| { | |
| "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": 9.595185279846191, | |
| "learning_rate": 9e-07, | |
| "loss": 9.1142, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 9.732986450195312, | |
| "learning_rate": 1.9e-06, | |
| "loss": 8.3446, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.25862068965517243, | |
| "grad_norm": 14.272214889526367, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 8.6592, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 15.0160493850708, | |
| "learning_rate": 3.8e-06, | |
| "loss": 7.6985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 16.610979080200195, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 6.9688, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 17.26924705505371, | |
| "learning_rate": 5.8e-06, | |
| "loss": 6.232, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.603448275862069, | |
| "grad_norm": 11.347734451293945, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 4.7271, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 4.237112045288086, | |
| "learning_rate": 7.8e-06, | |
| "loss": 3.7919, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7758620689655172, | |
| "grad_norm": 1.8833028078079224, | |
| "learning_rate": 8.8e-06, | |
| "loss": 3.3967, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 1.3788093328475952, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 3.1618, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "eval_loss": 3.1117007732391357, | |
| "eval_runtime": 40.0512, | |
| "eval_samples_per_second": 33.557, | |
| "eval_steps_per_second": 33.557, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9482758620689655, | |
| "grad_norm": 1.729278802871704, | |
| "learning_rate": 1.08e-05, | |
| "loss": 3.0865, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 1.905969500541687, | |
| "learning_rate": 1.18e-05, | |
| "loss": 3.0809, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1206896551724137, | |
| "grad_norm": 0.8360918760299683, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 3.0346, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 0.7653716206550598, | |
| "learning_rate": 1.3800000000000002e-05, | |
| "loss": 3.0106, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.293103448275862, | |
| "grad_norm": 0.94779372215271, | |
| "learning_rate": 1.48e-05, | |
| "loss": 3.0165, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.8457741737365723, | |
| "learning_rate": 1.58e-05, | |
| "loss": 3.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4655172413793103, | |
| "grad_norm": 1.4369837045669556, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 2.9903, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 1.8290436267852783, | |
| "learning_rate": 1.78e-05, | |
| "loss": 2.9852, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.6379310344827587, | |
| "grad_norm": 1.1530190706253052, | |
| "learning_rate": 1.88e-05, | |
| "loss": 2.99, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 1.1261711120605469, | |
| "learning_rate": 1.9800000000000004e-05, | |
| "loss": 2.9798, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "eval_loss": 2.9736363887786865, | |
| "eval_runtime": 39.6236, | |
| "eval_samples_per_second": 33.919, | |
| "eval_steps_per_second": 33.919, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.8103448275862069, | |
| "grad_norm": 0.903380811214447, | |
| "learning_rate": 2.08e-05, | |
| "loss": 2.9718, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.4889620244503021, | |
| "learning_rate": 2.18e-05, | |
| "loss": 2.9766, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9827586206896552, | |
| "grad_norm": 1.3861790895462036, | |
| "learning_rate": 2.2800000000000002e-05, | |
| "loss": 2.9658, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 0.7976490259170532, | |
| "learning_rate": 2.38e-05, | |
| "loss": 2.9588, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1551724137931036, | |
| "grad_norm": 0.698798418045044, | |
| "learning_rate": 2.48e-05, | |
| "loss": 2.9523, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.2413793103448274, | |
| "grad_norm": 1.0858148336410522, | |
| "learning_rate": 2.58e-05, | |
| "loss": 2.9496, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.3275862068965516, | |
| "grad_norm": 0.5658290386199951, | |
| "learning_rate": 2.6800000000000004e-05, | |
| "loss": 2.9421, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 0.5713534355163574, | |
| "learning_rate": 2.7800000000000005e-05, | |
| "loss": 2.9427, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.7386118769645691, | |
| "learning_rate": 2.88e-05, | |
| "loss": 2.9228, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 0.767816960811615, | |
| "learning_rate": 2.98e-05, | |
| "loss": 2.9144, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "eval_loss": 2.9074809551239014, | |
| "eval_runtime": 39.8997, | |
| "eval_samples_per_second": 33.684, | |
| "eval_steps_per_second": 33.684, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.6724137931034484, | |
| "grad_norm": 0.8676608204841614, | |
| "learning_rate": 3.08e-05, | |
| "loss": 2.8965, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 1.6954621076583862, | |
| "learning_rate": 3.18e-05, | |
| "loss": 2.8815, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.844827586206897, | |
| "grad_norm": 1.1631884574890137, | |
| "learning_rate": 3.2800000000000004e-05, | |
| "loss": 2.855, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.9310344827586206, | |
| "grad_norm": 1.625454306602478, | |
| "learning_rate": 3.38e-05, | |
| "loss": 2.781, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.0172413793103448, | |
| "grad_norm": 2.0763564109802246, | |
| "learning_rate": 3.48e-05, | |
| "loss": 2.7756, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.103448275862069, | |
| "grad_norm": 2.036031723022461, | |
| "learning_rate": 3.58e-05, | |
| "loss": 2.6458, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.189655172413793, | |
| "grad_norm": 1.366801142692566, | |
| "learning_rate": 3.68e-05, | |
| "loss": 2.5189, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.2758620689655173, | |
| "grad_norm": 2.034527540206909, | |
| "learning_rate": 3.7800000000000004e-05, | |
| "loss": 2.433, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.3620689655172415, | |
| "grad_norm": 3.8338165283203125, | |
| "learning_rate": 3.88e-05, | |
| "loss": 2.2885, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "grad_norm": 2.3443217277526855, | |
| "learning_rate": 3.9800000000000005e-05, | |
| "loss": 2.1714, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "eval_loss": 2.0944502353668213, | |
| "eval_runtime": 39.7668, | |
| "eval_samples_per_second": 33.797, | |
| "eval_steps_per_second": 33.797, | |
| "eval_wer": 1.0325047801147227, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.5344827586206895, | |
| "grad_norm": 4.349735260009766, | |
| "learning_rate": 4.08e-05, | |
| "loss": 2.0881, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.6206896551724137, | |
| "grad_norm": 2.450747489929199, | |
| "learning_rate": 4.18e-05, | |
| "loss": 1.9522, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.706896551724138, | |
| "grad_norm": 2.2519729137420654, | |
| "learning_rate": 4.2800000000000004e-05, | |
| "loss": 1.8395, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.793103448275862, | |
| "grad_norm": 2.693664789199829, | |
| "learning_rate": 4.38e-05, | |
| "loss": 1.7525, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.8793103448275863, | |
| "grad_norm": 1.9744929075241089, | |
| "learning_rate": 4.4800000000000005e-05, | |
| "loss": 1.6222, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.9655172413793105, | |
| "grad_norm": 3.802494764328003, | |
| "learning_rate": 4.58e-05, | |
| "loss": 1.5397, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.051724137931035, | |
| "grad_norm": 2.301044225692749, | |
| "learning_rate": 4.6800000000000006e-05, | |
| "loss": 1.4376, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 2.279372215270996, | |
| "learning_rate": 4.78e-05, | |
| "loss": 1.2829, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.224137931034483, | |
| "grad_norm": 3.314736843109131, | |
| "learning_rate": 4.88e-05, | |
| "loss": 1.1976, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.310344827586207, | |
| "grad_norm": 2.434694290161133, | |
| "learning_rate": 4.9800000000000004e-05, | |
| "loss": 1.1579, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.310344827586207, | |
| "eval_loss": 1.045101284980774, | |
| "eval_runtime": 39.7455, | |
| "eval_samples_per_second": 33.815, | |
| "eval_steps_per_second": 33.815, | |
| "eval_wer": 0.8299189656742239, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.396551724137931, | |
| "grad_norm": 1.8384031057357788, | |
| "learning_rate": 5.08e-05, | |
| "loss": 1.0684, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.482758620689655, | |
| "grad_norm": 3.599148988723755, | |
| "learning_rate": 5.1800000000000005e-05, | |
| "loss": 1.0319, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.568965517241379, | |
| "grad_norm": 2.066476583480835, | |
| "learning_rate": 5.28e-05, | |
| "loss": 0.9179, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.655172413793103, | |
| "grad_norm": 2.2173750400543213, | |
| "learning_rate": 5.380000000000001e-05, | |
| "loss": 0.8838, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.741379310344827, | |
| "grad_norm": 2.427091121673584, | |
| "learning_rate": 5.4800000000000004e-05, | |
| "loss": 0.8991, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.827586206896552, | |
| "grad_norm": 2.7432241439819336, | |
| "learning_rate": 5.580000000000001e-05, | |
| "loss": 0.8, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.913793103448276, | |
| "grad_norm": 3.254221200942993, | |
| "learning_rate": 5.68e-05, | |
| "loss": 0.7803, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.457448482513428, | |
| "learning_rate": 5.7799999999999995e-05, | |
| "loss": 0.8205, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 5.086206896551724, | |
| "grad_norm": 3.1023166179656982, | |
| "learning_rate": 5.88e-05, | |
| "loss": 0.6703, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.172413793103448, | |
| "grad_norm": 2.5916504859924316, | |
| "learning_rate": 5.9800000000000003e-05, | |
| "loss": 0.6087, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.172413793103448, | |
| "eval_loss": 0.6753795146942139, | |
| "eval_runtime": 39.7485, | |
| "eval_samples_per_second": 33.813, | |
| "eval_steps_per_second": 33.813, | |
| "eval_wer": 0.6440863152144223, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.258620689655173, | |
| "grad_norm": 2.1707613468170166, | |
| "learning_rate": 6.08e-05, | |
| "loss": 0.6569, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.344827586206897, | |
| "grad_norm": 2.4291555881500244, | |
| "learning_rate": 6.18e-05, | |
| "loss": 0.5627, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.431034482758621, | |
| "grad_norm": 2.249617338180542, | |
| "learning_rate": 6.280000000000001e-05, | |
| "loss": 0.5381, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.517241379310345, | |
| "grad_norm": 1.6661946773529053, | |
| "learning_rate": 6.38e-05, | |
| "loss": 0.6338, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.603448275862069, | |
| "grad_norm": 2.60294771194458, | |
| "learning_rate": 6.48e-05, | |
| "loss": 0.5181, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.689655172413794, | |
| "grad_norm": 3.3003089427948, | |
| "learning_rate": 6.58e-05, | |
| "loss": 0.5189, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.775862068965517, | |
| "grad_norm": 1.880764126777649, | |
| "learning_rate": 6.680000000000001e-05, | |
| "loss": 0.564, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.862068965517241, | |
| "grad_norm": 2.0575127601623535, | |
| "learning_rate": 6.780000000000001e-05, | |
| "loss": 0.4729, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.948275862068965, | |
| "grad_norm": 2.5159761905670166, | |
| "learning_rate": 6.879999999999999e-05, | |
| "loss": 0.4899, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 6.0344827586206895, | |
| "grad_norm": 1.4463504552841187, | |
| "learning_rate": 6.98e-05, | |
| "loss": 0.481, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.0344827586206895, | |
| "eval_loss": 0.5275412201881409, | |
| "eval_runtime": 39.9601, | |
| "eval_samples_per_second": 33.634, | |
| "eval_steps_per_second": 33.634, | |
| "eval_wer": 0.5760721114449604, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.120689655172414, | |
| "grad_norm": 1.788765549659729, | |
| "learning_rate": 7.08e-05, | |
| "loss": 0.3865, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.206896551724138, | |
| "grad_norm": 1.862762212753296, | |
| "learning_rate": 7.18e-05, | |
| "loss": 0.3726, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.293103448275862, | |
| "grad_norm": 1.6512093544006348, | |
| "learning_rate": 7.280000000000001e-05, | |
| "loss": 0.4116, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.379310344827586, | |
| "grad_norm": 2.098067045211792, | |
| "learning_rate": 7.38e-05, | |
| "loss": 0.3779, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.4655172413793105, | |
| "grad_norm": 3.3030078411102295, | |
| "learning_rate": 7.48e-05, | |
| "loss": 0.3728, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.551724137931035, | |
| "grad_norm": 2.1799120903015137, | |
| "learning_rate": 7.58e-05, | |
| "loss": 0.4047, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.637931034482759, | |
| "grad_norm": 1.862434983253479, | |
| "learning_rate": 7.680000000000001e-05, | |
| "loss": 0.313, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.724137931034483, | |
| "grad_norm": 6.29113245010376, | |
| "learning_rate": 7.780000000000001e-05, | |
| "loss": 0.4052, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.810344827586206, | |
| "grad_norm": 1.4220325946807861, | |
| "learning_rate": 7.88e-05, | |
| "loss": 0.3218, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 2.586819648742676, | |
| "learning_rate": 7.98e-05, | |
| "loss": 0.3072, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "eval_loss": 0.4836220443248749, | |
| "eval_runtime": 39.8762, | |
| "eval_samples_per_second": 33.704, | |
| "eval_steps_per_second": 33.704, | |
| "eval_wer": 0.5264499681325685, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.982758620689655, | |
| "grad_norm": 1.6589460372924805, | |
| "learning_rate": 8.080000000000001e-05, | |
| "loss": 0.3862, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 7.068965517241379, | |
| "grad_norm": 1.7299175262451172, | |
| "learning_rate": 8.18e-05, | |
| "loss": 0.2938, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 7.155172413793103, | |
| "grad_norm": 2.0545098781585693, | |
| "learning_rate": 8.28e-05, | |
| "loss": 0.249, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 7.241379310344827, | |
| "grad_norm": 24.935670852661133, | |
| "learning_rate": 8.38e-05, | |
| "loss": 0.3202, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.327586206896552, | |
| "grad_norm": 2.497840642929077, | |
| "learning_rate": 8.48e-05, | |
| "loss": 0.2803, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.413793103448276, | |
| "grad_norm": 2.698636531829834, | |
| "learning_rate": 8.58e-05, | |
| "loss": 0.2473, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.4561227560043335, | |
| "learning_rate": 8.680000000000001e-05, | |
| "loss": 0.3223, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.586206896551724, | |
| "grad_norm": 1.7760556936264038, | |
| "learning_rate": 8.78e-05, | |
| "loss": 0.2481, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.672413793103448, | |
| "grad_norm": 2.308103084564209, | |
| "learning_rate": 8.88e-05, | |
| "loss": 0.2545, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.758620689655173, | |
| "grad_norm": 1.4128385782241821, | |
| "learning_rate": 8.98e-05, | |
| "loss": 0.332, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.758620689655173, | |
| "eval_loss": 0.44030094146728516, | |
| "eval_runtime": 39.9401, | |
| "eval_samples_per_second": 33.65, | |
| "eval_steps_per_second": 33.65, | |
| "eval_wer": 0.5233542747883092, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.844827586206897, | |
| "grad_norm": 1.7903906106948853, | |
| "learning_rate": 9.080000000000001e-05, | |
| "loss": 0.2411, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 7.931034482758621, | |
| "grad_norm": 2.0804216861724854, | |
| "learning_rate": 9.180000000000001e-05, | |
| "loss": 0.2707, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 8.017241379310345, | |
| "grad_norm": 1.4420605897903442, | |
| "learning_rate": 9.28e-05, | |
| "loss": 0.3186, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 8.10344827586207, | |
| "grad_norm": 2.2910854816436768, | |
| "learning_rate": 9.38e-05, | |
| "loss": 0.1937, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 8.189655172413794, | |
| "grad_norm": 3.5892796516418457, | |
| "learning_rate": 9.48e-05, | |
| "loss": 0.2321, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.275862068965518, | |
| "grad_norm": 1.6509956121444702, | |
| "learning_rate": 9.58e-05, | |
| "loss": 0.2868, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.362068965517242, | |
| "grad_norm": 1.6983604431152344, | |
| "learning_rate": 9.680000000000001e-05, | |
| "loss": 0.2004, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.448275862068966, | |
| "grad_norm": 2.061176061630249, | |
| "learning_rate": 9.78e-05, | |
| "loss": 0.2025, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.53448275862069, | |
| "grad_norm": 1.7732270956039429, | |
| "learning_rate": 9.88e-05, | |
| "loss": 0.2598, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.620689655172415, | |
| "grad_norm": 1.8335466384887695, | |
| "learning_rate": 9.98e-05, | |
| "loss": 0.1876, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.620689655172415, | |
| "eval_loss": 0.4757933020591736, | |
| "eval_runtime": 39.8291, | |
| "eval_samples_per_second": 33.744, | |
| "eval_steps_per_second": 33.744, | |
| "eval_wer": 0.5221706273331512, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.706896551724139, | |
| "grad_norm": 2.52902889251709, | |
| "learning_rate": 9.939393939393939e-05, | |
| "loss": 0.2456, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.793103448275861, | |
| "grad_norm": 1.7294162511825562, | |
| "learning_rate": 9.863636363636364e-05, | |
| "loss": 0.2499, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 8.879310344827585, | |
| "grad_norm": 21.9121150970459, | |
| "learning_rate": 9.787878787878789e-05, | |
| "loss": 0.1854, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 8.96551724137931, | |
| "grad_norm": 3.9164559841156006, | |
| "learning_rate": 9.712121212121212e-05, | |
| "loss": 0.2576, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 9.051724137931034, | |
| "grad_norm": 1.239221215248108, | |
| "learning_rate": 9.636363636363637e-05, | |
| "loss": 0.2118, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 9.137931034482758, | |
| "grad_norm": 3.1416544914245605, | |
| "learning_rate": 9.560606060606061e-05, | |
| "loss": 0.1577, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 9.224137931034482, | |
| "grad_norm": 2.4253621101379395, | |
| "learning_rate": 9.484848484848486e-05, | |
| "loss": 0.2092, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 9.310344827586206, | |
| "grad_norm": 1.194345474243164, | |
| "learning_rate": 9.40909090909091e-05, | |
| "loss": 0.1876, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.39655172413793, | |
| "grad_norm": 2.411029100418091, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 0.1546, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.482758620689655, | |
| "grad_norm": 3.246082067489624, | |
| "learning_rate": 9.257575757575758e-05, | |
| "loss": 0.2232, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.482758620689655, | |
| "eval_loss": 0.45077577233314514, | |
| "eval_runtime": 39.9221, | |
| "eval_samples_per_second": 33.666, | |
| "eval_steps_per_second": 33.666, | |
| "eval_wer": 0.48921059819721385, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.568965517241379, | |
| "grad_norm": 1.3427454233169556, | |
| "learning_rate": 9.181818181818183e-05, | |
| "loss": 0.1777, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.655172413793103, | |
| "grad_norm": 1.5090447664260864, | |
| "learning_rate": 9.106060606060606e-05, | |
| "loss": 0.1646, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.741379310344827, | |
| "grad_norm": 1.3060975074768066, | |
| "learning_rate": 9.030303030303031e-05, | |
| "loss": 0.225, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 9.827586206896552, | |
| "grad_norm": 1.3011540174484253, | |
| "learning_rate": 8.954545454545455e-05, | |
| "loss": 0.1552, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 9.913793103448276, | |
| "grad_norm": 1.9938538074493408, | |
| "learning_rate": 8.87878787878788e-05, | |
| "loss": 0.1715, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.334385395050049, | |
| "learning_rate": 8.803030303030304e-05, | |
| "loss": 0.2092, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 10.086206896551724, | |
| "grad_norm": 1.011092185974121, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.14, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 10.172413793103448, | |
| "grad_norm": 2.517902135848999, | |
| "learning_rate": 8.651515151515152e-05, | |
| "loss": 0.1512, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 10.258620689655173, | |
| "grad_norm": 1.2418378591537476, | |
| "learning_rate": 8.575757575757576e-05, | |
| "loss": 0.1846, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 10.344827586206897, | |
| "grad_norm": 1.5885329246520996, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.1332, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.344827586206897, | |
| "eval_loss": 0.4394075274467468, | |
| "eval_runtime": 39.9367, | |
| "eval_samples_per_second": 33.653, | |
| "eval_steps_per_second": 33.653, | |
| "eval_wer": 0.4740052808886461, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.431034482758621, | |
| "grad_norm": 1.2539469003677368, | |
| "learning_rate": 8.424242424242424e-05, | |
| "loss": 0.1485, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 10.517241379310345, | |
| "grad_norm": 1.357601284980774, | |
| "learning_rate": 8.348484848484849e-05, | |
| "loss": 0.1988, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.60344827586207, | |
| "grad_norm": 2.0564587116241455, | |
| "learning_rate": 8.272727272727273e-05, | |
| "loss": 0.137, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 10.689655172413794, | |
| "grad_norm": 2.48364520072937, | |
| "learning_rate": 8.196969696969698e-05, | |
| "loss": 0.1245, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 10.775862068965518, | |
| "grad_norm": 1.015891671180725, | |
| "learning_rate": 8.121212121212121e-05, | |
| "loss": 0.1602, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 10.862068965517242, | |
| "grad_norm": 1.1023950576782227, | |
| "learning_rate": 8.045454545454546e-05, | |
| "loss": 0.1215, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 10.948275862068966, | |
| "grad_norm": 2.703427791595459, | |
| "learning_rate": 7.96969696969697e-05, | |
| "loss": 0.1621, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 11.03448275862069, | |
| "grad_norm": 1.1821691989898682, | |
| "learning_rate": 7.893939393939395e-05, | |
| "loss": 0.1651, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 11.120689655172415, | |
| "grad_norm": 0.930283784866333, | |
| "learning_rate": 7.818181818181818e-05, | |
| "loss": 0.1066, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 11.206896551724139, | |
| "grad_norm": 1.6548758745193481, | |
| "learning_rate": 7.742424242424243e-05, | |
| "loss": 0.1085, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.206896551724139, | |
| "eval_loss": 0.4466467499732971, | |
| "eval_runtime": 39.8633, | |
| "eval_samples_per_second": 33.715, | |
| "eval_steps_per_second": 33.715, | |
| "eval_wer": 0.46207775653282346, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.293103448275861, | |
| "grad_norm": 1.1760716438293457, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 0.1418, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 11.379310344827585, | |
| "grad_norm": 2.1062755584716797, | |
| "learning_rate": 7.59090909090909e-05, | |
| "loss": 0.1133, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 11.46551724137931, | |
| "grad_norm": 2.67399001121521, | |
| "learning_rate": 7.515151515151515e-05, | |
| "loss": 0.1318, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 11.551724137931034, | |
| "grad_norm": 1.0049142837524414, | |
| "learning_rate": 7.439393939393939e-05, | |
| "loss": 0.1474, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 11.637931034482758, | |
| "grad_norm": 1.586559772491455, | |
| "learning_rate": 7.363636363636364e-05, | |
| "loss": 0.0908, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 11.724137931034482, | |
| "grad_norm": 3.784040927886963, | |
| "learning_rate": 7.287878787878788e-05, | |
| "loss": 0.1521, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 11.810344827586206, | |
| "grad_norm": 1.125501275062561, | |
| "learning_rate": 7.212121212121213e-05, | |
| "loss": 0.1163, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 11.89655172413793, | |
| "grad_norm": 2.1989808082580566, | |
| "learning_rate": 7.136363636363636e-05, | |
| "loss": 0.1109, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 11.982758620689655, | |
| "grad_norm": 1.1287301778793335, | |
| "learning_rate": 7.060606060606061e-05, | |
| "loss": 0.152, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 12.068965517241379, | |
| "grad_norm": 1.538678765296936, | |
| "learning_rate": 6.984848484848485e-05, | |
| "loss": 0.098, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 12.068965517241379, | |
| "eval_loss": 0.42302384972572327, | |
| "eval_runtime": 40.1773, | |
| "eval_samples_per_second": 33.452, | |
| "eval_steps_per_second": 33.452, | |
| "eval_wer": 0.44933078393881454, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 12.155172413793103, | |
| "grad_norm": 1.400772213935852, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.092, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 12.241379310344827, | |
| "grad_norm": 3.6780846118927, | |
| "learning_rate": 6.833333333333333e-05, | |
| "loss": 0.1649, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 12.327586206896552, | |
| "grad_norm": 1.5424057245254517, | |
| "learning_rate": 6.757575757575758e-05, | |
| "loss": 0.091, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 12.413793103448276, | |
| "grad_norm": 1.4868180751800537, | |
| "learning_rate": 6.681818181818183e-05, | |
| "loss": 0.0869, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 1.1947145462036133, | |
| "learning_rate": 6.606060606060607e-05, | |
| "loss": 0.1499, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 12.586206896551724, | |
| "grad_norm": 1.0430784225463867, | |
| "learning_rate": 6.530303030303032e-05, | |
| "loss": 0.0954, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 12.672413793103448, | |
| "grad_norm": 2.4261584281921387, | |
| "learning_rate": 6.454545454545455e-05, | |
| "loss": 0.1032, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 12.758620689655173, | |
| "grad_norm": 1.033467411994934, | |
| "learning_rate": 6.37878787878788e-05, | |
| "loss": 0.1158, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 12.844827586206897, | |
| "grad_norm": 1.1535651683807373, | |
| "learning_rate": 6.303030303030302e-05, | |
| "loss": 0.0864, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 12.931034482758621, | |
| "grad_norm": 1.28826105594635, | |
| "learning_rate": 6.227272727272727e-05, | |
| "loss": 0.1219, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 12.931034482758621, | |
| "eval_loss": 0.418023020029068, | |
| "eval_runtime": 40.2192, | |
| "eval_samples_per_second": 33.417, | |
| "eval_steps_per_second": 33.417, | |
| "eval_wer": 0.44596194118182647, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.017241379310345, | |
| "grad_norm": 1.055411458015442, | |
| "learning_rate": 6.151515151515151e-05, | |
| "loss": 0.1289, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 13.10344827586207, | |
| "grad_norm": 1.1269094944000244, | |
| "learning_rate": 6.075757575757576e-05, | |
| "loss": 0.0776, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 13.189655172413794, | |
| "grad_norm": 1.7149118185043335, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0871, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 13.275862068965518, | |
| "grad_norm": 1.7456856966018677, | |
| "learning_rate": 5.9242424242424244e-05, | |
| "loss": 0.1087, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 13.362068965517242, | |
| "grad_norm": 1.3434715270996094, | |
| "learning_rate": 5.848484848484849e-05, | |
| "loss": 0.0821, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 13.448275862068966, | |
| "grad_norm": 2.103512763977051, | |
| "learning_rate": 5.772727272727273e-05, | |
| "loss": 0.0878, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 13.53448275862069, | |
| "grad_norm": 1.240224838256836, | |
| "learning_rate": 5.696969696969697e-05, | |
| "loss": 0.1044, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 13.620689655172415, | |
| "grad_norm": 0.7336703538894653, | |
| "learning_rate": 5.6212121212121215e-05, | |
| "loss": 0.0753, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 13.706896551724139, | |
| "grad_norm": 2.293342351913452, | |
| "learning_rate": 5.545454545454546e-05, | |
| "loss": 0.1059, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "grad_norm": 1.1853971481323242, | |
| "learning_rate": 5.46969696969697e-05, | |
| "loss": 0.1021, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "eval_loss": 0.41785839200019836, | |
| "eval_runtime": 40.2906, | |
| "eval_samples_per_second": 33.358, | |
| "eval_steps_per_second": 33.358, | |
| "eval_wer": 0.4405900027314941, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.879310344827585, | |
| "grad_norm": 1.331200361251831, | |
| "learning_rate": 5.393939393939394e-05, | |
| "loss": 0.0648, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 13.96551724137931, | |
| "grad_norm": 2.28397536277771, | |
| "learning_rate": 5.3181818181818186e-05, | |
| "loss": 0.1121, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 14.051724137931034, | |
| "grad_norm": 0.9436893463134766, | |
| "learning_rate": 5.242424242424243e-05, | |
| "loss": 0.0725, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 14.137931034482758, | |
| "grad_norm": 1.6113288402557373, | |
| "learning_rate": 5.166666666666667e-05, | |
| "loss": 0.0691, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 14.224137931034482, | |
| "grad_norm": 2.479888439178467, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.0979, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 14.310344827586206, | |
| "grad_norm": 1.006616473197937, | |
| "learning_rate": 5.015151515151515e-05, | |
| "loss": 0.0909, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 14.39655172413793, | |
| "grad_norm": 1.4571704864501953, | |
| "learning_rate": 4.93939393939394e-05, | |
| "loss": 0.0761, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 14.482758620689655, | |
| "grad_norm": 1.5729875564575195, | |
| "learning_rate": 4.863636363636364e-05, | |
| "loss": 0.0862, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 14.568965517241379, | |
| "grad_norm": 1.2180376052856445, | |
| "learning_rate": 4.787878787878788e-05, | |
| "loss": 0.0646, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 14.655172413793103, | |
| "grad_norm": 1.7464072704315186, | |
| "learning_rate": 4.712121212121212e-05, | |
| "loss": 0.0741, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 14.655172413793103, | |
| "eval_loss": 0.4113341271877289, | |
| "eval_runtime": 40.2841, | |
| "eval_samples_per_second": 33.363, | |
| "eval_steps_per_second": 33.363, | |
| "eval_wer": 0.4309387234817445, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 14.741379310344827, | |
| "grad_norm": 0.8571386337280273, | |
| "learning_rate": 4.6439393939393944e-05, | |
| "loss": 0.1315, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 14.827586206896552, | |
| "grad_norm": 1.331377387046814, | |
| "learning_rate": 4.5681818181818186e-05, | |
| "loss": 0.0603, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 14.913793103448276, | |
| "grad_norm": 1.5398732423782349, | |
| "learning_rate": 4.492424242424242e-05, | |
| "loss": 0.0796, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.689671754837036, | |
| "learning_rate": 4.4166666666666665e-05, | |
| "loss": 0.085, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 15.086206896551724, | |
| "grad_norm": 1.132613182067871, | |
| "learning_rate": 4.340909090909091e-05, | |
| "loss": 0.0544, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 15.172413793103448, | |
| "grad_norm": 1.5951859951019287, | |
| "learning_rate": 4.265151515151515e-05, | |
| "loss": 0.0601, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 15.258620689655173, | |
| "grad_norm": 0.5179944634437561, | |
| "learning_rate": 4.189393939393939e-05, | |
| "loss": 0.097, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 15.344827586206897, | |
| "grad_norm": 0.9744370579719543, | |
| "learning_rate": 4.113636363636364e-05, | |
| "loss": 0.0596, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 15.431034482758621, | |
| "grad_norm": 1.8794275522232056, | |
| "learning_rate": 4.0378787878787885e-05, | |
| "loss": 0.0677, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 15.517241379310345, | |
| "grad_norm": 0.748386025428772, | |
| "learning_rate": 3.962121212121213e-05, | |
| "loss": 0.0896, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.517241379310345, | |
| "eval_loss": 0.43920788168907166, | |
| "eval_runtime": 40.1997, | |
| "eval_samples_per_second": 33.433, | |
| "eval_steps_per_second": 33.433, | |
| "eval_wer": 0.4307566238732587, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.60344827586207, | |
| "grad_norm": 0.9639837145805359, | |
| "learning_rate": 3.8863636363636364e-05, | |
| "loss": 0.0604, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 15.689655172413794, | |
| "grad_norm": 1.9640839099884033, | |
| "learning_rate": 3.810606060606061e-05, | |
| "loss": 0.0711, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 15.775862068965518, | |
| "grad_norm": 1.4438735246658325, | |
| "learning_rate": 3.734848484848485e-05, | |
| "loss": 0.0867, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.862068965517242, | |
| "grad_norm": 1.0062426328659058, | |
| "learning_rate": 3.659090909090909e-05, | |
| "loss": 0.0605, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 15.948275862068966, | |
| "grad_norm": 1.6331523656845093, | |
| "learning_rate": 3.5833333333333335e-05, | |
| "loss": 0.0662, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 16.03448275862069, | |
| "grad_norm": 0.8070217370986938, | |
| "learning_rate": 3.507575757575758e-05, | |
| "loss": 0.0765, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 16.120689655172413, | |
| "grad_norm": 1.4137670993804932, | |
| "learning_rate": 3.431818181818182e-05, | |
| "loss": 0.0537, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 16.20689655172414, | |
| "grad_norm": 1.5437769889831543, | |
| "learning_rate": 3.356060606060606e-05, | |
| "loss": 0.0684, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 16.29310344827586, | |
| "grad_norm": 0.90281081199646, | |
| "learning_rate": 3.2803030303030305e-05, | |
| "loss": 0.0744, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 16.379310344827587, | |
| "grad_norm": 1.139837622642517, | |
| "learning_rate": 3.204545454545455e-05, | |
| "loss": 0.0492, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.379310344827587, | |
| "eval_loss": 0.4201890528202057, | |
| "eval_runtime": 40.1502, | |
| "eval_samples_per_second": 33.474, | |
| "eval_steps_per_second": 33.474, | |
| "eval_wer": 0.4313029226987162, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.46551724137931, | |
| "grad_norm": 1.679457426071167, | |
| "learning_rate": 3.128787878787879e-05, | |
| "loss": 0.0652, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 16.551724137931036, | |
| "grad_norm": 0.6661111116409302, | |
| "learning_rate": 3.0530303030303034e-05, | |
| "loss": 0.0649, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 16.637931034482758, | |
| "grad_norm": 1.1774355173110962, | |
| "learning_rate": 2.9772727272727273e-05, | |
| "loss": 0.0469, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 16.724137931034484, | |
| "grad_norm": 1.783923864364624, | |
| "learning_rate": 2.901515151515152e-05, | |
| "loss": 0.0752, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 16.810344827586206, | |
| "grad_norm": 1.176321268081665, | |
| "learning_rate": 2.825757575757576e-05, | |
| "loss": 0.0519, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 16.896551724137932, | |
| "grad_norm": 1.3150608539581299, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.0547, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 16.982758620689655, | |
| "grad_norm": 0.983769953250885, | |
| "learning_rate": 2.674242424242424e-05, | |
| "loss": 0.0799, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 17.06896551724138, | |
| "grad_norm": 0.996890127658844, | |
| "learning_rate": 2.5984848484848483e-05, | |
| "loss": 0.0577, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 17.155172413793103, | |
| "grad_norm": 2.3034253120422363, | |
| "learning_rate": 2.5227272727272726e-05, | |
| "loss": 0.0515, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 17.24137931034483, | |
| "grad_norm": 3.7528610229492188, | |
| "learning_rate": 2.4469696969696972e-05, | |
| "loss": 0.0759, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.24137931034483, | |
| "eval_loss": 0.43480169773101807, | |
| "eval_runtime": 40.017, | |
| "eval_samples_per_second": 33.586, | |
| "eval_steps_per_second": 33.586, | |
| "eval_wer": 0.4207411454065374, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.32758620689655, | |
| "grad_norm": 0.6646668314933777, | |
| "learning_rate": 2.3712121212121214e-05, | |
| "loss": 0.0419, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 17.413793103448278, | |
| "grad_norm": 1.3250740766525269, | |
| "learning_rate": 2.2954545454545457e-05, | |
| "loss": 0.0595, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 0.8094995021820068, | |
| "learning_rate": 2.21969696969697e-05, | |
| "loss": 0.0691, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 17.586206896551722, | |
| "grad_norm": 0.846946120262146, | |
| "learning_rate": 2.143939393939394e-05, | |
| "loss": 0.052, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 17.67241379310345, | |
| "grad_norm": 1.652417540550232, | |
| "learning_rate": 2.0681818181818182e-05, | |
| "loss": 0.0565, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 17.75862068965517, | |
| "grad_norm": 1.0080279111862183, | |
| "learning_rate": 1.9924242424242425e-05, | |
| "loss": 0.0745, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 17.844827586206897, | |
| "grad_norm": 0.7252691388130188, | |
| "learning_rate": 1.9166666666666667e-05, | |
| "loss": 0.0513, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 17.93103448275862, | |
| "grad_norm": 1.58548903465271, | |
| "learning_rate": 1.840909090909091e-05, | |
| "loss": 0.055, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 18.017241379310345, | |
| "grad_norm": 0.6634634733200073, | |
| "learning_rate": 1.7651515151515153e-05, | |
| "loss": 0.0658, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 18.103448275862068, | |
| "grad_norm": 1.1495524644851685, | |
| "learning_rate": 1.6893939393939395e-05, | |
| "loss": 0.0406, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.103448275862068, | |
| "eval_loss": 0.44191813468933105, | |
| "eval_runtime": 40.0967, | |
| "eval_samples_per_second": 33.519, | |
| "eval_steps_per_second": 33.519, | |
| "eval_wer": 0.42046799599380863, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.189655172413794, | |
| "grad_norm": 0.9788354635238647, | |
| "learning_rate": 1.6136363636363638e-05, | |
| "loss": 0.0381, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 18.275862068965516, | |
| "grad_norm": 1.093633770942688, | |
| "learning_rate": 1.5378787878787877e-05, | |
| "loss": 0.071, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 18.362068965517242, | |
| "grad_norm": 0.7164376974105835, | |
| "learning_rate": 1.4621212121212122e-05, | |
| "loss": 0.0439, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 18.448275862068964, | |
| "grad_norm": 0.9887032508850098, | |
| "learning_rate": 1.3863636363636364e-05, | |
| "loss": 0.0481, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 18.53448275862069, | |
| "grad_norm": 0.45052286982536316, | |
| "learning_rate": 1.3106060606060607e-05, | |
| "loss": 0.0571, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 18.620689655172413, | |
| "grad_norm": 1.167181134223938, | |
| "learning_rate": 1.234848484848485e-05, | |
| "loss": 0.0452, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 18.70689655172414, | |
| "grad_norm": 1.378661870956421, | |
| "learning_rate": 1.159090909090909e-05, | |
| "loss": 0.0643, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 18.79310344827586, | |
| "grad_norm": 0.854932963848114, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 0.0587, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 18.879310344827587, | |
| "grad_norm": 0.8007526397705078, | |
| "learning_rate": 1.0075757575757576e-05, | |
| "loss": 0.0395, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 18.96551724137931, | |
| "grad_norm": 3.317830801010132, | |
| "learning_rate": 9.318181818181819e-06, | |
| "loss": 0.074, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 18.96551724137931, | |
| "eval_loss": 0.43061742186546326, | |
| "eval_runtime": 40.0034, | |
| "eval_samples_per_second": 33.597, | |
| "eval_steps_per_second": 33.597, | |
| "eval_wer": 0.420012746972594, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 19.051724137931036, | |
| "grad_norm": 0.7710875272750854, | |
| "learning_rate": 8.56060606060606e-06, | |
| "loss": 0.046, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 19.137931034482758, | |
| "grad_norm": 0.5200530886650085, | |
| "learning_rate": 7.803030303030304e-06, | |
| "loss": 0.0394, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 19.224137931034484, | |
| "grad_norm": 1.3544327020645142, | |
| "learning_rate": 7.045454545454545e-06, | |
| "loss": 0.0582, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 19.310344827586206, | |
| "grad_norm": 0.8653574585914612, | |
| "learning_rate": 6.287878787878789e-06, | |
| "loss": 0.0606, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 19.396551724137932, | |
| "grad_norm": 1.5852700471878052, | |
| "learning_rate": 5.530303030303031e-06, | |
| "loss": 0.0367, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 19.482758620689655, | |
| "grad_norm": 2.2167246341705322, | |
| "learning_rate": 4.772727272727273e-06, | |
| "loss": 0.0782, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 19.56896551724138, | |
| "grad_norm": 0.5891330242156982, | |
| "learning_rate": 4.015151515151515e-06, | |
| "loss": 0.0416, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 19.655172413793103, | |
| "grad_norm": 1.1137330532073975, | |
| "learning_rate": 3.257575757575758e-06, | |
| "loss": 0.0515, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 19.74137931034483, | |
| "grad_norm": 0.8132285475730896, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0512, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 19.82758620689655, | |
| "grad_norm": 0.7994781136512756, | |
| "learning_rate": 1.7424242424242427e-06, | |
| "loss": 0.0378, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 19.82758620689655, | |
| "eval_loss": 0.4273350238800049, | |
| "eval_runtime": 40.0934, | |
| "eval_samples_per_second": 33.522, | |
| "eval_steps_per_second": 33.522, | |
| "eval_wer": 0.41728125284530637, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 19.913793103448278, | |
| "grad_norm": 0.9775754809379578, | |
| "learning_rate": 9.848484848484847e-07, | |
| "loss": 0.0489, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.8857516050338745, | |
| "learning_rate": 2.2727272727272726e-07, | |
| "loss": 0.0554, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2320, | |
| "total_flos": 2.1476719263248095e+18, | |
| "train_loss": 0.8618391515622879, | |
| "train_runtime": 3159.4128, | |
| "train_samples_per_second": 23.397, | |
| "train_steps_per_second": 0.734 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1476719263248095e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |