makhataei's picture
Training in progress, step 600
2ecb144
raw
history blame
72.3 kB
{
"best_metric": 43.17606786724215,
"best_model_checkpoint": "/media/makhataei/Backups/Whisper-Small-Common-Voice/checkpoint-300",
"epoch": 38.86162904808636,
"eval_steps": 100,
"global_step": 9900,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"learning_rate": 5e-06,
"loss": 0.3008,
"step": 25
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 0.2368,
"step": 50
},
{
"epoch": 0.29,
"learning_rate": 9.974874371859297e-06,
"loss": 0.2001,
"step": 75
},
{
"epoch": 0.39,
"learning_rate": 9.949748743718594e-06,
"loss": 0.1801,
"step": 100
},
{
"epoch": 0.39,
"eval_loss": 0.4975946843624115,
"eval_runtime": 1546.7344,
"eval_samples_per_second": 5.59,
"eval_steps_per_second": 0.559,
"eval_wer": 49.12602735195383,
"step": 100
},
{
"epoch": 0.49,
"learning_rate": 9.92462311557789e-06,
"loss": 0.1713,
"step": 125
},
{
"epoch": 0.59,
"learning_rate": 9.899497487437186e-06,
"loss": 0.1657,
"step": 150
},
{
"epoch": 0.69,
"learning_rate": 9.874371859296484e-06,
"loss": 0.1574,
"step": 175
},
{
"epoch": 0.79,
"learning_rate": 9.84924623115578e-06,
"loss": 0.1597,
"step": 200
},
{
"epoch": 0.79,
"eval_loss": 0.46244511008262634,
"eval_runtime": 1507.9912,
"eval_samples_per_second": 5.733,
"eval_steps_per_second": 0.574,
"eval_wer": 46.749681665591766,
"step": 200
},
{
"epoch": 0.88,
"learning_rate": 9.824120603015075e-06,
"loss": 0.1509,
"step": 225
},
{
"epoch": 0.98,
"learning_rate": 9.798994974874372e-06,
"loss": 0.1434,
"step": 250
},
{
"epoch": 1.08,
"learning_rate": 9.773869346733669e-06,
"loss": 0.0919,
"step": 275
},
{
"epoch": 1.18,
"learning_rate": 9.748743718592965e-06,
"loss": 0.0776,
"step": 300
},
{
"epoch": 1.18,
"eval_loss": 0.4794461131095886,
"eval_runtime": 1499.4378,
"eval_samples_per_second": 5.766,
"eval_steps_per_second": 0.577,
"eval_wer": 43.17606786724215,
"step": 300
},
{
"epoch": 1.28,
"learning_rate": 9.723618090452262e-06,
"loss": 0.0772,
"step": 325
},
{
"epoch": 1.37,
"learning_rate": 9.698492462311559e-06,
"loss": 0.0799,
"step": 350
},
{
"epoch": 1.47,
"learning_rate": 9.673366834170855e-06,
"loss": 0.0872,
"step": 375
},
{
"epoch": 1.57,
"learning_rate": 9.648241206030152e-06,
"loss": 0.083,
"step": 400
},
{
"epoch": 1.57,
"eval_loss": 0.48234912753105164,
"eval_runtime": 1500.1085,
"eval_samples_per_second": 5.764,
"eval_steps_per_second": 0.577,
"eval_wer": 43.802814572274315,
"step": 400
},
{
"epoch": 1.67,
"learning_rate": 9.623115577889448e-06,
"loss": 0.0807,
"step": 425
},
{
"epoch": 1.77,
"learning_rate": 9.597989949748745e-06,
"loss": 0.0815,
"step": 450
},
{
"epoch": 1.86,
"learning_rate": 9.57286432160804e-06,
"loss": 0.0803,
"step": 475
},
{
"epoch": 1.96,
"learning_rate": 9.547738693467338e-06,
"loss": 0.0786,
"step": 500
},
{
"epoch": 1.96,
"eval_loss": 0.48833438754081726,
"eval_runtime": 1502.1848,
"eval_samples_per_second": 5.756,
"eval_steps_per_second": 0.576,
"eval_wer": 44.39152651684278,
"step": 500
},
{
"epoch": 2.06,
"learning_rate": 9.522613065326633e-06,
"loss": 0.0538,
"step": 525
},
{
"epoch": 2.16,
"learning_rate": 9.49748743718593e-06,
"loss": 0.0338,
"step": 550
},
{
"epoch": 2.26,
"learning_rate": 9.472361809045228e-06,
"loss": 0.0315,
"step": 575
},
{
"epoch": 2.36,
"learning_rate": 9.447236180904523e-06,
"loss": 0.0331,
"step": 600
},
{
"epoch": 2.36,
"eval_loss": 0.538537859916687,
"eval_runtime": 1606.4592,
"eval_samples_per_second": 5.382,
"eval_steps_per_second": 0.538,
"eval_wer": 46.243653982900895,
"step": 600
},
{
"epoch": 2.45,
"learning_rate": 9.42211055276382e-06,
"loss": 0.0347,
"step": 625
},
{
"epoch": 2.55,
"learning_rate": 9.396984924623116e-06,
"loss": 0.0319,
"step": 650
},
{
"epoch": 2.65,
"learning_rate": 9.371859296482413e-06,
"loss": 0.035,
"step": 675
},
{
"epoch": 2.75,
"learning_rate": 9.34673366834171e-06,
"loss": 0.0353,
"step": 700
},
{
"epoch": 2.75,
"eval_loss": 0.5605480670928955,
"eval_runtime": 1551.3357,
"eval_samples_per_second": 5.573,
"eval_steps_per_second": 0.558,
"eval_wer": 44.943857386185115,
"step": 700
},
{
"epoch": 2.85,
"learning_rate": 9.321608040201006e-06,
"loss": 0.0393,
"step": 725
},
{
"epoch": 2.94,
"learning_rate": 9.296482412060303e-06,
"loss": 0.0362,
"step": 750
},
{
"epoch": 3.04,
"learning_rate": 9.271356783919598e-06,
"loss": 0.0265,
"step": 775
},
{
"epoch": 3.14,
"learning_rate": 9.246231155778896e-06,
"loss": 0.0139,
"step": 800
},
{
"epoch": 3.14,
"eval_loss": 0.5941358804702759,
"eval_runtime": 1606.3004,
"eval_samples_per_second": 5.383,
"eval_steps_per_second": 0.539,
"eval_wer": 45.281209174645696,
"step": 800
},
{
"epoch": 3.24,
"learning_rate": 9.221105527638193e-06,
"loss": 0.0161,
"step": 825
},
{
"epoch": 3.34,
"learning_rate": 9.195979899497488e-06,
"loss": 0.0148,
"step": 850
},
{
"epoch": 3.43,
"learning_rate": 9.170854271356784e-06,
"loss": 0.0169,
"step": 875
},
{
"epoch": 3.53,
"learning_rate": 9.14572864321608e-06,
"loss": 0.0152,
"step": 900
},
{
"epoch": 3.53,
"eval_loss": 0.5978226065635681,
"eval_runtime": 1599.7307,
"eval_samples_per_second": 5.405,
"eval_steps_per_second": 0.541,
"eval_wer": 49.293049560946564,
"step": 900
},
{
"epoch": 3.63,
"learning_rate": 9.120603015075377e-06,
"loss": 0.0148,
"step": 925
},
{
"epoch": 3.73,
"learning_rate": 9.095477386934674e-06,
"loss": 0.0154,
"step": 950
},
{
"epoch": 3.83,
"learning_rate": 9.07035175879397e-06,
"loss": 0.0178,
"step": 975
},
{
"epoch": 3.93,
"learning_rate": 9.045226130653267e-06,
"loss": 0.0155,
"step": 1000
},
{
"epoch": 3.93,
"eval_loss": 0.6114091873168945,
"eval_runtime": 1660.7592,
"eval_samples_per_second": 5.206,
"eval_steps_per_second": 0.521,
"eval_wer": 49.97767524929305,
"step": 1000
},
{
"epoch": 4.02,
"learning_rate": 9.020100502512564e-06,
"loss": 0.0135,
"step": 1025
},
{
"epoch": 4.12,
"learning_rate": 8.99497487437186e-06,
"loss": 0.0069,
"step": 1050
},
{
"epoch": 4.22,
"learning_rate": 8.969849246231157e-06,
"loss": 0.006,
"step": 1075
},
{
"epoch": 4.32,
"learning_rate": 8.944723618090452e-06,
"loss": 0.0063,
"step": 1100
},
{
"epoch": 4.32,
"eval_loss": 0.6466676592826843,
"eval_runtime": 1663.0802,
"eval_samples_per_second": 5.199,
"eval_steps_per_second": 0.52,
"eval_wer": 50.00413421309388,
"step": 1100
},
{
"epoch": 4.42,
"learning_rate": 8.91959798994975e-06,
"loss": 0.0077,
"step": 1125
},
{
"epoch": 4.51,
"learning_rate": 8.894472361809045e-06,
"loss": 0.0063,
"step": 1150
},
{
"epoch": 4.61,
"learning_rate": 8.869346733668342e-06,
"loss": 0.0087,
"step": 1175
},
{
"epoch": 4.71,
"learning_rate": 8.84422110552764e-06,
"loss": 0.0079,
"step": 1200
},
{
"epoch": 4.71,
"eval_loss": 0.6383361220359802,
"eval_runtime": 1613.3597,
"eval_samples_per_second": 5.359,
"eval_steps_per_second": 0.536,
"eval_wer": 48.08751302277125,
"step": 1200
},
{
"epoch": 4.81,
"learning_rate": 8.819095477386935e-06,
"loss": 0.0081,
"step": 1225
},
{
"epoch": 4.91,
"learning_rate": 8.793969849246232e-06,
"loss": 0.0085,
"step": 1250
},
{
"epoch": 5.0,
"learning_rate": 8.768844221105528e-06,
"loss": 0.0094,
"step": 1275
},
{
"epoch": 5.1,
"learning_rate": 8.743718592964825e-06,
"loss": 0.0046,
"step": 1300
},
{
"epoch": 5.1,
"eval_loss": 0.6500093340873718,
"eval_runtime": 1575.0065,
"eval_samples_per_second": 5.49,
"eval_steps_per_second": 0.549,
"eval_wer": 45.49949562600254,
"step": 1300
},
{
"epoch": 5.2,
"learning_rate": 8.71859296482412e-06,
"loss": 0.0046,
"step": 1325
},
{
"epoch": 5.3,
"learning_rate": 8.693467336683418e-06,
"loss": 0.0058,
"step": 1350
},
{
"epoch": 5.4,
"learning_rate": 8.668341708542715e-06,
"loss": 0.0055,
"step": 1375
},
{
"epoch": 5.5,
"learning_rate": 8.64321608040201e-06,
"loss": 0.0042,
"step": 1400
},
{
"epoch": 5.5,
"eval_loss": 0.647581934928894,
"eval_runtime": 1605.9286,
"eval_samples_per_second": 5.384,
"eval_steps_per_second": 0.539,
"eval_wer": 47.44919052107622,
"step": 1400
},
{
"epoch": 5.59,
"learning_rate": 8.618090452261308e-06,
"loss": 0.0047,
"step": 1425
},
{
"epoch": 5.69,
"learning_rate": 8.592964824120603e-06,
"loss": 0.0049,
"step": 1450
},
{
"epoch": 5.79,
"learning_rate": 8.5678391959799e-06,
"loss": 0.0048,
"step": 1475
},
{
"epoch": 5.89,
"learning_rate": 8.542713567839196e-06,
"loss": 0.0052,
"step": 1500
},
{
"epoch": 5.89,
"eval_loss": 0.668513298034668,
"eval_runtime": 1668.8488,
"eval_samples_per_second": 5.181,
"eval_steps_per_second": 0.518,
"eval_wer": 52.18699872666237,
"step": 1500
},
{
"epoch": 5.99,
"learning_rate": 8.517587939698493e-06,
"loss": 0.006,
"step": 1525
},
{
"epoch": 6.08,
"learning_rate": 8.49246231155779e-06,
"loss": 0.0052,
"step": 1550
},
{
"epoch": 6.18,
"learning_rate": 8.467336683417086e-06,
"loss": 0.0038,
"step": 1575
},
{
"epoch": 6.28,
"learning_rate": 8.442211055276383e-06,
"loss": 0.0023,
"step": 1600
},
{
"epoch": 6.28,
"eval_loss": 0.6793932318687439,
"eval_runtime": 1564.2392,
"eval_samples_per_second": 5.527,
"eval_steps_per_second": 0.553,
"eval_wer": 44.250963271650875,
"step": 1600
},
{
"epoch": 6.38,
"learning_rate": 8.41708542713568e-06,
"loss": 0.0025,
"step": 1625
},
{
"epoch": 6.48,
"learning_rate": 8.391959798994976e-06,
"loss": 0.0027,
"step": 1650
},
{
"epoch": 6.58,
"learning_rate": 8.366834170854273e-06,
"loss": 0.0034,
"step": 1675
},
{
"epoch": 6.67,
"learning_rate": 8.341708542713568e-06,
"loss": 0.0032,
"step": 1700
},
{
"epoch": 6.67,
"eval_loss": 0.6724051237106323,
"eval_runtime": 1596.3328,
"eval_samples_per_second": 5.416,
"eval_steps_per_second": 0.542,
"eval_wer": 45.71612839212184,
"step": 1700
},
{
"epoch": 6.77,
"learning_rate": 8.316582914572864e-06,
"loss": 0.0036,
"step": 1725
},
{
"epoch": 6.87,
"learning_rate": 8.291457286432163e-06,
"loss": 0.0029,
"step": 1750
},
{
"epoch": 6.97,
"learning_rate": 8.266331658291457e-06,
"loss": 0.0026,
"step": 1775
},
{
"epoch": 7.07,
"learning_rate": 8.241206030150754e-06,
"loss": 0.0021,
"step": 1800
},
{
"epoch": 7.07,
"eval_loss": 0.6820480227470398,
"eval_runtime": 1590.9785,
"eval_samples_per_second": 5.434,
"eval_steps_per_second": 0.544,
"eval_wer": 47.60132956293099,
"step": 1800
},
{
"epoch": 7.16,
"learning_rate": 8.21608040201005e-06,
"loss": 0.0025,
"step": 1825
},
{
"epoch": 7.26,
"learning_rate": 8.190954773869347e-06,
"loss": 0.0017,
"step": 1850
},
{
"epoch": 7.36,
"learning_rate": 8.165829145728644e-06,
"loss": 0.0015,
"step": 1875
},
{
"epoch": 7.46,
"learning_rate": 8.14070351758794e-06,
"loss": 0.0015,
"step": 1900
},
{
"epoch": 7.46,
"eval_loss": 0.6925253868103027,
"eval_runtime": 1624.9343,
"eval_samples_per_second": 5.321,
"eval_steps_per_second": 0.532,
"eval_wer": 46.67195845942683,
"step": 1900
},
{
"epoch": 7.56,
"learning_rate": 8.115577889447237e-06,
"loss": 0.0023,
"step": 1925
},
{
"epoch": 7.65,
"learning_rate": 8.090452261306532e-06,
"loss": 0.0014,
"step": 1950
},
{
"epoch": 7.75,
"learning_rate": 8.06532663316583e-06,
"loss": 0.002,
"step": 1975
},
{
"epoch": 7.85,
"learning_rate": 8.040201005025127e-06,
"loss": 0.0024,
"step": 2000
},
{
"epoch": 7.85,
"eval_loss": 0.710422158241272,
"eval_runtime": 1663.5693,
"eval_samples_per_second": 5.197,
"eval_steps_per_second": 0.52,
"eval_wer": 50.29022175919036,
"step": 2000
},
{
"epoch": 7.95,
"learning_rate": 8.015075376884422e-06,
"loss": 0.0027,
"step": 2025
},
{
"epoch": 8.05,
"learning_rate": 7.989949748743719e-06,
"loss": 0.0026,
"step": 2050
},
{
"epoch": 8.15,
"learning_rate": 7.965829145728644e-06,
"loss": 0.0029,
"step": 2075
},
{
"epoch": 8.24,
"learning_rate": 7.940703517587939e-06,
"loss": 0.0029,
"step": 2100
},
{
"epoch": 8.24,
"eval_loss": 0.6837413311004639,
"eval_runtime": 1582.0652,
"eval_samples_per_second": 5.465,
"eval_steps_per_second": 0.547,
"eval_wer": 46.41729093284384,
"step": 2100
},
{
"epoch": 8.34,
"learning_rate": 7.915577889447237e-06,
"loss": 0.0017,
"step": 2125
},
{
"epoch": 8.44,
"learning_rate": 7.890452261306534e-06,
"loss": 0.0015,
"step": 2150
},
{
"epoch": 8.54,
"learning_rate": 7.865326633165829e-06,
"loss": 0.0017,
"step": 2175
},
{
"epoch": 8.64,
"learning_rate": 7.840201005025127e-06,
"loss": 0.0016,
"step": 2200
},
{
"epoch": 8.64,
"eval_loss": 0.7190621495246887,
"eval_runtime": 1555.4401,
"eval_samples_per_second": 5.559,
"eval_steps_per_second": 0.556,
"eval_wer": 46.008830679168526,
"step": 2200
},
{
"epoch": 8.73,
"learning_rate": 7.815075376884422e-06,
"loss": 0.0013,
"step": 2225
},
{
"epoch": 8.83,
"learning_rate": 7.789949748743719e-06,
"loss": 0.0017,
"step": 2250
},
{
"epoch": 8.93,
"learning_rate": 7.764824120603015e-06,
"loss": 0.0018,
"step": 2275
},
{
"epoch": 9.03,
"learning_rate": 7.739698492462312e-06,
"loss": 0.0017,
"step": 2300
},
{
"epoch": 9.03,
"eval_loss": 0.7105358242988586,
"eval_runtime": 1601.1616,
"eval_samples_per_second": 5.4,
"eval_steps_per_second": 0.54,
"eval_wer": 47.59636850721834,
"step": 2300
},
{
"epoch": 9.13,
"learning_rate": 7.714572864321609e-06,
"loss": 0.0016,
"step": 2325
},
{
"epoch": 9.22,
"learning_rate": 7.689447236180905e-06,
"loss": 0.0015,
"step": 2350
},
{
"epoch": 9.32,
"learning_rate": 7.664321608040202e-06,
"loss": 0.0013,
"step": 2375
},
{
"epoch": 9.42,
"learning_rate": 7.639195979899499e-06,
"loss": 0.0014,
"step": 2400
},
{
"epoch": 9.42,
"eval_loss": 0.7293367981910706,
"eval_runtime": 1636.4789,
"eval_samples_per_second": 5.283,
"eval_steps_per_second": 0.529,
"eval_wer": 44.76029832481686,
"step": 2400
},
{
"epoch": 9.52,
"learning_rate": 7.614070351758794e-06,
"loss": 0.0015,
"step": 2425
},
{
"epoch": 9.62,
"learning_rate": 7.588944723618091e-06,
"loss": 0.0014,
"step": 2450
},
{
"epoch": 9.72,
"learning_rate": 7.563819095477387e-06,
"loss": 0.0012,
"step": 2475
},
{
"epoch": 9.81,
"learning_rate": 7.538693467336684e-06,
"loss": 0.0018,
"step": 2500
},
{
"epoch": 9.81,
"eval_loss": 0.7364875078201294,
"eval_runtime": 1682.5135,
"eval_samples_per_second": 5.139,
"eval_steps_per_second": 0.514,
"eval_wer": 49.896644672653004,
"step": 2500
},
{
"epoch": 9.91,
"learning_rate": 7.513567839195981e-06,
"loss": 0.0015,
"step": 2525
},
{
"epoch": 10.01,
"learning_rate": 7.488442211055277e-06,
"loss": 0.0016,
"step": 2550
},
{
"epoch": 10.11,
"learning_rate": 7.463316582914574e-06,
"loss": 0.0008,
"step": 2575
},
{
"epoch": 10.21,
"learning_rate": 7.43819095477387e-06,
"loss": 0.0008,
"step": 2600
},
{
"epoch": 10.21,
"eval_loss": 0.7378189563751221,
"eval_runtime": 1631.5066,
"eval_samples_per_second": 5.299,
"eval_steps_per_second": 0.53,
"eval_wer": 47.473995799639496,
"step": 2600
},
{
"epoch": 10.3,
"learning_rate": 7.4130653266331665e-06,
"loss": 0.0012,
"step": 2625
},
{
"epoch": 10.4,
"learning_rate": 7.387939698492463e-06,
"loss": 0.001,
"step": 2650
},
{
"epoch": 10.5,
"learning_rate": 7.362814070351759e-06,
"loss": 0.0024,
"step": 2675
},
{
"epoch": 10.6,
"learning_rate": 7.337688442211056e-06,
"loss": 0.0016,
"step": 2700
},
{
"epoch": 10.6,
"eval_loss": 0.7303053140640259,
"eval_runtime": 1617.5646,
"eval_samples_per_second": 5.345,
"eval_steps_per_second": 0.535,
"eval_wer": 45.96914223346728,
"step": 2700
},
{
"epoch": 10.7,
"learning_rate": 7.312562814070352e-06,
"loss": 0.0015,
"step": 2725
},
{
"epoch": 10.79,
"learning_rate": 7.287437185929649e-06,
"loss": 0.0013,
"step": 2750
},
{
"epoch": 10.89,
"learning_rate": 7.262311557788946e-06,
"loss": 0.0021,
"step": 2775
},
{
"epoch": 10.99,
"learning_rate": 7.237185929648242e-06,
"loss": 0.0011,
"step": 2800
},
{
"epoch": 10.99,
"eval_loss": 0.7329912781715393,
"eval_runtime": 1626.8373,
"eval_samples_per_second": 5.315,
"eval_steps_per_second": 0.532,
"eval_wer": 47.72535595574738,
"step": 2800
},
{
"epoch": 11.09,
"learning_rate": 7.212060301507539e-06,
"loss": 0.0017,
"step": 2825
},
{
"epoch": 11.19,
"learning_rate": 7.186934673366834e-06,
"loss": 0.0013,
"step": 2850
},
{
"epoch": 11.29,
"learning_rate": 7.161809045226131e-06,
"loss": 0.0013,
"step": 2875
},
{
"epoch": 11.38,
"learning_rate": 7.1366834170854284e-06,
"loss": 0.0014,
"step": 2900
},
{
"epoch": 11.38,
"eval_loss": 0.7448384761810303,
"eval_runtime": 1554.4308,
"eval_samples_per_second": 5.562,
"eval_steps_per_second": 0.556,
"eval_wer": 44.857865753832414,
"step": 2900
},
{
"epoch": 11.48,
"learning_rate": 7.111557788944724e-06,
"loss": 0.0014,
"step": 2925
},
{
"epoch": 11.58,
"learning_rate": 7.086432160804021e-06,
"loss": 0.0012,
"step": 2950
},
{
"epoch": 11.68,
"learning_rate": 7.061306532663317e-06,
"loss": 0.0012,
"step": 2975
},
{
"epoch": 11.78,
"learning_rate": 7.036180904522614e-06,
"loss": 0.0013,
"step": 3000
},
{
"epoch": 11.78,
"eval_loss": 0.7471309900283813,
"eval_runtime": 1597.219,
"eval_samples_per_second": 5.413,
"eval_steps_per_second": 0.542,
"eval_wer": 46.5115509913843,
"step": 3000
},
{
"epoch": 11.87,
"learning_rate": 7.011055276381911e-06,
"loss": 0.0021,
"step": 3025
},
{
"epoch": 11.97,
"learning_rate": 6.9859296482412065e-06,
"loss": 0.0021,
"step": 3050
},
{
"epoch": 12.07,
"learning_rate": 6.960804020100503e-06,
"loss": 0.0013,
"step": 3075
},
{
"epoch": 12.17,
"learning_rate": 6.935678391959799e-06,
"loss": 0.0015,
"step": 3100
},
{
"epoch": 12.17,
"eval_loss": 0.7513495683670044,
"eval_runtime": 1592.1737,
"eval_samples_per_second": 5.43,
"eval_steps_per_second": 0.543,
"eval_wer": 47.5699095434175,
"step": 3100
},
{
"epoch": 12.27,
"learning_rate": 6.910552763819096e-06,
"loss": 0.0008,
"step": 3125
},
{
"epoch": 12.37,
"learning_rate": 6.885427135678393e-06,
"loss": 0.001,
"step": 3150
},
{
"epoch": 12.46,
"learning_rate": 6.860301507537689e-06,
"loss": 0.0012,
"step": 3175
},
{
"epoch": 12.56,
"learning_rate": 6.835175879396986e-06,
"loss": 0.0014,
"step": 3200
},
{
"epoch": 12.56,
"eval_loss": 0.7380189299583435,
"eval_runtime": 1594.7949,
"eval_samples_per_second": 5.421,
"eval_steps_per_second": 0.542,
"eval_wer": 46.400754080468325,
"step": 3200
},
{
"epoch": 12.66,
"learning_rate": 6.810050251256282e-06,
"loss": 0.0013,
"step": 3225
},
{
"epoch": 12.76,
"learning_rate": 6.784924623115579e-06,
"loss": 0.0014,
"step": 3250
},
{
"epoch": 12.86,
"learning_rate": 6.759798994974875e-06,
"loss": 0.0018,
"step": 3275
},
{
"epoch": 12.95,
"learning_rate": 6.734673366834171e-06,
"loss": 0.0015,
"step": 3300
},
{
"epoch": 12.95,
"eval_loss": 0.7520471811294556,
"eval_runtime": 1574.6945,
"eval_samples_per_second": 5.491,
"eval_steps_per_second": 0.549,
"eval_wer": 45.965834862992175,
"step": 3300
},
{
"epoch": 13.05,
"learning_rate": 6.7095477386934685e-06,
"loss": 0.0026,
"step": 3325
},
{
"epoch": 13.15,
"learning_rate": 6.684422110552764e-06,
"loss": 0.0009,
"step": 3350
},
{
"epoch": 13.25,
"learning_rate": 6.659296482412061e-06,
"loss": 0.0009,
"step": 3375
},
{
"epoch": 13.35,
"learning_rate": 6.6341708542713575e-06,
"loss": 0.0009,
"step": 3400
},
{
"epoch": 13.35,
"eval_loss": 0.7481613159179688,
"eval_runtime": 1618.0061,
"eval_samples_per_second": 5.344,
"eval_steps_per_second": 0.535,
"eval_wer": 49.2269021514445,
"step": 3400
},
{
"epoch": 13.44,
"learning_rate": 6.609045226130654e-06,
"loss": 0.0009,
"step": 3425
},
{
"epoch": 13.54,
"learning_rate": 6.583919597989951e-06,
"loss": 0.0019,
"step": 3450
},
{
"epoch": 13.64,
"learning_rate": 6.5587939698492465e-06,
"loss": 0.0017,
"step": 3475
},
{
"epoch": 13.74,
"learning_rate": 6.533668341708543e-06,
"loss": 0.0021,
"step": 3500
},
{
"epoch": 13.74,
"eval_loss": 0.7618550658226013,
"eval_runtime": 1576.9068,
"eval_samples_per_second": 5.483,
"eval_steps_per_second": 0.549,
"eval_wer": 47.1234145292785,
"step": 3500
},
{
"epoch": 13.84,
"learning_rate": 6.508542713567839e-06,
"loss": 0.003,
"step": 3525
},
{
"epoch": 13.94,
"learning_rate": 6.483417085427136e-06,
"loss": 0.0017,
"step": 3550
},
{
"epoch": 14.03,
"learning_rate": 6.458291457286433e-06,
"loss": 0.0016,
"step": 3575
},
{
"epoch": 14.13,
"learning_rate": 6.433165829145729e-06,
"loss": 0.0013,
"step": 3600
},
{
"epoch": 14.13,
"eval_loss": 0.7452961802482605,
"eval_runtime": 1635.9285,
"eval_samples_per_second": 5.285,
"eval_steps_per_second": 0.529,
"eval_wer": 49.39557804567479,
"step": 3600
},
{
"epoch": 14.23,
"learning_rate": 6.408040201005025e-06,
"loss": 0.0011,
"step": 3625
},
{
"epoch": 14.33,
"learning_rate": 6.382914572864322e-06,
"loss": 0.0009,
"step": 3650
},
{
"epoch": 14.43,
"learning_rate": 6.357788944723619e-06,
"loss": 0.0017,
"step": 3675
},
{
"epoch": 14.52,
"learning_rate": 6.332663316582915e-06,
"loss": 0.001,
"step": 3700
},
{
"epoch": 14.52,
"eval_loss": 0.7582494020462036,
"eval_runtime": 1594.5778,
"eval_samples_per_second": 5.422,
"eval_steps_per_second": 0.542,
"eval_wer": 47.67409171338328,
"step": 3700
},
{
"epoch": 14.62,
"learning_rate": 6.307537688442211e-06,
"loss": 0.0012,
"step": 3725
},
{
"epoch": 14.72,
"learning_rate": 6.2824120603015085e-06,
"loss": 0.0014,
"step": 3750
},
{
"epoch": 14.82,
"learning_rate": 6.257286432160804e-06,
"loss": 0.001,
"step": 3775
},
{
"epoch": 14.92,
"learning_rate": 6.232160804020101e-06,
"loss": 0.0009,
"step": 3800
},
{
"epoch": 14.92,
"eval_loss": 0.7637260556221008,
"eval_runtime": 1592.8989,
"eval_samples_per_second": 5.428,
"eval_steps_per_second": 0.543,
"eval_wer": 46.971275487423725,
"step": 3800
},
{
"epoch": 15.01,
"learning_rate": 6.2070351758793975e-06,
"loss": 0.0008,
"step": 3825
},
{
"epoch": 15.11,
"learning_rate": 6.181909547738693e-06,
"loss": 0.0007,
"step": 3850
},
{
"epoch": 15.21,
"learning_rate": 6.156783919597991e-06,
"loss": 0.0003,
"step": 3875
},
{
"epoch": 15.31,
"learning_rate": 6.1316582914572865e-06,
"loss": 0.0014,
"step": 3900
},
{
"epoch": 15.31,
"eval_loss": 0.7722346782684326,
"eval_runtime": 1581.3202,
"eval_samples_per_second": 5.468,
"eval_steps_per_second": 0.547,
"eval_wer": 47.27059251542062,
"step": 3900
},
{
"epoch": 15.41,
"learning_rate": 6.106532663316583e-06,
"loss": 0.0007,
"step": 3925
},
{
"epoch": 15.51,
"learning_rate": 6.081407035175881e-06,
"loss": 0.0015,
"step": 3950
},
{
"epoch": 15.6,
"learning_rate": 6.056281407035176e-06,
"loss": 0.0012,
"step": 3975
},
{
"epoch": 15.7,
"learning_rate": 6.031155778894473e-06,
"loss": 0.001,
"step": 4000
},
{
"epoch": 15.7,
"eval_loss": 0.7691770195960999,
"eval_runtime": 1608.3109,
"eval_samples_per_second": 5.376,
"eval_steps_per_second": 0.538,
"eval_wer": 46.96631443171107,
"step": 4000
},
{
"epoch": 15.8,
"learning_rate": 6.006030150753769e-06,
"loss": 0.0006,
"step": 4025
},
{
"epoch": 15.9,
"learning_rate": 5.980904522613065e-06,
"loss": 0.0007,
"step": 4050
},
{
"epoch": 16.0,
"learning_rate": 5.955778894472363e-06,
"loss": 0.0008,
"step": 4075
},
{
"epoch": 16.09,
"learning_rate": 5.930653266331659e-06,
"loss": 0.0003,
"step": 4100
},
{
"epoch": 16.09,
"eval_loss": 0.774411678314209,
"eval_runtime": 1594.2731,
"eval_samples_per_second": 5.423,
"eval_steps_per_second": 0.543,
"eval_wer": 47.173025086405055,
"step": 4100
},
{
"epoch": 16.19,
"learning_rate": 5.905527638190955e-06,
"loss": 0.0004,
"step": 4125
},
{
"epoch": 16.29,
"learning_rate": 5.880402010050251e-06,
"loss": 0.0003,
"step": 4150
},
{
"epoch": 16.39,
"learning_rate": 5.8552763819095485e-06,
"loss": 0.0015,
"step": 4175
},
{
"epoch": 16.49,
"learning_rate": 5.830150753768845e-06,
"loss": 0.0004,
"step": 4200
},
{
"epoch": 16.49,
"eval_loss": 0.7842475175857544,
"eval_runtime": 1587.8752,
"eval_samples_per_second": 5.445,
"eval_steps_per_second": 0.545,
"eval_wer": 47.33508623968514,
"step": 4200
},
{
"epoch": 16.58,
"learning_rate": 5.805025125628141e-06,
"loss": 0.0005,
"step": 4225
},
{
"epoch": 16.68,
"learning_rate": 5.780904522613066e-06,
"loss": 0.0008,
"step": 4250
},
{
"epoch": 16.78,
"learning_rate": 5.755778894472362e-06,
"loss": 0.0003,
"step": 4275
},
{
"epoch": 16.88,
"learning_rate": 5.730653266331658e-06,
"loss": 0.0003,
"step": 4300
},
{
"epoch": 16.88,
"eval_loss": 0.7783875465393066,
"eval_runtime": 1573.9796,
"eval_samples_per_second": 5.493,
"eval_steps_per_second": 0.55,
"eval_wer": 47.07711134262704,
"step": 4300
},
{
"epoch": 16.98,
"learning_rate": 5.7055276381909555e-06,
"loss": 0.0005,
"step": 4325
},
{
"epoch": 17.08,
"learning_rate": 5.680402010050252e-06,
"loss": 0.0008,
"step": 4350
},
{
"epoch": 17.17,
"learning_rate": 5.655276381909548e-06,
"loss": 0.0005,
"step": 4375
},
{
"epoch": 17.27,
"learning_rate": 5.6301507537688445e-06,
"loss": 0.0002,
"step": 4400
},
{
"epoch": 17.27,
"eval_loss": 0.7879212498664856,
"eval_runtime": 1564.3111,
"eval_samples_per_second": 5.527,
"eval_steps_per_second": 0.553,
"eval_wer": 45.76408526401085,
"step": 4400
},
{
"epoch": 17.37,
"learning_rate": 5.605025125628141e-06,
"loss": 0.0004,
"step": 4425
},
{
"epoch": 17.47,
"learning_rate": 5.579899497487438e-06,
"loss": 0.0007,
"step": 4450
},
{
"epoch": 17.57,
"learning_rate": 5.554773869346734e-06,
"loss": 0.0005,
"step": 4475
},
{
"epoch": 17.66,
"learning_rate": 5.52964824120603e-06,
"loss": 0.0005,
"step": 4500
},
{
"epoch": 17.66,
"eval_loss": 0.7964885830879211,
"eval_runtime": 1592.8224,
"eval_samples_per_second": 5.428,
"eval_steps_per_second": 0.543,
"eval_wer": 50.02397843594451,
"step": 4500
},
{
"epoch": 17.76,
"learning_rate": 5.504522613065328e-06,
"loss": 0.0005,
"step": 4525
},
{
"epoch": 17.86,
"learning_rate": 5.479396984924623e-06,
"loss": 0.0005,
"step": 4550
},
{
"epoch": 17.96,
"learning_rate": 5.45427135678392e-06,
"loss": 0.0009,
"step": 4575
},
{
"epoch": 18.06,
"learning_rate": 5.429145728643217e-06,
"loss": 0.0004,
"step": 4600
},
{
"epoch": 18.06,
"eval_loss": 0.8001408576965332,
"eval_runtime": 1577.1711,
"eval_samples_per_second": 5.482,
"eval_steps_per_second": 0.548,
"eval_wer": 48.438094293132245,
"step": 4600
},
{
"epoch": 18.16,
"learning_rate": 5.404020100502513e-06,
"loss": 0.0007,
"step": 4625
},
{
"epoch": 18.25,
"learning_rate": 5.37889447236181e-06,
"loss": 0.0004,
"step": 4650
},
{
"epoch": 18.35,
"learning_rate": 5.353768844221106e-06,
"loss": 0.0002,
"step": 4675
},
{
"epoch": 18.45,
"learning_rate": 5.328643216080402e-06,
"loss": 0.0002,
"step": 4700
},
{
"epoch": 18.45,
"eval_loss": 0.8016232252120972,
"eval_runtime": 1606.7697,
"eval_samples_per_second": 5.381,
"eval_steps_per_second": 0.538,
"eval_wer": 49.00365464437499,
"step": 4700
},
{
"epoch": 18.55,
"learning_rate": 5.3035175879397e-06,
"loss": 0.0002,
"step": 4725
},
{
"epoch": 18.65,
"learning_rate": 5.2783919597989955e-06,
"loss": 0.0004,
"step": 4750
},
{
"epoch": 18.74,
"learning_rate": 5.253266331658292e-06,
"loss": 0.0002,
"step": 4775
},
{
"epoch": 18.84,
"learning_rate": 5.228140703517588e-06,
"loss": 0.0002,
"step": 4800
},
{
"epoch": 18.84,
"eval_loss": 0.8065798878669739,
"eval_runtime": 1608.099,
"eval_samples_per_second": 5.377,
"eval_steps_per_second": 0.538,
"eval_wer": 50.08681847497147,
"step": 4800
},
{
"epoch": 18.94,
"learning_rate": 5.2030150753768845e-06,
"loss": 0.0006,
"step": 4825
},
{
"epoch": 19.04,
"learning_rate": 5.177889447236182e-06,
"loss": 0.0002,
"step": 4850
},
{
"epoch": 19.14,
"learning_rate": 5.152763819095478e-06,
"loss": 0.0005,
"step": 4875
},
{
"epoch": 19.23,
"learning_rate": 5.127638190954774e-06,
"loss": 0.0009,
"step": 4900
},
{
"epoch": 19.23,
"eval_loss": 0.8020856380462646,
"eval_runtime": 1561.8925,
"eval_samples_per_second": 5.536,
"eval_steps_per_second": 0.554,
"eval_wer": 47.227596699244266,
"step": 4900
},
{
"epoch": 19.33,
"learning_rate": 5.10251256281407e-06,
"loss": 0.0005,
"step": 4925
},
{
"epoch": 19.43,
"learning_rate": 5.077386934673368e-06,
"loss": 0.0004,
"step": 4950
},
{
"epoch": 19.53,
"learning_rate": 5.052261306532664e-06,
"loss": 0.0007,
"step": 4975
},
{
"epoch": 19.63,
"learning_rate": 5.02713567839196e-06,
"loss": 0.0005,
"step": 5000
},
{
"epoch": 19.63,
"eval_loss": 0.8162133693695068,
"eval_runtime": 1588.3658,
"eval_samples_per_second": 5.443,
"eval_steps_per_second": 0.545,
"eval_wer": 47.34996940682311,
"step": 5000
},
{
"epoch": 19.73,
"learning_rate": 5.002010050251257e-06,
"loss": 0.0003,
"step": 5025
},
{
"epoch": 19.82,
"learning_rate": 4.976884422110553e-06,
"loss": 0.0004,
"step": 5050
},
{
"epoch": 19.92,
"learning_rate": 4.95175879396985e-06,
"loss": 0.0004,
"step": 5075
},
{
"epoch": 20.02,
"learning_rate": 4.9266331658291465e-06,
"loss": 0.0003,
"step": 5100
},
{
"epoch": 20.02,
"eval_loss": 0.8091246485710144,
"eval_runtime": 1607.5038,
"eval_samples_per_second": 5.379,
"eval_steps_per_second": 0.538,
"eval_wer": 48.72252815399117,
"step": 5100
},
{
"epoch": 20.12,
"learning_rate": 4.901507537688442e-06,
"loss": 0.0003,
"step": 5125
},
{
"epoch": 20.22,
"learning_rate": 4.876381909547739e-06,
"loss": 0.0003,
"step": 5150
},
{
"epoch": 20.31,
"learning_rate": 4.8512562814070355e-06,
"loss": 0.0003,
"step": 5175
},
{
"epoch": 20.41,
"learning_rate": 4.826130653266332e-06,
"loss": 0.0003,
"step": 5200
},
{
"epoch": 20.41,
"eval_loss": 0.8060252666473389,
"eval_runtime": 1624.2869,
"eval_samples_per_second": 5.323,
"eval_steps_per_second": 0.533,
"eval_wer": 51.50237303831589,
"step": 5200
},
{
"epoch": 20.51,
"learning_rate": 4.801005025125629e-06,
"loss": 0.0001,
"step": 5225
},
{
"epoch": 20.61,
"learning_rate": 4.7758793969849245e-06,
"loss": 0.0001,
"step": 5250
},
{
"epoch": 20.71,
"learning_rate": 4.750753768844221e-06,
"loss": 0.0003,
"step": 5275
},
{
"epoch": 20.8,
"learning_rate": 4.725628140703518e-06,
"loss": 0.0003,
"step": 5300
},
{
"epoch": 20.8,
"eval_loss": 0.8219889402389526,
"eval_runtime": 1625.3438,
"eval_samples_per_second": 5.319,
"eval_steps_per_second": 0.532,
"eval_wer": 51.487489871177914,
"step": 5300
},
{
"epoch": 20.9,
"learning_rate": 4.700502512562814e-06,
"loss": 0.0005,
"step": 5325
},
{
"epoch": 21.0,
"learning_rate": 4.675376884422111e-06,
"loss": 0.0004,
"step": 5350
},
{
"epoch": 21.1,
"learning_rate": 4.650251256281408e-06,
"loss": 0.0013,
"step": 5375
},
{
"epoch": 21.2,
"learning_rate": 4.625125628140703e-06,
"loss": 0.0003,
"step": 5400
},
{
"epoch": 21.2,
"eval_loss": 0.809751033782959,
"eval_runtime": 1570.6528,
"eval_samples_per_second": 5.505,
"eval_steps_per_second": 0.551,
"eval_wer": 45.861652693026414,
"step": 5400
},
{
"epoch": 21.3,
"learning_rate": 4.600000000000001e-06,
"loss": 0.0003,
"step": 5425
},
{
"epoch": 21.39,
"learning_rate": 4.574874371859297e-06,
"loss": 0.0003,
"step": 5450
},
{
"epoch": 21.49,
"learning_rate": 4.549748743718593e-06,
"loss": 0.0006,
"step": 5475
},
{
"epoch": 21.59,
"learning_rate": 4.52462311557789e-06,
"loss": 0.0003,
"step": 5500
},
{
"epoch": 21.59,
"eval_loss": 0.8131942749023438,
"eval_runtime": 1556.938,
"eval_samples_per_second": 5.553,
"eval_steps_per_second": 0.556,
"eval_wer": 44.87109523573283,
"step": 5500
},
{
"epoch": 21.69,
"learning_rate": 4.499497487437186e-06,
"loss": 0.0002,
"step": 5525
},
{
"epoch": 21.79,
"learning_rate": 4.474371859296483e-06,
"loss": 0.0003,
"step": 5550
},
{
"epoch": 21.88,
"learning_rate": 4.44924623115578e-06,
"loss": 0.0006,
"step": 5575
},
{
"epoch": 21.98,
"learning_rate": 4.4241206030150755e-06,
"loss": 0.0009,
"step": 5600
},
{
"epoch": 21.98,
"eval_loss": 0.8005769848823547,
"eval_runtime": 1557.485,
"eval_samples_per_second": 5.551,
"eval_steps_per_second": 0.555,
"eval_wer": 45.393659770799225,
"step": 5600
},
{
"epoch": 22.08,
"learning_rate": 4.398994974874372e-06,
"loss": 0.0009,
"step": 5625
},
{
"epoch": 22.18,
"learning_rate": 4.373869346733669e-06,
"loss": 0.0007,
"step": 5650
},
{
"epoch": 22.28,
"learning_rate": 4.348743718592965e-06,
"loss": 0.0011,
"step": 5675
},
{
"epoch": 22.37,
"learning_rate": 4.323618090452262e-06,
"loss": 0.0003,
"step": 5700
},
{
"epoch": 22.37,
"eval_loss": 0.8007655739784241,
"eval_runtime": 1588.3662,
"eval_samples_per_second": 5.443,
"eval_steps_per_second": 0.545,
"eval_wer": 45.61856096310628,
"step": 5700
},
{
"epoch": 22.47,
"learning_rate": 4.298492462311558e-06,
"loss": 0.0004,
"step": 5725
},
{
"epoch": 22.57,
"learning_rate": 4.273366834170854e-06,
"loss": 0.0004,
"step": 5750
},
{
"epoch": 22.67,
"learning_rate": 4.248241206030151e-06,
"loss": 0.0004,
"step": 5775
},
{
"epoch": 22.77,
"learning_rate": 4.223115577889448e-06,
"loss": 0.0002,
"step": 5800
},
{
"epoch": 22.77,
"eval_loss": 0.8081113696098328,
"eval_runtime": 1604.8728,
"eval_samples_per_second": 5.387,
"eval_steps_per_second": 0.539,
"eval_wer": 46.324684559540934,
"step": 5800
},
{
"epoch": 22.87,
"learning_rate": 4.197989949748744e-06,
"loss": 0.0005,
"step": 5825
},
{
"epoch": 22.96,
"learning_rate": 4.172864321608041e-06,
"loss": 0.0005,
"step": 5850
},
{
"epoch": 23.06,
"learning_rate": 4.147738693467337e-06,
"loss": 0.0008,
"step": 5875
},
{
"epoch": 23.16,
"learning_rate": 4.122613065326633e-06,
"loss": 0.0002,
"step": 5900
},
{
"epoch": 23.16,
"eval_loss": 0.8082166314125061,
"eval_runtime": 1581.3962,
"eval_samples_per_second": 5.467,
"eval_steps_per_second": 0.547,
"eval_wer": 46.12789601627226,
"step": 5900
},
{
"epoch": 23.26,
"learning_rate": 4.09748743718593e-06,
"loss": 0.0002,
"step": 5925
},
{
"epoch": 23.36,
"learning_rate": 4.0723618090452265e-06,
"loss": 0.0003,
"step": 5950
},
{
"epoch": 23.45,
"learning_rate": 4.047236180904523e-06,
"loss": 0.0003,
"step": 5975
},
{
"epoch": 23.55,
"learning_rate": 4.02211055276382e-06,
"loss": 0.0002,
"step": 6000
},
{
"epoch": 23.55,
"eval_loss": 0.8238235116004944,
"eval_runtime": 1599.956,
"eval_samples_per_second": 5.404,
"eval_steps_per_second": 0.541,
"eval_wer": 46.177506573398816,
"step": 6000
},
{
"epoch": 23.65,
"learning_rate": 3.9969849246231155e-06,
"loss": 0.0005,
"step": 6025
},
{
"epoch": 23.75,
"learning_rate": 3.971859296482413e-06,
"loss": 0.0007,
"step": 6050
},
{
"epoch": 23.85,
"learning_rate": 3.946733668341709e-06,
"loss": 0.0004,
"step": 6075
},
{
"epoch": 23.95,
"learning_rate": 3.921608040201005e-06,
"loss": 0.0005,
"step": 6100
},
{
"epoch": 23.95,
"eval_loss": 0.8119234442710876,
"eval_runtime": 1612.6898,
"eval_samples_per_second": 5.361,
"eval_steps_per_second": 0.536,
"eval_wer": 49.972714193580394,
"step": 6100
},
{
"epoch": 24.04,
"learning_rate": 3.896482412060302e-06,
"loss": 0.0006,
"step": 6125
},
{
"epoch": 24.14,
"learning_rate": 3.871356783919598e-06,
"loss": 0.0003,
"step": 6150
},
{
"epoch": 24.24,
"learning_rate": 3.846231155778895e-06,
"loss": 0.0009,
"step": 6175
},
{
"epoch": 24.34,
"learning_rate": 3.821105527638191e-06,
"loss": 0.0002,
"step": 6200
},
{
"epoch": 24.34,
"eval_loss": 0.8315232992172241,
"eval_runtime": 1606.8991,
"eval_samples_per_second": 5.381,
"eval_steps_per_second": 0.538,
"eval_wer": 49.086338906252585,
"step": 6200
},
{
"epoch": 24.44,
"learning_rate": 3.7959798994974876e-06,
"loss": 0.0003,
"step": 6225
},
{
"epoch": 24.53,
"learning_rate": 3.7708542713567843e-06,
"loss": 0.0003,
"step": 6250
},
{
"epoch": 24.63,
"learning_rate": 3.7457286432160805e-06,
"loss": 0.0004,
"step": 6275
},
{
"epoch": 24.73,
"learning_rate": 3.720603015075377e-06,
"loss": 0.0001,
"step": 6300
},
{
"epoch": 24.73,
"eval_loss": 0.8223620057106018,
"eval_runtime": 1589.1151,
"eval_samples_per_second": 5.441,
"eval_steps_per_second": 0.544,
"eval_wer": 47.22428932876917,
"step": 6300
},
{
"epoch": 24.83,
"learning_rate": 3.6954773869346737e-06,
"loss": 0.0002,
"step": 6325
},
{
"epoch": 24.93,
"learning_rate": 3.6703517587939703e-06,
"loss": 0.0003,
"step": 6350
},
{
"epoch": 25.02,
"learning_rate": 3.6452261306532665e-06,
"loss": 0.0001,
"step": 6375
},
{
"epoch": 25.12,
"learning_rate": 3.620100502512563e-06,
"loss": 0.0001,
"step": 6400
},
{
"epoch": 25.12,
"eval_loss": 0.8258624076843262,
"eval_runtime": 1596.5065,
"eval_samples_per_second": 5.416,
"eval_steps_per_second": 0.542,
"eval_wer": 47.1680640306924,
"step": 6400
},
{
"epoch": 25.22,
"learning_rate": 3.5949748743718593e-06,
"loss": 0.0002,
"step": 6425
},
{
"epoch": 25.32,
"learning_rate": 3.5698492462311564e-06,
"loss": 0.0001,
"step": 6450
},
{
"epoch": 25.42,
"learning_rate": 3.5447236180904526e-06,
"loss": 0.0004,
"step": 6475
},
{
"epoch": 25.52,
"learning_rate": 3.519597989949749e-06,
"loss": 0.0001,
"step": 6500
},
{
"epoch": 25.52,
"eval_loss": 0.8219370245933533,
"eval_runtime": 1629.406,
"eval_samples_per_second": 5.306,
"eval_steps_per_second": 0.531,
"eval_wer": 48.5736964826115,
"step": 6500
},
{
"epoch": 25.61,
"learning_rate": 3.4944723618090454e-06,
"loss": 0.0002,
"step": 6525
},
{
"epoch": 25.71,
"learning_rate": 3.4693467336683416e-06,
"loss": 0.0001,
"step": 6550
},
{
"epoch": 25.81,
"learning_rate": 3.4442211055276386e-06,
"loss": 0.0001,
"step": 6575
},
{
"epoch": 25.91,
"learning_rate": 3.419095477386935e-06,
"loss": 0.0002,
"step": 6600
},
{
"epoch": 25.91,
"eval_loss": 0.839992344379425,
"eval_runtime": 1605.3111,
"eval_samples_per_second": 5.386,
"eval_steps_per_second": 0.539,
"eval_wer": 48.90774090059698,
"step": 6600
},
{
"epoch": 26.01,
"learning_rate": 3.3939698492462315e-06,
"loss": 0.0009,
"step": 6625
},
{
"epoch": 26.1,
"learning_rate": 3.3688442211055276e-06,
"loss": 0.0004,
"step": 6650
},
{
"epoch": 26.2,
"learning_rate": 3.3437185929648243e-06,
"loss": 0.0002,
"step": 6675
},
{
"epoch": 26.3,
"learning_rate": 3.318592964824121e-06,
"loss": 0.0005,
"step": 6700
},
{
"epoch": 26.3,
"eval_loss": 0.8318980932235718,
"eval_runtime": 1614.1036,
"eval_samples_per_second": 5.357,
"eval_steps_per_second": 0.536,
"eval_wer": 47.55668006151709,
"step": 6700
},
{
"epoch": 26.4,
"learning_rate": 3.2934673366834175e-06,
"loss": 0.0002,
"step": 6725
},
{
"epoch": 26.5,
"learning_rate": 3.2683417085427137e-06,
"loss": 0.0002,
"step": 6750
},
{
"epoch": 26.59,
"learning_rate": 3.2432160804020103e-06,
"loss": 0.0006,
"step": 6775
},
{
"epoch": 26.69,
"learning_rate": 3.2180904522613065e-06,
"loss": 0.0001,
"step": 6800
},
{
"epoch": 26.69,
"eval_loss": 0.8393569588661194,
"eval_runtime": 1658.5859,
"eval_samples_per_second": 5.213,
"eval_steps_per_second": 0.522,
"eval_wer": 50.23565014635114,
"step": 6800
},
{
"epoch": 26.79,
"learning_rate": 3.1929648241206036e-06,
"loss": 0.0001,
"step": 6825
},
{
"epoch": 26.89,
"learning_rate": 3.1688442211055283e-06,
"loss": 0.0002,
"step": 6850
},
{
"epoch": 26.99,
"learning_rate": 3.1437185929648245e-06,
"loss": 0.0005,
"step": 6875
},
{
"epoch": 27.09,
"learning_rate": 3.1185929648241207e-06,
"loss": 0.0001,
"step": 6900
},
{
"epoch": 27.09,
"eval_loss": 0.8479753136634827,
"eval_runtime": 1618.6719,
"eval_samples_per_second": 5.341,
"eval_steps_per_second": 0.534,
"eval_wer": 48.46289957169552,
"step": 6900
},
{
"epoch": 27.18,
"learning_rate": 3.0934673366834173e-06,
"loss": 0.0004,
"step": 6925
},
{
"epoch": 27.28,
"learning_rate": 3.0683417085427135e-06,
"loss": 0.0001,
"step": 6950
},
{
"epoch": 27.38,
"learning_rate": 3.0432160804020105e-06,
"loss": 0.0003,
"step": 6975
},
{
"epoch": 27.48,
"learning_rate": 3.0180904522613067e-06,
"loss": 0.0001,
"step": 7000
},
{
"epoch": 27.48,
"eval_loss": 0.8498477339744568,
"eval_runtime": 1594.4093,
"eval_samples_per_second": 5.423,
"eval_steps_per_second": 0.543,
"eval_wer": 47.11514610309074,
"step": 7000
},
{
"epoch": 27.58,
"learning_rate": 2.9929648241206034e-06,
"loss": 0.0001,
"step": 7025
},
{
"epoch": 27.67,
"learning_rate": 2.9678391959798996e-06,
"loss": 0.0001,
"step": 7050
},
{
"epoch": 27.77,
"learning_rate": 2.942713567839196e-06,
"loss": 0.0003,
"step": 7075
},
{
"epoch": 27.87,
"learning_rate": 2.917587939698493e-06,
"loss": 0.0002,
"step": 7100
},
{
"epoch": 27.87,
"eval_loss": 0.8341818451881409,
"eval_runtime": 1613.2296,
"eval_samples_per_second": 5.359,
"eval_steps_per_second": 0.536,
"eval_wer": 48.9242777529725,
"step": 7100
},
{
"epoch": 27.97,
"learning_rate": 2.8924623115577894e-06,
"loss": 0.0002,
"step": 7125
},
{
"epoch": 28.07,
"learning_rate": 2.8673366834170856e-06,
"loss": 0.0004,
"step": 7150
},
{
"epoch": 28.16,
"learning_rate": 2.8422110552763822e-06,
"loss": 0.0005,
"step": 7175
},
{
"epoch": 28.26,
"learning_rate": 2.8170854271356784e-06,
"loss": 0.0003,
"step": 7200
},
{
"epoch": 28.26,
"eval_loss": 0.8183611035346985,
"eval_runtime": 1606.5293,
"eval_samples_per_second": 5.382,
"eval_steps_per_second": 0.538,
"eval_wer": 47.373121000148835,
"step": 7200
},
{
"epoch": 28.36,
"learning_rate": 2.7919597989949755e-06,
"loss": 0.0002,
"step": 7225
},
{
"epoch": 28.46,
"learning_rate": 2.7668341708542717e-06,
"loss": 0.0002,
"step": 7250
},
{
"epoch": 28.56,
"learning_rate": 2.7417085427135683e-06,
"loss": 0.0002,
"step": 7275
},
{
"epoch": 28.66,
"learning_rate": 2.7165829145728645e-06,
"loss": 0.0001,
"step": 7300
},
{
"epoch": 28.66,
"eval_loss": 0.8278088569641113,
"eval_runtime": 1611.8451,
"eval_samples_per_second": 5.364,
"eval_steps_per_second": 0.537,
"eval_wer": 47.92875923996626,
"step": 7300
},
{
"epoch": 28.75,
"learning_rate": 2.6914572864321607e-06,
"loss": 0.0001,
"step": 7325
},
{
"epoch": 28.85,
"learning_rate": 2.6663316582914573e-06,
"loss": 0.0001,
"step": 7350
},
{
"epoch": 28.95,
"learning_rate": 2.6412060301507544e-06,
"loss": 0.0002,
"step": 7375
},
{
"epoch": 29.05,
"learning_rate": 2.6160804020100506e-06,
"loss": 0.0002,
"step": 7400
},
{
"epoch": 29.05,
"eval_loss": 0.8438936471939087,
"eval_runtime": 1612.1092,
"eval_samples_per_second": 5.363,
"eval_steps_per_second": 0.537,
"eval_wer": 47.86095814522664,
"step": 7400
},
{
"epoch": 29.15,
"learning_rate": 2.5909547738693468e-06,
"loss": 0.0003,
"step": 7425
},
{
"epoch": 29.24,
"learning_rate": 2.5658291457286434e-06,
"loss": 0.0001,
"step": 7450
},
{
"epoch": 29.34,
"learning_rate": 2.5407035175879396e-06,
"loss": 0.0001,
"step": 7475
},
{
"epoch": 29.44,
"learning_rate": 2.5155778894472366e-06,
"loss": 0.0001,
"step": 7500
},
{
"epoch": 29.44,
"eval_loss": 0.8461014628410339,
"eval_runtime": 1631.5395,
"eval_samples_per_second": 5.299,
"eval_steps_per_second": 0.53,
"eval_wer": 49.94625522977956,
"step": 7500
},
{
"epoch": 29.54,
"learning_rate": 2.490452261306533e-06,
"loss": 0.0001,
"step": 7525
},
{
"epoch": 29.64,
"learning_rate": 2.4653266331658294e-06,
"loss": 0.0002,
"step": 7550
},
{
"epoch": 29.74,
"learning_rate": 2.4402010050251256e-06,
"loss": 0.0001,
"step": 7575
},
{
"epoch": 29.83,
"learning_rate": 2.4150753768844223e-06,
"loss": 0.0001,
"step": 7600
},
{
"epoch": 29.83,
"eval_loss": 0.8448674082756042,
"eval_runtime": 1642.4029,
"eval_samples_per_second": 5.264,
"eval_steps_per_second": 0.527,
"eval_wer": 48.48605116502125,
"step": 7600
},
{
"epoch": 29.93,
"learning_rate": 2.389949748743719e-06,
"loss": 0.0002,
"step": 7625
},
{
"epoch": 30.03,
"learning_rate": 2.3648241206030155e-06,
"loss": 0.0001,
"step": 7650
},
{
"epoch": 30.13,
"learning_rate": 2.3396984924623117e-06,
"loss": 0.0001,
"step": 7675
},
{
"epoch": 30.23,
"learning_rate": 2.314572864321608e-06,
"loss": 0.0001,
"step": 7700
},
{
"epoch": 30.23,
"eval_loss": 0.851162850856781,
"eval_runtime": 1646.9818,
"eval_samples_per_second": 5.25,
"eval_steps_per_second": 0.525,
"eval_wer": 49.000347273899884,
"step": 7700
},
{
"epoch": 30.32,
"learning_rate": 2.289447236180905e-06,
"loss": 0.0001,
"step": 7725
},
{
"epoch": 30.42,
"learning_rate": 2.264321608040201e-06,
"loss": 0.0001,
"step": 7750
},
{
"epoch": 30.52,
"learning_rate": 2.2391959798994978e-06,
"loss": 0.0001,
"step": 7775
},
{
"epoch": 30.62,
"learning_rate": 2.214070351758794e-06,
"loss": 0.0001,
"step": 7800
},
{
"epoch": 30.62,
"eval_loss": 0.8554991483688354,
"eval_runtime": 1638.7791,
"eval_samples_per_second": 5.276,
"eval_steps_per_second": 0.528,
"eval_wer": 48.27768682508971,
"step": 7800
},
{
"epoch": 30.72,
"learning_rate": 2.1889447236180906e-06,
"loss": 0.0001,
"step": 7825
},
{
"epoch": 30.81,
"learning_rate": 2.163819095477387e-06,
"loss": 0.0001,
"step": 7850
},
{
"epoch": 30.91,
"learning_rate": 2.1386934673366834e-06,
"loss": 0.0001,
"step": 7875
},
{
"epoch": 31.01,
"learning_rate": 2.11356783919598e-06,
"loss": 0.0001,
"step": 7900
},
{
"epoch": 31.01,
"eval_loss": 0.854286253452301,
"eval_runtime": 1639.4413,
"eval_samples_per_second": 5.274,
"eval_steps_per_second": 0.528,
"eval_wer": 48.67953233781482,
"step": 7900
},
{
"epoch": 31.11,
"learning_rate": 2.0884422110552766e-06,
"loss": 0.0001,
"step": 7925
},
{
"epoch": 31.21,
"learning_rate": 2.063316582914573e-06,
"loss": 0.0001,
"step": 7950
},
{
"epoch": 31.31,
"learning_rate": 2.0381909547738695e-06,
"loss": 0.0001,
"step": 7975
},
{
"epoch": 31.4,
"learning_rate": 2.013065326633166e-06,
"loss": 0.0001,
"step": 8000
},
{
"epoch": 31.4,
"eval_loss": 0.8566484451293945,
"eval_runtime": 1631.2411,
"eval_samples_per_second": 5.3,
"eval_steps_per_second": 0.53,
"eval_wer": 48.76552397016752,
"step": 8000
},
{
"epoch": 31.5,
"learning_rate": 1.9879396984924627e-06,
"loss": 0.0001,
"step": 8025
},
{
"epoch": 31.6,
"learning_rate": 1.962814070351759e-06,
"loss": 0.0001,
"step": 8050
},
{
"epoch": 31.7,
"learning_rate": 1.9376884422110555e-06,
"loss": 0.0001,
"step": 8075
},
{
"epoch": 31.8,
"learning_rate": 1.912562814070352e-06,
"loss": 0.0001,
"step": 8100
},
{
"epoch": 31.8,
"eval_loss": 0.8605026602745056,
"eval_runtime": 1651.0794,
"eval_samples_per_second": 5.237,
"eval_steps_per_second": 0.524,
"eval_wer": 48.67787865257726,
"step": 8100
},
{
"epoch": 31.89,
"learning_rate": 1.8874371859296483e-06,
"loss": 0.0,
"step": 8125
},
{
"epoch": 31.99,
"learning_rate": 1.862311557788945e-06,
"loss": 0.0,
"step": 8150
},
{
"epoch": 32.09,
"learning_rate": 1.8371859296482414e-06,
"loss": 0.0,
"step": 8175
},
{
"epoch": 32.19,
"learning_rate": 1.8120603015075378e-06,
"loss": 0.0,
"step": 8200
},
{
"epoch": 32.19,
"eval_loss": 0.8633646965026855,
"eval_runtime": 1649.2073,
"eval_samples_per_second": 5.243,
"eval_steps_per_second": 0.524,
"eval_wer": 49.36911908187396,
"step": 8200
},
{
"epoch": 32.29,
"learning_rate": 1.7869346733668344e-06,
"loss": 0.0,
"step": 8225
},
{
"epoch": 32.38,
"learning_rate": 1.7618090452261308e-06,
"loss": 0.0,
"step": 8250
},
{
"epoch": 32.48,
"learning_rate": 1.7366834170854272e-06,
"loss": 0.0,
"step": 8275
},
{
"epoch": 32.58,
"learning_rate": 1.7115577889447238e-06,
"loss": 0.0,
"step": 8300
},
{
"epoch": 32.58,
"eval_loss": 0.8662922978401184,
"eval_runtime": 1644.0444,
"eval_samples_per_second": 5.259,
"eval_steps_per_second": 0.526,
"eval_wer": 50.04382265879512,
"step": 8300
},
{
"epoch": 32.68,
"learning_rate": 1.6864321608040202e-06,
"loss": 0.0,
"step": 8325
},
{
"epoch": 32.78,
"learning_rate": 1.6613065326633169e-06,
"loss": 0.0,
"step": 8350
},
{
"epoch": 32.88,
"learning_rate": 1.6361809045226133e-06,
"loss": 0.0,
"step": 8375
},
{
"epoch": 32.97,
"learning_rate": 1.6110552763819095e-06,
"loss": 0.0,
"step": 8400
},
{
"epoch": 32.97,
"eval_loss": 0.8684667944908142,
"eval_runtime": 1656.8342,
"eval_samples_per_second": 5.218,
"eval_steps_per_second": 0.522,
"eval_wer": 49.727968778422714,
"step": 8400
},
{
"epoch": 33.07,
"learning_rate": 1.5859296482412063e-06,
"loss": 0.0,
"step": 8425
},
{
"epoch": 33.17,
"learning_rate": 1.5608040201005025e-06,
"loss": 0.0,
"step": 8450
},
{
"epoch": 33.27,
"learning_rate": 1.5356783919597993e-06,
"loss": 0.0,
"step": 8475
},
{
"epoch": 33.37,
"learning_rate": 1.5105527638190955e-06,
"loss": 0.0,
"step": 8500
},
{
"epoch": 33.37,
"eval_loss": 0.8704341053962708,
"eval_runtime": 1651.7464,
"eval_samples_per_second": 5.234,
"eval_steps_per_second": 0.524,
"eval_wer": 49.164062112417525,
"step": 8500
},
{
"epoch": 33.46,
"learning_rate": 1.485427135678392e-06,
"loss": 0.0,
"step": 8525
},
{
"epoch": 33.56,
"learning_rate": 1.4603015075376886e-06,
"loss": 0.0,
"step": 8550
},
{
"epoch": 33.66,
"learning_rate": 1.435175879396985e-06,
"loss": 0.0,
"step": 8575
},
{
"epoch": 33.76,
"learning_rate": 1.4100502512562816e-06,
"loss": 0.0,
"step": 8600
},
{
"epoch": 33.76,
"eval_loss": 0.8724026083946228,
"eval_runtime": 1643.6363,
"eval_samples_per_second": 5.26,
"eval_steps_per_second": 0.526,
"eval_wer": 48.841593491094905,
"step": 8600
},
{
"epoch": 33.86,
"learning_rate": 1.384924623115578e-06,
"loss": 0.0,
"step": 8625
},
{
"epoch": 33.95,
"learning_rate": 1.3597989949748744e-06,
"loss": 0.0,
"step": 8650
},
{
"epoch": 34.05,
"learning_rate": 1.334673366834171e-06,
"loss": 0.0,
"step": 8675
},
{
"epoch": 34.15,
"learning_rate": 1.3095477386934674e-06,
"loss": 0.0,
"step": 8700
},
{
"epoch": 34.15,
"eval_loss": 0.8736163377761841,
"eval_runtime": 1654.9145,
"eval_samples_per_second": 5.224,
"eval_steps_per_second": 0.523,
"eval_wer": 49.22855583668205,
"step": 8700
},
{
"epoch": 34.25,
"learning_rate": 1.284422110552764e-06,
"loss": 0.0,
"step": 8725
},
{
"epoch": 34.35,
"learning_rate": 1.2592964824120605e-06,
"loss": 0.0,
"step": 8750
},
{
"epoch": 34.45,
"learning_rate": 1.2341708542713569e-06,
"loss": 0.0,
"step": 8775
},
{
"epoch": 34.54,
"learning_rate": 1.2090452261306533e-06,
"loss": 0.0,
"step": 8800
},
{
"epoch": 34.54,
"eval_loss": 0.8755289316177368,
"eval_runtime": 1642.3091,
"eval_samples_per_second": 5.265,
"eval_steps_per_second": 0.527,
"eval_wer": 48.61338492831275,
"step": 8800
},
{
"epoch": 34.64,
"learning_rate": 1.18391959798995e-06,
"loss": 0.0,
"step": 8825
},
{
"epoch": 34.74,
"learning_rate": 1.1587939698492463e-06,
"loss": 0.0,
"step": 8850
},
{
"epoch": 34.84,
"learning_rate": 1.133668341708543e-06,
"loss": 0.0,
"step": 8875
},
{
"epoch": 34.94,
"learning_rate": 1.1085427135678393e-06,
"loss": 0.0,
"step": 8900
},
{
"epoch": 34.94,
"eval_loss": 0.8766536116600037,
"eval_runtime": 1656.0111,
"eval_samples_per_second": 5.221,
"eval_steps_per_second": 0.522,
"eval_wer": 48.92593143821005,
"step": 8900
},
{
"epoch": 35.03,
"learning_rate": 1.0834170854271357e-06,
"loss": 0.0,
"step": 8925
},
{
"epoch": 35.13,
"learning_rate": 1.0582914572864322e-06,
"loss": 0.0,
"step": 8950
},
{
"epoch": 35.23,
"learning_rate": 1.0331658291457288e-06,
"loss": 0.0,
"step": 8975
},
{
"epoch": 35.33,
"learning_rate": 1.0080402010050252e-06,
"loss": 0.0,
"step": 9000
},
{
"epoch": 35.33,
"eval_loss": 0.8778066039085388,
"eval_runtime": 1650.953,
"eval_samples_per_second": 5.237,
"eval_steps_per_second": 0.524,
"eval_wer": 48.98050305104926,
"step": 9000
},
{
"epoch": 35.43,
"learning_rate": 9.829145728643216e-07,
"loss": 0.0,
"step": 9025
},
{
"epoch": 35.53,
"learning_rate": 9.577889447236182e-07,
"loss": 0.0,
"step": 9050
},
{
"epoch": 35.62,
"learning_rate": 9.326633165829146e-07,
"loss": 0.0,
"step": 9075
},
{
"epoch": 35.72,
"learning_rate": 9.075376884422111e-07,
"loss": 0.0,
"step": 9100
},
{
"epoch": 35.72,
"eval_loss": 0.8791189193725586,
"eval_runtime": 1640.1356,
"eval_samples_per_second": 5.272,
"eval_steps_per_second": 0.527,
"eval_wer": 49.321162209984955,
"step": 9100
},
{
"epoch": 35.82,
"learning_rate": 8.824120603015077e-07,
"loss": 0.0,
"step": 9125
},
{
"epoch": 35.92,
"learning_rate": 8.572864321608041e-07,
"loss": 0.0,
"step": 9150
},
{
"epoch": 36.02,
"learning_rate": 8.321608040201006e-07,
"loss": 0.0,
"step": 9175
},
{
"epoch": 36.11,
"learning_rate": 8.070351758793971e-07,
"loss": 0.0,
"step": 9200
},
{
"epoch": 36.11,
"eval_loss": 0.8800622820854187,
"eval_runtime": 1655.7118,
"eval_samples_per_second": 5.222,
"eval_steps_per_second": 0.522,
"eval_wer": 49.37242645234906,
"step": 9200
},
{
"epoch": 36.21,
"learning_rate": 7.819095477386936e-07,
"loss": 0.0,
"step": 9225
},
{
"epoch": 36.31,
"learning_rate": 7.567839195979901e-07,
"loss": 0.0,
"step": 9250
},
{
"epoch": 36.41,
"learning_rate": 7.316582914572864e-07,
"loss": 0.0,
"step": 9275
},
{
"epoch": 36.51,
"learning_rate": 7.065326633165829e-07,
"loss": 0.0,
"step": 9300
},
{
"epoch": 36.51,
"eval_loss": 0.881312906742096,
"eval_runtime": 1661.7481,
"eval_samples_per_second": 5.203,
"eval_steps_per_second": 0.521,
"eval_wer": 49.43361280613848,
"step": 9300
},
{
"epoch": 36.6,
"learning_rate": 6.814070351758795e-07,
"loss": 0.0,
"step": 9325
},
{
"epoch": 36.7,
"learning_rate": 6.56281407035176e-07,
"loss": 0.0,
"step": 9350
},
{
"epoch": 36.8,
"learning_rate": 6.311557788944724e-07,
"loss": 0.0,
"step": 9375
},
{
"epoch": 36.9,
"learning_rate": 6.060301507537689e-07,
"loss": 0.0,
"step": 9400
},
{
"epoch": 36.9,
"eval_loss": 0.8819226622581482,
"eval_runtime": 1659.1244,
"eval_samples_per_second": 5.211,
"eval_steps_per_second": 0.521,
"eval_wer": 49.10452944386565,
"step": 9400
},
{
"epoch": 37.0,
"learning_rate": 5.809045226130654e-07,
"loss": 0.0,
"step": 9425
},
{
"epoch": 37.1,
"learning_rate": 5.557788944723619e-07,
"loss": 0.0,
"step": 9450
},
{
"epoch": 37.19,
"learning_rate": 5.306532663316583e-07,
"loss": 0.0,
"step": 9475
},
{
"epoch": 37.29,
"learning_rate": 5.055276381909549e-07,
"loss": 0.0,
"step": 9500
},
{
"epoch": 37.29,
"eval_loss": 0.8826147317886353,
"eval_runtime": 1662.0225,
"eval_samples_per_second": 5.202,
"eval_steps_per_second": 0.52,
"eval_wer": 49.26328322667064,
"step": 9500
},
{
"epoch": 37.39,
"learning_rate": 4.804020100502513e-07,
"loss": 0.0,
"step": 9525
},
{
"epoch": 37.49,
"learning_rate": 4.552763819095478e-07,
"loss": 0.0,
"step": 9550
},
{
"epoch": 37.59,
"learning_rate": 4.3015075376884424e-07,
"loss": 0.0,
"step": 9575
},
{
"epoch": 37.68,
"learning_rate": 4.0502512562814075e-07,
"loss": 0.0,
"step": 9600
},
{
"epoch": 37.68,
"eval_loss": 0.8831948041915894,
"eval_runtime": 1662.4658,
"eval_samples_per_second": 5.201,
"eval_steps_per_second": 0.52,
"eval_wer": 49.423690694713166,
"step": 9600
},
{
"epoch": 37.78,
"learning_rate": 3.7989949748743727e-07,
"loss": 0.0,
"step": 9625
},
{
"epoch": 37.88,
"learning_rate": 3.547738693467337e-07,
"loss": 0.0,
"step": 9650
},
{
"epoch": 37.98,
"learning_rate": 3.296482412060302e-07,
"loss": 0.0,
"step": 9675
},
{
"epoch": 38.08,
"learning_rate": 3.0452261306532665e-07,
"loss": 0.0,
"step": 9700
},
{
"epoch": 38.08,
"eval_loss": 0.8836556673049927,
"eval_runtime": 1645.335,
"eval_samples_per_second": 5.255,
"eval_steps_per_second": 0.526,
"eval_wer": 48.63157546592581,
"step": 9700
},
{
"epoch": 38.17,
"learning_rate": 2.7939698492462317e-07,
"loss": 0.0,
"step": 9725
},
{
"epoch": 38.27,
"learning_rate": 2.5427135678391963e-07,
"loss": 0.0,
"step": 9750
},
{
"epoch": 38.37,
"learning_rate": 2.291457286432161e-07,
"loss": 0.0,
"step": 9775
},
{
"epoch": 38.47,
"learning_rate": 2.0402010050251258e-07,
"loss": 0.0,
"step": 9800
},
{
"epoch": 38.47,
"eval_loss": 0.8840765953063965,
"eval_runtime": 1659.0687,
"eval_samples_per_second": 5.211,
"eval_steps_per_second": 0.521,
"eval_wer": 48.64645863306378,
"step": 9800
},
{
"epoch": 38.57,
"learning_rate": 1.7889447236180904e-07,
"loss": 0.0,
"step": 9825
},
{
"epoch": 38.67,
"learning_rate": 1.5376884422110556e-07,
"loss": 0.0,
"step": 9850
},
{
"epoch": 38.76,
"learning_rate": 1.2864321608040202e-07,
"loss": 0.0,
"step": 9875
},
{
"epoch": 38.86,
"learning_rate": 1.035175879396985e-07,
"loss": 0.0,
"step": 9900
},
{
"epoch": 38.86,
"eval_loss": 0.8842225670814514,
"eval_runtime": 1661.5936,
"eval_samples_per_second": 5.203,
"eval_steps_per_second": 0.521,
"eval_wer": 48.93419986439781,
"step": 9900
}
],
"logging_steps": 25,
"max_steps": 10000,
"num_train_epochs": 40,
"save_steps": 100,
"total_flos": 1.5989304973049856e+20,
"trial_name": null,
"trial_params": null
}