diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -1,1472 +1,5807 @@ { "best_metric": null, "best_model_checkpoint": null, - "epoch": 19.23076923076923, + "epoch": 16.913319238900634, "eval_steps": 1000, - "global_step": 5000, + "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { - "epoch": 0.09615384615384616, - "grad_norm": 14.387816429138184, + "epoch": 0.021141649048625793, + "grad_norm": 18.116064071655273, "learning_rate": 4.6000000000000004e-07, - "loss": 1.0674, + "loss": 1.1425, "step": 25 }, { - "epoch": 0.19230769230769232, - "grad_norm": 11.297602653503418, + "epoch": 0.042283298097251586, + "grad_norm": 12.63290023803711, "learning_rate": 9.600000000000001e-07, - "loss": 0.9654, + "loss": 0.9953, "step": 50 }, { - "epoch": 0.28846153846153844, - "grad_norm": 8.031381607055664, + "epoch": 0.06342494714587738, + "grad_norm": 8.746077537536621, "learning_rate": 1.46e-06, - "loss": 0.7994, + "loss": 0.7269, "step": 75 }, { - "epoch": 0.38461538461538464, - "grad_norm": 6.947944641113281, + "epoch": 0.08456659619450317, + "grad_norm": 6.509521484375, "learning_rate": 1.9600000000000003e-06, - "loss": 0.5243, + "loss": 0.4322, "step": 100 }, { - "epoch": 0.4807692307692308, - "grad_norm": 5.577055931091309, + "epoch": 0.10570824524312897, + "grad_norm": 6.591165542602539, "learning_rate": 2.46e-06, - "loss": 0.4378, + "loss": 0.3318, "step": 125 }, { - "epoch": 0.5769230769230769, - "grad_norm": 5.237943649291992, + "epoch": 0.12684989429175475, + "grad_norm": 5.082007884979248, "learning_rate": 2.96e-06, - "loss": 0.39, + "loss": 0.3381, "step": 150 }, { - "epoch": 0.6730769230769231, - "grad_norm": 5.476370811462402, + "epoch": 0.14799154334038056, + "grad_norm": 6.309211254119873, "learning_rate": 3.46e-06, - "loss": 0.3755, + "loss": 0.314, "step": 175 }, { - "epoch": 0.7692307692307693, - "grad_norm": 6.1683244705200195, + "epoch": 0.16913319238900634, + "grad_norm": 5.363051414489746, "learning_rate": 3.96e-06, - "loss": 0.3851, + "loss": 0.3026, "step": 200 }, { - "epoch": 0.8653846153846154, - "grad_norm": 5.542613506317139, + "epoch": 0.19027484143763213, + "grad_norm": 6.52091121673584, "learning_rate": 4.4600000000000005e-06, - "loss": 0.3461, + "loss": 0.3055, "step": 225 }, { - "epoch": 0.9615384615384616, - "grad_norm": 7.256478786468506, + "epoch": 0.21141649048625794, + "grad_norm": 4.084770679473877, "learning_rate": 4.960000000000001e-06, - "loss": 0.3369, + "loss": 0.2737, "step": 250 }, { - "epoch": 1.0576923076923077, - "grad_norm": 5.407237529754639, + "epoch": 0.23255813953488372, + "grad_norm": 5.420108795166016, "learning_rate": 5.460000000000001e-06, - "loss": 0.2975, + "loss": 0.2853, "step": 275 }, { - "epoch": 1.1538461538461537, - "grad_norm": 5.3987579345703125, + "epoch": 0.2536997885835095, + "grad_norm": 4.882087230682373, "learning_rate": 5.9600000000000005e-06, - "loss": 0.2739, + "loss": 0.2598, "step": 300 }, { - "epoch": 1.25, - "grad_norm": 4.13232946395874, + "epoch": 0.2748414376321353, + "grad_norm": 5.636622428894043, "learning_rate": 6.460000000000001e-06, - "loss": 0.2519, + "loss": 0.2711, "step": 325 }, { - "epoch": 1.3461538461538463, - "grad_norm": 4.822619438171387, + "epoch": 0.2959830866807611, + "grad_norm": 6.7103190422058105, "learning_rate": 6.96e-06, - "loss": 0.2663, + "loss": 0.238, "step": 350 }, { - "epoch": 1.4423076923076923, - "grad_norm": 4.323586463928223, + "epoch": 0.3171247357293869, + "grad_norm": 5.882842540740967, "learning_rate": 7.4600000000000006e-06, - "loss": 0.2482, + "loss": 0.2454, "step": 375 }, { - "epoch": 1.5384615384615383, - "grad_norm": 4.278296947479248, + "epoch": 0.3382663847780127, + "grad_norm": 5.11025857925415, "learning_rate": 7.960000000000002e-06, - "loss": 0.2525, + "loss": 0.2611, "step": 400 }, { - "epoch": 1.6346153846153846, - "grad_norm": 5.396093368530273, + "epoch": 0.3594080338266385, + "grad_norm": 5.972140312194824, "learning_rate": 8.46e-06, - "loss": 0.2575, + "loss": 0.2546, "step": 425 }, { - "epoch": 1.7307692307692308, - "grad_norm": 4.330433368682861, + "epoch": 0.38054968287526425, + "grad_norm": 4.569517135620117, "learning_rate": 8.96e-06, - "loss": 0.2502, + "loss": 0.2475, "step": 450 }, { - "epoch": 1.8269230769230769, - "grad_norm": 4.134779930114746, + "epoch": 0.40169133192389006, + "grad_norm": 5.826624393463135, "learning_rate": 9.460000000000001e-06, - "loss": 0.2615, + "loss": 0.246, "step": 475 }, { - "epoch": 1.9230769230769231, - "grad_norm": 4.702469348907471, + "epoch": 0.42283298097251587, + "grad_norm": 6.217140197753906, "learning_rate": 9.960000000000001e-06, - "loss": 0.2542, + "loss": 0.2592, "step": 500 }, { - "epoch": 2.019230769230769, - "grad_norm": 2.5487704277038574, - "learning_rate": 9.94888888888889e-06, - "loss": 0.2502, + "epoch": 0.4439746300211416, + "grad_norm": 5.610385417938232, + "learning_rate": 9.988205128205129e-06, + "loss": 0.2435, "step": 525 }, { - "epoch": 2.1153846153846154, - "grad_norm": 3.5824356079101562, - "learning_rate": 9.893333333333334e-06, - "loss": 0.1373, + "epoch": 0.46511627906976744, + "grad_norm": 4.2026567459106445, + "learning_rate": 9.975384615384616e-06, + "loss": 0.2227, "step": 550 }, { - "epoch": 2.2115384615384617, - "grad_norm": 3.7268402576446533, - "learning_rate": 9.837777777777778e-06, - "loss": 0.1367, + "epoch": 0.48625792811839325, + "grad_norm": 5.413006782531738, + "learning_rate": 9.962564102564102e-06, + "loss": 0.2344, "step": 575 }, { - "epoch": 2.3076923076923075, - "grad_norm": 3.243474245071411, - "learning_rate": 9.782222222222222e-06, - "loss": 0.1298, + "epoch": 0.507399577167019, + "grad_norm": 5.172091960906982, + "learning_rate": 9.94974358974359e-06, + "loss": 0.207, "step": 600 }, { - "epoch": 2.4038461538461537, - "grad_norm": 3.1370737552642822, - "learning_rate": 9.726666666666668e-06, - "loss": 0.135, + "epoch": 0.5285412262156448, + "grad_norm": 3.3823935985565186, + "learning_rate": 9.936923076923078e-06, + "loss": 0.2192, "step": 625 }, { - "epoch": 2.5, - "grad_norm": 2.788139581680298, - "learning_rate": 9.671111111111112e-06, - "loss": 0.1239, + "epoch": 0.5496828752642706, + "grad_norm": 6.060272693634033, + "learning_rate": 9.924102564102565e-06, + "loss": 0.2052, "step": 650 }, { - "epoch": 2.5961538461538463, - "grad_norm": 3.4849274158477783, - "learning_rate": 9.615555555555558e-06, - "loss": 0.1226, + "epoch": 0.5708245243128964, + "grad_norm": 5.090622901916504, + "learning_rate": 9.911282051282051e-06, + "loss": 0.2074, "step": 675 }, { - "epoch": 2.6923076923076925, - "grad_norm": 3.447479009628296, - "learning_rate": 9.56e-06, - "loss": 0.1281, + "epoch": 0.5919661733615222, + "grad_norm": 4.698369979858398, + "learning_rate": 9.898461538461538e-06, + "loss": 0.2305, "step": 700 }, { - "epoch": 2.7884615384615383, - "grad_norm": 3.737367868423462, - "learning_rate": 9.504444444444446e-06, - "loss": 0.121, + "epoch": 0.6131078224101479, + "grad_norm": 5.381127834320068, + "learning_rate": 9.885641025641027e-06, + "loss": 0.224, "step": 725 }, { - "epoch": 2.8846153846153846, - "grad_norm": 3.7365550994873047, - "learning_rate": 9.44888888888889e-06, - "loss": 0.1284, + "epoch": 0.6342494714587738, + "grad_norm": 4.093654155731201, + "learning_rate": 9.872820512820514e-06, + "loss": 0.189, "step": 750 }, { - "epoch": 2.980769230769231, - "grad_norm": 4.205287933349609, - "learning_rate": 9.393333333333334e-06, - "loss": 0.1393, + "epoch": 0.6553911205073996, + "grad_norm": 3.974806070327759, + "learning_rate": 9.86e-06, + "loss": 0.2005, "step": 775 }, { - "epoch": 3.076923076923077, - "grad_norm": 2.6460373401641846, - "learning_rate": 9.33777777777778e-06, - "loss": 0.0834, + "epoch": 0.6765327695560254, + "grad_norm": 4.721721172332764, + "learning_rate": 9.847179487179487e-06, + "loss": 0.2163, "step": 800 }, { - "epoch": 3.173076923076923, - "grad_norm": 2.1553828716278076, - "learning_rate": 9.282222222222222e-06, - "loss": 0.0541, + "epoch": 0.6976744186046512, + "grad_norm": 5.002155780792236, + "learning_rate": 9.834358974358976e-06, + "loss": 0.1996, "step": 825 }, { - "epoch": 3.269230769230769, - "grad_norm": 1.773935079574585, - "learning_rate": 9.226666666666668e-06, - "loss": 0.0559, + "epoch": 0.718816067653277, + "grad_norm": 5.590403079986572, + "learning_rate": 9.821538461538463e-06, + "loss": 0.2142, "step": 850 }, { - "epoch": 3.3653846153846154, - "grad_norm": 2.047421455383301, - "learning_rate": 9.171111111111112e-06, - "loss": 0.0618, + "epoch": 0.7399577167019028, + "grad_norm": 4.794340133666992, + "learning_rate": 9.80871794871795e-06, + "loss": 0.2069, "step": 875 }, { - "epoch": 3.4615384615384617, - "grad_norm": 2.5151689052581787, - "learning_rate": 9.115555555555556e-06, - "loss": 0.0604, + "epoch": 0.7610993657505285, + "grad_norm": 3.6653997898101807, + "learning_rate": 9.795897435897436e-06, + "loss": 0.1983, "step": 900 }, { - "epoch": 3.5576923076923075, - "grad_norm": 1.7141085863113403, - "learning_rate": 9.060000000000001e-06, - "loss": 0.0567, + "epoch": 0.7822410147991543, + "grad_norm": 4.958125114440918, + "learning_rate": 9.783076923076925e-06, + "loss": 0.213, "step": 925 }, { - "epoch": 3.6538461538461537, - "grad_norm": 2.3757855892181396, - "learning_rate": 9.004444444444445e-06, - "loss": 0.0567, + "epoch": 0.8033826638477801, + "grad_norm": 4.403783798217773, + "learning_rate": 9.770256410256412e-06, + "loss": 0.2215, "step": 950 }, { - "epoch": 3.75, - "grad_norm": 2.048637866973877, - "learning_rate": 8.94888888888889e-06, - "loss": 0.0595, + "epoch": 0.8245243128964059, + "grad_norm": 5.16312313079834, + "learning_rate": 9.757435897435898e-06, + "loss": 0.2096, "step": 975 }, { - "epoch": 3.8461538461538463, - "grad_norm": 2.78825306892395, - "learning_rate": 8.893333333333333e-06, - "loss": 0.0547, + "epoch": 0.8456659619450317, + "grad_norm": 4.566516876220703, + "learning_rate": 9.744615384615385e-06, + "loss": 0.1875, "step": 1000 }, { - "epoch": 3.8461538461538463, - "eval_loss": 0.2570088803768158, - "eval_runtime": 418.585, - "eval_samples_per_second": 8.698, - "eval_steps_per_second": 0.545, - "eval_wer": 0.16873172769038006, + "epoch": 0.8456659619450317, + "eval_loss": 0.13999927043914795, + "eval_runtime": 419.9778, + "eval_samples_per_second": 8.67, + "eval_steps_per_second": 0.543, + "eval_wer": 0.10993549584667502, "step": 1000 }, { - "epoch": 3.9423076923076925, - "grad_norm": 2.7837064266204834, - "learning_rate": 8.83777777777778e-06, - "loss": 0.0566, + "epoch": 0.8668076109936576, + "grad_norm": 4.138431072235107, + "learning_rate": 9.731794871794872e-06, + "loss": 0.2002, "step": 1025 }, { - "epoch": 4.038461538461538, - "grad_norm": 1.4186588525772095, - "learning_rate": 8.782222222222223e-06, - "loss": 0.0472, + "epoch": 0.8879492600422833, + "grad_norm": 5.494874477386475, + "learning_rate": 9.71897435897436e-06, + "loss": 0.1716, "step": 1050 }, { - "epoch": 4.134615384615385, - "grad_norm": 1.4111158847808838, - "learning_rate": 8.726666666666667e-06, - "loss": 0.026, + "epoch": 0.9090909090909091, + "grad_norm": 3.6946444511413574, + "learning_rate": 9.706153846153847e-06, + "loss": 0.1722, "step": 1075 }, { - "epoch": 4.230769230769231, - "grad_norm": 0.8355940580368042, - "learning_rate": 8.671111111111113e-06, - "loss": 0.025, + "epoch": 0.9302325581395349, + "grad_norm": 4.946837425231934, + "learning_rate": 9.693333333333334e-06, + "loss": 0.1735, "step": 1100 }, { - "epoch": 4.326923076923077, - "grad_norm": 2.8829832077026367, - "learning_rate": 8.615555555555555e-06, - "loss": 0.0281, + "epoch": 0.9513742071881607, + "grad_norm": 4.48101806640625, + "learning_rate": 9.680512820512821e-06, + "loss": 0.1836, "step": 1125 }, { - "epoch": 4.423076923076923, - "grad_norm": 1.8035304546356201, - "learning_rate": 8.560000000000001e-06, - "loss": 0.0275, + "epoch": 0.9725158562367865, + "grad_norm": 4.336252689361572, + "learning_rate": 9.667692307692308e-06, + "loss": 0.1827, "step": 1150 }, { - "epoch": 4.519230769230769, - "grad_norm": 2.558990240097046, - "learning_rate": 8.504444444444445e-06, - "loss": 0.0284, + "epoch": 0.9936575052854123, + "grad_norm": 5.601895332336426, + "learning_rate": 9.654871794871795e-06, + "loss": 0.1775, "step": 1175 }, { - "epoch": 4.615384615384615, - "grad_norm": 3.0063929557800293, - "learning_rate": 8.448888888888889e-06, - "loss": 0.0262, + "epoch": 1.014799154334038, + "grad_norm": 2.9764301776885986, + "learning_rate": 9.642051282051282e-06, + "loss": 0.136, "step": 1200 }, { - "epoch": 4.711538461538462, - "grad_norm": 1.181315541267395, - "learning_rate": 8.393333333333335e-06, - "loss": 0.0224, + "epoch": 1.0359408033826638, + "grad_norm": 3.8662917613983154, + "learning_rate": 9.62923076923077e-06, + "loss": 0.103, "step": 1225 }, { - "epoch": 4.8076923076923075, - "grad_norm": 1.4534687995910645, - "learning_rate": 8.337777777777777e-06, - "loss": 0.026, + "epoch": 1.0570824524312896, + "grad_norm": 2.8110098838806152, + "learning_rate": 9.616410256410257e-06, + "loss": 0.0959, "step": 1250 }, { - "epoch": 4.903846153846154, - "grad_norm": 1.941493034362793, - "learning_rate": 8.282222222222223e-06, - "loss": 0.0276, + "epoch": 1.0782241014799154, + "grad_norm": 3.119048833847046, + "learning_rate": 9.603589743589744e-06, + "loss": 0.1086, "step": 1275 }, { - "epoch": 5.0, - "grad_norm": 1.048115849494934, - "learning_rate": 8.226666666666667e-06, - "loss": 0.0278, + "epoch": 1.0993657505285412, + "grad_norm": 4.480291366577148, + "learning_rate": 9.59076923076923e-06, + "loss": 0.0961, "step": 1300 }, { - "epoch": 5.096153846153846, - "grad_norm": 1.3306798934936523, - "learning_rate": 8.171111111111113e-06, - "loss": 0.0136, + "epoch": 1.120507399577167, + "grad_norm": 2.7335333824157715, + "learning_rate": 9.577948717948719e-06, + "loss": 0.0915, "step": 1325 }, { - "epoch": 5.1923076923076925, - "grad_norm": 1.5832746028900146, - "learning_rate": 8.115555555555557e-06, - "loss": 0.0128, + "epoch": 1.1416490486257929, + "grad_norm": 3.7086055278778076, + "learning_rate": 9.565128205128206e-06, + "loss": 0.1077, "step": 1350 }, { - "epoch": 5.288461538461538, - "grad_norm": 1.175854206085205, - "learning_rate": 8.06e-06, - "loss": 0.0115, + "epoch": 1.1627906976744187, + "grad_norm": 3.4772427082061768, + "learning_rate": 9.552307692307693e-06, + "loss": 0.0892, "step": 1375 }, { - "epoch": 5.384615384615385, - "grad_norm": 1.0961897373199463, - "learning_rate": 8.004444444444445e-06, - "loss": 0.0128, + "epoch": 1.1839323467230445, + "grad_norm": 2.8368217945098877, + "learning_rate": 9.53948717948718e-06, + "loss": 0.1025, "step": 1400 }, { - "epoch": 5.480769230769231, - "grad_norm": 1.1296463012695312, - "learning_rate": 7.948888888888889e-06, - "loss": 0.0116, + "epoch": 1.20507399577167, + "grad_norm": 4.250027179718018, + "learning_rate": 9.526666666666668e-06, + "loss": 0.0903, "step": 1425 }, { - "epoch": 5.576923076923077, - "grad_norm": 1.3500127792358398, - "learning_rate": 7.893333333333335e-06, - "loss": 0.0127, + "epoch": 1.226215644820296, + "grad_norm": 3.0660858154296875, + "learning_rate": 9.513846153846155e-06, + "loss": 0.0975, "step": 1450 }, { - "epoch": 5.673076923076923, - "grad_norm": 1.6575067043304443, - "learning_rate": 7.837777777777779e-06, - "loss": 0.0126, + "epoch": 1.2473572938689217, + "grad_norm": 2.3566482067108154, + "learning_rate": 9.501025641025642e-06, + "loss": 0.0993, "step": 1475 }, { - "epoch": 5.769230769230769, - "grad_norm": 2.729482412338257, - "learning_rate": 7.782222222222223e-06, - "loss": 0.0125, + "epoch": 1.2684989429175475, + "grad_norm": 3.3394174575805664, + "learning_rate": 9.488205128205129e-06, + "loss": 0.0928, "step": 1500 }, { - "epoch": 5.865384615384615, - "grad_norm": 1.4455863237380981, - "learning_rate": 7.726666666666667e-06, - "loss": 0.0122, + "epoch": 1.2896405919661733, + "grad_norm": 3.662484645843506, + "learning_rate": 9.475384615384617e-06, + "loss": 0.0999, "step": 1525 }, { - "epoch": 5.961538461538462, - "grad_norm": 1.5944123268127441, - "learning_rate": 7.67111111111111e-06, - "loss": 0.0147, + "epoch": 1.3107822410147991, + "grad_norm": 3.5644421577453613, + "learning_rate": 9.462564102564104e-06, + "loss": 0.092, "step": 1550 }, { - "epoch": 6.0576923076923075, - "grad_norm": 0.8240692615509033, - "learning_rate": 7.6155555555555564e-06, - "loss": 0.0094, + "epoch": 1.331923890063425, + "grad_norm": 3.3423449993133545, + "learning_rate": 9.44974358974359e-06, + "loss": 0.1178, "step": 1575 }, { - "epoch": 6.153846153846154, - "grad_norm": 0.41078874468803406, - "learning_rate": 7.5600000000000005e-06, - "loss": 0.0063, + "epoch": 1.3530655391120507, + "grad_norm": 3.4482710361480713, + "learning_rate": 9.436923076923078e-06, + "loss": 0.1046, "step": 1600 }, { - "epoch": 6.25, - "grad_norm": 0.2937760651111603, - "learning_rate": 7.504444444444445e-06, - "loss": 0.008, + "epoch": 1.3742071881606766, + "grad_norm": 3.0157036781311035, + "learning_rate": 9.424102564102564e-06, + "loss": 0.0977, "step": 1625 }, { - "epoch": 6.346153846153846, - "grad_norm": 0.2572005093097687, - "learning_rate": 7.44888888888889e-06, - "loss": 0.006, + "epoch": 1.3953488372093024, + "grad_norm": 3.4354636669158936, + "learning_rate": 9.411282051282053e-06, + "loss": 0.0853, "step": 1650 }, { - "epoch": 6.4423076923076925, - "grad_norm": 0.5730597376823425, - "learning_rate": 7.393333333333333e-06, - "loss": 0.0083, + "epoch": 1.4164904862579282, + "grad_norm": 3.6379215717315674, + "learning_rate": 9.39846153846154e-06, + "loss": 0.1003, "step": 1675 }, { - "epoch": 6.538461538461538, - "grad_norm": 0.9147136211395264, - "learning_rate": 7.337777777777778e-06, - "loss": 0.0066, + "epoch": 1.437632135306554, + "grad_norm": 4.067347049713135, + "learning_rate": 9.385641025641027e-06, + "loss": 0.0795, "step": 1700 }, { - "epoch": 6.634615384615385, - "grad_norm": 0.45330244302749634, - "learning_rate": 7.282222222222222e-06, - "loss": 0.0065, + "epoch": 1.4587737843551798, + "grad_norm": 4.734683513641357, + "learning_rate": 9.372820512820513e-06, + "loss": 0.1152, "step": 1725 }, { - "epoch": 6.730769230769231, - "grad_norm": 0.6328800916671753, - "learning_rate": 7.226666666666667e-06, - "loss": 0.0078, + "epoch": 1.4799154334038054, + "grad_norm": 2.7462713718414307, + "learning_rate": 9.360000000000002e-06, + "loss": 0.0951, "step": 1750 }, { - "epoch": 6.826923076923077, - "grad_norm": 0.2872777581214905, - "learning_rate": 7.171111111111112e-06, - "loss": 0.0079, + "epoch": 1.5010570824524314, + "grad_norm": 3.8516879081726074, + "learning_rate": 9.347179487179487e-06, + "loss": 0.0895, "step": 1775 }, { - "epoch": 6.923076923076923, - "grad_norm": 1.7963801622390747, - "learning_rate": 7.115555555555557e-06, - "loss": 0.0078, + "epoch": 1.522198731501057, + "grad_norm": 4.691688537597656, + "learning_rate": 9.334358974358974e-06, + "loss": 0.1046, "step": 1800 }, { - "epoch": 7.019230769230769, - "grad_norm": 0.2437676042318344, - "learning_rate": 7.06e-06, - "loss": 0.0068, + "epoch": 1.543340380549683, + "grad_norm": 3.3214614391326904, + "learning_rate": 9.321538461538462e-06, + "loss": 0.0948, "step": 1825 }, { - "epoch": 7.115384615384615, - "grad_norm": 0.20286525785923004, - "learning_rate": 7.004444444444445e-06, - "loss": 0.0039, + "epoch": 1.5644820295983086, + "grad_norm": 3.6700942516326904, + "learning_rate": 9.30871794871795e-06, + "loss": 0.0933, "step": 1850 }, { - "epoch": 7.211538461538462, - "grad_norm": 0.16271603107452393, - "learning_rate": 6.948888888888889e-06, - "loss": 0.0038, + "epoch": 1.5856236786469344, + "grad_norm": 3.1985208988189697, + "learning_rate": 9.295897435897436e-06, + "loss": 0.0826, "step": 1875 }, { - "epoch": 7.3076923076923075, - "grad_norm": 0.12690773606300354, - "learning_rate": 6.893333333333334e-06, - "loss": 0.0037, + "epoch": 1.6067653276955602, + "grad_norm": 4.297844886779785, + "learning_rate": 9.283076923076923e-06, + "loss": 0.0886, "step": 1900 }, { - "epoch": 7.403846153846154, - "grad_norm": 0.41975799202919006, - "learning_rate": 6.837777777777779e-06, - "loss": 0.0042, + "epoch": 1.627906976744186, + "grad_norm": 1.9990788698196411, + "learning_rate": 9.270256410256411e-06, + "loss": 0.0846, "step": 1925 }, { - "epoch": 7.5, - "grad_norm": 0.2555113732814789, - "learning_rate": 6.782222222222222e-06, - "loss": 0.0033, + "epoch": 1.6490486257928119, + "grad_norm": 4.069921970367432, + "learning_rate": 9.257435897435898e-06, + "loss": 0.085, "step": 1950 }, { - "epoch": 7.596153846153846, - "grad_norm": 1.6203404664993286, - "learning_rate": 6.726666666666667e-06, - "loss": 0.0049, + "epoch": 1.6701902748414377, + "grad_norm": 3.8481404781341553, + "learning_rate": 9.244615384615385e-06, + "loss": 0.0806, "step": 1975 }, { - "epoch": 7.6923076923076925, - "grad_norm": 0.16367614269256592, - "learning_rate": 6.671111111111112e-06, - "loss": 0.0034, + "epoch": 1.6913319238900635, + "grad_norm": 2.80975604057312, + "learning_rate": 9.231794871794872e-06, + "loss": 0.0852, "step": 2000 }, { - "epoch": 7.6923076923076925, - "eval_loss": 0.31435152888298035, - "eval_runtime": 414.9289, - "eval_samples_per_second": 8.775, - "eval_steps_per_second": 0.549, - "eval_wer": 0.15889368416167804, + "epoch": 1.6913319238900635, + "eval_loss": 0.10433211177587509, + "eval_runtime": 421.4968, + "eval_samples_per_second": 8.638, + "eval_steps_per_second": 0.541, + "eval_wer": 0.08566522808482992, "step": 2000 }, { - "epoch": 7.788461538461538, - "grad_norm": 0.1244119256734848, - "learning_rate": 6.615555555555556e-06, - "loss": 0.0035, + "epoch": 1.712473572938689, + "grad_norm": 2.397152900695801, + "learning_rate": 9.21897435897436e-06, + "loss": 0.0969, "step": 2025 }, { - "epoch": 7.884615384615385, - "grad_norm": 0.7467623353004456, - "learning_rate": 6.560000000000001e-06, - "loss": 0.0042, + "epoch": 1.733615221987315, + "grad_norm": 3.2101306915283203, + "learning_rate": 9.206153846153847e-06, + "loss": 0.0975, "step": 2050 }, { - "epoch": 7.980769230769231, - "grad_norm": 0.26792147755622864, - "learning_rate": 6.504444444444446e-06, - "loss": 0.0043, + "epoch": 1.7547568710359407, + "grad_norm": 4.012678146362305, + "learning_rate": 9.193333333333334e-06, + "loss": 0.0766, "step": 2075 }, { - "epoch": 8.076923076923077, - "grad_norm": 0.0888567715883255, - "learning_rate": 6.448888888888889e-06, - "loss": 0.0029, + "epoch": 1.7758985200845667, + "grad_norm": 2.3309662342071533, + "learning_rate": 9.18051282051282e-06, + "loss": 0.0871, "step": 2100 }, { - "epoch": 8.173076923076923, - "grad_norm": 0.22752462327480316, - "learning_rate": 6.393333333333334e-06, - "loss": 0.003, + "epoch": 1.7970401691331923, + "grad_norm": 3.0771520137786865, + "learning_rate": 9.16769230769231e-06, + "loss": 0.0933, "step": 2125 }, { - "epoch": 8.26923076923077, - "grad_norm": 0.11363103985786438, - "learning_rate": 6.3377777777777786e-06, - "loss": 0.0023, + "epoch": 1.8181818181818183, + "grad_norm": 3.2476842403411865, + "learning_rate": 9.154871794871796e-06, + "loss": 0.0824, "step": 2150 }, { - "epoch": 8.365384615384615, - "grad_norm": 0.4889778196811676, - "learning_rate": 6.282222222222223e-06, - "loss": 0.0024, + "epoch": 1.839323467230444, + "grad_norm": 3.9469614028930664, + "learning_rate": 9.142051282051283e-06, + "loss": 0.0867, "step": 2175 }, { - "epoch": 8.461538461538462, - "grad_norm": 0.08544881641864777, - "learning_rate": 6.2266666666666675e-06, - "loss": 0.0023, + "epoch": 1.8604651162790697, + "grad_norm": 3.2466225624084473, + "learning_rate": 9.12923076923077e-06, + "loss": 0.0773, "step": 2200 }, { - "epoch": 8.557692307692308, - "grad_norm": 0.1377352476119995, - "learning_rate": 6.171111111111112e-06, - "loss": 0.0032, + "epoch": 1.8816067653276956, + "grad_norm": 2.797091245651245, + "learning_rate": 9.116410256410257e-06, + "loss": 0.0826, "step": 2225 }, { - "epoch": 8.653846153846153, - "grad_norm": 0.49440711736679077, - "learning_rate": 6.1155555555555555e-06, - "loss": 0.0022, + "epoch": 1.9027484143763214, + "grad_norm": 2.5921578407287598, + "learning_rate": 9.103589743589745e-06, + "loss": 0.0748, "step": 2250 }, { - "epoch": 8.75, - "grad_norm": 0.10170517861843109, - "learning_rate": 6.0600000000000004e-06, - "loss": 0.0029, + "epoch": 1.9238900634249472, + "grad_norm": 5.473248481750488, + "learning_rate": 9.090769230769232e-06, + "loss": 0.0747, "step": 2275 }, { - "epoch": 8.846153846153847, - "grad_norm": 0.0798933133482933, - "learning_rate": 6.004444444444445e-06, - "loss": 0.0025, + "epoch": 1.945031712473573, + "grad_norm": 3.044085741043091, + "learning_rate": 9.077948717948719e-06, + "loss": 0.0708, "step": 2300 }, { - "epoch": 8.942307692307692, - "grad_norm": 0.06800112873315811, - "learning_rate": 5.948888888888889e-06, - "loss": 0.0023, + "epoch": 1.9661733615221988, + "grad_norm": 3.948669672012329, + "learning_rate": 9.065128205128206e-06, + "loss": 0.0895, "step": 2325 }, { - "epoch": 9.038461538461538, - "grad_norm": 0.05597899481654167, - "learning_rate": 5.893333333333334e-06, - "loss": 0.0019, + "epoch": 1.9873150105708244, + "grad_norm": 3.0291192531585693, + "learning_rate": 9.052307692307694e-06, + "loss": 0.0803, "step": 2350 }, { - "epoch": 9.134615384615385, - "grad_norm": 0.055818960070610046, - "learning_rate": 5.837777777777777e-06, - "loss": 0.0016, + "epoch": 2.0084566596194504, + "grad_norm": 1.0824567079544067, + "learning_rate": 9.03948717948718e-06, + "loss": 0.06, "step": 2375 }, { - "epoch": 9.23076923076923, - "grad_norm": 0.07546485215425491, - "learning_rate": 5.782222222222222e-06, - "loss": 0.002, + "epoch": 2.029598308668076, + "grad_norm": 2.6235604286193848, + "learning_rate": 9.026666666666666e-06, + "loss": 0.0374, "step": 2400 }, { - "epoch": 9.326923076923077, - "grad_norm": 0.06564825028181076, - "learning_rate": 5.726666666666667e-06, - "loss": 0.0016, + "epoch": 2.050739957716702, + "grad_norm": 1.5301828384399414, + "learning_rate": 9.013846153846155e-06, + "loss": 0.036, "step": 2425 }, { - "epoch": 9.423076923076923, - "grad_norm": 0.0399746373295784, - "learning_rate": 5.671111111111112e-06, - "loss": 0.0016, + "epoch": 2.0718816067653276, + "grad_norm": 1.8513456583023071, + "learning_rate": 9.001025641025641e-06, + "loss": 0.0304, "step": 2450 }, { - "epoch": 9.51923076923077, - "grad_norm": 0.059612762182950974, - "learning_rate": 5.615555555555556e-06, - "loss": 0.0017, + "epoch": 2.0930232558139537, + "grad_norm": 1.9220322370529175, + "learning_rate": 8.988205128205128e-06, + "loss": 0.0337, "step": 2475 }, { - "epoch": 9.615384615384615, - "grad_norm": 0.06947290152311325, - "learning_rate": 5.560000000000001e-06, - "loss": 0.0016, + "epoch": 2.1141649048625792, + "grad_norm": 1.8854974508285522, + "learning_rate": 8.975384615384615e-06, + "loss": 0.0289, "step": 2500 }, { - "epoch": 9.711538461538462, - "grad_norm": 0.989651083946228, - "learning_rate": 5.504444444444444e-06, - "loss": 0.0017, + "epoch": 2.1353065539112053, + "grad_norm": 3.065642833709717, + "learning_rate": 8.962564102564104e-06, + "loss": 0.0364, "step": 2525 }, { - "epoch": 9.807692307692308, - "grad_norm": 0.07500547170639038, - "learning_rate": 5.448888888888889e-06, - "loss": 0.0016, + "epoch": 2.156448202959831, + "grad_norm": 1.0566080808639526, + "learning_rate": 8.94974358974359e-06, + "loss": 0.0339, "step": 2550 }, { - "epoch": 9.903846153846153, - "grad_norm": 0.09042783081531525, - "learning_rate": 5.393333333333334e-06, - "loss": 0.0017, + "epoch": 2.177589852008457, + "grad_norm": 1.686890721321106, + "learning_rate": 8.936923076923077e-06, + "loss": 0.0328, "step": 2575 }, { - "epoch": 10.0, - "grad_norm": 0.3791869282722473, - "learning_rate": 5.337777777777779e-06, - "loss": 0.0017, + "epoch": 2.1987315010570825, + "grad_norm": 2.0453619956970215, + "learning_rate": 8.924102564102564e-06, + "loss": 0.0369, "step": 2600 }, { - "epoch": 10.096153846153847, - "grad_norm": 0.043280228972435, - "learning_rate": 5.282222222222223e-06, - "loss": 0.0013, + "epoch": 2.219873150105708, + "grad_norm": 2.368619680404663, + "learning_rate": 8.911282051282053e-06, + "loss": 0.032, "step": 2625 }, { - "epoch": 10.192307692307692, - "grad_norm": 0.04278066009283066, - "learning_rate": 5.226666666666667e-06, - "loss": 0.0014, + "epoch": 2.241014799154334, + "grad_norm": 1.775297999382019, + "learning_rate": 8.89846153846154e-06, + "loss": 0.0315, "step": 2650 }, { - "epoch": 10.288461538461538, - "grad_norm": 0.043745964765548706, - "learning_rate": 5.171111111111111e-06, - "loss": 0.0013, + "epoch": 2.2621564482029597, + "grad_norm": 3.0013630390167236, + "learning_rate": 8.885641025641026e-06, + "loss": 0.0339, "step": 2675 }, { - "epoch": 10.384615384615385, - "grad_norm": 0.05145561695098877, - "learning_rate": 5.115555555555556e-06, - "loss": 0.0013, + "epoch": 2.2832980972515857, + "grad_norm": 1.0974608659744263, + "learning_rate": 8.872820512820513e-06, + "loss": 0.0351, "step": 2700 }, { - "epoch": 10.48076923076923, - "grad_norm": 0.04810141772031784, - "learning_rate": 5.060000000000001e-06, - "loss": 0.0013, + "epoch": 2.3044397463002113, + "grad_norm": 3.1183347702026367, + "learning_rate": 8.860000000000002e-06, + "loss": 0.0386, "step": 2725 }, { - "epoch": 10.576923076923077, - "grad_norm": 0.046568796038627625, - "learning_rate": 5.004444444444445e-06, - "loss": 0.002, + "epoch": 2.3255813953488373, + "grad_norm": 1.7240360975265503, + "learning_rate": 8.847179487179488e-06, + "loss": 0.0323, "step": 2750 }, { - "epoch": 10.673076923076923, - "grad_norm": 0.05206083878874779, - "learning_rate": 4.94888888888889e-06, - "loss": 0.0013, + "epoch": 2.346723044397463, + "grad_norm": 3.0883610248565674, + "learning_rate": 8.834358974358975e-06, + "loss": 0.0444, "step": 2775 }, { - "epoch": 10.76923076923077, - "grad_norm": 0.04524935781955719, - "learning_rate": 4.893333333333334e-06, - "loss": 0.0013, + "epoch": 2.367864693446089, + "grad_norm": 1.8349878787994385, + "learning_rate": 8.821538461538462e-06, + "loss": 0.0387, "step": 2800 }, { - "epoch": 10.865384615384615, - "grad_norm": 0.05039019510149956, - "learning_rate": 4.837777777777778e-06, - "loss": 0.0012, + "epoch": 2.3890063424947146, + "grad_norm": 3.4535439014434814, + "learning_rate": 8.80871794871795e-06, + "loss": 0.0422, "step": 2825 }, { - "epoch": 10.961538461538462, - "grad_norm": 0.048185594379901886, - "learning_rate": 4.7822222222222226e-06, - "loss": 0.0024, + "epoch": 2.41014799154334, + "grad_norm": 3.0113954544067383, + "learning_rate": 8.795897435897437e-06, + "loss": 0.0354, "step": 2850 }, { - "epoch": 11.057692307692308, - "grad_norm": 0.04220257326960564, - "learning_rate": 4.7266666666666674e-06, - "loss": 0.0012, + "epoch": 2.431289640591966, + "grad_norm": 1.868592619895935, + "learning_rate": 8.783076923076924e-06, + "loss": 0.0278, "step": 2875 }, { - "epoch": 11.153846153846153, - "grad_norm": 0.04200848937034607, - "learning_rate": 4.6711111111111115e-06, - "loss": 0.0011, + "epoch": 2.452431289640592, + "grad_norm": 2.4312257766723633, + "learning_rate": 8.770256410256411e-06, + "loss": 0.035, "step": 2900 }, { - "epoch": 11.25, - "grad_norm": 0.04376620426774025, - "learning_rate": 4.6155555555555555e-06, - "loss": 0.0011, + "epoch": 2.473572938689218, + "grad_norm": 2.4502596855163574, + "learning_rate": 8.757435897435898e-06, + "loss": 0.0314, "step": 2925 }, { - "epoch": 11.346153846153847, - "grad_norm": 0.03870835155248642, - "learning_rate": 4.56e-06, - "loss": 0.0011, + "epoch": 2.4947145877378434, + "grad_norm": 2.0330417156219482, + "learning_rate": 8.744615384615386e-06, + "loss": 0.0414, "step": 2950 }, { - "epoch": 11.442307692307692, - "grad_norm": 0.03991697356104851, - "learning_rate": 4.504444444444444e-06, - "loss": 0.0011, + "epoch": 2.5158562367864694, + "grad_norm": 1.8342368602752686, + "learning_rate": 8.731794871794873e-06, + "loss": 0.0486, "step": 2975 }, { - "epoch": 11.538461538461538, - "grad_norm": 0.03659369796514511, - "learning_rate": 4.448888888888889e-06, - "loss": 0.0011, + "epoch": 2.536997885835095, + "grad_norm": 3.516516923904419, + "learning_rate": 8.718974358974358e-06, + "loss": 0.0387, "step": 3000 }, { - "epoch": 11.538461538461538, - "eval_loss": 0.33577749133110046, - "eval_runtime": 414.076, - "eval_samples_per_second": 8.793, - "eval_steps_per_second": 0.551, - "eval_wer": 0.16153881850665924, + "epoch": 2.536997885835095, + "eval_loss": 0.09136591106653214, + "eval_runtime": 424.5707, + "eval_samples_per_second": 8.576, + "eval_steps_per_second": 0.537, + "eval_wer": 0.07568796695902362, "step": 3000 }, { - "epoch": 11.634615384615385, - "grad_norm": 0.035982731729745865, - "learning_rate": 4.393333333333334e-06, - "loss": 0.0011, + "epoch": 2.558139534883721, + "grad_norm": 2.264760732650757, + "learning_rate": 8.706153846153847e-06, + "loss": 0.0394, "step": 3025 }, { - "epoch": 11.73076923076923, - "grad_norm": 0.04060469567775726, - "learning_rate": 4.337777777777778e-06, - "loss": 0.0011, + "epoch": 2.5792811839323466, + "grad_norm": 2.3247904777526855, + "learning_rate": 8.693333333333334e-06, + "loss": 0.032, "step": 3050 }, { - "epoch": 11.826923076923077, - "grad_norm": 0.03556708246469498, - "learning_rate": 4.282222222222222e-06, - "loss": 0.0011, + "epoch": 2.6004228329809727, + "grad_norm": 2.2518162727355957, + "learning_rate": 8.68051282051282e-06, + "loss": 0.032, "step": 3075 }, { - "epoch": 11.923076923076923, - "grad_norm": 0.037598781287670135, - "learning_rate": 4.226666666666667e-06, - "loss": 0.001, + "epoch": 2.6215644820295982, + "grad_norm": 3.336057424545288, + "learning_rate": 8.667692307692307e-06, + "loss": 0.0354, "step": 3100 }, { - "epoch": 12.01923076923077, - "grad_norm": 0.03427188843488693, - "learning_rate": 4.171111111111111e-06, - "loss": 0.0011, + "epoch": 2.6427061310782243, + "grad_norm": 2.483700752258301, + "learning_rate": 8.654871794871796e-06, + "loss": 0.0371, "step": 3125 }, { - "epoch": 12.115384615384615, - "grad_norm": 0.03573931008577347, - "learning_rate": 4.115555555555556e-06, - "loss": 0.0009, + "epoch": 2.66384778012685, + "grad_norm": 1.6576588153839111, + "learning_rate": 8.642051282051283e-06, + "loss": 0.0335, "step": 3150 }, { - "epoch": 12.211538461538462, - "grad_norm": 0.03210974112153053, - "learning_rate": 4.060000000000001e-06, - "loss": 0.0009, + "epoch": 2.6849894291754755, + "grad_norm": 2.4956696033477783, + "learning_rate": 8.62923076923077e-06, + "loss": 0.0253, "step": 3175 }, { - "epoch": 12.307692307692308, - "grad_norm": 0.036103058606386185, - "learning_rate": 4.004444444444445e-06, - "loss": 0.001, + "epoch": 2.7061310782241015, + "grad_norm": 2.75431489944458, + "learning_rate": 8.616410256410256e-06, + "loss": 0.0318, "step": 3200 }, { - "epoch": 12.403846153846153, - "grad_norm": 0.036252211779356, - "learning_rate": 3.948888888888889e-06, - "loss": 0.001, + "epoch": 2.7272727272727275, + "grad_norm": 2.210634708404541, + "learning_rate": 8.603589743589745e-06, + "loss": 0.0394, "step": 3225 }, { - "epoch": 12.5, - "grad_norm": 0.03289297968149185, - "learning_rate": 3.893333333333333e-06, - "loss": 0.0009, + "epoch": 2.748414376321353, + "grad_norm": 2.500887632369995, + "learning_rate": 8.590769230769232e-06, + "loss": 0.0404, "step": 3250 }, { - "epoch": 12.596153846153847, - "grad_norm": 0.035281531512737274, - "learning_rate": 3.837777777777778e-06, - "loss": 0.001, + "epoch": 2.7695560253699787, + "grad_norm": 1.5044671297073364, + "learning_rate": 8.577948717948718e-06, + "loss": 0.031, "step": 3275 }, { - "epoch": 12.692307692307692, - "grad_norm": 0.036654986441135406, - "learning_rate": 3.782222222222223e-06, - "loss": 0.0009, + "epoch": 2.7906976744186047, + "grad_norm": 2.0523312091827393, + "learning_rate": 8.565128205128205e-06, + "loss": 0.0324, "step": 3300 }, { - "epoch": 12.788461538461538, - "grad_norm": 0.028989532962441444, - "learning_rate": 3.726666666666667e-06, - "loss": 0.0009, + "epoch": 2.8118393234672303, + "grad_norm": 3.076326608657837, + "learning_rate": 8.552307692307694e-06, + "loss": 0.0319, "step": 3325 }, { - "epoch": 12.884615384615385, - "grad_norm": 0.033068809658288956, - "learning_rate": 3.6711111111111113e-06, - "loss": 0.0009, + "epoch": 2.8329809725158563, + "grad_norm": 0.9975211024284363, + "learning_rate": 8.53948717948718e-06, + "loss": 0.0327, "step": 3350 }, { - "epoch": 12.98076923076923, - "grad_norm": 0.035067297518253326, - "learning_rate": 3.615555555555556e-06, - "loss": 0.0009, + "epoch": 2.854122621564482, + "grad_norm": 2.296332597732544, + "learning_rate": 8.526666666666667e-06, + "loss": 0.0429, "step": 3375 }, { - "epoch": 13.076923076923077, - "grad_norm": 0.03311024233698845, - "learning_rate": 3.5600000000000002e-06, - "loss": 0.0008, + "epoch": 2.875264270613108, + "grad_norm": 0.4409531354904175, + "learning_rate": 8.513846153846154e-06, + "loss": 0.0347, "step": 3400 }, { - "epoch": 13.173076923076923, - "grad_norm": 0.0331607460975647, - "learning_rate": 3.5044444444444447e-06, - "loss": 0.0009, + "epoch": 2.8964059196617336, + "grad_norm": 1.439736008644104, + "learning_rate": 8.501025641025643e-06, + "loss": 0.0337, "step": 3425 }, { - "epoch": 13.26923076923077, - "grad_norm": 0.031973276287317276, - "learning_rate": 3.4488888888888896e-06, - "loss": 0.0015, + "epoch": 2.9175475687103596, + "grad_norm": 0.9605826735496521, + "learning_rate": 8.48820512820513e-06, + "loss": 0.0351, "step": 3450 }, { - "epoch": 13.365384615384615, - "grad_norm": 0.031224718317389488, - "learning_rate": 3.3933333333333336e-06, - "loss": 0.0008, + "epoch": 2.938689217758985, + "grad_norm": 1.8650295734405518, + "learning_rate": 8.475384615384616e-06, + "loss": 0.0403, "step": 3475 }, { - "epoch": 13.461538461538462, - "grad_norm": 0.030508123338222504, - "learning_rate": 3.337777777777778e-06, - "loss": 0.0008, + "epoch": 2.9598308668076108, + "grad_norm": 2.824385643005371, + "learning_rate": 8.462564102564103e-06, + "loss": 0.0384, "step": 3500 }, { - "epoch": 13.557692307692308, - "grad_norm": 0.03208326920866966, - "learning_rate": 3.282222222222223e-06, - "loss": 0.0009, + "epoch": 2.980972515856237, + "grad_norm": 1.5251168012619019, + "learning_rate": 8.44974358974359e-06, + "loss": 0.0334, "step": 3525 }, { - "epoch": 13.653846153846153, - "grad_norm": 0.02819848246872425, - "learning_rate": 3.226666666666667e-06, - "loss": 0.0008, + "epoch": 3.0021141649048624, + "grad_norm": 0.8180655837059021, + "learning_rate": 8.436923076923079e-06, + "loss": 0.0313, "step": 3550 }, { - "epoch": 13.75, - "grad_norm": 0.02973184920847416, - "learning_rate": 3.1711111111111114e-06, - "loss": 0.0009, + "epoch": 3.0232558139534884, + "grad_norm": 2.1612069606781006, + "learning_rate": 8.424102564102565e-06, + "loss": 0.016, "step": 3575 }, { - "epoch": 13.846153846153847, - "grad_norm": 0.026373526081442833, - "learning_rate": 3.1155555555555555e-06, - "loss": 0.0008, + "epoch": 3.044397463002114, + "grad_norm": 1.3373942375183105, + "learning_rate": 8.411282051282052e-06, + "loss": 0.0144, "step": 3600 }, { - "epoch": 13.942307692307692, - "grad_norm": 0.030721429735422134, - "learning_rate": 3.0600000000000003e-06, - "loss": 0.0008, + "epoch": 3.06553911205074, + "grad_norm": 0.7444804310798645, + "learning_rate": 8.398461538461539e-06, + "loss": 0.0203, "step": 3625 }, { - "epoch": 14.038461538461538, - "grad_norm": 0.027194401249289513, - "learning_rate": 3.004444444444445e-06, - "loss": 0.0008, + "epoch": 3.0866807610993656, + "grad_norm": 1.6546038389205933, + "learning_rate": 8.385641025641026e-06, + "loss": 0.0109, "step": 3650 }, { - "epoch": 14.134615384615385, - "grad_norm": 0.02467515505850315, - "learning_rate": 2.948888888888889e-06, - "loss": 0.0008, + "epoch": 3.1078224101479917, + "grad_norm": 1.1087496280670166, + "learning_rate": 8.372820512820513e-06, + "loss": 0.0139, "step": 3675 }, { - "epoch": 14.23076923076923, - "grad_norm": 0.02789357490837574, - "learning_rate": 2.8933333333333337e-06, - "loss": 0.0008, + "epoch": 3.1289640591966172, + "grad_norm": 1.7939504384994507, + "learning_rate": 8.36e-06, + "loss": 0.0185, "step": 3700 }, { - "epoch": 14.326923076923077, - "grad_norm": 0.02929651364684105, - "learning_rate": 2.837777777777778e-06, - "loss": 0.0008, + "epoch": 3.1501057082452433, + "grad_norm": 1.848943829536438, + "learning_rate": 8.347179487179488e-06, + "loss": 0.0128, "step": 3725 }, { - "epoch": 14.423076923076923, - "grad_norm": 0.02698478475213051, - "learning_rate": 2.7822222222222222e-06, - "loss": 0.0008, + "epoch": 3.171247357293869, + "grad_norm": 2.3288862705230713, + "learning_rate": 8.334358974358975e-06, + "loss": 0.0132, "step": 3750 }, { - "epoch": 14.51923076923077, - "grad_norm": 0.026549380272626877, - "learning_rate": 2.726666666666667e-06, - "loss": 0.0008, + "epoch": 3.192389006342495, + "grad_norm": 2.1424756050109863, + "learning_rate": 8.321538461538462e-06, + "loss": 0.017, "step": 3775 }, { - "epoch": 14.615384615384615, - "grad_norm": 0.025898369029164314, - "learning_rate": 2.6711111111111116e-06, - "loss": 0.0008, + "epoch": 3.2135306553911205, + "grad_norm": 1.132733941078186, + "learning_rate": 8.308717948717949e-06, + "loss": 0.0181, "step": 3800 }, { - "epoch": 14.711538461538462, - "grad_norm": 0.024109981954097748, - "learning_rate": 2.6155555555555556e-06, - "loss": 0.0008, + "epoch": 3.234672304439746, + "grad_norm": 1.5112519264221191, + "learning_rate": 8.295897435897437e-06, + "loss": 0.012, "step": 3825 }, { - "epoch": 14.807692307692308, - "grad_norm": 0.028975041583180428, - "learning_rate": 2.56e-06, - "loss": 0.0007, + "epoch": 3.255813953488372, + "grad_norm": 1.1612873077392578, + "learning_rate": 8.283076923076924e-06, + "loss": 0.0142, "step": 3850 }, { - "epoch": 14.903846153846153, - "grad_norm": 0.029065443202853203, - "learning_rate": 2.504444444444445e-06, - "loss": 0.0008, + "epoch": 3.276955602536998, + "grad_norm": 2.4286863803863525, + "learning_rate": 8.27025641025641e-06, + "loss": 0.0106, "step": 3875 }, { - "epoch": 15.0, - "grad_norm": 0.03024737536907196, - "learning_rate": 2.448888888888889e-06, - "loss": 0.0013, + "epoch": 3.2980972515856237, + "grad_norm": 2.1390466690063477, + "learning_rate": 8.257435897435898e-06, + "loss": 0.0176, "step": 3900 }, { - "epoch": 15.096153846153847, - "grad_norm": 0.020746273919939995, - "learning_rate": 2.3933333333333334e-06, - "loss": 0.0007, + "epoch": 3.3192389006342493, + "grad_norm": 3.464160680770874, + "learning_rate": 8.244615384615386e-06, + "loss": 0.0143, "step": 3925 }, { - "epoch": 15.192307692307692, - "grad_norm": 0.022498076781630516, - "learning_rate": 2.337777777777778e-06, - "loss": 0.0007, + "epoch": 3.3403805496828753, + "grad_norm": 1.063372254371643, + "learning_rate": 8.231794871794873e-06, + "loss": 0.0196, "step": 3950 }, { - "epoch": 15.288461538461538, - "grad_norm": 0.025430675595998764, - "learning_rate": 2.2822222222222223e-06, - "loss": 0.0007, + "epoch": 3.361522198731501, + "grad_norm": 0.763058066368103, + "learning_rate": 8.21897435897436e-06, + "loss": 0.0152, "step": 3975 }, { - "epoch": 15.384615384615385, - "grad_norm": 0.024518582969903946, - "learning_rate": 2.226666666666667e-06, - "loss": 0.0007, + "epoch": 3.382663847780127, + "grad_norm": 0.30352699756622314, + "learning_rate": 8.206153846153847e-06, + "loss": 0.0153, "step": 4000 }, { - "epoch": 15.384615384615385, - "eval_loss": 0.34604960680007935, - "eval_runtime": 414.6061, - "eval_samples_per_second": 8.782, - "eval_steps_per_second": 0.55, - "eval_wer": 0.1614460067752564, + "epoch": 3.382663847780127, + "eval_loss": 0.08595944941043854, + "eval_runtime": 425.6418, + "eval_samples_per_second": 8.554, + "eval_steps_per_second": 0.536, + "eval_wer": 0.08181354123161168, "step": 4000 }, { - "epoch": 15.48076923076923, - "grad_norm": 0.028795845806598663, - "learning_rate": 2.1711111111111113e-06, - "loss": 0.0007, + "epoch": 3.4038054968287526, + "grad_norm": 0.6997278332710266, + "learning_rate": 8.193333333333335e-06, + "loss": 0.0121, "step": 4025 }, { - "epoch": 15.576923076923077, - "grad_norm": 0.02770942822098732, - "learning_rate": 2.1155555555555557e-06, - "loss": 0.0007, + "epoch": 3.4249471458773786, + "grad_norm": 2.2743613719940186, + "learning_rate": 8.180512820512822e-06, + "loss": 0.0148, "step": 4050 }, { - "epoch": 15.673076923076923, - "grad_norm": 0.026291629299521446, - "learning_rate": 2.06e-06, - "loss": 0.0008, + "epoch": 3.446088794926004, + "grad_norm": 1.2777513265609741, + "learning_rate": 8.167692307692309e-06, + "loss": 0.0142, "step": 4075 }, { - "epoch": 15.76923076923077, - "grad_norm": 0.024974102154374123, - "learning_rate": 2.0044444444444446e-06, - "loss": 0.0007, + "epoch": 3.46723044397463, + "grad_norm": 1.5009987354278564, + "learning_rate": 8.154871794871796e-06, + "loss": 0.0191, "step": 4100 }, { - "epoch": 15.865384615384615, - "grad_norm": 0.02725864201784134, - "learning_rate": 1.948888888888889e-06, - "loss": 0.0007, + "epoch": 3.488372093023256, + "grad_norm": 1.3110395669937134, + "learning_rate": 8.142051282051282e-06, + "loss": 0.0133, "step": 4125 }, { - "epoch": 15.961538461538462, - "grad_norm": 0.027727024629712105, - "learning_rate": 1.8933333333333333e-06, - "loss": 0.0007, + "epoch": 3.5095137420718814, + "grad_norm": 1.5313085317611694, + "learning_rate": 8.129230769230771e-06, + "loss": 0.0141, "step": 4150 }, { - "epoch": 16.057692307692307, - "grad_norm": 0.023184822872281075, - "learning_rate": 1.837777777777778e-06, - "loss": 0.0007, + "epoch": 3.5306553911205074, + "grad_norm": 1.2553045749664307, + "learning_rate": 8.116410256410258e-06, + "loss": 0.0169, "step": 4175 }, { - "epoch": 16.153846153846153, - "grad_norm": 0.020365368574857712, - "learning_rate": 1.7822222222222225e-06, - "loss": 0.0007, + "epoch": 3.5517970401691334, + "grad_norm": 1.313785433769226, + "learning_rate": 8.103589743589745e-06, + "loss": 0.0193, "step": 4200 }, { - "epoch": 16.25, - "grad_norm": 0.023292195051908493, - "learning_rate": 1.7266666666666667e-06, - "loss": 0.0006, + "epoch": 3.572938689217759, + "grad_norm": 1.203377366065979, + "learning_rate": 8.090769230769231e-06, + "loss": 0.0151, "step": 4225 }, { - "epoch": 16.346153846153847, - "grad_norm": 0.02580893039703369, - "learning_rate": 1.6711111111111112e-06, - "loss": 0.0006, + "epoch": 3.5940803382663846, + "grad_norm": 0.6605008244514465, + "learning_rate": 8.077948717948718e-06, + "loss": 0.0181, "step": 4250 }, { - "epoch": 16.442307692307693, - "grad_norm": 0.02440100722014904, - "learning_rate": 1.6155555555555559e-06, - "loss": 0.0007, + "epoch": 3.6152219873150107, + "grad_norm": 1.213782787322998, + "learning_rate": 8.065128205128205e-06, + "loss": 0.019, "step": 4275 }, { - "epoch": 16.53846153846154, - "grad_norm": 0.02583976835012436, - "learning_rate": 1.56e-06, - "loss": 0.0007, + "epoch": 3.6363636363636362, + "grad_norm": 3.021005153656006, + "learning_rate": 8.052307692307692e-06, + "loss": 0.0169, "step": 4300 }, { - "epoch": 16.634615384615383, - "grad_norm": 0.024059804156422615, - "learning_rate": 1.5044444444444446e-06, - "loss": 0.0007, + "epoch": 3.6575052854122623, + "grad_norm": 1.8544044494628906, + "learning_rate": 8.03948717948718e-06, + "loss": 0.0131, "step": 4325 }, { - "epoch": 16.73076923076923, - "grad_norm": 0.023279106244444847, - "learning_rate": 1.4488888888888892e-06, - "loss": 0.0007, + "epoch": 3.678646934460888, + "grad_norm": 1.1160951852798462, + "learning_rate": 8.026666666666667e-06, + "loss": 0.0174, "step": 4350 }, { - "epoch": 16.826923076923077, - "grad_norm": 0.022877462208271027, - "learning_rate": 1.3933333333333335e-06, - "loss": 0.0007, + "epoch": 3.699788583509514, + "grad_norm": 0.6091210246086121, + "learning_rate": 8.013846153846154e-06, + "loss": 0.0125, "step": 4375 }, { - "epoch": 16.923076923076923, - "grad_norm": 0.02401687018573284, - "learning_rate": 1.337777777777778e-06, - "loss": 0.0006, + "epoch": 3.7209302325581395, + "grad_norm": 1.5282201766967773, + "learning_rate": 8.00102564102564e-06, + "loss": 0.0193, "step": 4400 }, { - "epoch": 17.01923076923077, - "grad_norm": 0.021333323791623116, - "learning_rate": 1.2822222222222222e-06, - "loss": 0.0006, + "epoch": 3.7420718816067655, + "grad_norm": 0.671487033367157, + "learning_rate": 7.98820512820513e-06, + "loss": 0.0142, "step": 4425 }, { - "epoch": 17.115384615384617, - "grad_norm": 0.019655166193842888, - "learning_rate": 1.2266666666666666e-06, - "loss": 0.0006, + "epoch": 3.763213530655391, + "grad_norm": 1.4481194019317627, + "learning_rate": 7.975384615384616e-06, + "loss": 0.0205, "step": 4450 }, { - "epoch": 17.21153846153846, - "grad_norm": 0.02106211706995964, - "learning_rate": 1.171111111111111e-06, - "loss": 0.0006, + "epoch": 3.7843551797040167, + "grad_norm": 1.552512288093567, + "learning_rate": 7.962564102564103e-06, + "loss": 0.0156, "step": 4475 }, { - "epoch": 17.307692307692307, - "grad_norm": 0.022571563720703125, - "learning_rate": 1.1155555555555558e-06, - "loss": 0.0006, + "epoch": 3.8054968287526427, + "grad_norm": 2.821441411972046, + "learning_rate": 7.94974358974359e-06, + "loss": 0.0205, "step": 4500 }, { - "epoch": 17.403846153846153, - "grad_norm": 0.023800240829586983, - "learning_rate": 1.06e-06, - "loss": 0.0006, + "epoch": 3.8266384778012688, + "grad_norm": 1.934550404548645, + "learning_rate": 7.936923076923078e-06, + "loss": 0.014, "step": 4525 }, { - "epoch": 17.5, - "grad_norm": 0.0228471290320158, - "learning_rate": 1.0044444444444445e-06, - "loss": 0.0006, + "epoch": 3.8477801268498943, + "grad_norm": 1.8110979795455933, + "learning_rate": 7.924102564102565e-06, + "loss": 0.015, "step": 4550 }, { - "epoch": 17.596153846153847, - "grad_norm": 0.021810224279761314, - "learning_rate": 9.488888888888889e-07, - "loss": 0.0006, + "epoch": 3.86892177589852, + "grad_norm": 1.9725172519683838, + "learning_rate": 7.911282051282052e-06, + "loss": 0.0149, "step": 4575 }, { - "epoch": 17.692307692307693, - "grad_norm": 0.02282741293311119, - "learning_rate": 8.933333333333334e-07, - "loss": 0.0008, + "epoch": 3.890063424947146, + "grad_norm": 1.2784725427627563, + "learning_rate": 7.898461538461539e-06, + "loss": 0.0146, "step": 4600 }, { - "epoch": 17.78846153846154, - "grad_norm": 0.024272125214338303, - "learning_rate": 8.37777777777778e-07, - "loss": 0.0006, + "epoch": 3.9112050739957716, + "grad_norm": 2.2306292057037354, + "learning_rate": 7.885641025641027e-06, + "loss": 0.0163, "step": 4625 }, { - "epoch": 17.884615384615383, - "grad_norm": 0.026012342423200607, - "learning_rate": 7.822222222222223e-07, - "loss": 0.0006, + "epoch": 3.9323467230443976, + "grad_norm": 1.8566030263900757, + "learning_rate": 7.872820512820514e-06, + "loss": 0.015, "step": 4650 }, { - "epoch": 17.98076923076923, - "grad_norm": 0.02298705466091633, - "learning_rate": 7.266666666666668e-07, - "loss": 0.0006, + "epoch": 3.953488372093023, + "grad_norm": 1.6263699531555176, + "learning_rate": 7.860000000000001e-06, + "loss": 0.0145, "step": 4675 }, { - "epoch": 18.076923076923077, - "grad_norm": 0.021737568080425262, - "learning_rate": 6.711111111111111e-07, - "loss": 0.0006, + "epoch": 3.974630021141649, + "grad_norm": 1.2644612789154053, + "learning_rate": 7.847179487179488e-06, + "loss": 0.0158, "step": 4700 }, { - "epoch": 18.173076923076923, - "grad_norm": 0.021294375881552696, - "learning_rate": 6.155555555555556e-07, - "loss": 0.0006, + "epoch": 3.995771670190275, + "grad_norm": 0.5040600299835205, + "learning_rate": 7.834871794871795e-06, + "loss": 0.0162, "step": 4725 }, { - "epoch": 18.26923076923077, - "grad_norm": 0.0215039923787117, - "learning_rate": 5.6e-07, - "loss": 0.0006, + "epoch": 4.016913319238901, + "grad_norm": 0.6542588472366333, + "learning_rate": 7.822051282051282e-06, + "loss": 0.009, "step": 4750 }, { - "epoch": 18.365384615384617, - "grad_norm": 0.02031785622239113, - "learning_rate": 5.044444444444445e-07, - "loss": 0.0006, + "epoch": 4.038054968287526, + "grad_norm": 1.0600502490997314, + "learning_rate": 7.80923076923077e-06, + "loss": 0.0125, "step": 4775 }, { - "epoch": 18.46153846153846, - "grad_norm": 0.02249339409172535, - "learning_rate": 4.488888888888889e-07, - "loss": 0.0006, + "epoch": 4.059196617336152, + "grad_norm": 0.7481040358543396, + "learning_rate": 7.796410256410257e-06, + "loss": 0.0065, "step": 4800 }, { - "epoch": 18.557692307692307, - "grad_norm": 0.022298617288470268, - "learning_rate": 3.9333333333333336e-07, - "loss": 0.0006, + "epoch": 4.080338266384778, + "grad_norm": 0.9099847078323364, + "learning_rate": 7.783589743589744e-06, + "loss": 0.0057, "step": 4825 }, { - "epoch": 18.653846153846153, - "grad_norm": 0.019332217052578926, - "learning_rate": 3.3777777777777777e-07, - "loss": 0.0006, + "epoch": 4.101479915433404, + "grad_norm": 0.5291013717651367, + "learning_rate": 7.770769230769231e-06, + "loss": 0.005, "step": 4850 }, { - "epoch": 18.75, - "grad_norm": 0.02518359012901783, - "learning_rate": 2.822222222222222e-07, - "loss": 0.0006, + "epoch": 4.12262156448203, + "grad_norm": 1.499250888824463, + "learning_rate": 7.75794871794872e-06, + "loss": 0.0083, "step": 4875 }, { - "epoch": 18.846153846153847, - "grad_norm": 0.022782141342759132, - "learning_rate": 2.266666666666667e-07, - "loss": 0.0006, + "epoch": 4.143763213530655, + "grad_norm": 0.21640999615192413, + "learning_rate": 7.745128205128206e-06, + "loss": 0.0075, "step": 4900 }, { - "epoch": 18.942307692307693, - "grad_norm": 0.023741982877254486, - "learning_rate": 1.7111111111111114e-07, - "loss": 0.0006, + "epoch": 4.164904862579281, + "grad_norm": 0.20196348428726196, + "learning_rate": 7.732307692307693e-06, + "loss": 0.0088, "step": 4925 }, { - "epoch": 19.03846153846154, - "grad_norm": 0.02083681896328926, - "learning_rate": 1.1555555555555556e-07, - "loss": 0.0006, + "epoch": 4.186046511627907, + "grad_norm": 0.6973519921302795, + "learning_rate": 7.71948717948718e-06, + "loss": 0.0079, "step": 4950 }, { - "epoch": 19.134615384615383, - "grad_norm": 0.020423738285899162, - "learning_rate": 6.000000000000001e-08, - "loss": 0.0006, + "epoch": 4.207188160676533, + "grad_norm": 0.38807210326194763, + "learning_rate": 7.706666666666669e-06, + "loss": 0.005, "step": 4975 }, { - "epoch": 19.23076923076923, - "grad_norm": 0.0213633980602026, - "learning_rate": 4.444444444444445e-09, - "loss": 0.0006, + "epoch": 4.2283298097251585, + "grad_norm": 0.19292686879634857, + "learning_rate": 7.693846153846154e-06, + "loss": 0.008, "step": 5000 }, { - "epoch": 19.23076923076923, - "eval_loss": 0.35056987404823303, - "eval_runtime": 416.1932, - "eval_samples_per_second": 8.748, - "eval_steps_per_second": 0.548, - "eval_wer": 0.16172444196946495, + "epoch": 4.2283298097251585, + "eval_loss": 0.08777374029159546, + "eval_runtime": 416.8384, + "eval_samples_per_second": 8.735, + "eval_steps_per_second": 0.547, + "eval_wer": 0.06984082788064411, "step": 5000 }, { - "epoch": 19.23076923076923, - "step": 5000, - "total_flos": 4.616818092859392e+19, - "train_loss": 0.05540717707388103, - "train_runtime": 18195.7149, - "train_samples_per_second": 8.793, + "epoch": 4.249471458773784, + "grad_norm": 0.18094445765018463, + "learning_rate": 7.68102564102564e-06, + "loss": 0.0059, + "step": 5025 + }, + { + "epoch": 4.2706131078224105, + "grad_norm": 1.249588131904602, + "learning_rate": 7.668205128205129e-06, + "loss": 0.0063, + "step": 5050 + }, + { + "epoch": 4.291754756871036, + "grad_norm": 1.3055378198623657, + "learning_rate": 7.655384615384616e-06, + "loss": 0.0074, + "step": 5075 + }, + { + "epoch": 4.312896405919662, + "grad_norm": 1.6589789390563965, + "learning_rate": 7.642564102564103e-06, + "loss": 0.0094, + "step": 5100 + }, + { + "epoch": 4.334038054968287, + "grad_norm": 0.46987977623939514, + "learning_rate": 7.62974358974359e-06, + "loss": 0.0117, + "step": 5125 + }, + { + "epoch": 4.355179704016914, + "grad_norm": 0.4591355323791504, + "learning_rate": 7.616923076923077e-06, + "loss": 0.0079, + "step": 5150 + }, + { + "epoch": 4.376321353065539, + "grad_norm": 1.6983137130737305, + "learning_rate": 7.604102564102565e-06, + "loss": 0.0053, + "step": 5175 + }, + { + "epoch": 4.397463002114165, + "grad_norm": 0.3046073913574219, + "learning_rate": 7.591282051282052e-06, + "loss": 0.0104, + "step": 5200 + }, + { + "epoch": 4.4186046511627906, + "grad_norm": 0.7796456217765808, + "learning_rate": 7.578461538461539e-06, + "loss": 0.0059, + "step": 5225 + }, + { + "epoch": 4.439746300211416, + "grad_norm": 0.7012127637863159, + "learning_rate": 7.565641025641026e-06, + "loss": 0.0076, + "step": 5250 + }, + { + "epoch": 4.460887949260043, + "grad_norm": 0.6566082835197449, + "learning_rate": 7.552820512820514e-06, + "loss": 0.0085, + "step": 5275 + }, + { + "epoch": 4.482029598308668, + "grad_norm": 1.1547282934188843, + "learning_rate": 7.540000000000001e-06, + "loss": 0.0082, + "step": 5300 + }, + { + "epoch": 4.503171247357294, + "grad_norm": 1.3107972145080566, + "learning_rate": 7.5271794871794875e-06, + "loss": 0.007, + "step": 5325 + }, + { + "epoch": 4.524312896405919, + "grad_norm": 1.5770258903503418, + "learning_rate": 7.514358974358975e-06, + "loss": 0.0118, + "step": 5350 + }, + { + "epoch": 4.545454545454545, + "grad_norm": 2.0194594860076904, + "learning_rate": 7.501538461538462e-06, + "loss": 0.0104, + "step": 5375 + }, + { + "epoch": 4.5665961945031714, + "grad_norm": 1.2425894737243652, + "learning_rate": 7.48871794871795e-06, + "loss": 0.009, + "step": 5400 + }, + { + "epoch": 4.587737843551797, + "grad_norm": 1.568264365196228, + "learning_rate": 7.4758974358974365e-06, + "loss": 0.0095, + "step": 5425 + }, + { + "epoch": 4.608879492600423, + "grad_norm": 0.7294540405273438, + "learning_rate": 7.463076923076924e-06, + "loss": 0.0092, + "step": 5450 + }, + { + "epoch": 4.630021141649049, + "grad_norm": 1.0853856801986694, + "learning_rate": 7.450256410256411e-06, + "loss": 0.0078, + "step": 5475 + }, + { + "epoch": 4.651162790697675, + "grad_norm": 1.0923691987991333, + "learning_rate": 7.437435897435899e-06, + "loss": 0.0071, + "step": 5500 + }, + { + "epoch": 4.6723044397463, + "grad_norm": 1.096602201461792, + "learning_rate": 7.4246153846153855e-06, + "loss": 0.0072, + "step": 5525 + }, + { + "epoch": 4.693446088794926, + "grad_norm": 1.0058438777923584, + "learning_rate": 7.411794871794873e-06, + "loss": 0.009, + "step": 5550 + }, + { + "epoch": 4.7145877378435515, + "grad_norm": 0.3397851884365082, + "learning_rate": 7.39897435897436e-06, + "loss": 0.006, + "step": 5575 + }, + { + "epoch": 4.735729386892178, + "grad_norm": 0.9336525201797485, + "learning_rate": 7.386153846153846e-06, + "loss": 0.0144, + "step": 5600 + }, + { + "epoch": 4.7568710359408035, + "grad_norm": 0.1965409815311432, + "learning_rate": 7.373333333333334e-06, + "loss": 0.0057, + "step": 5625 + }, + { + "epoch": 4.778012684989429, + "grad_norm": 0.3345631957054138, + "learning_rate": 7.3605128205128204e-06, + "loss": 0.008, + "step": 5650 + }, + { + "epoch": 4.799154334038055, + "grad_norm": 0.8351041674613953, + "learning_rate": 7.347692307692308e-06, + "loss": 0.0063, + "step": 5675 + }, + { + "epoch": 4.82029598308668, + "grad_norm": 0.6171631217002869, + "learning_rate": 7.334871794871795e-06, + "loss": 0.0056, + "step": 5700 + }, + { + "epoch": 4.841437632135307, + "grad_norm": 0.22757196426391602, + "learning_rate": 7.322051282051283e-06, + "loss": 0.0098, + "step": 5725 + }, + { + "epoch": 4.862579281183932, + "grad_norm": 2.954468011856079, + "learning_rate": 7.309230769230769e-06, + "loss": 0.0056, + "step": 5750 + }, + { + "epoch": 4.883720930232558, + "grad_norm": 1.6333624124526978, + "learning_rate": 7.296410256410257e-06, + "loss": 0.0067, + "step": 5775 + }, + { + "epoch": 4.904862579281184, + "grad_norm": 0.10707265883684158, + "learning_rate": 7.283589743589744e-06, + "loss": 0.0055, + "step": 5800 + }, + { + "epoch": 4.92600422832981, + "grad_norm": 0.8286884427070618, + "learning_rate": 7.270769230769232e-06, + "loss": 0.0081, + "step": 5825 + }, + { + "epoch": 4.947145877378436, + "grad_norm": 0.42702218890190125, + "learning_rate": 7.257948717948718e-06, + "loss": 0.0064, + "step": 5850 + }, + { + "epoch": 4.968287526427061, + "grad_norm": 1.5272870063781738, + "learning_rate": 7.245128205128206e-06, + "loss": 0.006, + "step": 5875 + }, + { + "epoch": 4.989429175475687, + "grad_norm": 1.0080416202545166, + "learning_rate": 7.232307692307693e-06, + "loss": 0.0064, + "step": 5900 + }, + { + "epoch": 5.010570824524313, + "grad_norm": 1.6429482698440552, + "learning_rate": 7.21948717948718e-06, + "loss": 0.004, + "step": 5925 + }, + { + "epoch": 5.031712473572939, + "grad_norm": 0.593721866607666, + "learning_rate": 7.206666666666667e-06, + "loss": 0.0033, + "step": 5950 + }, + { + "epoch": 5.052854122621564, + "grad_norm": 0.20310665667057037, + "learning_rate": 7.193846153846154e-06, + "loss": 0.0045, + "step": 5975 + }, + { + "epoch": 5.07399577167019, + "grad_norm": 1.1912012100219727, + "learning_rate": 7.181025641025642e-06, + "loss": 0.005, + "step": 6000 + }, + { + "epoch": 5.07399577167019, + "eval_loss": 0.08779177069664001, + "eval_runtime": 422.2642, + "eval_samples_per_second": 8.623, + "eval_steps_per_second": 0.54, + "eval_wer": 0.07452782031648801, + "step": 6000 + }, + { + "epoch": 5.0951374207188165, + "grad_norm": 0.7110195159912109, + "learning_rate": 7.168205128205129e-06, + "loss": 0.0077, + "step": 6025 + }, + { + "epoch": 5.116279069767442, + "grad_norm": 0.11154381185770035, + "learning_rate": 7.155384615384616e-06, + "loss": 0.004, + "step": 6050 + }, + { + "epoch": 5.137420718816068, + "grad_norm": 0.07287934422492981, + "learning_rate": 7.142564102564103e-06, + "loss": 0.003, + "step": 6075 + }, + { + "epoch": 5.158562367864693, + "grad_norm": 0.07207977771759033, + "learning_rate": 7.129743589743591e-06, + "loss": 0.0029, + "step": 6100 + }, + { + "epoch": 5.179704016913319, + "grad_norm": 0.09618539363145828, + "learning_rate": 7.116923076923078e-06, + "loss": 0.0045, + "step": 6125 + }, + { + "epoch": 5.200845665961945, + "grad_norm": 0.13538002967834473, + "learning_rate": 7.104102564102565e-06, + "loss": 0.0036, + "step": 6150 + }, + { + "epoch": 5.221987315010571, + "grad_norm": 0.08731002360582352, + "learning_rate": 7.091282051282052e-06, + "loss": 0.0039, + "step": 6175 + }, + { + "epoch": 5.2431289640591965, + "grad_norm": 1.615174412727356, + "learning_rate": 7.078461538461538e-06, + "loss": 0.0037, + "step": 6200 + }, + { + "epoch": 5.264270613107822, + "grad_norm": 0.844247579574585, + "learning_rate": 7.065641025641026e-06, + "loss": 0.0044, + "step": 6225 + }, + { + "epoch": 5.2854122621564485, + "grad_norm": 0.36915040016174316, + "learning_rate": 7.052820512820513e-06, + "loss": 0.0039, + "step": 6250 + }, + { + "epoch": 5.306553911205074, + "grad_norm": 0.6561350226402283, + "learning_rate": 7.04e-06, + "loss": 0.0077, + "step": 6275 + }, + { + "epoch": 5.3276955602537, + "grad_norm": 0.10375604033470154, + "learning_rate": 7.027179487179487e-06, + "loss": 0.0027, + "step": 6300 + }, + { + "epoch": 5.348837209302325, + "grad_norm": 0.11399765312671661, + "learning_rate": 7.014358974358975e-06, + "loss": 0.005, + "step": 6325 + }, + { + "epoch": 5.369978858350952, + "grad_norm": 0.39025041460990906, + "learning_rate": 7.001538461538462e-06, + "loss": 0.0032, + "step": 6350 + }, + { + "epoch": 5.391120507399577, + "grad_norm": 0.7788843512535095, + "learning_rate": 6.988717948717949e-06, + "loss": 0.0047, + "step": 6375 + }, + { + "epoch": 5.412262156448203, + "grad_norm": 5.279190540313721, + "learning_rate": 6.975897435897436e-06, + "loss": 0.0046, + "step": 6400 + }, + { + "epoch": 5.4334038054968286, + "grad_norm": 0.554918646812439, + "learning_rate": 6.963076923076924e-06, + "loss": 0.0039, + "step": 6425 + }, + { + "epoch": 5.454545454545454, + "grad_norm": 0.062186941504478455, + "learning_rate": 6.950256410256411e-06, + "loss": 0.0027, + "step": 6450 + }, + { + "epoch": 5.475687103594081, + "grad_norm": 0.22088727355003357, + "learning_rate": 6.937435897435898e-06, + "loss": 0.004, + "step": 6475 + }, + { + "epoch": 5.496828752642706, + "grad_norm": 1.3965909481048584, + "learning_rate": 6.924615384615385e-06, + "loss": 0.0046, + "step": 6500 + }, + { + "epoch": 5.517970401691332, + "grad_norm": 0.13480918109416962, + "learning_rate": 6.911794871794872e-06, + "loss": 0.0044, + "step": 6525 + }, + { + "epoch": 5.539112050739957, + "grad_norm": 0.2856655716896057, + "learning_rate": 6.89897435897436e-06, + "loss": 0.005, + "step": 6550 + }, + { + "epoch": 5.560253699788584, + "grad_norm": 0.8899367451667786, + "learning_rate": 6.8861538461538465e-06, + "loss": 0.0035, + "step": 6575 + }, + { + "epoch": 5.5813953488372094, + "grad_norm": 1.383070707321167, + "learning_rate": 6.873333333333334e-06, + "loss": 0.0054, + "step": 6600 + }, + { + "epoch": 5.602536997885835, + "grad_norm": 0.2242678701877594, + "learning_rate": 6.860512820512821e-06, + "loss": 0.0074, + "step": 6625 + }, + { + "epoch": 5.623678646934461, + "grad_norm": 0.3287247121334076, + "learning_rate": 6.847692307692309e-06, + "loss": 0.0068, + "step": 6650 + }, + { + "epoch": 5.644820295983086, + "grad_norm": 0.47697022557258606, + "learning_rate": 6.8348717948717955e-06, + "loss": 0.0036, + "step": 6675 + }, + { + "epoch": 5.665961945031713, + "grad_norm": 0.23843590915203094, + "learning_rate": 6.822051282051283e-06, + "loss": 0.0053, + "step": 6700 + }, + { + "epoch": 5.687103594080338, + "grad_norm": 0.14214707911014557, + "learning_rate": 6.80923076923077e-06, + "loss": 0.0023, + "step": 6725 + }, + { + "epoch": 5.708245243128964, + "grad_norm": 0.9868163466453552, + "learning_rate": 6.796410256410258e-06, + "loss": 0.003, + "step": 6750 + }, + { + "epoch": 5.72938689217759, + "grad_norm": 0.6320108771324158, + "learning_rate": 6.7835897435897445e-06, + "loss": 0.0061, + "step": 6775 + }, + { + "epoch": 5.750528541226216, + "grad_norm": 0.08067617565393448, + "learning_rate": 6.770769230769232e-06, + "loss": 0.0055, + "step": 6800 + }, + { + "epoch": 5.7716701902748415, + "grad_norm": 1.7972437143325806, + "learning_rate": 6.757948717948718e-06, + "loss": 0.0039, + "step": 6825 + }, + { + "epoch": 5.792811839323467, + "grad_norm": 0.9350019693374634, + "learning_rate": 6.745128205128205e-06, + "loss": 0.0055, + "step": 6850 + }, + { + "epoch": 5.813953488372093, + "grad_norm": 0.21952944993972778, + "learning_rate": 6.732307692307693e-06, + "loss": 0.0037, + "step": 6875 + }, + { + "epoch": 5.835095137420719, + "grad_norm": 2.4713776111602783, + "learning_rate": 6.7194871794871794e-06, + "loss": 0.0056, + "step": 6900 + }, + { + "epoch": 5.856236786469345, + "grad_norm": 0.4809707999229431, + "learning_rate": 6.706666666666667e-06, + "loss": 0.0046, + "step": 6925 + }, + { + "epoch": 5.87737843551797, + "grad_norm": 1.1874123811721802, + "learning_rate": 6.693846153846154e-06, + "loss": 0.0046, + "step": 6950 + }, + { + "epoch": 5.898520084566596, + "grad_norm": 0.13787618279457092, + "learning_rate": 6.681025641025642e-06, + "loss": 0.004, + "step": 6975 + }, + { + "epoch": 5.9196617336152215, + "grad_norm": 1.4916746616363525, + "learning_rate": 6.6682051282051284e-06, + "loss": 0.0033, + "step": 7000 + }, + { + "epoch": 5.9196617336152215, + "eval_loss": 0.08342708647251129, + "eval_runtime": 419.2861, + "eval_samples_per_second": 8.684, + "eval_steps_per_second": 0.544, + "eval_wer": 0.0651074295790988, + "step": 7000 + }, + { + "epoch": 5.940803382663848, + "grad_norm": 2.867432117462158, + "learning_rate": 6.655384615384616e-06, + "loss": 0.0097, + "step": 7025 + }, + { + "epoch": 5.961945031712474, + "grad_norm": 0.590003490447998, + "learning_rate": 6.642564102564103e-06, + "loss": 0.0032, + "step": 7050 + }, + { + "epoch": 5.983086680761099, + "grad_norm": 2.920905351638794, + "learning_rate": 6.629743589743591e-06, + "loss": 0.005, + "step": 7075 + }, + { + "epoch": 6.004228329809725, + "grad_norm": 0.9033246040344238, + "learning_rate": 6.616923076923077e-06, + "loss": 0.0062, + "step": 7100 + }, + { + "epoch": 6.025369978858351, + "grad_norm": 5.353052139282227, + "learning_rate": 6.604102564102565e-06, + "loss": 0.004, + "step": 7125 + }, + { + "epoch": 6.046511627906977, + "grad_norm": 0.0608515739440918, + "learning_rate": 6.591282051282052e-06, + "loss": 0.0027, + "step": 7150 + }, + { + "epoch": 6.067653276955602, + "grad_norm": 0.09016130864620209, + "learning_rate": 6.578461538461539e-06, + "loss": 0.0046, + "step": 7175 + }, + { + "epoch": 6.088794926004228, + "grad_norm": 1.249212384223938, + "learning_rate": 6.565641025641026e-06, + "loss": 0.0044, + "step": 7200 + }, + { + "epoch": 6.1099365750528545, + "grad_norm": 0.2739676237106323, + "learning_rate": 6.552820512820513e-06, + "loss": 0.0036, + "step": 7225 + }, + { + "epoch": 6.13107822410148, + "grad_norm": 1.4266902208328247, + "learning_rate": 6.540000000000001e-06, + "loss": 0.0056, + "step": 7250 + }, + { + "epoch": 6.152219873150106, + "grad_norm": 1.1306654214859009, + "learning_rate": 6.527179487179488e-06, + "loss": 0.0023, + "step": 7275 + }, + { + "epoch": 6.173361522198731, + "grad_norm": 0.24447208642959595, + "learning_rate": 6.514358974358975e-06, + "loss": 0.0025, + "step": 7300 + }, + { + "epoch": 6.194503171247358, + "grad_norm": 1.2917050123214722, + "learning_rate": 6.501538461538462e-06, + "loss": 0.0033, + "step": 7325 + }, + { + "epoch": 6.215644820295983, + "grad_norm": 0.2974170744419098, + "learning_rate": 6.48871794871795e-06, + "loss": 0.0024, + "step": 7350 + }, + { + "epoch": 6.236786469344609, + "grad_norm": 1.0729445219039917, + "learning_rate": 6.475897435897437e-06, + "loss": 0.0021, + "step": 7375 + }, + { + "epoch": 6.2579281183932345, + "grad_norm": 0.10066499561071396, + "learning_rate": 6.463076923076924e-06, + "loss": 0.0019, + "step": 7400 + }, + { + "epoch": 6.27906976744186, + "grad_norm": 0.07153913378715515, + "learning_rate": 6.45025641025641e-06, + "loss": 0.0027, + "step": 7425 + }, + { + "epoch": 6.3002114164904865, + "grad_norm": 0.6892584562301636, + "learning_rate": 6.437435897435897e-06, + "loss": 0.0023, + "step": 7450 + }, + { + "epoch": 6.321353065539112, + "grad_norm": 1.0604571104049683, + "learning_rate": 6.424615384615385e-06, + "loss": 0.0039, + "step": 7475 + }, + { + "epoch": 6.342494714587738, + "grad_norm": 1.8776297569274902, + "learning_rate": 6.411794871794872e-06, + "loss": 0.002, + "step": 7500 + }, + { + "epoch": 6.363636363636363, + "grad_norm": 0.16421526670455933, + "learning_rate": 6.398974358974359e-06, + "loss": 0.0027, + "step": 7525 + }, + { + "epoch": 6.38477801268499, + "grad_norm": 0.1744580715894699, + "learning_rate": 6.386153846153846e-06, + "loss": 0.0044, + "step": 7550 + }, + { + "epoch": 6.405919661733615, + "grad_norm": 0.06431788951158524, + "learning_rate": 6.373333333333334e-06, + "loss": 0.0025, + "step": 7575 + }, + { + "epoch": 6.427061310782241, + "grad_norm": 0.1876668632030487, + "learning_rate": 6.360512820512821e-06, + "loss": 0.0032, + "step": 7600 + }, + { + "epoch": 6.4482029598308666, + "grad_norm": 0.1662682741880417, + "learning_rate": 6.347692307692308e-06, + "loss": 0.0027, + "step": 7625 + }, + { + "epoch": 6.469344608879492, + "grad_norm": 0.08980126678943634, + "learning_rate": 6.334871794871795e-06, + "loss": 0.0035, + "step": 7650 + }, + { + "epoch": 6.490486257928119, + "grad_norm": 0.13998441398143768, + "learning_rate": 6.322051282051283e-06, + "loss": 0.0015, + "step": 7675 + }, + { + "epoch": 6.511627906976744, + "grad_norm": 0.08528197556734085, + "learning_rate": 6.30923076923077e-06, + "loss": 0.0039, + "step": 7700 + }, + { + "epoch": 6.53276955602537, + "grad_norm": 0.07632896304130554, + "learning_rate": 6.296410256410257e-06, + "loss": 0.004, + "step": 7725 + }, + { + "epoch": 6.553911205073996, + "grad_norm": 1.8177125453948975, + "learning_rate": 6.283589743589744e-06, + "loss": 0.0045, + "step": 7750 + }, + { + "epoch": 6.575052854122622, + "grad_norm": 0.3318551778793335, + "learning_rate": 6.270769230769231e-06, + "loss": 0.003, + "step": 7775 + }, + { + "epoch": 6.5961945031712474, + "grad_norm": 0.18770574033260345, + "learning_rate": 6.257948717948719e-06, + "loss": 0.0026, + "step": 7800 + }, + { + "epoch": 6.617336152219873, + "grad_norm": 0.120570108294487, + "learning_rate": 6.2451282051282055e-06, + "loss": 0.0053, + "step": 7825 + }, + { + "epoch": 6.638477801268499, + "grad_norm": 8.554120063781738, + "learning_rate": 6.232307692307693e-06, + "loss": 0.0038, + "step": 7850 + }, + { + "epoch": 6.659619450317125, + "grad_norm": 0.31921058893203735, + "learning_rate": 6.21948717948718e-06, + "loss": 0.0031, + "step": 7875 + }, + { + "epoch": 6.680761099365751, + "grad_norm": 0.34284916520118713, + "learning_rate": 6.206666666666668e-06, + "loss": 0.0037, + "step": 7900 + }, + { + "epoch": 6.701902748414376, + "grad_norm": 1.419690489768982, + "learning_rate": 6.1938461538461545e-06, + "loss": 0.005, + "step": 7925 + }, + { + "epoch": 6.723044397463002, + "grad_norm": 0.1808280348777771, + "learning_rate": 6.181025641025642e-06, + "loss": 0.0036, + "step": 7950 + }, + { + "epoch": 6.7441860465116275, + "grad_norm": 1.7059262990951538, + "learning_rate": 6.168205128205129e-06, + "loss": 0.002, + "step": 7975 + }, + { + "epoch": 6.765327695560254, + "grad_norm": 1.1756020784378052, + "learning_rate": 6.155384615384617e-06, + "loss": 0.0029, + "step": 8000 + }, + { + "epoch": 6.765327695560254, + "eval_loss": 0.08152683824300766, + "eval_runtime": 415.4126, + "eval_samples_per_second": 8.765, + "eval_steps_per_second": 0.549, + "eval_wer": 0.06269432456262472, + "step": 8000 + }, + { + "epoch": 6.7864693446088795, + "grad_norm": 0.07017653435468674, + "learning_rate": 6.142564102564103e-06, + "loss": 0.002, + "step": 8025 + }, + { + "epoch": 6.807610993657505, + "grad_norm": 0.10346279293298721, + "learning_rate": 6.1297435897435895e-06, + "loss": 0.0039, + "step": 8050 + }, + { + "epoch": 6.828752642706131, + "grad_norm": 1.0720205307006836, + "learning_rate": 6.116923076923077e-06, + "loss": 0.0028, + "step": 8075 + }, + { + "epoch": 6.849894291754757, + "grad_norm": 0.05146399512887001, + "learning_rate": 6.104102564102564e-06, + "loss": 0.0024, + "step": 8100 + }, + { + "epoch": 6.871035940803383, + "grad_norm": 0.5876528024673462, + "learning_rate": 6.091282051282052e-06, + "loss": 0.0034, + "step": 8125 + }, + { + "epoch": 6.892177589852008, + "grad_norm": 0.4982081353664398, + "learning_rate": 6.0784615384615384e-06, + "loss": 0.0032, + "step": 8150 + }, + { + "epoch": 6.913319238900634, + "grad_norm": 0.23112235963344574, + "learning_rate": 6.065641025641026e-06, + "loss": 0.0019, + "step": 8175 + }, + { + "epoch": 6.93446088794926, + "grad_norm": 0.19416943192481995, + "learning_rate": 6.052820512820513e-06, + "loss": 0.0045, + "step": 8200 + }, + { + "epoch": 6.955602536997886, + "grad_norm": 1.9558351039886475, + "learning_rate": 6.040000000000001e-06, + "loss": 0.0028, + "step": 8225 + }, + { + "epoch": 6.976744186046512, + "grad_norm": 0.15180020034313202, + "learning_rate": 6.0271794871794874e-06, + "loss": 0.0026, + "step": 8250 + }, + { + "epoch": 6.997885835095137, + "grad_norm": 0.434987872838974, + "learning_rate": 6.014358974358975e-06, + "loss": 0.0039, + "step": 8275 + }, + { + "epoch": 7.019027484143764, + "grad_norm": 0.07390391826629639, + "learning_rate": 6.001538461538462e-06, + "loss": 0.0026, + "step": 8300 + }, + { + "epoch": 7.040169133192389, + "grad_norm": 0.3644394874572754, + "learning_rate": 5.98871794871795e-06, + "loss": 0.0022, + "step": 8325 + }, + { + "epoch": 7.061310782241015, + "grad_norm": 0.0214530136436224, + "learning_rate": 5.9758974358974364e-06, + "loss": 0.0017, + "step": 8350 + }, + { + "epoch": 7.08245243128964, + "grad_norm": 0.05759245902299881, + "learning_rate": 5.963076923076923e-06, + "loss": 0.0011, + "step": 8375 + }, + { + "epoch": 7.103594080338266, + "grad_norm": 0.40693360567092896, + "learning_rate": 5.950256410256411e-06, + "loss": 0.0015, + "step": 8400 + }, + { + "epoch": 7.1247357293868925, + "grad_norm": 0.03377969190478325, + "learning_rate": 5.937435897435898e-06, + "loss": 0.0017, + "step": 8425 + }, + { + "epoch": 7.145877378435518, + "grad_norm": 0.07641027122735977, + "learning_rate": 5.9246153846153854e-06, + "loss": 0.0017, + "step": 8450 + }, + { + "epoch": 7.167019027484144, + "grad_norm": 0.03931877389550209, + "learning_rate": 5.911794871794872e-06, + "loss": 0.0042, + "step": 8475 + }, + { + "epoch": 7.188160676532769, + "grad_norm": 0.8600488901138306, + "learning_rate": 5.89897435897436e-06, + "loss": 0.0012, + "step": 8500 + }, + { + "epoch": 7.209302325581396, + "grad_norm": 0.256000816822052, + "learning_rate": 5.886153846153847e-06, + "loss": 0.0023, + "step": 8525 + }, + { + "epoch": 7.230443974630021, + "grad_norm": 0.22403153777122498, + "learning_rate": 5.873333333333334e-06, + "loss": 0.0014, + "step": 8550 + }, + { + "epoch": 7.251585623678647, + "grad_norm": 0.7728956341743469, + "learning_rate": 5.860512820512821e-06, + "loss": 0.003, + "step": 8575 + }, + { + "epoch": 7.2727272727272725, + "grad_norm": 1.0295345783233643, + "learning_rate": 5.848205128205129e-06, + "loss": 0.0045, + "step": 8600 + }, + { + "epoch": 7.293868921775898, + "grad_norm": 0.1343035101890564, + "learning_rate": 5.835384615384616e-06, + "loss": 0.0021, + "step": 8625 + }, + { + "epoch": 7.3150105708245245, + "grad_norm": 0.027379866689443588, + "learning_rate": 5.822564102564103e-06, + "loss": 0.0016, + "step": 8650 + }, + { + "epoch": 7.33615221987315, + "grad_norm": 0.0675494521856308, + "learning_rate": 5.809743589743591e-06, + "loss": 0.0011, + "step": 8675 + }, + { + "epoch": 7.357293868921776, + "grad_norm": 0.05748479440808296, + "learning_rate": 5.796923076923078e-06, + "loss": 0.0011, + "step": 8700 + }, + { + "epoch": 7.378435517970401, + "grad_norm": 0.07236622273921967, + "learning_rate": 5.784102564102565e-06, + "loss": 0.0012, + "step": 8725 + }, + { + "epoch": 7.399577167019028, + "grad_norm": 0.05686989426612854, + "learning_rate": 5.771282051282052e-06, + "loss": 0.0035, + "step": 8750 + }, + { + "epoch": 7.420718816067653, + "grad_norm": 2.922973871231079, + "learning_rate": 5.75846153846154e-06, + "loss": 0.0042, + "step": 8775 + }, + { + "epoch": 7.441860465116279, + "grad_norm": 0.06991543620824814, + "learning_rate": 5.745641025641027e-06, + "loss": 0.0021, + "step": 8800 + }, + { + "epoch": 7.4630021141649046, + "grad_norm": 0.3402734100818634, + "learning_rate": 5.732820512820513e-06, + "loss": 0.0028, + "step": 8825 + }, + { + "epoch": 7.484143763213531, + "grad_norm": 0.06663922965526581, + "learning_rate": 5.72e-06, + "loss": 0.0024, + "step": 8850 + }, + { + "epoch": 7.505285412262157, + "grad_norm": 0.0772562250494957, + "learning_rate": 5.707179487179487e-06, + "loss": 0.0026, + "step": 8875 + }, + { + "epoch": 7.526427061310782, + "grad_norm": 0.13530157506465912, + "learning_rate": 5.694358974358975e-06, + "loss": 0.002, + "step": 8900 + }, + { + "epoch": 7.547568710359408, + "grad_norm": 0.04225620627403259, + "learning_rate": 5.681538461538462e-06, + "loss": 0.003, + "step": 8925 + }, + { + "epoch": 7.568710359408033, + "grad_norm": 0.06277480721473694, + "learning_rate": 5.668717948717949e-06, + "loss": 0.0041, + "step": 8950 + }, + { + "epoch": 7.58985200845666, + "grad_norm": 0.06269287317991257, + "learning_rate": 5.655897435897436e-06, + "loss": 0.0023, + "step": 8975 + }, + { + "epoch": 7.6109936575052854, + "grad_norm": 0.06426391750574112, + "learning_rate": 5.643076923076923e-06, + "loss": 0.0014, + "step": 9000 + }, + { + "epoch": 7.6109936575052854, + "eval_loss": 0.08534783869981766, + "eval_runtime": 415.5854, + "eval_samples_per_second": 8.761, + "eval_steps_per_second": 0.549, + "eval_wer": 0.06274073042832615, + "step": 9000 + }, + { + "epoch": 7.632135306553911, + "grad_norm": 0.04878684878349304, + "learning_rate": 5.630256410256411e-06, + "loss": 0.0013, + "step": 9025 + }, + { + "epoch": 7.653276955602537, + "grad_norm": 0.073283351957798, + "learning_rate": 5.6174358974358974e-06, + "loss": 0.0015, + "step": 9050 + }, + { + "epoch": 7.674418604651163, + "grad_norm": 0.04749315604567528, + "learning_rate": 5.604615384615385e-06, + "loss": 0.0015, + "step": 9075 + }, + { + "epoch": 7.695560253699789, + "grad_norm": 0.13554809987545013, + "learning_rate": 5.591794871794872e-06, + "loss": 0.0014, + "step": 9100 + }, + { + "epoch": 7.716701902748414, + "grad_norm": 0.03620721027255058, + "learning_rate": 5.57897435897436e-06, + "loss": 0.0023, + "step": 9125 + }, + { + "epoch": 7.73784355179704, + "grad_norm": 0.9968867301940918, + "learning_rate": 5.566153846153846e-06, + "loss": 0.0026, + "step": 9150 + }, + { + "epoch": 7.758985200845666, + "grad_norm": 0.055259935557842255, + "learning_rate": 5.553333333333334e-06, + "loss": 0.0016, + "step": 9175 + }, + { + "epoch": 7.780126849894292, + "grad_norm": 1.2477086782455444, + "learning_rate": 5.540512820512821e-06, + "loss": 0.0026, + "step": 9200 + }, + { + "epoch": 7.8012684989429175, + "grad_norm": 0.42810365557670593, + "learning_rate": 5.527692307692309e-06, + "loss": 0.0022, + "step": 9225 + }, + { + "epoch": 7.822410147991543, + "grad_norm": 0.7914177775382996, + "learning_rate": 5.514871794871795e-06, + "loss": 0.0026, + "step": 9250 + }, + { + "epoch": 7.843551797040169, + "grad_norm": 0.7927113771438599, + "learning_rate": 5.502051282051283e-06, + "loss": 0.0026, + "step": 9275 + }, + { + "epoch": 7.864693446088795, + "grad_norm": 0.09316157549619675, + "learning_rate": 5.48923076923077e-06, + "loss": 0.0032, + "step": 9300 + }, + { + "epoch": 7.885835095137421, + "grad_norm": 0.21055510640144348, + "learning_rate": 5.476410256410258e-06, + "loss": 0.0023, + "step": 9325 + }, + { + "epoch": 7.906976744186046, + "grad_norm": 1.5093400478363037, + "learning_rate": 5.463589743589744e-06, + "loss": 0.0028, + "step": 9350 + }, + { + "epoch": 7.928118393234672, + "grad_norm": 0.9146241545677185, + "learning_rate": 5.450769230769232e-06, + "loss": 0.0015, + "step": 9375 + }, + { + "epoch": 7.949260042283298, + "grad_norm": 0.6534460186958313, + "learning_rate": 5.437948717948719e-06, + "loss": 0.0028, + "step": 9400 + }, + { + "epoch": 7.970401691331924, + "grad_norm": 0.056375931948423386, + "learning_rate": 5.425128205128205e-06, + "loss": 0.0015, + "step": 9425 + }, + { + "epoch": 7.99154334038055, + "grad_norm": 0.08614702522754669, + "learning_rate": 5.4123076923076925e-06, + "loss": 0.0028, + "step": 9450 + }, + { + "epoch": 8.012684989429175, + "grad_norm": 0.04644578695297241, + "learning_rate": 5.399487179487179e-06, + "loss": 0.0033, + "step": 9475 + }, + { + "epoch": 8.033826638477802, + "grad_norm": 0.12249679118394852, + "learning_rate": 5.386666666666667e-06, + "loss": 0.002, + "step": 9500 + }, + { + "epoch": 8.054968287526426, + "grad_norm": 0.18292202055454254, + "learning_rate": 5.373846153846154e-06, + "loss": 0.0009, + "step": 9525 + }, + { + "epoch": 8.076109936575053, + "grad_norm": 0.94189453125, + "learning_rate": 5.3610256410256415e-06, + "loss": 0.0029, + "step": 9550 + }, + { + "epoch": 8.09725158562368, + "grad_norm": 0.05061868205666542, + "learning_rate": 5.348205128205128e-06, + "loss": 0.0011, + "step": 9575 + }, + { + "epoch": 8.118393234672304, + "grad_norm": 0.6493793725967407, + "learning_rate": 5.335384615384615e-06, + "loss": 0.0006, + "step": 9600 + }, + { + "epoch": 8.13953488372093, + "grad_norm": 0.5912912487983704, + "learning_rate": 5.322564102564103e-06, + "loss": 0.0015, + "step": 9625 + }, + { + "epoch": 8.160676532769555, + "grad_norm": 0.036978259682655334, + "learning_rate": 5.30974358974359e-06, + "loss": 0.0021, + "step": 9650 + }, + { + "epoch": 8.181818181818182, + "grad_norm": 0.13611486554145813, + "learning_rate": 5.296923076923077e-06, + "loss": 0.002, + "step": 9675 + }, + { + "epoch": 8.202959830866808, + "grad_norm": 1.0610445737838745, + "learning_rate": 5.284102564102564e-06, + "loss": 0.0016, + "step": 9700 + }, + { + "epoch": 8.224101479915433, + "grad_norm": 0.024419546127319336, + "learning_rate": 5.271282051282052e-06, + "loss": 0.004, + "step": 9725 + }, + { + "epoch": 8.24524312896406, + "grad_norm": 0.07428953051567078, + "learning_rate": 5.258461538461539e-06, + "loss": 0.0014, + "step": 9750 + }, + { + "epoch": 8.266384778012686, + "grad_norm": 1.252059817314148, + "learning_rate": 5.245641025641026e-06, + "loss": 0.0018, + "step": 9775 + }, + { + "epoch": 8.28752642706131, + "grad_norm": 0.14112520217895508, + "learning_rate": 5.232820512820513e-06, + "loss": 0.0015, + "step": 9800 + }, + { + "epoch": 8.308668076109937, + "grad_norm": 0.05485629662871361, + "learning_rate": 5.220000000000001e-06, + "loss": 0.0017, + "step": 9825 + }, + { + "epoch": 8.329809725158562, + "grad_norm": 0.9850168228149414, + "learning_rate": 5.207179487179488e-06, + "loss": 0.0034, + "step": 9850 + }, + { + "epoch": 8.350951374207188, + "grad_norm": 0.2707412838935852, + "learning_rate": 5.194358974358975e-06, + "loss": 0.0024, + "step": 9875 + }, + { + "epoch": 8.372093023255815, + "grad_norm": 0.7773647308349609, + "learning_rate": 5.181538461538462e-06, + "loss": 0.0023, + "step": 9900 + }, + { + "epoch": 8.39323467230444, + "grad_norm": 0.2679407000541687, + "learning_rate": 5.16871794871795e-06, + "loss": 0.0023, + "step": 9925 + }, + { + "epoch": 8.414376321353066, + "grad_norm": 0.1327216774225235, + "learning_rate": 5.155897435897437e-06, + "loss": 0.0027, + "step": 9950 + }, + { + "epoch": 8.43551797040169, + "grad_norm": 0.5626333951950073, + "learning_rate": 5.143076923076924e-06, + "loss": 0.0015, + "step": 9975 + }, + { + "epoch": 8.456659619450317, + "grad_norm": 0.06341422349214554, + "learning_rate": 5.130256410256411e-06, + "loss": 0.0013, + "step": 10000 + }, + { + "epoch": 8.456659619450317, + "eval_loss": 0.08607128262519836, + "eval_runtime": 417.0751, + "eval_samples_per_second": 8.73, + "eval_steps_per_second": 0.547, + "eval_wer": 0.06413290639936887, + "step": 10000 + }, + { + "epoch": 8.477801268498943, + "grad_norm": 0.04358547925949097, + "learning_rate": 5.117435897435897e-06, + "loss": 0.0018, + "step": 10025 + }, + { + "epoch": 8.498942917547568, + "grad_norm": 0.11723045259714127, + "learning_rate": 5.104615384615385e-06, + "loss": 0.0034, + "step": 10050 + }, + { + "epoch": 8.520084566596195, + "grad_norm": 0.034698598086833954, + "learning_rate": 5.091794871794872e-06, + "loss": 0.001, + "step": 10075 + }, + { + "epoch": 8.541226215644821, + "grad_norm": 0.2308579683303833, + "learning_rate": 5.078974358974359e-06, + "loss": 0.0015, + "step": 10100 + }, + { + "epoch": 8.562367864693446, + "grad_norm": 0.03924379497766495, + "learning_rate": 5.066153846153846e-06, + "loss": 0.0014, + "step": 10125 + }, + { + "epoch": 8.583509513742072, + "grad_norm": 0.018929792568087578, + "learning_rate": 5.053333333333334e-06, + "loss": 0.0014, + "step": 10150 + }, + { + "epoch": 8.604651162790697, + "grad_norm": 0.06054197624325752, + "learning_rate": 5.040512820512821e-06, + "loss": 0.0032, + "step": 10175 + }, + { + "epoch": 8.625792811839323, + "grad_norm": 0.04145154356956482, + "learning_rate": 5.027692307692308e-06, + "loss": 0.002, + "step": 10200 + }, + { + "epoch": 8.64693446088795, + "grad_norm": 0.05452277883887291, + "learning_rate": 5.014871794871795e-06, + "loss": 0.0014, + "step": 10225 + }, + { + "epoch": 8.668076109936575, + "grad_norm": 0.04759210720658302, + "learning_rate": 5.002051282051282e-06, + "loss": 0.002, + "step": 10250 + }, + { + "epoch": 8.689217758985201, + "grad_norm": 0.029456498101353645, + "learning_rate": 4.98923076923077e-06, + "loss": 0.001, + "step": 10275 + }, + { + "epoch": 8.710359408033828, + "grad_norm": 0.062097664922475815, + "learning_rate": 4.9764102564102564e-06, + "loss": 0.001, + "step": 10300 + }, + { + "epoch": 8.731501057082452, + "grad_norm": 0.07016895711421967, + "learning_rate": 4.963589743589744e-06, + "loss": 0.0011, + "step": 10325 + }, + { + "epoch": 8.752642706131079, + "grad_norm": 2.872891426086426, + "learning_rate": 4.950769230769231e-06, + "loss": 0.0018, + "step": 10350 + }, + { + "epoch": 8.773784355179703, + "grad_norm": 0.20757004618644714, + "learning_rate": 4.937948717948719e-06, + "loss": 0.0016, + "step": 10375 + }, + { + "epoch": 8.79492600422833, + "grad_norm": 0.10146087408065796, + "learning_rate": 4.9251282051282054e-06, + "loss": 0.0032, + "step": 10400 + }, + { + "epoch": 8.816067653276956, + "grad_norm": 0.07840722799301147, + "learning_rate": 4.912307692307693e-06, + "loss": 0.001, + "step": 10425 + }, + { + "epoch": 8.837209302325581, + "grad_norm": 0.0831177830696106, + "learning_rate": 4.89948717948718e-06, + "loss": 0.0012, + "step": 10450 + }, + { + "epoch": 8.858350951374208, + "grad_norm": 0.6827990412712097, + "learning_rate": 4.886666666666668e-06, + "loss": 0.0016, + "step": 10475 + }, + { + "epoch": 8.879492600422832, + "grad_norm": 0.25397488474845886, + "learning_rate": 4.873846153846154e-06, + "loss": 0.0017, + "step": 10500 + }, + { + "epoch": 8.900634249471459, + "grad_norm": 0.07114572823047638, + "learning_rate": 4.861025641025641e-06, + "loss": 0.0009, + "step": 10525 + }, + { + "epoch": 8.921775898520085, + "grad_norm": 0.05897723138332367, + "learning_rate": 4.848205128205128e-06, + "loss": 0.0016, + "step": 10550 + }, + { + "epoch": 8.94291754756871, + "grad_norm": 0.048447199165821075, + "learning_rate": 4.835384615384616e-06, + "loss": 0.0015, + "step": 10575 + }, + { + "epoch": 8.964059196617336, + "grad_norm": 0.04842350259423256, + "learning_rate": 4.8225641025641026e-06, + "loss": 0.0016, + "step": 10600 + }, + { + "epoch": 8.985200845665961, + "grad_norm": 0.06469196826219559, + "learning_rate": 4.80974358974359e-06, + "loss": 0.0011, + "step": 10625 + }, + { + "epoch": 9.006342494714588, + "grad_norm": 0.021318404003977776, + "learning_rate": 4.796923076923077e-06, + "loss": 0.0038, + "step": 10650 + }, + { + "epoch": 9.027484143763214, + "grad_norm": 0.03229638934135437, + "learning_rate": 4.784102564102565e-06, + "loss": 0.0014, + "step": 10675 + }, + { + "epoch": 9.048625792811839, + "grad_norm": 0.08799826353788376, + "learning_rate": 4.7712820512820516e-06, + "loss": 0.0026, + "step": 10700 + }, + { + "epoch": 9.069767441860465, + "grad_norm": 0.11259903013706207, + "learning_rate": 4.758461538461539e-06, + "loss": 0.0015, + "step": 10725 + }, + { + "epoch": 9.090909090909092, + "grad_norm": 0.14656835794448853, + "learning_rate": 4.745641025641026e-06, + "loss": 0.0019, + "step": 10750 + }, + { + "epoch": 9.112050739957716, + "grad_norm": 0.5436356663703918, + "learning_rate": 4.732820512820514e-06, + "loss": 0.001, + "step": 10775 + }, + { + "epoch": 9.133192389006343, + "grad_norm": 0.02085413970053196, + "learning_rate": 4.7200000000000005e-06, + "loss": 0.001, + "step": 10800 + }, + { + "epoch": 9.154334038054968, + "grad_norm": 0.04881500452756882, + "learning_rate": 4.707179487179487e-06, + "loss": 0.001, + "step": 10825 + }, + { + "epoch": 9.175475687103594, + "grad_norm": 0.04903466999530792, + "learning_rate": 4.694358974358974e-06, + "loss": 0.0008, + "step": 10850 + }, + { + "epoch": 9.19661733615222, + "grad_norm": 0.020907558500766754, + "learning_rate": 4.681538461538462e-06, + "loss": 0.0014, + "step": 10875 + }, + { + "epoch": 9.217758985200845, + "grad_norm": 0.13889507949352264, + "learning_rate": 4.668717948717949e-06, + "loss": 0.0014, + "step": 10900 + }, + { + "epoch": 9.238900634249472, + "grad_norm": 0.050613515079021454, + "learning_rate": 4.655897435897436e-06, + "loss": 0.0032, + "step": 10925 + }, + { + "epoch": 9.260042283298096, + "grad_norm": 1.744171380996704, + "learning_rate": 4.643076923076923e-06, + "loss": 0.0011, + "step": 10950 + }, + { + "epoch": 9.281183932346723, + "grad_norm": 0.013857314363121986, + "learning_rate": 4.630256410256411e-06, + "loss": 0.0011, + "step": 10975 + }, + { + "epoch": 9.30232558139535, + "grad_norm": 0.03641614317893982, + "learning_rate": 4.617435897435898e-06, + "loss": 0.0005, + "step": 11000 + }, + { + "epoch": 9.30232558139535, + "eval_loss": 0.08570627868175507, + "eval_runtime": 422.6796, + "eval_samples_per_second": 8.614, + "eval_steps_per_second": 0.539, + "eval_wer": 0.06334400668244466, + "step": 11000 + }, + { + "epoch": 9.323467230443974, + "grad_norm": 0.32879745960235596, + "learning_rate": 4.604615384615385e-06, + "loss": 0.0007, + "step": 11025 + }, + { + "epoch": 9.3446088794926, + "grad_norm": 0.21095559000968933, + "learning_rate": 4.591794871794872e-06, + "loss": 0.0014, + "step": 11050 + }, + { + "epoch": 9.365750528541227, + "grad_norm": 0.01756209135055542, + "learning_rate": 4.57897435897436e-06, + "loss": 0.0025, + "step": 11075 + }, + { + "epoch": 9.386892177589852, + "grad_norm": 0.04053978621959686, + "learning_rate": 4.566153846153847e-06, + "loss": 0.0005, + "step": 11100 + }, + { + "epoch": 9.408033826638478, + "grad_norm": 1.196875810623169, + "learning_rate": 4.5533333333333335e-06, + "loss": 0.0015, + "step": 11125 + }, + { + "epoch": 9.429175475687103, + "grad_norm": 0.027264421805739403, + "learning_rate": 4.54051282051282e-06, + "loss": 0.0005, + "step": 11150 + }, + { + "epoch": 9.45031712473573, + "grad_norm": 0.021927664056420326, + "learning_rate": 4.527692307692308e-06, + "loss": 0.0013, + "step": 11175 + }, + { + "epoch": 9.471458773784356, + "grad_norm": 0.04593001306056976, + "learning_rate": 4.514871794871795e-06, + "loss": 0.0013, + "step": 11200 + }, + { + "epoch": 9.49260042283298, + "grad_norm": 0.8032304644584656, + "learning_rate": 4.5020512820512825e-06, + "loss": 0.0015, + "step": 11225 + }, + { + "epoch": 9.513742071881607, + "grad_norm": 0.6102417707443237, + "learning_rate": 4.489230769230769e-06, + "loss": 0.0009, + "step": 11250 + }, + { + "epoch": 9.534883720930232, + "grad_norm": 0.029139434918761253, + "learning_rate": 4.476410256410257e-06, + "loss": 0.0011, + "step": 11275 + }, + { + "epoch": 9.556025369978858, + "grad_norm": 0.01795080117881298, + "learning_rate": 4.463589743589744e-06, + "loss": 0.0008, + "step": 11300 + }, + { + "epoch": 9.577167019027485, + "grad_norm": 0.029438558965921402, + "learning_rate": 4.4507692307692315e-06, + "loss": 0.0009, + "step": 11325 + }, + { + "epoch": 9.59830866807611, + "grad_norm": 0.03739332780241966, + "learning_rate": 4.437948717948718e-06, + "loss": 0.0006, + "step": 11350 + }, + { + "epoch": 9.619450317124736, + "grad_norm": 0.11360177397727966, + "learning_rate": 4.425128205128206e-06, + "loss": 0.001, + "step": 11375 + }, + { + "epoch": 9.640591966173362, + "grad_norm": 0.03182990849018097, + "learning_rate": 4.412307692307693e-06, + "loss": 0.0008, + "step": 11400 + }, + { + "epoch": 9.661733615221987, + "grad_norm": 0.023038147017359734, + "learning_rate": 4.39948717948718e-06, + "loss": 0.001, + "step": 11425 + }, + { + "epoch": 9.682875264270614, + "grad_norm": 0.03209986910223961, + "learning_rate": 4.3866666666666665e-06, + "loss": 0.0007, + "step": 11450 + }, + { + "epoch": 9.704016913319238, + "grad_norm": 0.028583193197846413, + "learning_rate": 4.373846153846154e-06, + "loss": 0.0004, + "step": 11475 + }, + { + "epoch": 9.725158562367865, + "grad_norm": 0.023485548794269562, + "learning_rate": 4.361025641025641e-06, + "loss": 0.0007, + "step": 11500 + }, + { + "epoch": 9.746300211416491, + "grad_norm": 0.3672647476196289, + "learning_rate": 4.348205128205129e-06, + "loss": 0.0008, + "step": 11525 + }, + { + "epoch": 9.767441860465116, + "grad_norm": 0.8596897125244141, + "learning_rate": 4.3353846153846154e-06, + "loss": 0.0012, + "step": 11550 + }, + { + "epoch": 9.788583509513742, + "grad_norm": 1.4410885572433472, + "learning_rate": 4.322564102564103e-06, + "loss": 0.001, + "step": 11575 + }, + { + "epoch": 9.809725158562367, + "grad_norm": 0.023211924359202385, + "learning_rate": 4.30974358974359e-06, + "loss": 0.0005, + "step": 11600 + }, + { + "epoch": 9.830866807610994, + "grad_norm": 0.032794274389743805, + "learning_rate": 4.296923076923078e-06, + "loss": 0.0008, + "step": 11625 + }, + { + "epoch": 9.85200845665962, + "grad_norm": 0.037845127284526825, + "learning_rate": 4.2841025641025644e-06, + "loss": 0.0012, + "step": 11650 + }, + { + "epoch": 9.873150105708245, + "grad_norm": 0.0936095118522644, + "learning_rate": 4.271282051282052e-06, + "loss": 0.0012, + "step": 11675 + }, + { + "epoch": 9.894291754756871, + "grad_norm": 0.06250233203172684, + "learning_rate": 4.258461538461539e-06, + "loss": 0.0005, + "step": 11700 + }, + { + "epoch": 9.915433403805498, + "grad_norm": 0.040198661386966705, + "learning_rate": 4.245641025641026e-06, + "loss": 0.0009, + "step": 11725 + }, + { + "epoch": 9.936575052854122, + "grad_norm": 0.02659662999212742, + "learning_rate": 4.232820512820513e-06, + "loss": 0.0004, + "step": 11750 + }, + { + "epoch": 9.957716701902749, + "grad_norm": 0.021418403834104538, + "learning_rate": 4.22e-06, + "loss": 0.0004, + "step": 11775 + }, + { + "epoch": 9.978858350951374, + "grad_norm": 0.2249915599822998, + "learning_rate": 4.207179487179487e-06, + "loss": 0.0005, + "step": 11800 + }, + { + "epoch": 10.0, + "grad_norm": 0.8315058350563049, + "learning_rate": 4.194358974358975e-06, + "loss": 0.0011, + "step": 11825 + }, + { + "epoch": 10.021141649048626, + "grad_norm": 0.1930394470691681, + "learning_rate": 4.1815384615384616e-06, + "loss": 0.001, + "step": 11850 + }, + { + "epoch": 10.042283298097251, + "grad_norm": 0.01425615418702364, + "learning_rate": 4.168717948717949e-06, + "loss": 0.0012, + "step": 11875 + }, + { + "epoch": 10.063424947145878, + "grad_norm": 0.023017780855298042, + "learning_rate": 4.155897435897436e-06, + "loss": 0.0011, + "step": 11900 + }, + { + "epoch": 10.084566596194502, + "grad_norm": 0.029264431446790695, + "learning_rate": 4.143076923076924e-06, + "loss": 0.0011, + "step": 11925 + }, + { + "epoch": 10.105708245243129, + "grad_norm": 0.04874871298670769, + "learning_rate": 4.1302564102564106e-06, + "loss": 0.0005, + "step": 11950 + }, + { + "epoch": 10.126849894291755, + "grad_norm": 0.745508074760437, + "learning_rate": 4.117435897435898e-06, + "loss": 0.0015, + "step": 11975 + }, + { + "epoch": 10.14799154334038, + "grad_norm": 0.03195907920598984, + "learning_rate": 4.104615384615385e-06, + "loss": 0.0005, + "step": 12000 + }, + { + "epoch": 10.14799154334038, + "eval_loss": 0.08556105196475983, + "eval_runtime": 420.9801, + "eval_samples_per_second": 8.649, + "eval_steps_per_second": 0.542, + "eval_wer": 0.06195183071140192, + "step": 12000 + }, + { + "epoch": 10.169133192389006, + "grad_norm": 0.029029415920376778, + "learning_rate": 4.091794871794872e-06, + "loss": 0.0004, + "step": 12025 + }, + { + "epoch": 10.190274841437633, + "grad_norm": 0.029037466272711754, + "learning_rate": 4.078974358974359e-06, + "loss": 0.0004, + "step": 12050 + }, + { + "epoch": 10.211416490486258, + "grad_norm": 0.1034269854426384, + "learning_rate": 4.066153846153846e-06, + "loss": 0.0009, + "step": 12075 + }, + { + "epoch": 10.232558139534884, + "grad_norm": 0.025130081921815872, + "learning_rate": 4.053333333333333e-06, + "loss": 0.0009, + "step": 12100 + }, + { + "epoch": 10.253699788583509, + "grad_norm": 0.05509277433156967, + "learning_rate": 4.040512820512821e-06, + "loss": 0.0003, + "step": 12125 + }, + { + "epoch": 10.274841437632135, + "grad_norm": 0.022796448320150375, + "learning_rate": 4.027692307692308e-06, + "loss": 0.0004, + "step": 12150 + }, + { + "epoch": 10.295983086680762, + "grad_norm": 0.03756018355488777, + "learning_rate": 4.014871794871795e-06, + "loss": 0.0008, + "step": 12175 + }, + { + "epoch": 10.317124735729386, + "grad_norm": 0.6810747385025024, + "learning_rate": 4.002051282051282e-06, + "loss": 0.0008, + "step": 12200 + }, + { + "epoch": 10.338266384778013, + "grad_norm": 0.2094302922487259, + "learning_rate": 3.98923076923077e-06, + "loss": 0.0016, + "step": 12225 + }, + { + "epoch": 10.359408033826638, + "grad_norm": 0.02640734240412712, + "learning_rate": 3.976410256410257e-06, + "loss": 0.0007, + "step": 12250 + }, + { + "epoch": 10.380549682875264, + "grad_norm": 0.015162879601120949, + "learning_rate": 3.963589743589744e-06, + "loss": 0.0006, + "step": 12275 + }, + { + "epoch": 10.40169133192389, + "grad_norm": 0.023550117388367653, + "learning_rate": 3.950769230769231e-06, + "loss": 0.0007, + "step": 12300 + }, + { + "epoch": 10.422832980972515, + "grad_norm": 1.2513926029205322, + "learning_rate": 3.937948717948718e-06, + "loss": 0.0012, + "step": 12325 + }, + { + "epoch": 10.443974630021142, + "grad_norm": 1.0258477926254272, + "learning_rate": 3.925128205128205e-06, + "loss": 0.0012, + "step": 12350 + }, + { + "epoch": 10.465116279069768, + "grad_norm": 1.8391358852386475, + "learning_rate": 3.9123076923076925e-06, + "loss": 0.0008, + "step": 12375 + }, + { + "epoch": 10.486257928118393, + "grad_norm": 0.02079303376376629, + "learning_rate": 3.899487179487179e-06, + "loss": 0.001, + "step": 12400 + }, + { + "epoch": 10.50739957716702, + "grad_norm": 0.3406016230583191, + "learning_rate": 3.886666666666667e-06, + "loss": 0.0013, + "step": 12425 + }, + { + "epoch": 10.528541226215644, + "grad_norm": 0.08087711036205292, + "learning_rate": 3.873846153846154e-06, + "loss": 0.0018, + "step": 12450 + }, + { + "epoch": 10.54968287526427, + "grad_norm": 0.03195007145404816, + "learning_rate": 3.8610256410256415e-06, + "loss": 0.0007, + "step": 12475 + }, + { + "epoch": 10.570824524312897, + "grad_norm": 0.0327339693903923, + "learning_rate": 3.848205128205128e-06, + "loss": 0.0005, + "step": 12500 + }, + { + "epoch": 10.591966173361522, + "grad_norm": 0.050105396658182144, + "learning_rate": 3.835384615384616e-06, + "loss": 0.0005, + "step": 12525 + }, + { + "epoch": 10.613107822410148, + "grad_norm": 0.04684900864958763, + "learning_rate": 3.822564102564103e-06, + "loss": 0.0007, + "step": 12550 + }, + { + "epoch": 10.634249471458773, + "grad_norm": 0.01622653938829899, + "learning_rate": 3.80974358974359e-06, + "loss": 0.0011, + "step": 12575 + }, + { + "epoch": 10.6553911205074, + "grad_norm": 0.03434791415929794, + "learning_rate": 3.7969230769230773e-06, + "loss": 0.0007, + "step": 12600 + }, + { + "epoch": 10.676532769556026, + "grad_norm": 0.016083180904388428, + "learning_rate": 3.784102564102564e-06, + "loss": 0.0005, + "step": 12625 + }, + { + "epoch": 10.69767441860465, + "grad_norm": 0.019794505089521408, + "learning_rate": 3.7712820512820514e-06, + "loss": 0.0005, + "step": 12650 + }, + { + "epoch": 10.718816067653277, + "grad_norm": 0.019432729110121727, + "learning_rate": 3.7584615384615386e-06, + "loss": 0.0006, + "step": 12675 + }, + { + "epoch": 10.739957716701904, + "grad_norm": 0.021112432703375816, + "learning_rate": 3.745641025641026e-06, + "loss": 0.0007, + "step": 12700 + }, + { + "epoch": 10.761099365750528, + "grad_norm": 0.02420271746814251, + "learning_rate": 3.732820512820513e-06, + "loss": 0.0012, + "step": 12725 + }, + { + "epoch": 10.782241014799155, + "grad_norm": 0.044272564351558685, + "learning_rate": 3.7200000000000004e-06, + "loss": 0.0004, + "step": 12750 + }, + { + "epoch": 10.80338266384778, + "grad_norm": 0.6784161329269409, + "learning_rate": 3.7071794871794876e-06, + "loss": 0.0013, + "step": 12775 + }, + { + "epoch": 10.824524312896406, + "grad_norm": 0.110283263027668, + "learning_rate": 3.694358974358975e-06, + "loss": 0.0006, + "step": 12800 + }, + { + "epoch": 10.845665961945032, + "grad_norm": 0.026112297549843788, + "learning_rate": 3.681538461538462e-06, + "loss": 0.0011, + "step": 12825 + }, + { + "epoch": 10.866807610993657, + "grad_norm": 0.03150051832199097, + "learning_rate": 3.6687179487179494e-06, + "loss": 0.0004, + "step": 12850 + }, + { + "epoch": 10.887949260042284, + "grad_norm": 0.08290383219718933, + "learning_rate": 3.655897435897436e-06, + "loss": 0.0008, + "step": 12875 + }, + { + "epoch": 10.909090909090908, + "grad_norm": 0.018157586455345154, + "learning_rate": 3.6430769230769234e-06, + "loss": 0.0008, + "step": 12900 + }, + { + "epoch": 10.930232558139535, + "grad_norm": 0.04915007948875427, + "learning_rate": 3.6302564102564103e-06, + "loss": 0.0007, + "step": 12925 + }, + { + "epoch": 10.951374207188161, + "grad_norm": 0.023294847458600998, + "learning_rate": 3.6174358974358975e-06, + "loss": 0.0014, + "step": 12950 + }, + { + "epoch": 10.972515856236786, + "grad_norm": 0.013834277167916298, + "learning_rate": 3.6046153846153848e-06, + "loss": 0.0005, + "step": 12975 + }, + { + "epoch": 10.993657505285412, + "grad_norm": 0.018607119098305702, + "learning_rate": 3.591794871794872e-06, + "loss": 0.0007, + "step": 13000 + }, + { + "epoch": 10.993657505285412, + "eval_loss": 0.08658287674188614, + "eval_runtime": 413.9659, + "eval_samples_per_second": 8.795, + "eval_steps_per_second": 0.551, + "eval_wer": 0.060513248874657755, + "step": 13000 + }, + { + "epoch": 11.014799154334039, + "grad_norm": 0.034428298473358154, + "learning_rate": 3.5789743589743593e-06, + "loss": 0.0002, + "step": 13025 + }, + { + "epoch": 11.035940803382664, + "grad_norm": 0.024136802181601524, + "learning_rate": 3.5661538461538465e-06, + "loss": 0.001, + "step": 13050 + }, + { + "epoch": 11.05708245243129, + "grad_norm": 0.014073341153562069, + "learning_rate": 3.5533333333333338e-06, + "loss": 0.0002, + "step": 13075 + }, + { + "epoch": 11.078224101479915, + "grad_norm": 0.015313107520341873, + "learning_rate": 3.540512820512821e-06, + "loss": 0.0002, + "step": 13100 + }, + { + "epoch": 11.099365750528541, + "grad_norm": 0.014939980581402779, + "learning_rate": 3.5276923076923083e-06, + "loss": 0.0006, + "step": 13125 + }, + { + "epoch": 11.120507399577168, + "grad_norm": 0.054304271936416626, + "learning_rate": 3.5148717948717955e-06, + "loss": 0.0006, + "step": 13150 + }, + { + "epoch": 11.141649048625792, + "grad_norm": 0.01899879425764084, + "learning_rate": 3.5020512820512823e-06, + "loss": 0.0005, + "step": 13175 + }, + { + "epoch": 11.162790697674419, + "grad_norm": 0.008164430968463421, + "learning_rate": 3.4892307692307696e-06, + "loss": 0.0002, + "step": 13200 + }, + { + "epoch": 11.183932346723044, + "grad_norm": 0.12134576588869095, + "learning_rate": 3.4764102564102564e-06, + "loss": 0.0002, + "step": 13225 + }, + { + "epoch": 11.20507399577167, + "grad_norm": 0.019934862852096558, + "learning_rate": 3.4635897435897436e-06, + "loss": 0.0011, + "step": 13250 + }, + { + "epoch": 11.226215644820297, + "grad_norm": 0.01381596364080906, + "learning_rate": 3.450769230769231e-06, + "loss": 0.001, + "step": 13275 + }, + { + "epoch": 11.247357293868921, + "grad_norm": 0.1572095900774002, + "learning_rate": 3.437948717948718e-06, + "loss": 0.0003, + "step": 13300 + }, + { + "epoch": 11.268498942917548, + "grad_norm": 0.03256658464670181, + "learning_rate": 3.4251282051282054e-06, + "loss": 0.0002, + "step": 13325 + }, + { + "epoch": 11.289640591966174, + "grad_norm": 0.012259602546691895, + "learning_rate": 3.4123076923076926e-06, + "loss": 0.0005, + "step": 13350 + }, + { + "epoch": 11.310782241014799, + "grad_norm": 0.7077483534812927, + "learning_rate": 3.39948717948718e-06, + "loss": 0.002, + "step": 13375 + }, + { + "epoch": 11.331923890063425, + "grad_norm": 0.013183564879000187, + "learning_rate": 3.386666666666667e-06, + "loss": 0.0003, + "step": 13400 + }, + { + "epoch": 11.35306553911205, + "grad_norm": 0.015149732120335102, + "learning_rate": 3.3738461538461544e-06, + "loss": 0.0007, + "step": 13425 + }, + { + "epoch": 11.374207188160677, + "grad_norm": 1.2345651388168335, + "learning_rate": 3.3610256410256416e-06, + "loss": 0.0007, + "step": 13450 + }, + { + "epoch": 11.395348837209303, + "grad_norm": 0.015832841396331787, + "learning_rate": 3.3482051282051285e-06, + "loss": 0.0004, + "step": 13475 + }, + { + "epoch": 11.416490486257928, + "grad_norm": 0.02144164778292179, + "learning_rate": 3.3353846153846157e-06, + "loss": 0.0005, + "step": 13500 + }, + { + "epoch": 11.437632135306554, + "grad_norm": 0.02886483073234558, + "learning_rate": 3.322564102564103e-06, + "loss": 0.0002, + "step": 13525 + }, + { + "epoch": 11.458773784355179, + "grad_norm": 0.01857246272265911, + "learning_rate": 3.3097435897435898e-06, + "loss": 0.0007, + "step": 13550 + }, + { + "epoch": 11.479915433403805, + "grad_norm": 0.06685657054185867, + "learning_rate": 3.296923076923077e-06, + "loss": 0.0016, + "step": 13575 + }, + { + "epoch": 11.501057082452432, + "grad_norm": 0.058960702270269394, + "learning_rate": 3.2841025641025643e-06, + "loss": 0.0006, + "step": 13600 + }, + { + "epoch": 11.522198731501057, + "grad_norm": 0.016561295837163925, + "learning_rate": 3.2712820512820515e-06, + "loss": 0.0008, + "step": 13625 + }, + { + "epoch": 11.543340380549683, + "grad_norm": 0.019466349855065346, + "learning_rate": 3.2584615384615388e-06, + "loss": 0.0004, + "step": 13650 + }, + { + "epoch": 11.56448202959831, + "grad_norm": 0.014084039255976677, + "learning_rate": 3.245641025641026e-06, + "loss": 0.0002, + "step": 13675 + }, + { + "epoch": 11.585623678646934, + "grad_norm": 0.022753361612558365, + "learning_rate": 3.2328205128205133e-06, + "loss": 0.0002, + "step": 13700 + }, + { + "epoch": 11.60676532769556, + "grad_norm": 0.012626050040125847, + "learning_rate": 3.2200000000000005e-06, + "loss": 0.0007, + "step": 13725 + }, + { + "epoch": 11.627906976744185, + "grad_norm": 0.9902653694152832, + "learning_rate": 3.2071794871794878e-06, + "loss": 0.0009, + "step": 13750 + }, + { + "epoch": 11.649048625792812, + "grad_norm": 0.034385230392217636, + "learning_rate": 3.194358974358975e-06, + "loss": 0.0003, + "step": 13775 + }, + { + "epoch": 11.670190274841438, + "grad_norm": 0.013236723840236664, + "learning_rate": 3.181538461538462e-06, + "loss": 0.0004, + "step": 13800 + }, + { + "epoch": 11.691331923890063, + "grad_norm": 0.014498353935778141, + "learning_rate": 3.168717948717949e-06, + "loss": 0.0009, + "step": 13825 + }, + { + "epoch": 11.71247357293869, + "grad_norm": 0.012556152418255806, + "learning_rate": 3.155897435897436e-06, + "loss": 0.0007, + "step": 13850 + }, + { + "epoch": 11.733615221987314, + "grad_norm": 0.017172476276755333, + "learning_rate": 3.143076923076923e-06, + "loss": 0.0002, + "step": 13875 + }, + { + "epoch": 11.75475687103594, + "grad_norm": 0.04857848584651947, + "learning_rate": 3.1302564102564104e-06, + "loss": 0.0007, + "step": 13900 + }, + { + "epoch": 11.775898520084567, + "grad_norm": 0.2465539425611496, + "learning_rate": 3.1174358974358976e-06, + "loss": 0.001, + "step": 13925 + }, + { + "epoch": 11.797040169133192, + "grad_norm": 0.01584937795996666, + "learning_rate": 3.104615384615385e-06, + "loss": 0.0005, + "step": 13950 + }, + { + "epoch": 11.818181818181818, + "grad_norm": 0.022607628256082535, + "learning_rate": 3.091794871794872e-06, + "loss": 0.0003, + "step": 13975 + }, + { + "epoch": 11.839323467230443, + "grad_norm": 0.021190233528614044, + "learning_rate": 3.0789743589743594e-06, + "loss": 0.0005, + "step": 14000 + }, + { + "epoch": 11.839323467230443, + "eval_loss": 0.08707120269536972, + "eval_runtime": 422.3648, + "eval_samples_per_second": 8.621, + "eval_steps_per_second": 0.54, + "eval_wer": 0.061394960322984825, + "step": 14000 + }, + { + "epoch": 11.86046511627907, + "grad_norm": 0.02160489559173584, + "learning_rate": 3.0661538461538466e-06, + "loss": 0.0013, + "step": 14025 + }, + { + "epoch": 11.881606765327696, + "grad_norm": 0.017302557826042175, + "learning_rate": 3.053333333333334e-06, + "loss": 0.0006, + "step": 14050 + }, + { + "epoch": 11.90274841437632, + "grad_norm": 0.03932559862732887, + "learning_rate": 3.040512820512821e-06, + "loss": 0.0006, + "step": 14075 + }, + { + "epoch": 11.923890063424947, + "grad_norm": 0.16547563672065735, + "learning_rate": 3.027692307692308e-06, + "loss": 0.001, + "step": 14100 + }, + { + "epoch": 11.945031712473574, + "grad_norm": 0.05746368691325188, + "learning_rate": 3.014871794871795e-06, + "loss": 0.0005, + "step": 14125 + }, + { + "epoch": 11.966173361522198, + "grad_norm": 0.017229914665222168, + "learning_rate": 3.002051282051282e-06, + "loss": 0.0016, + "step": 14150 + }, + { + "epoch": 11.987315010570825, + "grad_norm": 0.017080632969737053, + "learning_rate": 2.9892307692307693e-06, + "loss": 0.0005, + "step": 14175 + }, + { + "epoch": 12.00845665961945, + "grad_norm": 0.14502297341823578, + "learning_rate": 2.9764102564102565e-06, + "loss": 0.0004, + "step": 14200 + }, + { + "epoch": 12.029598308668076, + "grad_norm": 0.013931095600128174, + "learning_rate": 2.9635897435897438e-06, + "loss": 0.0002, + "step": 14225 + }, + { + "epoch": 12.050739957716702, + "grad_norm": 0.04354655370116234, + "learning_rate": 2.950769230769231e-06, + "loss": 0.0004, + "step": 14250 + }, + { + "epoch": 12.071881606765327, + "grad_norm": 0.015747781842947006, + "learning_rate": 2.9379487179487183e-06, + "loss": 0.0006, + "step": 14275 + }, + { + "epoch": 12.093023255813954, + "grad_norm": 0.08688750118017197, + "learning_rate": 2.9251282051282055e-06, + "loss": 0.0009, + "step": 14300 + }, + { + "epoch": 12.11416490486258, + "grad_norm": 0.01061871089041233, + "learning_rate": 2.9123076923076928e-06, + "loss": 0.0013, + "step": 14325 + }, + { + "epoch": 12.135306553911205, + "grad_norm": 0.01094849780201912, + "learning_rate": 2.89948717948718e-06, + "loss": 0.0006, + "step": 14350 + }, + { + "epoch": 12.156448202959831, + "grad_norm": 0.012437337078154087, + "learning_rate": 2.8866666666666673e-06, + "loss": 0.0005, + "step": 14375 + }, + { + "epoch": 12.177589852008456, + "grad_norm": 0.014366303570568562, + "learning_rate": 2.873846153846154e-06, + "loss": 0.0005, + "step": 14400 + }, + { + "epoch": 12.198731501057082, + "grad_norm": 0.017898643389344215, + "learning_rate": 2.8610256410256413e-06, + "loss": 0.0005, + "step": 14425 + }, + { + "epoch": 12.219873150105709, + "grad_norm": 0.016448453068733215, + "learning_rate": 2.848205128205128e-06, + "loss": 0.0002, + "step": 14450 + }, + { + "epoch": 12.241014799154334, + "grad_norm": 0.017254604026675224, + "learning_rate": 2.8353846153846154e-06, + "loss": 0.0003, + "step": 14475 + }, + { + "epoch": 12.26215644820296, + "grad_norm": 0.01735294982790947, + "learning_rate": 2.8225641025641027e-06, + "loss": 0.0005, + "step": 14500 + }, + { + "epoch": 12.283298097251585, + "grad_norm": 0.017333608120679855, + "learning_rate": 2.80974358974359e-06, + "loss": 0.0005, + "step": 14525 + }, + { + "epoch": 12.304439746300211, + "grad_norm": 0.010371909476816654, + "learning_rate": 2.796923076923077e-06, + "loss": 0.0004, + "step": 14550 + }, + { + "epoch": 12.325581395348838, + "grad_norm": 0.051518309861421585, + "learning_rate": 2.7841025641025644e-06, + "loss": 0.0003, + "step": 14575 + }, + { + "epoch": 12.346723044397462, + "grad_norm": 0.015523129142820835, + "learning_rate": 2.7712820512820516e-06, + "loss": 0.0002, + "step": 14600 + }, + { + "epoch": 12.367864693446089, + "grad_norm": 0.013449127785861492, + "learning_rate": 2.758461538461539e-06, + "loss": 0.0003, + "step": 14625 + }, + { + "epoch": 12.389006342494715, + "grad_norm": 0.28945910930633545, + "learning_rate": 2.745641025641026e-06, + "loss": 0.0008, + "step": 14650 + }, + { + "epoch": 12.41014799154334, + "grad_norm": 0.01732316054403782, + "learning_rate": 2.7328205128205134e-06, + "loss": 0.0002, + "step": 14675 + }, + { + "epoch": 12.431289640591967, + "grad_norm": 0.0187962856143713, + "learning_rate": 2.7200000000000002e-06, + "loss": 0.0006, + "step": 14700 + }, + { + "epoch": 12.452431289640591, + "grad_norm": 0.010457909666001797, + "learning_rate": 2.7071794871794875e-06, + "loss": 0.0008, + "step": 14725 + }, + { + "epoch": 12.473572938689218, + "grad_norm": 0.017735542729496956, + "learning_rate": 2.6943589743589743e-06, + "loss": 0.0006, + "step": 14750 + }, + { + "epoch": 12.494714587737844, + "grad_norm": 0.01583629846572876, + "learning_rate": 2.6815384615384615e-06, + "loss": 0.0008, + "step": 14775 + }, + { + "epoch": 12.515856236786469, + "grad_norm": 0.024372195824980736, + "learning_rate": 2.6687179487179488e-06, + "loss": 0.0003, + "step": 14800 + }, + { + "epoch": 12.536997885835095, + "grad_norm": 0.03954934701323509, + "learning_rate": 2.655897435897436e-06, + "loss": 0.0002, + "step": 14825 + }, + { + "epoch": 12.55813953488372, + "grad_norm": 0.07841933518648148, + "learning_rate": 2.6430769230769233e-06, + "loss": 0.0003, + "step": 14850 + }, + { + "epoch": 12.579281183932347, + "grad_norm": 0.023459434509277344, + "learning_rate": 2.6302564102564105e-06, + "loss": 0.0008, + "step": 14875 + }, + { + "epoch": 12.600422832980973, + "grad_norm": 0.02575707994401455, + "learning_rate": 2.6174358974358978e-06, + "loss": 0.0002, + "step": 14900 + }, + { + "epoch": 12.621564482029598, + "grad_norm": 0.01122299861162901, + "learning_rate": 2.604615384615385e-06, + "loss": 0.0008, + "step": 14925 + }, + { + "epoch": 12.642706131078224, + "grad_norm": 0.013861022889614105, + "learning_rate": 2.5917948717948723e-06, + "loss": 0.0002, + "step": 14950 + }, + { + "epoch": 12.66384778012685, + "grad_norm": 0.015051736496388912, + "learning_rate": 2.5789743589743595e-06, + "loss": 0.0009, + "step": 14975 + }, + { + "epoch": 12.684989429175475, + "grad_norm": 0.053748294711112976, + "learning_rate": 2.5661538461538463e-06, + "loss": 0.0002, + "step": 15000 + }, + { + "epoch": 12.684989429175475, + "eval_loss": 0.08498353511095047, + "eval_runtime": 422.6177, + "eval_samples_per_second": 8.615, + "eval_steps_per_second": 0.539, + "eval_wer": 0.05963153742633069, + "step": 15000 + }, + { + "epoch": 12.706131078224102, + "grad_norm": 0.008815528824925423, + "learning_rate": 2.5533333333333336e-06, + "loss": 0.0002, + "step": 15025 + }, + { + "epoch": 12.727272727272727, + "grad_norm": 0.01506373006850481, + "learning_rate": 2.5405128205128204e-06, + "loss": 0.0007, + "step": 15050 + }, + { + "epoch": 12.748414376321353, + "grad_norm": 0.11881545186042786, + "learning_rate": 2.5276923076923077e-06, + "loss": 0.0002, + "step": 15075 + }, + { + "epoch": 12.76955602536998, + "grad_norm": 0.01836121454834938, + "learning_rate": 2.514871794871795e-06, + "loss": 0.0004, + "step": 15100 + }, + { + "epoch": 12.790697674418604, + "grad_norm": 0.019225163385272026, + "learning_rate": 2.502051282051282e-06, + "loss": 0.0006, + "step": 15125 + }, + { + "epoch": 12.81183932346723, + "grad_norm": 0.02618946135044098, + "learning_rate": 2.4892307692307694e-06, + "loss": 0.0002, + "step": 15150 + }, + { + "epoch": 12.832980972515855, + "grad_norm": 0.02191704511642456, + "learning_rate": 2.4764102564102567e-06, + "loss": 0.0002, + "step": 15175 + }, + { + "epoch": 12.854122621564482, + "grad_norm": 0.010356785729527473, + "learning_rate": 2.463589743589744e-06, + "loss": 0.0004, + "step": 15200 + }, + { + "epoch": 12.875264270613108, + "grad_norm": 0.011190637946128845, + "learning_rate": 2.450769230769231e-06, + "loss": 0.0002, + "step": 15225 + }, + { + "epoch": 12.896405919661733, + "grad_norm": 0.4306105375289917, + "learning_rate": 2.4379487179487184e-06, + "loss": 0.0003, + "step": 15250 + }, + { + "epoch": 12.91754756871036, + "grad_norm": 0.01129214372485876, + "learning_rate": 2.4251282051282056e-06, + "loss": 0.0004, + "step": 15275 + }, + { + "epoch": 12.938689217758984, + "grad_norm": 0.013043936342000961, + "learning_rate": 2.4123076923076925e-06, + "loss": 0.0005, + "step": 15300 + }, + { + "epoch": 12.95983086680761, + "grad_norm": 0.03151563182473183, + "learning_rate": 2.3994871794871797e-06, + "loss": 0.0002, + "step": 15325 + }, + { + "epoch": 12.980972515856237, + "grad_norm": 0.01353500597178936, + "learning_rate": 2.386666666666667e-06, + "loss": 0.0002, + "step": 15350 + }, + { + "epoch": 13.002114164904862, + "grad_norm": 0.010209338739514351, + "learning_rate": 2.3738461538461542e-06, + "loss": 0.0009, + "step": 15375 + }, + { + "epoch": 13.023255813953488, + "grad_norm": 0.01054307445883751, + "learning_rate": 2.3610256410256415e-06, + "loss": 0.0002, + "step": 15400 + }, + { + "epoch": 13.044397463002115, + "grad_norm": 0.0059889694675803185, + "learning_rate": 2.3482051282051287e-06, + "loss": 0.0001, + "step": 15425 + }, + { + "epoch": 13.06553911205074, + "grad_norm": 0.008033188059926033, + "learning_rate": 2.3353846153846155e-06, + "loss": 0.0001, + "step": 15450 + }, + { + "epoch": 13.086680761099366, + "grad_norm": 0.009764975868165493, + "learning_rate": 2.3225641025641028e-06, + "loss": 0.0001, + "step": 15475 + }, + { + "epoch": 13.10782241014799, + "grad_norm": 0.015029883943498135, + "learning_rate": 2.30974358974359e-06, + "loss": 0.0001, + "step": 15500 + }, + { + "epoch": 13.128964059196617, + "grad_norm": 0.8327123522758484, + "learning_rate": 2.2969230769230773e-06, + "loss": 0.0003, + "step": 15525 + }, + { + "epoch": 13.150105708245244, + "grad_norm": 0.010631484910845757, + "learning_rate": 2.2841025641025645e-06, + "loss": 0.0005, + "step": 15550 + }, + { + "epoch": 13.171247357293868, + "grad_norm": 0.011425021104514599, + "learning_rate": 2.2712820512820518e-06, + "loss": 0.0001, + "step": 15575 + }, + { + "epoch": 13.192389006342495, + "grad_norm": 0.009755819104611874, + "learning_rate": 2.2584615384615386e-06, + "loss": 0.0002, + "step": 15600 + }, + { + "epoch": 13.213530655391121, + "grad_norm": 0.011913259513676167, + "learning_rate": 2.245641025641026e-06, + "loss": 0.0007, + "step": 15625 + }, + { + "epoch": 13.234672304439746, + "grad_norm": 0.0059737637639045715, + "learning_rate": 2.232820512820513e-06, + "loss": 0.0002, + "step": 15650 + }, + { + "epoch": 13.255813953488373, + "grad_norm": 0.015654481947422028, + "learning_rate": 2.2200000000000003e-06, + "loss": 0.0009, + "step": 15675 + }, + { + "epoch": 13.276955602536997, + "grad_norm": 0.010202393867075443, + "learning_rate": 2.2071794871794876e-06, + "loss": 0.0002, + "step": 15700 + }, + { + "epoch": 13.298097251585624, + "grad_norm": 0.008887005969882011, + "learning_rate": 2.194358974358975e-06, + "loss": 0.0002, + "step": 15725 + }, + { + "epoch": 13.31923890063425, + "grad_norm": 0.010564594529569149, + "learning_rate": 2.1815384615384617e-06, + "loss": 0.0001, + "step": 15750 + }, + { + "epoch": 13.340380549682875, + "grad_norm": 0.007872478105127811, + "learning_rate": 2.168717948717949e-06, + "loss": 0.0001, + "step": 15775 + }, + { + "epoch": 13.361522198731501, + "grad_norm": 0.014362242072820663, + "learning_rate": 2.155897435897436e-06, + "loss": 0.0001, + "step": 15800 + }, + { + "epoch": 13.382663847780126, + "grad_norm": 0.013115731067955494, + "learning_rate": 2.1430769230769234e-06, + "loss": 0.0001, + "step": 15825 + }, + { + "epoch": 13.403805496828753, + "grad_norm": 0.007929886691272259, + "learning_rate": 2.1302564102564107e-06, + "loss": 0.0004, + "step": 15850 + }, + { + "epoch": 13.424947145877379, + "grad_norm": 0.01074293628334999, + "learning_rate": 2.117435897435898e-06, + "loss": 0.0001, + "step": 15875 + }, + { + "epoch": 13.446088794926004, + "grad_norm": 0.007108274381607771, + "learning_rate": 2.1046153846153847e-06, + "loss": 0.0002, + "step": 15900 + }, + { + "epoch": 13.46723044397463, + "grad_norm": 0.006991777569055557, + "learning_rate": 2.091794871794872e-06, + "loss": 0.0003, + "step": 15925 + }, + { + "epoch": 13.488372093023255, + "grad_norm": 0.006929017137736082, + "learning_rate": 2.0789743589743592e-06, + "loss": 0.0001, + "step": 15950 + }, + { + "epoch": 13.509513742071881, + "grad_norm": 0.016624469310045242, + "learning_rate": 2.0661538461538465e-06, + "loss": 0.0002, + "step": 15975 + }, + { + "epoch": 13.530655391120508, + "grad_norm": 0.009122622199356556, + "learning_rate": 2.0533333333333337e-06, + "loss": 0.0004, + "step": 16000 + }, + { + "epoch": 13.530655391120508, + "eval_loss": 0.08493294566869736, + "eval_runtime": 416.438, + "eval_samples_per_second": 8.743, + "eval_steps_per_second": 0.548, + "eval_wer": 0.06004919021764351, + "step": 16000 + }, + { + "epoch": 13.551797040169133, + "grad_norm": 0.007773926947265863, + "learning_rate": 2.040512820512821e-06, + "loss": 0.0003, + "step": 16025 + }, + { + "epoch": 13.572938689217759, + "grad_norm": 0.023034438490867615, + "learning_rate": 2.027692307692308e-06, + "loss": 0.0004, + "step": 16050 + }, + { + "epoch": 13.594080338266386, + "grad_norm": 0.012188156135380268, + "learning_rate": 2.014871794871795e-06, + "loss": 0.0001, + "step": 16075 + }, + { + "epoch": 13.61522198731501, + "grad_norm": 0.009154338389635086, + "learning_rate": 2.0020512820512823e-06, + "loss": 0.0001, + "step": 16100 + }, + { + "epoch": 13.636363636363637, + "grad_norm": 0.007401302456855774, + "learning_rate": 1.9892307692307695e-06, + "loss": 0.0002, + "step": 16125 + }, + { + "epoch": 13.657505285412261, + "grad_norm": 0.011268429458141327, + "learning_rate": 1.9764102564102568e-06, + "loss": 0.0005, + "step": 16150 + }, + { + "epoch": 13.678646934460888, + "grad_norm": 0.010454394854605198, + "learning_rate": 1.963589743589744e-06, + "loss": 0.0001, + "step": 16175 + }, + { + "epoch": 13.699788583509514, + "grad_norm": 0.011891184374690056, + "learning_rate": 1.950769230769231e-06, + "loss": 0.0003, + "step": 16200 + }, + { + "epoch": 13.720930232558139, + "grad_norm": 0.016350483521819115, + "learning_rate": 1.937948717948718e-06, + "loss": 0.0001, + "step": 16225 + }, + { + "epoch": 13.742071881606766, + "grad_norm": 0.13966374099254608, + "learning_rate": 1.9251282051282054e-06, + "loss": 0.0004, + "step": 16250 + }, + { + "epoch": 13.763213530655392, + "grad_norm": 0.008222724311053753, + "learning_rate": 1.9123076923076926e-06, + "loss": 0.0001, + "step": 16275 + }, + { + "epoch": 13.784355179704017, + "grad_norm": 0.011430777609348297, + "learning_rate": 1.8994871794871796e-06, + "loss": 0.0002, + "step": 16300 + }, + { + "epoch": 13.805496828752643, + "grad_norm": 0.009091757237911224, + "learning_rate": 1.8866666666666669e-06, + "loss": 0.0006, + "step": 16325 + }, + { + "epoch": 13.826638477801268, + "grad_norm": 0.009097087197005749, + "learning_rate": 1.873846153846154e-06, + "loss": 0.0001, + "step": 16350 + }, + { + "epoch": 13.847780126849894, + "grad_norm": 0.0094605116173625, + "learning_rate": 1.8610256410256412e-06, + "loss": 0.0001, + "step": 16375 + }, + { + "epoch": 13.86892177589852, + "grad_norm": 0.006917539052665234, + "learning_rate": 1.8482051282051284e-06, + "loss": 0.0001, + "step": 16400 + }, + { + "epoch": 13.890063424947146, + "grad_norm": 0.006485483143478632, + "learning_rate": 1.8353846153846155e-06, + "loss": 0.0004, + "step": 16425 + }, + { + "epoch": 13.911205073995772, + "grad_norm": 0.010646677576005459, + "learning_rate": 1.8225641025641027e-06, + "loss": 0.0003, + "step": 16450 + }, + { + "epoch": 13.932346723044397, + "grad_norm": 0.009001165628433228, + "learning_rate": 1.80974358974359e-06, + "loss": 0.0002, + "step": 16475 + }, + { + "epoch": 13.953488372093023, + "grad_norm": 0.007851574569940567, + "learning_rate": 1.796923076923077e-06, + "loss": 0.0003, + "step": 16500 + }, + { + "epoch": 13.97463002114165, + "grad_norm": 0.005184480920433998, + "learning_rate": 1.7841025641025642e-06, + "loss": 0.0001, + "step": 16525 + }, + { + "epoch": 13.995771670190274, + "grad_norm": 0.009073914960026741, + "learning_rate": 1.7712820512820515e-06, + "loss": 0.0005, + "step": 16550 + }, + { + "epoch": 14.0169133192389, + "grad_norm": 0.006329812109470367, + "learning_rate": 1.7584615384615385e-06, + "loss": 0.0002, + "step": 16575 + }, + { + "epoch": 14.038054968287527, + "grad_norm": 0.00448222178965807, + "learning_rate": 1.7456410256410258e-06, + "loss": 0.0001, + "step": 16600 + }, + { + "epoch": 14.059196617336152, + "grad_norm": 0.006917804013937712, + "learning_rate": 1.732820512820513e-06, + "loss": 0.0001, + "step": 16625 + }, + { + "epoch": 14.080338266384778, + "grad_norm": 0.011933974921703339, + "learning_rate": 1.72e-06, + "loss": 0.0002, + "step": 16650 + }, + { + "epoch": 14.101479915433403, + "grad_norm": 0.006539530120790005, + "learning_rate": 1.7071794871794873e-06, + "loss": 0.0001, + "step": 16675 + }, + { + "epoch": 14.12262156448203, + "grad_norm": 0.007711876183748245, + "learning_rate": 1.6943589743589745e-06, + "loss": 0.0001, + "step": 16700 + }, + { + "epoch": 14.143763213530656, + "grad_norm": 0.00857022125273943, + "learning_rate": 1.6820512820512822e-06, + "loss": 0.0007, + "step": 16725 + }, + { + "epoch": 14.16490486257928, + "grad_norm": 0.00736591499298811, + "learning_rate": 1.6692307692307694e-06, + "loss": 0.0001, + "step": 16750 + }, + { + "epoch": 14.186046511627907, + "grad_norm": 0.008430298417806625, + "learning_rate": 1.6564102564102567e-06, + "loss": 0.0001, + "step": 16775 + }, + { + "epoch": 14.207188160676532, + "grad_norm": 0.006714751478284597, + "learning_rate": 1.643589743589744e-06, + "loss": 0.0002, + "step": 16800 + }, + { + "epoch": 14.228329809725158, + "grad_norm": 0.007295829243957996, + "learning_rate": 1.6307692307692307e-06, + "loss": 0.0001, + "step": 16825 + }, + { + "epoch": 14.249471458773785, + "grad_norm": 0.005545494146645069, + "learning_rate": 1.617948717948718e-06, + "loss": 0.0001, + "step": 16850 + }, + { + "epoch": 14.27061310782241, + "grad_norm": 0.006525352597236633, + "learning_rate": 1.6051282051282052e-06, + "loss": 0.0003, + "step": 16875 + }, + { + "epoch": 14.291754756871036, + "grad_norm": 0.006453375332057476, + "learning_rate": 1.5923076923076925e-06, + "loss": 0.0001, + "step": 16900 + }, + { + "epoch": 14.312896405919663, + "grad_norm": 0.005989603232592344, + "learning_rate": 1.5794871794871797e-06, + "loss": 0.0001, + "step": 16925 + }, + { + "epoch": 14.334038054968287, + "grad_norm": 0.006138788536190987, + "learning_rate": 1.566666666666667e-06, + "loss": 0.0001, + "step": 16950 + }, + { + "epoch": 14.355179704016914, + "grad_norm": 0.006993785500526428, + "learning_rate": 1.5538461538461538e-06, + "loss": 0.0001, + "step": 16975 + }, + { + "epoch": 14.376321353065538, + "grad_norm": 0.007072498090565205, + "learning_rate": 1.541025641025641e-06, + "loss": 0.0001, + "step": 17000 + }, + { + "epoch": 14.376321353065538, + "eval_loss": 0.0867857038974762, + "eval_runtime": 413.9388, + "eval_samples_per_second": 8.796, + "eval_steps_per_second": 0.551, + "eval_wer": 0.05916747876931644, + "step": 17000 + }, + { + "epoch": 14.397463002114165, + "grad_norm": 0.008112003095448017, + "learning_rate": 1.5282051282051283e-06, + "loss": 0.0001, + "step": 17025 + }, + { + "epoch": 14.418604651162791, + "grad_norm": 0.007216181140393019, + "learning_rate": 1.5153846153846156e-06, + "loss": 0.0002, + "step": 17050 + }, + { + "epoch": 14.439746300211416, + "grad_norm": 0.003465973073616624, + "learning_rate": 1.5025641025641028e-06, + "loss": 0.0001, + "step": 17075 + }, + { + "epoch": 14.460887949260043, + "grad_norm": 0.005432849284261465, + "learning_rate": 1.48974358974359e-06, + "loss": 0.0001, + "step": 17100 + }, + { + "epoch": 14.482029598308667, + "grad_norm": 0.00709709245711565, + "learning_rate": 1.476923076923077e-06, + "loss": 0.0001, + "step": 17125 + }, + { + "epoch": 14.503171247357294, + "grad_norm": 0.006040381733328104, + "learning_rate": 1.4641025641025641e-06, + "loss": 0.0001, + "step": 17150 + }, + { + "epoch": 14.52431289640592, + "grad_norm": 0.00656255753710866, + "learning_rate": 1.4512820512820514e-06, + "loss": 0.0001, + "step": 17175 + }, + { + "epoch": 14.545454545454545, + "grad_norm": 0.006466065067797899, + "learning_rate": 1.4384615384615386e-06, + "loss": 0.0001, + "step": 17200 + }, + { + "epoch": 14.566596194503171, + "grad_norm": 0.008718462660908699, + "learning_rate": 1.4261538461538462e-06, + "loss": 0.0003, + "step": 17225 + }, + { + "epoch": 14.587737843551796, + "grad_norm": 0.0089957807213068, + "learning_rate": 1.4133333333333335e-06, + "loss": 0.0001, + "step": 17250 + }, + { + "epoch": 14.608879492600423, + "grad_norm": 0.004027772229164839, + "learning_rate": 1.4005128205128207e-06, + "loss": 0.0003, + "step": 17275 + }, + { + "epoch": 14.630021141649049, + "grad_norm": 0.006016540341079235, + "learning_rate": 1.3876923076923076e-06, + "loss": 0.0001, + "step": 17300 + }, + { + "epoch": 14.651162790697674, + "grad_norm": 0.006955642718821764, + "learning_rate": 1.3748717948717948e-06, + "loss": 0.0001, + "step": 17325 + }, + { + "epoch": 14.6723044397463, + "grad_norm": 0.006230717524886131, + "learning_rate": 1.362051282051282e-06, + "loss": 0.0001, + "step": 17350 + }, + { + "epoch": 14.693446088794927, + "grad_norm": 0.007248835172504187, + "learning_rate": 1.3492307692307693e-06, + "loss": 0.0001, + "step": 17375 + }, + { + "epoch": 14.714587737843551, + "grad_norm": 0.009279552847146988, + "learning_rate": 1.3364102564102566e-06, + "loss": 0.0001, + "step": 17400 + }, + { + "epoch": 14.735729386892178, + "grad_norm": 0.004904840141534805, + "learning_rate": 1.3235897435897438e-06, + "loss": 0.0001, + "step": 17425 + }, + { + "epoch": 14.756871035940803, + "grad_norm": 0.008709315210580826, + "learning_rate": 1.3107692307692308e-06, + "loss": 0.0002, + "step": 17450 + }, + { + "epoch": 14.778012684989429, + "grad_norm": 0.005629656370729208, + "learning_rate": 1.2979487179487179e-06, + "loss": 0.0001, + "step": 17475 + }, + { + "epoch": 14.799154334038056, + "grad_norm": 0.007197321858257055, + "learning_rate": 1.2851282051282051e-06, + "loss": 0.0001, + "step": 17500 + }, + { + "epoch": 14.82029598308668, + "grad_norm": 0.013610289432108402, + "learning_rate": 1.2723076923076924e-06, + "loss": 0.0001, + "step": 17525 + }, + { + "epoch": 14.841437632135307, + "grad_norm": 0.006126283202320337, + "learning_rate": 1.2594871794871796e-06, + "loss": 0.0004, + "step": 17550 + }, + { + "epoch": 14.862579281183933, + "grad_norm": 0.006340153515338898, + "learning_rate": 1.2466666666666667e-06, + "loss": 0.0004, + "step": 17575 + }, + { + "epoch": 14.883720930232558, + "grad_norm": 0.007039876654744148, + "learning_rate": 1.233846153846154e-06, + "loss": 0.0001, + "step": 17600 + }, + { + "epoch": 14.904862579281184, + "grad_norm": 0.006753341294825077, + "learning_rate": 1.2210256410256412e-06, + "loss": 0.0001, + "step": 17625 + }, + { + "epoch": 14.926004228329809, + "grad_norm": 0.007074939087033272, + "learning_rate": 1.2082051282051282e-06, + "loss": 0.0001, + "step": 17650 + }, + { + "epoch": 14.947145877378436, + "grad_norm": 0.006931220646947622, + "learning_rate": 1.1953846153846154e-06, + "loss": 0.0002, + "step": 17675 + }, + { + "epoch": 14.968287526427062, + "grad_norm": 0.006897748447954655, + "learning_rate": 1.1825641025641027e-06, + "loss": 0.0001, + "step": 17700 + }, + { + "epoch": 14.989429175475687, + "grad_norm": 0.0052500138990581036, + "learning_rate": 1.1697435897435897e-06, + "loss": 0.0003, + "step": 17725 + }, + { + "epoch": 15.010570824524313, + "grad_norm": 0.003996185027062893, + "learning_rate": 1.156923076923077e-06, + "loss": 0.0002, + "step": 17750 + }, + { + "epoch": 15.031712473572938, + "grad_norm": 0.006482282187789679, + "learning_rate": 1.1441025641025642e-06, + "loss": 0.0001, + "step": 17775 + }, + { + "epoch": 15.052854122621564, + "grad_norm": 0.004399996716529131, + "learning_rate": 1.1312820512820513e-06, + "loss": 0.0001, + "step": 17800 + }, + { + "epoch": 15.073995771670191, + "grad_norm": 0.004176503047347069, + "learning_rate": 1.1184615384615385e-06, + "loss": 0.0001, + "step": 17825 + }, + { + "epoch": 15.095137420718816, + "grad_norm": 0.005335643887519836, + "learning_rate": 1.1056410256410258e-06, + "loss": 0.0001, + "step": 17850 + }, + { + "epoch": 15.116279069767442, + "grad_norm": 0.006689351052045822, + "learning_rate": 1.0928205128205128e-06, + "loss": 0.0001, + "step": 17875 + }, + { + "epoch": 15.137420718816067, + "grad_norm": 0.0035949235316365957, + "learning_rate": 1.08e-06, + "loss": 0.0001, + "step": 17900 + }, + { + "epoch": 15.158562367864693, + "grad_norm": 0.004826567135751247, + "learning_rate": 1.0671794871794873e-06, + "loss": 0.0001, + "step": 17925 + }, + { + "epoch": 15.17970401691332, + "grad_norm": 0.006563639268279076, + "learning_rate": 1.0543589743589743e-06, + "loss": 0.0001, + "step": 17950 + }, + { + "epoch": 15.200845665961944, + "grad_norm": 0.005385232623666525, + "learning_rate": 1.0415384615384616e-06, + "loss": 0.0001, + "step": 17975 + }, + { + "epoch": 15.221987315010571, + "grad_norm": 0.004666761960834265, + "learning_rate": 1.0287179487179488e-06, + "loss": 0.0002, + "step": 18000 + }, + { + "epoch": 15.221987315010571, + "eval_loss": 0.08731851726770401, + "eval_runtime": 418.9511, + "eval_samples_per_second": 8.691, + "eval_steps_per_second": 0.544, + "eval_wer": 0.05930669636642071, + "step": 18000 + }, + { + "epoch": 15.243128964059197, + "grad_norm": 0.004958902020007372, + "learning_rate": 1.0158974358974359e-06, + "loss": 0.0001, + "step": 18025 + }, + { + "epoch": 15.264270613107822, + "grad_norm": 0.005769228097051382, + "learning_rate": 1.003076923076923e-06, + "loss": 0.0001, + "step": 18050 + }, + { + "epoch": 15.285412262156449, + "grad_norm": 0.004522492177784443, + "learning_rate": 9.902564102564103e-07, + "loss": 0.0001, + "step": 18075 + }, + { + "epoch": 15.306553911205073, + "grad_norm": 0.00491857435554266, + "learning_rate": 9.774358974358974e-07, + "loss": 0.0001, + "step": 18100 + }, + { + "epoch": 15.3276955602537, + "grad_norm": 0.007983505725860596, + "learning_rate": 9.646153846153846e-07, + "loss": 0.0001, + "step": 18125 + }, + { + "epoch": 15.348837209302326, + "grad_norm": 0.004594389349222183, + "learning_rate": 9.517948717948719e-07, + "loss": 0.0001, + "step": 18150 + }, + { + "epoch": 15.369978858350951, + "grad_norm": 0.005598857067525387, + "learning_rate": 9.38974358974359e-07, + "loss": 0.0001, + "step": 18175 + }, + { + "epoch": 15.391120507399577, + "grad_norm": 0.004582141991704702, + "learning_rate": 9.261538461538462e-07, + "loss": 0.0001, + "step": 18200 + }, + { + "epoch": 15.412262156448204, + "grad_norm": 0.008931068703532219, + "learning_rate": 9.133333333333334e-07, + "loss": 0.0001, + "step": 18225 + }, + { + "epoch": 15.433403805496829, + "grad_norm": 0.0040603941306471825, + "learning_rate": 9.005128205128206e-07, + "loss": 0.0001, + "step": 18250 + }, + { + "epoch": 15.454545454545455, + "grad_norm": 0.005295175593346357, + "learning_rate": 8.876923076923077e-07, + "loss": 0.0001, + "step": 18275 + }, + { + "epoch": 15.47568710359408, + "grad_norm": 0.005524117033928633, + "learning_rate": 8.748717948717949e-07, + "loss": 0.0001, + "step": 18300 + }, + { + "epoch": 15.496828752642706, + "grad_norm": 0.005777684506028891, + "learning_rate": 8.620512820512822e-07, + "loss": 0.0001, + "step": 18325 + }, + { + "epoch": 15.517970401691333, + "grad_norm": 0.004244440235197544, + "learning_rate": 8.492307692307692e-07, + "loss": 0.0001, + "step": 18350 + }, + { + "epoch": 15.539112050739957, + "grad_norm": 0.008006569929420948, + "learning_rate": 8.364102564102565e-07, + "loss": 0.0001, + "step": 18375 + }, + { + "epoch": 15.560253699788584, + "grad_norm": 0.00508382823318243, + "learning_rate": 8.235897435897437e-07, + "loss": 0.0001, + "step": 18400 + }, + { + "epoch": 15.581395348837209, + "grad_norm": 0.005526789929717779, + "learning_rate": 8.107692307692308e-07, + "loss": 0.0001, + "step": 18425 + }, + { + "epoch": 15.602536997885835, + "grad_norm": 0.0036265768576413393, + "learning_rate": 7.97948717948718e-07, + "loss": 0.0001, + "step": 18450 + }, + { + "epoch": 15.623678646934462, + "grad_norm": 0.004917936399579048, + "learning_rate": 7.851282051282053e-07, + "loss": 0.0002, + "step": 18475 + }, + { + "epoch": 15.644820295983086, + "grad_norm": 0.005852636881172657, + "learning_rate": 7.723076923076923e-07, + "loss": 0.0001, + "step": 18500 + }, + { + "epoch": 15.665961945031713, + "grad_norm": 0.0046966904774308205, + "learning_rate": 7.594871794871795e-07, + "loss": 0.0001, + "step": 18525 + }, + { + "epoch": 15.687103594080337, + "grad_norm": 0.004561580251902342, + "learning_rate": 7.466666666666668e-07, + "loss": 0.0001, + "step": 18550 + }, + { + "epoch": 15.708245243128964, + "grad_norm": 0.006141614634543657, + "learning_rate": 7.338461538461538e-07, + "loss": 0.0001, + "step": 18575 + }, + { + "epoch": 15.72938689217759, + "grad_norm": 0.005293000023812056, + "learning_rate": 7.210256410256411e-07, + "loss": 0.0001, + "step": 18600 + }, + { + "epoch": 15.750528541226215, + "grad_norm": 0.024633372202515602, + "learning_rate": 7.082051282051283e-07, + "loss": 0.0001, + "step": 18625 + }, + { + "epoch": 15.771670190274842, + "grad_norm": 0.007387631572782993, + "learning_rate": 6.953846153846154e-07, + "loss": 0.0001, + "step": 18650 + }, + { + "epoch": 15.792811839323468, + "grad_norm": 0.005313596688210964, + "learning_rate": 6.825641025641026e-07, + "loss": 0.0002, + "step": 18675 + }, + { + "epoch": 15.813953488372093, + "grad_norm": 0.004781834315508604, + "learning_rate": 6.697435897435899e-07, + "loss": 0.0001, + "step": 18700 + }, + { + "epoch": 15.83509513742072, + "grad_norm": 0.11602696031332016, + "learning_rate": 6.569230769230769e-07, + "loss": 0.0001, + "step": 18725 + }, + { + "epoch": 15.856236786469344, + "grad_norm": 0.005565613973885775, + "learning_rate": 6.441025641025641e-07, + "loss": 0.0001, + "step": 18750 + }, + { + "epoch": 15.87737843551797, + "grad_norm": 0.005617073271423578, + "learning_rate": 6.312820512820514e-07, + "loss": 0.0001, + "step": 18775 + }, + { + "epoch": 15.898520084566597, + "grad_norm": 0.005743805319070816, + "learning_rate": 6.184615384615385e-07, + "loss": 0.0001, + "step": 18800 + }, + { + "epoch": 15.919661733615222, + "grad_norm": 0.005840159021317959, + "learning_rate": 6.056410256410257e-07, + "loss": 0.0001, + "step": 18825 + }, + { + "epoch": 15.940803382663848, + "grad_norm": 0.0066063860431313515, + "learning_rate": 5.928205128205128e-07, + "loss": 0.0001, + "step": 18850 + }, + { + "epoch": 15.961945031712474, + "grad_norm": 0.0054490105248987675, + "learning_rate": 5.800000000000001e-07, + "loss": 0.0001, + "step": 18875 + }, + { + "epoch": 15.9830866807611, + "grad_norm": 0.006493702996522188, + "learning_rate": 5.671794871794872e-07, + "loss": 0.0007, + "step": 18900 + }, + { + "epoch": 16.004228329809724, + "grad_norm": 0.004811755381524563, + "learning_rate": 5.543589743589743e-07, + "loss": 0.0001, + "step": 18925 + }, + { + "epoch": 16.02536997885835, + "grad_norm": 0.005250418093055487, + "learning_rate": 5.415384615384616e-07, + "loss": 0.0001, + "step": 18950 + }, + { + "epoch": 16.046511627906977, + "grad_norm": 0.004121173173189163, + "learning_rate": 5.287179487179487e-07, + "loss": 0.0001, + "step": 18975 + }, + { + "epoch": 16.067653276955603, + "grad_norm": 0.005048220045864582, + "learning_rate": 5.158974358974359e-07, + "loss": 0.0001, + "step": 19000 + }, + { + "epoch": 16.067653276955603, + "eval_loss": 0.08753260225057602, + "eval_runtime": 419.2956, + "eval_samples_per_second": 8.684, + "eval_steps_per_second": 0.544, + "eval_wer": 0.05847139078379507, + "step": 19000 + }, + { + "epoch": 16.08879492600423, + "grad_norm": 0.005383364390581846, + "learning_rate": 5.030769230769231e-07, + "loss": 0.0001, + "step": 19025 + }, + { + "epoch": 16.109936575052853, + "grad_norm": 0.003624641802161932, + "learning_rate": 4.902564102564103e-07, + "loss": 0.0001, + "step": 19050 + }, + { + "epoch": 16.13107822410148, + "grad_norm": 0.003508235327899456, + "learning_rate": 4.774358974358974e-07, + "loss": 0.0001, + "step": 19075 + }, + { + "epoch": 16.152219873150106, + "grad_norm": 0.004268943332135677, + "learning_rate": 4.6461538461538465e-07, + "loss": 0.0001, + "step": 19100 + }, + { + "epoch": 16.173361522198732, + "grad_norm": 0.003968693781644106, + "learning_rate": 4.5179487179487185e-07, + "loss": 0.0001, + "step": 19125 + }, + { + "epoch": 16.19450317124736, + "grad_norm": 0.004143453668802977, + "learning_rate": 4.38974358974359e-07, + "loss": 0.0001, + "step": 19150 + }, + { + "epoch": 16.21564482029598, + "grad_norm": 0.004426514729857445, + "learning_rate": 4.261538461538462e-07, + "loss": 0.0001, + "step": 19175 + }, + { + "epoch": 16.236786469344608, + "grad_norm": 0.005651060491800308, + "learning_rate": 4.133333333333334e-07, + "loss": 0.0001, + "step": 19200 + }, + { + "epoch": 16.257928118393234, + "grad_norm": 0.003962742630392313, + "learning_rate": 4.005128205128205e-07, + "loss": 0.0001, + "step": 19225 + }, + { + "epoch": 16.27906976744186, + "grad_norm": 0.004281258676201105, + "learning_rate": 3.876923076923077e-07, + "loss": 0.0001, + "step": 19250 + }, + { + "epoch": 16.300211416490487, + "grad_norm": 0.003360031172633171, + "learning_rate": 3.748717948717949e-07, + "loss": 0.0001, + "step": 19275 + }, + { + "epoch": 16.32135306553911, + "grad_norm": 0.005053969100117683, + "learning_rate": 3.6205128205128206e-07, + "loss": 0.0001, + "step": 19300 + }, + { + "epoch": 16.342494714587737, + "grad_norm": 0.005359134636819363, + "learning_rate": 3.4923076923076925e-07, + "loss": 0.0001, + "step": 19325 + }, + { + "epoch": 16.363636363636363, + "grad_norm": 0.004870879463851452, + "learning_rate": 3.3641025641025645e-07, + "loss": 0.0001, + "step": 19350 + }, + { + "epoch": 16.38477801268499, + "grad_norm": 0.005147048272192478, + "learning_rate": 3.235897435897436e-07, + "loss": 0.0001, + "step": 19375 + }, + { + "epoch": 16.405919661733616, + "grad_norm": 0.005027804523706436, + "learning_rate": 3.107692307692308e-07, + "loss": 0.0001, + "step": 19400 + }, + { + "epoch": 16.427061310782243, + "grad_norm": 0.012405909597873688, + "learning_rate": 2.97948717948718e-07, + "loss": 0.0001, + "step": 19425 + }, + { + "epoch": 16.448202959830866, + "grad_norm": 0.004649542272090912, + "learning_rate": 2.8512820512820517e-07, + "loss": 0.0001, + "step": 19450 + }, + { + "epoch": 16.469344608879492, + "grad_norm": 0.004960506223142147, + "learning_rate": 2.7230769230769237e-07, + "loss": 0.0001, + "step": 19475 + }, + { + "epoch": 16.49048625792812, + "grad_norm": 0.00558600015938282, + "learning_rate": 2.594871794871795e-07, + "loss": 0.0001, + "step": 19500 + }, + { + "epoch": 16.511627906976745, + "grad_norm": 0.00477565685287118, + "learning_rate": 2.466666666666667e-07, + "loss": 0.0001, + "step": 19525 + }, + { + "epoch": 16.53276955602537, + "grad_norm": 0.007015578914433718, + "learning_rate": 2.3384615384615387e-07, + "loss": 0.0001, + "step": 19550 + }, + { + "epoch": 16.553911205073994, + "grad_norm": 0.0036124507896602154, + "learning_rate": 2.2102564102564104e-07, + "loss": 0.0001, + "step": 19575 + }, + { + "epoch": 16.57505285412262, + "grad_norm": 0.005142053589224815, + "learning_rate": 2.082051282051282e-07, + "loss": 0.0001, + "step": 19600 + }, + { + "epoch": 16.596194503171247, + "grad_norm": 0.005827662535011768, + "learning_rate": 1.953846153846154e-07, + "loss": 0.0001, + "step": 19625 + }, + { + "epoch": 16.617336152219874, + "grad_norm": 0.006833038758486509, + "learning_rate": 1.8256410256410257e-07, + "loss": 0.0001, + "step": 19650 + }, + { + "epoch": 16.6384778012685, + "grad_norm": 0.0065496135503053665, + "learning_rate": 1.6974358974358974e-07, + "loss": 0.0001, + "step": 19675 + }, + { + "epoch": 16.659619450317123, + "grad_norm": 0.0090131014585495, + "learning_rate": 1.5692307692307694e-07, + "loss": 0.0001, + "step": 19700 + }, + { + "epoch": 16.68076109936575, + "grad_norm": 0.0042783962562680244, + "learning_rate": 1.441025641025641e-07, + "loss": 0.0001, + "step": 19725 + }, + { + "epoch": 16.701902748414376, + "grad_norm": 0.005526359658688307, + "learning_rate": 1.3128205128205127e-07, + "loss": 0.0001, + "step": 19750 + }, + { + "epoch": 16.723044397463003, + "grad_norm": 0.004802173003554344, + "learning_rate": 1.1846153846153847e-07, + "loss": 0.0001, + "step": 19775 + }, + { + "epoch": 16.74418604651163, + "grad_norm": 0.003740393090993166, + "learning_rate": 1.0564102564102565e-07, + "loss": 0.0001, + "step": 19800 + }, + { + "epoch": 16.765327695560252, + "grad_norm": 0.004182030912488699, + "learning_rate": 9.282051282051283e-08, + "loss": 0.0001, + "step": 19825 + }, + { + "epoch": 16.78646934460888, + "grad_norm": 0.004386850632727146, + "learning_rate": 8e-08, + "loss": 0.0001, + "step": 19850 + }, + { + "epoch": 16.807610993657505, + "grad_norm": 0.005020872224122286, + "learning_rate": 6.717948717948718e-08, + "loss": 0.0001, + "step": 19875 + }, + { + "epoch": 16.82875264270613, + "grad_norm": 0.0037505626678466797, + "learning_rate": 5.435897435897436e-08, + "loss": 0.0001, + "step": 19900 + }, + { + "epoch": 16.849894291754758, + "grad_norm": 0.004394158720970154, + "learning_rate": 4.153846153846154e-08, + "loss": 0.0001, + "step": 19925 + }, + { + "epoch": 16.87103594080338, + "grad_norm": 0.004075503908097744, + "learning_rate": 2.8717948717948722e-08, + "loss": 0.0001, + "step": 19950 + }, + { + "epoch": 16.892177589852007, + "grad_norm": 0.0045666322112083435, + "learning_rate": 1.5897435897435897e-08, + "loss": 0.0001, + "step": 19975 + }, + { + "epoch": 16.913319238900634, + "grad_norm": 0.006132797803729773, + "learning_rate": 3.0769230769230774e-09, + "loss": 0.0001, + "step": 20000 + }, + { + "epoch": 16.913319238900634, + "eval_loss": 0.08780684322118759, + "eval_runtime": 420.9914, + "eval_samples_per_second": 8.649, + "eval_steps_per_second": 0.542, + "eval_wer": 0.05902826117221217, + "step": 20000 + }, + { + "epoch": 16.913319238900634, + "step": 20000, + "total_flos": 1.8463463044153344e+20, + "train_loss": 0.02644415222366224, + "train_runtime": 72777.8888, + "train_samples_per_second": 8.794, "train_steps_per_second": 0.275 } ], "logging_steps": 25, - "max_steps": 5000, + "max_steps": 20000, "num_input_tokens_seen": 0, - "num_train_epochs": 20, + "num_train_epochs": 17, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { @@ -1480,7 +5815,7 @@ "attributes": {} } }, - "total_flos": 4.616818092859392e+19, + "total_flos": 1.8463463044153344e+20, "train_batch_size": 16, "trial_name": null, "trial_params": null