| { | |
| "best_metric": 0.5367517471313477, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-bigcgen-combined-20hrs-model/checkpoint-2400", | |
| "epoch": 2.282943525385054, | |
| "eval_steps": 200, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.019015021867275148, | |
| "grad_norm": 131.49729919433594, | |
| "learning_rate": 6.649999999999999e-07, | |
| "loss": 13.9909, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.038030043734550295, | |
| "grad_norm": 104.89710235595703, | |
| "learning_rate": 1.5399999999999999e-06, | |
| "loss": 10.094, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05704506560182544, | |
| "grad_norm": 103.4134292602539, | |
| "learning_rate": 2.415e-06, | |
| "loss": 6.8224, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07606008746910059, | |
| "grad_norm": 58.73570251464844, | |
| "learning_rate": 3.29e-06, | |
| "loss": 5.5754, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09507510933637574, | |
| "grad_norm": 84.27243041992188, | |
| "learning_rate": 4.164999999999999e-06, | |
| "loss": 4.2985, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.11409013120365089, | |
| "grad_norm": 49.697967529296875, | |
| "learning_rate": 5.039999999999999e-06, | |
| "loss": 4.0458, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13310515307092602, | |
| "grad_norm": 77.4672622680664, | |
| "learning_rate": 5.915e-06, | |
| "loss": 4.0797, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.15212017493820118, | |
| "grad_norm": 56.156681060791016, | |
| "learning_rate": 6.789999999999999e-06, | |
| "loss": 3.7679, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15212017493820118, | |
| "eval_loss": 0.8942375183105469, | |
| "eval_runtime": 545.4273, | |
| "eval_samples_per_second": 1.679, | |
| "eval_steps_per_second": 0.84, | |
| "eval_wer": 0.64549276039234, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17113519680547631, | |
| "grad_norm": 71.11776733398438, | |
| "learning_rate": 7.664999999999999e-06, | |
| "loss": 4.0009, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.19015021867275148, | |
| "grad_norm": 64.66726684570312, | |
| "learning_rate": 8.54e-06, | |
| "loss": 3.4266, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2091652405400266, | |
| "grad_norm": 58.9884033203125, | |
| "learning_rate": 9.415e-06, | |
| "loss": 3.8021, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.22818026240730177, | |
| "grad_norm": 61.482627868652344, | |
| "learning_rate": 1.0289999999999998e-05, | |
| "loss": 3.167, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2471952842745769, | |
| "grad_norm": 57.6241569519043, | |
| "learning_rate": 1.1164999999999999e-05, | |
| "loss": 3.0943, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.26621030614185204, | |
| "grad_norm": 55.91292953491211, | |
| "learning_rate": 1.2039999999999998e-05, | |
| "loss": 3.1799, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2852253280091272, | |
| "grad_norm": 64.21235656738281, | |
| "learning_rate": 1.2915e-05, | |
| "loss": 3.17, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.30424034987640236, | |
| "grad_norm": 38.54878616333008, | |
| "learning_rate": 1.379e-05, | |
| "loss": 3.4733, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.30424034987640236, | |
| "eval_loss": 0.7609195113182068, | |
| "eval_runtime": 542.694, | |
| "eval_samples_per_second": 1.688, | |
| "eval_steps_per_second": 0.844, | |
| "eval_wer": 0.5650397010742644, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3232553717436775, | |
| "grad_norm": 62.31101608276367, | |
| "learning_rate": 1.4664999999999998e-05, | |
| "loss": 3.2382, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.34227039361095263, | |
| "grad_norm": 44.90293502807617, | |
| "learning_rate": 1.554e-05, | |
| "loss": 3.476, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3612854154782278, | |
| "grad_norm": 61.89760971069336, | |
| "learning_rate": 1.6414999999999998e-05, | |
| "loss": 2.7826, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.38030043734550295, | |
| "grad_norm": 64.2639389038086, | |
| "learning_rate": 1.729e-05, | |
| "loss": 3.1842, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3993154592127781, | |
| "grad_norm": 58.504661560058594, | |
| "learning_rate": 1.7426111111111108e-05, | |
| "loss": 2.7755, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4183304810800532, | |
| "grad_norm": 52.18159484863281, | |
| "learning_rate": 1.7328888888888887e-05, | |
| "loss": 2.8868, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4373455029473284, | |
| "grad_norm": 63.206600189208984, | |
| "learning_rate": 1.7231666666666666e-05, | |
| "loss": 3.1849, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.45636052481460354, | |
| "grad_norm": 43.44758605957031, | |
| "learning_rate": 1.7134444444444442e-05, | |
| "loss": 2.6698, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45636052481460354, | |
| "eval_loss": 0.6957600712776184, | |
| "eval_runtime": 538.2917, | |
| "eval_samples_per_second": 1.702, | |
| "eval_steps_per_second": 0.851, | |
| "eval_wer": 0.5374824848201775, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4753755466818787, | |
| "grad_norm": 59.37651062011719, | |
| "learning_rate": 1.703722222222222e-05, | |
| "loss": 2.9604, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4943905685491538, | |
| "grad_norm": 49.53156661987305, | |
| "learning_rate": 1.6939999999999997e-05, | |
| "loss": 2.6304, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.513405590416429, | |
| "grad_norm": 47.221038818359375, | |
| "learning_rate": 1.6842777777777776e-05, | |
| "loss": 2.9998, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5324206122837041, | |
| "grad_norm": 55.7159309387207, | |
| "learning_rate": 1.6745555555555555e-05, | |
| "loss": 2.7585, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5514356341509793, | |
| "grad_norm": 48.83000946044922, | |
| "learning_rate": 1.664833333333333e-05, | |
| "loss": 2.7687, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.5704506560182544, | |
| "grad_norm": 41.48577880859375, | |
| "learning_rate": 1.655111111111111e-05, | |
| "loss": 2.4255, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5894656778855296, | |
| "grad_norm": 58.95222091674805, | |
| "learning_rate": 1.6453888888888886e-05, | |
| "loss": 2.9675, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.6084806997528047, | |
| "grad_norm": 65.72980499267578, | |
| "learning_rate": 1.6356666666666665e-05, | |
| "loss": 2.5647, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6084806997528047, | |
| "eval_loss": 0.6684515476226807, | |
| "eval_runtime": 543.3556, | |
| "eval_samples_per_second": 1.686, | |
| "eval_steps_per_second": 0.843, | |
| "eval_wer": 0.5496263428304531, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6274957216200798, | |
| "grad_norm": 55.358070373535156, | |
| "learning_rate": 1.6259444444444445e-05, | |
| "loss": 2.713, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.646510743487355, | |
| "grad_norm": 61.67774963378906, | |
| "learning_rate": 1.616222222222222e-05, | |
| "loss": 3.1045, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6655257653546301, | |
| "grad_norm": 58.04518127441406, | |
| "learning_rate": 1.6065e-05, | |
| "loss": 2.8299, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.6845407872219053, | |
| "grad_norm": 57.893192291259766, | |
| "learning_rate": 1.5967777777777775e-05, | |
| "loss": 2.7055, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7035558090891805, | |
| "grad_norm": 53.013309478759766, | |
| "learning_rate": 1.5870555555555554e-05, | |
| "loss": 2.6981, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.7225708309564556, | |
| "grad_norm": 45.20479202270508, | |
| "learning_rate": 1.577333333333333e-05, | |
| "loss": 3.0628, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7415858528237308, | |
| "grad_norm": 46.2863655090332, | |
| "learning_rate": 1.567611111111111e-05, | |
| "loss": 2.6797, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.7606008746910059, | |
| "grad_norm": 42.149932861328125, | |
| "learning_rate": 1.557888888888889e-05, | |
| "loss": 2.1636, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7606008746910059, | |
| "eval_loss": 0.6228325963020325, | |
| "eval_runtime": 556.097, | |
| "eval_samples_per_second": 1.647, | |
| "eval_steps_per_second": 0.824, | |
| "eval_wer": 0.5102755721625408, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.779615896558281, | |
| "grad_norm": 49.531471252441406, | |
| "learning_rate": 1.5481666666666668e-05, | |
| "loss": 2.8739, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.7986309184255562, | |
| "grad_norm": 39.97299575805664, | |
| "learning_rate": 1.5384444444444444e-05, | |
| "loss": 2.595, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8176459402928313, | |
| "grad_norm": 42.683372497558594, | |
| "learning_rate": 1.528722222222222e-05, | |
| "loss": 2.4377, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.8366609621601064, | |
| "grad_norm": 27.87632942199707, | |
| "learning_rate": 1.5189999999999999e-05, | |
| "loss": 2.2694, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8556759840273817, | |
| "grad_norm": 40.27122116088867, | |
| "learning_rate": 1.5092777777777776e-05, | |
| "loss": 2.6472, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.8746910058946568, | |
| "grad_norm": 53.14582824707031, | |
| "learning_rate": 1.4995555555555555e-05, | |
| "loss": 2.5136, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.893706027761932, | |
| "grad_norm": 66.0662612915039, | |
| "learning_rate": 1.4898333333333333e-05, | |
| "loss": 2.5476, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.9127210496292071, | |
| "grad_norm": 48.282222747802734, | |
| "learning_rate": 1.4801111111111109e-05, | |
| "loss": 2.7265, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9127210496292071, | |
| "eval_loss": 0.5869264602661133, | |
| "eval_runtime": 539.0867, | |
| "eval_samples_per_second": 1.699, | |
| "eval_steps_per_second": 0.85, | |
| "eval_wer": 0.4705744978981784, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9317360714964822, | |
| "grad_norm": 36.0697135925293, | |
| "learning_rate": 1.4703888888888888e-05, | |
| "loss": 2.1289, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.9507510933637574, | |
| "grad_norm": 36.52973556518555, | |
| "learning_rate": 1.4606666666666665e-05, | |
| "loss": 2.2548, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9697661152310325, | |
| "grad_norm": 36.967796325683594, | |
| "learning_rate": 1.4509444444444444e-05, | |
| "loss": 2.263, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.9887811370983076, | |
| "grad_norm": 44.840248107910156, | |
| "learning_rate": 1.4412222222222222e-05, | |
| "loss": 2.6958, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.008366609621601, | |
| "grad_norm": 45.30537796020508, | |
| "learning_rate": 1.4314999999999998e-05, | |
| "loss": 2.3112, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.0273816314888762, | |
| "grad_norm": 41.10464859008789, | |
| "learning_rate": 1.4217777777777777e-05, | |
| "loss": 2.1152, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.0463966533561513, | |
| "grad_norm": 38.10401153564453, | |
| "learning_rate": 1.4120555555555554e-05, | |
| "loss": 1.7989, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.0654116752234266, | |
| "grad_norm": 30.451261520385742, | |
| "learning_rate": 1.4023333333333332e-05, | |
| "loss": 1.5404, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0654116752234266, | |
| "eval_loss": 0.598966121673584, | |
| "eval_runtime": 530.9607, | |
| "eval_samples_per_second": 1.725, | |
| "eval_steps_per_second": 0.863, | |
| "eval_wer": 0.45422699673049977, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0844266970907017, | |
| "grad_norm": 33.9229736328125, | |
| "learning_rate": 1.3926111111111111e-05, | |
| "loss": 1.6963, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.1034417189579768, | |
| "grad_norm": 37.29084396362305, | |
| "learning_rate": 1.3828888888888888e-05, | |
| "loss": 1.9026, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.122456740825252, | |
| "grad_norm": 37.309139251708984, | |
| "learning_rate": 1.3731666666666664e-05, | |
| "loss": 1.7798, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.141471762692527, | |
| "grad_norm": 27.37919807434082, | |
| "learning_rate": 1.3634444444444443e-05, | |
| "loss": 1.8952, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1604867845598021, | |
| "grad_norm": 47.561519622802734, | |
| "learning_rate": 1.3537222222222221e-05, | |
| "loss": 1.7325, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.1795018064270775, | |
| "grad_norm": 46.943992614746094, | |
| "learning_rate": 1.3439999999999998e-05, | |
| "loss": 1.71, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.1985168282943526, | |
| "grad_norm": 48.82526397705078, | |
| "learning_rate": 1.3342777777777778e-05, | |
| "loss": 1.8316, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.2175318501616277, | |
| "grad_norm": 45.330326080322266, | |
| "learning_rate": 1.3245555555555555e-05, | |
| "loss": 1.9844, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2175318501616277, | |
| "eval_loss": 0.5892675518989563, | |
| "eval_runtime": 532.0087, | |
| "eval_samples_per_second": 1.722, | |
| "eval_steps_per_second": 0.861, | |
| "eval_wer": 0.4642690331620738, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2365468720289028, | |
| "grad_norm": 34.17546463012695, | |
| "learning_rate": 1.314833333333333e-05, | |
| "loss": 1.7767, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.2555618938961781, | |
| "grad_norm": 34.69303512573242, | |
| "learning_rate": 1.305111111111111e-05, | |
| "loss": 1.8796, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.2745769157634532, | |
| "grad_norm": 35.748783111572266, | |
| "learning_rate": 1.2953888888888887e-05, | |
| "loss": 1.7917, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.2935919376307283, | |
| "grad_norm": 51.4835319519043, | |
| "learning_rate": 1.2856666666666667e-05, | |
| "loss": 2.0249, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3126069594980034, | |
| "grad_norm": 21.467344284057617, | |
| "learning_rate": 1.2759444444444444e-05, | |
| "loss": 1.8292, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.3316219813652785, | |
| "grad_norm": 35.42878723144531, | |
| "learning_rate": 1.266222222222222e-05, | |
| "loss": 1.852, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.3506370032325536, | |
| "grad_norm": 47.91001510620117, | |
| "learning_rate": 1.2564999999999999e-05, | |
| "loss": 1.7353, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.3696520250998288, | |
| "grad_norm": 30.44899559020996, | |
| "learning_rate": 1.2467777777777777e-05, | |
| "loss": 1.6926, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.3696520250998288, | |
| "eval_loss": 0.5730497241020203, | |
| "eval_runtime": 512.6986, | |
| "eval_samples_per_second": 1.787, | |
| "eval_steps_per_second": 0.893, | |
| "eval_wer": 0.4412657636618403, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.388667046967104, | |
| "grad_norm": 34.98470687866211, | |
| "learning_rate": 1.2370555555555554e-05, | |
| "loss": 1.956, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.4076820688343792, | |
| "grad_norm": 28.451519012451172, | |
| "learning_rate": 1.2273333333333333e-05, | |
| "loss": 1.6875, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.4266970907016543, | |
| "grad_norm": 54.90264129638672, | |
| "learning_rate": 1.217611111111111e-05, | |
| "loss": 1.68, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.4457121125689294, | |
| "grad_norm": 51.0279655456543, | |
| "learning_rate": 1.2078888888888887e-05, | |
| "loss": 1.7196, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4647271344362047, | |
| "grad_norm": 37.92000961303711, | |
| "learning_rate": 1.1981666666666666e-05, | |
| "loss": 1.9185, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.4837421563034798, | |
| "grad_norm": 37.440956115722656, | |
| "learning_rate": 1.1884444444444443e-05, | |
| "loss": 1.7995, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.502757178170755, | |
| "grad_norm": 29.802404403686523, | |
| "learning_rate": 1.178722222222222e-05, | |
| "loss": 1.9844, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.52177220003803, | |
| "grad_norm": 44.70022964477539, | |
| "learning_rate": 1.169e-05, | |
| "loss": 1.8654, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.52177220003803, | |
| "eval_loss": 0.5549585223197937, | |
| "eval_runtime": 524.5769, | |
| "eval_samples_per_second": 1.746, | |
| "eval_steps_per_second": 0.873, | |
| "eval_wer": 0.4599486221391873, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5407872219053051, | |
| "grad_norm": 37.87806701660156, | |
| "learning_rate": 1.1592777777777776e-05, | |
| "loss": 1.6249, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.5598022437725803, | |
| "grad_norm": 48.078773498535156, | |
| "learning_rate": 1.1495555555555553e-05, | |
| "loss": 1.8834, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.5788172656398554, | |
| "grad_norm": 33.26875305175781, | |
| "learning_rate": 1.1398333333333332e-05, | |
| "loss": 1.8992, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.5978322875071307, | |
| "grad_norm": 34.35662841796875, | |
| "learning_rate": 1.130111111111111e-05, | |
| "loss": 1.8728, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.6168473093744058, | |
| "grad_norm": 29.405532836914062, | |
| "learning_rate": 1.1203888888888889e-05, | |
| "loss": 1.7595, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.635862331241681, | |
| "grad_norm": 33.51266098022461, | |
| "learning_rate": 1.1106666666666666e-05, | |
| "loss": 1.8762, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.6548773531089562, | |
| "grad_norm": 38.80997848510742, | |
| "learning_rate": 1.1009444444444442e-05, | |
| "loss": 2.0687, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.6738923749762313, | |
| "grad_norm": 43.26702880859375, | |
| "learning_rate": 1.0912222222222221e-05, | |
| "loss": 1.8045, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6738923749762313, | |
| "eval_loss": 0.5445375442504883, | |
| "eval_runtime": 530.1331, | |
| "eval_samples_per_second": 1.728, | |
| "eval_steps_per_second": 0.864, | |
| "eval_wer": 0.41779542269967307, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6929073968435064, | |
| "grad_norm": 36.24801254272461, | |
| "learning_rate": 1.0814999999999999e-05, | |
| "loss": 2.0172, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.7119224187107815, | |
| "grad_norm": 32.35186767578125, | |
| "learning_rate": 1.0717777777777776e-05, | |
| "loss": 1.6788, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.7309374405780567, | |
| "grad_norm": 39.30363845825195, | |
| "learning_rate": 1.0620555555555556e-05, | |
| "loss": 1.5118, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.7499524624453318, | |
| "grad_norm": 37.63472366333008, | |
| "learning_rate": 1.0523333333333333e-05, | |
| "loss": 1.8138, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.7689674843126069, | |
| "grad_norm": 35.43053436279297, | |
| "learning_rate": 1.0426111111111109e-05, | |
| "loss": 1.7971, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.787982506179882, | |
| "grad_norm": 41.22832107543945, | |
| "learning_rate": 1.0328888888888888e-05, | |
| "loss": 1.7849, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.8069975280471573, | |
| "grad_norm": 34.19541931152344, | |
| "learning_rate": 1.0231666666666666e-05, | |
| "loss": 1.6543, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.8260125499144324, | |
| "grad_norm": 44.1766242980957, | |
| "learning_rate": 1.0134444444444445e-05, | |
| "loss": 1.8258, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.8260125499144324, | |
| "eval_loss": 0.5367517471313477, | |
| "eval_runtime": 534.8861, | |
| "eval_samples_per_second": 1.713, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.4210649229332088, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.8450275717817075, | |
| "grad_norm": 53.05873107910156, | |
| "learning_rate": 1.0037222222222222e-05, | |
| "loss": 1.651, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.8640425936489828, | |
| "grad_norm": 49.550086975097656, | |
| "learning_rate": 9.939999999999998e-06, | |
| "loss": 1.8205, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.883057615516258, | |
| "grad_norm": 39.620155334472656, | |
| "learning_rate": 9.842777777777777e-06, | |
| "loss": 1.8859, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.902072637383533, | |
| "grad_norm": 24.356233596801758, | |
| "learning_rate": 9.745555555555555e-06, | |
| "loss": 1.8407, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9210876592508082, | |
| "grad_norm": 36.06525421142578, | |
| "learning_rate": 9.648333333333332e-06, | |
| "loss": 1.5886, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.9401026811180833, | |
| "grad_norm": 32.340152740478516, | |
| "learning_rate": 9.551111111111111e-06, | |
| "loss": 1.6434, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.9591177029853584, | |
| "grad_norm": 22.598873138427734, | |
| "learning_rate": 9.453888888888889e-06, | |
| "loss": 1.9292, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.9781327248526335, | |
| "grad_norm": 28.097835540771484, | |
| "learning_rate": 9.356666666666665e-06, | |
| "loss": 1.543, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.9781327248526335, | |
| "eval_loss": 0.5370936393737793, | |
| "eval_runtime": 535.3346, | |
| "eval_samples_per_second": 1.711, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.42445119103222795, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.9971477467199086, | |
| "grad_norm": 40.16139221191406, | |
| "learning_rate": 9.259444444444444e-06, | |
| "loss": 2.2658, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.016733219243202, | |
| "grad_norm": 34.17202377319336, | |
| "learning_rate": 9.162222222222221e-06, | |
| "loss": 1.237, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.0357482411104773, | |
| "grad_norm": 45.333221435546875, | |
| "learning_rate": 9.064999999999999e-06, | |
| "loss": 0.977, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.0547632629777524, | |
| "grad_norm": 39.80405807495117, | |
| "learning_rate": 8.967777777777778e-06, | |
| "loss": 1.0369, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.0737782848450275, | |
| "grad_norm": 37.61433029174805, | |
| "learning_rate": 8.870555555555555e-06, | |
| "loss": 0.9978, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.0927933067123026, | |
| "grad_norm": 44.44002914428711, | |
| "learning_rate": 8.773333333333331e-06, | |
| "loss": 1.0163, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.1118083285795777, | |
| "grad_norm": 35.349544525146484, | |
| "learning_rate": 8.67611111111111e-06, | |
| "loss": 0.9823, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.1308233504468532, | |
| "grad_norm": 25.33350372314453, | |
| "learning_rate": 8.578888888888888e-06, | |
| "loss": 0.9667, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.1308233504468532, | |
| "eval_loss": 0.5587082505226135, | |
| "eval_runtime": 547.2043, | |
| "eval_samples_per_second": 1.674, | |
| "eval_steps_per_second": 0.837, | |
| "eval_wer": 0.4544605324614666, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.1498383723141283, | |
| "grad_norm": 31.81471061706543, | |
| "learning_rate": 8.481666666666667e-06, | |
| "loss": 1.3269, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 2.1688533941814034, | |
| "grad_norm": 18.468244552612305, | |
| "learning_rate": 8.384444444444443e-06, | |
| "loss": 1.0429, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.1878684160486785, | |
| "grad_norm": 20.408119201660156, | |
| "learning_rate": 8.287222222222222e-06, | |
| "loss": 1.2085, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 2.2068834379159536, | |
| "grad_norm": 45.54349136352539, | |
| "learning_rate": 8.19e-06, | |
| "loss": 0.9542, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.2258984597832288, | |
| "grad_norm": 38.3132209777832, | |
| "learning_rate": 8.092777777777777e-06, | |
| "loss": 1.0642, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 2.244913481650504, | |
| "grad_norm": 29.104040145874023, | |
| "learning_rate": 7.995555555555554e-06, | |
| "loss": 1.0053, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.263928503517779, | |
| "grad_norm": 33.763031005859375, | |
| "learning_rate": 7.898333333333332e-06, | |
| "loss": 1.3275, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 2.282943525385054, | |
| "grad_norm": 32.66128921508789, | |
| "learning_rate": 7.801111111111111e-06, | |
| "loss": 1.0216, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.282943525385054, | |
| "eval_loss": 0.5617040395736694, | |
| "eval_runtime": 528.9401, | |
| "eval_samples_per_second": 1.732, | |
| "eval_steps_per_second": 0.866, | |
| "eval_wer": 0.4096216721158337, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.282943525385054, | |
| "step": 3000, | |
| "total_flos": 2.450268261974016e+19, | |
| "train_loss": 2.4239137687683106, | |
| "train_runtime": 14144.0718, | |
| "train_samples_per_second": 2.828, | |
| "train_steps_per_second": 0.354 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.450268261974016e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |