| { | |
| "best_metric": 0.3816666007041931, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-swagen-combined-20hrs-model/checkpoint-3000", | |
| "epoch": 2.2291021671826625, | |
| "eval_steps": 200, | |
| "global_step": 3600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015479876160990712, | |
| "grad_norm": 138.2083740234375, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 10.5632, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.030959752321981424, | |
| "grad_norm": 90.87299346923828, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 7.9292, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04643962848297214, | |
| "grad_norm": 79.41549682617188, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 5.8787, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06191950464396285, | |
| "grad_norm": 68.67957305908203, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 4.1248, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07739938080495357, | |
| "grad_norm": 65.86703491210938, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 3.5039, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.09287925696594428, | |
| "grad_norm": 62.98320388793945, | |
| "learning_rate": 2.9e-06, | |
| "loss": 3.201, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10835913312693499, | |
| "grad_norm": 62.89395523071289, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 3.181, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1238390092879257, | |
| "grad_norm": 62.861602783203125, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 2.7478, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1238390092879257, | |
| "eval_loss": 0.797576904296875, | |
| "eval_runtime": 563.0008, | |
| "eval_samples_per_second": 1.931, | |
| "eval_steps_per_second": 0.966, | |
| "eval_wer": 0.4823457792207792, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1393188854489164, | |
| "grad_norm": 63.16032409667969, | |
| "learning_rate": 4.4e-06, | |
| "loss": 2.6638, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.15479876160990713, | |
| "grad_norm": 62.850440979003906, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 2.1734, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17027863777089783, | |
| "grad_norm": 52.81850051879883, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 2.3938, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.18575851393188855, | |
| "grad_norm": 44.25548553466797, | |
| "learning_rate": 5.9e-06, | |
| "loss": 2.2972, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.20123839009287925, | |
| "grad_norm": 68.06084442138672, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 2.02, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.21671826625386997, | |
| "grad_norm": 68.29243469238281, | |
| "learning_rate": 6.9e-06, | |
| "loss": 2.0561, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.23219814241486067, | |
| "grad_norm": 65.92505645751953, | |
| "learning_rate": 7.4e-06, | |
| "loss": 2.1382, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.2476780185758514, | |
| "grad_norm": 50.10117721557617, | |
| "learning_rate": 7.9e-06, | |
| "loss": 1.6879, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2476780185758514, | |
| "eval_loss": 0.6310324668884277, | |
| "eval_runtime": 559.0623, | |
| "eval_samples_per_second": 1.944, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 0.38950892857142855, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 56.55495071411133, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.9753, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.2786377708978328, | |
| "grad_norm": 65.00907135009766, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 1.8005, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 34.83129119873047, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.7464, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.30959752321981426, | |
| "grad_norm": 69.51799011230469, | |
| "learning_rate": 9.9e-06, | |
| "loss": 1.991, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32507739938080493, | |
| "grad_norm": 59.508968353271484, | |
| "learning_rate": 9.995828988529718e-06, | |
| "loss": 1.9345, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.34055727554179566, | |
| "grad_norm": 38.56643295288086, | |
| "learning_rate": 9.990615224191867e-06, | |
| "loss": 1.7101, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3560371517027864, | |
| "grad_norm": 61.955345153808594, | |
| "learning_rate": 9.985401459854015e-06, | |
| "loss": 1.8107, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3715170278637771, | |
| "grad_norm": 44.76349639892578, | |
| "learning_rate": 9.980187695516164e-06, | |
| "loss": 1.7394, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3715170278637771, | |
| "eval_loss": 0.5530644655227661, | |
| "eval_runtime": 582.5012, | |
| "eval_samples_per_second": 1.866, | |
| "eval_steps_per_second": 0.934, | |
| "eval_wer": 0.3317775974025974, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.38699690402476783, | |
| "grad_norm": 53.88811111450195, | |
| "learning_rate": 9.974973931178311e-06, | |
| "loss": 1.4966, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4024767801857585, | |
| "grad_norm": 36.39348602294922, | |
| "learning_rate": 9.96976016684046e-06, | |
| "loss": 1.6524, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4179566563467492, | |
| "grad_norm": 50.56556701660156, | |
| "learning_rate": 9.964546402502608e-06, | |
| "loss": 1.6449, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.43343653250773995, | |
| "grad_norm": 49.1718635559082, | |
| "learning_rate": 9.959332638164755e-06, | |
| "loss": 1.6947, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.44891640866873067, | |
| "grad_norm": 44.10185623168945, | |
| "learning_rate": 9.954118873826903e-06, | |
| "loss": 1.6989, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.46439628482972134, | |
| "grad_norm": 47.85836410522461, | |
| "learning_rate": 9.948905109489052e-06, | |
| "loss": 1.5363, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.47987616099071206, | |
| "grad_norm": 50.4083251953125, | |
| "learning_rate": 9.9436913451512e-06, | |
| "loss": 1.5924, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.4953560371517028, | |
| "grad_norm": 68.3183364868164, | |
| "learning_rate": 9.938477580813348e-06, | |
| "loss": 1.4248, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4953560371517028, | |
| "eval_loss": 0.5066032409667969, | |
| "eval_runtime": 568.3038, | |
| "eval_samples_per_second": 1.913, | |
| "eval_steps_per_second": 0.957, | |
| "eval_wer": 0.2980925324675325, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5108359133126935, | |
| "grad_norm": 39.81132888793945, | |
| "learning_rate": 9.933263816475496e-06, | |
| "loss": 1.5851, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 35.51476287841797, | |
| "learning_rate": 9.928050052137645e-06, | |
| "loss": 1.5689, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.541795665634675, | |
| "grad_norm": 43.64626693725586, | |
| "learning_rate": 9.922836287799792e-06, | |
| "loss": 1.2399, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.5572755417956656, | |
| "grad_norm": 46.96986770629883, | |
| "learning_rate": 9.91762252346194e-06, | |
| "loss": 1.4891, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5727554179566563, | |
| "grad_norm": 48.96132278442383, | |
| "learning_rate": 9.912408759124087e-06, | |
| "loss": 1.5456, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 46.15045166015625, | |
| "learning_rate": 9.907194994786236e-06, | |
| "loss": 1.4521, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.6037151702786377, | |
| "grad_norm": 41.92116928100586, | |
| "learning_rate": 9.901981230448384e-06, | |
| "loss": 1.4187, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "grad_norm": 34.402976989746094, | |
| "learning_rate": 9.896767466110533e-06, | |
| "loss": 1.2897, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "eval_loss": 0.4649743437767029, | |
| "eval_runtime": 564.2182, | |
| "eval_samples_per_second": 1.927, | |
| "eval_steps_per_second": 0.964, | |
| "eval_wer": 0.2688717532467532, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6346749226006192, | |
| "grad_norm": 31.557924270629883, | |
| "learning_rate": 9.89155370177268e-06, | |
| "loss": 1.4588, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.6501547987616099, | |
| "grad_norm": 46.03443908691406, | |
| "learning_rate": 9.88633993743483e-06, | |
| "loss": 1.3974, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6656346749226006, | |
| "grad_norm": 56.12520980834961, | |
| "learning_rate": 9.881126173096977e-06, | |
| "loss": 1.4078, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.6811145510835913, | |
| "grad_norm": 49.19106674194336, | |
| "learning_rate": 9.875912408759124e-06, | |
| "loss": 1.2417, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6965944272445821, | |
| "grad_norm": 51.47815704345703, | |
| "learning_rate": 9.870698644421273e-06, | |
| "loss": 1.4088, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.7120743034055728, | |
| "grad_norm": 36.05540084838867, | |
| "learning_rate": 9.865484880083421e-06, | |
| "loss": 1.4249, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7275541795665634, | |
| "grad_norm": 50.757347106933594, | |
| "learning_rate": 9.860271115745568e-06, | |
| "loss": 1.3757, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.7430340557275542, | |
| "grad_norm": 43.31879425048828, | |
| "learning_rate": 9.855057351407717e-06, | |
| "loss": 1.357, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7430340557275542, | |
| "eval_loss": 0.4457937777042389, | |
| "eval_runtime": 572.8759, | |
| "eval_samples_per_second": 1.897, | |
| "eval_steps_per_second": 0.95, | |
| "eval_wer": 0.27384334415584416, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7585139318885449, | |
| "grad_norm": 52.25995635986328, | |
| "learning_rate": 9.849843587069865e-06, | |
| "loss": 1.0826, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.7739938080495357, | |
| "grad_norm": 43.19371032714844, | |
| "learning_rate": 9.844629822732014e-06, | |
| "loss": 1.4315, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 28.991748809814453, | |
| "learning_rate": 9.839416058394161e-06, | |
| "loss": 1.2291, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.804953560371517, | |
| "grad_norm": 38.30876922607422, | |
| "learning_rate": 9.83420229405631e-06, | |
| "loss": 1.3047, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8204334365325078, | |
| "grad_norm": 44.77728271484375, | |
| "learning_rate": 9.828988529718458e-06, | |
| "loss": 1.4421, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.8359133126934984, | |
| "grad_norm": 64.83729553222656, | |
| "learning_rate": 9.823774765380605e-06, | |
| "loss": 1.3138, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.8513931888544891, | |
| "grad_norm": 36.1328010559082, | |
| "learning_rate": 9.818561001042753e-06, | |
| "loss": 1.2132, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.8668730650154799, | |
| "grad_norm": 40.22114562988281, | |
| "learning_rate": 9.813347236704902e-06, | |
| "loss": 1.3443, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8668730650154799, | |
| "eval_loss": 0.43931862711906433, | |
| "eval_runtime": 580.0591, | |
| "eval_samples_per_second": 1.874, | |
| "eval_steps_per_second": 0.938, | |
| "eval_wer": 0.270900974025974, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 27.64898681640625, | |
| "learning_rate": 9.80813347236705e-06, | |
| "loss": 1.1846, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.8978328173374613, | |
| "grad_norm": 19.260305404663086, | |
| "learning_rate": 9.802919708029199e-06, | |
| "loss": 1.1317, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.913312693498452, | |
| "grad_norm": 30.223257064819336, | |
| "learning_rate": 9.797705943691346e-06, | |
| "loss": 1.3269, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.9287925696594427, | |
| "grad_norm": 35.81090545654297, | |
| "learning_rate": 9.792492179353495e-06, | |
| "loss": 1.0545, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9442724458204335, | |
| "grad_norm": 32.93012237548828, | |
| "learning_rate": 9.787278415015643e-06, | |
| "loss": 1.1677, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.9597523219814241, | |
| "grad_norm": 46.6026496887207, | |
| "learning_rate": 9.78206465067779e-06, | |
| "loss": 1.2468, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.9752321981424149, | |
| "grad_norm": 29.324565887451172, | |
| "learning_rate": 9.776850886339937e-06, | |
| "loss": 1.081, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.9907120743034056, | |
| "grad_norm": 41.78209686279297, | |
| "learning_rate": 9.771637122002087e-06, | |
| "loss": 1.2284, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9907120743034056, | |
| "eval_loss": 0.41507846117019653, | |
| "eval_runtime": 580.4169, | |
| "eval_samples_per_second": 1.873, | |
| "eval_steps_per_second": 0.937, | |
| "eval_wer": 0.27445211038961037, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.0061919504643964, | |
| "grad_norm": 25.455608367919922, | |
| "learning_rate": 9.766423357664234e-06, | |
| "loss": 0.9931, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.021671826625387, | |
| "grad_norm": 22.430744171142578, | |
| "learning_rate": 9.761209593326383e-06, | |
| "loss": 0.6197, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.0371517027863777, | |
| "grad_norm": 21.164098739624023, | |
| "learning_rate": 9.75599582898853e-06, | |
| "loss": 0.6202, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 23.99432373046875, | |
| "learning_rate": 9.75078206465068e-06, | |
| "loss": 0.7075, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.068111455108359, | |
| "grad_norm": 32.372291564941406, | |
| "learning_rate": 9.745568300312827e-06, | |
| "loss": 0.6919, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.08359133126935, | |
| "grad_norm": 29.520769119262695, | |
| "learning_rate": 9.740354535974975e-06, | |
| "loss": 0.5938, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.0990712074303406, | |
| "grad_norm": 34.81272888183594, | |
| "learning_rate": 9.735140771637122e-06, | |
| "loss": 0.7783, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.1145510835913313, | |
| "grad_norm": 26.11049461364746, | |
| "learning_rate": 9.729927007299271e-06, | |
| "loss": 0.6748, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.1145510835913313, | |
| "eval_loss": 0.41134530305862427, | |
| "eval_runtime": 569.9787, | |
| "eval_samples_per_second": 1.907, | |
| "eval_steps_per_second": 0.954, | |
| "eval_wer": 0.2520292207792208, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.130030959752322, | |
| "grad_norm": 31.540014266967773, | |
| "learning_rate": 9.724713242961419e-06, | |
| "loss": 0.5986, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.1455108359133126, | |
| "grad_norm": 29.100893020629883, | |
| "learning_rate": 9.719499478623568e-06, | |
| "loss": 0.6972, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.1609907120743035, | |
| "grad_norm": 24.08401870727539, | |
| "learning_rate": 9.714285714285715e-06, | |
| "loss": 0.7194, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 17.05337142944336, | |
| "learning_rate": 9.709071949947864e-06, | |
| "loss": 0.7229, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.1919504643962848, | |
| "grad_norm": 34.69645309448242, | |
| "learning_rate": 9.703858185610012e-06, | |
| "loss": 0.7934, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.2074303405572755, | |
| "grad_norm": 23.91748046875, | |
| "learning_rate": 9.69864442127216e-06, | |
| "loss": 0.6275, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.2229102167182662, | |
| "grad_norm": 29.427534103393555, | |
| "learning_rate": 9.693430656934307e-06, | |
| "loss": 0.6677, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.238390092879257, | |
| "grad_norm": 25.043569564819336, | |
| "learning_rate": 9.688216892596456e-06, | |
| "loss": 0.7242, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.238390092879257, | |
| "eval_loss": 0.4232771694660187, | |
| "eval_runtime": 576.1378, | |
| "eval_samples_per_second": 1.887, | |
| "eval_steps_per_second": 0.944, | |
| "eval_wer": 0.26704545454545453, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.2538699690402477, | |
| "grad_norm": 37.21471405029297, | |
| "learning_rate": 9.683003128258603e-06, | |
| "loss": 0.6917, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.2693498452012384, | |
| "grad_norm": 19.723859786987305, | |
| "learning_rate": 9.677789363920752e-06, | |
| "loss": 0.6583, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.284829721362229, | |
| "grad_norm": 25.643213272094727, | |
| "learning_rate": 9.6725755995829e-06, | |
| "loss": 0.6923, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.3003095975232197, | |
| "grad_norm": 27.70577621459961, | |
| "learning_rate": 9.667361835245049e-06, | |
| "loss": 0.5808, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 23.764432907104492, | |
| "learning_rate": 9.662148070907196e-06, | |
| "loss": 0.5764, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.3312693498452013, | |
| "grad_norm": 32.69343185424805, | |
| "learning_rate": 9.656934306569344e-06, | |
| "loss": 0.6372, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.346749226006192, | |
| "grad_norm": 20.59185791015625, | |
| "learning_rate": 9.651720542231491e-06, | |
| "loss": 0.5799, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.3622291021671826, | |
| "grad_norm": 20.03351593017578, | |
| "learning_rate": 9.64650677789364e-06, | |
| "loss": 0.6472, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.3622291021671826, | |
| "eval_loss": 0.42063385248184204, | |
| "eval_runtime": 574.5155, | |
| "eval_samples_per_second": 1.892, | |
| "eval_steps_per_second": 0.947, | |
| "eval_wer": 0.25923295454545453, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.3777089783281733, | |
| "grad_norm": 23.296613693237305, | |
| "learning_rate": 9.641293013555788e-06, | |
| "loss": 0.6845, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.3931888544891642, | |
| "grad_norm": 30.750635147094727, | |
| "learning_rate": 9.636079249217935e-06, | |
| "loss": 0.6608, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.4086687306501549, | |
| "grad_norm": 25.292583465576172, | |
| "learning_rate": 9.630865484880084e-06, | |
| "loss": 0.7074, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.4241486068111455, | |
| "grad_norm": 20.848609924316406, | |
| "learning_rate": 9.625651720542232e-06, | |
| "loss": 0.5744, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.4396284829721362, | |
| "grad_norm": 24.94603729248047, | |
| "learning_rate": 9.620437956204381e-06, | |
| "loss": 0.5716, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.4551083591331269, | |
| "grad_norm": 35.606422424316406, | |
| "learning_rate": 9.615224191866528e-06, | |
| "loss": 0.6942, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 33.53971481323242, | |
| "learning_rate": 9.610010427528676e-06, | |
| "loss": 0.6933, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.4860681114551084, | |
| "grad_norm": 26.1285457611084, | |
| "learning_rate": 9.604796663190825e-06, | |
| "loss": 0.6917, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.4860681114551084, | |
| "eval_loss": 0.39901289343833923, | |
| "eval_runtime": 584.9316, | |
| "eval_samples_per_second": 1.858, | |
| "eval_steps_per_second": 0.93, | |
| "eval_wer": 0.27901785714285715, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.501547987616099, | |
| "grad_norm": 11.766839027404785, | |
| "learning_rate": 9.599582898852972e-06, | |
| "loss": 0.5896, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.5170278637770898, | |
| "grad_norm": 30.828351974487305, | |
| "learning_rate": 9.59436913451512e-06, | |
| "loss": 0.7007, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.5325077399380804, | |
| "grad_norm": 21.049291610717773, | |
| "learning_rate": 9.589155370177269e-06, | |
| "loss": 0.855, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "grad_norm": 17.50288963317871, | |
| "learning_rate": 9.583941605839416e-06, | |
| "loss": 0.5517, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.5634674922600618, | |
| "grad_norm": 40.06394577026367, | |
| "learning_rate": 9.578727841501565e-06, | |
| "loss": 0.6987, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 26.996692657470703, | |
| "learning_rate": 9.573514077163713e-06, | |
| "loss": 0.6327, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.5944272445820433, | |
| "grad_norm": 15.66840934753418, | |
| "learning_rate": 9.56830031282586e-06, | |
| "loss": 0.6396, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.609907120743034, | |
| "grad_norm": 21.837093353271484, | |
| "learning_rate": 9.56308654848801e-06, | |
| "loss": 0.7172, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.609907120743034, | |
| "eval_loss": 0.3971739113330841, | |
| "eval_runtime": 571.0488, | |
| "eval_samples_per_second": 1.904, | |
| "eval_steps_per_second": 0.953, | |
| "eval_wer": 0.3005275974025974, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.6253869969040249, | |
| "grad_norm": 22.674137115478516, | |
| "learning_rate": 9.557872784150157e-06, | |
| "loss": 0.5473, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.6408668730650153, | |
| "grad_norm": 30.52071762084961, | |
| "learning_rate": 9.552659019812304e-06, | |
| "loss": 0.594, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.6563467492260062, | |
| "grad_norm": 4.544656276702881, | |
| "learning_rate": 9.547445255474453e-06, | |
| "loss": 0.6123, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.671826625386997, | |
| "grad_norm": 31.003204345703125, | |
| "learning_rate": 9.542231491136601e-06, | |
| "loss": 0.6435, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.6873065015479876, | |
| "grad_norm": 24.042932510375977, | |
| "learning_rate": 9.53701772679875e-06, | |
| "loss": 0.6526, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.7027863777089784, | |
| "grad_norm": 25.72020149230957, | |
| "learning_rate": 9.531803962460897e-06, | |
| "loss": 0.6013, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.718266253869969, | |
| "grad_norm": 32.10405349731445, | |
| "learning_rate": 9.526590198123047e-06, | |
| "loss": 0.7614, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.7337461300309598, | |
| "grad_norm": 28.47552490234375, | |
| "learning_rate": 9.521376433785194e-06, | |
| "loss": 0.6105, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.7337461300309598, | |
| "eval_loss": 0.39262855052948, | |
| "eval_runtime": 546.1891, | |
| "eval_samples_per_second": 1.99, | |
| "eval_steps_per_second": 0.996, | |
| "eval_wer": 0.23133116883116883, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.7492260061919505, | |
| "grad_norm": 29.18010139465332, | |
| "learning_rate": 9.516162669447341e-06, | |
| "loss": 0.7792, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 23.199079513549805, | |
| "learning_rate": 9.510948905109489e-06, | |
| "loss": 0.5944, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.780185758513932, | |
| "grad_norm": 19.647172927856445, | |
| "learning_rate": 9.505735140771638e-06, | |
| "loss": 0.7278, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.7956656346749225, | |
| "grad_norm": 26.793537139892578, | |
| "learning_rate": 9.500521376433785e-06, | |
| "loss": 0.6646, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.8111455108359134, | |
| "grad_norm": 30.784910202026367, | |
| "learning_rate": 9.495307612095935e-06, | |
| "loss": 0.7125, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.826625386996904, | |
| "grad_norm": 27.740474700927734, | |
| "learning_rate": 9.490093847758082e-06, | |
| "loss": 0.6095, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 58.08521270751953, | |
| "learning_rate": 9.484880083420231e-06, | |
| "loss": 0.657, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.8575851393188856, | |
| "grad_norm": 34.06402587890625, | |
| "learning_rate": 9.479666319082379e-06, | |
| "loss": 0.7442, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.8575851393188856, | |
| "eval_loss": 0.3816666007041931, | |
| "eval_runtime": 538.3038, | |
| "eval_samples_per_second": 2.019, | |
| "eval_steps_per_second": 1.011, | |
| "eval_wer": 0.22899756493506493, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.873065015479876, | |
| "grad_norm": 20.675968170166016, | |
| "learning_rate": 9.474452554744526e-06, | |
| "loss": 0.7677, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 1.888544891640867, | |
| "grad_norm": 30.081329345703125, | |
| "learning_rate": 9.469238790406673e-06, | |
| "loss": 0.592, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.9040247678018576, | |
| "grad_norm": 26.64177703857422, | |
| "learning_rate": 9.464025026068823e-06, | |
| "loss": 0.6696, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 1.9195046439628483, | |
| "grad_norm": 28.337804794311523, | |
| "learning_rate": 9.45881126173097e-06, | |
| "loss": 0.6002, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.9349845201238391, | |
| "grad_norm": 15.685124397277832, | |
| "learning_rate": 9.453597497393119e-06, | |
| "loss": 0.6714, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 1.9504643962848296, | |
| "grad_norm": 28.92962646484375, | |
| "learning_rate": 9.448383733055267e-06, | |
| "loss": 0.6535, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.9659442724458205, | |
| "grad_norm": 29.42474365234375, | |
| "learning_rate": 9.443169968717416e-06, | |
| "loss": 0.5845, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 1.9814241486068112, | |
| "grad_norm": 17.18292999267578, | |
| "learning_rate": 9.437956204379563e-06, | |
| "loss": 0.7074, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.9814241486068112, | |
| "eval_loss": 0.3848791718482971, | |
| "eval_runtime": 535.7133, | |
| "eval_samples_per_second": 2.029, | |
| "eval_steps_per_second": 1.015, | |
| "eval_wer": 0.2262581168831169, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.9969040247678018, | |
| "grad_norm": 30.033981323242188, | |
| "learning_rate": 9.43274244004171e-06, | |
| "loss": 0.7561, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 2.0123839009287927, | |
| "grad_norm": 11.292386054992676, | |
| "learning_rate": 9.427528675703858e-06, | |
| "loss": 0.324, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.027863777089783, | |
| "grad_norm": 14.040888786315918, | |
| "learning_rate": 9.422314911366007e-06, | |
| "loss": 0.2739, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 2.043343653250774, | |
| "grad_norm": 10.198466300964355, | |
| "learning_rate": 9.417101147028155e-06, | |
| "loss": 0.2792, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.0588235294117645, | |
| "grad_norm": 11.23904800415039, | |
| "learning_rate": 9.411887382690304e-06, | |
| "loss": 0.2426, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 2.0743034055727554, | |
| "grad_norm": 15.241331100463867, | |
| "learning_rate": 9.406673618352451e-06, | |
| "loss": 0.2859, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.0897832817337463, | |
| "grad_norm": 12.682251930236816, | |
| "learning_rate": 9.4014598540146e-06, | |
| "loss": 0.2581, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 8.495072364807129, | |
| "learning_rate": 9.396246089676748e-06, | |
| "loss": 0.2659, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "eval_loss": 0.400187611579895, | |
| "eval_runtime": 545.9468, | |
| "eval_samples_per_second": 1.991, | |
| "eval_steps_per_second": 0.996, | |
| "eval_wer": 0.24857954545454544, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.1207430340557276, | |
| "grad_norm": 10.508811950683594, | |
| "learning_rate": 9.391032325338895e-06, | |
| "loss": 0.2665, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 2.136222910216718, | |
| "grad_norm": 27.000606536865234, | |
| "learning_rate": 9.385818561001043e-06, | |
| "loss": 0.2768, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.151702786377709, | |
| "grad_norm": 25.08758544921875, | |
| "learning_rate": 9.380604796663192e-06, | |
| "loss": 0.2693, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 2.1671826625387, | |
| "grad_norm": 16.024877548217773, | |
| "learning_rate": 9.375391032325339e-06, | |
| "loss": 0.2919, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.1826625386996903, | |
| "grad_norm": 11.162302017211914, | |
| "learning_rate": 9.370177267987488e-06, | |
| "loss": 0.2464, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 2.198142414860681, | |
| "grad_norm": 14.112415313720703, | |
| "learning_rate": 9.364963503649636e-06, | |
| "loss": 0.3192, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 2.2136222910216716, | |
| "grad_norm": 14.262118339538574, | |
| "learning_rate": 9.359749739311785e-06, | |
| "loss": 0.2935, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 2.2291021671826625, | |
| "grad_norm": 14.042037963867188, | |
| "learning_rate": 9.354535974973932e-06, | |
| "loss": 0.3132, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.2291021671826625, | |
| "eval_loss": 0.39583250880241394, | |
| "eval_runtime": 534.4305, | |
| "eval_samples_per_second": 2.034, | |
| "eval_steps_per_second": 1.018, | |
| "eval_wer": 0.22280844155844157, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.2291021671826625, | |
| "step": 3600, | |
| "total_flos": 2.939138012676096e+19, | |
| "train_loss": 1.2231303668022155, | |
| "train_runtime": 17395.8198, | |
| "train_samples_per_second": 22.279, | |
| "train_steps_per_second": 2.785 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 48450, | |
| "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": 2.939138012676096e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |