xls-r-es-test-lm / trainer_state.json
tomascufaro's picture
copy repo from tomascufaro
36107b3
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 10.0,
"global_step": 71700,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 3.7125e-06,
"loss": 13.0763,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 7.35e-06,
"loss": 9.212,
"step": 200
},
{
"epoch": 0.04,
"learning_rate": 1.1099999999999999e-05,
"loss": 3.7882,
"step": 300
},
{
"epoch": 0.06,
"learning_rate": 1.485e-05,
"loss": 3.1078,
"step": 400
},
{
"epoch": 0.07,
"learning_rate": 1.8599999999999998e-05,
"loss": 2.9613,
"step": 500
},
{
"epoch": 0.07,
"eval_loss": 2.9646778106689453,
"eval_runtime": 822.6775,
"eval_samples_per_second": 18.768,
"eval_steps_per_second": 2.346,
"eval_wer": 1.0,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 2.2349999999999998e-05,
"loss": 2.934,
"step": 600
},
{
"epoch": 0.1,
"learning_rate": 2.6099999999999997e-05,
"loss": 2.9101,
"step": 700
},
{
"epoch": 0.11,
"learning_rate": 2.985e-05,
"loss": 2.8886,
"step": 800
},
{
"epoch": 0.13,
"learning_rate": 3.36e-05,
"loss": 2.8442,
"step": 900
},
{
"epoch": 0.14,
"learning_rate": 3.735e-05,
"loss": 2.604,
"step": 1000
},
{
"epoch": 0.14,
"eval_loss": 1.8299518823623657,
"eval_runtime": 822.8035,
"eval_samples_per_second": 18.765,
"eval_steps_per_second": 2.346,
"eval_wer": 0.9561869374839984,
"step": 1000
},
{
"epoch": 0.15,
"learning_rate": 4.11e-05,
"loss": 1.7999,
"step": 1100
},
{
"epoch": 0.17,
"learning_rate": 4.484999999999999e-05,
"loss": 1.4168,
"step": 1200
},
{
"epoch": 0.18,
"learning_rate": 4.8599999999999995e-05,
"loss": 1.2936,
"step": 1300
},
{
"epoch": 0.2,
"learning_rate": 5.234999999999999e-05,
"loss": 1.2409,
"step": 1400
},
{
"epoch": 0.21,
"learning_rate": 5.6099999999999995e-05,
"loss": 1.177,
"step": 1500
},
{
"epoch": 0.21,
"eval_loss": 0.36516621708869934,
"eval_runtime": 841.4928,
"eval_samples_per_second": 18.348,
"eval_steps_per_second": 2.294,
"eval_wer": 0.3077155367658161,
"step": 1500
},
{
"epoch": 0.22,
"learning_rate": 5.985e-05,
"loss": 1.1222,
"step": 1600
},
{
"epoch": 0.24,
"learning_rate": 6.359999999999999e-05,
"loss": 1.1053,
"step": 1700
},
{
"epoch": 0.25,
"learning_rate": 6.735e-05,
"loss": 1.0908,
"step": 1800
},
{
"epoch": 0.26,
"learning_rate": 7.11e-05,
"loss": 1.0798,
"step": 1900
},
{
"epoch": 0.28,
"learning_rate": 7.484999999999999e-05,
"loss": 1.0745,
"step": 2000
},
{
"epoch": 0.28,
"eval_loss": 0.27073442935943604,
"eval_runtime": 839.6609,
"eval_samples_per_second": 18.388,
"eval_steps_per_second": 2.299,
"eval_wer": 0.25043163153437625,
"step": 2000
},
{
"epoch": 0.29,
"learning_rate": 7.489670014347202e-05,
"loss": 1.0408,
"step": 2100
},
{
"epoch": 0.31,
"learning_rate": 7.478909612625537e-05,
"loss": 1.0591,
"step": 2200
},
{
"epoch": 0.32,
"learning_rate": 7.468149210903873e-05,
"loss": 1.0234,
"step": 2300
},
{
"epoch": 0.33,
"learning_rate": 7.457388809182208e-05,
"loss": 1.0357,
"step": 2400
},
{
"epoch": 0.35,
"learning_rate": 7.446628407460545e-05,
"loss": 1.0103,
"step": 2500
},
{
"epoch": 0.35,
"eval_loss": 0.2338443398475647,
"eval_runtime": 817.1264,
"eval_samples_per_second": 18.895,
"eval_steps_per_second": 2.362,
"eval_wer": 0.21574355506830611,
"step": 2500
},
{
"epoch": 0.36,
"learning_rate": 7.43586800573888e-05,
"loss": 1.009,
"step": 2600
},
{
"epoch": 0.38,
"learning_rate": 7.425107604017216e-05,
"loss": 1.0083,
"step": 2700
},
{
"epoch": 0.39,
"learning_rate": 7.414347202295551e-05,
"loss": 1.0148,
"step": 2800
},
{
"epoch": 0.4,
"learning_rate": 7.403586800573888e-05,
"loss": 1.0232,
"step": 2900
},
{
"epoch": 0.42,
"learning_rate": 7.392826398852223e-05,
"loss": 0.9858,
"step": 3000
},
{
"epoch": 0.42,
"eval_loss": 0.23211880028247833,
"eval_runtime": 838.7422,
"eval_samples_per_second": 18.409,
"eval_steps_per_second": 2.301,
"eval_wer": 0.21287476449001833,
"step": 3000
},
{
"epoch": 0.43,
"learning_rate": 7.382065997130559e-05,
"loss": 0.9887,
"step": 3100
},
{
"epoch": 0.45,
"learning_rate": 7.371305595408894e-05,
"loss": 0.9783,
"step": 3200
},
{
"epoch": 0.46,
"learning_rate": 7.360545193687231e-05,
"loss": 0.9763,
"step": 3300
},
{
"epoch": 0.47,
"learning_rate": 7.349784791965566e-05,
"loss": 0.9819,
"step": 3400
},
{
"epoch": 0.49,
"learning_rate": 7.339024390243902e-05,
"loss": 0.974,
"step": 3500
},
{
"epoch": 0.49,
"eval_loss": 0.2164456844329834,
"eval_runtime": 833.2825,
"eval_samples_per_second": 18.529,
"eval_steps_per_second": 2.316,
"eval_wer": 0.20313925779070302,
"step": 3500
},
{
"epoch": 0.5,
"learning_rate": 7.328263988522237e-05,
"loss": 0.9872,
"step": 3600
},
{
"epoch": 0.52,
"learning_rate": 7.317503586800574e-05,
"loss": 0.9782,
"step": 3700
},
{
"epoch": 0.53,
"learning_rate": 7.30674318507891e-05,
"loss": 0.9706,
"step": 3800
},
{
"epoch": 0.54,
"learning_rate": 7.295982783357245e-05,
"loss": 0.9726,
"step": 3900
},
{
"epoch": 0.56,
"learning_rate": 7.28522238163558e-05,
"loss": 0.9699,
"step": 4000
},
{
"epoch": 0.56,
"eval_loss": 0.2077634185552597,
"eval_runtime": 836.414,
"eval_samples_per_second": 18.46,
"eval_steps_per_second": 2.307,
"eval_wer": 0.1969880981297061,
"step": 4000
},
{
"epoch": 0.57,
"learning_rate": 7.274461979913917e-05,
"loss": 0.972,
"step": 4100
},
{
"epoch": 0.59,
"learning_rate": 7.263701578192252e-05,
"loss": 0.9727,
"step": 4200
},
{
"epoch": 0.6,
"learning_rate": 7.252941176470588e-05,
"loss": 0.9597,
"step": 4300
},
{
"epoch": 0.61,
"learning_rate": 7.24228837876614e-05,
"loss": 0.9539,
"step": 4400
},
{
"epoch": 0.63,
"learning_rate": 7.231527977044475e-05,
"loss": 0.9513,
"step": 4500
},
{
"epoch": 0.63,
"eval_loss": 0.2172592580318451,
"eval_runtime": 817.4817,
"eval_samples_per_second": 18.887,
"eval_steps_per_second": 2.361,
"eval_wer": 0.21385291047666563,
"step": 4500
},
{
"epoch": 0.64,
"learning_rate": 7.220767575322811e-05,
"loss": 0.965,
"step": 4600
},
{
"epoch": 0.66,
"learning_rate": 7.210007173601146e-05,
"loss": 0.9609,
"step": 4700
},
{
"epoch": 0.67,
"learning_rate": 7.199246771879483e-05,
"loss": 0.9571,
"step": 4800
},
{
"epoch": 0.68,
"learning_rate": 7.188486370157819e-05,
"loss": 0.964,
"step": 4900
},
{
"epoch": 0.7,
"learning_rate": 7.177725968436154e-05,
"loss": 0.9657,
"step": 5000
},
{
"epoch": 0.7,
"eval_loss": 0.20500008761882782,
"eval_runtime": 815.4377,
"eval_samples_per_second": 18.935,
"eval_steps_per_second": 2.367,
"eval_wer": 0.19793998516369174,
"step": 5000
},
{
"epoch": 0.71,
"learning_rate": 7.16696556671449e-05,
"loss": 0.9494,
"step": 5100
},
{
"epoch": 0.73,
"learning_rate": 7.156205164992825e-05,
"loss": 0.9494,
"step": 5200
},
{
"epoch": 0.74,
"learning_rate": 7.145444763271162e-05,
"loss": 0.9647,
"step": 5300
},
{
"epoch": 0.75,
"learning_rate": 7.134684361549497e-05,
"loss": 0.9569,
"step": 5400
},
{
"epoch": 0.77,
"learning_rate": 7.123923959827832e-05,
"loss": 0.9484,
"step": 5500
},
{
"epoch": 0.77,
"eval_loss": 0.20082205533981323,
"eval_runtime": 821.6118,
"eval_samples_per_second": 18.792,
"eval_steps_per_second": 2.349,
"eval_wer": 0.1918610376225144,
"step": 5500
},
{
"epoch": 0.78,
"learning_rate": 7.113163558106168e-05,
"loss": 0.9571,
"step": 5600
},
{
"epoch": 0.79,
"learning_rate": 7.102510760401721e-05,
"loss": 0.9474,
"step": 5700
},
{
"epoch": 0.81,
"learning_rate": 7.091750358680057e-05,
"loss": 0.9477,
"step": 5800
},
{
"epoch": 0.82,
"learning_rate": 7.080989956958392e-05,
"loss": 0.9316,
"step": 5900
},
{
"epoch": 0.84,
"learning_rate": 7.070229555236729e-05,
"loss": 0.9317,
"step": 6000
},
{
"epoch": 0.84,
"eval_loss": 0.20117491483688354,
"eval_runtime": 813.1073,
"eval_samples_per_second": 18.989,
"eval_steps_per_second": 2.374,
"eval_wer": 0.19111922220982216,
"step": 6000
},
{
"epoch": 0.85,
"learning_rate": 7.059469153515064e-05,
"loss": 0.9495,
"step": 6100
},
{
"epoch": 0.86,
"learning_rate": 7.0487087517934e-05,
"loss": 0.9396,
"step": 6200
},
{
"epoch": 0.88,
"learning_rate": 7.037948350071735e-05,
"loss": 0.949,
"step": 6300
},
{
"epoch": 0.89,
"learning_rate": 7.027187948350072e-05,
"loss": 0.9571,
"step": 6400
},
{
"epoch": 0.91,
"learning_rate": 7.016427546628408e-05,
"loss": 0.9366,
"step": 6500
},
{
"epoch": 0.91,
"eval_loss": 0.2023884803056717,
"eval_runtime": 827.9444,
"eval_samples_per_second": 18.649,
"eval_steps_per_second": 2.331,
"eval_wer": 0.19762487773175166,
"step": 6500
},
{
"epoch": 0.92,
"learning_rate": 7.005667144906743e-05,
"loss": 0.9284,
"step": 6600
},
{
"epoch": 0.93,
"learning_rate": 6.994906743185078e-05,
"loss": 0.9334,
"step": 6700
},
{
"epoch": 0.95,
"learning_rate": 6.984146341463414e-05,
"loss": 0.934,
"step": 6800
},
{
"epoch": 0.96,
"learning_rate": 6.97338593974175e-05,
"loss": 0.9275,
"step": 6900
},
{
"epoch": 0.98,
"learning_rate": 6.962625538020086e-05,
"loss": 0.9242,
"step": 7000
},
{
"epoch": 0.98,
"eval_loss": 0.20615938305854797,
"eval_runtime": 833.0567,
"eval_samples_per_second": 18.534,
"eval_steps_per_second": 2.317,
"eval_wer": 0.20280445614426668,
"step": 7000
},
{
"epoch": 0.99,
"learning_rate": 6.951865136298421e-05,
"loss": 0.947,
"step": 7100
},
{
"epoch": 1.0,
"learning_rate": 6.941104734576757e-05,
"loss": 0.9186,
"step": 7200
},
{
"epoch": 1.02,
"learning_rate": 6.930344332855092e-05,
"loss": 0.9107,
"step": 7300
},
{
"epoch": 1.03,
"learning_rate": 6.919583931133428e-05,
"loss": 0.934,
"step": 7400
},
{
"epoch": 1.05,
"learning_rate": 6.908823529411765e-05,
"loss": 0.9138,
"step": 7500
},
{
"epoch": 1.05,
"eval_loss": 0.19241651892662048,
"eval_runtime": 827.2513,
"eval_samples_per_second": 18.664,
"eval_steps_per_second": 2.333,
"eval_wer": 0.18634009282539765,
"step": 7500
},
{
"epoch": 1.06,
"learning_rate": 6.8980631276901e-05,
"loss": 0.916,
"step": 7600
},
{
"epoch": 1.07,
"learning_rate": 6.887302725968435e-05,
"loss": 0.9286,
"step": 7700
},
{
"epoch": 1.09,
"learning_rate": 6.876542324246771e-05,
"loss": 0.9283,
"step": 7800
},
{
"epoch": 1.1,
"learning_rate": 6.865781922525106e-05,
"loss": 0.892,
"step": 7900
},
{
"epoch": 1.12,
"learning_rate": 6.855021520803443e-05,
"loss": 0.921,
"step": 8000
},
{
"epoch": 1.12,
"eval_loss": 0.1934651881456375,
"eval_runtime": 817.2756,
"eval_samples_per_second": 18.892,
"eval_steps_per_second": 2.362,
"eval_wer": 0.18360916174858366,
"step": 8000
},
{
"epoch": 1.13,
"learning_rate": 6.844261119081778e-05,
"loss": 0.9301,
"step": 8100
},
{
"epoch": 1.14,
"learning_rate": 6.833500717360114e-05,
"loss": 0.9306,
"step": 8200
},
{
"epoch": 1.16,
"learning_rate": 6.822740315638449e-05,
"loss": 0.9132,
"step": 8300
},
{
"epoch": 1.17,
"learning_rate": 6.811979913916785e-05,
"loss": 0.918,
"step": 8400
},
{
"epoch": 1.19,
"learning_rate": 6.801219512195121e-05,
"loss": 0.9117,
"step": 8500
},
{
"epoch": 1.19,
"eval_loss": 0.18866156041622162,
"eval_runtime": 834.8273,
"eval_samples_per_second": 18.495,
"eval_steps_per_second": 2.312,
"eval_wer": 0.18145592763032647,
"step": 8500
},
{
"epoch": 1.2,
"learning_rate": 6.790459110473457e-05,
"loss": 0.9252,
"step": 8600
},
{
"epoch": 1.21,
"learning_rate": 6.779698708751792e-05,
"loss": 0.9189,
"step": 8700
},
{
"epoch": 1.23,
"learning_rate": 6.768938307030128e-05,
"loss": 0.9154,
"step": 8800
},
{
"epoch": 1.24,
"learning_rate": 6.758177905308465e-05,
"loss": 0.9132,
"step": 8900
},
{
"epoch": 1.26,
"learning_rate": 6.7474175035868e-05,
"loss": 0.9064,
"step": 9000
},
{
"epoch": 1.26,
"eval_loss": 0.19089102745056152,
"eval_runtime": 834.6909,
"eval_samples_per_second": 18.498,
"eval_steps_per_second": 2.312,
"eval_wer": 0.18389144548969663,
"step": 9000
},
{
"epoch": 1.27,
"learning_rate": 6.736657101865135e-05,
"loss": 0.9238,
"step": 9100
},
{
"epoch": 1.28,
"learning_rate": 6.725896700143471e-05,
"loss": 0.9098,
"step": 9200
},
{
"epoch": 1.3,
"learning_rate": 6.715136298421808e-05,
"loss": 0.9031,
"step": 9300
},
{
"epoch": 1.31,
"learning_rate": 6.704375896700143e-05,
"loss": 0.9075,
"step": 9400
},
{
"epoch": 1.32,
"learning_rate": 6.693615494978478e-05,
"loss": 0.9118,
"step": 9500
},
{
"epoch": 1.32,
"eval_loss": 0.1868850290775299,
"eval_runtime": 823.4598,
"eval_samples_per_second": 18.75,
"eval_steps_per_second": 2.344,
"eval_wer": 0.18301833531369602,
"step": 9500
},
{
"epoch": 1.34,
"learning_rate": 6.682855093256814e-05,
"loss": 0.9069,
"step": 9600
},
{
"epoch": 1.35,
"learning_rate": 6.672094691535151e-05,
"loss": 0.9261,
"step": 9700
},
{
"epoch": 1.37,
"learning_rate": 6.661441893830703e-05,
"loss": 0.9241,
"step": 9800
},
{
"epoch": 1.38,
"learning_rate": 6.650681492109038e-05,
"loss": 0.9093,
"step": 9900
},
{
"epoch": 1.39,
"learning_rate": 6.639921090387374e-05,
"loss": 0.9121,
"step": 10000
},
{
"epoch": 1.39,
"eval_loss": 0.18625491857528687,
"eval_runtime": 823.3284,
"eval_samples_per_second": 18.753,
"eval_steps_per_second": 2.344,
"eval_wer": 0.18022175685522782,
"step": 10000
},
{
"epoch": 1.41,
"learning_rate": 6.629268292682926e-05,
"loss": 0.9026,
"step": 10100
},
{
"epoch": 1.42,
"learning_rate": 6.618507890961261e-05,
"loss": 0.9278,
"step": 10200
},
{
"epoch": 1.44,
"learning_rate": 6.607747489239597e-05,
"loss": 0.9078,
"step": 10300
},
{
"epoch": 1.45,
"learning_rate": 6.596987087517934e-05,
"loss": 0.9142,
"step": 10400
},
{
"epoch": 1.46,
"learning_rate": 6.586226685796269e-05,
"loss": 0.9048,
"step": 10500
},
{
"epoch": 1.46,
"eval_loss": 0.18454793095588684,
"eval_runtime": 828.7597,
"eval_samples_per_second": 18.63,
"eval_steps_per_second": 2.329,
"eval_wer": 0.17907949241444504,
"step": 10500
},
{
"epoch": 1.48,
"learning_rate": 6.575466284074604e-05,
"loss": 0.8931,
"step": 10600
},
{
"epoch": 1.49,
"learning_rate": 6.56470588235294e-05,
"loss": 0.9064,
"step": 10700
},
{
"epoch": 1.51,
"learning_rate": 6.553945480631277e-05,
"loss": 0.8957,
"step": 10800
},
{
"epoch": 1.52,
"learning_rate": 6.543185078909612e-05,
"loss": 0.9053,
"step": 10900
},
{
"epoch": 1.53,
"learning_rate": 6.532424677187947e-05,
"loss": 0.8955,
"step": 11000
},
{
"epoch": 1.53,
"eval_loss": 0.18630339205265045,
"eval_runtime": 811.839,
"eval_samples_per_second": 19.019,
"eval_steps_per_second": 2.377,
"eval_wer": 0.17739235470593256,
"step": 11000
},
{
"epoch": 1.55,
"learning_rate": 6.521664275466283e-05,
"loss": 0.9038,
"step": 11100
},
{
"epoch": 1.56,
"learning_rate": 6.51090387374462e-05,
"loss": 0.9045,
"step": 11200
},
{
"epoch": 1.58,
"learning_rate": 6.500143472022955e-05,
"loss": 0.8885,
"step": 11300
},
{
"epoch": 1.59,
"learning_rate": 6.48938307030129e-05,
"loss": 0.9193,
"step": 11400
},
{
"epoch": 1.6,
"learning_rate": 6.478622668579626e-05,
"loss": 0.8947,
"step": 11500
},
{
"epoch": 1.6,
"eval_loss": 0.19065053761005402,
"eval_runtime": 805.1782,
"eval_samples_per_second": 19.176,
"eval_steps_per_second": 2.397,
"eval_wer": 0.1813574565578452,
"step": 11500
},
{
"epoch": 1.62,
"learning_rate": 6.467862266857963e-05,
"loss": 0.9008,
"step": 11600
},
{
"epoch": 1.63,
"learning_rate": 6.457101865136298e-05,
"loss": 0.913,
"step": 11700
},
{
"epoch": 1.65,
"learning_rate": 6.446341463414634e-05,
"loss": 0.9007,
"step": 11800
},
{
"epoch": 1.66,
"learning_rate": 6.435581061692969e-05,
"loss": 0.9043,
"step": 11900
},
{
"epoch": 1.67,
"learning_rate": 6.424928263988523e-05,
"loss": 0.9073,
"step": 12000
},
{
"epoch": 1.67,
"eval_loss": 0.1891753375530243,
"eval_runtime": 812.1877,
"eval_samples_per_second": 19.01,
"eval_steps_per_second": 2.376,
"eval_wer": 0.1853488173624195,
"step": 12000
},
{
"epoch": 1.69,
"learning_rate": 6.414167862266858e-05,
"loss": 0.888,
"step": 12100
},
{
"epoch": 1.7,
"learning_rate": 6.403407460545193e-05,
"loss": 0.8932,
"step": 12200
},
{
"epoch": 1.72,
"learning_rate": 6.392647058823529e-05,
"loss": 0.8922,
"step": 12300
},
{
"epoch": 1.73,
"learning_rate": 6.381886657101864e-05,
"loss": 0.8855,
"step": 12400
},
{
"epoch": 1.74,
"learning_rate": 6.371126255380201e-05,
"loss": 0.8927,
"step": 12500
},
{
"epoch": 1.74,
"eval_loss": 0.18209430575370789,
"eval_runtime": 809.9382,
"eval_samples_per_second": 19.063,
"eval_steps_per_second": 2.383,
"eval_wer": 0.17497653106105862,
"step": 12500
},
{
"epoch": 1.76,
"learning_rate": 6.360365853658536e-05,
"loss": 0.9054,
"step": 12600
},
{
"epoch": 1.77,
"learning_rate": 6.349605451936872e-05,
"loss": 0.8961,
"step": 12700
},
{
"epoch": 1.79,
"learning_rate": 6.338845050215207e-05,
"loss": 0.897,
"step": 12800
},
{
"epoch": 1.8,
"learning_rate": 6.328084648493543e-05,
"loss": 0.9045,
"step": 12900
},
{
"epoch": 1.81,
"learning_rate": 6.317324246771878e-05,
"loss": 0.8732,
"step": 13000
},
{
"epoch": 1.81,
"eval_loss": 0.1814534217119217,
"eval_runtime": 824.3836,
"eval_samples_per_second": 18.729,
"eval_steps_per_second": 2.341,
"eval_wer": 0.17678839879471409,
"step": 13000
},
{
"epoch": 1.83,
"learning_rate": 6.306563845050215e-05,
"loss": 0.8986,
"step": 13100
},
{
"epoch": 1.84,
"learning_rate": 6.29580344332855e-05,
"loss": 0.8983,
"step": 13200
},
{
"epoch": 1.85,
"learning_rate": 6.285043041606886e-05,
"loss": 0.8941,
"step": 13300
},
{
"epoch": 1.87,
"learning_rate": 6.274282639885221e-05,
"loss": 0.8875,
"step": 13400
},
{
"epoch": 1.88,
"learning_rate": 6.263522238163557e-05,
"loss": 0.8761,
"step": 13500
},
{
"epoch": 1.88,
"eval_loss": 0.18223263323307037,
"eval_runtime": 827.5314,
"eval_samples_per_second": 18.658,
"eval_steps_per_second": 2.332,
"eval_wer": 0.1748518010359157,
"step": 13500
},
{
"epoch": 1.9,
"learning_rate": 6.252761836441893e-05,
"loss": 0.8878,
"step": 13600
},
{
"epoch": 1.91,
"learning_rate": 6.242001434720229e-05,
"loss": 0.8831,
"step": 13700
},
{
"epoch": 1.92,
"learning_rate": 6.231241032998564e-05,
"loss": 0.9091,
"step": 13800
},
{
"epoch": 1.94,
"learning_rate": 6.2204806312769e-05,
"loss": 0.8877,
"step": 13900
},
{
"epoch": 1.95,
"learning_rate": 6.209720229555236e-05,
"loss": 0.8751,
"step": 14000
},
{
"epoch": 1.95,
"eval_loss": 0.17893458902835846,
"eval_runtime": 826.4134,
"eval_samples_per_second": 18.683,
"eval_steps_per_second": 2.335,
"eval_wer": 0.1715169140478832,
"step": 14000
},
{
"epoch": 1.97,
"learning_rate": 6.198959827833572e-05,
"loss": 0.8934,
"step": 14100
},
{
"epoch": 1.98,
"learning_rate": 6.188199426111907e-05,
"loss": 0.8865,
"step": 14200
},
{
"epoch": 1.99,
"learning_rate": 6.177439024390243e-05,
"loss": 0.8837,
"step": 14300
},
{
"epoch": 2.01,
"learning_rate": 6.16667862266858e-05,
"loss": 0.8999,
"step": 14400
},
{
"epoch": 2.02,
"learning_rate": 6.155918220946915e-05,
"loss": 0.8889,
"step": 14500
},
{
"epoch": 2.02,
"eval_loss": 0.1818888783454895,
"eval_runtime": 806.8131,
"eval_samples_per_second": 19.137,
"eval_steps_per_second": 2.392,
"eval_wer": 0.17912544558160298,
"step": 14500
},
{
"epoch": 2.04,
"learning_rate": 6.14515781922525e-05,
"loss": 0.8681,
"step": 14600
},
{
"epoch": 2.05,
"learning_rate": 6.134397417503586e-05,
"loss": 0.8923,
"step": 14700
},
{
"epoch": 2.06,
"learning_rate": 6.123637015781923e-05,
"loss": 0.8817,
"step": 14800
},
{
"epoch": 2.08,
"learning_rate": 6.112876614060258e-05,
"loss": 0.8716,
"step": 14900
},
{
"epoch": 2.09,
"learning_rate": 6.10222381635581e-05,
"loss": 0.8864,
"step": 15000
},
{
"epoch": 2.09,
"eval_loss": 0.1825788915157318,
"eval_runtime": 813.2986,
"eval_samples_per_second": 18.984,
"eval_steps_per_second": 2.373,
"eval_wer": 0.17936177615555804,
"step": 15000
},
{
"epoch": 2.11,
"learning_rate": 6.091463414634146e-05,
"loss": 0.8873,
"step": 15100
},
{
"epoch": 2.12,
"learning_rate": 6.080703012912482e-05,
"loss": 0.8774,
"step": 15200
},
{
"epoch": 2.13,
"learning_rate": 6.069942611190817e-05,
"loss": 0.8758,
"step": 15300
},
{
"epoch": 2.15,
"learning_rate": 6.0591822094691526e-05,
"loss": 0.8959,
"step": 15400
},
{
"epoch": 2.16,
"learning_rate": 6.048421807747488e-05,
"loss": 0.886,
"step": 15500
},
{
"epoch": 2.16,
"eval_loss": 0.17876982688903809,
"eval_runtime": 809.7859,
"eval_samples_per_second": 19.067,
"eval_steps_per_second": 2.383,
"eval_wer": 0.17760242632722595,
"step": 15500
},
{
"epoch": 2.18,
"learning_rate": 6.037661406025825e-05,
"loss": 0.8871,
"step": 15600
},
{
"epoch": 2.19,
"learning_rate": 6.02690100430416e-05,
"loss": 0.8768,
"step": 15700
},
{
"epoch": 2.2,
"learning_rate": 6.0161406025824957e-05,
"loss": 0.8819,
"step": 15800
},
{
"epoch": 2.22,
"learning_rate": 6.005380200860831e-05,
"loss": 0.8948,
"step": 15900
},
{
"epoch": 2.23,
"learning_rate": 5.994619799139168e-05,
"loss": 0.8915,
"step": 16000
},
{
"epoch": 2.23,
"eval_loss": 0.17562659084796906,
"eval_runtime": 802.4278,
"eval_samples_per_second": 19.242,
"eval_steps_per_second": 2.405,
"eval_wer": 0.1719436220286354,
"step": 16000
},
{
"epoch": 2.25,
"learning_rate": 5.983859397417503e-05,
"loss": 0.8683,
"step": 16100
},
{
"epoch": 2.26,
"learning_rate": 5.973098995695839e-05,
"loss": 0.8887,
"step": 16200
},
{
"epoch": 2.27,
"learning_rate": 5.962338593974174e-05,
"loss": 0.8751,
"step": 16300
},
{
"epoch": 2.29,
"learning_rate": 5.95157819225251e-05,
"loss": 0.8675,
"step": 16400
},
{
"epoch": 2.3,
"learning_rate": 5.940817790530846e-05,
"loss": 0.8689,
"step": 16500
},
{
"epoch": 2.3,
"eval_loss": 0.17687593400478363,
"eval_runtime": 801.4845,
"eval_samples_per_second": 19.264,
"eval_steps_per_second": 2.408,
"eval_wer": 0.17114272397245436,
"step": 16500
},
{
"epoch": 2.32,
"learning_rate": 5.930057388809182e-05,
"loss": 0.8742,
"step": 16600
},
{
"epoch": 2.33,
"learning_rate": 5.919296987087517e-05,
"loss": 0.8684,
"step": 16700
},
{
"epoch": 2.34,
"learning_rate": 5.908536585365853e-05,
"loss": 0.8757,
"step": 16800
},
{
"epoch": 2.36,
"learning_rate": 5.897776183644189e-05,
"loss": 0.8763,
"step": 16900
},
{
"epoch": 2.37,
"learning_rate": 5.887015781922524e-05,
"loss": 0.879,
"step": 17000
},
{
"epoch": 2.37,
"eval_loss": 0.17770837247371674,
"eval_runtime": 806.5154,
"eval_samples_per_second": 19.144,
"eval_steps_per_second": 2.393,
"eval_wer": 0.17385396083477211,
"step": 17000
},
{
"epoch": 2.38,
"learning_rate": 5.87625538020086e-05,
"loss": 0.8734,
"step": 17100
},
{
"epoch": 2.4,
"learning_rate": 5.8654949784791964e-05,
"loss": 0.8804,
"step": 17200
},
{
"epoch": 2.41,
"learning_rate": 5.8548421807747485e-05,
"loss": 0.8634,
"step": 17300
},
{
"epoch": 2.43,
"learning_rate": 5.844081779053084e-05,
"loss": 0.874,
"step": 17400
},
{
"epoch": 2.44,
"learning_rate": 5.833428981348637e-05,
"loss": 0.8692,
"step": 17500
},
{
"epoch": 2.44,
"eval_loss": 0.17646534740924835,
"eval_runtime": 806.9866,
"eval_samples_per_second": 19.133,
"eval_steps_per_second": 2.392,
"eval_wer": 0.17046655594141627,
"step": 17500
},
{
"epoch": 2.45,
"learning_rate": 5.822668579626972e-05,
"loss": 0.8693,
"step": 17600
},
{
"epoch": 2.47,
"learning_rate": 5.811908177905308e-05,
"loss": 0.8628,
"step": 17700
},
{
"epoch": 2.48,
"learning_rate": 5.801147776183643e-05,
"loss": 0.883,
"step": 17800
},
{
"epoch": 2.5,
"learning_rate": 5.79038737446198e-05,
"loss": 0.8781,
"step": 17900
},
{
"epoch": 2.51,
"learning_rate": 5.779626972740315e-05,
"loss": 0.8504,
"step": 18000
},
{
"epoch": 2.51,
"eval_loss": 0.16994699835777283,
"eval_runtime": 804.8476,
"eval_samples_per_second": 19.184,
"eval_steps_per_second": 2.398,
"eval_wer": 0.1651556827655929,
"step": 18000
},
{
"epoch": 2.52,
"learning_rate": 5.768866571018651e-05,
"loss": 0.8572,
"step": 18100
},
{
"epoch": 2.54,
"learning_rate": 5.758106169296986e-05,
"loss": 0.8619,
"step": 18200
},
{
"epoch": 2.55,
"learning_rate": 5.747345767575323e-05,
"loss": 0.8654,
"step": 18300
},
{
"epoch": 2.57,
"learning_rate": 5.7365853658536584e-05,
"loss": 0.8645,
"step": 18400
},
{
"epoch": 2.58,
"learning_rate": 5.725824964131994e-05,
"loss": 0.8728,
"step": 18500
},
{
"epoch": 2.58,
"eval_loss": 0.1705409586429596,
"eval_runtime": 809.8007,
"eval_samples_per_second": 19.066,
"eval_steps_per_second": 2.383,
"eval_wer": 0.16938337414412227,
"step": 18500
},
{
"epoch": 2.59,
"learning_rate": 5.715064562410329e-05,
"loss": 0.8772,
"step": 18600
},
{
"epoch": 2.61,
"learning_rate": 5.704304160688665e-05,
"loss": 0.8617,
"step": 18700
},
{
"epoch": 2.62,
"learning_rate": 5.693543758967001e-05,
"loss": 0.8644,
"step": 18800
},
{
"epoch": 2.64,
"learning_rate": 5.682783357245337e-05,
"loss": 0.8778,
"step": 18900
},
{
"epoch": 2.65,
"learning_rate": 5.672022955523672e-05,
"loss": 0.8523,
"step": 19000
},
{
"epoch": 2.65,
"eval_loss": 0.16738072037696838,
"eval_runtime": 810.091,
"eval_samples_per_second": 19.06,
"eval_steps_per_second": 2.382,
"eval_wer": 0.1645123384253819,
"step": 19000
},
{
"epoch": 2.66,
"learning_rate": 5.6612625538020084e-05,
"loss": 0.8495,
"step": 19100
},
{
"epoch": 2.68,
"learning_rate": 5.650502152080344e-05,
"loss": 0.8569,
"step": 19200
},
{
"epoch": 2.69,
"learning_rate": 5.639741750358679e-05,
"loss": 0.8759,
"step": 19300
},
{
"epoch": 2.71,
"learning_rate": 5.628981348637015e-05,
"loss": 0.8593,
"step": 19400
},
{
"epoch": 2.72,
"learning_rate": 5.6182209469153515e-05,
"loss": 0.8513,
"step": 19500
},
{
"epoch": 2.72,
"eval_loss": 0.1660722941160202,
"eval_runtime": 806.603,
"eval_samples_per_second": 19.142,
"eval_steps_per_second": 2.393,
"eval_wer": 0.16107241562670274,
"step": 19500
},
{
"epoch": 2.73,
"learning_rate": 5.607460545193687e-05,
"loss": 0.8504,
"step": 19600
},
{
"epoch": 2.75,
"learning_rate": 5.596700143472022e-05,
"loss": 0.8468,
"step": 19700
},
{
"epoch": 2.76,
"learning_rate": 5.585939741750358e-05,
"loss": 0.8609,
"step": 19800
},
{
"epoch": 2.78,
"learning_rate": 5.5751793400286945e-05,
"loss": 0.8711,
"step": 19900
},
{
"epoch": 2.79,
"learning_rate": 5.56441893830703e-05,
"loss": 0.8498,
"step": 20000
},
{
"epoch": 2.79,
"eval_loss": 0.16595293581485748,
"eval_runtime": 807.8038,
"eval_samples_per_second": 19.114,
"eval_steps_per_second": 2.389,
"eval_wer": 0.16308779024348613,
"step": 20000
},
{
"epoch": 2.8,
"learning_rate": 5.5536585365853654e-05,
"loss": 0.8533,
"step": 20100
},
{
"epoch": 2.82,
"learning_rate": 5.542898134863701e-05,
"loss": 0.8572,
"step": 20200
},
{
"epoch": 2.83,
"learning_rate": 5.532137733142037e-05,
"loss": 0.8492,
"step": 20300
},
{
"epoch": 2.85,
"learning_rate": 5.521377331420373e-05,
"loss": 0.8502,
"step": 20400
},
{
"epoch": 2.86,
"learning_rate": 5.5106169296987085e-05,
"loss": 0.8432,
"step": 20500
},
{
"epoch": 2.86,
"eval_loss": 0.16357631981372833,
"eval_runtime": 817.1731,
"eval_samples_per_second": 18.894,
"eval_steps_per_second": 2.362,
"eval_wer": 0.1609805092923869,
"step": 20500
},
{
"epoch": 2.87,
"learning_rate": 5.499856527977044e-05,
"loss": 0.8457,
"step": 20600
},
{
"epoch": 2.89,
"learning_rate": 5.48909612625538e-05,
"loss": 0.8668,
"step": 20700
},
{
"epoch": 2.9,
"learning_rate": 5.4783357245337154e-05,
"loss": 0.8588,
"step": 20800
},
{
"epoch": 2.91,
"learning_rate": 5.467575322812051e-05,
"loss": 0.8502,
"step": 20900
},
{
"epoch": 2.93,
"learning_rate": 5.456814921090387e-05,
"loss": 0.8492,
"step": 21000
},
{
"epoch": 2.93,
"eval_loss": 0.1707760989665985,
"eval_runtime": 811.1583,
"eval_samples_per_second": 19.035,
"eval_steps_per_second": 2.379,
"eval_wer": 0.16875315928024212,
"step": 21000
},
{
"epoch": 2.94,
"learning_rate": 5.446054519368723e-05,
"loss": 0.85,
"step": 21100
},
{
"epoch": 2.96,
"learning_rate": 5.4352941176470585e-05,
"loss": 0.8605,
"step": 21200
},
{
"epoch": 2.97,
"learning_rate": 5.424533715925394e-05,
"loss": 0.8649,
"step": 21300
},
{
"epoch": 2.98,
"learning_rate": 5.4137733142037293e-05,
"loss": 0.8518,
"step": 21400
},
{
"epoch": 3.0,
"learning_rate": 5.403012912482066e-05,
"loss": 0.8561,
"step": 21500
},
{
"epoch": 3.0,
"eval_loss": 0.16626086831092834,
"eval_runtime": 808.8282,
"eval_samples_per_second": 19.089,
"eval_steps_per_second": 2.386,
"eval_wer": 0.16042250654832632,
"step": 21500
},
{
"epoch": 3.01,
"learning_rate": 5.3922525107604016e-05,
"loss": 0.8347,
"step": 21600
},
{
"epoch": 3.03,
"learning_rate": 5.381492109038737e-05,
"loss": 0.8588,
"step": 21700
},
{
"epoch": 3.04,
"learning_rate": 5.3707317073170724e-05,
"loss": 0.8658,
"step": 21800
},
{
"epoch": 3.05,
"learning_rate": 5.3599713055954085e-05,
"loss": 0.8473,
"step": 21900
},
{
"epoch": 3.07,
"learning_rate": 5.3492109038737446e-05,
"loss": 0.842,
"step": 22000
},
{
"epoch": 3.07,
"eval_loss": 0.16903015971183777,
"eval_runtime": 823.293,
"eval_samples_per_second": 18.754,
"eval_steps_per_second": 2.344,
"eval_wer": 0.16249039907043308,
"step": 22000
},
{
"epoch": 3.08,
"learning_rate": 5.33845050215208e-05,
"loss": 0.8483,
"step": 22100
},
{
"epoch": 3.1,
"learning_rate": 5.327797704447632e-05,
"loss": 0.8483,
"step": 22200
},
{
"epoch": 3.11,
"learning_rate": 5.317037302725968e-05,
"loss": 0.8439,
"step": 22300
},
{
"epoch": 3.12,
"learning_rate": 5.306276901004304e-05,
"loss": 0.8532,
"step": 22400
},
{
"epoch": 3.14,
"learning_rate": 5.295516499282639e-05,
"loss": 0.857,
"step": 22500
},
{
"epoch": 3.14,
"eval_loss": 0.16415256261825562,
"eval_runtime": 819.3591,
"eval_samples_per_second": 18.844,
"eval_steps_per_second": 2.355,
"eval_wer": 0.1604553302391534,
"step": 22500
},
{
"epoch": 3.15,
"learning_rate": 5.2847560975609746e-05,
"loss": 0.8507,
"step": 22600
},
{
"epoch": 3.17,
"learning_rate": 5.2739956958393114e-05,
"loss": 0.8499,
"step": 22700
},
{
"epoch": 3.18,
"learning_rate": 5.263235294117647e-05,
"loss": 0.8489,
"step": 22800
},
{
"epoch": 3.19,
"learning_rate": 5.252474892395982e-05,
"loss": 0.8525,
"step": 22900
},
{
"epoch": 3.21,
"learning_rate": 5.2417144906743177e-05,
"loss": 0.8518,
"step": 23000
},
{
"epoch": 3.21,
"eval_loss": 0.16263364255428314,
"eval_runtime": 811.1154,
"eval_samples_per_second": 19.036,
"eval_steps_per_second": 2.379,
"eval_wer": 0.15849247352769336,
"step": 23000
},
{
"epoch": 3.22,
"learning_rate": 5.2309540889526544e-05,
"loss": 0.8428,
"step": 23100
},
{
"epoch": 3.24,
"learning_rate": 5.22019368723099e-05,
"loss": 0.8552,
"step": 23200
},
{
"epoch": 3.25,
"learning_rate": 5.209433285509325e-05,
"loss": 0.8458,
"step": 23300
},
{
"epoch": 3.26,
"learning_rate": 5.198672883787661e-05,
"loss": 0.8472,
"step": 23400
},
{
"epoch": 3.28,
"learning_rate": 5.187912482065997e-05,
"loss": 0.8506,
"step": 23500
},
{
"epoch": 3.28,
"eval_loss": 0.16511571407318115,
"eval_runtime": 816.5059,
"eval_samples_per_second": 18.91,
"eval_steps_per_second": 2.364,
"eval_wer": 0.16054067183530385,
"step": 23500
},
{
"epoch": 3.29,
"learning_rate": 5.177152080344332e-05,
"loss": 0.8373,
"step": 23600
},
{
"epoch": 3.31,
"learning_rate": 5.1663916786226684e-05,
"loss": 0.8516,
"step": 23700
},
{
"epoch": 3.32,
"learning_rate": 5.155631276901004e-05,
"loss": 0.8548,
"step": 23800
},
{
"epoch": 3.33,
"learning_rate": 5.14487087517934e-05,
"loss": 0.8385,
"step": 23900
},
{
"epoch": 3.35,
"learning_rate": 5.134110473457675e-05,
"loss": 0.8394,
"step": 24000
},
{
"epoch": 3.35,
"eval_loss": 0.16468414664268494,
"eval_runtime": 816.8132,
"eval_samples_per_second": 18.903,
"eval_steps_per_second": 2.363,
"eval_wer": 0.15853842669485127,
"step": 24000
},
{
"epoch": 3.36,
"learning_rate": 5.123350071736011e-05,
"loss": 0.8477,
"step": 24100
},
{
"epoch": 3.38,
"learning_rate": 5.112589670014346e-05,
"loss": 0.8494,
"step": 24200
},
{
"epoch": 3.39,
"learning_rate": 5.101829268292683e-05,
"loss": 0.8375,
"step": 24300
},
{
"epoch": 3.4,
"learning_rate": 5.0910688665710184e-05,
"loss": 0.8443,
"step": 24400
},
{
"epoch": 3.42,
"learning_rate": 5.080308464849354e-05,
"loss": 0.8431,
"step": 24500
},
{
"epoch": 3.42,
"eval_loss": 0.16316334903240204,
"eval_runtime": 813.384,
"eval_samples_per_second": 18.982,
"eval_steps_per_second": 2.373,
"eval_wer": 0.15731082065791807,
"step": 24500
},
{
"epoch": 3.43,
"learning_rate": 5.069548063127689e-05,
"loss": 0.8393,
"step": 24600
},
{
"epoch": 3.44,
"learning_rate": 5.058895265423242e-05,
"loss": 0.8525,
"step": 24700
},
{
"epoch": 3.46,
"learning_rate": 5.0481348637015775e-05,
"loss": 0.8444,
"step": 24800
},
{
"epoch": 3.47,
"learning_rate": 5.0373744619799136e-05,
"loss": 0.8505,
"step": 24900
},
{
"epoch": 3.49,
"learning_rate": 5.026614060258249e-05,
"loss": 0.8566,
"step": 25000
},
{
"epoch": 3.49,
"eval_loss": 0.16140349209308624,
"eval_runtime": 811.9119,
"eval_samples_per_second": 19.017,
"eval_steps_per_second": 2.377,
"eval_wer": 0.15503942125268333,
"step": 25000
},
{
"epoch": 3.5,
"learning_rate": 5.015853658536585e-05,
"loss": 0.8414,
"step": 25100
},
{
"epoch": 3.51,
"learning_rate": 5.0050932568149206e-05,
"loss": 0.8468,
"step": 25200
},
{
"epoch": 3.53,
"learning_rate": 4.994332855093256e-05,
"loss": 0.8417,
"step": 25300
},
{
"epoch": 3.54,
"learning_rate": 4.983572453371592e-05,
"loss": 0.8457,
"step": 25400
},
{
"epoch": 3.56,
"learning_rate": 4.972812051649928e-05,
"loss": 0.8534,
"step": 25500
},
{
"epoch": 3.56,
"eval_loss": 0.16451936960220337,
"eval_runtime": 819.141,
"eval_samples_per_second": 18.849,
"eval_steps_per_second": 2.356,
"eval_wer": 0.15887322834128761,
"step": 25500
},
{
"epoch": 3.57,
"learning_rate": 4.9620516499282636e-05,
"loss": 0.8566,
"step": 25600
},
{
"epoch": 3.58,
"learning_rate": 4.951291248206599e-05,
"loss": 0.8438,
"step": 25700
},
{
"epoch": 3.6,
"learning_rate": 4.940638450502151e-05,
"loss": 0.8485,
"step": 25800
},
{
"epoch": 3.61,
"learning_rate": 4.929878048780487e-05,
"loss": 0.8311,
"step": 25900
},
{
"epoch": 3.63,
"learning_rate": 4.9191176470588234e-05,
"loss": 0.8386,
"step": 26000
},
{
"epoch": 3.63,
"eval_loss": 0.16321870684623718,
"eval_runtime": 820.2843,
"eval_samples_per_second": 18.823,
"eval_steps_per_second": 2.353,
"eval_wer": 0.15823644873924203,
"step": 26000
},
{
"epoch": 3.64,
"learning_rate": 4.908357245337159e-05,
"loss": 0.8507,
"step": 26100
},
{
"epoch": 3.65,
"learning_rate": 4.897596843615494e-05,
"loss": 0.8255,
"step": 26200
},
{
"epoch": 3.67,
"learning_rate": 4.8868364418938304e-05,
"loss": 0.8322,
"step": 26300
},
{
"epoch": 3.68,
"learning_rate": 4.876076040172166e-05,
"loss": 0.844,
"step": 26400
},
{
"epoch": 3.7,
"learning_rate": 4.865315638450501e-05,
"loss": 0.8357,
"step": 26500
},
{
"epoch": 3.7,
"eval_loss": 0.16306844353675842,
"eval_runtime": 813.5415,
"eval_samples_per_second": 18.979,
"eval_steps_per_second": 2.372,
"eval_wer": 0.15558429452041306,
"step": 26500
},
{
"epoch": 3.71,
"learning_rate": 4.8545552367288374e-05,
"loss": 0.8245,
"step": 26600
},
{
"epoch": 3.72,
"learning_rate": 4.8437948350071735e-05,
"loss": 0.8225,
"step": 26700
},
{
"epoch": 3.74,
"learning_rate": 4.833034433285509e-05,
"loss": 0.8503,
"step": 26800
},
{
"epoch": 3.75,
"learning_rate": 4.822274031563844e-05,
"loss": 0.84,
"step": 26900
},
{
"epoch": 3.77,
"learning_rate": 4.81151362984218e-05,
"loss": 0.8299,
"step": 27000
},
{
"epoch": 3.77,
"eval_loss": 0.1612057387828827,
"eval_runtime": 811.2359,
"eval_samples_per_second": 19.033,
"eval_steps_per_second": 2.379,
"eval_wer": 0.15496720913286374,
"step": 27000
},
{
"epoch": 3.78,
"learning_rate": 4.8007532281205165e-05,
"loss": 0.8448,
"step": 27100
},
{
"epoch": 3.79,
"learning_rate": 4.789992826398852e-05,
"loss": 0.8287,
"step": 27200
},
{
"epoch": 3.81,
"learning_rate": 4.7792324246771874e-05,
"loss": 0.8402,
"step": 27300
},
{
"epoch": 3.82,
"learning_rate": 4.768472022955523e-05,
"loss": 0.8424,
"step": 27400
},
{
"epoch": 3.84,
"learning_rate": 4.757711621233859e-05,
"loss": 0.8421,
"step": 27500
},
{
"epoch": 3.84,
"eval_loss": 0.1601605862379074,
"eval_runtime": 814.3515,
"eval_samples_per_second": 18.96,
"eval_steps_per_second": 2.37,
"eval_wer": 0.15522323392131504,
"step": 27500
},
{
"epoch": 3.85,
"learning_rate": 4.746951219512195e-05,
"loss": 0.8389,
"step": 27600
},
{
"epoch": 3.86,
"learning_rate": 4.7361908177905305e-05,
"loss": 0.846,
"step": 27700
},
{
"epoch": 3.88,
"learning_rate": 4.725430416068866e-05,
"loss": 0.8323,
"step": 27800
},
{
"epoch": 3.89,
"learning_rate": 4.714670014347202e-05,
"loss": 0.8355,
"step": 27900
},
{
"epoch": 3.91,
"learning_rate": 4.7039096126255374e-05,
"loss": 0.8375,
"step": 28000
},
{
"epoch": 3.91,
"eval_loss": 0.15917135775089264,
"eval_runtime": 812.5996,
"eval_samples_per_second": 19.001,
"eval_steps_per_second": 2.375,
"eval_wer": 0.1536542614997801,
"step": 28000
},
{
"epoch": 3.92,
"learning_rate": 4.6931492109038735e-05,
"loss": 0.8505,
"step": 28100
},
{
"epoch": 3.93,
"learning_rate": 4.682388809182209e-05,
"loss": 0.8399,
"step": 28200
},
{
"epoch": 3.95,
"learning_rate": 4.671628407460545e-05,
"loss": 0.8144,
"step": 28300
},
{
"epoch": 3.96,
"learning_rate": 4.6608680057388805e-05,
"loss": 0.8366,
"step": 28400
},
{
"epoch": 3.97,
"learning_rate": 4.650107604017216e-05,
"loss": 0.8328,
"step": 28500
},
{
"epoch": 3.97,
"eval_loss": 0.15874174237251282,
"eval_runtime": 815.2612,
"eval_samples_per_second": 18.939,
"eval_steps_per_second": 2.367,
"eval_wer": 0.15370677940510344,
"step": 28500
},
{
"epoch": 3.99,
"learning_rate": 4.6393472022955513e-05,
"loss": 0.836,
"step": 28600
},
{
"epoch": 4.0,
"learning_rate": 4.628586800573888e-05,
"loss": 0.8237,
"step": 28700
},
{
"epoch": 4.02,
"learning_rate": 4.6178263988522236e-05,
"loss": 0.8168,
"step": 28800
},
{
"epoch": 4.03,
"learning_rate": 4.607065997130559e-05,
"loss": 0.8184,
"step": 28900
},
{
"epoch": 4.04,
"learning_rate": 4.5963055954088944e-05,
"loss": 0.8155,
"step": 29000
},
{
"epoch": 4.04,
"eval_loss": 0.15868720412254333,
"eval_runtime": 816.5093,
"eval_samples_per_second": 18.91,
"eval_steps_per_second": 2.364,
"eval_wer": 0.1519671237912676,
"step": 29000
},
{
"epoch": 4.06,
"learning_rate": 4.585545193687231e-05,
"loss": 0.8367,
"step": 29100
},
{
"epoch": 4.07,
"learning_rate": 4.5747847919655666e-05,
"loss": 0.8119,
"step": 29200
},
{
"epoch": 4.09,
"learning_rate": 4.564024390243902e-05,
"loss": 0.8144,
"step": 29300
},
{
"epoch": 4.1,
"learning_rate": 4.5532639885222375e-05,
"loss": 0.8174,
"step": 29400
},
{
"epoch": 4.11,
"learning_rate": 4.5425035868005736e-05,
"loss": 0.8335,
"step": 29500
},
{
"epoch": 4.11,
"eval_loss": 0.16237002611160278,
"eval_runtime": 806.8941,
"eval_samples_per_second": 19.135,
"eval_steps_per_second": 2.392,
"eval_wer": 0.15564994190206724,
"step": 29500
},
{
"epoch": 4.13,
"learning_rate": 4.531743185078909e-05,
"loss": 0.8293,
"step": 29600
},
{
"epoch": 4.14,
"learning_rate": 4.520982783357245e-05,
"loss": 0.8238,
"step": 29700
},
{
"epoch": 4.16,
"learning_rate": 4.5102223816355805e-05,
"loss": 0.8136,
"step": 29800
},
{
"epoch": 4.17,
"learning_rate": 4.4994619799139166e-05,
"loss": 0.8462,
"step": 29900
},
{
"epoch": 4.18,
"learning_rate": 4.488701578192252e-05,
"loss": 0.8138,
"step": 30000
},
{
"epoch": 4.18,
"eval_loss": 0.1581183820962906,
"eval_runtime": 811.5914,
"eval_samples_per_second": 19.024,
"eval_steps_per_second": 2.378,
"eval_wer": 0.1546652311772545,
"step": 30000
},
{
"epoch": 4.2,
"learning_rate": 4.478048780487804e-05,
"loss": 0.8258,
"step": 30100
},
{
"epoch": 4.21,
"learning_rate": 4.4672883787661397e-05,
"loss": 0.8245,
"step": 30200
},
{
"epoch": 4.23,
"learning_rate": 4.4565279770444764e-05,
"loss": 0.8335,
"step": 30300
},
{
"epoch": 4.24,
"learning_rate": 4.445767575322812e-05,
"loss": 0.8256,
"step": 30400
},
{
"epoch": 4.25,
"learning_rate": 4.435007173601147e-05,
"loss": 0.8195,
"step": 30500
},
{
"epoch": 4.25,
"eval_loss": 0.15603557229042053,
"eval_runtime": 808.6853,
"eval_samples_per_second": 19.093,
"eval_steps_per_second": 2.387,
"eval_wer": 0.1506804351108456,
"step": 30500
},
{
"epoch": 4.27,
"learning_rate": 4.424246771879483e-05,
"loss": 0.8098,
"step": 30600
},
{
"epoch": 4.28,
"learning_rate": 4.413486370157819e-05,
"loss": 0.822,
"step": 30700
},
{
"epoch": 4.3,
"learning_rate": 4.402725968436155e-05,
"loss": 0.8236,
"step": 30800
},
{
"epoch": 4.31,
"learning_rate": 4.3919655667144904e-05,
"loss": 0.8192,
"step": 30900
},
{
"epoch": 4.32,
"learning_rate": 4.381205164992826e-05,
"loss": 0.8092,
"step": 31000
},
{
"epoch": 4.32,
"eval_loss": 0.15605725347995758,
"eval_runtime": 804.1445,
"eval_samples_per_second": 19.201,
"eval_steps_per_second": 2.4,
"eval_wer": 0.15339167197316336,
"step": 31000
},
{
"epoch": 4.34,
"learning_rate": 4.370444763271162e-05,
"loss": 0.8265,
"step": 31100
},
{
"epoch": 4.35,
"learning_rate": 4.359684361549497e-05,
"loss": 0.8164,
"step": 31200
},
{
"epoch": 4.37,
"learning_rate": 4.348923959827833e-05,
"loss": 0.8123,
"step": 31300
},
{
"epoch": 4.38,
"learning_rate": 4.338163558106169e-05,
"loss": 0.8176,
"step": 31400
},
{
"epoch": 4.39,
"learning_rate": 4.327403156384505e-05,
"loss": 0.8191,
"step": 31500
},
{
"epoch": 4.39,
"eval_loss": 0.154859259724617,
"eval_runtime": 813.9946,
"eval_samples_per_second": 18.968,
"eval_steps_per_second": 2.371,
"eval_wer": 0.1493412285251003,
"step": 31500
},
{
"epoch": 4.41,
"learning_rate": 4.3166427546628404e-05,
"loss": 0.8223,
"step": 31600
},
{
"epoch": 4.42,
"learning_rate": 4.305882352941176e-05,
"loss": 0.8231,
"step": 31700
},
{
"epoch": 4.44,
"learning_rate": 4.295121951219511e-05,
"loss": 0.8134,
"step": 31800
},
{
"epoch": 4.45,
"learning_rate": 4.284361549497848e-05,
"loss": 0.8144,
"step": 31900
},
{
"epoch": 4.46,
"learning_rate": 4.2736011477761835e-05,
"loss": 0.8008,
"step": 32000
},
{
"epoch": 4.46,
"eval_loss": 0.15402169525623322,
"eval_runtime": 815.8948,
"eval_samples_per_second": 18.924,
"eval_steps_per_second": 2.366,
"eval_wer": 0.14928871061977694,
"step": 32000
},
{
"epoch": 4.48,
"learning_rate": 4.262840746054519e-05,
"loss": 0.8184,
"step": 32100
},
{
"epoch": 4.49,
"learning_rate": 4.252080344332854e-05,
"loss": 0.811,
"step": 32200
},
{
"epoch": 4.5,
"learning_rate": 4.2413199426111904e-05,
"loss": 0.8307,
"step": 32300
},
{
"epoch": 4.52,
"learning_rate": 4.2305595408895265e-05,
"loss": 0.8204,
"step": 32400
},
{
"epoch": 4.53,
"learning_rate": 4.219799139167862e-05,
"loss": 0.8138,
"step": 32500
},
{
"epoch": 4.53,
"eval_loss": 0.15436123311519623,
"eval_runtime": 819.1951,
"eval_samples_per_second": 18.848,
"eval_steps_per_second": 2.356,
"eval_wer": 0.14926245166711527,
"step": 32500
},
{
"epoch": 4.55,
"learning_rate": 4.2090387374461974e-05,
"loss": 0.8136,
"step": 32600
},
{
"epoch": 4.56,
"learning_rate": 4.1982783357245335e-05,
"loss": 0.8176,
"step": 32700
},
{
"epoch": 4.57,
"learning_rate": 4.187517934002869e-05,
"loss": 0.8104,
"step": 32800
},
{
"epoch": 4.59,
"learning_rate": 4.176757532281205e-05,
"loss": 0.8092,
"step": 32900
},
{
"epoch": 4.6,
"learning_rate": 4.1659971305595405e-05,
"loss": 0.8173,
"step": 33000
},
{
"epoch": 4.6,
"eval_loss": 0.15529733896255493,
"eval_runtime": 812.8116,
"eval_samples_per_second": 18.996,
"eval_steps_per_second": 2.374,
"eval_wer": 0.1510808841389361,
"step": 33000
},
{
"epoch": 4.62,
"learning_rate": 4.1552367288378766e-05,
"loss": 0.8246,
"step": 33100
},
{
"epoch": 4.63,
"learning_rate": 4.144476327116212e-05,
"loss": 0.823,
"step": 33200
},
{
"epoch": 4.64,
"learning_rate": 4.1337159253945474e-05,
"loss": 0.8041,
"step": 33300
},
{
"epoch": 4.66,
"learning_rate": 4.122955523672883e-05,
"loss": 0.8152,
"step": 33400
},
{
"epoch": 4.67,
"learning_rate": 4.1121951219512196e-05,
"loss": 0.8081,
"step": 33500
},
{
"epoch": 4.67,
"eval_loss": 0.15408268570899963,
"eval_runtime": 820.8329,
"eval_samples_per_second": 18.81,
"eval_steps_per_second": 2.351,
"eval_wer": 0.14840903570561087,
"step": 33500
},
{
"epoch": 4.69,
"learning_rate": 4.101434720229555e-05,
"loss": 0.8054,
"step": 33600
},
{
"epoch": 4.7,
"learning_rate": 4.0906743185078905e-05,
"loss": 0.8189,
"step": 33700
},
{
"epoch": 4.71,
"learning_rate": 4.079913916786226e-05,
"loss": 0.8148,
"step": 33800
},
{
"epoch": 4.73,
"learning_rate": 4.069153515064563e-05,
"loss": 0.8136,
"step": 33900
},
{
"epoch": 4.74,
"learning_rate": 4.058393113342898e-05,
"loss": 0.8192,
"step": 34000
},
{
"epoch": 4.74,
"eval_loss": 0.15597496926784515,
"eval_runtime": 818.1096,
"eval_samples_per_second": 18.873,
"eval_steps_per_second": 2.359,
"eval_wer": 0.1506476114200185,
"step": 34000
},
{
"epoch": 4.76,
"learning_rate": 4.0476327116212335e-05,
"loss": 0.8072,
"step": 34100
},
{
"epoch": 4.77,
"learning_rate": 4.036872309899569e-05,
"loss": 0.816,
"step": 34200
},
{
"epoch": 4.78,
"learning_rate": 4.026219512195122e-05,
"loss": 0.8189,
"step": 34300
},
{
"epoch": 4.8,
"learning_rate": 4.015459110473457e-05,
"loss": 0.8101,
"step": 34400
},
{
"epoch": 4.81,
"learning_rate": 4.004698708751793e-05,
"loss": 0.8068,
"step": 34500
},
{
"epoch": 4.81,
"eval_loss": 0.15400581061840057,
"eval_runtime": 813.2994,
"eval_samples_per_second": 18.984,
"eval_steps_per_second": 2.373,
"eval_wer": 0.15033250398807843,
"step": 34500
},
{
"epoch": 4.83,
"learning_rate": 3.993938307030128e-05,
"loss": 0.8034,
"step": 34600
},
{
"epoch": 4.84,
"learning_rate": 3.983177905308465e-05,
"loss": 0.803,
"step": 34700
},
{
"epoch": 4.85,
"learning_rate": 3.9724175035868e-05,
"loss": 0.7913,
"step": 34800
},
{
"epoch": 4.87,
"learning_rate": 3.961657101865136e-05,
"loss": 0.8055,
"step": 34900
},
{
"epoch": 4.88,
"learning_rate": 3.9510043041606886e-05,
"loss": 0.8105,
"step": 35000
},
{
"epoch": 4.88,
"eval_loss": 0.1528894603252411,
"eval_runtime": 803.1219,
"eval_samples_per_second": 19.225,
"eval_steps_per_second": 2.403,
"eval_wer": 0.1483433883239567,
"step": 35000
},
{
"epoch": 4.9,
"learning_rate": 3.940351506456241e-05,
"loss": 0.8254,
"step": 35100
},
{
"epoch": 4.91,
"learning_rate": 3.929591104734576e-05,
"loss": 0.7991,
"step": 35200
},
{
"epoch": 4.92,
"learning_rate": 3.9188307030129116e-05,
"loss": 0.811,
"step": 35300
},
{
"epoch": 4.94,
"learning_rate": 3.9080703012912484e-05,
"loss": 0.8067,
"step": 35400
},
{
"epoch": 4.95,
"learning_rate": 3.897309899569584e-05,
"loss": 0.7976,
"step": 35500
},
{
"epoch": 4.95,
"eval_loss": 0.15067319571971893,
"eval_runtime": 830.1567,
"eval_samples_per_second": 18.599,
"eval_steps_per_second": 2.325,
"eval_wer": 0.14513979609923258,
"step": 35500
},
{
"epoch": 4.97,
"learning_rate": 3.886549497847919e-05,
"loss": 0.8179,
"step": 35600
},
{
"epoch": 4.98,
"learning_rate": 3.8757890961262546e-05,
"loss": 0.8064,
"step": 35700
},
{
"epoch": 4.99,
"learning_rate": 3.865028694404591e-05,
"loss": 0.8118,
"step": 35800
},
{
"epoch": 5.01,
"learning_rate": 3.854268292682927e-05,
"loss": 0.8247,
"step": 35900
},
{
"epoch": 5.02,
"learning_rate": 3.843507890961262e-05,
"loss": 0.8143,
"step": 36000
},
{
"epoch": 5.02,
"eval_loss": 0.1505265235900879,
"eval_runtime": 815.1828,
"eval_samples_per_second": 18.941,
"eval_steps_per_second": 2.368,
"eval_wer": 0.14620328368203034,
"step": 36000
},
{
"epoch": 5.03,
"learning_rate": 3.832747489239598e-05,
"loss": 0.8003,
"step": 36100
},
{
"epoch": 5.05,
"learning_rate": 3.821987087517934e-05,
"loss": 0.789,
"step": 36200
},
{
"epoch": 5.06,
"learning_rate": 3.811226685796269e-05,
"loss": 0.8047,
"step": 36300
},
{
"epoch": 5.08,
"learning_rate": 3.8004662840746053e-05,
"loss": 0.7983,
"step": 36400
},
{
"epoch": 5.09,
"learning_rate": 3.789705882352941e-05,
"loss": 0.8053,
"step": 36500
},
{
"epoch": 5.09,
"eval_loss": 0.15166980028152466,
"eval_runtime": 808.3392,
"eval_samples_per_second": 19.101,
"eval_steps_per_second": 2.388,
"eval_wer": 0.14757531395860277,
"step": 36500
},
{
"epoch": 5.1,
"learning_rate": 3.778945480631277e-05,
"loss": 0.7996,
"step": 36600
},
{
"epoch": 5.12,
"learning_rate": 3.768185078909612e-05,
"loss": 0.7984,
"step": 36700
},
{
"epoch": 5.13,
"learning_rate": 3.757424677187948e-05,
"loss": 0.7851,
"step": 36800
},
{
"epoch": 5.15,
"learning_rate": 3.746664275466284e-05,
"loss": 0.7984,
"step": 36900
},
{
"epoch": 5.16,
"learning_rate": 3.735903873744619e-05,
"loss": 0.785,
"step": 37000
},
{
"epoch": 5.16,
"eval_loss": 0.1525907963514328,
"eval_runtime": 816.4679,
"eval_samples_per_second": 18.911,
"eval_steps_per_second": 2.364,
"eval_wer": 0.14781164453255782,
"step": 37000
},
{
"epoch": 5.17,
"learning_rate": 3.7251434720229554e-05,
"loss": 0.7975,
"step": 37100
},
{
"epoch": 5.19,
"learning_rate": 3.714383070301291e-05,
"loss": 0.8007,
"step": 37200
},
{
"epoch": 5.2,
"learning_rate": 3.703622668579627e-05,
"loss": 0.8003,
"step": 37300
},
{
"epoch": 5.22,
"learning_rate": 3.6928622668579623e-05,
"loss": 0.7983,
"step": 37400
},
{
"epoch": 5.23,
"learning_rate": 3.6821018651362984e-05,
"loss": 0.7936,
"step": 37500
},
{
"epoch": 5.23,
"eval_loss": 0.14893749356269836,
"eval_runtime": 821.3523,
"eval_samples_per_second": 18.798,
"eval_steps_per_second": 2.35,
"eval_wer": 0.14213971075763643,
"step": 37500
},
{
"epoch": 5.24,
"learning_rate": 3.671341463414634e-05,
"loss": 0.7987,
"step": 37600
},
{
"epoch": 5.26,
"learning_rate": 3.66058106169297e-05,
"loss": 0.7858,
"step": 37700
},
{
"epoch": 5.27,
"learning_rate": 3.6498206599713054e-05,
"loss": 0.7979,
"step": 37800
},
{
"epoch": 5.29,
"learning_rate": 3.639060258249641e-05,
"loss": 0.7955,
"step": 37900
},
{
"epoch": 5.3,
"learning_rate": 3.628299856527977e-05,
"loss": 0.807,
"step": 38000
},
{
"epoch": 5.3,
"eval_loss": 0.1482589840888977,
"eval_runtime": 813.343,
"eval_samples_per_second": 18.983,
"eval_steps_per_second": 2.373,
"eval_wer": 0.14198872177983182,
"step": 38000
},
{
"epoch": 5.31,
"learning_rate": 3.617647058823529e-05,
"loss": 0.7998,
"step": 38100
},
{
"epoch": 5.33,
"learning_rate": 3.6068866571018645e-05,
"loss": 0.7901,
"step": 38200
},
{
"epoch": 5.34,
"learning_rate": 3.5961262553802006e-05,
"loss": 0.8109,
"step": 38300
},
{
"epoch": 5.36,
"learning_rate": 3.585365853658536e-05,
"loss": 0.7955,
"step": 38400
},
{
"epoch": 5.37,
"learning_rate": 3.574605451936872e-05,
"loss": 0.8092,
"step": 38500
},
{
"epoch": 5.37,
"eval_loss": 0.14809782803058624,
"eval_runtime": 810.6526,
"eval_samples_per_second": 19.046,
"eval_steps_per_second": 2.381,
"eval_wer": 0.14350517629604342,
"step": 38500
},
{
"epoch": 5.38,
"learning_rate": 3.5638450502152076e-05,
"loss": 0.8044,
"step": 38600
},
{
"epoch": 5.4,
"learning_rate": 3.553084648493544e-05,
"loss": 0.803,
"step": 38700
},
{
"epoch": 5.41,
"learning_rate": 3.542324246771879e-05,
"loss": 0.8014,
"step": 38800
},
{
"epoch": 5.43,
"learning_rate": 3.531563845050215e-05,
"loss": 0.7905,
"step": 38900
},
{
"epoch": 5.44,
"learning_rate": 3.5208034433285507e-05,
"loss": 0.793,
"step": 39000
},
{
"epoch": 5.44,
"eval_loss": 0.15028123557567596,
"eval_runtime": 824.2677,
"eval_samples_per_second": 18.732,
"eval_steps_per_second": 2.341,
"eval_wer": 0.14376776582266015,
"step": 39000
},
{
"epoch": 5.45,
"learning_rate": 3.510043041606887e-05,
"loss": 0.8069,
"step": 39100
},
{
"epoch": 5.47,
"learning_rate": 3.499282639885222e-05,
"loss": 0.7981,
"step": 39200
},
{
"epoch": 5.48,
"learning_rate": 3.4885222381635576e-05,
"loss": 0.8026,
"step": 39300
},
{
"epoch": 5.5,
"learning_rate": 3.477761836441894e-05,
"loss": 0.7884,
"step": 39400
},
{
"epoch": 5.51,
"learning_rate": 3.467001434720229e-05,
"loss": 0.814,
"step": 39500
},
{
"epoch": 5.51,
"eval_loss": 0.14950084686279297,
"eval_runtime": 812.6951,
"eval_samples_per_second": 18.999,
"eval_steps_per_second": 2.375,
"eval_wer": 0.1479823277248587,
"step": 39500
},
{
"epoch": 5.52,
"learning_rate": 3.4562410329985646e-05,
"loss": 0.8009,
"step": 39600
},
{
"epoch": 5.54,
"learning_rate": 3.445480631276901e-05,
"loss": 0.7933,
"step": 39700
},
{
"epoch": 5.55,
"learning_rate": 3.434720229555236e-05,
"loss": 0.7975,
"step": 39800
},
{
"epoch": 5.56,
"learning_rate": 3.423959827833572e-05,
"loss": 0.7919,
"step": 39900
},
{
"epoch": 5.58,
"learning_rate": 3.4131994261119076e-05,
"loss": 0.807,
"step": 40000
},
{
"epoch": 5.58,
"eval_loss": 0.14722611010074615,
"eval_runtime": 812.7319,
"eval_samples_per_second": 18.998,
"eval_steps_per_second": 2.375,
"eval_wer": 0.14235634711709524,
"step": 40000
},
{
"epoch": 5.59,
"learning_rate": 3.402439024390244e-05,
"loss": 0.7951,
"step": 40100
},
{
"epoch": 5.61,
"learning_rate": 3.391678622668579e-05,
"loss": 0.789,
"step": 40200
},
{
"epoch": 5.62,
"learning_rate": 3.380918220946915e-05,
"loss": 0.7916,
"step": 40300
},
{
"epoch": 5.63,
"learning_rate": 3.370157819225251e-05,
"loss": 0.7892,
"step": 40400
},
{
"epoch": 5.65,
"learning_rate": 3.359397417503587e-05,
"loss": 0.7913,
"step": 40500
},
{
"epoch": 5.65,
"eval_loss": 0.14706110954284668,
"eval_runtime": 811.5569,
"eval_samples_per_second": 19.025,
"eval_steps_per_second": 2.378,
"eval_wer": 0.14223818183011772,
"step": 40500
},
{
"epoch": 5.66,
"learning_rate": 3.348637015781922e-05,
"loss": 0.7909,
"step": 40600
},
{
"epoch": 5.68,
"learning_rate": 3.3378766140602584e-05,
"loss": 0.7738,
"step": 40700
},
{
"epoch": 5.69,
"learning_rate": 3.327116212338594e-05,
"loss": 0.7921,
"step": 40800
},
{
"epoch": 5.7,
"learning_rate": 3.316355810616929e-05,
"loss": 0.7953,
"step": 40900
},
{
"epoch": 5.72,
"learning_rate": 3.305595408895265e-05,
"loss": 0.7844,
"step": 41000
},
{
"epoch": 5.72,
"eval_loss": 0.14731355011463165,
"eval_runtime": 811.7297,
"eval_samples_per_second": 19.021,
"eval_steps_per_second": 2.378,
"eval_wer": 0.14219222866295977,
"step": 41000
},
{
"epoch": 5.73,
"learning_rate": 3.294835007173601e-05,
"loss": 0.8009,
"step": 41100
},
{
"epoch": 5.75,
"learning_rate": 3.284074605451937e-05,
"loss": 0.7839,
"step": 41200
},
{
"epoch": 5.76,
"learning_rate": 3.273314203730272e-05,
"loss": 0.7789,
"step": 41300
},
{
"epoch": 5.77,
"learning_rate": 3.262553802008608e-05,
"loss": 0.7778,
"step": 41400
},
{
"epoch": 5.79,
"learning_rate": 3.2519010043041605e-05,
"loss": 0.7888,
"step": 41500
},
{
"epoch": 5.79,
"eval_loss": 0.14446745812892914,
"eval_runtime": 834.7197,
"eval_samples_per_second": 18.497,
"eval_steps_per_second": 2.312,
"eval_wer": 0.1384962810758293,
"step": 41500
},
{
"epoch": 5.8,
"learning_rate": 3.241140602582496e-05,
"loss": 0.7826,
"step": 41600
},
{
"epoch": 5.82,
"learning_rate": 3.230380200860832e-05,
"loss": 0.7831,
"step": 41700
},
{
"epoch": 5.83,
"learning_rate": 3.2196197991391675e-05,
"loss": 0.7849,
"step": 41800
},
{
"epoch": 5.84,
"learning_rate": 3.2088593974175036e-05,
"loss": 0.7847,
"step": 41900
},
{
"epoch": 5.86,
"learning_rate": 3.198098995695839e-05,
"loss": 0.7806,
"step": 42000
},
{
"epoch": 5.86,
"eval_loss": 0.14347364008426666,
"eval_runtime": 811.264,
"eval_samples_per_second": 19.032,
"eval_steps_per_second": 2.379,
"eval_wer": 0.13944816810981495,
"step": 42000
},
{
"epoch": 5.87,
"learning_rate": 3.187338593974175e-05,
"loss": 0.7943,
"step": 42100
},
{
"epoch": 5.89,
"learning_rate": 3.1765781922525106e-05,
"loss": 0.7856,
"step": 42200
},
{
"epoch": 5.9,
"learning_rate": 3.165817790530846e-05,
"loss": 0.7857,
"step": 42300
},
{
"epoch": 5.91,
"learning_rate": 3.155057388809182e-05,
"loss": 0.7947,
"step": 42400
},
{
"epoch": 5.93,
"learning_rate": 3.1442969870875175e-05,
"loss": 0.7773,
"step": 42500
},
{
"epoch": 5.93,
"eval_loss": 0.14614766836166382,
"eval_runtime": 825.4168,
"eval_samples_per_second": 18.706,
"eval_steps_per_second": 2.338,
"eval_wer": 0.14243512397508026,
"step": 42500
},
{
"epoch": 5.94,
"learning_rate": 3.133536585365853e-05,
"loss": 0.7914,
"step": 42600
},
{
"epoch": 5.96,
"learning_rate": 3.122776183644189e-05,
"loss": 0.7926,
"step": 42700
},
{
"epoch": 5.97,
"learning_rate": 3.1120157819225245e-05,
"loss": 0.7728,
"step": 42800
},
{
"epoch": 5.98,
"learning_rate": 3.1012553802008606e-05,
"loss": 0.7851,
"step": 42900
},
{
"epoch": 6.0,
"learning_rate": 3.090494978479196e-05,
"loss": 0.786,
"step": 43000
},
{
"epoch": 6.0,
"eval_loss": 0.14499257504940033,
"eval_runtime": 807.5088,
"eval_samples_per_second": 19.121,
"eval_steps_per_second": 2.39,
"eval_wer": 0.14128629479613206,
"step": 43000
},
{
"epoch": 6.01,
"learning_rate": 3.079734576757532e-05,
"loss": 0.7538,
"step": 43100
},
{
"epoch": 6.03,
"learning_rate": 3.0689741750358676e-05,
"loss": 0.7764,
"step": 43200
},
{
"epoch": 6.04,
"learning_rate": 3.058213773314204e-05,
"loss": 0.7839,
"step": 43300
},
{
"epoch": 6.05,
"learning_rate": 3.047453371592539e-05,
"loss": 0.7692,
"step": 43400
},
{
"epoch": 6.07,
"learning_rate": 3.036692969870875e-05,
"loss": 0.7784,
"step": 43500
},
{
"epoch": 6.07,
"eval_loss": 0.1462818831205368,
"eval_runtime": 804.4116,
"eval_samples_per_second": 19.194,
"eval_steps_per_second": 2.399,
"eval_wer": 0.14242855923691483,
"step": 43500
},
{
"epoch": 6.08,
"learning_rate": 3.0259325681492106e-05,
"loss": 0.7912,
"step": 43600
},
{
"epoch": 6.09,
"learning_rate": 3.0151721664275464e-05,
"loss": 0.7906,
"step": 43700
},
{
"epoch": 6.11,
"learning_rate": 3.004411764705882e-05,
"loss": 0.7782,
"step": 43800
},
{
"epoch": 6.12,
"learning_rate": 2.993651362984218e-05,
"loss": 0.7835,
"step": 43900
},
{
"epoch": 6.14,
"learning_rate": 2.98299856527977e-05,
"loss": 0.7937,
"step": 44000
},
{
"epoch": 6.14,
"eval_loss": 0.14378893375396729,
"eval_runtime": 811.2224,
"eval_samples_per_second": 19.033,
"eval_steps_per_second": 2.379,
"eval_wer": 0.1386407053154685,
"step": 44000
},
{
"epoch": 6.15,
"learning_rate": 2.972238163558106e-05,
"loss": 0.7766,
"step": 44100
},
{
"epoch": 6.16,
"learning_rate": 2.9614777618364416e-05,
"loss": 0.7805,
"step": 44200
},
{
"epoch": 6.18,
"learning_rate": 2.9507173601147774e-05,
"loss": 0.7807,
"step": 44300
},
{
"epoch": 6.19,
"learning_rate": 2.9399569583931128e-05,
"loss": 0.7847,
"step": 44400
},
{
"epoch": 6.21,
"learning_rate": 2.929196556671449e-05,
"loss": 0.7738,
"step": 44500
},
{
"epoch": 6.21,
"eval_loss": 0.143657848238945,
"eval_runtime": 802.0352,
"eval_samples_per_second": 19.251,
"eval_steps_per_second": 2.406,
"eval_wer": 0.138319033145363,
"step": 44500
},
{
"epoch": 6.22,
"learning_rate": 2.9184361549497843e-05,
"loss": 0.7704,
"step": 44600
},
{
"epoch": 6.23,
"learning_rate": 2.9076757532281204e-05,
"loss": 0.7778,
"step": 44700
},
{
"epoch": 6.25,
"learning_rate": 2.896915351506456e-05,
"loss": 0.7779,
"step": 44800
},
{
"epoch": 6.26,
"learning_rate": 2.8861549497847916e-05,
"loss": 0.759,
"step": 44900
},
{
"epoch": 6.28,
"learning_rate": 2.8753945480631274e-05,
"loss": 0.7728,
"step": 45000
},
{
"epoch": 6.28,
"eval_loss": 0.14244486391544342,
"eval_runtime": 813.5497,
"eval_samples_per_second": 18.979,
"eval_steps_per_second": 2.372,
"eval_wer": 0.13710455658476062,
"step": 45000
},
{
"epoch": 6.29,
"learning_rate": 2.8646341463414632e-05,
"loss": 0.7848,
"step": 45100
},
{
"epoch": 6.3,
"learning_rate": 2.8538737446197986e-05,
"loss": 0.7784,
"step": 45200
},
{
"epoch": 6.32,
"learning_rate": 2.8431133428981347e-05,
"loss": 0.7865,
"step": 45300
},
{
"epoch": 6.33,
"learning_rate": 2.83235294117647e-05,
"loss": 0.7771,
"step": 45400
},
{
"epoch": 6.35,
"learning_rate": 2.8215925394548062e-05,
"loss": 0.7681,
"step": 45500
},
{
"epoch": 6.35,
"eval_loss": 0.14163097739219666,
"eval_runtime": 820.3528,
"eval_samples_per_second": 18.821,
"eval_steps_per_second": 2.353,
"eval_wer": 0.13761660616166324,
"step": 45500
},
{
"epoch": 6.36,
"learning_rate": 2.8108321377331417e-05,
"loss": 0.7908,
"step": 45600
},
{
"epoch": 6.37,
"learning_rate": 2.8000717360114774e-05,
"loss": 0.7805,
"step": 45700
},
{
"epoch": 6.39,
"learning_rate": 2.7893113342898132e-05,
"loss": 0.7737,
"step": 45800
},
{
"epoch": 6.4,
"learning_rate": 2.778550932568149e-05,
"loss": 0.7797,
"step": 45900
},
{
"epoch": 6.42,
"learning_rate": 2.7677905308464844e-05,
"loss": 0.776,
"step": 46000
},
{
"epoch": 6.42,
"eval_loss": 0.1414603292942047,
"eval_runtime": 800.7729,
"eval_samples_per_second": 19.281,
"eval_steps_per_second": 2.41,
"eval_wer": 0.13797110202259583,
"step": 46000
},
{
"epoch": 6.43,
"learning_rate": 2.7570301291248205e-05,
"loss": 0.7765,
"step": 46100
},
{
"epoch": 6.44,
"learning_rate": 2.746269727403156e-05,
"loss": 0.7826,
"step": 46200
},
{
"epoch": 6.46,
"learning_rate": 2.735509325681492e-05,
"loss": 0.7674,
"step": 46300
},
{
"epoch": 6.47,
"learning_rate": 2.7248565279770445e-05,
"loss": 0.7837,
"step": 46400
},
{
"epoch": 6.49,
"learning_rate": 2.71409612625538e-05,
"loss": 0.7773,
"step": 46500
},
{
"epoch": 6.49,
"eval_loss": 0.14155656099319458,
"eval_runtime": 815.4547,
"eval_samples_per_second": 18.934,
"eval_steps_per_second": 2.367,
"eval_wer": 0.13709799184659519,
"step": 46500
},
{
"epoch": 6.5,
"learning_rate": 2.7033357245337157e-05,
"loss": 0.774,
"step": 46600
},
{
"epoch": 6.51,
"learning_rate": 2.6925753228120515e-05,
"loss": 0.7668,
"step": 46700
},
{
"epoch": 6.53,
"learning_rate": 2.6818149210903873e-05,
"loss": 0.7745,
"step": 46800
},
{
"epoch": 6.54,
"learning_rate": 2.671054519368723e-05,
"loss": 0.7788,
"step": 46900
},
{
"epoch": 6.56,
"learning_rate": 2.6602941176470588e-05,
"loss": 0.7692,
"step": 47000
},
{
"epoch": 6.56,
"eval_loss": 0.13975241780281067,
"eval_runtime": 812.8703,
"eval_samples_per_second": 18.994,
"eval_steps_per_second": 2.374,
"eval_wer": 0.13453774396208207,
"step": 47000
},
{
"epoch": 6.57,
"learning_rate": 2.6495337159253942e-05,
"loss": 0.7724,
"step": 47100
},
{
"epoch": 6.58,
"learning_rate": 2.63877331420373e-05,
"loss": 0.777,
"step": 47200
},
{
"epoch": 6.6,
"learning_rate": 2.6280129124820658e-05,
"loss": 0.7681,
"step": 47300
},
{
"epoch": 6.61,
"learning_rate": 2.6172525107604012e-05,
"loss": 0.7646,
"step": 47400
},
{
"epoch": 6.62,
"learning_rate": 2.6064921090387373e-05,
"loss": 0.7642,
"step": 47500
},
{
"epoch": 6.62,
"eval_loss": 0.13808567821979523,
"eval_runtime": 814.9147,
"eval_samples_per_second": 18.947,
"eval_steps_per_second": 2.368,
"eval_wer": 0.13412416545766073,
"step": 47500
},
{
"epoch": 6.64,
"learning_rate": 2.5958393113342898e-05,
"loss": 0.7622,
"step": 47600
},
{
"epoch": 6.65,
"learning_rate": 2.5850789096126252e-05,
"loss": 0.7735,
"step": 47700
},
{
"epoch": 6.67,
"learning_rate": 2.5743185078909613e-05,
"loss": 0.7763,
"step": 47800
},
{
"epoch": 6.68,
"learning_rate": 2.5635581061692967e-05,
"loss": 0.7657,
"step": 47900
},
{
"epoch": 6.69,
"learning_rate": 2.5527977044476325e-05,
"loss": 0.7692,
"step": 48000
},
{
"epoch": 6.69,
"eval_loss": 0.13921351730823517,
"eval_runtime": 801.0595,
"eval_samples_per_second": 19.274,
"eval_steps_per_second": 2.409,
"eval_wer": 0.13341517373579556,
"step": 48000
},
{
"epoch": 6.71,
"learning_rate": 2.5420373027259683e-05,
"loss": 0.7663,
"step": 48100
},
{
"epoch": 6.72,
"learning_rate": 2.531276901004304e-05,
"loss": 0.7796,
"step": 48200
},
{
"epoch": 6.74,
"learning_rate": 2.5205164992826395e-05,
"loss": 0.7734,
"step": 48300
},
{
"epoch": 6.75,
"learning_rate": 2.5097560975609756e-05,
"loss": 0.7673,
"step": 48400
},
{
"epoch": 6.76,
"learning_rate": 2.498995695839311e-05,
"loss": 0.7667,
"step": 48500
},
{
"epoch": 6.76,
"eval_loss": 0.13920722901821136,
"eval_runtime": 801.6435,
"eval_samples_per_second": 19.26,
"eval_steps_per_second": 2.408,
"eval_wer": 0.13482002770319507,
"step": 48500
},
{
"epoch": 6.78,
"learning_rate": 2.488235294117647e-05,
"loss": 0.7734,
"step": 48600
},
{
"epoch": 6.79,
"learning_rate": 2.4774748923959825e-05,
"loss": 0.771,
"step": 48700
},
{
"epoch": 6.81,
"learning_rate": 2.4667144906743183e-05,
"loss": 0.7671,
"step": 48800
},
{
"epoch": 6.82,
"learning_rate": 2.455954088952654e-05,
"loss": 0.7468,
"step": 48900
},
{
"epoch": 6.83,
"learning_rate": 2.44519368723099e-05,
"loss": 0.7712,
"step": 49000
},
{
"epoch": 6.83,
"eval_loss": 0.13978517055511475,
"eval_runtime": 804.4861,
"eval_samples_per_second": 19.192,
"eval_steps_per_second": 2.399,
"eval_wer": 0.13331013792514884,
"step": 49000
},
{
"epoch": 6.85,
"learning_rate": 2.4344332855093253e-05,
"loss": 0.7815,
"step": 49100
},
{
"epoch": 6.86,
"learning_rate": 2.4236728837876614e-05,
"loss": 0.769,
"step": 49200
},
{
"epoch": 6.88,
"learning_rate": 2.4129124820659968e-05,
"loss": 0.7736,
"step": 49300
},
{
"epoch": 6.89,
"learning_rate": 2.402152080344333e-05,
"loss": 0.7579,
"step": 49400
},
{
"epoch": 6.9,
"learning_rate": 2.3913916786226683e-05,
"loss": 0.7628,
"step": 49500
},
{
"epoch": 6.9,
"eval_loss": 0.13921290636062622,
"eval_runtime": 803.687,
"eval_samples_per_second": 19.211,
"eval_steps_per_second": 2.401,
"eval_wer": 0.13439331972244287,
"step": 49500
},
{
"epoch": 6.92,
"learning_rate": 2.3806312769010044e-05,
"loss": 0.7559,
"step": 49600
},
{
"epoch": 6.93,
"learning_rate": 2.36987087517934e-05,
"loss": 0.7539,
"step": 49700
},
{
"epoch": 6.95,
"learning_rate": 2.3591104734576756e-05,
"loss": 0.7681,
"step": 49800
},
{
"epoch": 6.96,
"learning_rate": 2.3483500717360114e-05,
"loss": 0.7641,
"step": 49900
},
{
"epoch": 6.97,
"learning_rate": 2.337589670014347e-05,
"loss": 0.7622,
"step": 50000
},
{
"epoch": 6.97,
"eval_loss": 0.13767020404338837,
"eval_runtime": 814.2982,
"eval_samples_per_second": 18.961,
"eval_steps_per_second": 2.37,
"eval_wer": 0.13285060625356956,
"step": 50000
},
{
"epoch": 6.99,
"learning_rate": 2.3268292682926826e-05,
"loss": 0.7791,
"step": 50100
},
{
"epoch": 7.0,
"learning_rate": 2.3160688665710187e-05,
"loss": 0.7716,
"step": 50200
},
{
"epoch": 7.02,
"learning_rate": 2.305308464849354e-05,
"loss": 0.7569,
"step": 50300
},
{
"epoch": 7.03,
"learning_rate": 2.2945480631276902e-05,
"loss": 0.7608,
"step": 50400
},
{
"epoch": 7.04,
"learning_rate": 2.2837876614060257e-05,
"loss": 0.7639,
"step": 50500
},
{
"epoch": 7.04,
"eval_loss": 0.13606122136116028,
"eval_runtime": 802.291,
"eval_samples_per_second": 19.245,
"eval_steps_per_second": 2.406,
"eval_wer": 0.13157704704947842,
"step": 50500
},
{
"epoch": 7.06,
"learning_rate": 2.2730272596843614e-05,
"loss": 0.7624,
"step": 50600
},
{
"epoch": 7.07,
"learning_rate": 2.2622668579626972e-05,
"loss": 0.7655,
"step": 50700
},
{
"epoch": 7.09,
"learning_rate": 2.2515064562410326e-05,
"loss": 0.764,
"step": 50800
},
{
"epoch": 7.1,
"learning_rate": 2.2407460545193684e-05,
"loss": 0.7602,
"step": 50900
},
{
"epoch": 7.11,
"learning_rate": 2.229985652797704e-05,
"loss": 0.742,
"step": 51000
},
{
"epoch": 7.11,
"eval_loss": 0.13759128749370575,
"eval_runtime": 807.5614,
"eval_samples_per_second": 19.119,
"eval_steps_per_second": 2.39,
"eval_wer": 0.13273900570475747,
"step": 51000
},
{
"epoch": 7.13,
"learning_rate": 2.21922525107604e-05,
"loss": 0.7677,
"step": 51100
},
{
"epoch": 7.14,
"learning_rate": 2.2084648493543754e-05,
"loss": 0.7551,
"step": 51200
},
{
"epoch": 7.15,
"learning_rate": 2.1977044476327115e-05,
"loss": 0.7607,
"step": 51300
},
{
"epoch": 7.17,
"learning_rate": 2.186944045911047e-05,
"loss": 0.7659,
"step": 51400
},
{
"epoch": 7.18,
"learning_rate": 2.176183644189383e-05,
"loss": 0.7526,
"step": 51500
},
{
"epoch": 7.18,
"eval_loss": 0.13867902755737305,
"eval_runtime": 802.6792,
"eval_samples_per_second": 19.236,
"eval_steps_per_second": 2.404,
"eval_wer": 0.13418981283931491,
"step": 51500
},
{
"epoch": 7.2,
"learning_rate": 2.1654232424677184e-05,
"loss": 0.77,
"step": 51600
},
{
"epoch": 7.21,
"learning_rate": 2.1546628407460545e-05,
"loss": 0.7771,
"step": 51700
},
{
"epoch": 7.22,
"learning_rate": 2.14390243902439e-05,
"loss": 0.7605,
"step": 51800
},
{
"epoch": 7.24,
"learning_rate": 2.1331420373027257e-05,
"loss": 0.7613,
"step": 51900
},
{
"epoch": 7.25,
"learning_rate": 2.1223816355810615e-05,
"loss": 0.7606,
"step": 52000
},
{
"epoch": 7.25,
"eval_loss": 0.13629250228405,
"eval_runtime": 804.4746,
"eval_samples_per_second": 19.193,
"eval_steps_per_second": 2.399,
"eval_wer": 0.13158361178764386,
"step": 52000
},
{
"epoch": 7.27,
"learning_rate": 2.1116212338593973e-05,
"loss": 0.7551,
"step": 52100
},
{
"epoch": 7.28,
"learning_rate": 2.1008608321377327e-05,
"loss": 0.7688,
"step": 52200
},
{
"epoch": 7.29,
"learning_rate": 2.0901004304160688e-05,
"loss": 0.7755,
"step": 52300
},
{
"epoch": 7.31,
"learning_rate": 2.0793400286944042e-05,
"loss": 0.7499,
"step": 52400
},
{
"epoch": 7.32,
"learning_rate": 2.0685796269727403e-05,
"loss": 0.7626,
"step": 52500
},
{
"epoch": 7.32,
"eval_loss": 0.13645130395889282,
"eval_runtime": 803.1619,
"eval_samples_per_second": 19.224,
"eval_steps_per_second": 2.403,
"eval_wer": 0.13132102226102713,
"step": 52500
},
{
"epoch": 7.34,
"learning_rate": 2.0578192252510758e-05,
"loss": 0.7711,
"step": 52600
},
{
"epoch": 7.35,
"learning_rate": 2.0470588235294115e-05,
"loss": 0.7542,
"step": 52700
},
{
"epoch": 7.36,
"learning_rate": 2.0362984218077473e-05,
"loss": 0.77,
"step": 52800
},
{
"epoch": 7.38,
"learning_rate": 2.025538020086083e-05,
"loss": 0.7581,
"step": 52900
},
{
"epoch": 7.39,
"learning_rate": 2.0147776183644185e-05,
"loss": 0.752,
"step": 53000
},
{
"epoch": 7.39,
"eval_loss": 0.13542360067367554,
"eval_runtime": 801.8612,
"eval_samples_per_second": 19.255,
"eval_steps_per_second": 2.407,
"eval_wer": 0.13090744375660576,
"step": 53000
},
{
"epoch": 7.41,
"learning_rate": 2.0040172166427546e-05,
"loss": 0.7595,
"step": 53100
},
{
"epoch": 7.42,
"learning_rate": 1.99325681492109e-05,
"loss": 0.7534,
"step": 53200
},
{
"epoch": 7.43,
"learning_rate": 1.982496413199426e-05,
"loss": 0.7587,
"step": 53300
},
{
"epoch": 7.45,
"learning_rate": 1.9717360114777615e-05,
"loss": 0.7501,
"step": 53400
},
{
"epoch": 7.46,
"learning_rate": 1.9609756097560973e-05,
"loss": 0.7562,
"step": 53500
},
{
"epoch": 7.46,
"eval_loss": 0.13623282313346863,
"eval_runtime": 801.4119,
"eval_samples_per_second": 19.266,
"eval_steps_per_second": 2.408,
"eval_wer": 0.1311700332832225,
"step": 53500
},
{
"epoch": 7.48,
"learning_rate": 1.950215208034433e-05,
"loss": 0.752,
"step": 53600
},
{
"epoch": 7.49,
"learning_rate": 1.939454806312769e-05,
"loss": 0.7588,
"step": 53700
},
{
"epoch": 7.5,
"learning_rate": 1.9288020086083213e-05,
"loss": 0.7438,
"step": 53800
},
{
"epoch": 7.52,
"learning_rate": 1.9181492109038735e-05,
"loss": 0.7584,
"step": 53900
},
{
"epoch": 7.53,
"learning_rate": 1.9073888091822093e-05,
"loss": 0.7557,
"step": 54000
},
{
"epoch": 7.53,
"eval_loss": 0.13583356142044067,
"eval_runtime": 818.349,
"eval_samples_per_second": 18.867,
"eval_steps_per_second": 2.358,
"eval_wer": 0.13247641617814074,
"step": 54000
},
{
"epoch": 7.55,
"learning_rate": 1.896628407460545e-05,
"loss": 0.7588,
"step": 54100
},
{
"epoch": 7.56,
"learning_rate": 1.8858680057388808e-05,
"loss": 0.7581,
"step": 54200
},
{
"epoch": 7.57,
"learning_rate": 1.8751076040172162e-05,
"loss": 0.7467,
"step": 54300
},
{
"epoch": 7.59,
"learning_rate": 1.864347202295552e-05,
"loss": 0.7537,
"step": 54400
},
{
"epoch": 7.6,
"learning_rate": 1.8535868005738878e-05,
"loss": 0.7588,
"step": 54500
},
{
"epoch": 7.6,
"eval_loss": 0.1342582106590271,
"eval_runtime": 808.1466,
"eval_samples_per_second": 19.105,
"eval_steps_per_second": 2.388,
"eval_wer": 0.13114377433056082,
"step": 54500
},
{
"epoch": 7.62,
"learning_rate": 1.8428263988522235e-05,
"loss": 0.7486,
"step": 54600
},
{
"epoch": 7.63,
"learning_rate": 1.8320659971305593e-05,
"loss": 0.7517,
"step": 54700
},
{
"epoch": 7.64,
"learning_rate": 1.821305595408895e-05,
"loss": 0.7471,
"step": 54800
},
{
"epoch": 7.66,
"learning_rate": 1.8105451936872308e-05,
"loss": 0.7553,
"step": 54900
},
{
"epoch": 7.67,
"learning_rate": 1.7997847919655666e-05,
"loss": 0.7485,
"step": 55000
},
{
"epoch": 7.67,
"eval_loss": 0.134577676653862,
"eval_runtime": 803.8799,
"eval_samples_per_second": 19.207,
"eval_steps_per_second": 2.401,
"eval_wer": 0.13014593412941725,
"step": 55000
},
{
"epoch": 7.68,
"learning_rate": 1.7890243902439024e-05,
"loss": 0.745,
"step": 55100
},
{
"epoch": 7.7,
"learning_rate": 1.7782639885222378e-05,
"loss": 0.7502,
"step": 55200
},
{
"epoch": 7.71,
"learning_rate": 1.7675035868005736e-05,
"loss": 0.7539,
"step": 55300
},
{
"epoch": 7.73,
"learning_rate": 1.7567431850789093e-05,
"loss": 0.7441,
"step": 55400
},
{
"epoch": 7.74,
"learning_rate": 1.745982783357245e-05,
"loss": 0.7466,
"step": 55500
},
{
"epoch": 7.74,
"eval_loss": 0.13543131947517395,
"eval_runtime": 801.1575,
"eval_samples_per_second": 19.272,
"eval_steps_per_second": 2.409,
"eval_wer": 0.13142605807167382,
"step": 55500
},
{
"epoch": 7.75,
"learning_rate": 1.735222381635581e-05,
"loss": 0.7463,
"step": 55600
},
{
"epoch": 7.77,
"learning_rate": 1.7244619799139166e-05,
"loss": 0.7499,
"step": 55700
},
{
"epoch": 7.78,
"learning_rate": 1.7137015781922524e-05,
"loss": 0.7455,
"step": 55800
},
{
"epoch": 7.8,
"learning_rate": 1.702941176470588e-05,
"loss": 0.7471,
"step": 55900
},
{
"epoch": 7.81,
"learning_rate": 1.692180774748924e-05,
"loss": 0.7558,
"step": 56000
},
{
"epoch": 7.81,
"eval_loss": 0.13593655824661255,
"eval_runtime": 803.2169,
"eval_samples_per_second": 19.223,
"eval_steps_per_second": 2.403,
"eval_wer": 0.13252236934529865,
"step": 56000
},
{
"epoch": 7.82,
"learning_rate": 1.6814203730272593e-05,
"loss": 0.7403,
"step": 56100
},
{
"epoch": 7.84,
"learning_rate": 1.670659971305595e-05,
"loss": 0.751,
"step": 56200
},
{
"epoch": 7.85,
"learning_rate": 1.659899569583931e-05,
"loss": 0.7513,
"step": 56300
},
{
"epoch": 7.87,
"learning_rate": 1.6491391678622666e-05,
"loss": 0.7562,
"step": 56400
},
{
"epoch": 7.88,
"learning_rate": 1.6383787661406024e-05,
"loss": 0.7578,
"step": 56500
},
{
"epoch": 7.88,
"eval_loss": 0.13627561926841736,
"eval_runtime": 800.366,
"eval_samples_per_second": 19.291,
"eval_steps_per_second": 2.411,
"eval_wer": 0.1333889147831339,
"step": 56500
},
{
"epoch": 7.89,
"learning_rate": 1.6276183644189382e-05,
"loss": 0.7484,
"step": 56600
},
{
"epoch": 7.91,
"learning_rate": 1.6169655667144907e-05,
"loss": 0.7497,
"step": 56700
},
{
"epoch": 7.92,
"learning_rate": 1.6062051649928264e-05,
"loss": 0.7507,
"step": 56800
},
{
"epoch": 7.94,
"learning_rate": 1.5954447632711622e-05,
"loss": 0.752,
"step": 56900
},
{
"epoch": 7.95,
"learning_rate": 1.5846843615494976e-05,
"loss": 0.7411,
"step": 57000
},
{
"epoch": 7.95,
"eval_loss": 0.1345953643321991,
"eval_runtime": 802.4506,
"eval_samples_per_second": 19.241,
"eval_steps_per_second": 2.405,
"eval_wer": 0.130126239914921,
"step": 57000
},
{
"epoch": 7.96,
"learning_rate": 1.5739239598278334e-05,
"loss": 0.7551,
"step": 57100
},
{
"epoch": 7.98,
"learning_rate": 1.5631635581061692e-05,
"loss": 0.7395,
"step": 57200
},
{
"epoch": 7.99,
"learning_rate": 1.552403156384505e-05,
"loss": 0.7463,
"step": 57300
},
{
"epoch": 8.01,
"learning_rate": 1.5416427546628407e-05,
"loss": 0.7454,
"step": 57400
},
{
"epoch": 8.02,
"learning_rate": 1.5308823529411765e-05,
"loss": 0.7478,
"step": 57500
},
{
"epoch": 8.02,
"eval_loss": 0.1355125904083252,
"eval_runtime": 802.8161,
"eval_samples_per_second": 19.232,
"eval_steps_per_second": 2.404,
"eval_wer": 0.1305201242048461,
"step": 57500
},
{
"epoch": 8.03,
"learning_rate": 1.5201219512195122e-05,
"loss": 0.7326,
"step": 57600
},
{
"epoch": 8.05,
"learning_rate": 1.5093615494978478e-05,
"loss": 0.7387,
"step": 57700
},
{
"epoch": 8.06,
"learning_rate": 1.4986011477761834e-05,
"loss": 0.752,
"step": 57800
},
{
"epoch": 8.08,
"learning_rate": 1.4878407460545192e-05,
"loss": 0.7525,
"step": 57900
},
{
"epoch": 8.09,
"learning_rate": 1.4770803443328548e-05,
"loss": 0.7451,
"step": 58000
},
{
"epoch": 8.09,
"eval_loss": 0.13486774265766144,
"eval_runtime": 805.7495,
"eval_samples_per_second": 19.162,
"eval_steps_per_second": 2.395,
"eval_wer": 0.1301918872965752,
"step": 58000
},
{
"epoch": 8.1,
"learning_rate": 1.4663199426111906e-05,
"loss": 0.7433,
"step": 58100
},
{
"epoch": 8.12,
"learning_rate": 1.4555595408895263e-05,
"loss": 0.7376,
"step": 58200
},
{
"epoch": 8.13,
"learning_rate": 1.4447991391678621e-05,
"loss": 0.7458,
"step": 58300
},
{
"epoch": 8.15,
"learning_rate": 1.4340387374461979e-05,
"loss": 0.7462,
"step": 58400
},
{
"epoch": 8.16,
"learning_rate": 1.4232783357245335e-05,
"loss": 0.7383,
"step": 58500
},
{
"epoch": 8.16,
"eval_loss": 0.13493329286575317,
"eval_runtime": 809.3737,
"eval_samples_per_second": 19.076,
"eval_steps_per_second": 2.385,
"eval_wer": 0.1294106834548904,
"step": 58500
},
{
"epoch": 8.17,
"learning_rate": 1.4125179340028692e-05,
"loss": 0.7496,
"step": 58600
},
{
"epoch": 8.19,
"learning_rate": 1.401757532281205e-05,
"loss": 0.7477,
"step": 58700
},
{
"epoch": 8.2,
"learning_rate": 1.3911047345767575e-05,
"loss": 0.7267,
"step": 58800
},
{
"epoch": 8.21,
"learning_rate": 1.3803443328550933e-05,
"loss": 0.7356,
"step": 58900
},
{
"epoch": 8.23,
"learning_rate": 1.3695839311334289e-05,
"loss": 0.7482,
"step": 59000
},
{
"epoch": 8.23,
"eval_loss": 0.13410241901874542,
"eval_runtime": 812.7329,
"eval_samples_per_second": 18.998,
"eval_steps_per_second": 2.375,
"eval_wer": 0.12928595342974744,
"step": 59000
},
{
"epoch": 8.24,
"learning_rate": 1.3588235294117646e-05,
"loss": 0.7469,
"step": 59100
},
{
"epoch": 8.26,
"learning_rate": 1.348170731707317e-05,
"loss": 0.7443,
"step": 59200
},
{
"epoch": 8.27,
"learning_rate": 1.3374103299856527e-05,
"loss": 0.7337,
"step": 59300
},
{
"epoch": 8.28,
"learning_rate": 1.3266499282639883e-05,
"loss": 0.7433,
"step": 59400
},
{
"epoch": 8.3,
"learning_rate": 1.315889526542324e-05,
"loss": 0.742,
"step": 59500
},
{
"epoch": 8.3,
"eval_loss": 0.13384965062141418,
"eval_runtime": 809.2173,
"eval_samples_per_second": 19.08,
"eval_steps_per_second": 2.385,
"eval_wer": 0.12963388455251462,
"step": 59500
},
{
"epoch": 8.31,
"learning_rate": 1.3051291248206598e-05,
"loss": 0.7417,
"step": 59600
},
{
"epoch": 8.33,
"learning_rate": 1.2943687230989956e-05,
"loss": 0.7331,
"step": 59700
},
{
"epoch": 8.34,
"learning_rate": 1.2836083213773312e-05,
"loss": 0.7503,
"step": 59800
},
{
"epoch": 8.35,
"learning_rate": 1.272847919655667e-05,
"loss": 0.7415,
"step": 59900
},
{
"epoch": 8.37,
"learning_rate": 1.2620875179340027e-05,
"loss": 0.7343,
"step": 60000
},
{
"epoch": 8.37,
"eval_loss": 0.13476364314556122,
"eval_runtime": 804.2449,
"eval_samples_per_second": 19.198,
"eval_steps_per_second": 2.4,
"eval_wer": 0.13065141896815446,
"step": 60000
},
{
"epoch": 8.38,
"learning_rate": 1.2513271162123385e-05,
"loss": 0.7448,
"step": 60100
},
{
"epoch": 8.4,
"learning_rate": 1.2405667144906741e-05,
"loss": 0.7515,
"step": 60200
},
{
"epoch": 8.41,
"learning_rate": 1.2298063127690099e-05,
"loss": 0.7417,
"step": 60300
},
{
"epoch": 8.42,
"learning_rate": 1.2190459110473456e-05,
"loss": 0.7348,
"step": 60400
},
{
"epoch": 8.44,
"learning_rate": 1.2082855093256814e-05,
"loss": 0.7385,
"step": 60500
},
{
"epoch": 8.44,
"eval_loss": 0.13243480026721954,
"eval_runtime": 804.2456,
"eval_samples_per_second": 19.198,
"eval_steps_per_second": 2.4,
"eval_wer": 0.12820933637061885,
"step": 60500
},
{
"epoch": 8.45,
"learning_rate": 1.1975251076040172e-05,
"loss": 0.7419,
"step": 60600
},
{
"epoch": 8.47,
"learning_rate": 1.1867647058823528e-05,
"loss": 0.7426,
"step": 60700
},
{
"epoch": 8.48,
"learning_rate": 1.1760043041606885e-05,
"loss": 0.7422,
"step": 60800
},
{
"epoch": 8.49,
"learning_rate": 1.1652439024390243e-05,
"loss": 0.7386,
"step": 60900
},
{
"epoch": 8.51,
"learning_rate": 1.15448350071736e-05,
"loss": 0.7567,
"step": 61000
},
{
"epoch": 8.51,
"eval_loss": 0.1333577036857605,
"eval_runtime": 804.5993,
"eval_samples_per_second": 19.19,
"eval_steps_per_second": 2.399,
"eval_wer": 0.12808460634547592,
"step": 61000
},
{
"epoch": 8.52,
"learning_rate": 1.1437230989956957e-05,
"loss": 0.7259,
"step": 61100
},
{
"epoch": 8.54,
"learning_rate": 1.1329626972740314e-05,
"loss": 0.7452,
"step": 61200
},
{
"epoch": 8.55,
"learning_rate": 1.1222022955523672e-05,
"loss": 0.7302,
"step": 61300
},
{
"epoch": 8.56,
"learning_rate": 1.111441893830703e-05,
"loss": 0.7325,
"step": 61400
},
{
"epoch": 8.58,
"learning_rate": 1.1006814921090387e-05,
"loss": 0.7342,
"step": 61500
},
{
"epoch": 8.58,
"eval_loss": 0.13381603360176086,
"eval_runtime": 806.0918,
"eval_samples_per_second": 19.154,
"eval_steps_per_second": 2.394,
"eval_wer": 0.12890519861615318,
"step": 61500
},
{
"epoch": 8.59,
"learning_rate": 1.0899210903873743e-05,
"loss": 0.7546,
"step": 61600
},
{
"epoch": 8.61,
"learning_rate": 1.0791606886657101e-05,
"loss": 0.7306,
"step": 61700
},
{
"epoch": 8.62,
"learning_rate": 1.0684002869440459e-05,
"loss": 0.7295,
"step": 61800
},
{
"epoch": 8.63,
"learning_rate": 1.0576398852223816e-05,
"loss": 0.7411,
"step": 61900
},
{
"epoch": 8.65,
"learning_rate": 1.0468794835007172e-05,
"loss": 0.7401,
"step": 62000
},
{
"epoch": 8.65,
"eval_loss": 0.13308760523796082,
"eval_runtime": 808.6331,
"eval_samples_per_second": 19.094,
"eval_steps_per_second": 2.387,
"eval_wer": 0.12849162011173185,
"step": 62000
},
{
"epoch": 8.66,
"learning_rate": 1.036119081779053e-05,
"loss": 0.722,
"step": 62100
},
{
"epoch": 8.68,
"learning_rate": 1.0253586800573888e-05,
"loss": 0.7385,
"step": 62200
},
{
"epoch": 8.69,
"learning_rate": 1.0145982783357245e-05,
"loss": 0.7333,
"step": 62300
},
{
"epoch": 8.7,
"learning_rate": 1.0038378766140601e-05,
"loss": 0.7312,
"step": 62400
},
{
"epoch": 8.72,
"learning_rate": 9.930774748923959e-06,
"loss": 0.7362,
"step": 62500
},
{
"epoch": 8.72,
"eval_loss": 0.13290312886238098,
"eval_runtime": 812.3851,
"eval_samples_per_second": 19.006,
"eval_steps_per_second": 2.376,
"eval_wer": 0.1283012427049347,
"step": 62500
},
{
"epoch": 8.73,
"learning_rate": 9.823170731707317e-06,
"loss": 0.7429,
"step": 62600
},
{
"epoch": 8.74,
"learning_rate": 9.715566714490674e-06,
"loss": 0.7538,
"step": 62700
},
{
"epoch": 8.76,
"learning_rate": 9.607962697274032e-06,
"loss": 0.7421,
"step": 62800
},
{
"epoch": 8.77,
"learning_rate": 9.500358680057388e-06,
"loss": 0.7367,
"step": 62900
},
{
"epoch": 8.79,
"learning_rate": 9.392754662840746e-06,
"loss": 0.7241,
"step": 63000
},
{
"epoch": 8.79,
"eval_loss": 0.13225644826889038,
"eval_runtime": 811.1164,
"eval_samples_per_second": 19.035,
"eval_steps_per_second": 2.379,
"eval_wer": 0.12769072205555082,
"step": 63000
},
{
"epoch": 8.8,
"learning_rate": 9.285150645624102e-06,
"loss": 0.7423,
"step": 63100
},
{
"epoch": 8.81,
"learning_rate": 9.17754662840746e-06,
"loss": 0.7327,
"step": 63200
},
{
"epoch": 8.83,
"learning_rate": 9.071018651362982e-06,
"loss": 0.734,
"step": 63300
},
{
"epoch": 8.84,
"learning_rate": 8.96341463414634e-06,
"loss": 0.7229,
"step": 63400
},
{
"epoch": 8.86,
"learning_rate": 8.855810616929698e-06,
"loss": 0.7244,
"step": 63500
},
{
"epoch": 8.86,
"eval_loss": 0.13174521923065186,
"eval_runtime": 804.0262,
"eval_samples_per_second": 19.203,
"eval_steps_per_second": 2.4,
"eval_wer": 0.1268963887375352,
"step": 63500
},
{
"epoch": 8.87,
"learning_rate": 8.748206599713055e-06,
"loss": 0.7311,
"step": 63600
},
{
"epoch": 8.88,
"learning_rate": 8.640602582496411e-06,
"loss": 0.7328,
"step": 63700
},
{
"epoch": 8.9,
"learning_rate": 8.532998565279769e-06,
"loss": 0.7224,
"step": 63800
},
{
"epoch": 8.91,
"learning_rate": 8.425394548063127e-06,
"loss": 0.7382,
"step": 63900
},
{
"epoch": 8.93,
"learning_rate": 8.317790530846484e-06,
"loss": 0.7274,
"step": 64000
},
{
"epoch": 8.93,
"eval_loss": 0.1307557225227356,
"eval_runtime": 801.064,
"eval_samples_per_second": 19.274,
"eval_steps_per_second": 2.409,
"eval_wer": 0.1259641959180458,
"step": 64000
},
{
"epoch": 8.94,
"learning_rate": 8.210186513629842e-06,
"loss": 0.7299,
"step": 64100
},
{
"epoch": 8.95,
"learning_rate": 8.102582496413198e-06,
"loss": 0.7183,
"step": 64200
},
{
"epoch": 8.97,
"learning_rate": 7.994978479196556e-06,
"loss": 0.7378,
"step": 64300
},
{
"epoch": 8.98,
"learning_rate": 7.887374461979913e-06,
"loss": 0.73,
"step": 64400
},
{
"epoch": 9.0,
"learning_rate": 7.779770444763271e-06,
"loss": 0.7411,
"step": 64500
},
{
"epoch": 9.0,
"eval_loss": 0.13092704117298126,
"eval_runtime": 805.3626,
"eval_samples_per_second": 19.171,
"eval_steps_per_second": 2.396,
"eval_wer": 0.12557687636628614,
"step": 64500
},
{
"epoch": 9.01,
"learning_rate": 7.672166427546627e-06,
"loss": 0.7213,
"step": 64600
},
{
"epoch": 9.02,
"learning_rate": 7.564562410329985e-06,
"loss": 0.7277,
"step": 64700
},
{
"epoch": 9.04,
"learning_rate": 7.456958393113342e-06,
"loss": 0.7331,
"step": 64800
},
{
"epoch": 9.05,
"learning_rate": 7.349354375896699e-06,
"loss": 0.7252,
"step": 64900
},
{
"epoch": 9.07,
"learning_rate": 7.241750358680057e-06,
"loss": 0.7255,
"step": 65000
},
{
"epoch": 9.07,
"eval_loss": 0.13162025809288025,
"eval_runtime": 804.768,
"eval_samples_per_second": 19.186,
"eval_steps_per_second": 2.398,
"eval_wer": 0.12646968075678303,
"step": 65000
},
{
"epoch": 9.08,
"learning_rate": 7.134146341463414e-06,
"loss": 0.716,
"step": 65100
},
{
"epoch": 9.09,
"learning_rate": 7.026542324246771e-06,
"loss": 0.7248,
"step": 65200
},
{
"epoch": 9.11,
"learning_rate": 6.918938307030129e-06,
"loss": 0.7194,
"step": 65300
},
{
"epoch": 9.12,
"learning_rate": 6.811334289813486e-06,
"loss": 0.7221,
"step": 65400
},
{
"epoch": 9.14,
"learning_rate": 6.7037302725968435e-06,
"loss": 0.7406,
"step": 65500
},
{
"epoch": 9.14,
"eval_loss": 0.13147445023059845,
"eval_runtime": 803.4274,
"eval_samples_per_second": 19.218,
"eval_steps_per_second": 2.402,
"eval_wer": 0.12702111876267816,
"step": 65500
},
{
"epoch": 9.15,
"learning_rate": 6.5961262553802e-06,
"loss": 0.7439,
"step": 65600
},
{
"epoch": 9.16,
"learning_rate": 6.488522238163558e-06,
"loss": 0.7451,
"step": 65700
},
{
"epoch": 9.18,
"learning_rate": 6.380918220946915e-06,
"loss": 0.7286,
"step": 65800
},
{
"epoch": 9.19,
"learning_rate": 6.2733142037302725e-06,
"loss": 0.7307,
"step": 65900
},
{
"epoch": 9.21,
"learning_rate": 6.166786226685796e-06,
"loss": 0.7418,
"step": 66000
},
{
"epoch": 9.21,
"eval_loss": 0.13148032128810883,
"eval_runtime": 801.1993,
"eval_samples_per_second": 19.271,
"eval_steps_per_second": 2.409,
"eval_wer": 0.12692264769019687,
"step": 66000
},
{
"epoch": 9.22,
"learning_rate": 6.0591822094691534e-06,
"loss": 0.716,
"step": 66100
},
{
"epoch": 9.23,
"learning_rate": 5.951578192252511e-06,
"loss": 0.7214,
"step": 66200
},
{
"epoch": 9.25,
"learning_rate": 5.843974175035868e-06,
"loss": 0.7255,
"step": 66300
},
{
"epoch": 9.26,
"learning_rate": 5.736370157819226e-06,
"loss": 0.7193,
"step": 66400
},
{
"epoch": 9.27,
"learning_rate": 5.6287661406025816e-06,
"loss": 0.7301,
"step": 66500
},
{
"epoch": 9.27,
"eval_loss": 0.13154751062393188,
"eval_runtime": 802.8155,
"eval_samples_per_second": 19.232,
"eval_steps_per_second": 2.404,
"eval_wer": 0.12730340250379113,
"step": 66500
},
{
"epoch": 9.29,
"learning_rate": 5.521162123385938e-06,
"loss": 0.7266,
"step": 66600
},
{
"epoch": 9.3,
"learning_rate": 5.413558106169296e-06,
"loss": 0.7423,
"step": 66700
},
{
"epoch": 9.32,
"learning_rate": 5.305954088952654e-06,
"loss": 0.7215,
"step": 66800
},
{
"epoch": 9.33,
"learning_rate": 5.1983500717360105e-06,
"loss": 0.7283,
"step": 66900
},
{
"epoch": 9.34,
"learning_rate": 5.090746054519368e-06,
"loss": 0.7248,
"step": 67000
},
{
"epoch": 9.34,
"eval_loss": 0.13229568302631378,
"eval_runtime": 806.5335,
"eval_samples_per_second": 19.144,
"eval_steps_per_second": 2.393,
"eval_wer": 0.12737561462361074,
"step": 67000
},
{
"epoch": 9.36,
"learning_rate": 4.983142037302725e-06,
"loss": 0.7317,
"step": 67100
},
{
"epoch": 9.37,
"learning_rate": 4.875538020086083e-06,
"loss": 0.7348,
"step": 67200
},
{
"epoch": 9.39,
"learning_rate": 4.7679340028694395e-06,
"loss": 0.7416,
"step": 67300
},
{
"epoch": 9.4,
"learning_rate": 4.660329985652797e-06,
"loss": 0.7261,
"step": 67400
},
{
"epoch": 9.41,
"learning_rate": 4.552725968436154e-06,
"loss": 0.7423,
"step": 67500
},
{
"epoch": 9.41,
"eval_loss": 0.13094688951969147,
"eval_runtime": 816.5792,
"eval_samples_per_second": 18.908,
"eval_steps_per_second": 2.364,
"eval_wer": 0.12666005816358014,
"step": 67500
},
{
"epoch": 9.43,
"learning_rate": 4.445121951219512e-06,
"loss": 0.7408,
"step": 67600
},
{
"epoch": 9.44,
"learning_rate": 4.338593974175036e-06,
"loss": 0.717,
"step": 67700
},
{
"epoch": 9.46,
"learning_rate": 4.230989956958393e-06,
"loss": 0.7278,
"step": 67800
},
{
"epoch": 9.47,
"learning_rate": 4.12338593974175e-06,
"loss": 0.7272,
"step": 67900
},
{
"epoch": 9.48,
"learning_rate": 4.015781922525107e-06,
"loss": 0.7152,
"step": 68000
},
{
"epoch": 9.48,
"eval_loss": 0.13121984899044037,
"eval_runtime": 805.1304,
"eval_samples_per_second": 19.177,
"eval_steps_per_second": 2.397,
"eval_wer": 0.12709333088249775,
"step": 68000
},
{
"epoch": 9.5,
"learning_rate": 3.908177905308465e-06,
"loss": 0.7371,
"step": 68100
},
{
"epoch": 9.51,
"learning_rate": 3.8005738880918216e-06,
"loss": 0.7389,
"step": 68200
},
{
"epoch": 9.53,
"learning_rate": 3.6929698708751792e-06,
"loss": 0.7443,
"step": 68300
},
{
"epoch": 9.54,
"learning_rate": 3.5853658536585365e-06,
"loss": 0.7277,
"step": 68400
},
{
"epoch": 9.55,
"learning_rate": 3.4777618364418937e-06,
"loss": 0.7295,
"step": 68500
},
{
"epoch": 9.55,
"eval_loss": 0.13064497709274292,
"eval_runtime": 808.6417,
"eval_samples_per_second": 19.094,
"eval_steps_per_second": 2.387,
"eval_wer": 0.1261742675393392,
"step": 68500
},
{
"epoch": 9.57,
"learning_rate": 3.370157819225251e-06,
"loss": 0.7175,
"step": 68600
},
{
"epoch": 9.58,
"learning_rate": 3.2625538020086082e-06,
"loss": 0.7126,
"step": 68700
},
{
"epoch": 9.6,
"learning_rate": 3.1549497847919655e-06,
"loss": 0.7339,
"step": 68800
},
{
"epoch": 9.61,
"learning_rate": 3.0473457675753227e-06,
"loss": 0.7205,
"step": 68900
},
{
"epoch": 9.62,
"learning_rate": 2.93974175035868e-06,
"loss": 0.7231,
"step": 69000
},
{
"epoch": 9.62,
"eval_loss": 0.13077928125858307,
"eval_runtime": 804.8055,
"eval_samples_per_second": 19.185,
"eval_steps_per_second": 2.398,
"eval_wer": 0.12634495073164007,
"step": 69000
},
{
"epoch": 9.64,
"learning_rate": 2.832137733142037e-06,
"loss": 0.7401,
"step": 69100
},
{
"epoch": 9.65,
"learning_rate": 2.724533715925394e-06,
"loss": 0.7268,
"step": 69200
},
{
"epoch": 9.67,
"learning_rate": 2.6169296987087513e-06,
"loss": 0.7164,
"step": 69300
},
{
"epoch": 9.68,
"learning_rate": 2.5093256814921085e-06,
"loss": 0.7316,
"step": 69400
},
{
"epoch": 9.69,
"learning_rate": 2.4017216642754657e-06,
"loss": 0.7344,
"step": 69500
},
{
"epoch": 9.69,
"eval_loss": 0.13128259778022766,
"eval_runtime": 805.2253,
"eval_samples_per_second": 19.175,
"eval_steps_per_second": 2.397,
"eval_wer": 0.1266797523780764,
"step": 69500
},
{
"epoch": 9.71,
"learning_rate": 2.2941176470588234e-06,
"loss": 0.7165,
"step": 69600
},
{
"epoch": 9.72,
"learning_rate": 2.1865136298421807e-06,
"loss": 0.7431,
"step": 69700
},
{
"epoch": 9.74,
"learning_rate": 2.078909612625538e-06,
"loss": 0.7128,
"step": 69800
},
{
"epoch": 9.75,
"learning_rate": 1.971305595408895e-06,
"loss": 0.7319,
"step": 69900
},
{
"epoch": 9.76,
"learning_rate": 1.8637015781922524e-06,
"loss": 0.7264,
"step": 70000
},
{
"epoch": 9.76,
"eval_loss": 0.13052190840244293,
"eval_runtime": 835.8187,
"eval_samples_per_second": 18.473,
"eval_steps_per_second": 2.309,
"eval_wer": 0.1263186917789784,
"step": 70000
},
{
"epoch": 9.78,
"learning_rate": 1.7560975609756096e-06,
"loss": 0.7184,
"step": 70100
},
{
"epoch": 9.79,
"learning_rate": 1.6484935437589669e-06,
"loss": 0.7312,
"step": 70200
},
{
"epoch": 9.8,
"learning_rate": 1.5408895265423241e-06,
"loss": 0.7083,
"step": 70300
},
{
"epoch": 9.82,
"learning_rate": 1.4332855093256814e-06,
"loss": 0.7293,
"step": 70400
},
{
"epoch": 9.83,
"learning_rate": 1.3256814921090388e-06,
"loss": 0.7309,
"step": 70500
},
{
"epoch": 9.83,
"eval_loss": 0.13029050827026367,
"eval_runtime": 805.5283,
"eval_samples_per_second": 19.168,
"eval_steps_per_second": 2.396,
"eval_wer": 0.1262136559683317,
"step": 70500
},
{
"epoch": 9.85,
"learning_rate": 1.2180774748923956e-06,
"loss": 0.7254,
"step": 70600
},
{
"epoch": 9.86,
"learning_rate": 1.110473457675753e-06,
"loss": 0.73,
"step": 70700
},
{
"epoch": 9.87,
"learning_rate": 1.0028694404591103e-06,
"loss": 0.7331,
"step": 70800
},
{
"epoch": 9.89,
"learning_rate": 8.952654232424676e-07,
"loss": 0.7285,
"step": 70900
},
{
"epoch": 9.9,
"learning_rate": 7.876614060258248e-07,
"loss": 0.73,
"step": 71000
},
{
"epoch": 9.9,
"eval_loss": 0.13025349378585815,
"eval_runtime": 810.0942,
"eval_samples_per_second": 19.06,
"eval_steps_per_second": 2.382,
"eval_wer": 0.12609549068135417,
"step": 71000
},
{
"epoch": 9.92,
"learning_rate": 6.800573888091822e-07,
"loss": 0.7271,
"step": 71100
},
{
"epoch": 9.93,
"learning_rate": 5.724533715925393e-07,
"loss": 0.7131,
"step": 71200
},
{
"epoch": 9.94,
"learning_rate": 4.648493543758967e-07,
"loss": 0.7218,
"step": 71300
},
{
"epoch": 9.96,
"learning_rate": 3.572453371592539e-07,
"loss": 0.7226,
"step": 71400
},
{
"epoch": 9.97,
"learning_rate": 2.4964131994261116e-07,
"loss": 0.7353,
"step": 71500
},
{
"epoch": 9.97,
"eval_loss": 0.13038034737110138,
"eval_runtime": 826.6805,
"eval_samples_per_second": 18.677,
"eval_steps_per_second": 2.335,
"eval_wer": 0.12604297277603083,
"step": 71500
},
{
"epoch": 9.99,
"learning_rate": 1.420373027259684e-07,
"loss": 0.7333,
"step": 71600
},
{
"epoch": 10.0,
"learning_rate": 3.443328550932568e-08,
"loss": 0.7394,
"step": 71700
},
{
"epoch": 10.0,
"step": 71700,
"total_flos": 3.7403945946937406e+20,
"train_loss": 0.8738376914307662,
"train_runtime": 268775.0566,
"train_samples_per_second": 8.537,
"train_steps_per_second": 0.267
}
],
"max_steps": 71700,
"num_train_epochs": 10,
"total_flos": 3.7403945946937406e+20,
"trial_name": null,
"trial_params": null
}