wav2vec2-large-xls-r-1b-nl-lm / trainer_state.json
RuudVelo's picture
Upload trainer_state.json
7bfeeb7
raw
history blame
79.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 47.47152290548906,
"global_step": 46000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"learning_rate": 9.900000000000002e-06,
"loss": 5.0766,
"step": 100
},
{
"epoch": 0.21,
"learning_rate": 1.9900000000000003e-05,
"loss": 2.8354,
"step": 200
},
{
"epoch": 0.31,
"learning_rate": 2.9900000000000002e-05,
"loss": 1.7963,
"step": 300
},
{
"epoch": 0.41,
"learning_rate": 3.99e-05,
"loss": 1.3003,
"step": 400
},
{
"epoch": 0.52,
"learning_rate": 4.99e-05,
"loss": 1.2223,
"step": 500
},
{
"epoch": 0.52,
"eval_loss": 0.38657668232917786,
"eval_runtime": 473.1633,
"eval_samples_per_second": 22.142,
"eval_steps_per_second": 2.769,
"eval_wer": 0.3425097378671439,
"step": 500
},
{
"epoch": 0.62,
"learning_rate": 4.989676746611053e-05,
"loss": 1.1977,
"step": 600
},
{
"epoch": 0.72,
"learning_rate": 4.979249217935349e-05,
"loss": 1.1405,
"step": 700
},
{
"epoch": 0.83,
"learning_rate": 4.9688216892596456e-05,
"loss": 1.1294,
"step": 800
},
{
"epoch": 0.93,
"learning_rate": 4.958394160583942e-05,
"loss": 1.108,
"step": 900
},
{
"epoch": 1.03,
"learning_rate": 4.947966631908238e-05,
"loss": 1.0748,
"step": 1000
},
{
"epoch": 1.03,
"eval_loss": 0.25742873549461365,
"eval_runtime": 470.5268,
"eval_samples_per_second": 22.267,
"eval_steps_per_second": 2.784,
"eval_wer": 0.2169386251184335,
"step": 1000
},
{
"epoch": 1.14,
"learning_rate": 4.937539103232534e-05,
"loss": 1.0635,
"step": 1100
},
{
"epoch": 1.24,
"learning_rate": 4.9271115745568305e-05,
"loss": 1.0427,
"step": 1200
},
{
"epoch": 1.34,
"learning_rate": 4.916684045881127e-05,
"loss": 1.0402,
"step": 1300
},
{
"epoch": 1.44,
"learning_rate": 4.906256517205422e-05,
"loss": 1.0445,
"step": 1400
},
{
"epoch": 1.55,
"learning_rate": 4.895933263816476e-05,
"loss": 1.0416,
"step": 1500
},
{
"epoch": 1.55,
"eval_loss": 0.21774140000343323,
"eval_runtime": 466.1858,
"eval_samples_per_second": 22.474,
"eval_steps_per_second": 2.81,
"eval_wer": 0.19455732182334984,
"step": 1500
},
{
"epoch": 1.65,
"learning_rate": 4.885505735140772e-05,
"loss": 1.039,
"step": 1600
},
{
"epoch": 1.75,
"learning_rate": 4.875078206465068e-05,
"loss": 1.0274,
"step": 1700
},
{
"epoch": 1.86,
"learning_rate": 4.864650677789364e-05,
"loss": 1.0116,
"step": 1800
},
{
"epoch": 1.96,
"learning_rate": 4.85422314911366e-05,
"loss": 1.0057,
"step": 1900
},
{
"epoch": 2.06,
"learning_rate": 4.843795620437956e-05,
"loss": 0.9951,
"step": 2000
},
{
"epoch": 2.06,
"eval_loss": 0.2008274346590042,
"eval_runtime": 468.5372,
"eval_samples_per_second": 22.361,
"eval_steps_per_second": 2.796,
"eval_wer": 0.17599747341825456,
"step": 2000
},
{
"epoch": 2.17,
"learning_rate": 4.8333680917622525e-05,
"loss": 0.9822,
"step": 2100
},
{
"epoch": 2.27,
"learning_rate": 4.822940563086549e-05,
"loss": 0.9847,
"step": 2200
},
{
"epoch": 2.37,
"learning_rate": 4.812513034410845e-05,
"loss": 0.9909,
"step": 2300
},
{
"epoch": 2.48,
"learning_rate": 4.802085505735141e-05,
"loss": 0.983,
"step": 2400
},
{
"epoch": 2.58,
"learning_rate": 4.7916579770594374e-05,
"loss": 0.975,
"step": 2500
},
{
"epoch": 2.58,
"eval_loss": 0.19611841440200806,
"eval_runtime": 467.2599,
"eval_samples_per_second": 22.422,
"eval_steps_per_second": 2.804,
"eval_wer": 0.17505000526371198,
"step": 2500
},
{
"epoch": 2.68,
"learning_rate": 4.7812304483837336e-05,
"loss": 0.9699,
"step": 2600
},
{
"epoch": 2.79,
"learning_rate": 4.77080291970803e-05,
"loss": 0.9829,
"step": 2700
},
{
"epoch": 2.89,
"learning_rate": 4.7603753910323254e-05,
"loss": 0.9676,
"step": 2800
},
{
"epoch": 2.99,
"learning_rate": 4.7499478623566216e-05,
"loss": 0.9824,
"step": 2900
},
{
"epoch": 3.1,
"learning_rate": 4.739520333680918e-05,
"loss": 0.9461,
"step": 3000
},
{
"epoch": 3.1,
"eval_loss": 0.19894707202911377,
"eval_runtime": 467.4994,
"eval_samples_per_second": 22.411,
"eval_steps_per_second": 2.802,
"eval_wer": 0.1782082324455206,
"step": 3000
},
{
"epoch": 3.2,
"learning_rate": 4.729092805005214e-05,
"loss": 0.9551,
"step": 3100
},
{
"epoch": 3.3,
"learning_rate": 4.71866527632951e-05,
"loss": 0.961,
"step": 3200
},
{
"epoch": 3.41,
"learning_rate": 4.7082377476538065e-05,
"loss": 0.9533,
"step": 3300
},
{
"epoch": 3.51,
"learning_rate": 4.697810218978103e-05,
"loss": 0.9522,
"step": 3400
},
{
"epoch": 3.61,
"learning_rate": 4.687382690302399e-05,
"loss": 0.9381,
"step": 3500
},
{
"epoch": 3.61,
"eval_loss": 0.1928330808877945,
"eval_runtime": 470.0863,
"eval_samples_per_second": 22.287,
"eval_steps_per_second": 2.787,
"eval_wer": 0.1699126223813033,
"step": 3500
},
{
"epoch": 3.72,
"learning_rate": 4.676955161626695e-05,
"loss": 0.9567,
"step": 3600
},
{
"epoch": 3.82,
"learning_rate": 4.666527632950991e-05,
"loss": 0.9431,
"step": 3700
},
{
"epoch": 3.92,
"learning_rate": 4.656100104275287e-05,
"loss": 0.9537,
"step": 3800
},
{
"epoch": 4.02,
"learning_rate": 4.645672575599583e-05,
"loss": 0.9193,
"step": 3900
},
{
"epoch": 4.13,
"learning_rate": 4.6352450469238794e-05,
"loss": 0.934,
"step": 4000
},
{
"epoch": 4.13,
"eval_loss": 0.1922750324010849,
"eval_runtime": 484.6325,
"eval_samples_per_second": 21.618,
"eval_steps_per_second": 2.703,
"eval_wer": 0.16326981787556585,
"step": 4000
},
{
"epoch": 4.23,
"learning_rate": 4.6248175182481756e-05,
"loss": 0.9179,
"step": 4100
},
{
"epoch": 4.33,
"learning_rate": 4.614389989572472e-05,
"loss": 0.9291,
"step": 4200
},
{
"epoch": 4.44,
"learning_rate": 4.6039624608967674e-05,
"loss": 0.9111,
"step": 4300
},
{
"epoch": 4.54,
"learning_rate": 4.5935349322210636e-05,
"loss": 0.924,
"step": 4400
},
{
"epoch": 4.64,
"learning_rate": 4.58310740354536e-05,
"loss": 0.9322,
"step": 4500
},
{
"epoch": 4.64,
"eval_loss": 0.1871235966682434,
"eval_runtime": 466.8923,
"eval_samples_per_second": 22.44,
"eval_steps_per_second": 2.806,
"eval_wer": 0.1634277292346563,
"step": 4500
},
{
"epoch": 4.75,
"learning_rate": 4.572679874869656e-05,
"loss": 0.936,
"step": 4600
},
{
"epoch": 4.85,
"learning_rate": 4.562252346193952e-05,
"loss": 0.926,
"step": 4700
},
{
"epoch": 4.95,
"learning_rate": 4.551824817518248e-05,
"loss": 0.9285,
"step": 4800
},
{
"epoch": 5.06,
"learning_rate": 4.541397288842544e-05,
"loss": 0.9293,
"step": 4900
},
{
"epoch": 5.16,
"learning_rate": 4.53096976016684e-05,
"loss": 0.9012,
"step": 5000
},
{
"epoch": 5.16,
"eval_loss": 0.1889733076095581,
"eval_runtime": 464.7576,
"eval_samples_per_second": 22.543,
"eval_steps_per_second": 2.819,
"eval_wer": 0.1701863354037267,
"step": 5000
},
{
"epoch": 5.26,
"learning_rate": 4.5205422314911365e-05,
"loss": 0.8889,
"step": 5100
},
{
"epoch": 5.37,
"learning_rate": 4.510114702815433e-05,
"loss": 0.907,
"step": 5200
},
{
"epoch": 5.47,
"learning_rate": 4.499687174139729e-05,
"loss": 0.9002,
"step": 5300
},
{
"epoch": 5.57,
"learning_rate": 4.489259645464025e-05,
"loss": 0.8984,
"step": 5400
},
{
"epoch": 5.68,
"learning_rate": 4.4788321167883214e-05,
"loss": 0.9045,
"step": 5500
},
{
"epoch": 5.68,
"eval_loss": 0.188216432929039,
"eval_runtime": 466.8913,
"eval_samples_per_second": 22.44,
"eval_steps_per_second": 2.806,
"eval_wer": 0.1739867354458364,
"step": 5500
},
{
"epoch": 5.78,
"learning_rate": 4.468404588112618e-05,
"loss": 0.8995,
"step": 5600
},
{
"epoch": 5.88,
"learning_rate": 4.457977059436914e-05,
"loss": 0.9113,
"step": 5700
},
{
"epoch": 5.99,
"learning_rate": 4.4475495307612094e-05,
"loss": 0.9044,
"step": 5800
},
{
"epoch": 6.09,
"learning_rate": 4.437122002085506e-05,
"loss": 0.8822,
"step": 5900
},
{
"epoch": 6.19,
"learning_rate": 4.426694473409802e-05,
"loss": 0.8826,
"step": 6000
},
{
"epoch": 6.19,
"eval_loss": 0.18564490973949432,
"eval_runtime": 465.7133,
"eval_samples_per_second": 22.497,
"eval_steps_per_second": 2.813,
"eval_wer": 0.15750078955679545,
"step": 6000
},
{
"epoch": 6.3,
"learning_rate": 4.416266944734098e-05,
"loss": 0.8986,
"step": 6100
},
{
"epoch": 6.4,
"learning_rate": 4.4058394160583944e-05,
"loss": 0.8936,
"step": 6200
},
{
"epoch": 6.5,
"learning_rate": 4.395516162669447e-05,
"loss": 0.889,
"step": 6300
},
{
"epoch": 6.6,
"learning_rate": 4.3850886339937434e-05,
"loss": 0.883,
"step": 6400
},
{
"epoch": 6.71,
"learning_rate": 4.37466110531804e-05,
"loss": 0.8848,
"step": 6500
},
{
"epoch": 6.71,
"eval_loss": 0.1860707700252533,
"eval_runtime": 479.2504,
"eval_samples_per_second": 21.861,
"eval_steps_per_second": 2.733,
"eval_wer": 0.16168017686072217,
"step": 6500
},
{
"epoch": 6.81,
"learning_rate": 4.364233576642336e-05,
"loss": 0.8986,
"step": 6600
},
{
"epoch": 6.91,
"learning_rate": 4.353806047966632e-05,
"loss": 0.8862,
"step": 6700
},
{
"epoch": 7.02,
"learning_rate": 4.3433785192909284e-05,
"loss": 0.8704,
"step": 6800
},
{
"epoch": 7.12,
"learning_rate": 4.3329509906152246e-05,
"loss": 0.8687,
"step": 6900
},
{
"epoch": 7.22,
"learning_rate": 4.322523461939521e-05,
"loss": 0.8723,
"step": 7000
},
{
"epoch": 7.22,
"eval_loss": 0.19270716607570648,
"eval_runtime": 470.8538,
"eval_samples_per_second": 22.251,
"eval_steps_per_second": 2.782,
"eval_wer": 0.16464891041162227,
"step": 7000
},
{
"epoch": 7.33,
"learning_rate": 4.3120959332638164e-05,
"loss": 0.8732,
"step": 7100
},
{
"epoch": 7.43,
"learning_rate": 4.3016684045881126e-05,
"loss": 0.8726,
"step": 7200
},
{
"epoch": 7.53,
"learning_rate": 4.291240875912409e-05,
"loss": 0.8728,
"step": 7300
},
{
"epoch": 7.64,
"learning_rate": 4.280813347236705e-05,
"loss": 0.8617,
"step": 7400
},
{
"epoch": 7.74,
"learning_rate": 4.270385818561001e-05,
"loss": 0.8725,
"step": 7500
},
{
"epoch": 7.74,
"eval_loss": 0.17976288497447968,
"eval_runtime": 466.2208,
"eval_samples_per_second": 22.472,
"eval_steps_per_second": 2.81,
"eval_wer": 0.15313190862196022,
"step": 7500
},
{
"epoch": 7.84,
"learning_rate": 4.2599582898852975e-05,
"loss": 0.8712,
"step": 7600
},
{
"epoch": 7.95,
"learning_rate": 4.249530761209594e-05,
"loss": 0.8639,
"step": 7700
},
{
"epoch": 8.05,
"learning_rate": 4.23910323253389e-05,
"loss": 0.8909,
"step": 7800
},
{
"epoch": 8.15,
"learning_rate": 4.228675703858186e-05,
"loss": 0.866,
"step": 7900
},
{
"epoch": 8.26,
"learning_rate": 4.2182481751824824e-05,
"loss": 0.8573,
"step": 8000
},
{
"epoch": 8.26,
"eval_loss": 0.17806729674339294,
"eval_runtime": 469.9178,
"eval_samples_per_second": 22.295,
"eval_steps_per_second": 2.788,
"eval_wer": 0.15872197073376146,
"step": 8000
},
{
"epoch": 8.36,
"learning_rate": 4.207820646506778e-05,
"loss": 0.8568,
"step": 8100
},
{
"epoch": 8.46,
"learning_rate": 4.197393117831074e-05,
"loss": 0.863,
"step": 8200
},
{
"epoch": 8.57,
"learning_rate": 4.1869655891553704e-05,
"loss": 0.859,
"step": 8300
},
{
"epoch": 8.67,
"learning_rate": 4.1765380604796666e-05,
"loss": 0.8374,
"step": 8400
},
{
"epoch": 8.77,
"learning_rate": 4.166110531803963e-05,
"loss": 0.8633,
"step": 8500
},
{
"epoch": 8.77,
"eval_loss": 0.18520139157772064,
"eval_runtime": 466.6826,
"eval_samples_per_second": 22.45,
"eval_steps_per_second": 2.807,
"eval_wer": 0.162838193494052,
"step": 8500
},
{
"epoch": 8.87,
"learning_rate": 4.155683003128259e-05,
"loss": 0.8656,
"step": 8600
},
{
"epoch": 8.98,
"learning_rate": 4.145255474452555e-05,
"loss": 0.8522,
"step": 8700
},
{
"epoch": 9.08,
"learning_rate": 4.1348279457768515e-05,
"loss": 0.8341,
"step": 8800
},
{
"epoch": 9.18,
"learning_rate": 4.124400417101148e-05,
"loss": 0.8484,
"step": 8900
},
{
"epoch": 9.29,
"learning_rate": 4.113972888425443e-05,
"loss": 0.8603,
"step": 9000
},
{
"epoch": 9.29,
"eval_loss": 0.1832893192768097,
"eval_runtime": 472.9551,
"eval_samples_per_second": 22.152,
"eval_steps_per_second": 2.77,
"eval_wer": 0.16013264554163595,
"step": 9000
},
{
"epoch": 9.39,
"learning_rate": 4.1035453597497395e-05,
"loss": 0.8473,
"step": 9100
},
{
"epoch": 9.49,
"learning_rate": 4.093117831074036e-05,
"loss": 0.8488,
"step": 9200
},
{
"epoch": 9.6,
"learning_rate": 4.082690302398332e-05,
"loss": 0.8492,
"step": 9300
},
{
"epoch": 9.7,
"learning_rate": 4.072367049009385e-05,
"loss": 0.8364,
"step": 9400
},
{
"epoch": 9.8,
"learning_rate": 4.061939520333681e-05,
"loss": 0.8421,
"step": 9500
},
{
"epoch": 9.8,
"eval_loss": 0.17876620590686798,
"eval_runtime": 467.77,
"eval_samples_per_second": 22.398,
"eval_steps_per_second": 2.801,
"eval_wer": 0.15428992525529003,
"step": 9500
},
{
"epoch": 9.91,
"learning_rate": 4.051511991657977e-05,
"loss": 0.8468,
"step": 9600
},
{
"epoch": 10.01,
"learning_rate": 4.0410844629822735e-05,
"loss": 0.8396,
"step": 9700
},
{
"epoch": 10.11,
"learning_rate": 4.03065693430657e-05,
"loss": 0.8346,
"step": 9800
},
{
"epoch": 10.22,
"learning_rate": 4.020229405630866e-05,
"loss": 0.8411,
"step": 9900
},
{
"epoch": 10.32,
"learning_rate": 4.009801876955162e-05,
"loss": 0.8404,
"step": 10000
},
{
"epoch": 10.32,
"eval_loss": 0.184404194355011,
"eval_runtime": 465.0709,
"eval_samples_per_second": 22.528,
"eval_steps_per_second": 2.817,
"eval_wer": 0.15555321612801348,
"step": 10000
},
{
"epoch": 10.42,
"learning_rate": 3.9993743482794584e-05,
"loss": 0.8382,
"step": 10100
},
{
"epoch": 10.53,
"learning_rate": 3.988946819603754e-05,
"loss": 0.8454,
"step": 10200
},
{
"epoch": 10.63,
"learning_rate": 3.97851929092805e-05,
"loss": 0.8472,
"step": 10300
},
{
"epoch": 10.73,
"learning_rate": 3.9680917622523464e-05,
"loss": 0.8273,
"step": 10400
},
{
"epoch": 10.84,
"learning_rate": 3.957664233576642e-05,
"loss": 0.8342,
"step": 10500
},
{
"epoch": 10.84,
"eval_loss": 0.17697852849960327,
"eval_runtime": 469.3266,
"eval_samples_per_second": 22.323,
"eval_steps_per_second": 2.791,
"eval_wer": 0.15379513633014003,
"step": 10500
},
{
"epoch": 10.94,
"learning_rate": 3.947236704900938e-05,
"loss": 0.8465,
"step": 10600
},
{
"epoch": 11.04,
"learning_rate": 3.9368091762252344e-05,
"loss": 0.8563,
"step": 10700
},
{
"epoch": 11.15,
"learning_rate": 3.9263816475495307e-05,
"loss": 0.823,
"step": 10800
},
{
"epoch": 11.25,
"learning_rate": 3.915954118873827e-05,
"loss": 0.8325,
"step": 10900
},
{
"epoch": 11.35,
"learning_rate": 3.905526590198123e-05,
"loss": 0.8161,
"step": 11000
},
{
"epoch": 11.35,
"eval_loss": 0.1820838451385498,
"eval_runtime": 468.1432,
"eval_samples_per_second": 22.38,
"eval_steps_per_second": 2.798,
"eval_wer": 0.1567112327613433,
"step": 11000
},
{
"epoch": 11.46,
"learning_rate": 3.895203336809176e-05,
"loss": 0.8227,
"step": 11100
},
{
"epoch": 11.56,
"learning_rate": 3.884775808133472e-05,
"loss": 0.8296,
"step": 11200
},
{
"epoch": 11.66,
"learning_rate": 3.8743482794577684e-05,
"loss": 0.8237,
"step": 11300
},
{
"epoch": 11.76,
"learning_rate": 3.8639207507820647e-05,
"loss": 0.8171,
"step": 11400
},
{
"epoch": 11.87,
"learning_rate": 3.853493222106361e-05,
"loss": 0.8371,
"step": 11500
},
{
"epoch": 11.87,
"eval_loss": 0.19089022278785706,
"eval_runtime": 465.3394,
"eval_samples_per_second": 22.515,
"eval_steps_per_second": 2.815,
"eval_wer": 0.16290135803768818,
"step": 11500
},
{
"epoch": 11.97,
"learning_rate": 3.843065693430657e-05,
"loss": 0.8331,
"step": 11600
},
{
"epoch": 12.07,
"learning_rate": 3.832638164754953e-05,
"loss": 0.8088,
"step": 11700
},
{
"epoch": 12.18,
"learning_rate": 3.8222106360792496e-05,
"loss": 0.8176,
"step": 11800
},
{
"epoch": 12.28,
"learning_rate": 3.811783107403545e-05,
"loss": 0.8034,
"step": 11900
},
{
"epoch": 12.38,
"learning_rate": 3.801355578727841e-05,
"loss": 0.8083,
"step": 12000
},
{
"epoch": 12.38,
"eval_loss": 0.17782799899578094,
"eval_runtime": 470.9072,
"eval_samples_per_second": 22.249,
"eval_steps_per_second": 2.782,
"eval_wer": 0.14979471523318244,
"step": 12000
},
{
"epoch": 12.49,
"learning_rate": 3.7909280500521376e-05,
"loss": 0.8128,
"step": 12100
},
{
"epoch": 12.59,
"learning_rate": 3.780500521376434e-05,
"loss": 0.8234,
"step": 12200
},
{
"epoch": 12.69,
"learning_rate": 3.77007299270073e-05,
"loss": 0.808,
"step": 12300
},
{
"epoch": 12.8,
"learning_rate": 3.759645464025026e-05,
"loss": 0.8225,
"step": 12400
},
{
"epoch": 12.9,
"learning_rate": 3.7492179353493225e-05,
"loss": 0.806,
"step": 12500
},
{
"epoch": 12.9,
"eval_loss": 0.18016664683818817,
"eval_runtime": 471.6501,
"eval_samples_per_second": 22.214,
"eval_steps_per_second": 2.777,
"eval_wer": 0.1547426044846826,
"step": 12500
},
{
"epoch": 13.0,
"learning_rate": 3.738790406673619e-05,
"loss": 0.8376,
"step": 12600
},
{
"epoch": 13.11,
"learning_rate": 3.728362877997915e-05,
"loss": 0.8087,
"step": 12700
},
{
"epoch": 13.21,
"learning_rate": 3.7179353493222105e-05,
"loss": 0.7941,
"step": 12800
},
{
"epoch": 13.31,
"learning_rate": 3.707507820646507e-05,
"loss": 0.8064,
"step": 12900
},
{
"epoch": 13.42,
"learning_rate": 3.69718456725756e-05,
"loss": 0.8013,
"step": 13000
},
{
"epoch": 13.42,
"eval_loss": 0.1859445571899414,
"eval_runtime": 467.0913,
"eval_samples_per_second": 22.43,
"eval_steps_per_second": 2.805,
"eval_wer": 0.15842720286345932,
"step": 13000
},
{
"epoch": 13.52,
"learning_rate": 3.6867570385818565e-05,
"loss": 0.8004,
"step": 13100
},
{
"epoch": 13.62,
"learning_rate": 3.676329509906152e-05,
"loss": 0.8201,
"step": 13200
},
{
"epoch": 13.73,
"learning_rate": 3.665901981230448e-05,
"loss": 0.8141,
"step": 13300
},
{
"epoch": 13.83,
"learning_rate": 3.6554744525547445e-05,
"loss": 0.8086,
"step": 13400
},
{
"epoch": 13.93,
"learning_rate": 3.645046923879041e-05,
"loss": 0.7913,
"step": 13500
},
{
"epoch": 13.93,
"eval_loss": 0.18746201694011688,
"eval_runtime": 466.4675,
"eval_samples_per_second": 22.46,
"eval_steps_per_second": 2.808,
"eval_wer": 0.15165806927044953,
"step": 13500
},
{
"epoch": 14.04,
"learning_rate": 3.634619395203337e-05,
"loss": 0.7865,
"step": 13600
},
{
"epoch": 14.14,
"learning_rate": 3.624191866527633e-05,
"loss": 0.7845,
"step": 13700
},
{
"epoch": 14.24,
"learning_rate": 3.6137643378519294e-05,
"loss": 0.8024,
"step": 13800
},
{
"epoch": 14.34,
"learning_rate": 3.6033368091762256e-05,
"loss": 0.8109,
"step": 13900
},
{
"epoch": 14.45,
"learning_rate": 3.592909280500522e-05,
"loss": 0.8063,
"step": 14000
},
{
"epoch": 14.45,
"eval_loss": 0.17992526292800903,
"eval_runtime": 465.4116,
"eval_samples_per_second": 22.511,
"eval_steps_per_second": 2.815,
"eval_wer": 0.15713232971891777,
"step": 14000
},
{
"epoch": 14.55,
"learning_rate": 3.582481751824818e-05,
"loss": 0.8058,
"step": 14100
},
{
"epoch": 14.65,
"learning_rate": 3.5720542231491136e-05,
"loss": 0.8093,
"step": 14200
},
{
"epoch": 14.76,
"learning_rate": 3.56162669447341e-05,
"loss": 0.7969,
"step": 14300
},
{
"epoch": 14.86,
"learning_rate": 3.551199165797706e-05,
"loss": 0.8031,
"step": 14400
},
{
"epoch": 14.96,
"learning_rate": 3.540771637122002e-05,
"loss": 0.7991,
"step": 14500
},
{
"epoch": 14.96,
"eval_loss": 0.17917674779891968,
"eval_runtime": 479.1894,
"eval_samples_per_second": 21.864,
"eval_steps_per_second": 2.734,
"eval_wer": 0.15384777344983683,
"step": 14500
},
{
"epoch": 15.07,
"learning_rate": 3.5303441084462985e-05,
"loss": 0.7973,
"step": 14600
},
{
"epoch": 15.17,
"learning_rate": 3.519916579770595e-05,
"loss": 0.7975,
"step": 14700
},
{
"epoch": 15.27,
"learning_rate": 3.509489051094891e-05,
"loss": 0.7785,
"step": 14800
},
{
"epoch": 15.38,
"learning_rate": 3.499061522419187e-05,
"loss": 0.8042,
"step": 14900
},
{
"epoch": 15.48,
"learning_rate": 3.4886339937434834e-05,
"loss": 0.7843,
"step": 15000
},
{
"epoch": 15.48,
"eval_loss": 0.17534053325653076,
"eval_runtime": 464.1652,
"eval_samples_per_second": 22.572,
"eval_steps_per_second": 2.822,
"eval_wer": 0.14635224760501106,
"step": 15000
},
{
"epoch": 15.58,
"learning_rate": 3.478206465067779e-05,
"loss": 0.7885,
"step": 15100
},
{
"epoch": 15.69,
"learning_rate": 3.467778936392075e-05,
"loss": 0.7886,
"step": 15200
},
{
"epoch": 15.79,
"learning_rate": 3.4573514077163714e-05,
"loss": 0.7995,
"step": 15300
},
{
"epoch": 15.89,
"learning_rate": 3.4469238790406676e-05,
"loss": 0.7894,
"step": 15400
},
{
"epoch": 16.0,
"learning_rate": 3.436496350364964e-05,
"loss": 0.7905,
"step": 15500
},
{
"epoch": 16.0,
"eval_loss": 0.1784074753522873,
"eval_runtime": 467.7669,
"eval_samples_per_second": 22.398,
"eval_steps_per_second": 2.801,
"eval_wer": 0.15082640277923992,
"step": 15500
},
{
"epoch": 16.1,
"learning_rate": 3.42606882168926e-05,
"loss": 0.7932,
"step": 15600
},
{
"epoch": 16.2,
"learning_rate": 3.415641293013556e-05,
"loss": 0.7954,
"step": 15700
},
{
"epoch": 16.31,
"learning_rate": 3.4052137643378525e-05,
"loss": 0.778,
"step": 15800
},
{
"epoch": 16.41,
"learning_rate": 3.394786235662149e-05,
"loss": 0.7963,
"step": 15900
},
{
"epoch": 16.51,
"learning_rate": 3.384358706986444e-05,
"loss": 0.7808,
"step": 16000
},
{
"epoch": 16.51,
"eval_loss": 0.17709574103355408,
"eval_runtime": 467.7784,
"eval_samples_per_second": 22.397,
"eval_steps_per_second": 2.8,
"eval_wer": 0.148531424360459,
"step": 16000
},
{
"epoch": 16.61,
"learning_rate": 3.3739311783107405e-05,
"loss": 0.7811,
"step": 16100
},
{
"epoch": 16.72,
"learning_rate": 3.363503649635036e-05,
"loss": 0.7732,
"step": 16200
},
{
"epoch": 16.82,
"learning_rate": 3.353076120959332e-05,
"loss": 0.7809,
"step": 16300
},
{
"epoch": 16.92,
"learning_rate": 3.3426485922836285e-05,
"loss": 0.7819,
"step": 16400
},
{
"epoch": 17.03,
"learning_rate": 3.332221063607925e-05,
"loss": 0.7743,
"step": 16500
},
{
"epoch": 17.03,
"eval_loss": 0.17948681116104126,
"eval_runtime": 462.7998,
"eval_samples_per_second": 22.638,
"eval_steps_per_second": 2.831,
"eval_wer": 0.14913148752500263,
"step": 16500
},
{
"epoch": 17.13,
"learning_rate": 3.321793534932221e-05,
"loss": 0.7669,
"step": 16600
},
{
"epoch": 17.23,
"learning_rate": 3.311366006256517e-05,
"loss": 0.7713,
"step": 16700
},
{
"epoch": 17.34,
"learning_rate": 3.3009384775808134e-05,
"loss": 0.7757,
"step": 16800
},
{
"epoch": 17.44,
"learning_rate": 3.29051094890511e-05,
"loss": 0.7736,
"step": 16900
},
{
"epoch": 17.54,
"learning_rate": 3.280083420229406e-05,
"loss": 0.7833,
"step": 17000
},
{
"epoch": 17.54,
"eval_loss": 0.1721695214509964,
"eval_runtime": 464.6724,
"eval_samples_per_second": 22.547,
"eval_steps_per_second": 2.819,
"eval_wer": 0.14838404042530792,
"step": 17000
},
{
"epoch": 17.65,
"learning_rate": 3.2696558915537014e-05,
"loss": 0.7752,
"step": 17100
},
{
"epoch": 17.75,
"learning_rate": 3.259228362877998e-05,
"loss": 0.7772,
"step": 17200
},
{
"epoch": 17.85,
"learning_rate": 3.248800834202294e-05,
"loss": 0.7805,
"step": 17300
},
{
"epoch": 17.96,
"learning_rate": 3.23837330552659e-05,
"loss": 0.7744,
"step": 17400
},
{
"epoch": 18.06,
"learning_rate": 3.2279457768508863e-05,
"loss": 0.7763,
"step": 17500
},
{
"epoch": 18.06,
"eval_loss": 0.1766623854637146,
"eval_runtime": 465.5186,
"eval_samples_per_second": 22.506,
"eval_steps_per_second": 2.814,
"eval_wer": 0.15183703547741867,
"step": 17500
},
{
"epoch": 18.16,
"learning_rate": 3.2175182481751826e-05,
"loss": 0.7705,
"step": 17600
},
{
"epoch": 18.27,
"learning_rate": 3.207090719499479e-05,
"loss": 0.7576,
"step": 17700
},
{
"epoch": 18.37,
"learning_rate": 3.196663190823775e-05,
"loss": 0.7605,
"step": 17800
},
{
"epoch": 18.47,
"learning_rate": 3.186235662148071e-05,
"loss": 0.7672,
"step": 17900
},
{
"epoch": 18.58,
"learning_rate": 3.1758081334723675e-05,
"loss": 0.7698,
"step": 18000
},
{
"epoch": 18.58,
"eval_loss": 0.17199581861495972,
"eval_runtime": 465.8946,
"eval_samples_per_second": 22.488,
"eval_steps_per_second": 2.812,
"eval_wer": 0.14599431519107275,
"step": 18000
},
{
"epoch": 18.68,
"learning_rate": 3.1654848800834204e-05,
"loss": 0.7677,
"step": 18100
},
{
"epoch": 18.78,
"learning_rate": 3.1550573514077166e-05,
"loss": 0.7732,
"step": 18200
},
{
"epoch": 18.89,
"learning_rate": 3.144629822732013e-05,
"loss": 0.7687,
"step": 18300
},
{
"epoch": 18.99,
"learning_rate": 3.134202294056309e-05,
"loss": 0.7687,
"step": 18400
},
{
"epoch": 19.09,
"learning_rate": 3.123879040667362e-05,
"loss": 0.7571,
"step": 18500
},
{
"epoch": 19.09,
"eval_loss": 0.1734933853149414,
"eval_runtime": 481.2248,
"eval_samples_per_second": 21.772,
"eval_steps_per_second": 2.722,
"eval_wer": 0.14779450468470365,
"step": 18500
},
{
"epoch": 19.2,
"learning_rate": 3.113451511991658e-05,
"loss": 0.7649,
"step": 18600
},
{
"epoch": 19.3,
"learning_rate": 3.1030239833159544e-05,
"loss": 0.7691,
"step": 18700
},
{
"epoch": 19.4,
"learning_rate": 3.0925964546402506e-05,
"loss": 0.7709,
"step": 18800
},
{
"epoch": 19.5,
"learning_rate": 3.082168925964546e-05,
"loss": 0.7758,
"step": 18900
},
{
"epoch": 19.61,
"learning_rate": 3.0717413972888424e-05,
"loss": 0.7673,
"step": 19000
},
{
"epoch": 19.61,
"eval_loss": 0.18169456720352173,
"eval_runtime": 475.7022,
"eval_samples_per_second": 22.024,
"eval_steps_per_second": 2.754,
"eval_wer": 0.15106853352984526,
"step": 19000
},
{
"epoch": 19.71,
"learning_rate": 3.0613138686131386e-05,
"loss": 0.7466,
"step": 19100
},
{
"epoch": 19.81,
"learning_rate": 3.0508863399374348e-05,
"loss": 0.7556,
"step": 19200
},
{
"epoch": 19.92,
"learning_rate": 3.040458811261731e-05,
"loss": 0.7424,
"step": 19300
},
{
"epoch": 20.02,
"learning_rate": 3.0300312825860273e-05,
"loss": 0.7455,
"step": 19400
},
{
"epoch": 20.12,
"learning_rate": 3.0196037539103235e-05,
"loss": 0.7415,
"step": 19500
},
{
"epoch": 20.12,
"eval_loss": 0.17630091309547424,
"eval_runtime": 467.192,
"eval_samples_per_second": 22.425,
"eval_steps_per_second": 2.804,
"eval_wer": 0.14813138225076325,
"step": 19500
},
{
"epoch": 20.23,
"learning_rate": 3.0091762252346194e-05,
"loss": 0.7479,
"step": 19600
},
{
"epoch": 20.33,
"learning_rate": 2.9987486965589156e-05,
"loss": 0.7506,
"step": 19700
},
{
"epoch": 20.43,
"learning_rate": 2.9883211678832118e-05,
"loss": 0.7523,
"step": 19800
},
{
"epoch": 20.54,
"learning_rate": 2.977893639207508e-05,
"loss": 0.7646,
"step": 19900
},
{
"epoch": 20.64,
"learning_rate": 2.9674661105318043e-05,
"loss": 0.751,
"step": 20000
},
{
"epoch": 20.64,
"eval_loss": 0.1741507649421692,
"eval_runtime": 472.297,
"eval_samples_per_second": 22.183,
"eval_steps_per_second": 2.774,
"eval_wer": 0.1483629855774292,
"step": 20000
},
{
"epoch": 20.74,
"learning_rate": 2.9570385818561e-05,
"loss": 0.7533,
"step": 20100
},
{
"epoch": 20.85,
"learning_rate": 2.9466110531803964e-05,
"loss": 0.7505,
"step": 20200
},
{
"epoch": 20.95,
"learning_rate": 2.9361835245046926e-05,
"loss": 0.7481,
"step": 20300
},
{
"epoch": 21.05,
"learning_rate": 2.925755995828989e-05,
"loss": 0.7741,
"step": 20400
},
{
"epoch": 21.16,
"learning_rate": 2.9153284671532847e-05,
"loss": 0.7563,
"step": 20500
},
{
"epoch": 21.16,
"eval_loss": 0.18097689747810364,
"eval_runtime": 466.3942,
"eval_samples_per_second": 22.464,
"eval_steps_per_second": 2.809,
"eval_wer": 0.16113275081587536,
"step": 20500
},
{
"epoch": 21.26,
"learning_rate": 2.904900938477581e-05,
"loss": 0.755,
"step": 20600
},
{
"epoch": 21.36,
"learning_rate": 2.8944734098018772e-05,
"loss": 0.7452,
"step": 20700
},
{
"epoch": 21.47,
"learning_rate": 2.8840458811261734e-05,
"loss": 0.7559,
"step": 20800
},
{
"epoch": 21.57,
"learning_rate": 2.8736183524504696e-05,
"loss": 0.7326,
"step": 20900
},
{
"epoch": 21.67,
"learning_rate": 2.8631908237747655e-05,
"loss": 0.7423,
"step": 21000
},
{
"epoch": 21.67,
"eval_loss": 0.1817416250705719,
"eval_runtime": 467.2447,
"eval_samples_per_second": 22.423,
"eval_steps_per_second": 2.804,
"eval_wer": 0.15571112748710392,
"step": 21000
},
{
"epoch": 21.77,
"learning_rate": 2.8527632950990617e-05,
"loss": 0.744,
"step": 21100
},
{
"epoch": 21.88,
"learning_rate": 2.842335766423358e-05,
"loss": 0.7539,
"step": 21200
},
{
"epoch": 21.98,
"learning_rate": 2.8319082377476542e-05,
"loss": 0.7502,
"step": 21300
},
{
"epoch": 22.08,
"learning_rate": 2.8214807090719504e-05,
"loss": 0.7374,
"step": 21400
},
{
"epoch": 22.19,
"learning_rate": 2.8110531803962463e-05,
"loss": 0.7242,
"step": 21500
},
{
"epoch": 22.19,
"eval_loss": 0.1689848154783249,
"eval_runtime": 467.3401,
"eval_samples_per_second": 22.418,
"eval_steps_per_second": 2.803,
"eval_wer": 0.14457311295925887,
"step": 21500
},
{
"epoch": 22.29,
"learning_rate": 2.8006256517205425e-05,
"loss": 0.7461,
"step": 21600
},
{
"epoch": 22.39,
"learning_rate": 2.7901981230448388e-05,
"loss": 0.7269,
"step": 21700
},
{
"epoch": 22.5,
"learning_rate": 2.779770594369135e-05,
"loss": 0.744,
"step": 21800
},
{
"epoch": 22.6,
"learning_rate": 2.7693430656934312e-05,
"loss": 0.7437,
"step": 21900
},
{
"epoch": 22.7,
"learning_rate": 2.759019812304484e-05,
"loss": 0.7251,
"step": 22000
},
{
"epoch": 22.7,
"eval_loss": 0.16843970119953156,
"eval_runtime": 466.906,
"eval_samples_per_second": 22.439,
"eval_steps_per_second": 2.806,
"eval_wer": 0.14457311295925887,
"step": 22000
},
{
"epoch": 22.81,
"learning_rate": 2.7485922836287803e-05,
"loss": 0.7233,
"step": 22100
},
{
"epoch": 22.91,
"learning_rate": 2.7381647549530765e-05,
"loss": 0.7169,
"step": 22200
},
{
"epoch": 23.01,
"learning_rate": 2.7277372262773728e-05,
"loss": 0.7374,
"step": 22300
},
{
"epoch": 23.12,
"learning_rate": 2.7173096976016687e-05,
"loss": 0.7289,
"step": 22400
},
{
"epoch": 23.22,
"learning_rate": 2.706882168925965e-05,
"loss": 0.7302,
"step": 22500
},
{
"epoch": 23.22,
"eval_loss": 0.17354781925678253,
"eval_runtime": 469.7694,
"eval_samples_per_second": 22.302,
"eval_steps_per_second": 2.789,
"eval_wer": 0.14295188967259712,
"step": 22500
},
{
"epoch": 23.32,
"learning_rate": 2.696454640250261e-05,
"loss": 0.7307,
"step": 22600
},
{
"epoch": 23.43,
"learning_rate": 2.6860271115745573e-05,
"loss": 0.7343,
"step": 22700
},
{
"epoch": 23.53,
"learning_rate": 2.6755995828988532e-05,
"loss": 0.7346,
"step": 22800
},
{
"epoch": 23.63,
"learning_rate": 2.665172054223149e-05,
"loss": 0.7349,
"step": 22900
},
{
"epoch": 23.74,
"learning_rate": 2.654744525547445e-05,
"loss": 0.733,
"step": 23000
},
{
"epoch": 23.74,
"eval_loss": 0.17200088500976562,
"eval_runtime": 479.405,
"eval_samples_per_second": 21.854,
"eval_steps_per_second": 2.733,
"eval_wer": 0.14537319717865038,
"step": 23000
},
{
"epoch": 23.84,
"learning_rate": 2.6443169968717412e-05,
"loss": 0.7285,
"step": 23100
},
{
"epoch": 23.94,
"learning_rate": 2.6338894681960374e-05,
"loss": 0.7238,
"step": 23200
},
{
"epoch": 24.05,
"learning_rate": 2.6234619395203337e-05,
"loss": 0.7355,
"step": 23300
},
{
"epoch": 24.15,
"learning_rate": 2.61303441084463e-05,
"loss": 0.725,
"step": 23400
},
{
"epoch": 24.25,
"learning_rate": 2.6026068821689258e-05,
"loss": 0.7128,
"step": 23500
},
{
"epoch": 24.25,
"eval_loss": 0.16681526601314545,
"eval_runtime": 465.5348,
"eval_samples_per_second": 22.505,
"eval_steps_per_second": 2.814,
"eval_wer": 0.1382987682913991,
"step": 23500
},
{
"epoch": 24.35,
"learning_rate": 2.592179353493222e-05,
"loss": 0.7104,
"step": 23600
},
{
"epoch": 24.46,
"learning_rate": 2.5817518248175182e-05,
"loss": 0.7139,
"step": 23700
},
{
"epoch": 24.56,
"learning_rate": 2.5713242961418145e-05,
"loss": 0.7086,
"step": 23800
},
{
"epoch": 24.66,
"learning_rate": 2.5608967674661103e-05,
"loss": 0.7124,
"step": 23900
},
{
"epoch": 24.77,
"learning_rate": 2.5504692387904066e-05,
"loss": 0.7184,
"step": 24000
},
{
"epoch": 24.77,
"eval_loss": 0.16353647410869598,
"eval_runtime": 467.6672,
"eval_samples_per_second": 22.403,
"eval_steps_per_second": 2.801,
"eval_wer": 0.1377408148226129,
"step": 24000
},
{
"epoch": 24.87,
"learning_rate": 2.5400417101147028e-05,
"loss": 0.7174,
"step": 24100
},
{
"epoch": 24.97,
"learning_rate": 2.529614181438999e-05,
"loss": 0.7164,
"step": 24200
},
{
"epoch": 25.08,
"learning_rate": 2.5191866527632953e-05,
"loss": 0.6949,
"step": 24300
},
{
"epoch": 25.18,
"learning_rate": 2.508759124087591e-05,
"loss": 0.7121,
"step": 24400
},
{
"epoch": 25.28,
"learning_rate": 2.4983315954118874e-05,
"loss": 0.7015,
"step": 24500
},
{
"epoch": 25.28,
"eval_loss": 0.1646386831998825,
"eval_runtime": 470.6383,
"eval_samples_per_second": 22.261,
"eval_steps_per_second": 2.783,
"eval_wer": 0.1389409411517002,
"step": 24500
},
{
"epoch": 25.39,
"learning_rate": 2.4879040667361836e-05,
"loss": 0.7104,
"step": 24600
},
{
"epoch": 25.49,
"learning_rate": 2.4774765380604798e-05,
"loss": 0.7083,
"step": 24700
},
{
"epoch": 25.59,
"learning_rate": 2.467049009384776e-05,
"loss": 0.7085,
"step": 24800
},
{
"epoch": 25.7,
"learning_rate": 2.456621480709072e-05,
"loss": 0.7215,
"step": 24900
},
{
"epoch": 25.8,
"learning_rate": 2.446193952033368e-05,
"loss": 0.7198,
"step": 25000
},
{
"epoch": 25.8,
"eval_loss": 0.17749331891536713,
"eval_runtime": 481.0207,
"eval_samples_per_second": 21.781,
"eval_steps_per_second": 2.723,
"eval_wer": 0.14620486366986,
"step": 25000
},
{
"epoch": 25.9,
"learning_rate": 2.4357664233576644e-05,
"loss": 0.7108,
"step": 25100
},
{
"epoch": 26.01,
"learning_rate": 2.4253388946819606e-05,
"loss": 0.7197,
"step": 25200
},
{
"epoch": 26.11,
"learning_rate": 2.414911366006257e-05,
"loss": 0.6984,
"step": 25300
},
{
"epoch": 26.21,
"learning_rate": 2.4044838373305527e-05,
"loss": 0.6976,
"step": 25400
},
{
"epoch": 26.32,
"learning_rate": 2.394056308654849e-05,
"loss": 0.7178,
"step": 25500
},
{
"epoch": 26.32,
"eval_loss": 0.17049279808998108,
"eval_runtime": 463.0645,
"eval_samples_per_second": 22.625,
"eval_steps_per_second": 2.829,
"eval_wer": 0.1418780924307822,
"step": 25500
},
{
"epoch": 26.42,
"learning_rate": 2.3836287799791452e-05,
"loss": 0.7164,
"step": 25600
},
{
"epoch": 26.52,
"learning_rate": 2.3732012513034414e-05,
"loss": 0.7322,
"step": 25700
},
{
"epoch": 26.63,
"learning_rate": 2.3627737226277373e-05,
"loss": 0.7158,
"step": 25800
},
{
"epoch": 26.73,
"learning_rate": 2.3523461939520335e-05,
"loss": 0.7146,
"step": 25900
},
{
"epoch": 26.83,
"learning_rate": 2.3419186652763297e-05,
"loss": 0.7199,
"step": 26000
},
{
"epoch": 26.83,
"eval_loss": 0.16493017971515656,
"eval_runtime": 466.7445,
"eval_samples_per_second": 22.447,
"eval_steps_per_second": 2.807,
"eval_wer": 0.14164648910411623,
"step": 26000
},
{
"epoch": 26.93,
"learning_rate": 2.331595411887383e-05,
"loss": 0.7001,
"step": 26100
},
{
"epoch": 27.04,
"learning_rate": 2.321167883211679e-05,
"loss": 0.6918,
"step": 26200
},
{
"epoch": 27.14,
"learning_rate": 2.310740354535975e-05,
"loss": 0.6864,
"step": 26300
},
{
"epoch": 27.24,
"learning_rate": 2.3003128258602713e-05,
"loss": 0.7008,
"step": 26400
},
{
"epoch": 27.35,
"learning_rate": 2.2898852971845675e-05,
"loss": 0.6981,
"step": 26500
},
{
"epoch": 27.35,
"eval_loss": 0.1723548024892807,
"eval_runtime": 467.3195,
"eval_samples_per_second": 22.419,
"eval_steps_per_second": 2.803,
"eval_wer": 0.14182545531108537,
"step": 26500
},
{
"epoch": 27.45,
"learning_rate": 2.2794577685088637e-05,
"loss": 0.7041,
"step": 26600
},
{
"epoch": 27.55,
"learning_rate": 2.2690302398331596e-05,
"loss": 0.6864,
"step": 26700
},
{
"epoch": 27.66,
"learning_rate": 2.258602711157456e-05,
"loss": 0.6989,
"step": 26800
},
{
"epoch": 27.76,
"learning_rate": 2.2481751824817517e-05,
"loss": 0.687,
"step": 26900
},
{
"epoch": 27.86,
"learning_rate": 2.237747653806048e-05,
"loss": 0.6886,
"step": 27000
},
{
"epoch": 27.86,
"eval_loss": 0.16325753927230835,
"eval_runtime": 468.8775,
"eval_samples_per_second": 22.345,
"eval_steps_per_second": 2.794,
"eval_wer": 0.13823560374776292,
"step": 27000
},
{
"epoch": 27.97,
"learning_rate": 2.2273201251303442e-05,
"loss": 0.7044,
"step": 27100
},
{
"epoch": 28.07,
"learning_rate": 2.2168925964546404e-05,
"loss": 0.6713,
"step": 27200
},
{
"epoch": 28.17,
"learning_rate": 2.2064650677789363e-05,
"loss": 0.6889,
"step": 27300
},
{
"epoch": 28.28,
"learning_rate": 2.1960375391032325e-05,
"loss": 0.6769,
"step": 27400
},
{
"epoch": 28.38,
"learning_rate": 2.185714285714286e-05,
"loss": 0.6922,
"step": 27500
},
{
"epoch": 28.38,
"eval_loss": 0.16981077194213867,
"eval_runtime": 479.6579,
"eval_samples_per_second": 21.843,
"eval_steps_per_second": 2.731,
"eval_wer": 0.14196231182229707,
"step": 27500
},
{
"epoch": 28.48,
"learning_rate": 2.175286757038582e-05,
"loss": 0.6962,
"step": 27600
},
{
"epoch": 28.59,
"learning_rate": 2.164859228362878e-05,
"loss": 0.6907,
"step": 27700
},
{
"epoch": 28.69,
"learning_rate": 2.154431699687174e-05,
"loss": 0.6967,
"step": 27800
},
{
"epoch": 28.79,
"learning_rate": 2.1440041710114703e-05,
"loss": 0.6699,
"step": 27900
},
{
"epoch": 28.9,
"learning_rate": 2.1335766423357665e-05,
"loss": 0.6833,
"step": 28000
},
{
"epoch": 28.9,
"eval_loss": 0.1610790491104126,
"eval_runtime": 471.6243,
"eval_samples_per_second": 22.215,
"eval_steps_per_second": 2.778,
"eval_wer": 0.13509843141383304,
"step": 28000
},
{
"epoch": 29.0,
"learning_rate": 2.1231491136600624e-05,
"loss": 0.694,
"step": 28100
},
{
"epoch": 29.1,
"learning_rate": 2.1127215849843587e-05,
"loss": 0.6839,
"step": 28200
},
{
"epoch": 29.21,
"learning_rate": 2.102294056308655e-05,
"loss": 0.6791,
"step": 28300
},
{
"epoch": 29.31,
"learning_rate": 2.091866527632951e-05,
"loss": 0.6855,
"step": 28400
},
{
"epoch": 29.41,
"learning_rate": 2.0814389989572473e-05,
"loss": 0.6798,
"step": 28500
},
{
"epoch": 29.41,
"eval_loss": 0.16391155123710632,
"eval_runtime": 478.5296,
"eval_samples_per_second": 21.894,
"eval_steps_per_second": 2.738,
"eval_wer": 0.1364669965259501,
"step": 28500
},
{
"epoch": 29.51,
"learning_rate": 2.0710114702815432e-05,
"loss": 0.6755,
"step": 28600
},
{
"epoch": 29.62,
"learning_rate": 2.0605839416058394e-05,
"loss": 0.6892,
"step": 28700
},
{
"epoch": 29.72,
"learning_rate": 2.0501564129301357e-05,
"loss": 0.6883,
"step": 28800
},
{
"epoch": 29.82,
"learning_rate": 2.039728884254432e-05,
"loss": 0.685,
"step": 28900
},
{
"epoch": 29.93,
"learning_rate": 2.029301355578728e-05,
"loss": 0.6711,
"step": 29000
},
{
"epoch": 29.93,
"eval_loss": 0.16675817966461182,
"eval_runtime": 467.8784,
"eval_samples_per_second": 22.393,
"eval_steps_per_second": 2.8,
"eval_wer": 0.13579324139383092,
"step": 29000
},
{
"epoch": 30.03,
"learning_rate": 2.018873826903024e-05,
"loss": 0.6585,
"step": 29100
},
{
"epoch": 30.13,
"learning_rate": 2.0084462982273202e-05,
"loss": 0.6604,
"step": 29200
},
{
"epoch": 30.24,
"learning_rate": 1.9980187695516165e-05,
"loss": 0.6662,
"step": 29300
},
{
"epoch": 30.34,
"learning_rate": 1.9875912408759127e-05,
"loss": 0.6639,
"step": 29400
},
{
"epoch": 30.44,
"learning_rate": 1.9771637122002086e-05,
"loss": 0.6762,
"step": 29500
},
{
"epoch": 30.44,
"eval_loss": 0.1681944876909256,
"eval_runtime": 470.2202,
"eval_samples_per_second": 22.281,
"eval_steps_per_second": 2.786,
"eval_wer": 0.13550900094746815,
"step": 29500
},
{
"epoch": 30.55,
"learning_rate": 1.9667361835245048e-05,
"loss": 0.6732,
"step": 29600
},
{
"epoch": 30.65,
"learning_rate": 1.956308654848801e-05,
"loss": 0.6729,
"step": 29700
},
{
"epoch": 30.75,
"learning_rate": 1.9458811261730973e-05,
"loss": 0.6591,
"step": 29800
},
{
"epoch": 30.86,
"learning_rate": 1.935453597497393e-05,
"loss": 0.6612,
"step": 29900
},
{
"epoch": 30.96,
"learning_rate": 1.9250260688216894e-05,
"loss": 0.6594,
"step": 30000
},
{
"epoch": 30.96,
"eval_loss": 0.16289828717708588,
"eval_runtime": 468.3015,
"eval_samples_per_second": 22.372,
"eval_steps_per_second": 2.797,
"eval_wer": 0.13452995052110747,
"step": 30000
},
{
"epoch": 31.06,
"learning_rate": 1.9145985401459853e-05,
"loss": 0.6624,
"step": 30100
},
{
"epoch": 31.17,
"learning_rate": 1.9041710114702815e-05,
"loss": 0.6696,
"step": 30200
},
{
"epoch": 31.27,
"learning_rate": 1.8937434827945777e-05,
"loss": 0.6672,
"step": 30300
},
{
"epoch": 31.37,
"learning_rate": 1.883315954118874e-05,
"loss": 0.666,
"step": 30400
},
{
"epoch": 31.48,
"learning_rate": 1.87288842544317e-05,
"loss": 0.6664,
"step": 30500
},
{
"epoch": 31.48,
"eval_loss": 0.16248419880867004,
"eval_runtime": 467.0084,
"eval_samples_per_second": 22.434,
"eval_steps_per_second": 2.805,
"eval_wer": 0.1321402252868723,
"step": 30500
},
{
"epoch": 31.58,
"learning_rate": 1.862460896767466e-05,
"loss": 0.6811,
"step": 30600
},
{
"epoch": 31.68,
"learning_rate": 1.8520333680917623e-05,
"loss": 0.653,
"step": 30700
},
{
"epoch": 31.79,
"learning_rate": 1.8416058394160585e-05,
"loss": 0.6717,
"step": 30800
},
{
"epoch": 31.89,
"learning_rate": 1.8311783107403547e-05,
"loss": 0.6654,
"step": 30900
},
{
"epoch": 31.99,
"learning_rate": 1.8207507820646506e-05,
"loss": 0.6838,
"step": 31000
},
{
"epoch": 31.99,
"eval_loss": 0.15968269109725952,
"eval_runtime": 469.2514,
"eval_samples_per_second": 22.327,
"eval_steps_per_second": 2.792,
"eval_wer": 0.13722497104958417,
"step": 31000
},
{
"epoch": 32.09,
"learning_rate": 1.810323253388947e-05,
"loss": 0.6643,
"step": 31100
},
{
"epoch": 32.2,
"learning_rate": 1.799895724713243e-05,
"loss": 0.6728,
"step": 31200
},
{
"epoch": 32.3,
"learning_rate": 1.7894681960375393e-05,
"loss": 0.6599,
"step": 31300
},
{
"epoch": 32.4,
"learning_rate": 1.7790406673618355e-05,
"loss": 0.6645,
"step": 31400
},
{
"epoch": 32.51,
"learning_rate": 1.7686131386861314e-05,
"loss": 0.6603,
"step": 31500
},
{
"epoch": 32.51,
"eval_loss": 0.1583111584186554,
"eval_runtime": 484.3394,
"eval_samples_per_second": 21.632,
"eval_steps_per_second": 2.705,
"eval_wer": 0.13024528897778714,
"step": 31500
},
{
"epoch": 32.61,
"learning_rate": 1.7581856100104276e-05,
"loss": 0.6659,
"step": 31600
},
{
"epoch": 32.71,
"learning_rate": 1.747758081334724e-05,
"loss": 0.6551,
"step": 31700
},
{
"epoch": 32.82,
"learning_rate": 1.73733055265902e-05,
"loss": 0.6581,
"step": 31800
},
{
"epoch": 32.92,
"learning_rate": 1.7269030239833163e-05,
"loss": 0.6526,
"step": 31900
},
{
"epoch": 33.02,
"learning_rate": 1.7164754953076122e-05,
"loss": 0.6468,
"step": 32000
},
{
"epoch": 33.02,
"eval_loss": 0.15948732197284698,
"eval_runtime": 467.9806,
"eval_samples_per_second": 22.388,
"eval_steps_per_second": 2.799,
"eval_wer": 0.13224549952626594,
"step": 32000
},
{
"epoch": 33.13,
"learning_rate": 1.706047966631908e-05,
"loss": 0.642,
"step": 32100
},
{
"epoch": 33.23,
"learning_rate": 1.6956204379562043e-05,
"loss": 0.6546,
"step": 32200
},
{
"epoch": 33.33,
"learning_rate": 1.685297184567258e-05,
"loss": 0.6438,
"step": 32300
},
{
"epoch": 33.44,
"learning_rate": 1.6748696558915537e-05,
"loss": 0.6466,
"step": 32400
},
{
"epoch": 33.54,
"learning_rate": 1.66444212721585e-05,
"loss": 0.6464,
"step": 32500
},
{
"epoch": 33.54,
"eval_loss": 0.16093206405639648,
"eval_runtime": 468.2076,
"eval_samples_per_second": 22.377,
"eval_steps_per_second": 2.798,
"eval_wer": 0.13148752500263186,
"step": 32500
},
{
"epoch": 33.64,
"learning_rate": 1.6540145985401462e-05,
"loss": 0.6528,
"step": 32600
},
{
"epoch": 33.75,
"learning_rate": 1.6435870698644424e-05,
"loss": 0.641,
"step": 32700
},
{
"epoch": 33.85,
"learning_rate": 1.6331595411887386e-05,
"loss": 0.6416,
"step": 32800
},
{
"epoch": 33.95,
"learning_rate": 1.6227320125130342e-05,
"loss": 0.6398,
"step": 32900
},
{
"epoch": 34.06,
"learning_rate": 1.6123044838373304e-05,
"loss": 0.6623,
"step": 33000
},
{
"epoch": 34.06,
"eval_loss": 0.16215749084949493,
"eval_runtime": 466.7215,
"eval_samples_per_second": 22.448,
"eval_steps_per_second": 2.807,
"eval_wer": 0.13661438046110116,
"step": 33000
},
{
"epoch": 34.16,
"learning_rate": 1.6018769551616266e-05,
"loss": 0.6433,
"step": 33100
},
{
"epoch": 34.26,
"learning_rate": 1.591449426485923e-05,
"loss": 0.6484,
"step": 33200
},
{
"epoch": 34.37,
"learning_rate": 1.581021897810219e-05,
"loss": 0.6483,
"step": 33300
},
{
"epoch": 34.47,
"learning_rate": 1.570594369134515e-05,
"loss": 0.6439,
"step": 33400
},
{
"epoch": 34.57,
"learning_rate": 1.5601668404588112e-05,
"loss": 0.6414,
"step": 33500
},
{
"epoch": 34.57,
"eval_loss": 0.15873852372169495,
"eval_runtime": 469.8368,
"eval_samples_per_second": 22.299,
"eval_steps_per_second": 2.788,
"eval_wer": 0.13302452889777872,
"step": 33500
},
{
"epoch": 34.67,
"learning_rate": 1.5497393117831074e-05,
"loss": 0.6458,
"step": 33600
},
{
"epoch": 34.78,
"learning_rate": 1.5393117831074037e-05,
"loss": 0.6423,
"step": 33700
},
{
"epoch": 34.88,
"learning_rate": 1.5288842544317e-05,
"loss": 0.6416,
"step": 33800
},
{
"epoch": 34.98,
"learning_rate": 1.518456725755996e-05,
"loss": 0.6481,
"step": 33900
},
{
"epoch": 35.09,
"learning_rate": 1.508029197080292e-05,
"loss": 0.6242,
"step": 34000
},
{
"epoch": 35.09,
"eval_loss": 0.16140839457511902,
"eval_runtime": 481.1715,
"eval_samples_per_second": 21.774,
"eval_steps_per_second": 2.723,
"eval_wer": 0.1336982840298979,
"step": 34000
},
{
"epoch": 35.19,
"learning_rate": 1.4976016684045882e-05,
"loss": 0.6335,
"step": 34100
},
{
"epoch": 35.29,
"learning_rate": 1.4871741397288843e-05,
"loss": 0.6365,
"step": 34200
},
{
"epoch": 35.4,
"learning_rate": 1.4767466110531805e-05,
"loss": 0.6392,
"step": 34300
},
{
"epoch": 35.5,
"learning_rate": 1.4663190823774767e-05,
"loss": 0.6355,
"step": 34400
},
{
"epoch": 35.6,
"learning_rate": 1.4558915537017728e-05,
"loss": 0.632,
"step": 34500
},
{
"epoch": 35.6,
"eval_loss": 0.15684087574481964,
"eval_runtime": 469.3002,
"eval_samples_per_second": 22.325,
"eval_steps_per_second": 2.791,
"eval_wer": 0.12724497315506894,
"step": 34500
},
{
"epoch": 35.71,
"learning_rate": 1.4455683003128258e-05,
"loss": 0.6337,
"step": 34600
},
{
"epoch": 35.81,
"learning_rate": 1.435140771637122e-05,
"loss": 0.6304,
"step": 34700
},
{
"epoch": 35.91,
"learning_rate": 1.4247132429614183e-05,
"loss": 0.6398,
"step": 34800
},
{
"epoch": 36.02,
"learning_rate": 1.4142857142857143e-05,
"loss": 0.6341,
"step": 34900
},
{
"epoch": 36.12,
"learning_rate": 1.4038581856100106e-05,
"loss": 0.6346,
"step": 35000
},
{
"epoch": 36.12,
"eval_loss": 0.15833768248558044,
"eval_runtime": 469.6695,
"eval_samples_per_second": 22.307,
"eval_steps_per_second": 2.789,
"eval_wer": 0.12739235709022,
"step": 35000
},
{
"epoch": 36.22,
"learning_rate": 1.3934306569343066e-05,
"loss": 0.6286,
"step": 35100
},
{
"epoch": 36.33,
"learning_rate": 1.3830031282586029e-05,
"loss": 0.6331,
"step": 35200
},
{
"epoch": 36.43,
"learning_rate": 1.372575599582899e-05,
"loss": 0.6256,
"step": 35300
},
{
"epoch": 36.53,
"learning_rate": 1.3621480709071951e-05,
"loss": 0.6289,
"step": 35400
},
{
"epoch": 36.64,
"learning_rate": 1.3517205422314914e-05,
"loss": 0.6143,
"step": 35500
},
{
"epoch": 36.64,
"eval_loss": 0.15764719247817993,
"eval_runtime": 468.9196,
"eval_samples_per_second": 22.343,
"eval_steps_per_second": 2.794,
"eval_wer": 0.12639225181598063,
"step": 35500
},
{
"epoch": 36.74,
"learning_rate": 1.3412930135557874e-05,
"loss": 0.6414,
"step": 35600
},
{
"epoch": 36.84,
"learning_rate": 1.3308654848800836e-05,
"loss": 0.6184,
"step": 35700
},
{
"epoch": 36.95,
"learning_rate": 1.3204379562043795e-05,
"loss": 0.6227,
"step": 35800
},
{
"epoch": 37.05,
"learning_rate": 1.3100104275286756e-05,
"loss": 0.6488,
"step": 35900
},
{
"epoch": 37.15,
"learning_rate": 1.2995828988529718e-05,
"loss": 0.6208,
"step": 36000
},
{
"epoch": 37.15,
"eval_loss": 0.16206438839435577,
"eval_runtime": 468.3673,
"eval_samples_per_second": 22.369,
"eval_steps_per_second": 2.797,
"eval_wer": 0.12628697757658702,
"step": 36000
},
{
"epoch": 37.25,
"learning_rate": 1.2891553701772679e-05,
"loss": 0.6335,
"step": 36100
},
{
"epoch": 37.36,
"learning_rate": 1.2787278415015641e-05,
"loss": 0.6227,
"step": 36200
},
{
"epoch": 37.46,
"learning_rate": 1.2683003128258603e-05,
"loss": 0.6291,
"step": 36300
},
{
"epoch": 37.56,
"learning_rate": 1.2578727841501564e-05,
"loss": 0.6203,
"step": 36400
},
{
"epoch": 37.67,
"learning_rate": 1.2474452554744526e-05,
"loss": 0.6185,
"step": 36500
},
{
"epoch": 37.67,
"eval_loss": 0.16232742369174957,
"eval_runtime": 467.766,
"eval_samples_per_second": 22.398,
"eval_steps_per_second": 2.801,
"eval_wer": 0.12702389725234234,
"step": 36500
},
{
"epoch": 37.77,
"learning_rate": 1.2370177267987487e-05,
"loss": 0.6137,
"step": 36600
},
{
"epoch": 37.87,
"learning_rate": 1.2265901981230449e-05,
"loss": 0.6131,
"step": 36700
},
{
"epoch": 37.98,
"learning_rate": 1.216162669447341e-05,
"loss": 0.6076,
"step": 36800
},
{
"epoch": 38.08,
"learning_rate": 1.2057351407716372e-05,
"loss": 0.6058,
"step": 36900
},
{
"epoch": 38.18,
"learning_rate": 1.1953076120959334e-05,
"loss": 0.6128,
"step": 37000
},
{
"epoch": 38.18,
"eval_loss": 0.16044454276561737,
"eval_runtime": 479.5033,
"eval_samples_per_second": 21.85,
"eval_steps_per_second": 2.732,
"eval_wer": 0.12678176650173703,
"step": 37000
},
{
"epoch": 38.29,
"learning_rate": 1.1848800834202295e-05,
"loss": 0.6095,
"step": 37100
},
{
"epoch": 38.39,
"learning_rate": 1.1746611053180396e-05,
"loss": 0.62,
"step": 37200
},
{
"epoch": 38.49,
"learning_rate": 1.1642335766423359e-05,
"loss": 0.607,
"step": 37300
},
{
"epoch": 38.6,
"learning_rate": 1.153806047966632e-05,
"loss": 0.62,
"step": 37400
},
{
"epoch": 38.7,
"learning_rate": 1.1433785192909282e-05,
"loss": 0.6151,
"step": 37500
},
{
"epoch": 38.7,
"eval_loss": 0.15926694869995117,
"eval_runtime": 481.5864,
"eval_samples_per_second": 21.755,
"eval_steps_per_second": 2.72,
"eval_wer": 0.12461311717022845,
"step": 37500
},
{
"epoch": 38.8,
"learning_rate": 1.1329509906152242e-05,
"loss": 0.6178,
"step": 37600
},
{
"epoch": 38.91,
"learning_rate": 1.1225234619395204e-05,
"loss": 0.615,
"step": 37700
},
{
"epoch": 39.01,
"learning_rate": 1.1120959332638167e-05,
"loss": 0.6096,
"step": 37800
},
{
"epoch": 39.11,
"learning_rate": 1.1016684045881126e-05,
"loss": 0.6134,
"step": 37900
},
{
"epoch": 39.22,
"learning_rate": 1.0912408759124088e-05,
"loss": 0.6082,
"step": 38000
},
{
"epoch": 39.22,
"eval_loss": 0.15319351851940155,
"eval_runtime": 480.0879,
"eval_samples_per_second": 21.823,
"eval_steps_per_second": 2.729,
"eval_wer": 0.12383408779871566,
"step": 38000
},
{
"epoch": 39.32,
"learning_rate": 1.0808133472367048e-05,
"loss": 0.6099,
"step": 38100
},
{
"epoch": 39.42,
"learning_rate": 1.070385818561001e-05,
"loss": 0.606,
"step": 38200
},
{
"epoch": 39.53,
"learning_rate": 1.0599582898852973e-05,
"loss": 0.6115,
"step": 38300
},
{
"epoch": 39.63,
"learning_rate": 1.0495307612095933e-05,
"loss": 0.6105,
"step": 38400
},
{
"epoch": 39.73,
"learning_rate": 1.0391032325338896e-05,
"loss": 0.6,
"step": 38500
},
{
"epoch": 39.73,
"eval_loss": 0.15240420401096344,
"eval_runtime": 471.9993,
"eval_samples_per_second": 22.197,
"eval_steps_per_second": 2.775,
"eval_wer": 0.1224339404147805,
"step": 38500
},
{
"epoch": 39.83,
"learning_rate": 1.0286757038581856e-05,
"loss": 0.5997,
"step": 38600
},
{
"epoch": 39.94,
"learning_rate": 1.0182481751824819e-05,
"loss": 0.5957,
"step": 38700
},
{
"epoch": 40.04,
"learning_rate": 1.0078206465067779e-05,
"loss": 0.613,
"step": 38800
},
{
"epoch": 40.14,
"learning_rate": 9.973931178310741e-06,
"loss": 0.5961,
"step": 38900
},
{
"epoch": 40.25,
"learning_rate": 9.869655891553702e-06,
"loss": 0.6032,
"step": 39000
},
{
"epoch": 40.25,
"eval_loss": 0.1521444320678711,
"eval_runtime": 467.5416,
"eval_samples_per_second": 22.409,
"eval_steps_per_second": 2.802,
"eval_wer": 0.12116012211811769,
"step": 39000
},
{
"epoch": 40.35,
"learning_rate": 9.765380604796663e-06,
"loss": 0.5952,
"step": 39100
},
{
"epoch": 40.45,
"learning_rate": 9.661105318039625e-06,
"loss": 0.6034,
"step": 39200
},
{
"epoch": 40.56,
"learning_rate": 9.556830031282587e-06,
"loss": 0.5917,
"step": 39300
},
{
"epoch": 40.66,
"learning_rate": 9.452554744525548e-06,
"loss": 0.602,
"step": 39400
},
{
"epoch": 40.76,
"learning_rate": 9.34827945776851e-06,
"loss": 0.6016,
"step": 39500
},
{
"epoch": 40.76,
"eval_loss": 0.15509599447250366,
"eval_runtime": 467.848,
"eval_samples_per_second": 22.394,
"eval_steps_per_second": 2.8,
"eval_wer": 0.12151805453205601,
"step": 39500
},
{
"epoch": 40.87,
"learning_rate": 9.24400417101147e-06,
"loss": 0.6019,
"step": 39600
},
{
"epoch": 40.97,
"learning_rate": 9.139728884254433e-06,
"loss": 0.5999,
"step": 39700
},
{
"epoch": 41.07,
"learning_rate": 9.035453597497393e-06,
"loss": 0.5809,
"step": 39800
},
{
"epoch": 41.18,
"learning_rate": 8.931178310740356e-06,
"loss": 0.5907,
"step": 39900
},
{
"epoch": 41.28,
"learning_rate": 8.826903023983318e-06,
"loss": 0.6009,
"step": 40000
},
{
"epoch": 41.28,
"eval_loss": 0.1522541046142578,
"eval_runtime": 472.1828,
"eval_samples_per_second": 22.188,
"eval_steps_per_second": 2.774,
"eval_wer": 0.12153910937993473,
"step": 40000
},
{
"epoch": 41.38,
"learning_rate": 8.722627737226277e-06,
"loss": 0.5898,
"step": 40100
},
{
"epoch": 41.49,
"learning_rate": 8.618352450469239e-06,
"loss": 0.6018,
"step": 40200
},
{
"epoch": 41.59,
"learning_rate": 8.5140771637122e-06,
"loss": 0.6043,
"step": 40300
},
{
"epoch": 41.69,
"learning_rate": 8.409801876955162e-06,
"loss": 0.585,
"step": 40400
},
{
"epoch": 41.8,
"learning_rate": 8.305526590198124e-06,
"loss": 0.5875,
"step": 40500
},
{
"epoch": 41.8,
"eval_loss": 0.1540576070547104,
"eval_runtime": 466.9539,
"eval_samples_per_second": 22.437,
"eval_steps_per_second": 2.805,
"eval_wer": 0.1216022739235709,
"step": 40500
},
{
"epoch": 41.9,
"learning_rate": 8.201251303441085e-06,
"loss": 0.5941,
"step": 40600
},
{
"epoch": 42.0,
"learning_rate": 8.096976016684047e-06,
"loss": 0.6136,
"step": 40700
},
{
"epoch": 42.11,
"learning_rate": 7.992700729927007e-06,
"loss": 0.5969,
"step": 40800
},
{
"epoch": 42.21,
"learning_rate": 7.88842544316997e-06,
"loss": 0.5959,
"step": 40900
},
{
"epoch": 42.31,
"learning_rate": 7.78415015641293e-06,
"loss": 0.608,
"step": 41000
},
{
"epoch": 42.31,
"eval_loss": 0.1535973846912384,
"eval_runtime": 466.2489,
"eval_samples_per_second": 22.471,
"eval_steps_per_second": 2.81,
"eval_wer": 0.12093904621539109,
"step": 41000
},
{
"epoch": 42.41,
"learning_rate": 7.679874869655892e-06,
"loss": 0.5965,
"step": 41100
},
{
"epoch": 42.52,
"learning_rate": 7.575599582898853e-06,
"loss": 0.5773,
"step": 41200
},
{
"epoch": 42.62,
"learning_rate": 7.472367049009385e-06,
"loss": 0.5844,
"step": 41300
},
{
"epoch": 42.72,
"learning_rate": 7.3680917622523466e-06,
"loss": 0.5906,
"step": 41400
},
{
"epoch": 42.83,
"learning_rate": 7.263816475495308e-06,
"loss": 0.5876,
"step": 41500
},
{
"epoch": 42.83,
"eval_loss": 0.15671317279338837,
"eval_runtime": 479.4414,
"eval_samples_per_second": 21.853,
"eval_steps_per_second": 2.732,
"eval_wer": 0.12110748499842089,
"step": 41500
},
{
"epoch": 42.93,
"learning_rate": 7.159541188738269e-06,
"loss": 0.5746,
"step": 41600
},
{
"epoch": 43.03,
"learning_rate": 7.055265901981231e-06,
"loss": 0.5863,
"step": 41700
},
{
"epoch": 43.14,
"learning_rate": 6.950990615224193e-06,
"loss": 0.5768,
"step": 41800
},
{
"epoch": 43.24,
"learning_rate": 6.8467153284671545e-06,
"loss": 0.5852,
"step": 41900
},
{
"epoch": 43.34,
"learning_rate": 6.742440041710114e-06,
"loss": 0.5714,
"step": 42000
},
{
"epoch": 43.34,
"eval_loss": 0.15320713818073273,
"eval_runtime": 486.9148,
"eval_samples_per_second": 21.517,
"eval_steps_per_second": 2.69,
"eval_wer": 0.12170754816296453,
"step": 42000
},
{
"epoch": 43.45,
"learning_rate": 6.638164754953076e-06,
"loss": 0.5857,
"step": 42100
},
{
"epoch": 43.55,
"learning_rate": 6.5349322210636086e-06,
"loss": 0.592,
"step": 42200
},
{
"epoch": 43.65,
"learning_rate": 6.43065693430657e-06,
"loss": 0.5853,
"step": 42300
},
{
"epoch": 43.76,
"learning_rate": 6.326381647549531e-06,
"loss": 0.5834,
"step": 42400
},
{
"epoch": 43.86,
"learning_rate": 6.222106360792492e-06,
"loss": 0.5756,
"step": 42500
},
{
"epoch": 43.86,
"eval_loss": 0.15160232782363892,
"eval_runtime": 477.2403,
"eval_samples_per_second": 21.953,
"eval_steps_per_second": 2.745,
"eval_wer": 0.11955995367933467,
"step": 42500
},
{
"epoch": 43.96,
"learning_rate": 6.117831074035453e-06,
"loss": 0.5835,
"step": 42600
},
{
"epoch": 44.07,
"learning_rate": 6.013555787278416e-06,
"loss": 0.5761,
"step": 42700
},
{
"epoch": 44.17,
"learning_rate": 5.909280500521377e-06,
"loss": 0.5703,
"step": 42800
},
{
"epoch": 44.27,
"learning_rate": 5.8050052137643385e-06,
"loss": 0.5838,
"step": 42900
},
{
"epoch": 44.38,
"learning_rate": 5.7007299270073e-06,
"loss": 0.5719,
"step": 43000
},
{
"epoch": 44.38,
"eval_loss": 0.14908714592456818,
"eval_runtime": 478.7642,
"eval_samples_per_second": 21.883,
"eval_steps_per_second": 2.736,
"eval_wer": 0.11914938414569955,
"step": 43000
},
{
"epoch": 44.48,
"learning_rate": 5.5964546402502604e-06,
"loss": 0.5849,
"step": 43100
},
{
"epoch": 44.58,
"learning_rate": 5.492179353493222e-06,
"loss": 0.5731,
"step": 43200
},
{
"epoch": 44.69,
"learning_rate": 5.387904066736184e-06,
"loss": 0.5722,
"step": 43300
},
{
"epoch": 44.79,
"learning_rate": 5.2836287799791455e-06,
"loss": 0.578,
"step": 43400
},
{
"epoch": 44.89,
"learning_rate": 5.179353493222107e-06,
"loss": 0.5829,
"step": 43500
},
{
"epoch": 44.89,
"eval_loss": 0.1497020274400711,
"eval_runtime": 463.7767,
"eval_samples_per_second": 22.591,
"eval_steps_per_second": 2.825,
"eval_wer": 0.11930729550478998,
"step": 43500
},
{
"epoch": 44.99,
"learning_rate": 5.0750782064650675e-06,
"loss": 0.5936,
"step": 43600
},
{
"epoch": 45.1,
"learning_rate": 4.970802919708029e-06,
"loss": 0.5714,
"step": 43700
},
{
"epoch": 45.2,
"learning_rate": 4.866527632950991e-06,
"loss": 0.5593,
"step": 43800
},
{
"epoch": 45.3,
"learning_rate": 4.762252346193953e-06,
"loss": 0.5766,
"step": 43900
},
{
"epoch": 45.41,
"learning_rate": 4.657977059436914e-06,
"loss": 0.5664,
"step": 44000
},
{
"epoch": 45.41,
"eval_loss": 0.14872093498706818,
"eval_runtime": 471.3612,
"eval_samples_per_second": 22.227,
"eval_steps_per_second": 2.779,
"eval_wer": 0.11728603010843247,
"step": 44000
},
{
"epoch": 45.51,
"learning_rate": 4.553701772679875e-06,
"loss": 0.5697,
"step": 44100
},
{
"epoch": 45.61,
"learning_rate": 4.449426485922836e-06,
"loss": 0.5718,
"step": 44200
},
{
"epoch": 45.72,
"learning_rate": 4.345151199165797e-06,
"loss": 0.5744,
"step": 44300
},
{
"epoch": 45.82,
"learning_rate": 4.24087591240876e-06,
"loss": 0.5672,
"step": 44400
},
{
"epoch": 45.92,
"learning_rate": 4.137643378519291e-06,
"loss": 0.5707,
"step": 44500
},
{
"epoch": 45.92,
"eval_loss": 0.1470126211643219,
"eval_runtime": 478.6316,
"eval_samples_per_second": 21.889,
"eval_steps_per_second": 2.737,
"eval_wer": 0.11641225392146541,
"step": 44500
},
{
"epoch": 46.03,
"learning_rate": 4.033368091762252e-06,
"loss": 0.5555,
"step": 44600
},
{
"epoch": 46.13,
"learning_rate": 3.929092805005214e-06,
"loss": 0.5699,
"step": 44700
},
{
"epoch": 46.23,
"learning_rate": 3.824817518248176e-06,
"loss": 0.573,
"step": 44800
},
{
"epoch": 46.34,
"learning_rate": 3.7205422314911366e-06,
"loss": 0.5697,
"step": 44900
},
{
"epoch": 46.44,
"learning_rate": 3.616266944734098e-06,
"loss": 0.5696,
"step": 45000
},
{
"epoch": 46.44,
"eval_loss": 0.14786458015441895,
"eval_runtime": 467.8513,
"eval_samples_per_second": 22.394,
"eval_steps_per_second": 2.8,
"eval_wer": 0.11606484893146647,
"step": 45000
},
{
"epoch": 46.54,
"learning_rate": 3.5119916579770594e-06,
"loss": 0.5634,
"step": 45100
},
{
"epoch": 46.65,
"learning_rate": 3.4077163712200212e-06,
"loss": 0.5742,
"step": 45200
},
{
"epoch": 46.75,
"learning_rate": 3.3044838373305525e-06,
"loss": 0.5612,
"step": 45300
},
{
"epoch": 46.85,
"learning_rate": 3.2002085505735143e-06,
"loss": 0.5548,
"step": 45400
},
{
"epoch": 46.96,
"learning_rate": 3.0959332638164757e-06,
"loss": 0.5767,
"step": 45500
},
{
"epoch": 46.96,
"eval_loss": 0.1492297500371933,
"eval_runtime": 465.8572,
"eval_samples_per_second": 22.49,
"eval_steps_per_second": 2.812,
"eval_wer": 0.11754921570691652,
"step": 45500
},
{
"epoch": 47.06,
"learning_rate": 2.991657977059437e-06,
"loss": 0.5831,
"step": 45600
},
{
"epoch": 47.16,
"learning_rate": 2.8873826903023986e-06,
"loss": 0.5653,
"step": 45700
},
{
"epoch": 47.27,
"learning_rate": 2.78310740354536e-06,
"loss": 0.5744,
"step": 45800
},
{
"epoch": 47.37,
"learning_rate": 2.6788321167883214e-06,
"loss": 0.5587,
"step": 45900
},
{
"epoch": 47.47,
"learning_rate": 2.574556830031283e-06,
"loss": 0.5573,
"step": 46000
},
{
"epoch": 47.47,
"eval_loss": 0.14706237614154816,
"eval_runtime": 467.4225,
"eval_samples_per_second": 22.414,
"eval_steps_per_second": 2.803,
"eval_wer": 0.1164754184651016,
"step": 46000
}
],
"max_steps": 48450,
"num_train_epochs": 50,
"total_flos": 5.920193480786585e+20,
"trial_name": null,
"trial_params": null
}