "step": 4500
    },
    {
      "epoch": 13.8,
      "eval_loss": 0.3983347713947296,
      "eval_runtime": 2686.1318,
      "eval_samples_per_second": 3.896,
      "eval_steps_per_second": 0.122,
      "eval_wer": 59.94622929557675,
      "step": 4500
    },
    {
      "epoch": 13.88,
      "learning_rate": 9.636363636363636e-08,
      "loss": 0.1635,
      "step": 4525
    },
    {
      "epoch": 13.96,
      "learning_rate": 9.131313131313132e-08,
      "loss": 0.1769,
      "step": 4550
    },
    {
      "epoch": 14.03,
      "learning_rate": 8.626262626262626e-08,
      "loss": 0.1747,
      "step": 4575
    },
    {
      "epoch": 14.11,
      "learning_rate": 8.12121212121212e-08,
      "loss": 0.1684,
      "step": 4600
    },
    {
      "epoch": 14.19,
      "learning_rate": 7.616161616161616e-08,
      "loss": 0.1664,
      "step": 4625
    },
    {
      "epoch": 14.26,
      "learning_rate": 7.111111111111111e-08,
      "loss": 0.1736,
      "step": 4650
    },
    {
      "epoch": 14.34,
      "learning_rate": 6.606060606060605e-08,
      "loss": 0.1609,
      "step": 4675
    },
    {
      "epoch": 14.42,
      "learning_rate": 6.101010101010101e-08,
      "loss": 0.1662,
      "step": 4700
    },
    {
      "epoch": 14.49,
      "learning_rate": 5.5959595959595956e-08,
      "loss": 0.1746,
      "step": 4725
    },
    {
      "epoch": 14.57,
      "learning_rate": 5.090909090909091e-08,
      "loss": 0.1677,
      "step": 4750
    },
    {
      "epoch": 14.65,
      "learning_rate": 4.585858585858585e-08,
      "loss": 0.1674,
      "step": 4775
    },
    {
      "epoch": 14.72,
      "learning_rate": 4.0808080808080803e-08,
      "loss": 0.168,
      "step": 4800
    },
    {
      "epoch": 14.8,
      "learning_rate": 3.5757575757575755e-08,
      "loss": 0.1691,
      "step": 4825
    },
    {
      "epoch": 14.88,
      "learning_rate": 3.0707070707070706e-08,
      "loss": 0.1641,
      "step": 4850
    },
    {
      "epoch": 14.95,
      "learning_rate": 2.5656565656565654e-08,
      "loss": 0.1678,
      "step": 4875
    },
    {
      "epoch": 15.03,
      "learning_rate": 2.0606060606060606e-08,
      "loss": 0.1672,
      "step": 4900
    },
    {
      "epoch": 15.11,
      "learning_rate": 1.5555555555555554e-08,
      "loss": 0.1656,
      "step": 4925
    },
    {
      "epoch": 15.18,
      "learning_rate": 1.0505050505050505e-08,
      "loss": 0.1637,
      "step": 4950
    },
    {
      "epoch": 15.26,
      "learning_rate": 5.454545454545455e-09,
      "loss": 0.1614,
      "step": 4975
    },
    {
      "epoch": 15.34,
      "learning_rate": 4.0404040404040403e-10,
      "loss": 0.1616,
      "step": 5000
    },
    {
      "epoch": 15.34,
      "eval_loss": 0.3983960449695587,
      "eval_runtime": 2696.9238,
      "eval_samples_per_second": 3.881,
      "eval_steps_per_second": 0.122,
      "eval_wer": 59.830041341295356,
      "step": 5000
    },
    {
      "epoch": 15.34,
      "step": 5000,
      "total_flos": 9.22520949202944e+19,
      "train_loss": 0.23562390689849855,
      "train_runtime": 67707.284,
      "train_samples_per_second": 4.726,
      "train_steps_per_second": 0.074
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 16,
  "save_steps": 500,
  "total_flos": 9.22520949202944e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}