| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 100, | |
| "global_step": 2320, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08620689655172414, | |
| "grad_norm": 9.590513229370117, | |
| "learning_rate": 9e-07, | |
| "loss": 9.1142, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 9.729034423828125, | |
| "learning_rate": 1.9e-06, | |
| "loss": 8.3444, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.25862068965517243, | |
| "grad_norm": 14.263296127319336, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 8.6571, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 15.165566444396973, | |
| "learning_rate": 3.8e-06, | |
| "loss": 7.6933, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 16.664026260375977, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 6.9526, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 17.301790237426758, | |
| "learning_rate": 5.8e-06, | |
| "loss": 6.2005, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.603448275862069, | |
| "grad_norm": 11.270517349243164, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 4.7052, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 4.181836128234863, | |
| "learning_rate": 7.8e-06, | |
| "loss": 3.7839, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7758620689655172, | |
| "grad_norm": 1.8710567951202393, | |
| "learning_rate": 8.8e-06, | |
| "loss": 3.3947, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 1.407578706741333, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 3.1612, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "eval_loss": 3.118124485015869, | |
| "eval_runtime": 39.9765, | |
| "eval_samples_per_second": 33.62, | |
| "eval_steps_per_second": 33.62, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9482758620689655, | |
| "grad_norm": 1.6558985710144043, | |
| "learning_rate": 1.08e-05, | |
| "loss": 3.0858, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 1.8191890716552734, | |
| "learning_rate": 1.18e-05, | |
| "loss": 3.0791, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1206896551724137, | |
| "grad_norm": 0.9610480070114136, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 3.034, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 0.7719367742538452, | |
| "learning_rate": 1.3800000000000002e-05, | |
| "loss": 3.0098, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.293103448275862, | |
| "grad_norm": 0.7965385913848877, | |
| "learning_rate": 1.48e-05, | |
| "loss": 3.0138, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.9387674927711487, | |
| "learning_rate": 1.58e-05, | |
| "loss": 2.9973, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4655172413793103, | |
| "grad_norm": 0.8531718254089355, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 2.9875, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 0.5282021760940552, | |
| "learning_rate": 1.78e-05, | |
| "loss": 2.9905, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.6379310344827587, | |
| "grad_norm": 1.1362160444259644, | |
| "learning_rate": 1.88e-05, | |
| "loss": 2.9881, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 1.0381989479064941, | |
| "learning_rate": 1.9800000000000004e-05, | |
| "loss": 2.978, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "eval_loss": 2.9721522331237793, | |
| "eval_runtime": 39.4986, | |
| "eval_samples_per_second": 34.026, | |
| "eval_steps_per_second": 34.026, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.8103448275862069, | |
| "grad_norm": 1.0819815397262573, | |
| "learning_rate": 2.08e-05, | |
| "loss": 2.9728, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.5016360878944397, | |
| "learning_rate": 2.18e-05, | |
| "loss": 2.9769, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9827586206896552, | |
| "grad_norm": 1.476746678352356, | |
| "learning_rate": 2.2800000000000002e-05, | |
| "loss": 2.9638, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 0.416255384683609, | |
| "learning_rate": 2.38e-05, | |
| "loss": 2.9566, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1551724137931036, | |
| "grad_norm": 0.5918602347373962, | |
| "learning_rate": 2.48e-05, | |
| "loss": 2.9507, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.2413793103448274, | |
| "grad_norm": 0.8085893988609314, | |
| "learning_rate": 2.58e-05, | |
| "loss": 2.9469, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.3275862068965516, | |
| "grad_norm": 0.8484460115432739, | |
| "learning_rate": 2.6800000000000004e-05, | |
| "loss": 2.9438, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 0.4717480540275574, | |
| "learning_rate": 2.7800000000000005e-05, | |
| "loss": 2.9421, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.0097577571868896, | |
| "learning_rate": 2.88e-05, | |
| "loss": 2.9234, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 1.3988488912582397, | |
| "learning_rate": 2.98e-05, | |
| "loss": 2.9185, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "eval_loss": 2.909769058227539, | |
| "eval_runtime": 39.5502, | |
| "eval_samples_per_second": 33.982, | |
| "eval_steps_per_second": 33.982, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.6724137931034484, | |
| "grad_norm": 2.138706922531128, | |
| "learning_rate": 3.08e-05, | |
| "loss": 2.9028, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 0.854528546333313, | |
| "learning_rate": 3.18e-05, | |
| "loss": 2.8889, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.844827586206897, | |
| "grad_norm": 0.8741695284843445, | |
| "learning_rate": 3.2800000000000004e-05, | |
| "loss": 2.8698, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.9310344827586206, | |
| "grad_norm": 2.8528709411621094, | |
| "learning_rate": 3.38e-05, | |
| "loss": 2.7851, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.0172413793103448, | |
| "grad_norm": 2.0051960945129395, | |
| "learning_rate": 3.48e-05, | |
| "loss": 2.7537, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.103448275862069, | |
| "grad_norm": 1.6813557147979736, | |
| "learning_rate": 3.58e-05, | |
| "loss": 2.6125, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.189655172413793, | |
| "grad_norm": 2.679445266723633, | |
| "learning_rate": 3.68e-05, | |
| "loss": 2.4925, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.2758620689655173, | |
| "grad_norm": 1.8131614923477173, | |
| "learning_rate": 3.7800000000000004e-05, | |
| "loss": 2.4007, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.3620689655172415, | |
| "grad_norm": 2.168797016143799, | |
| "learning_rate": 3.88e-05, | |
| "loss": 2.2533, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "grad_norm": 2.1998753547668457, | |
| "learning_rate": 3.9800000000000005e-05, | |
| "loss": 2.1282, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "eval_loss": 2.0066494941711426, | |
| "eval_runtime": 39.5789, | |
| "eval_samples_per_second": 33.958, | |
| "eval_steps_per_second": 33.958, | |
| "eval_wer": 1.0246744969498316, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.5344827586206895, | |
| "grad_norm": 4.27025842666626, | |
| "learning_rate": 4.08e-05, | |
| "loss": 2.0277, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.6206896551724137, | |
| "grad_norm": 2.3157191276550293, | |
| "learning_rate": 4.18e-05, | |
| "loss": 1.8844, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.706896551724138, | |
| "grad_norm": 3.1167590618133545, | |
| "learning_rate": 4.2800000000000004e-05, | |
| "loss": 1.7777, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.793103448275862, | |
| "grad_norm": 4.118265151977539, | |
| "learning_rate": 4.38e-05, | |
| "loss": 1.7015, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.8793103448275863, | |
| "grad_norm": 2.6515376567840576, | |
| "learning_rate": 4.4800000000000005e-05, | |
| "loss": 1.5855, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.9655172413793105, | |
| "grad_norm": 3.6305439472198486, | |
| "learning_rate": 4.58e-05, | |
| "loss": 1.5015, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.051724137931035, | |
| "grad_norm": 1.8111392259597778, | |
| "learning_rate": 4.6800000000000006e-05, | |
| "loss": 1.3926, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 2.117809295654297, | |
| "learning_rate": 4.78e-05, | |
| "loss": 1.2363, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.224137931034483, | |
| "grad_norm": 2.7456188201904297, | |
| "learning_rate": 4.88e-05, | |
| "loss": 1.156, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.310344827586207, | |
| "grad_norm": 2.3859684467315674, | |
| "learning_rate": 4.9800000000000004e-05, | |
| "loss": 1.1234, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.310344827586207, | |
| "eval_loss": 1.0196667909622192, | |
| "eval_runtime": 39.5574, | |
| "eval_samples_per_second": 33.976, | |
| "eval_steps_per_second": 33.976, | |
| "eval_wer": 0.8392970955112447, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.396551724137931, | |
| "grad_norm": 2.3013603687286377, | |
| "learning_rate": 5.08e-05, | |
| "loss": 1.0335, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.482758620689655, | |
| "grad_norm": 5.266668796539307, | |
| "learning_rate": 5.1800000000000005e-05, | |
| "loss": 1.002, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.568965517241379, | |
| "grad_norm": 2.516477346420288, | |
| "learning_rate": 5.28e-05, | |
| "loss": 0.915, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.655172413793103, | |
| "grad_norm": 2.389324903488159, | |
| "learning_rate": 5.380000000000001e-05, | |
| "loss": 0.8828, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.741379310344827, | |
| "grad_norm": 2.378732204437256, | |
| "learning_rate": 5.4800000000000004e-05, | |
| "loss": 0.8822, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.827586206896552, | |
| "grad_norm": 2.126372814178467, | |
| "learning_rate": 5.580000000000001e-05, | |
| "loss": 0.7924, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.913793103448276, | |
| "grad_norm": 2.4130356311798096, | |
| "learning_rate": 5.68e-05, | |
| "loss": 0.7577, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.412507057189941, | |
| "learning_rate": 5.7799999999999995e-05, | |
| "loss": 0.8124, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 5.086206896551724, | |
| "grad_norm": 2.4421167373657227, | |
| "learning_rate": 5.88e-05, | |
| "loss": 0.6599, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.172413793103448, | |
| "grad_norm": 2.1765213012695312, | |
| "learning_rate": 5.9800000000000003e-05, | |
| "loss": 0.602, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.172413793103448, | |
| "eval_loss": 0.6714467406272888, | |
| "eval_runtime": 39.6477, | |
| "eval_samples_per_second": 33.899, | |
| "eval_steps_per_second": 33.899, | |
| "eval_wer": 0.6600200309569334, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.258620689655173, | |
| "grad_norm": 2.5929956436157227, | |
| "learning_rate": 6.08e-05, | |
| "loss": 0.6666, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.344827586206897, | |
| "grad_norm": 2.023226261138916, | |
| "learning_rate": 6.18e-05, | |
| "loss": 0.557, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.431034482758621, | |
| "grad_norm": 2.348583936691284, | |
| "learning_rate": 6.280000000000001e-05, | |
| "loss": 0.536, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.517241379310345, | |
| "grad_norm": 1.9607300758361816, | |
| "learning_rate": 6.38e-05, | |
| "loss": 0.618, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.603448275862069, | |
| "grad_norm": 2.136683940887451, | |
| "learning_rate": 6.48e-05, | |
| "loss": 0.5107, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.689655172413794, | |
| "grad_norm": 2.5568346977233887, | |
| "learning_rate": 6.58e-05, | |
| "loss": 0.512, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.775862068965517, | |
| "grad_norm": 1.6805388927459717, | |
| "learning_rate": 6.680000000000001e-05, | |
| "loss": 0.5663, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.862068965517241, | |
| "grad_norm": 3.271713972091675, | |
| "learning_rate": 6.780000000000001e-05, | |
| "loss": 0.4748, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.948275862068965, | |
| "grad_norm": 12.830371856689453, | |
| "learning_rate": 6.879999999999999e-05, | |
| "loss": 0.4823, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 6.0344827586206895, | |
| "grad_norm": 1.5290584564208984, | |
| "learning_rate": 6.98e-05, | |
| "loss": 0.5032, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.0344827586206895, | |
| "eval_loss": 0.5284600257873535, | |
| "eval_runtime": 39.6104, | |
| "eval_samples_per_second": 33.931, | |
| "eval_steps_per_second": 33.931, | |
| "eval_wer": 0.5658745333697532, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.120689655172414, | |
| "grad_norm": 1.902786374092102, | |
| "learning_rate": 7.08e-05, | |
| "loss": 0.387, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.206896551724138, | |
| "grad_norm": 2.5520503520965576, | |
| "learning_rate": 7.18e-05, | |
| "loss": 0.3752, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.293103448275862, | |
| "grad_norm": 1.8766177892684937, | |
| "learning_rate": 7.280000000000001e-05, | |
| "loss": 0.422, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.379310344827586, | |
| "grad_norm": 1.9960404634475708, | |
| "learning_rate": 7.38e-05, | |
| "loss": 0.3703, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.4655172413793105, | |
| "grad_norm": 2.9510915279388428, | |
| "learning_rate": 7.48e-05, | |
| "loss": 0.3777, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.551724137931035, | |
| "grad_norm": 1.5135278701782227, | |
| "learning_rate": 7.58e-05, | |
| "loss": 0.3978, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.637931034482759, | |
| "grad_norm": 2.198090076446533, | |
| "learning_rate": 7.680000000000001e-05, | |
| "loss": 0.306, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.724137931034483, | |
| "grad_norm": 4.511343955993652, | |
| "learning_rate": 7.780000000000001e-05, | |
| "loss": 0.4127, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.810344827586206, | |
| "grad_norm": 1.5059940814971924, | |
| "learning_rate": 7.88e-05, | |
| "loss": 0.3259, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 2.252959728240967, | |
| "learning_rate": 7.98e-05, | |
| "loss": 0.3101, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "eval_loss": 0.48188281059265137, | |
| "eval_runtime": 39.7499, | |
| "eval_samples_per_second": 33.811, | |
| "eval_steps_per_second": 33.811, | |
| "eval_wer": 0.528179914413184, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.982758620689655, | |
| "grad_norm": 2.095763683319092, | |
| "learning_rate": 8.080000000000001e-05, | |
| "loss": 0.3926, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 7.068965517241379, | |
| "grad_norm": 1.679442048072815, | |
| "learning_rate": 8.18e-05, | |
| "loss": 0.3035, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 7.155172413793103, | |
| "grad_norm": 1.725831151008606, | |
| "learning_rate": 8.28e-05, | |
| "loss": 0.2546, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 7.241379310344827, | |
| "grad_norm": 4.802426338195801, | |
| "learning_rate": 8.38e-05, | |
| "loss": 0.3201, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.327586206896552, | |
| "grad_norm": 1.871408462524414, | |
| "learning_rate": 8.48e-05, | |
| "loss": 0.2891, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.413793103448276, | |
| "grad_norm": 2.0499789714813232, | |
| "learning_rate": 8.58e-05, | |
| "loss": 0.2513, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.3894842863082886, | |
| "learning_rate": 8.680000000000001e-05, | |
| "loss": 0.3401, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.586206896551724, | |
| "grad_norm": 1.56522798538208, | |
| "learning_rate": 8.78e-05, | |
| "loss": 0.2398, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.672413793103448, | |
| "grad_norm": 1.908964991569519, | |
| "learning_rate": 8.88e-05, | |
| "loss": 0.2656, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.758620689655173, | |
| "grad_norm": 1.5925772190093994, | |
| "learning_rate": 8.98e-05, | |
| "loss": 0.3432, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.758620689655173, | |
| "eval_loss": 0.46526795625686646, | |
| "eval_runtime": 39.52, | |
| "eval_samples_per_second": 34.008, | |
| "eval_steps_per_second": 34.008, | |
| "eval_wer": 0.5271783665665118, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.844827586206897, | |
| "grad_norm": 1.7011960744857788, | |
| "learning_rate": 9.080000000000001e-05, | |
| "loss": 0.2539, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 7.931034482758621, | |
| "grad_norm": 1.718232274055481, | |
| "learning_rate": 9.180000000000001e-05, | |
| "loss": 0.2653, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 8.017241379310345, | |
| "grad_norm": 1.6879725456237793, | |
| "learning_rate": 9.28e-05, | |
| "loss": 0.3313, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 8.10344827586207, | |
| "grad_norm": 2.626966714859009, | |
| "learning_rate": 9.38e-05, | |
| "loss": 0.2116, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 8.189655172413794, | |
| "grad_norm": 2.0977375507354736, | |
| "learning_rate": 9.48e-05, | |
| "loss": 0.2445, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.275862068965518, | |
| "grad_norm": 1.6694329977035522, | |
| "learning_rate": 9.58e-05, | |
| "loss": 0.2928, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.362068965517242, | |
| "grad_norm": 1.8011162281036377, | |
| "learning_rate": 9.680000000000001e-05, | |
| "loss": 0.2008, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.448275862068966, | |
| "grad_norm": 2.6468820571899414, | |
| "learning_rate": 9.78e-05, | |
| "loss": 0.2089, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.53448275862069, | |
| "grad_norm": 1.5480060577392578, | |
| "learning_rate": 9.88e-05, | |
| "loss": 0.2593, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.620689655172415, | |
| "grad_norm": 2.0152957439422607, | |
| "learning_rate": 9.98e-05, | |
| "loss": 0.1922, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.620689655172415, | |
| "eval_loss": 0.46724727749824524, | |
| "eval_runtime": 39.7331, | |
| "eval_samples_per_second": 33.826, | |
| "eval_steps_per_second": 33.826, | |
| "eval_wer": 0.49175999271601567, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.706896551724139, | |
| "grad_norm": 2.7271480560302734, | |
| "learning_rate": 9.939393939393939e-05, | |
| "loss": 0.2466, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.793103448275861, | |
| "grad_norm": 1.4664121866226196, | |
| "learning_rate": 9.863636363636364e-05, | |
| "loss": 0.259, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 8.879310344827585, | |
| "grad_norm": 2.0848429203033447, | |
| "learning_rate": 9.787878787878789e-05, | |
| "loss": 0.1964, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 8.96551724137931, | |
| "grad_norm": 2.68035888671875, | |
| "learning_rate": 9.712121212121212e-05, | |
| "loss": 0.2528, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 9.051724137931034, | |
| "grad_norm": 1.2563536167144775, | |
| "learning_rate": 9.636363636363637e-05, | |
| "loss": 0.2322, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 9.137931034482758, | |
| "grad_norm": 1.8890479803085327, | |
| "learning_rate": 9.560606060606061e-05, | |
| "loss": 0.1597, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 9.224137931034482, | |
| "grad_norm": 2.1262826919555664, | |
| "learning_rate": 9.484848484848486e-05, | |
| "loss": 0.2046, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 9.310344827586206, | |
| "grad_norm": 1.9750289916992188, | |
| "learning_rate": 9.40909090909091e-05, | |
| "loss": 0.2035, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.39655172413793, | |
| "grad_norm": 1.9659441709518433, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 0.1623, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.482758620689655, | |
| "grad_norm": 3.0013935565948486, | |
| "learning_rate": 9.257575757575758e-05, | |
| "loss": 0.2284, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.482758620689655, | |
| "eval_loss": 0.48344284296035767, | |
| "eval_runtime": 39.5675, | |
| "eval_samples_per_second": 33.967, | |
| "eval_steps_per_second": 33.967, | |
| "eval_wer": 0.48702540289538376, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.568965517241379, | |
| "grad_norm": 1.5623351335525513, | |
| "learning_rate": 9.181818181818183e-05, | |
| "loss": 0.182, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.655172413793103, | |
| "grad_norm": 1.5138722658157349, | |
| "learning_rate": 9.106060606060606e-05, | |
| "loss": 0.1741, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.741379310344827, | |
| "grad_norm": 1.3799549341201782, | |
| "learning_rate": 9.030303030303031e-05, | |
| "loss": 0.2227, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 9.827586206896552, | |
| "grad_norm": 1.6967511177062988, | |
| "learning_rate": 8.954545454545455e-05, | |
| "loss": 0.1617, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 9.913793103448276, | |
| "grad_norm": 2.4861340522766113, | |
| "learning_rate": 8.87878787878788e-05, | |
| "loss": 0.1712, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 4.324175834655762, | |
| "learning_rate": 8.803030303030304e-05, | |
| "loss": 0.2196, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 10.086206896551724, | |
| "grad_norm": 1.889611840248108, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.1392, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 10.172413793103448, | |
| "grad_norm": 1.7731642723083496, | |
| "learning_rate": 8.651515151515152e-05, | |
| "loss": 0.147, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 10.258620689655173, | |
| "grad_norm": 1.3258129358291626, | |
| "learning_rate": 8.575757575757576e-05, | |
| "loss": 0.1793, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 10.344827586206897, | |
| "grad_norm": 2.139970302581787, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.1372, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.344827586206897, | |
| "eval_loss": 0.43795427680015564, | |
| "eval_runtime": 39.566, | |
| "eval_samples_per_second": 33.969, | |
| "eval_steps_per_second": 33.969, | |
| "eval_wer": 0.4727305836292452, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.431034482758621, | |
| "grad_norm": 2.1916654109954834, | |
| "learning_rate": 8.424242424242424e-05, | |
| "loss": 0.1442, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 10.517241379310345, | |
| "grad_norm": 1.1983482837677002, | |
| "learning_rate": 8.348484848484849e-05, | |
| "loss": 0.1926, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.60344827586207, | |
| "grad_norm": 1.9001710414886475, | |
| "learning_rate": 8.272727272727273e-05, | |
| "loss": 0.1368, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 10.689655172413794, | |
| "grad_norm": 1.9396876096725464, | |
| "learning_rate": 8.196969696969698e-05, | |
| "loss": 0.131, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 10.775862068965518, | |
| "grad_norm": 1.2405284643173218, | |
| "learning_rate": 8.121212121212121e-05, | |
| "loss": 0.162, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 10.862068965517242, | |
| "grad_norm": 1.1841609477996826, | |
| "learning_rate": 8.045454545454546e-05, | |
| "loss": 0.1199, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 10.948275862068966, | |
| "grad_norm": 1.863655924797058, | |
| "learning_rate": 7.96969696969697e-05, | |
| "loss": 0.1571, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 11.03448275862069, | |
| "grad_norm": 1.102363109588623, | |
| "learning_rate": 7.893939393939395e-05, | |
| "loss": 0.1672, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 11.120689655172415, | |
| "grad_norm": 1.411650538444519, | |
| "learning_rate": 7.818181818181818e-05, | |
| "loss": 0.1004, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 11.206896551724139, | |
| "grad_norm": 1.5993527173995972, | |
| "learning_rate": 7.742424242424243e-05, | |
| "loss": 0.1105, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.206896551724139, | |
| "eval_loss": 0.45091673731803894, | |
| "eval_runtime": 39.5057, | |
| "eval_samples_per_second": 34.02, | |
| "eval_steps_per_second": 34.02, | |
| "eval_wer": 0.45943731220977874, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.293103448275861, | |
| "grad_norm": 0.9610430598258972, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 0.1514, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 11.379310344827585, | |
| "grad_norm": 1.1394908428192139, | |
| "learning_rate": 7.59090909090909e-05, | |
| "loss": 0.1087, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 11.46551724137931, | |
| "grad_norm": 2.085495948791504, | |
| "learning_rate": 7.515151515151515e-05, | |
| "loss": 0.1328, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 11.551724137931034, | |
| "grad_norm": 1.0012094974517822, | |
| "learning_rate": 7.439393939393939e-05, | |
| "loss": 0.1406, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 11.637931034482758, | |
| "grad_norm": 1.3768068552017212, | |
| "learning_rate": 7.363636363636364e-05, | |
| "loss": 0.1068, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 11.724137931034482, | |
| "grad_norm": 3.6142425537109375, | |
| "learning_rate": 7.287878787878788e-05, | |
| "loss": 0.1447, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 11.810344827586206, | |
| "grad_norm": 1.16116201877594, | |
| "learning_rate": 7.212121212121213e-05, | |
| "loss": 0.1099, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 11.89655172413793, | |
| "grad_norm": 1.4963595867156982, | |
| "learning_rate": 7.136363636363636e-05, | |
| "loss": 0.098, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 11.982758620689655, | |
| "grad_norm": 1.1868886947631836, | |
| "learning_rate": 7.060606060606061e-05, | |
| "loss": 0.1603, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 12.068965517241379, | |
| "grad_norm": 1.6679799556732178, | |
| "learning_rate": 6.984848484848485e-05, | |
| "loss": 0.0992, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 12.068965517241379, | |
| "eval_loss": 0.41960862278938293, | |
| "eval_runtime": 39.5319, | |
| "eval_samples_per_second": 33.998, | |
| "eval_steps_per_second": 33.998, | |
| "eval_wer": 0.4544295729764181, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 12.155172413793103, | |
| "grad_norm": 0.9624470472335815, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.092, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 12.241379310344827, | |
| "grad_norm": 3.408127546310425, | |
| "learning_rate": 6.833333333333333e-05, | |
| "loss": 0.1716, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 12.327586206896552, | |
| "grad_norm": 1.4163497686386108, | |
| "learning_rate": 6.757575757575758e-05, | |
| "loss": 0.0885, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 12.413793103448276, | |
| "grad_norm": 1.494627594947815, | |
| "learning_rate": 6.681818181818183e-05, | |
| "loss": 0.0891, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 1.1766793727874756, | |
| "learning_rate": 6.606060606060607e-05, | |
| "loss": 0.138, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 12.586206896551724, | |
| "grad_norm": 1.2334414720535278, | |
| "learning_rate": 6.530303030303032e-05, | |
| "loss": 0.0946, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 12.672413793103448, | |
| "grad_norm": 2.576019763946533, | |
| "learning_rate": 6.454545454545455e-05, | |
| "loss": 0.1043, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 12.758620689655173, | |
| "grad_norm": 0.9484899640083313, | |
| "learning_rate": 6.37878787878788e-05, | |
| "loss": 0.1155, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 12.844827586206897, | |
| "grad_norm": 1.0187716484069824, | |
| "learning_rate": 6.303030303030302e-05, | |
| "loss": 0.0828, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 12.931034482758621, | |
| "grad_norm": 1.0715346336364746, | |
| "learning_rate": 6.227272727272727e-05, | |
| "loss": 0.1226, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 12.931034482758621, | |
| "eval_loss": 0.4237208664417267, | |
| "eval_runtime": 39.6767, | |
| "eval_samples_per_second": 33.874, | |
| "eval_steps_per_second": 33.874, | |
| "eval_wer": 0.4321223709369025, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.017241379310345, | |
| "grad_norm": 0.8969095945358276, | |
| "learning_rate": 6.151515151515151e-05, | |
| "loss": 0.1272, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 13.10344827586207, | |
| "grad_norm": 1.4273531436920166, | |
| "learning_rate": 6.075757575757576e-05, | |
| "loss": 0.0804, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 13.189655172413794, | |
| "grad_norm": 1.5268211364746094, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0907, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 13.275862068965518, | |
| "grad_norm": 1.1891446113586426, | |
| "learning_rate": 5.9242424242424244e-05, | |
| "loss": 0.122, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 13.362068965517242, | |
| "grad_norm": 1.2826952934265137, | |
| "learning_rate": 5.848484848484849e-05, | |
| "loss": 0.079, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 13.448275862068966, | |
| "grad_norm": 1.4693151712417603, | |
| "learning_rate": 5.772727272727273e-05, | |
| "loss": 0.0953, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 13.53448275862069, | |
| "grad_norm": 1.1638069152832031, | |
| "learning_rate": 5.696969696969697e-05, | |
| "loss": 0.1076, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 13.620689655172415, | |
| "grad_norm": 2.33617901802063, | |
| "learning_rate": 5.6212121212121215e-05, | |
| "loss": 0.0712, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 13.706896551724139, | |
| "grad_norm": 2.3127667903900146, | |
| "learning_rate": 5.545454545454546e-05, | |
| "loss": 0.0947, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "grad_norm": 0.9676732420921326, | |
| "learning_rate": 5.46969696969697e-05, | |
| "loss": 0.1013, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "eval_loss": 0.4112694263458252, | |
| "eval_runtime": 39.48, | |
| "eval_samples_per_second": 34.043, | |
| "eval_steps_per_second": 34.043, | |
| "eval_wer": 0.4298461258308295, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.879310344827585, | |
| "grad_norm": 1.8281100988388062, | |
| "learning_rate": 5.393939393939394e-05, | |
| "loss": 0.0696, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 13.96551724137931, | |
| "grad_norm": 2.354693651199341, | |
| "learning_rate": 5.3181818181818186e-05, | |
| "loss": 0.1129, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 14.051724137931034, | |
| "grad_norm": 1.0465914011001587, | |
| "learning_rate": 5.242424242424243e-05, | |
| "loss": 0.0744, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 14.137931034482758, | |
| "grad_norm": 1.0246185064315796, | |
| "learning_rate": 5.166666666666667e-05, | |
| "loss": 0.0656, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 14.224137931034482, | |
| "grad_norm": 2.083202838897705, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.0869, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 14.310344827586206, | |
| "grad_norm": 1.176715612411499, | |
| "learning_rate": 5.015151515151515e-05, | |
| "loss": 0.0821, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 14.39655172413793, | |
| "grad_norm": 1.6385036706924438, | |
| "learning_rate": 4.93939393939394e-05, | |
| "loss": 0.0695, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 14.482758620689655, | |
| "grad_norm": 2.735548973083496, | |
| "learning_rate": 4.863636363636364e-05, | |
| "loss": 0.0875, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 14.568965517241379, | |
| "grad_norm": 1.3907177448272705, | |
| "learning_rate": 4.787878787878788e-05, | |
| "loss": 0.0773, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 14.655172413793103, | |
| "grad_norm": 1.582645297050476, | |
| "learning_rate": 4.712121212121212e-05, | |
| "loss": 0.0661, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 14.655172413793103, | |
| "eval_loss": 0.4037941098213196, | |
| "eval_runtime": 39.716, | |
| "eval_samples_per_second": 33.84, | |
| "eval_steps_per_second": 33.84, | |
| "eval_wer": 0.42756988072475643, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 14.741379310344827, | |
| "grad_norm": 1.1024818420410156, | |
| "learning_rate": 4.636363636363636e-05, | |
| "loss": 0.1285, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 14.827586206896552, | |
| "grad_norm": 1.0617228746414185, | |
| "learning_rate": 4.5606060606060606e-05, | |
| "loss": 0.0638, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 14.913793103448276, | |
| "grad_norm": 1.8958903551101685, | |
| "learning_rate": 4.484848484848485e-05, | |
| "loss": 0.0772, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 2.5858068466186523, | |
| "learning_rate": 4.409090909090909e-05, | |
| "loss": 0.094, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 15.086206896551724, | |
| "grad_norm": 0.6185240745544434, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 0.0506, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 15.172413793103448, | |
| "grad_norm": 1.9594557285308838, | |
| "learning_rate": 4.257575757575758e-05, | |
| "loss": 0.0667, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 15.258620689655173, | |
| "grad_norm": 0.6401175260543823, | |
| "learning_rate": 4.181818181818182e-05, | |
| "loss": 0.0909, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 15.344827586206897, | |
| "grad_norm": 1.2541193962097168, | |
| "learning_rate": 4.106060606060606e-05, | |
| "loss": 0.0573, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 15.431034482758621, | |
| "grad_norm": 1.74609375, | |
| "learning_rate": 4.0303030303030305e-05, | |
| "loss": 0.0666, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 15.517241379310345, | |
| "grad_norm": 1.3555305004119873, | |
| "learning_rate": 3.954545454545455e-05, | |
| "loss": 0.0901, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.517241379310345, | |
| "eval_loss": 0.43212294578552246, | |
| "eval_runtime": 40.1884, | |
| "eval_samples_per_second": 33.443, | |
| "eval_steps_per_second": 33.443, | |
| "eval_wer": 0.42247109168715286, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.60344827586207, | |
| "grad_norm": 0.8425617218017578, | |
| "learning_rate": 3.878787878787879e-05, | |
| "loss": 0.0567, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 15.689655172413794, | |
| "grad_norm": 1.145479679107666, | |
| "learning_rate": 3.803030303030303e-05, | |
| "loss": 0.0682, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 15.775862068965518, | |
| "grad_norm": 1.0084822177886963, | |
| "learning_rate": 3.7272727272727276e-05, | |
| "loss": 0.0847, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.862068965517242, | |
| "grad_norm": 0.986998438835144, | |
| "learning_rate": 3.651515151515152e-05, | |
| "loss": 0.0623, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 15.948275862068966, | |
| "grad_norm": 1.5837907791137695, | |
| "learning_rate": 3.575757575757576e-05, | |
| "loss": 0.0719, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 16.03448275862069, | |
| "grad_norm": 0.8865498900413513, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.0766, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 16.120689655172413, | |
| "grad_norm": 1.3217542171478271, | |
| "learning_rate": 3.424242424242424e-05, | |
| "loss": 0.0598, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 16.20689655172414, | |
| "grad_norm": 1.5297330617904663, | |
| "learning_rate": 3.348484848484848e-05, | |
| "loss": 0.0632, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 16.29310344827586, | |
| "grad_norm": 0.6415010094642639, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 0.0732, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 16.379310344827587, | |
| "grad_norm": 2.088172435760498, | |
| "learning_rate": 3.1969696969696974e-05, | |
| "loss": 0.053, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.379310344827587, | |
| "eval_loss": 0.4076109826564789, | |
| "eval_runtime": 40.2143, | |
| "eval_samples_per_second": 33.421, | |
| "eval_steps_per_second": 33.421, | |
| "eval_wer": 0.42356368933806793, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.46551724137931, | |
| "grad_norm": 1.5730756521224976, | |
| "learning_rate": 3.121212121212122e-05, | |
| "loss": 0.0722, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 16.551724137931036, | |
| "grad_norm": 0.7452704906463623, | |
| "learning_rate": 3.0454545454545456e-05, | |
| "loss": 0.0743, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 16.637931034482758, | |
| "grad_norm": 0.890084445476532, | |
| "learning_rate": 2.96969696969697e-05, | |
| "loss": 0.0462, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 16.724137931034484, | |
| "grad_norm": 2.400942325592041, | |
| "learning_rate": 2.893939393939394e-05, | |
| "loss": 0.076, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 16.810344827586206, | |
| "grad_norm": 1.0481088161468506, | |
| "learning_rate": 2.818181818181818e-05, | |
| "loss": 0.0577, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 16.896551724137932, | |
| "grad_norm": 1.7372900247573853, | |
| "learning_rate": 2.7424242424242424e-05, | |
| "loss": 0.0539, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 16.982758620689655, | |
| "grad_norm": 0.7519413828849792, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.076, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 17.06896551724138, | |
| "grad_norm": 1.1846277713775635, | |
| "learning_rate": 2.590909090909091e-05, | |
| "loss": 0.0553, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 17.155172413793103, | |
| "grad_norm": 0.7746219038963318, | |
| "learning_rate": 2.5151515151515155e-05, | |
| "loss": 0.0617, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 17.24137931034483, | |
| "grad_norm": 1.951456904411316, | |
| "learning_rate": 2.4393939393939395e-05, | |
| "loss": 0.0805, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.24137931034483, | |
| "eval_loss": 0.43357911705970764, | |
| "eval_runtime": 40.2275, | |
| "eval_samples_per_second": 33.41, | |
| "eval_steps_per_second": 33.41, | |
| "eval_wer": 0.4156423563689338, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.32758620689655, | |
| "grad_norm": 0.5796881318092346, | |
| "learning_rate": 2.3636363636363637e-05, | |
| "loss": 0.0466, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 17.413793103448278, | |
| "grad_norm": 1.2703924179077148, | |
| "learning_rate": 2.287878787878788e-05, | |
| "loss": 0.0557, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 1.242013692855835, | |
| "learning_rate": 2.2121212121212123e-05, | |
| "loss": 0.0719, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 17.586206896551722, | |
| "grad_norm": 1.1339091062545776, | |
| "learning_rate": 2.1363636363636362e-05, | |
| "loss": 0.0488, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 17.67241379310345, | |
| "grad_norm": 1.4959948062896729, | |
| "learning_rate": 2.0606060606060608e-05, | |
| "loss": 0.0522, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 17.75862068965517, | |
| "grad_norm": 0.9912744760513306, | |
| "learning_rate": 1.984848484848485e-05, | |
| "loss": 0.0777, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 17.844827586206897, | |
| "grad_norm": 1.1574323177337646, | |
| "learning_rate": 1.9090909090909094e-05, | |
| "loss": 0.0517, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 17.93103448275862, | |
| "grad_norm": 1.5049959421157837, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.0565, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 18.017241379310345, | |
| "grad_norm": 0.6796301603317261, | |
| "learning_rate": 1.7575757575757576e-05, | |
| "loss": 0.0618, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 18.103448275862068, | |
| "grad_norm": 0.8808640837669373, | |
| "learning_rate": 1.6818181818181818e-05, | |
| "loss": 0.049, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.103448275862068, | |
| "eval_loss": 0.4193398654460907, | |
| "eval_runtime": 40.1393, | |
| "eval_samples_per_second": 33.483, | |
| "eval_steps_per_second": 33.483, | |
| "eval_wer": 0.41136301556951654, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.189655172413794, | |
| "grad_norm": 1.1478749513626099, | |
| "learning_rate": 1.606060606060606e-05, | |
| "loss": 0.0467, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 18.275862068965516, | |
| "grad_norm": 0.9449758529663086, | |
| "learning_rate": 1.5303030303030304e-05, | |
| "loss": 0.074, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 18.362068965517242, | |
| "grad_norm": 0.7945724725723267, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 0.0415, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 18.448275862068964, | |
| "grad_norm": 1.1694364547729492, | |
| "learning_rate": 1.3787878787878789e-05, | |
| "loss": 0.0547, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 18.53448275862069, | |
| "grad_norm": 0.7521646022796631, | |
| "learning_rate": 1.3030303030303032e-05, | |
| "loss": 0.0544, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 18.620689655172413, | |
| "grad_norm": 0.9526033997535706, | |
| "learning_rate": 1.2272727272727273e-05, | |
| "loss": 0.0507, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 18.70689655172414, | |
| "grad_norm": 1.0075325965881348, | |
| "learning_rate": 1.1515151515151517e-05, | |
| "loss": 0.0595, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 18.79310344827586, | |
| "grad_norm": 0.6182655096054077, | |
| "learning_rate": 1.0757575757575758e-05, | |
| "loss": 0.0593, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 18.879310344827587, | |
| "grad_norm": 0.5009527206420898, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0415, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 18.96551724137931, | |
| "grad_norm": 2.459338426589966, | |
| "learning_rate": 9.242424242424244e-06, | |
| "loss": 0.0717, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 18.96551724137931, | |
| "eval_loss": 0.4138683080673218, | |
| "eval_runtime": 40.2569, | |
| "eval_samples_per_second": 33.386, | |
| "eval_steps_per_second": 33.386, | |
| "eval_wer": 0.4090867704634435, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 19.051724137931036, | |
| "grad_norm": 1.2094837427139282, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 0.0486, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 19.137931034482758, | |
| "grad_norm": 1.1685006618499756, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 0.0376, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 19.224137931034484, | |
| "grad_norm": 0.764283299446106, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 0.0598, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 19.310344827586206, | |
| "grad_norm": 0.6357612013816833, | |
| "learning_rate": 6.212121212121212e-06, | |
| "loss": 0.0494, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 19.396551724137932, | |
| "grad_norm": 2.4395949840545654, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.0395, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 19.482758620689655, | |
| "grad_norm": 2.8990790843963623, | |
| "learning_rate": 4.696969696969697e-06, | |
| "loss": 0.0643, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 19.56896551724138, | |
| "grad_norm": 0.5647782683372498, | |
| "learning_rate": 3.939393939393939e-06, | |
| "loss": 0.0458, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 19.655172413793103, | |
| "grad_norm": 0.6177101731300354, | |
| "learning_rate": 3.1818181818181817e-06, | |
| "loss": 0.0521, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 19.74137931034483, | |
| "grad_norm": 0.9529480934143066, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 0.0497, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 19.82758620689655, | |
| "grad_norm": 0.7542730569839478, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.0389, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 19.82758620689655, | |
| "eval_loss": 0.4216426610946655, | |
| "eval_runtime": 39.4724, | |
| "eval_samples_per_second": 34.049, | |
| "eval_steps_per_second": 34.049, | |
| "eval_wer": 0.4087225712464718, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 19.913793103448278, | |
| "grad_norm": 0.9613261222839355, | |
| "learning_rate": 9.09090909090909e-07, | |
| "loss": 0.0483, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.2119308710098267, | |
| "learning_rate": 1.5151515151515152e-07, | |
| "loss": 0.0557, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2320, | |
| "total_flos": 2.1476719263248095e+18, | |
| "train_loss": 0.8590125822430027, | |
| "train_runtime": 3151.1477, | |
| "train_samples_per_second": 23.458, | |
| "train_steps_per_second": 0.736 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2320, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1476719263248095e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |