| { | |
| "best_metric": 2.176206588745117, | |
| "best_model_checkpoint": "./Hubert-common_voice_JSUT-ja-demo-japanese/checkpoint-10300", | |
| "epoch": 20.0, | |
| "eval_steps": 100, | |
| "global_step": 10340, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.19342359767891681, | |
| "eval_cer": 8.48504498730911, | |
| "eval_loss": 84.69584655761719, | |
| "eval_runtime": 236.8633, | |
| "eval_samples_per_second": 23.055, | |
| "eval_steps_per_second": 2.884, | |
| "eval_wer": 1.0115257958287596, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38684719535783363, | |
| "eval_cer": 8.344315240030694, | |
| "eval_loss": 83.78863525390625, | |
| "eval_runtime": 232.6848, | |
| "eval_samples_per_second": 23.47, | |
| "eval_steps_per_second": 2.935, | |
| "eval_wer": 1.008964507866813, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5802707930367504, | |
| "eval_cer": 4.815741763569977, | |
| "eval_loss": 81.74565124511719, | |
| "eval_runtime": 231.6247, | |
| "eval_samples_per_second": 23.577, | |
| "eval_steps_per_second": 2.949, | |
| "eval_wer": 1.000365898280278, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7736943907156673, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 75.43038940429688, | |
| "eval_runtime": 227.0361, | |
| "eval_samples_per_second": 24.053, | |
| "eval_steps_per_second": 3.008, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9671179883945842, | |
| "grad_norm": 140.793212890625, | |
| "learning_rate": 1.188e-06, | |
| "loss": 66.0277, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9671179883945842, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 63.1251106262207, | |
| "eval_runtime": 226.7903, | |
| "eval_samples_per_second": 24.08, | |
| "eval_steps_per_second": 3.012, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1605415860735009, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 57.10499572753906, | |
| "eval_runtime": 255.6303, | |
| "eval_samples_per_second": 21.363, | |
| "eval_steps_per_second": 2.672, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3539651837524178, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 55.67994689941406, | |
| "eval_runtime": 269.8181, | |
| "eval_samples_per_second": 20.24, | |
| "eval_steps_per_second": 2.531, | |
| "eval_wer": 1.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5473887814313345, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 55.04756546020508, | |
| "eval_runtime": 241.9741, | |
| "eval_samples_per_second": 22.569, | |
| "eval_steps_per_second": 2.823, | |
| "eval_wer": 1.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7408123791102514, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 54.408546447753906, | |
| "eval_runtime": 256.9455, | |
| "eval_samples_per_second": 21.254, | |
| "eval_steps_per_second": 2.658, | |
| "eval_wer": 1.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9342359767891684, | |
| "grad_norm": 141.13522338867188, | |
| "learning_rate": 2.3855999999999997e-06, | |
| "loss": 46.3141, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9342359767891684, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 53.68925857543945, | |
| "eval_runtime": 260.4063, | |
| "eval_samples_per_second": 20.971, | |
| "eval_steps_per_second": 2.623, | |
| "eval_wer": 1.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 52.97110366821289, | |
| "eval_runtime": 254.697, | |
| "eval_samples_per_second": 21.441, | |
| "eval_steps_per_second": 2.682, | |
| "eval_wer": 1.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3210831721470018, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 52.13261032104492, | |
| "eval_runtime": 246.5799, | |
| "eval_samples_per_second": 22.147, | |
| "eval_steps_per_second": 2.77, | |
| "eval_wer": 1.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.514506769825919, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 51.254920959472656, | |
| "eval_runtime": 260.2756, | |
| "eval_samples_per_second": 20.982, | |
| "eval_steps_per_second": 2.624, | |
| "eval_wer": 1.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7079303675048356, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 50.26494598388672, | |
| "eval_runtime": 261.2096, | |
| "eval_samples_per_second": 20.907, | |
| "eval_steps_per_second": 2.615, | |
| "eval_wer": 1.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9013539651837523, | |
| "grad_norm": 120.8663101196289, | |
| "learning_rate": 3.5856e-06, | |
| "loss": 42.8642, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.9013539651837523, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 49.20811080932617, | |
| "eval_runtime": 258.4268, | |
| "eval_samples_per_second": 21.132, | |
| "eval_steps_per_second": 2.643, | |
| "eval_wer": 1.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.094777562862669, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 48.105064392089844, | |
| "eval_runtime": 247.347, | |
| "eval_samples_per_second": 22.078, | |
| "eval_steps_per_second": 2.761, | |
| "eval_wer": 1.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.288201160541586, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 46.87884521484375, | |
| "eval_runtime": 262.3223, | |
| "eval_samples_per_second": 20.818, | |
| "eval_steps_per_second": 2.604, | |
| "eval_wer": 1.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.481624758220503, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 45.54108810424805, | |
| "eval_runtime": 228.7269, | |
| "eval_samples_per_second": 23.876, | |
| "eval_steps_per_second": 2.986, | |
| "eval_wer": 1.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.6750483558994196, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 44.151554107666016, | |
| "eval_runtime": 228.7844, | |
| "eval_samples_per_second": 23.87, | |
| "eval_steps_per_second": 2.985, | |
| "eval_wer": 1.0, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.8684719535783367, | |
| "grad_norm": 172.9249725341797, | |
| "learning_rate": 4.7856e-06, | |
| "loss": 38.3378, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.8684719535783367, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 42.60865783691406, | |
| "eval_runtime": 227.6693, | |
| "eval_samples_per_second": 23.987, | |
| "eval_steps_per_second": 3.0, | |
| "eval_wer": 1.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.061895551257253, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 40.9814567565918, | |
| "eval_runtime": 230.7363, | |
| "eval_samples_per_second": 23.668, | |
| "eval_steps_per_second": 2.96, | |
| "eval_wer": 1.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 39.240055084228516, | |
| "eval_runtime": 230.4939, | |
| "eval_samples_per_second": 23.693, | |
| "eval_steps_per_second": 2.963, | |
| "eval_wer": 1.0, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.448742746615087, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 37.40217590332031, | |
| "eval_runtime": 228.8585, | |
| "eval_samples_per_second": 23.862, | |
| "eval_steps_per_second": 2.984, | |
| "eval_wer": 1.0, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.6421663442940035, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 35.430931091308594, | |
| "eval_runtime": 229.2781, | |
| "eval_samples_per_second": 23.818, | |
| "eval_steps_per_second": 2.979, | |
| "eval_wer": 1.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.835589941972921, | |
| "grad_norm": 113.81422424316406, | |
| "learning_rate": 5.9856e-06, | |
| "loss": 31.9192, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.835589941972921, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 33.417510986328125, | |
| "eval_runtime": 233.5128, | |
| "eval_samples_per_second": 23.386, | |
| "eval_steps_per_second": 2.925, | |
| "eval_wer": 1.0, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.029013539651838, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 31.266008377075195, | |
| "eval_runtime": 235.6664, | |
| "eval_samples_per_second": 23.173, | |
| "eval_steps_per_second": 2.898, | |
| "eval_wer": 1.0, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.222437137330754, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 29.01472282409668, | |
| "eval_runtime": 238.0062, | |
| "eval_samples_per_second": 22.945, | |
| "eval_steps_per_second": 2.87, | |
| "eval_wer": 1.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.415860735009671, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 26.68852996826172, | |
| "eval_runtime": 267.9224, | |
| "eval_samples_per_second": 20.383, | |
| "eval_steps_per_second": 2.549, | |
| "eval_wer": 1.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.609284332688588, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 24.301000595092773, | |
| "eval_runtime": 235.2693, | |
| "eval_samples_per_second": 23.212, | |
| "eval_steps_per_second": 2.903, | |
| "eval_wer": 1.0, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.802707930367505, | |
| "grad_norm": 106.30471801757812, | |
| "learning_rate": 7.1856e-06, | |
| "loss": 23.4284, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.802707930367505, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 21.88077735900879, | |
| "eval_runtime": 232.75, | |
| "eval_samples_per_second": 23.463, | |
| "eval_steps_per_second": 2.934, | |
| "eval_wer": 1.0, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.996131528046422, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 19.473451614379883, | |
| "eval_runtime": 233.7728, | |
| "eval_samples_per_second": 23.36, | |
| "eval_steps_per_second": 2.922, | |
| "eval_wer": 1.0, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.189555125725338, | |
| "eval_cer": 0.9908760508984813, | |
| "eval_loss": 17.129289627075195, | |
| "eval_runtime": 240.7161, | |
| "eval_samples_per_second": 22.686, | |
| "eval_steps_per_second": 2.837, | |
| "eval_wer": 1.0, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.382978723404255, | |
| "eval_cer": 0.9907917260454173, | |
| "eval_loss": 14.863801002502441, | |
| "eval_runtime": 233.6375, | |
| "eval_samples_per_second": 23.374, | |
| "eval_steps_per_second": 2.923, | |
| "eval_wer": 1.0, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.576402321083172, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 12.806206703186035, | |
| "eval_runtime": 251.782, | |
| "eval_samples_per_second": 21.689, | |
| "eval_steps_per_second": 2.713, | |
| "eval_wer": 1.0, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.769825918762089, | |
| "grad_norm": 82.23467254638672, | |
| "learning_rate": 8.3856e-06, | |
| "loss": 13.9431, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.769825918762089, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 10.964253425598145, | |
| "eval_runtime": 233.2904, | |
| "eval_samples_per_second": 23.409, | |
| "eval_steps_per_second": 2.928, | |
| "eval_wer": 1.0, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.963249516441006, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 9.411906242370605, | |
| "eval_runtime": 235.0547, | |
| "eval_samples_per_second": 23.233, | |
| "eval_steps_per_second": 2.906, | |
| "eval_wer": 1.0, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 7.156673114119923, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 8.164007186889648, | |
| "eval_runtime": 287.0729, | |
| "eval_samples_per_second": 19.023, | |
| "eval_steps_per_second": 2.379, | |
| "eval_wer": 1.0, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.350096711798839, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 7.22973108291626, | |
| "eval_runtime": 237.3392, | |
| "eval_samples_per_second": 23.009, | |
| "eval_steps_per_second": 2.878, | |
| "eval_wer": 1.0, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.543520309477756, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 6.571568489074707, | |
| "eval_runtime": 232.6033, | |
| "eval_samples_per_second": 23.478, | |
| "eval_steps_per_second": 2.936, | |
| "eval_wer": 1.0, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.7369439071566735, | |
| "grad_norm": 4.932778358459473, | |
| "learning_rate": 9.585600000000002e-06, | |
| "loss": 7.4585, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.7369439071566735, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 6.141250133514404, | |
| "eval_runtime": 234.2968, | |
| "eval_samples_per_second": 23.308, | |
| "eval_steps_per_second": 2.915, | |
| "eval_wer": 1.0, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.93036750483559, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 5.885389804840088, | |
| "eval_runtime": 236.1436, | |
| "eval_samples_per_second": 23.126, | |
| "eval_steps_per_second": 2.892, | |
| "eval_wer": 1.0, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 8.123791102514506, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 5.770660877227783, | |
| "eval_runtime": 234.5762, | |
| "eval_samples_per_second": 23.28, | |
| "eval_steps_per_second": 2.912, | |
| "eval_wer": 1.0, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.317214700193423, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 5.680200576782227, | |
| "eval_runtime": 231.827, | |
| "eval_samples_per_second": 23.556, | |
| "eval_steps_per_second": 2.946, | |
| "eval_wer": 1.0, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.51063829787234, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 5.597055435180664, | |
| "eval_runtime": 232.7784, | |
| "eval_samples_per_second": 23.46, | |
| "eval_steps_per_second": 2.934, | |
| "eval_wer": 1.0, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.704061895551257, | |
| "grad_norm": 1.6451424360275269, | |
| "learning_rate": 1.07856e-05, | |
| "loss": 5.7398, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.704061895551257, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 5.533324241638184, | |
| "eval_runtime": 232.5892, | |
| "eval_samples_per_second": 23.479, | |
| "eval_steps_per_second": 2.937, | |
| "eval_wer": 1.0, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.897485493230175, | |
| "eval_cer": 0.9907074011923535, | |
| "eval_loss": 5.475062370300293, | |
| "eval_runtime": 233.1057, | |
| "eval_samples_per_second": 23.427, | |
| "eval_steps_per_second": 2.93, | |
| "eval_wer": 1.0, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 5.425434589385986, | |
| "eval_runtime": 232.7779, | |
| "eval_samples_per_second": 23.46, | |
| "eval_steps_per_second": 2.934, | |
| "eval_wer": 1.0, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.284332688588007, | |
| "eval_cer": 0.9908423209572558, | |
| "eval_loss": 5.377471923828125, | |
| "eval_runtime": 234.4379, | |
| "eval_samples_per_second": 23.294, | |
| "eval_steps_per_second": 2.913, | |
| "eval_wer": 1.1319063300402488, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.477756286266924, | |
| "eval_cer": 0.9907495636188854, | |
| "eval_loss": 5.343258380889893, | |
| "eval_runtime": 239.0036, | |
| "eval_samples_per_second": 22.849, | |
| "eval_steps_per_second": 2.858, | |
| "eval_wer": 1.33205268935236, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.671179883945841, | |
| "grad_norm": 2.292585611343384, | |
| "learning_rate": 1.19856e-05, | |
| "loss": 5.4159, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.671179883945841, | |
| "eval_cer": 0.9906146438539831, | |
| "eval_loss": 5.311874866485596, | |
| "eval_runtime": 235.3723, | |
| "eval_samples_per_second": 23.202, | |
| "eval_steps_per_second": 2.902, | |
| "eval_wer": 1.686242224661544, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.864603481624759, | |
| "eval_cer": 0.9909688082368516, | |
| "eval_loss": 5.269064426422119, | |
| "eval_runtime": 234.5017, | |
| "eval_samples_per_second": 23.288, | |
| "eval_steps_per_second": 2.913, | |
| "eval_wer": 1.4255396999634102, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 10.058027079303676, | |
| "eval_cer": 0.9909182133250133, | |
| "eval_loss": 5.236879825592041, | |
| "eval_runtime": 235.3148, | |
| "eval_samples_per_second": 23.207, | |
| "eval_steps_per_second": 2.902, | |
| "eval_wer": 1.4043175997072814, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.251450676982591, | |
| "eval_cer": 0.9909603757515453, | |
| "eval_loss": 5.194947719573975, | |
| "eval_runtime": 232.4545, | |
| "eval_samples_per_second": 23.493, | |
| "eval_steps_per_second": 2.938, | |
| "eval_wer": 1.5686059275521405, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.444874274661508, | |
| "eval_cer": 0.9908170235013366, | |
| "eval_loss": 5.151918888092041, | |
| "eval_runtime": 231.5581, | |
| "eval_samples_per_second": 23.584, | |
| "eval_steps_per_second": 2.95, | |
| "eval_wer": 1.5166483717526527, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.638297872340425, | |
| "grad_norm": 2.0632503032684326, | |
| "learning_rate": 1.3185600000000001e-05, | |
| "loss": 5.2163, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.638297872340425, | |
| "eval_cer": 0.9909603757515453, | |
| "eval_loss": 5.108114242553711, | |
| "eval_runtime": 232.6627, | |
| "eval_samples_per_second": 23.472, | |
| "eval_steps_per_second": 2.936, | |
| "eval_wer": 1.247713135748262, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.831721470019342, | |
| "eval_cer": 0.9907664285894981, | |
| "eval_loss": 5.055335998535156, | |
| "eval_runtime": 231.9668, | |
| "eval_samples_per_second": 23.542, | |
| "eval_steps_per_second": 2.944, | |
| "eval_wer": 1.5124405415294548, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 11.02514506769826, | |
| "eval_cer": 0.9908760508984813, | |
| "eval_loss": 5.012271881103516, | |
| "eval_runtime": 231.643, | |
| "eval_samples_per_second": 23.575, | |
| "eval_steps_per_second": 2.949, | |
| "eval_wer": 1.5495792169776803, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 11.218568665377177, | |
| "eval_cer": 0.9885824148951421, | |
| "eval_loss": 4.942389965057373, | |
| "eval_runtime": 232.3058, | |
| "eval_samples_per_second": 23.508, | |
| "eval_steps_per_second": 2.94, | |
| "eval_wer": 1.7621661178192463, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.411992263056092, | |
| "eval_cer": 0.9830675695047602, | |
| "eval_loss": 4.8753485679626465, | |
| "eval_runtime": 233.7911, | |
| "eval_samples_per_second": 23.358, | |
| "eval_steps_per_second": 2.921, | |
| "eval_wer": 1.5404317599707282, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.60541586073501, | |
| "grad_norm": 1.9146788120269775, | |
| "learning_rate": 1.43856e-05, | |
| "loss": 4.9465, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.60541586073501, | |
| "eval_cer": 0.9749808160959279, | |
| "eval_loss": 4.77677059173584, | |
| "eval_runtime": 232.0203, | |
| "eval_samples_per_second": 23.537, | |
| "eval_steps_per_second": 2.944, | |
| "eval_wer": 1.853457738748628, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.798839458413926, | |
| "eval_cer": 0.9713126849876464, | |
| "eval_loss": 4.6841044425964355, | |
| "eval_runtime": 231.7273, | |
| "eval_samples_per_second": 23.566, | |
| "eval_steps_per_second": 2.947, | |
| "eval_wer": 1.8395536040980607, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.992263056092844, | |
| "eval_cer": 0.9697020802941251, | |
| "eval_loss": 4.582820892333984, | |
| "eval_runtime": 232.9124, | |
| "eval_samples_per_second": 23.447, | |
| "eval_steps_per_second": 2.932, | |
| "eval_wer": 1.7444200512257593, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 12.18568665377176, | |
| "eval_cer": 0.968909426675324, | |
| "eval_loss": 4.485307693481445, | |
| "eval_runtime": 232.9842, | |
| "eval_samples_per_second": 23.439, | |
| "eval_steps_per_second": 2.932, | |
| "eval_wer": 1.801317233809001, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.379110251450676, | |
| "eval_cer": 0.9556029648618337, | |
| "eval_loss": 4.395504951477051, | |
| "eval_runtime": 233.0172, | |
| "eval_samples_per_second": 23.436, | |
| "eval_steps_per_second": 2.931, | |
| "eval_wer": 1.827844859129162, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.572533849129593, | |
| "grad_norm": 4.459765434265137, | |
| "learning_rate": 1.55856e-05, | |
| "loss": 4.5094, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.572533849129593, | |
| "eval_cer": 0.9123021528134987, | |
| "eval_loss": 4.284241676330566, | |
| "eval_runtime": 230.5488, | |
| "eval_samples_per_second": 23.687, | |
| "eval_steps_per_second": 2.962, | |
| "eval_wer": 1.8728503476033662, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.76595744680851, | |
| "eval_cer": 0.8650380726711584, | |
| "eval_loss": 4.181938171386719, | |
| "eval_runtime": 232.0403, | |
| "eval_samples_per_second": 23.535, | |
| "eval_steps_per_second": 2.943, | |
| "eval_wer": 1.9094401756311745, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.959381044487428, | |
| "eval_cer": 0.848628456264915, | |
| "eval_loss": 4.074057579040527, | |
| "eval_runtime": 231.4468, | |
| "eval_samples_per_second": 23.595, | |
| "eval_steps_per_second": 2.951, | |
| "eval_wer": 1.9134650567142335, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 13.152804642166345, | |
| "eval_cer": 0.8385937987503057, | |
| "eval_loss": 3.9648523330688477, | |
| "eval_runtime": 231.3067, | |
| "eval_samples_per_second": 23.609, | |
| "eval_steps_per_second": 2.953, | |
| "eval_wer": 1.9191364800585438, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.346228239845262, | |
| "eval_cer": 0.8189292430157941, | |
| "eval_loss": 3.8640658855438232, | |
| "eval_runtime": 237.0711, | |
| "eval_samples_per_second": 23.035, | |
| "eval_steps_per_second": 2.881, | |
| "eval_wer": 1.9195023783388219, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.539651837524177, | |
| "grad_norm": 4.586193084716797, | |
| "learning_rate": 1.67856e-05, | |
| "loss": 4.0097, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.539651837524177, | |
| "eval_cer": 0.8014487009756386, | |
| "eval_loss": 3.7686750888824463, | |
| "eval_runtime": 244.328, | |
| "eval_samples_per_second": 22.351, | |
| "eval_steps_per_second": 2.795, | |
| "eval_wer": 1.9275521405049396, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.733075435203094, | |
| "eval_cer": 0.7963301823946571, | |
| "eval_loss": 3.680776596069336, | |
| "eval_runtime": 230.3109, | |
| "eval_samples_per_second": 23.711, | |
| "eval_steps_per_second": 2.966, | |
| "eval_wer": 1.9259055982436883, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.926499032882012, | |
| "eval_cer": 0.779220669707983, | |
| "eval_loss": 3.60208797454834, | |
| "eval_runtime": 229.5487, | |
| "eval_samples_per_second": 23.79, | |
| "eval_steps_per_second": 2.975, | |
| "eval_wer": 1.9275521405049396, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 14.119922630560929, | |
| "eval_cer": 0.777466712764253, | |
| "eval_loss": 3.55332088470459, | |
| "eval_runtime": 231.4776, | |
| "eval_samples_per_second": 23.592, | |
| "eval_steps_per_second": 2.951, | |
| "eval_wer": 1.9366995975118917, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 14.313346228239846, | |
| "eval_cer": 0.775071886937237, | |
| "eval_loss": 3.476841449737549, | |
| "eval_runtime": 240.5885, | |
| "eval_samples_per_second": 22.699, | |
| "eval_steps_per_second": 2.839, | |
| "eval_wer": 1.9321258690084158, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.506769825918763, | |
| "grad_norm": 4.310749053955078, | |
| "learning_rate": 1.7985600000000003e-05, | |
| "loss": 3.5619, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.506769825918763, | |
| "eval_cer": 0.7672043781463711, | |
| "eval_loss": 3.4284844398498535, | |
| "eval_runtime": 240.9512, | |
| "eval_samples_per_second": 22.664, | |
| "eval_steps_per_second": 2.835, | |
| "eval_wer": 1.938529088913282, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.700193423597678, | |
| "eval_cer": 0.7659816677769439, | |
| "eval_loss": 3.362793445587158, | |
| "eval_runtime": 230.1401, | |
| "eval_samples_per_second": 23.729, | |
| "eval_steps_per_second": 2.968, | |
| "eval_wer": 1.9361507500914745, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.893617021276595, | |
| "eval_cer": 0.7618919124033426, | |
| "eval_loss": 3.2909657955169678, | |
| "eval_runtime": 233.6336, | |
| "eval_samples_per_second": 23.374, | |
| "eval_steps_per_second": 2.923, | |
| "eval_wer": 1.9313940724478595, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 15.087040618955513, | |
| "eval_cer": 0.7486360455016907, | |
| "eval_loss": 3.2242627143859863, | |
| "eval_runtime": 241.9894, | |
| "eval_samples_per_second": 22.567, | |
| "eval_steps_per_second": 2.822, | |
| "eval_wer": 1.928832784485913, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 15.28046421663443, | |
| "eval_cer": 0.7431802275084536, | |
| "eval_loss": 3.164518117904663, | |
| "eval_runtime": 232.2322, | |
| "eval_samples_per_second": 23.515, | |
| "eval_steps_per_second": 2.941, | |
| "eval_wer": 1.9308452250274424, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.473887814313347, | |
| "grad_norm": 5.67767333984375, | |
| "learning_rate": 1.91856e-05, | |
| "loss": 3.2379, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.473887814313347, | |
| "eval_cer": 0.7382556560895193, | |
| "eval_loss": 3.1185944080352783, | |
| "eval_runtime": 235.8886, | |
| "eval_samples_per_second": 23.151, | |
| "eval_steps_per_second": 2.895, | |
| "eval_wer": 1.9332235638492499, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.667311411992262, | |
| "eval_cer": 0.7374714349560246, | |
| "eval_loss": 3.078275203704834, | |
| "eval_runtime": 234.7574, | |
| "eval_samples_per_second": 23.262, | |
| "eval_steps_per_second": 2.909, | |
| "eval_wer": 1.9348701061105014, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.86073500967118, | |
| "eval_cer": 0.727942726559799, | |
| "eval_loss": 3.0145950317382812, | |
| "eval_runtime": 231.4339, | |
| "eval_samples_per_second": 23.596, | |
| "eval_steps_per_second": 2.951, | |
| "eval_wer": 1.9321258690084158, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 16.054158607350097, | |
| "eval_cer": 0.7300086854598656, | |
| "eval_loss": 2.9523487091064453, | |
| "eval_runtime": 235.1315, | |
| "eval_samples_per_second": 23.225, | |
| "eval_steps_per_second": 2.905, | |
| "eval_wer": 1.9308452250274424, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 16.247582205029012, | |
| "eval_cer": 0.7254467109091063, | |
| "eval_loss": 2.918687105178833, | |
| "eval_runtime": 238.3374, | |
| "eval_samples_per_second": 22.913, | |
| "eval_steps_per_second": 2.866, | |
| "eval_wer": 1.9273691913648006, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.44100580270793, | |
| "grad_norm": 8.755329132080078, | |
| "learning_rate": 2.03856e-05, | |
| "loss": 2.9448, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.44100580270793, | |
| "eval_cer": 0.7177394193390618, | |
| "eval_loss": 2.8671371936798096, | |
| "eval_runtime": 229.0989, | |
| "eval_samples_per_second": 23.837, | |
| "eval_steps_per_second": 2.981, | |
| "eval_wer": 1.929015733626052, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.634429400386846, | |
| "eval_cer": 0.7115752725800876, | |
| "eval_loss": 2.8188540935516357, | |
| "eval_runtime": 230.6192, | |
| "eval_samples_per_second": 23.68, | |
| "eval_steps_per_second": 2.962, | |
| "eval_wer": 1.9348701061105014, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.827852998065765, | |
| "eval_cer": 0.70778065419221, | |
| "eval_loss": 2.76908802986145, | |
| "eval_runtime": 239.0894, | |
| "eval_samples_per_second": 22.841, | |
| "eval_steps_per_second": 2.857, | |
| "eval_wer": 1.9365166483717526, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 17.02127659574468, | |
| "eval_cer": 0.7069205406909579, | |
| "eval_loss": 2.731661081314087, | |
| "eval_runtime": 231.1522, | |
| "eval_samples_per_second": 23.625, | |
| "eval_steps_per_second": 2.955, | |
| "eval_wer": 1.9420051225759238, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 17.214700193423596, | |
| "eval_cer": 0.7056388029243859, | |
| "eval_loss": 2.683185577392578, | |
| "eval_runtime": 229.2327, | |
| "eval_samples_per_second": 23.823, | |
| "eval_steps_per_second": 2.98, | |
| "eval_wer": 1.9489571899012075, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.408123791102515, | |
| "grad_norm": 4.358935832977295, | |
| "learning_rate": 2.1585600000000002e-05, | |
| "loss": 2.6749, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.408123791102515, | |
| "eval_cer": 0.7020128342426364, | |
| "eval_loss": 2.6420364379882812, | |
| "eval_runtime": 229.374, | |
| "eval_samples_per_second": 23.808, | |
| "eval_steps_per_second": 2.978, | |
| "eval_wer": 1.978412001463593, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.60154738878143, | |
| "eval_cer": 0.6991204917825431, | |
| "eval_loss": 2.601982831954956, | |
| "eval_runtime": 229.7146, | |
| "eval_samples_per_second": 23.773, | |
| "eval_steps_per_second": 2.973, | |
| "eval_wer": 1.9414562751555067, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.79497098646035, | |
| "eval_cer": 0.6994577911947989, | |
| "eval_loss": 2.5666821002960205, | |
| "eval_runtime": 234.5424, | |
| "eval_samples_per_second": 23.284, | |
| "eval_steps_per_second": 2.912, | |
| "eval_wer": 1.9762166117819246, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.988394584139265, | |
| "eval_cer": 0.6771285701034666, | |
| "eval_loss": 2.517096757888794, | |
| "eval_runtime": 247.5848, | |
| "eval_samples_per_second": 22.057, | |
| "eval_steps_per_second": 2.759, | |
| "eval_wer": 1.9857299670691548, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 18.181818181818183, | |
| "eval_cer": 0.6774658695157224, | |
| "eval_loss": 2.492238759994507, | |
| "eval_runtime": 229.5938, | |
| "eval_samples_per_second": 23.785, | |
| "eval_steps_per_second": 2.975, | |
| "eval_wer": 1.9890230515916576, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 18.3752417794971, | |
| "grad_norm": 6.681089878082275, | |
| "learning_rate": 2.27856e-05, | |
| "loss": 2.4473, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.3752417794971, | |
| "eval_cer": 0.6682575955611397, | |
| "eval_loss": 2.445500373840332, | |
| "eval_runtime": 229.8314, | |
| "eval_samples_per_second": 23.761, | |
| "eval_steps_per_second": 2.972, | |
| "eval_wer": 1.9882912550311014, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.568665377176014, | |
| "eval_cer": 0.6620681513462463, | |
| "eval_loss": 2.4191782474517822, | |
| "eval_runtime": 230.0128, | |
| "eval_samples_per_second": 23.742, | |
| "eval_steps_per_second": 2.969, | |
| "eval_wer": 1.9815221368459568, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.762088974854933, | |
| "eval_cer": 0.6523454957879736, | |
| "eval_loss": 2.386597156524658, | |
| "eval_runtime": 240.9084, | |
| "eval_samples_per_second": 22.668, | |
| "eval_steps_per_second": 2.835, | |
| "eval_wer": 1.9904866447127698, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.95551257253385, | |
| "eval_cer": 0.6539392355108822, | |
| "eval_loss": 2.335400342941284, | |
| "eval_runtime": 233.6249, | |
| "eval_samples_per_second": 23.375, | |
| "eval_steps_per_second": 2.923, | |
| "eval_wer": 1.991401390413465, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 19.148936170212767, | |
| "eval_cer": 0.6515612746544789, | |
| "eval_loss": 2.3113534450531006, | |
| "eval_runtime": 252.8489, | |
| "eval_samples_per_second": 21.598, | |
| "eval_steps_per_second": 2.701, | |
| "eval_wer": 1.9924990852542992, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 19.342359767891683, | |
| "grad_norm": 8.786458015441895, | |
| "learning_rate": 2.39856e-05, | |
| "loss": 2.2307, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.342359767891683, | |
| "eval_cer": 0.645447722807343, | |
| "eval_loss": 2.269487142562866, | |
| "eval_runtime": 236.1588, | |
| "eval_samples_per_second": 23.124, | |
| "eval_steps_per_second": 2.892, | |
| "eval_wer": 1.9903036955726308, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.535783365570598, | |
| "eval_cer": 0.6464258911028847, | |
| "eval_loss": 2.246647834777832, | |
| "eval_runtime": 241.6204, | |
| "eval_samples_per_second": 22.602, | |
| "eval_steps_per_second": 2.827, | |
| "eval_wer": 1.9924990852542992, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.729206963249517, | |
| "eval_cer": 0.6422939733027515, | |
| "eval_loss": 2.2167210578918457, | |
| "eval_runtime": 230.6902, | |
| "eval_samples_per_second": 23.672, | |
| "eval_steps_per_second": 2.961, | |
| "eval_wer": 1.9928649835345773, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.922630560928432, | |
| "eval_cer": 0.641256777610065, | |
| "eval_loss": 2.176206588745117, | |
| "eval_runtime": 236.5184, | |
| "eval_samples_per_second": 23.089, | |
| "eval_steps_per_second": 2.888, | |
| "eval_wer": 1.991401390413465, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10340, | |
| "total_flos": 1.4857862206321902e+19, | |
| "train_loss": 15.407913101757057, | |
| "train_runtime": 65838.1292, | |
| "train_samples_per_second": 5.022, | |
| "train_steps_per_second": 0.157 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10340, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4857862206321902e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |