| { | |
| "best_metric": 0.36100757122039795, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-swagen-combined-30hrs-model/checkpoint-2400", | |
| "epoch": 1.2421608196212357, | |
| "eval_steps": 200, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01034875297526648, | |
| "grad_norm": 139.34642028808594, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 10.2474, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02069750595053296, | |
| "grad_norm": 89.0820541381836, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 8.2652, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03104625892579944, | |
| "grad_norm": 85.51993560791016, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 6.0026, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04139501190106592, | |
| "grad_norm": 91.87001037597656, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 4.5811, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0517437648763324, | |
| "grad_norm": 79.70684814453125, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 3.2507, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06209251785159888, | |
| "grad_norm": 72.15753936767578, | |
| "learning_rate": 2.9e-06, | |
| "loss": 3.2154, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07244127082686536, | |
| "grad_norm": 60.531593322753906, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 2.8239, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.08279002380213184, | |
| "grad_norm": 60.771507263183594, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 2.7508, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08279002380213184, | |
| "eval_loss": 0.8133686184883118, | |
| "eval_runtime": 566.5007, | |
| "eval_samples_per_second": 1.919, | |
| "eval_steps_per_second": 0.96, | |
| "eval_wer": 0.4877232142857143, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09313877677739832, | |
| "grad_norm": 51.035518646240234, | |
| "learning_rate": 4.4e-06, | |
| "loss": 2.2578, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1034875297526648, | |
| "grad_norm": 69.9592514038086, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 2.4931, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.11383628272793128, | |
| "grad_norm": 55.8157844543457, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 2.2654, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.12418503570319776, | |
| "grad_norm": 57.47214889526367, | |
| "learning_rate": 5.9e-06, | |
| "loss": 2.1863, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.13453378867846424, | |
| "grad_norm": 48.41241455078125, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 2.1504, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.1448825416537307, | |
| "grad_norm": 61.10322570800781, | |
| "learning_rate": 6.9e-06, | |
| "loss": 2.1106, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1552312946289972, | |
| "grad_norm": 49.85371017456055, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.9696, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.16558004760426367, | |
| "grad_norm": 78.61709594726562, | |
| "learning_rate": 7.9e-06, | |
| "loss": 1.8748, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.16558004760426367, | |
| "eval_loss": 0.6291218996047974, | |
| "eval_runtime": 561.4046, | |
| "eval_samples_per_second": 1.936, | |
| "eval_steps_per_second": 0.969, | |
| "eval_wer": 0.3898133116883117, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.17592880057953017, | |
| "grad_norm": 58.551658630371094, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.848, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.18627755355479664, | |
| "grad_norm": 37.77573013305664, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 1.9659, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.19662630653006313, | |
| "grad_norm": 67.77395629882812, | |
| "learning_rate": 9.4e-06, | |
| "loss": 2.0312, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.2069750595053296, | |
| "grad_norm": 48.11214828491211, | |
| "learning_rate": 9.9e-06, | |
| "loss": 1.795, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2173238124805961, | |
| "grad_norm": 54.6898307800293, | |
| "learning_rate": 9.997220291869355e-06, | |
| "loss": 1.8759, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.22767256545586256, | |
| "grad_norm": 56.6167106628418, | |
| "learning_rate": 9.993745656706047e-06, | |
| "loss": 1.9263, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.23802131843112906, | |
| "grad_norm": 43.32560729980469, | |
| "learning_rate": 9.990271021542739e-06, | |
| "loss": 1.889, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.24837007140639553, | |
| "grad_norm": 49.59733581542969, | |
| "learning_rate": 9.986796386379432e-06, | |
| "loss": 1.6214, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.24837007140639553, | |
| "eval_loss": 0.5559969544410706, | |
| "eval_runtime": 561.0579, | |
| "eval_samples_per_second": 1.937, | |
| "eval_steps_per_second": 0.97, | |
| "eval_wer": 0.3431412337662338, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.258718824381662, | |
| "grad_norm": 40.30105972290039, | |
| "learning_rate": 9.983321751216123e-06, | |
| "loss": 1.6189, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.2690675773569285, | |
| "grad_norm": 42.67938995361328, | |
| "learning_rate": 9.979847116052815e-06, | |
| "loss": 1.3983, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.279416330332195, | |
| "grad_norm": 36.10395431518555, | |
| "learning_rate": 9.976372480889508e-06, | |
| "loss": 1.6815, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.2897650833074614, | |
| "grad_norm": 42.29354476928711, | |
| "learning_rate": 9.9728978457262e-06, | |
| "loss": 1.5022, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3001138362827279, | |
| "grad_norm": 50.72279357910156, | |
| "learning_rate": 9.969423210562891e-06, | |
| "loss": 1.5644, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.3104625892579944, | |
| "grad_norm": 49.18547821044922, | |
| "learning_rate": 9.965948575399585e-06, | |
| "loss": 1.6269, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3208113422332609, | |
| "grad_norm": 42.0450325012207, | |
| "learning_rate": 9.962473940236276e-06, | |
| "loss": 1.4277, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.33116009520852735, | |
| "grad_norm": 58.25082015991211, | |
| "learning_rate": 9.958999305072968e-06, | |
| "loss": 1.559, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.33116009520852735, | |
| "eval_loss": 0.49684250354766846, | |
| "eval_runtime": 559.6818, | |
| "eval_samples_per_second": 1.942, | |
| "eval_steps_per_second": 0.972, | |
| "eval_wer": 0.29525162337662336, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.34150884818379384, | |
| "grad_norm": 37.699710845947266, | |
| "learning_rate": 9.955524669909661e-06, | |
| "loss": 1.2958, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.35185760115906034, | |
| "grad_norm": 30.309236526489258, | |
| "learning_rate": 9.952050034746353e-06, | |
| "loss": 1.4174, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.36220635413432684, | |
| "grad_norm": 43.847652435302734, | |
| "learning_rate": 9.948575399583044e-06, | |
| "loss": 1.5185, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.3725551071095933, | |
| "grad_norm": 57.50359344482422, | |
| "learning_rate": 9.945100764419738e-06, | |
| "loss": 1.5072, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.38290386008485977, | |
| "grad_norm": 43.106422424316406, | |
| "learning_rate": 9.941626129256429e-06, | |
| "loss": 1.4437, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.39325261306012627, | |
| "grad_norm": 38.97978591918945, | |
| "learning_rate": 9.93815149409312e-06, | |
| "loss": 1.4267, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.40360136603539276, | |
| "grad_norm": 39.729026794433594, | |
| "learning_rate": 9.934676858929814e-06, | |
| "loss": 1.4166, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.4139501190106592, | |
| "grad_norm": 45.119964599609375, | |
| "learning_rate": 9.931202223766506e-06, | |
| "loss": 1.3616, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4139501190106592, | |
| "eval_loss": 0.4720439910888672, | |
| "eval_runtime": 559.355, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 0.2872362012987013, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4242988719859257, | |
| "grad_norm": 39.95614242553711, | |
| "learning_rate": 9.927727588603197e-06, | |
| "loss": 1.4632, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.4346476249611922, | |
| "grad_norm": 26.40887451171875, | |
| "learning_rate": 9.92425295343989e-06, | |
| "loss": 1.2984, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.44499637793645863, | |
| "grad_norm": 42.91450500488281, | |
| "learning_rate": 9.920778318276582e-06, | |
| "loss": 1.5535, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.4553451309117251, | |
| "grad_norm": 61.96196746826172, | |
| "learning_rate": 9.917303683113274e-06, | |
| "loss": 1.2644, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4656938838869916, | |
| "grad_norm": 54.67252731323242, | |
| "learning_rate": 9.913829047949967e-06, | |
| "loss": 1.1995, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.4760426368622581, | |
| "grad_norm": 46.1182975769043, | |
| "learning_rate": 9.910354412786658e-06, | |
| "loss": 1.5119, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.48639138983752456, | |
| "grad_norm": 43.22040557861328, | |
| "learning_rate": 9.90687977762335e-06, | |
| "loss": 1.244, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.49674014281279105, | |
| "grad_norm": 45.609161376953125, | |
| "learning_rate": 9.903405142460043e-06, | |
| "loss": 1.3078, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.49674014281279105, | |
| "eval_loss": 0.4577382802963257, | |
| "eval_runtime": 550.5837, | |
| "eval_samples_per_second": 1.974, | |
| "eval_steps_per_second": 0.988, | |
| "eval_wer": 0.29778814935064934, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5070888957880575, | |
| "grad_norm": 57.24745178222656, | |
| "learning_rate": 9.899930507296735e-06, | |
| "loss": 1.2625, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.517437648763324, | |
| "grad_norm": 29.896320343017578, | |
| "learning_rate": 9.896455872133426e-06, | |
| "loss": 1.2438, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5277864017385905, | |
| "grad_norm": 56.462646484375, | |
| "learning_rate": 9.89298123697012e-06, | |
| "loss": 1.3759, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.538135154713857, | |
| "grad_norm": 39.78219985961914, | |
| "learning_rate": 9.889506601806811e-06, | |
| "loss": 1.3728, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5484839076891235, | |
| "grad_norm": 45.62682342529297, | |
| "learning_rate": 9.886031966643503e-06, | |
| "loss": 1.3165, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.55883266066439, | |
| "grad_norm": 41.924949645996094, | |
| "learning_rate": 9.882557331480196e-06, | |
| "loss": 1.3968, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.5691814136396565, | |
| "grad_norm": 39.86012649536133, | |
| "learning_rate": 9.879082696316888e-06, | |
| "loss": 1.1669, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.5795301666149228, | |
| "grad_norm": 30.276840209960938, | |
| "learning_rate": 9.87560806115358e-06, | |
| "loss": 1.2579, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5795301666149228, | |
| "eval_loss": 0.42178860306739807, | |
| "eval_runtime": 565.0754, | |
| "eval_samples_per_second": 1.924, | |
| "eval_steps_per_second": 0.963, | |
| "eval_wer": 0.2757711038961039, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5898789195901893, | |
| "grad_norm": 27.80983543395996, | |
| "learning_rate": 9.872133425990272e-06, | |
| "loss": 1.1973, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.6002276725654558, | |
| "grad_norm": 54.13436508178711, | |
| "learning_rate": 9.868658790826964e-06, | |
| "loss": 1.3214, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6105764255407223, | |
| "grad_norm": 31.381301879882812, | |
| "learning_rate": 9.865184155663656e-06, | |
| "loss": 1.3413, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.6209251785159888, | |
| "grad_norm": 35.9166259765625, | |
| "learning_rate": 9.861709520500349e-06, | |
| "loss": 1.2117, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6312739314912553, | |
| "grad_norm": 37.55325698852539, | |
| "learning_rate": 9.85823488533704e-06, | |
| "loss": 1.2422, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.6416226844665218, | |
| "grad_norm": 32.475807189941406, | |
| "learning_rate": 9.854760250173732e-06, | |
| "loss": 1.2666, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.6519714374417883, | |
| "grad_norm": 35.703243255615234, | |
| "learning_rate": 9.851285615010425e-06, | |
| "loss": 1.1171, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.6623201904170547, | |
| "grad_norm": 41.8420524597168, | |
| "learning_rate": 9.847810979847117e-06, | |
| "loss": 1.214, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6623201904170547, | |
| "eval_loss": 0.41559898853302, | |
| "eval_runtime": 555.4115, | |
| "eval_samples_per_second": 1.957, | |
| "eval_steps_per_second": 0.979, | |
| "eval_wer": 0.26542207792207795, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6726689433923212, | |
| "grad_norm": 38.391361236572266, | |
| "learning_rate": 9.844336344683808e-06, | |
| "loss": 1.2246, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.6830176963675877, | |
| "grad_norm": 57.34391784667969, | |
| "learning_rate": 9.840861709520502e-06, | |
| "loss": 1.0536, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.6933664493428542, | |
| "grad_norm": 19.72743034362793, | |
| "learning_rate": 9.837387074357193e-06, | |
| "loss": 1.2277, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.7037152023181207, | |
| "grad_norm": 38.91509246826172, | |
| "learning_rate": 9.833912439193885e-06, | |
| "loss": 1.2657, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7140639552933872, | |
| "grad_norm": 51.42982864379883, | |
| "learning_rate": 9.830437804030578e-06, | |
| "loss": 1.2865, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.7244127082686537, | |
| "grad_norm": 37.256927490234375, | |
| "learning_rate": 9.82696316886727e-06, | |
| "loss": 1.1125, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.73476146124392, | |
| "grad_norm": 37.06630325317383, | |
| "learning_rate": 9.823488533703961e-06, | |
| "loss": 1.2179, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.7451102142191866, | |
| "grad_norm": 28.06951332092285, | |
| "learning_rate": 9.820013898540655e-06, | |
| "loss": 1.0719, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7451102142191866, | |
| "eval_loss": 0.40048521757125854, | |
| "eval_runtime": 544.8903, | |
| "eval_samples_per_second": 1.995, | |
| "eval_steps_per_second": 0.998, | |
| "eval_wer": 0.2315340909090909, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.755458967194453, | |
| "grad_norm": 43.74811553955078, | |
| "learning_rate": 9.816539263377346e-06, | |
| "loss": 1.1466, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.7658077201697195, | |
| "grad_norm": 67.86227416992188, | |
| "learning_rate": 9.813064628214038e-06, | |
| "loss": 1.2406, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.776156473144986, | |
| "grad_norm": 21.888675689697266, | |
| "learning_rate": 9.809589993050731e-06, | |
| "loss": 1.0934, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.7865052261202525, | |
| "grad_norm": 34.76344680786133, | |
| "learning_rate": 9.806115357887423e-06, | |
| "loss": 1.3778, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.796853979095519, | |
| "grad_norm": 37.02476501464844, | |
| "learning_rate": 9.802640722724114e-06, | |
| "loss": 1.1266, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.8072027320707855, | |
| "grad_norm": 25.008710861206055, | |
| "learning_rate": 9.799166087560807e-06, | |
| "loss": 0.9663, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.8175514850460519, | |
| "grad_norm": 43.19198989868164, | |
| "learning_rate": 9.795691452397499e-06, | |
| "loss": 1.2192, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.8279002380213184, | |
| "grad_norm": 42.85536193847656, | |
| "learning_rate": 9.79221681723419e-06, | |
| "loss": 1.0432, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8279002380213184, | |
| "eval_loss": 0.3864189684391022, | |
| "eval_runtime": 555.4403, | |
| "eval_samples_per_second": 1.957, | |
| "eval_steps_per_second": 0.979, | |
| "eval_wer": 0.24330357142857142, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8382489909965849, | |
| "grad_norm": 42.24909591674805, | |
| "learning_rate": 9.788742182070884e-06, | |
| "loss": 1.2698, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.8485977439718514, | |
| "grad_norm": 30.07208824157715, | |
| "learning_rate": 9.785267546907575e-06, | |
| "loss": 1.0744, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.8589464969471179, | |
| "grad_norm": 40.065677642822266, | |
| "learning_rate": 9.781792911744267e-06, | |
| "loss": 1.0424, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.8692952499223844, | |
| "grad_norm": 33.75371170043945, | |
| "learning_rate": 9.77831827658096e-06, | |
| "loss": 1.1718, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8796440028976509, | |
| "grad_norm": 29.951263427734375, | |
| "learning_rate": 9.774843641417652e-06, | |
| "loss": 1.0589, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.8899927558729173, | |
| "grad_norm": 48.64168930053711, | |
| "learning_rate": 9.771369006254343e-06, | |
| "loss": 1.253, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.9003415088481838, | |
| "grad_norm": 50.58803939819336, | |
| "learning_rate": 9.767894371091037e-06, | |
| "loss": 1.1886, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.9106902618234503, | |
| "grad_norm": 40.60319900512695, | |
| "learning_rate": 9.764419735927728e-06, | |
| "loss": 0.9825, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.9106902618234503, | |
| "eval_loss": 0.3742503523826599, | |
| "eval_runtime": 564.0293, | |
| "eval_samples_per_second": 1.927, | |
| "eval_steps_per_second": 0.964, | |
| "eval_wer": 0.22067775974025974, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.9210390147987167, | |
| "grad_norm": 42.79588317871094, | |
| "learning_rate": 9.760945100764422e-06, | |
| "loss": 1.0184, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.9313877677739832, | |
| "grad_norm": 59.941490173339844, | |
| "learning_rate": 9.757470465601113e-06, | |
| "loss": 1.2345, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.9417365207492497, | |
| "grad_norm": 45.79833984375, | |
| "learning_rate": 9.753995830437805e-06, | |
| "loss": 1.0212, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.9520852737245162, | |
| "grad_norm": 41.137176513671875, | |
| "learning_rate": 9.750521195274498e-06, | |
| "loss": 1.2013, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.9624340266997827, | |
| "grad_norm": 36.94132614135742, | |
| "learning_rate": 9.74704656011119e-06, | |
| "loss": 1.1972, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.9727827796750491, | |
| "grad_norm": 27.98811912536621, | |
| "learning_rate": 9.743571924947881e-06, | |
| "loss": 1.088, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.9831315326503156, | |
| "grad_norm": 20.012388229370117, | |
| "learning_rate": 9.740097289784574e-06, | |
| "loss": 0.9881, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.9934802856255821, | |
| "grad_norm": 40.9871940612793, | |
| "learning_rate": 9.736622654621266e-06, | |
| "loss": 1.0952, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9934802856255821, | |
| "eval_loss": 0.36100757122039795, | |
| "eval_runtime": 567.8727, | |
| "eval_samples_per_second": 1.914, | |
| "eval_steps_per_second": 0.958, | |
| "eval_wer": 0.2234172077922078, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.0041395011901066, | |
| "grad_norm": 36.53849411010742, | |
| "learning_rate": 9.733148019457958e-06, | |
| "loss": 0.8935, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.014488254165373, | |
| "grad_norm": 34.85436248779297, | |
| "learning_rate": 9.72967338429465e-06, | |
| "loss": 0.6152, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.0248370071406396, | |
| "grad_norm": 22.534700393676758, | |
| "learning_rate": 9.726198749131342e-06, | |
| "loss": 0.5735, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.035185760115906, | |
| "grad_norm": 23.778539657592773, | |
| "learning_rate": 9.722724113968034e-06, | |
| "loss": 0.5496, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0455345130911726, | |
| "grad_norm": 36.943050384521484, | |
| "learning_rate": 9.719249478804727e-06, | |
| "loss": 0.6067, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.055883266066439, | |
| "grad_norm": 21.048763275146484, | |
| "learning_rate": 9.715774843641419e-06, | |
| "loss": 0.5577, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.0662320190417054, | |
| "grad_norm": 31.97542953491211, | |
| "learning_rate": 9.71230020847811e-06, | |
| "loss": 0.5266, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.076580772016972, | |
| "grad_norm": 36.05984115600586, | |
| "learning_rate": 9.708825573314804e-06, | |
| "loss": 0.6001, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.076580772016972, | |
| "eval_loss": 0.388786643743515, | |
| "eval_runtime": 572.4362, | |
| "eval_samples_per_second": 1.899, | |
| "eval_steps_per_second": 0.95, | |
| "eval_wer": 0.24228896103896103, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.0869295249922384, | |
| "grad_norm": 35.275272369384766, | |
| "learning_rate": 9.705350938151495e-06, | |
| "loss": 0.6265, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.097278277967505, | |
| "grad_norm": 23.014320373535156, | |
| "learning_rate": 9.701876302988187e-06, | |
| "loss": 0.5507, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.1076270309427714, | |
| "grad_norm": 18.49266242980957, | |
| "learning_rate": 9.69840166782488e-06, | |
| "loss": 0.6563, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.117975783918038, | |
| "grad_norm": 22.754758834838867, | |
| "learning_rate": 9.694927032661572e-06, | |
| "loss": 0.633, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.1283245368933044, | |
| "grad_norm": 24.042457580566406, | |
| "learning_rate": 9.691452397498263e-06, | |
| "loss": 0.5714, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.1386732898685707, | |
| "grad_norm": 31.35757827758789, | |
| "learning_rate": 9.687977762334956e-06, | |
| "loss": 0.6269, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.1490220428438374, | |
| "grad_norm": 15.74731731414795, | |
| "learning_rate": 9.684503127171648e-06, | |
| "loss": 0.633, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.1593707958191037, | |
| "grad_norm": 25.903976440429688, | |
| "learning_rate": 9.68102849200834e-06, | |
| "loss": 0.5491, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.1593707958191037, | |
| "eval_loss": 0.37300390005111694, | |
| "eval_runtime": 573.7758, | |
| "eval_samples_per_second": 1.894, | |
| "eval_steps_per_second": 0.948, | |
| "eval_wer": 0.22646103896103897, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.1697195487943703, | |
| "grad_norm": 36.39768981933594, | |
| "learning_rate": 9.677553856845033e-06, | |
| "loss": 0.5838, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.1800683017696367, | |
| "grad_norm": 36.48637390136719, | |
| "learning_rate": 9.674079221681724e-06, | |
| "loss": 0.5071, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.1904170547449033, | |
| "grad_norm": 43.22209167480469, | |
| "learning_rate": 9.670604586518416e-06, | |
| "loss": 0.5702, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.2007658077201697, | |
| "grad_norm": 26.855955123901367, | |
| "learning_rate": 9.66712995135511e-06, | |
| "loss": 0.6014, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.211114560695436, | |
| "grad_norm": 28.16461181640625, | |
| "learning_rate": 9.663655316191801e-06, | |
| "loss": 0.6213, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.2214633136707027, | |
| "grad_norm": 19.875337600708008, | |
| "learning_rate": 9.660180681028492e-06, | |
| "loss": 0.5896, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.231812066645969, | |
| "grad_norm": 24.210161209106445, | |
| "learning_rate": 9.656706045865186e-06, | |
| "loss": 0.6079, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.2421608196212357, | |
| "grad_norm": 21.59490394592285, | |
| "learning_rate": 9.653231410701877e-06, | |
| "loss": 0.6732, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2421608196212357, | |
| "eval_loss": 0.3701952397823334, | |
| "eval_runtime": 580.4786, | |
| "eval_samples_per_second": 1.873, | |
| "eval_steps_per_second": 0.937, | |
| "eval_wer": 0.2200689935064935, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2421608196212357, | |
| "step": 3000, | |
| "total_flos": 2.449860020011008e+19, | |
| "train_loss": 1.494960307121277, | |
| "train_runtime": 14532.737, | |
| "train_samples_per_second": 39.895, | |
| "train_steps_per_second": 4.985 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 72450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "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": 2.449860020011008e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |