boumehdi's picture
Upload 9 files
cf16445
raw
history blame
19.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 27.657824933687003,
"global_step": 5200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 0.9903,
"step": 100
},
{
"epoch": 0.53,
"eval_loss": 0.3881553113460541,
"eval_runtime": 209.4743,
"eval_samples_per_second": 19.344,
"eval_steps_per_second": 2.42,
"eval_wer": 0.4149805202965942,
"step": 100
},
{
"epoch": 1.06,
"learning_rate": 9.933897408778425e-05,
"loss": 0.6655,
"step": 200
},
{
"epoch": 1.06,
"eval_loss": 0.33381059765815735,
"eval_runtime": 209.3188,
"eval_samples_per_second": 19.358,
"eval_steps_per_second": 2.422,
"eval_wer": 0.35063466130451176,
"step": 200
},
{
"epoch": 1.59,
"learning_rate": 9.867794817556849e-05,
"loss": 0.5789,
"step": 300
},
{
"epoch": 1.59,
"eval_loss": 0.30856332182884216,
"eval_runtime": 210.7573,
"eval_samples_per_second": 19.226,
"eval_steps_per_second": 2.406,
"eval_wer": 0.34303129320095516,
"step": 300
},
{
"epoch": 2.13,
"learning_rate": 9.801692226335273e-05,
"loss": 0.5539,
"step": 400
},
{
"epoch": 2.13,
"eval_loss": 0.3059796392917633,
"eval_runtime": 210.5815,
"eval_samples_per_second": 19.242,
"eval_steps_per_second": 2.408,
"eval_wer": 0.31789619203217295,
"step": 400
},
{
"epoch": 2.66,
"learning_rate": 9.735589635113697e-05,
"loss": 0.5041,
"step": 500
},
{
"epoch": 2.66,
"eval_loss": 0.3157837688922882,
"eval_runtime": 211.1395,
"eval_samples_per_second": 19.191,
"eval_steps_per_second": 2.401,
"eval_wer": 0.3104813371873822,
"step": 500
},
{
"epoch": 3.19,
"learning_rate": 9.669487043892121e-05,
"loss": 0.4771,
"step": 600
},
{
"epoch": 3.19,
"eval_loss": 0.3155499994754791,
"eval_runtime": 213.4785,
"eval_samples_per_second": 18.981,
"eval_steps_per_second": 2.375,
"eval_wer": 0.32072389091366094,
"step": 600
},
{
"epoch": 3.72,
"learning_rate": 9.603384452670545e-05,
"loss": 0.448,
"step": 700
},
{
"epoch": 3.72,
"eval_loss": 0.29987651109695435,
"eval_runtime": 213.1151,
"eval_samples_per_second": 19.013,
"eval_steps_per_second": 2.379,
"eval_wer": 0.3066482342591429,
"step": 700
},
{
"epoch": 4.25,
"learning_rate": 9.537281861448969e-05,
"loss": 0.4454,
"step": 800
},
{
"epoch": 4.25,
"eval_loss": 0.3030799925327301,
"eval_runtime": 214.1216,
"eval_samples_per_second": 18.924,
"eval_steps_per_second": 2.368,
"eval_wer": 0.32084956641950485,
"step": 800
},
{
"epoch": 4.79,
"learning_rate": 9.471179270227393e-05,
"loss": 0.3857,
"step": 900
},
{
"epoch": 4.79,
"eval_loss": 0.29331761598587036,
"eval_runtime": 214.8248,
"eval_samples_per_second": 18.862,
"eval_steps_per_second": 2.36,
"eval_wer": 0.2947718989568933,
"step": 900
},
{
"epoch": 5.32,
"learning_rate": 9.405076679005818e-05,
"loss": 0.3722,
"step": 1000
},
{
"epoch": 5.32,
"eval_loss": 0.28789493441581726,
"eval_runtime": 214.8964,
"eval_samples_per_second": 18.856,
"eval_steps_per_second": 2.359,
"eval_wer": 0.28578610028905366,
"step": 1000
},
{
"epoch": 5.85,
"learning_rate": 9.338974087784242e-05,
"loss": 0.371,
"step": 1100
},
{
"epoch": 5.85,
"eval_loss": 0.2818315625190735,
"eval_runtime": 215.1931,
"eval_samples_per_second": 18.83,
"eval_steps_per_second": 2.356,
"eval_wer": 0.2755435465627749,
"step": 1100
},
{
"epoch": 6.38,
"learning_rate": 9.272871496562666e-05,
"loss": 0.358,
"step": 1200
},
{
"epoch": 6.38,
"eval_loss": 0.30174919962882996,
"eval_runtime": 215.3959,
"eval_samples_per_second": 18.812,
"eval_steps_per_second": 2.354,
"eval_wer": 0.2778685434208873,
"step": 1200
},
{
"epoch": 6.91,
"learning_rate": 9.20676890534109e-05,
"loss": 0.3459,
"step": 1300
},
{
"epoch": 6.91,
"eval_loss": 0.2872300148010254,
"eval_runtime": 216.2384,
"eval_samples_per_second": 18.739,
"eval_steps_per_second": 2.345,
"eval_wer": 0.27422395375141384,
"step": 1300
},
{
"epoch": 7.45,
"learning_rate": 9.140666314119514e-05,
"loss": 0.3293,
"step": 1400
},
{
"epoch": 7.45,
"eval_loss": 0.31064674258232117,
"eval_runtime": 216.4276,
"eval_samples_per_second": 18.722,
"eval_steps_per_second": 2.343,
"eval_wer": 0.2762347618449164,
"step": 1400
},
{
"epoch": 7.98,
"learning_rate": 9.074563722897938e-05,
"loss": 0.3305,
"step": 1500
},
{
"epoch": 7.98,
"eval_loss": 0.2984163761138916,
"eval_runtime": 216.6335,
"eval_samples_per_second": 18.704,
"eval_steps_per_second": 2.34,
"eval_wer": 0.2746638180218675,
"step": 1500
},
{
"epoch": 8.51,
"learning_rate": 9.008461131676362e-05,
"loss": 0.322,
"step": 1600
},
{
"epoch": 8.51,
"eval_loss": 0.3066250681877136,
"eval_runtime": 217.5565,
"eval_samples_per_second": 18.625,
"eval_steps_per_second": 2.33,
"eval_wer": 0.2688199070001257,
"step": 1600
},
{
"epoch": 9.04,
"learning_rate": 8.942358540454786e-05,
"loss": 0.3051,
"step": 1700
},
{
"epoch": 9.04,
"eval_loss": 0.3064703941345215,
"eval_runtime": 218.1465,
"eval_samples_per_second": 18.575,
"eval_steps_per_second": 2.324,
"eval_wer": 0.26743747643584265,
"step": 1700
},
{
"epoch": 9.57,
"learning_rate": 8.87625594923321e-05,
"loss": 0.2906,
"step": 1800
},
{
"epoch": 9.57,
"eval_loss": 0.2988126277923584,
"eval_runtime": 218.1544,
"eval_samples_per_second": 18.574,
"eval_steps_per_second": 2.324,
"eval_wer": 0.26209626743747644,
"step": 1800
},
{
"epoch": 10.11,
"learning_rate": 8.810153358011635e-05,
"loss": 0.2908,
"step": 1900
},
{
"epoch": 10.11,
"eval_loss": 0.31721261143684387,
"eval_runtime": 217.8764,
"eval_samples_per_second": 18.598,
"eval_steps_per_second": 2.327,
"eval_wer": 0.26448410204851075,
"step": 1900
},
{
"epoch": 10.64,
"learning_rate": 8.744050766790059e-05,
"loss": 0.2644,
"step": 2000
},
{
"epoch": 10.64,
"eval_loss": 0.328941285610199,
"eval_runtime": 217.9159,
"eval_samples_per_second": 18.594,
"eval_steps_per_second": 2.327,
"eval_wer": 0.25851451552092497,
"step": 2000
},
{
"epoch": 11.17,
"learning_rate": 8.677948175568483e-05,
"loss": 0.2895,
"step": 2100
},
{
"epoch": 11.17,
"eval_loss": 0.32557472586631775,
"eval_runtime": 218.4939,
"eval_samples_per_second": 18.545,
"eval_steps_per_second": 2.32,
"eval_wer": 0.25681789619203216,
"step": 2100
},
{
"epoch": 11.7,
"learning_rate": 8.611845584346907e-05,
"loss": 0.2764,
"step": 2200
},
{
"epoch": 11.7,
"eval_loss": 0.31102919578552246,
"eval_runtime": 221.5496,
"eval_samples_per_second": 18.289,
"eval_steps_per_second": 2.288,
"eval_wer": 0.2536760085459344,
"step": 2200
},
{
"epoch": 12.23,
"learning_rate": 8.545742993125331e-05,
"loss": 0.2712,
"step": 2300
},
{
"epoch": 12.23,
"eval_loss": 0.31744903326034546,
"eval_runtime": 229.9509,
"eval_samples_per_second": 17.621,
"eval_steps_per_second": 2.205,
"eval_wer": 0.25920573080306647,
"step": 2300
},
{
"epoch": 12.76,
"learning_rate": 8.479640401903755e-05,
"loss": 0.2688,
"step": 2400
},
{
"epoch": 12.76,
"eval_loss": 0.3221331238746643,
"eval_runtime": 231.8744,
"eval_samples_per_second": 17.475,
"eval_steps_per_second": 2.187,
"eval_wer": 0.25826316450923714,
"step": 2400
},
{
"epoch": 13.3,
"learning_rate": 8.413537810682179e-05,
"loss": 0.2509,
"step": 2500
},
{
"epoch": 13.3,
"eval_loss": 0.32597509026527405,
"eval_runtime": 232.2681,
"eval_samples_per_second": 17.445,
"eval_steps_per_second": 2.183,
"eval_wer": 0.2532989820284027,
"step": 2500
},
{
"epoch": 13.83,
"learning_rate": 8.34809624537282e-05,
"loss": 0.2419,
"step": 2600
},
{
"epoch": 13.83,
"eval_loss": 0.3077153265476227,
"eval_runtime": 232.0814,
"eval_samples_per_second": 17.459,
"eval_steps_per_second": 2.185,
"eval_wer": 0.25530979012190524,
"step": 2600
},
{
"epoch": 14.36,
"learning_rate": 8.281993654151243e-05,
"loss": 0.2429,
"step": 2700
},
{
"epoch": 14.36,
"eval_loss": 0.32647523283958435,
"eval_runtime": 232.8137,
"eval_samples_per_second": 17.404,
"eval_steps_per_second": 2.178,
"eval_wer": 0.2588287042855347,
"step": 2700
},
{
"epoch": 14.89,
"learning_rate": 8.215891062929667e-05,
"loss": 0.2358,
"step": 2800
},
{
"epoch": 14.89,
"eval_loss": 0.3333515226840973,
"eval_runtime": 232.1937,
"eval_samples_per_second": 17.451,
"eval_steps_per_second": 2.184,
"eval_wer": 0.25380168405177833,
"step": 2800
},
{
"epoch": 15.42,
"learning_rate": 8.149788471708092e-05,
"loss": 0.2415,
"step": 2900
},
{
"epoch": 15.42,
"eval_loss": 0.3471778631210327,
"eval_runtime": 232.185,
"eval_samples_per_second": 17.452,
"eval_steps_per_second": 2.184,
"eval_wer": 0.24921452808847555,
"step": 2900
},
{
"epoch": 15.95,
"learning_rate": 8.083685880486515e-05,
"loss": 0.2384,
"step": 3000
},
{
"epoch": 15.95,
"eval_loss": 0.3481573760509491,
"eval_runtime": 232.291,
"eval_samples_per_second": 17.444,
"eval_steps_per_second": 2.183,
"eval_wer": 0.2548070880985296,
"step": 3000
},
{
"epoch": 16.49,
"learning_rate": 8.017583289264939e-05,
"loss": 0.2316,
"step": 3100
},
{
"epoch": 16.49,
"eval_loss": 0.3469015955924988,
"eval_runtime": 232.5701,
"eval_samples_per_second": 17.423,
"eval_steps_per_second": 2.18,
"eval_wer": 0.24695236898328515,
"step": 3100
},
{
"epoch": 17.02,
"learning_rate": 7.951480698043363e-05,
"loss": 0.225,
"step": 3200
},
{
"epoch": 17.02,
"eval_loss": 0.3405754566192627,
"eval_runtime": 220.3171,
"eval_samples_per_second": 18.392,
"eval_steps_per_second": 2.301,
"eval_wer": 0.2588915420384567,
"step": 3200
},
{
"epoch": 17.55,
"learning_rate": 7.885378106821787e-05,
"loss": 0.2108,
"step": 3300
},
{
"epoch": 17.55,
"eval_loss": 0.34463852643966675,
"eval_runtime": 223.0068,
"eval_samples_per_second": 18.17,
"eval_steps_per_second": 2.273,
"eval_wer": 0.25072263415860246,
"step": 3300
},
{
"epoch": 18.08,
"learning_rate": 7.819275515600211e-05,
"loss": 0.2179,
"step": 3400
},
{
"epoch": 18.08,
"eval_loss": 0.34099045395851135,
"eval_runtime": 221.6762,
"eval_samples_per_second": 18.279,
"eval_steps_per_second": 2.287,
"eval_wer": 0.248586150559256,
"step": 3400
},
{
"epoch": 18.62,
"learning_rate": 7.753172924378636e-05,
"loss": 0.2056,
"step": 3500
},
{
"epoch": 18.62,
"eval_loss": 0.33803310990333557,
"eval_runtime": 221.1741,
"eval_samples_per_second": 18.32,
"eval_steps_per_second": 2.292,
"eval_wer": 0.248586150559256,
"step": 3500
},
{
"epoch": 19.15,
"learning_rate": 7.687731359069276e-05,
"loss": 0.2088,
"step": 3600
},
{
"epoch": 19.15,
"eval_loss": 0.33515065908432007,
"eval_runtime": 221.8587,
"eval_samples_per_second": 18.264,
"eval_steps_per_second": 2.285,
"eval_wer": 0.24437602111348497,
"step": 3600
},
{
"epoch": 19.68,
"learning_rate": 7.6216287678477e-05,
"loss": 0.1994,
"step": 3700
},
{
"epoch": 19.68,
"eval_loss": 0.3439195454120636,
"eval_runtime": 222.8182,
"eval_samples_per_second": 18.185,
"eval_steps_per_second": 2.275,
"eval_wer": 0.2422395375141385,
"step": 3700
},
{
"epoch": 20.21,
"learning_rate": 7.555526176626125e-05,
"loss": 0.2027,
"step": 3800
},
{
"epoch": 20.21,
"eval_loss": 0.37243127822875977,
"eval_runtime": 222.5733,
"eval_samples_per_second": 18.205,
"eval_steps_per_second": 2.278,
"eval_wer": 0.24337061706673369,
"step": 3800
},
{
"epoch": 20.74,
"learning_rate": 7.489423585404549e-05,
"loss": 0.2044,
"step": 3900
},
{
"epoch": 20.74,
"eval_loss": 0.3538868725299835,
"eval_runtime": 222.1787,
"eval_samples_per_second": 18.238,
"eval_steps_per_second": 2.282,
"eval_wer": 0.2501570943823049,
"step": 3900
},
{
"epoch": 21.28,
"learning_rate": 7.423320994182971e-05,
"loss": 0.1932,
"step": 4000
},
{
"epoch": 21.28,
"eval_loss": 0.34956350922584534,
"eval_runtime": 221.2804,
"eval_samples_per_second": 18.312,
"eval_steps_per_second": 2.291,
"eval_wer": 0.25304763101671485,
"step": 4000
},
{
"epoch": 21.81,
"learning_rate": 7.357218402961397e-05,
"loss": 0.1903,
"step": 4100
},
{
"epoch": 21.81,
"eval_loss": 0.3472049832344055,
"eval_runtime": 221.4938,
"eval_samples_per_second": 18.294,
"eval_steps_per_second": 2.289,
"eval_wer": 0.2465753424657534,
"step": 4100
},
{
"epoch": 22.34,
"learning_rate": 7.291115811739821e-05,
"loss": 0.1895,
"step": 4200
},
{
"epoch": 22.34,
"eval_loss": 0.34313011169433594,
"eval_runtime": 222.6794,
"eval_samples_per_second": 18.197,
"eval_steps_per_second": 2.277,
"eval_wer": 0.24286791504335806,
"step": 4200
},
{
"epoch": 22.87,
"learning_rate": 7.225013220518244e-05,
"loss": 0.1865,
"step": 4300
},
{
"epoch": 22.87,
"eval_loss": 0.3476735055446625,
"eval_runtime": 222.0492,
"eval_samples_per_second": 18.248,
"eval_steps_per_second": 2.283,
"eval_wer": 0.24481588538393867,
"step": 4300
},
{
"epoch": 23.4,
"learning_rate": 7.158910629296669e-05,
"loss": 0.1851,
"step": 4400
},
{
"epoch": 23.4,
"eval_loss": 0.3553401231765747,
"eval_runtime": 223.2051,
"eval_samples_per_second": 18.154,
"eval_steps_per_second": 2.271,
"eval_wer": 0.23922332537388463,
"step": 4400
},
{
"epoch": 23.93,
"learning_rate": 7.092808038075093e-05,
"loss": 0.179,
"step": 4500
},
{
"epoch": 23.93,
"eval_loss": 0.3559369742870331,
"eval_runtime": 226.903,
"eval_samples_per_second": 17.858,
"eval_steps_per_second": 2.234,
"eval_wer": 0.24255372627874827,
"step": 4500
},
{
"epoch": 24.47,
"learning_rate": 7.026705446853516e-05,
"loss": 0.1797,
"step": 4600
},
{
"epoch": 24.47,
"eval_loss": 0.3819045424461365,
"eval_runtime": 222.3201,
"eval_samples_per_second": 18.226,
"eval_steps_per_second": 2.28,
"eval_wer": 0.24644966695990952,
"step": 4600
},
{
"epoch": 25.0,
"learning_rate": 6.960602855631942e-05,
"loss": 0.1889,
"step": 4700
},
{
"epoch": 25.0,
"eval_loss": 0.3539634346961975,
"eval_runtime": 224.8629,
"eval_samples_per_second": 18.02,
"eval_steps_per_second": 2.255,
"eval_wer": 0.24073143144401157,
"step": 4700
},
{
"epoch": 25.53,
"learning_rate": 6.894500264410366e-05,
"loss": 0.1679,
"step": 4800
},
{
"epoch": 25.53,
"eval_loss": 0.3614364564418793,
"eval_runtime": 233.686,
"eval_samples_per_second": 17.34,
"eval_steps_per_second": 2.17,
"eval_wer": 0.2404172426794018,
"step": 4800
},
{
"epoch": 26.06,
"learning_rate": 6.828397673188788e-05,
"loss": 0.1667,
"step": 4900
},
{
"epoch": 26.06,
"eval_loss": 0.3456764817237854,
"eval_runtime": 234.7624,
"eval_samples_per_second": 17.26,
"eval_steps_per_second": 2.16,
"eval_wer": 0.24230237526706044,
"step": 4900
},
{
"epoch": 26.59,
"learning_rate": 6.762295081967214e-05,
"loss": 0.1652,
"step": 5000
},
{
"epoch": 26.59,
"eval_loss": 0.35390254855155945,
"eval_runtime": 235.3962,
"eval_samples_per_second": 17.214,
"eval_steps_per_second": 2.154,
"eval_wer": 0.23991454065602613,
"step": 5000
},
{
"epoch": 27.13,
"learning_rate": 6.696192490745638e-05,
"loss": 0.1745,
"step": 5100
},
{
"epoch": 27.13,
"eval_loss": 0.3428182601928711,
"eval_runtime": 234.2975,
"eval_samples_per_second": 17.294,
"eval_steps_per_second": 2.164,
"eval_wer": 0.2343219806459721,
"step": 5100
},
{
"epoch": 27.66,
"learning_rate": 6.630089899524061e-05,
"loss": 0.1596,
"step": 5200
},
{
"epoch": 27.66,
"eval_loss": 0.33925893902778625,
"eval_runtime": 236.9973,
"eval_samples_per_second": 17.097,
"eval_steps_per_second": 2.139,
"eval_wer": 0.234447656151816,
"step": 5200
}
],
"max_steps": 15228,
"num_train_epochs": 81,
"total_flos": 2.2833398907064197e+19,
"trial_name": null,
"trial_params": null
}