| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "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": 2.0973, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.0002999616623572683, | |
| "loss": 0.5502, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.00029984666902607135, | |
| "loss": 0.2896, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.0002996550787873857, | |
| "loss": 0.2299, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.0002993869895761197, | |
| "loss": 0.2005, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_cer": 0.2034279967643908, | |
| "eval_loss": 0.2542494535446167, | |
| "eval_runtime": 84.4226, | |
| "eval_samples_per_second": 58.634, | |
| "eval_steps_per_second": 0.118, | |
| "eval_wer": 0.5028282828282828, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.000299042538431052, | |
| "loss": 0.1818, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.00029862190142478177, | |
| "loss": 0.168, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.00029812529357372587, | |
| "loss": 0.1575, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.00029755296872820933, | |
| "loss": 0.1488, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0002969052194427048, | |
| "loss": 0.1409, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "eval_cer": 0.15823877921339718, | |
| "eval_loss": 0.17229020595550537, | |
| "eval_runtime": 83.5893, | |
| "eval_samples_per_second": 59.218, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.41454545454545455, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.0002961823768262882, | |
| "loss": 0.1342, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0002953848103733858, | |
| "loss": 0.128, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00029451292777490066, | |
| "loss": 0.1221, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0002935671747098137, | |
| "loss": 0.1175, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.00029254803461736643, | |
| "loss": 0.1114, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_cer": 0.1419514221376883, | |
| "eval_loss": 0.13527829945087433, | |
| "eval_runtime": 83.6072, | |
| "eval_samples_per_second": 59.205, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.37757575757575756, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 0.00029145602844994243, | |
| "loss": 0.1085, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 0.0002902917144067724, | |
| "loss": 0.1043, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.00028905568764860047, | |
| "loss": 0.1012, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 0.00028774857999345685, | |
| "loss": 0.099, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 0.0002863710595936922, | |
| "loss": 0.096, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_cer": 0.13088913666076388, | |
| "eval_loss": 0.11353930085897446, | |
| "eval_runtime": 83.6733, | |
| "eval_samples_per_second": 59.159, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.35434343434343435, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 0.0002849238305944389, | |
| "loss": 0.0934, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 0.00028340763277367477, | |
| "loss": 0.0916, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.0002818232411640713, | |
| "loss": 0.0891, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 0.00028017146565682144, | |
| "loss": 0.0882, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 0.00027845315058764886, | |
| "loss": 0.0852, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_cer": 0.12312804704750661, | |
| "eval_loss": 0.10132956504821777, | |
| "eval_runtime": 83.2401, | |
| "eval_samples_per_second": 59.467, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.3373737373737374, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 0.00027666917430520975, | |
| "loss": 0.0836, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 0.00027482044872210895, | |
| "loss": 0.0819, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.0002729079188487587, | |
| "loss": 0.0806, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 0.00027093256231031885, | |
| "loss": 0.0773, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 0.00026889538884696597, | |
| "loss": 0.0769, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_cer": 0.1182309088126626, | |
| "eval_loss": 0.09128542244434357, | |
| "eval_runtime": 83.8171, | |
| "eval_samples_per_second": 59.057, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.3333333333333333, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 0.0002667974397977457, | |
| "loss": 0.0757, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 0.0002646397875682729, | |
| "loss": 0.0745, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 0.00026242353508255185, | |
| "loss": 0.0737, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 0.0002601498152191957, | |
| "loss": 0.073, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.0002578197902323352, | |
| "loss": 0.072, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_cer": 0.11523578408867319, | |
| "eval_loss": 0.0870400071144104, | |
| "eval_runtime": 83.456, | |
| "eval_samples_per_second": 59.313, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.32666666666666666, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 0.00025543465115751026, | |
| "loss": 0.0707, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.0002529956172028505, | |
| "loss": 0.07, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 0.0002505039351258541, | |
| "loss": 0.0692, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 0.0002479608785960846, | |
| "loss": 0.0682, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 0.0002453677475441111, | |
| "loss": 0.0672, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_cer": 0.11036050807809186, | |
| "eval_loss": 0.08000758290290833, | |
| "eval_runtime": 84.1634, | |
| "eval_samples_per_second": 58.814, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.31292929292929295, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.00024272586749702474, | |
| "loss": 0.0666, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 0.0002400365889008706, | |
| "loss": 0.0662, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 0.00023730128643034235, | |
| "loss": 0.0631, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 0.00023452135828609167, | |
| "loss": 0.0635, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 0.0002316982254800121, | |
| "loss": 0.0625, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_cer": 0.10987953914431255, | |
| "eval_loss": 0.0780956819653511, | |
| "eval_runtime": 83.8176, | |
| "eval_samples_per_second": 59.057, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.30646464646464644, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 0.00022883333110886237, | |
| "loss": 0.0622, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 0.00022592813961660067, | |
| "loss": 0.0615, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 0.00022298413604580696, | |
| "loss": 0.0608, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 0.00022000282527857588, | |
| "loss": 0.0604, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 0.0002169857312672683, | |
| "loss": 0.0597, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "eval_cer": 0.108261734548873, | |
| "eval_loss": 0.07320722192525864, | |
| "eval_runtime": 84.043, | |
| "eval_samples_per_second": 58.898, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.30484848484848487, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 0.00021393439625551483, | |
| "loss": 0.0598, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 0.00021085037998986924, | |
| "loss": 0.0588, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 0.00020773525892251514, | |
| "loss": 0.0587, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 0.00020459062540543316, | |
| "loss": 0.0582, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 0.00020141808687644067, | |
| "loss": 0.0572, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "eval_cer": 0.10843663234297457, | |
| "eval_loss": 0.0711025819182396, | |
| "eval_runtime": 83.6072, | |
| "eval_samples_per_second": 59.205, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.30262626262626263, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.00019821926503751995, | |
| "loss": 0.0568, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 0.00019499579502585537, | |
| "loss": 0.0549, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "learning_rate": 0.00019174932457800242, | |
| "loss": 0.0544, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 0.0001884815131876167, | |
| "loss": 0.0542, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 0.00018519403125717278, | |
| "loss": 0.0537, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "eval_cer": 0.10559454318882402, | |
| "eval_loss": 0.06793416291475296, | |
| "eval_runtime": 83.7622, | |
| "eval_samples_per_second": 59.096, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.29474747474747476, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 0.00018188855924410722, | |
| "loss": 0.0539, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 0.00017856678680182127, | |
| "loss": 0.0534, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 0.0001752304119159834, | |
| "loss": 0.053, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 0.00017188114003657205, | |
| "loss": 0.0528, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 0.00016852068320610358, | |
| "loss": 0.052, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "eval_cer": 0.10421722306027416, | |
| "eval_loss": 0.06535057723522186, | |
| "eval_runtime": 83.4596, | |
| "eval_samples_per_second": 59.31, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.2911111111111111, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 0.00016515075918448972, | |
| "loss": 0.0519, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 0.00016177309057097285, | |
| "loss": 0.0511, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 0.00015838940392358722, | |
| "loss": 0.0512, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 0.00015500142887659688, | |
| "loss": 0.0506, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.00015161089725636095, | |
| "loss": 0.0506, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.10323342296845281, | |
| "eval_loss": 0.06469610333442688, | |
| "eval_runtime": 83.521, | |
| "eval_samples_per_second": 59.267, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.2909090909090909, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 0.00014821954219607845, | |
| "loss": 0.0483, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "learning_rate": 0.0001448290972498651, | |
| "loss": 0.0483, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 0.00014144129550661485, | |
| "loss": 0.048, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 0.0001380578687040995, | |
| "loss": 0.0478, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 0.00013468054634375843, | |
| "loss": 0.0481, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "eval_cer": 0.10148444502743709, | |
| "eval_loss": 0.06252569705247879, | |
| "eval_runtime": 83.6066, | |
| "eval_samples_per_second": 59.206, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.2886868686868687, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.00013131105480663235, | |
| "loss": 0.0476, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "learning_rate": 0.000127951116470891, | |
| "loss": 0.0472, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 0.00012460244883140783, | |
| "loss": 0.0469, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "learning_rate": 0.0001212667636218309, | |
| "loss": 0.0465, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "learning_rate": 0.00011794576593959775, | |
| "loss": 0.0462, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "eval_cer": 0.10179051616711485, | |
| "eval_loss": 0.06165655702352524, | |
| "eval_runtime": 83.7971, | |
| "eval_samples_per_second": 59.071, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.28383838383838383, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "learning_rate": 0.00011464115337434394, | |
| "loss": 0.0463, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 0.00011135461514014796, | |
| "loss": 0.0456, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "learning_rate": 0.00010808783121205837, | |
| "loss": 0.0456, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 0.00010484247146734352, | |
| "loss": 0.0454, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "learning_rate": 0.00010162019483190237, | |
| "loss": 0.0433, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "eval_cer": 0.10122209833628473, | |
| "eval_loss": 0.05922077223658562, | |
| "eval_runtime": 83.6915, | |
| "eval_samples_per_second": 59.146, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.2824242424242424, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 9.842264843227404e-05, | |
| "loss": 0.0439, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 9.52514667536784e-05, | |
| "loss": 0.043, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 9.210827080451842e-05, | |
| "loss": 0.0431, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "learning_rate": 8.899466728777203e-05, | |
| "loss": 0.0434, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 8.591224777969557e-05, | |
| "loss": 0.0428, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "eval_cer": 0.0995168448437944, | |
| "eval_loss": 0.05855976790189743, | |
| "eval_runtime": 83.8689, | |
| "eval_samples_per_second": 59.021, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.2791919191919192, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 8.286258791626041e-05, | |
| "loss": 0.0428, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "learning_rate": 7.984724658773716e-05, | |
| "loss": 0.0427, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "learning_rate": 7.686776514184009e-05, | |
| "loss": 0.0424, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "learning_rate": 7.392566659583846e-05, | |
| "loss": 0.0422, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "learning_rate": 7.102245485803813e-05, | |
| "loss": 0.0421, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "eval_cer": 0.10063181828119193, | |
| "eval_loss": 0.05791032314300537, | |
| "eval_runtime": 83.4355, | |
| "eval_samples_per_second": 59.327, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.2822222222222222, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "learning_rate": 6.81596139590308e-05, | |
| "loss": 0.042, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 6.533860729310434e-05, | |
| "loss": 0.0416, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.256087687020127e-05, | |
| "loss": 0.0416, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 5.98278425788092e-05, | |
| "loss": 0.0402, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "learning_rate": 5.71409014601578e-05, | |
| "loss": 0.0399, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "eval_cer": 0.09980105375920946, | |
| "eval_loss": 0.05731820687651634, | |
| "eval_runtime": 83.6985, | |
| "eval_samples_per_second": 59.141, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.2775757575757576, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "learning_rate": 5.4501426994095876e-05, | |
| "loss": 0.0397, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "learning_rate": 5.191076839701103e-05, | |
| "loss": 0.0397, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 4.9370249932153075e-05, | |
| "loss": 0.0398, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "learning_rate": 4.6881170232712164e-05, | |
| "loss": 0.0397, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 4.444480163799822e-05, | |
| "loss": 0.0396, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "eval_cer": 0.0988828403401762, | |
| "eval_loss": 0.05615651234984398, | |
| "eval_runtime": 84.0673, | |
| "eval_samples_per_second": 58.881, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.27656565656565657, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 7.57, | |
| "learning_rate": 4.2062389543061265e-05, | |
| "loss": 0.0395, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 3.9735151762084384e-05, | |
| "loss": 0.0394, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 3.746427790587557e-05, | |
| "loss": 0.0389, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 3.525092877377602e-05, | |
| "loss": 0.0392, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 3.309623576029597e-05, | |
| "loss": 0.0388, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "eval_cer": 0.09962615596510789, | |
| "eval_loss": 0.05532450973987579, | |
| "eval_runtime": 83.843, | |
| "eval_samples_per_second": 59.039, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.281010101010101, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 3.1001300276781274e-05, | |
| "loss": 0.039, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 2.8967193188406938e-05, | |
| "loss": 0.0386, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "learning_rate": 2.699495426678389e-05, | |
| "loss": 0.0379, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 2.5085591658461056e-05, | |
| "loss": 0.0376, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 2.3240081369591984e-05, | |
| "loss": 0.0375, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "eval_cer": 0.09870794254607464, | |
| "eval_loss": 0.055045340210199356, | |
| "eval_runtime": 83.5002, | |
| "eval_samples_per_second": 59.281, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.27636363636363637, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 2.1459366767031522e-05, | |
| "loss": 0.0377, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "learning_rate": 1.9744358096116225e-05, | |
| "loss": 0.0375, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 8.43, | |
| "learning_rate": 1.8095932015375496e-05, | |
| "loss": 0.0374, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "learning_rate": 1.65149311484114e-05, | |
| "loss": 0.0373, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 8.57, | |
| "learning_rate": 1.500216365317587e-05, | |
| "loss": 0.0372, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 8.57, | |
| "eval_cer": 0.09803021359393105, | |
| "eval_loss": 0.05470244958996773, | |
| "eval_runtime": 83.9735, | |
| "eval_samples_per_second": 58.947, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.27454545454545454, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 1.355840280886582e-05, | |
| "loss": 0.0374, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "learning_rate": 1.2184386620647097e-05, | |
| "loss": 0.0375, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 1.0880817442409478e-05, | |
| "loss": 0.0374, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "learning_rate": 9.648361617745371e-06, | |
| "loss": 0.0372, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "learning_rate": 8.487649139335962e-06, | |
| "loss": 0.0373, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "eval_cer": 0.09853304475197307, | |
| "eval_loss": 0.054629139602184296, | |
| "eval_runtime": 83.9435, | |
| "eval_samples_per_second": 58.968, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.27575757575757576, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 7.399273326918692e-06, | |
| "loss": 0.0372, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "learning_rate": 6.383790524001009e-06, | |
| "loss": 0.0367, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 9.15, | |
| "learning_rate": 5.441719813474849e-06, | |
| "loss": 0.0367, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "learning_rate": 4.57354275227797e-06, | |
| "loss": 0.0366, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "learning_rate": 3.7797031252369767e-06, | |
| "loss": 0.0363, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "eval_cer": 0.09787717802409217, | |
| "eval_loss": 0.05455449968576431, | |
| "eval_runtime": 83.6734, | |
| "eval_samples_per_second": 59.159, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 0.2739393939393939, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "learning_rate": 3.0606067182186776e-06, | |
| "loss": 0.0362, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "learning_rate": 2.4166211107049584e-06, | |
| "loss": 0.0365, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 1.8480754878977489e-06, | |
| "loss": 0.0365, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "learning_rate": 1.3552604724498928e-06, | |
| "loss": 0.0366, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "learning_rate": 9.384279759080127e-07, | |
| "loss": 0.0366, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "eval_cer": 0.09768041800572791, | |
| "eval_loss": 0.05442119389772415, | |
| "eval_runtime": 83.7582, | |
| "eval_samples_per_second": 59.099, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 0.2719191919191919, | |
| "step": 135000 | |
| } | |
| ], | |
| "max_steps": 139950, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.424533742064886e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |