| { | |
| "best_metric": 0.4174646901919629, | |
| "best_model_checkpoint": "/scratch/lingjzhu_root/lingjzhu1/lingjzhu/g2p/mt5_small_finetuned_from_pretrained/checkpoint-5000", | |
| "epoch": 9.646302250803858, | |
| "global_step": 135000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0003, | |
| "loss": 6.4541, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.0002999616623572683, | |
| "loss": 2.0742, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.00029984666902607135, | |
| "loss": 1.6562, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.0002996550787873857, | |
| "loss": 1.4386, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.0002993869895761197, | |
| "loss": 1.2918, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_cer": 0.4174646901919629, | |
| "eval_loss": 1.3500438928604126, | |
| "eval_runtime": 51.1023, | |
| "eval_samples_per_second": 96.864, | |
| "eval_steps_per_second": 0.763, | |
| "eval_wer": 0.8442424242424242, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.000299042538431052, | |
| "loss": 1.1816, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.00029862190142478177, | |
| "loss": 1.0922, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.00029812529357372587, | |
| "loss": 1.0252, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.00029755296872820933, | |
| "loss": 0.9679, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0002969052194427048, | |
| "loss": 0.9178, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "eval_cer": 0.327889282434737, | |
| "eval_loss": 1.0114015340805054, | |
| "eval_runtime": 50.1773, | |
| "eval_samples_per_second": 98.65, | |
| "eval_steps_per_second": 0.777, | |
| "eval_wer": 0.7397979797979798, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.0002961823768262882, | |
| "loss": 0.8744, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0002953848103733858, | |
| "loss": 0.8375, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00029451292777490066, | |
| "loss": 0.8037, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0002935671747098137, | |
| "loss": 0.7722, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.00029254803461736643, | |
| "loss": 0.7378, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_cer": 0.28204119113210024, | |
| "eval_loss": 0.8321653008460999, | |
| "eval_runtime": 50.1597, | |
| "eval_samples_per_second": 98.685, | |
| "eval_steps_per_second": 0.778, | |
| "eval_wer": 0.6737373737373737, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 0.00029145602844994243, | |
| "loss": 0.7151, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 0.0002902917144067724, | |
| "loss": 0.6922, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.00028905568764860047, | |
| "loss": 0.6723, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 0.00028774857999345685, | |
| "loss": 0.6533, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 0.0002863710595936922, | |
| "loss": 0.6377, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_cer": 0.24406401679128953, | |
| "eval_loss": 0.7227747440338135, | |
| "eval_runtime": 49.3788, | |
| "eval_samples_per_second": 100.245, | |
| "eval_steps_per_second": 0.79, | |
| "eval_wer": 0.623030303030303, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 0.0002849238305944389, | |
| "loss": 0.618, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 0.00028340763277367477, | |
| "loss": 0.6064, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.0002818232411640713, | |
| "loss": 0.5906, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 0.00028017146565682144, | |
| "loss": 0.5806, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 0.00027845315058764886, | |
| "loss": 0.5641, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_cer": 0.22388386024749662, | |
| "eval_loss": 0.6477043032646179, | |
| "eval_runtime": 49.7718, | |
| "eval_samples_per_second": 99.454, | |
| "eval_steps_per_second": 0.784, | |
| "eval_wer": 0.5822222222222222, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 0.00027666917430520975, | |
| "loss": 0.5522, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 0.00027482044872210895, | |
| "loss": 0.5415, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.0002729079188487587, | |
| "loss": 0.5308, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 0.00027093256231031885, | |
| "loss": 0.5134, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 0.00026889538884696597, | |
| "loss": 0.5042, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_cer": 0.20149547422274694, | |
| "eval_loss": 0.593482494354248, | |
| "eval_runtime": 50.469, | |
| "eval_samples_per_second": 98.08, | |
| "eval_steps_per_second": 0.773, | |
| "eval_wer": 0.5438383838383838, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 0.0002667974397977457, | |
| "loss": 0.496, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 0.0002646397875682729, | |
| "loss": 0.4864, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 0.00026242353508255185, | |
| "loss": 0.4802, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 0.0002601498152191957, | |
| "loss": 0.4743, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.0002578197902323352, | |
| "loss": 0.4677, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_cer": 0.18538195810923083, | |
| "eval_loss": 0.5494824051856995, | |
| "eval_runtime": 49.5862, | |
| "eval_samples_per_second": 99.826, | |
| "eval_steps_per_second": 0.787, | |
| "eval_wer": 0.5167676767676768, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 0.00025543465115751026, | |
| "loss": 0.4599, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.0002529956172028505, | |
| "loss": 0.4532, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 0.0002505039351258541, | |
| "loss": 0.4471, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 0.0002479608785960846, | |
| "loss": 0.4414, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 0.0002453677475441111, | |
| "loss": 0.4346, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_cer": 0.17329135510953692, | |
| "eval_loss": 0.5157131552696228, | |
| "eval_runtime": 49.3694, | |
| "eval_samples_per_second": 100.265, | |
| "eval_steps_per_second": 0.79, | |
| "eval_wer": 0.4888888888888889, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.00024272586749702474, | |
| "loss": 0.4295, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 0.0002400365889008706, | |
| "loss": 0.4248, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 0.00023730128643034235, | |
| "loss": 0.4108, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 0.00023452135828609167, | |
| "loss": 0.4082, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 0.0002316982254800121, | |
| "loss": 0.4023, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_cer": 0.1658139839958022, | |
| "eval_loss": 0.48504093289375305, | |
| "eval_runtime": 49.7275, | |
| "eval_samples_per_second": 99.543, | |
| "eval_steps_per_second": 0.784, | |
| "eval_wer": 0.47333333333333333, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 0.00022883333110886237, | |
| "loss": 0.3987, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 0.00022592813961660067, | |
| "loss": 0.3942, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 0.00022298413604580696, | |
| "loss": 0.3901, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 0.00022000282527857588, | |
| "loss": 0.3865, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 0.0002169857312672683, | |
| "loss": 0.3832, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "eval_cer": 0.15555992828720103, | |
| "eval_loss": 0.46218162775039673, | |
| "eval_runtime": 49.258, | |
| "eval_samples_per_second": 100.491, | |
| "eval_steps_per_second": 0.792, | |
| "eval_wer": 0.45656565656565656, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 0.00021393439625551483, | |
| "loss": 0.3814, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 0.00021085037998986924, | |
| "loss": 0.3766, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 0.00020773525892251514, | |
| "loss": 0.3733, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 0.00020459062540543316, | |
| "loss": 0.3704, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 0.00020141808687644067, | |
| "loss": 0.3679, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "eval_cer": 0.14858542131269403, | |
| "eval_loss": 0.4401688873767853, | |
| "eval_runtime": 49.2401, | |
| "eval_samples_per_second": 100.528, | |
| "eval_steps_per_second": 0.792, | |
| "eval_wer": 0.44161616161616163, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.00019821926503751995, | |
| "loss": 0.3642, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 0.00019499579502585537, | |
| "loss": 0.3541, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "learning_rate": 0.00019174932457800242, | |
| "loss": 0.3519, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 0.0001884815131876167, | |
| "loss": 0.3489, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 0.00018519403125717278, | |
| "loss": 0.3449, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "eval_cer": 0.1472080108443745, | |
| "eval_loss": 0.42405223846435547, | |
| "eval_runtime": 49.3556, | |
| "eval_samples_per_second": 100.292, | |
| "eval_steps_per_second": 0.79, | |
| "eval_wer": 0.4298989898989899, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 0.00018188855924410722, | |
| "loss": 0.3446, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 0.00017856678680182127, | |
| "loss": 0.3427, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 0.0001752304119159834, | |
| "loss": 0.3398, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 0.00017188114003657205, | |
| "loss": 0.3389, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 0.00016852068320610358, | |
| "loss": 0.3356, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "eval_cer": 0.1398399580217762, | |
| "eval_loss": 0.41292211413383484, | |
| "eval_runtime": 50.0206, | |
| "eval_samples_per_second": 98.959, | |
| "eval_steps_per_second": 0.78, | |
| "eval_wer": 0.4202020202020202, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 0.00016515075918448972, | |
| "loss": 0.334, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 0.00016177309057097285, | |
| "loss": 0.3314, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 0.00015838940392358722, | |
| "loss": 0.3287, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 0.00015500142887659688, | |
| "loss": 0.3268, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.00015161089725636095, | |
| "loss": 0.3248, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.13376186103458831, | |
| "eval_loss": 0.39875853061676025, | |
| "eval_runtime": 50.019, | |
| "eval_samples_per_second": 98.962, | |
| "eval_steps_per_second": 0.78, | |
| "eval_wer": 0.40969696969696967, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 0.00014821954219607845, | |
| "loss": 0.3166, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "learning_rate": 0.0001448290972498651, | |
| "loss": 0.3157, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 0.00014144129550661485, | |
| "loss": 0.3147, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 0.0001380578687040995, | |
| "loss": 0.3128, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 0.00013468054634375843, | |
| "loss": 0.3136, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "eval_cer": 0.13196904105995014, | |
| "eval_loss": 0.3886343240737915, | |
| "eval_runtime": 50.1519, | |
| "eval_samples_per_second": 98.7, | |
| "eval_steps_per_second": 0.778, | |
| "eval_wer": 0.4026262626262626, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.00013131105480663235, | |
| "loss": 0.3104, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "learning_rate": 0.000127951116470891, | |
| "loss": 0.3102, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 0.00012460244883140783, | |
| "loss": 0.3076, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "learning_rate": 0.0001212667636218309, | |
| "loss": 0.3068, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "learning_rate": 0.00011794576593959775, | |
| "loss": 0.3054, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "eval_cer": 0.12906117451571997, | |
| "eval_loss": 0.38137927651405334, | |
| "eval_runtime": 50.2174, | |
| "eval_samples_per_second": 98.571, | |
| "eval_steps_per_second": 0.777, | |
| "eval_wer": 0.39656565656565657, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "learning_rate": 0.00011464115337434394, | |
| "loss": 0.304, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 0.00011135461514014796, | |
| "loss": 0.3026, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "learning_rate": 0.00010808783121205837, | |
| "loss": 0.3019, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 0.00010484247146734352, | |
| "loss": 0.3013, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "learning_rate": 0.00010162019483190237, | |
| "loss": 0.2939, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "eval_cer": 0.12759630941449124, | |
| "eval_loss": 0.37229374051094055, | |
| "eval_runtime": 49.9246, | |
| "eval_samples_per_second": 99.149, | |
| "eval_steps_per_second": 0.781, | |
| "eval_wer": 0.39111111111111113, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 9.842264843227404e-05, | |
| "loss": 0.2946, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 9.52514667536784e-05, | |
| "loss": 0.2929, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 9.210827080451842e-05, | |
| "loss": 0.2915, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "learning_rate": 8.899466728777203e-05, | |
| "loss": 0.2911, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 8.591224777969557e-05, | |
| "loss": 0.2896, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "eval_cer": 0.12425116970571516, | |
| "eval_loss": 0.36660394072532654, | |
| "eval_runtime": 49.8757, | |
| "eval_samples_per_second": 99.247, | |
| "eval_steps_per_second": 0.782, | |
| "eval_wer": 0.3882828282828283, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 8.286258791626041e-05, | |
| "loss": 0.289, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "learning_rate": 7.984724658773716e-05, | |
| "loss": 0.2888, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "learning_rate": 7.686776514184009e-05, | |
| "loss": 0.2871, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "learning_rate": 7.392566659583846e-05, | |
| "loss": 0.2869, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "learning_rate": 7.102245485803813e-05, | |
| "loss": 0.2858, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "eval_cer": 0.1241199877563514, | |
| "eval_loss": 0.3614996075630188, | |
| "eval_runtime": 50.0083, | |
| "eval_samples_per_second": 98.984, | |
| "eval_steps_per_second": 0.78, | |
| "eval_wer": 0.38303030303030305, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "learning_rate": 6.81596139590308e-05, | |
| "loss": 0.285, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 6.533860729310434e-05, | |
| "loss": 0.2844, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.256087687020127e-05, | |
| "loss": 0.2837, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 5.98278425788092e-05, | |
| "loss": 0.2838, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "learning_rate": 5.71409014601578e-05, | |
| "loss": 0.2825, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "eval_cer": 0.12335475971839609, | |
| "eval_loss": 0.357832670211792, | |
| "eval_runtime": 51.2963, | |
| "eval_samples_per_second": 96.498, | |
| "eval_steps_per_second": 0.76, | |
| "eval_wer": 0.3808080808080808, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "learning_rate": 5.4501426994095876e-05, | |
| "loss": 0.2818, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "learning_rate": 5.191076839701103e-05, | |
| "loss": 0.2801, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 4.9370249932153075e-05, | |
| "loss": 0.2808, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "learning_rate": 4.6881170232712164e-05, | |
| "loss": 0.2799, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 4.444480163799822e-05, | |
| "loss": 0.2792, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "eval_cer": 0.11998775635139272, | |
| "eval_loss": 0.35440635681152344, | |
| "eval_runtime": 51.3774, | |
| "eval_samples_per_second": 96.346, | |
| "eval_steps_per_second": 0.759, | |
| "eval_wer": 0.37535353535353533, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 7.57, | |
| "learning_rate": 4.2062389543061265e-05, | |
| "loss": 0.28, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 3.9735151762084384e-05, | |
| "loss": 0.2783, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 3.746427790587557e-05, | |
| "loss": 0.2787, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 3.525092877377602e-05, | |
| "loss": 0.2765, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 3.309623576029597e-05, | |
| "loss": 0.277, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "eval_cer": 0.12156193974375792, | |
| "eval_loss": 0.35234031081199646, | |
| "eval_runtime": 51.2794, | |
| "eval_samples_per_second": 96.53, | |
| "eval_steps_per_second": 0.761, | |
| "eval_wer": 0.37737373737373736, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 3.1001300276781274e-05, | |
| "loss": 0.2759, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 2.8967193188406938e-05, | |
| "loss": 0.2759, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "learning_rate": 2.699495426678389e-05, | |
| "loss": 0.2733, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 2.5085591658461056e-05, | |
| "loss": 0.2731, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 2.3240081369591984e-05, | |
| "loss": 0.273, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "eval_cer": 0.12014080195898377, | |
| "eval_loss": 0.3501618802547455, | |
| "eval_runtime": 51.2122, | |
| "eval_samples_per_second": 96.657, | |
| "eval_steps_per_second": 0.762, | |
| "eval_wer": 0.37353535353535355, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 2.1459366767031522e-05, | |
| "loss": 0.2727, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "learning_rate": 1.9744358096116225e-05, | |
| "loss": 0.2741, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 8.43, | |
| "learning_rate": 1.8095932015375496e-05, | |
| "loss": 0.2728, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "learning_rate": 1.65149311484114e-05, | |
| "loss": 0.2724, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 8.57, | |
| "learning_rate": 1.500216365317587e-05, | |
| "loss": 0.2714, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 8.57, | |
| "eval_cer": 0.11911321002230094, | |
| "eval_loss": 0.3485568165779114, | |
| "eval_runtime": 52.142, | |
| "eval_samples_per_second": 94.933, | |
| "eval_steps_per_second": 0.748, | |
| "eval_wer": 0.3713131313131313, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 1.355840280886582e-05, | |
| "loss": 0.2723, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "learning_rate": 1.2184386620647097e-05, | |
| "loss": 0.2718, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 1.0880817442409478e-05, | |
| "loss": 0.272, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "learning_rate": 9.648361617745371e-06, | |
| "loss": 0.2714, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "learning_rate": 8.487649139335962e-06, | |
| "loss": 0.2715, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "eval_cer": 0.11834798198434562, | |
| "eval_loss": 0.34796079993247986, | |
| "eval_runtime": 50.1634, | |
| "eval_samples_per_second": 98.678, | |
| "eval_steps_per_second": 0.777, | |
| "eval_wer": 0.37232323232323233, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 7.399273326918692e-06, | |
| "loss": 0.2725, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "learning_rate": 6.383790524001009e-06, | |
| "loss": 0.2692, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 9.15, | |
| "learning_rate": 5.441719813474849e-06, | |
| "loss": 0.271, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "learning_rate": 4.57354275227797e-06, | |
| "loss": 0.2699, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "learning_rate": 3.7797031252369767e-06, | |
| "loss": 0.2691, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "eval_cer": 0.11874152783243692, | |
| "eval_loss": 0.3474748134613037, | |
| "eval_runtime": 50.5688, | |
| "eval_samples_per_second": 97.887, | |
| "eval_steps_per_second": 0.771, | |
| "eval_wer": 0.37353535353535355, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "learning_rate": 3.0606067182186776e-06, | |
| "loss": 0.2689, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "learning_rate": 2.4166211107049584e-06, | |
| "loss": 0.2692, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 1.8480754878977489e-06, | |
| "loss": 0.269, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "learning_rate": 1.3552604724498928e-06, | |
| "loss": 0.2695, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "learning_rate": 9.384279759080127e-07, | |
| "loss": 0.269, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "eval_cer": 0.11850102759193669, | |
| "eval_loss": 0.347343772649765, | |
| "eval_runtime": 49.946, | |
| "eval_samples_per_second": 99.107, | |
| "eval_steps_per_second": 0.781, | |
| "eval_wer": 0.37232323232323233, | |
| "step": 135000 | |
| } | |
| ], | |
| "max_steps": 139950, | |
| "num_train_epochs": 10, | |
| "total_flos": 9.106275040360243e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |