| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 24.12123763418228, | |
| "eval_steps": 100, | |
| "global_step": 57300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.6430506706237793, | |
| "learning_rate": 9.5e-06, | |
| "loss": 5.6141, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "eval_cer": 0.9770320871299519, | |
| "eval_loss": 2.355087995529175, | |
| "eval_runtime": 329.8418, | |
| "eval_samples_per_second": 28.735, | |
| "eval_steps_per_second": 3.593, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.838773012161255, | |
| "learning_rate": 1.9500000000000003e-05, | |
| "loss": 2.1498, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_cer": 0.6385162879824802, | |
| "eval_loss": 1.961960792541504, | |
| "eval_runtime": 334.0577, | |
| "eval_samples_per_second": 28.372, | |
| "eval_steps_per_second": 3.547, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.848980665206909, | |
| "learning_rate": 2.95e-05, | |
| "loss": 2.6461, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_cer": 0.5852479371162645, | |
| "eval_loss": 2.6564223766326904, | |
| "eval_runtime": 350.7705, | |
| "eval_samples_per_second": 27.021, | |
| "eval_steps_per_second": 3.378, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 3.9299519062042236, | |
| "learning_rate": 3.9500000000000005e-05, | |
| "loss": 2.359, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_cer": 0.5907913261116108, | |
| "eval_loss": 2.6195201873779297, | |
| "eval_runtime": 341.3159, | |
| "eval_samples_per_second": 27.769, | |
| "eval_steps_per_second": 3.472, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 4.236645698547363, | |
| "learning_rate": 4.9500000000000004e-05, | |
| "loss": 4.2363, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_cer": 0.624139650385202, | |
| "eval_loss": 2.457566261291504, | |
| "eval_runtime": 356.3471, | |
| "eval_samples_per_second": 26.598, | |
| "eval_steps_per_second": 3.325, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 13.769584655761719, | |
| "learning_rate": 4.9932862190812725e-05, | |
| "loss": 2.6933, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_cer": 0.6483369833013961, | |
| "eval_loss": 3.4720070362091064, | |
| "eval_runtime": 345.5879, | |
| "eval_samples_per_second": 27.426, | |
| "eval_steps_per_second": 3.429, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 4.292990207672119, | |
| "learning_rate": 4.986289752650177e-05, | |
| "loss": 4.1075, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_cer": 0.6181098705564898, | |
| "eval_loss": 3.3849761486053467, | |
| "eval_runtime": 353.3198, | |
| "eval_samples_per_second": 26.826, | |
| "eval_steps_per_second": 3.354, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 4.32611083984375, | |
| "learning_rate": 4.97922261484099e-05, | |
| "loss": 2.6907, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_cer": 0.6191510891243988, | |
| "eval_loss": 2.5412750244140625, | |
| "eval_runtime": 343.4967, | |
| "eval_samples_per_second": 27.593, | |
| "eval_steps_per_second": 3.45, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 3.173403263092041, | |
| "learning_rate": 4.972155477031802e-05, | |
| "loss": 4.2294, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_cer": 0.5840918423213797, | |
| "eval_loss": 2.6817102432250977, | |
| "eval_runtime": 354.5341, | |
| "eval_samples_per_second": 26.734, | |
| "eval_steps_per_second": 3.342, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 4.8158440589904785, | |
| "learning_rate": 4.965088339222615e-05, | |
| "loss": 2.2949, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_cer": 0.5736649915920379, | |
| "eval_loss": 1.9513429403305054, | |
| "eval_runtime": 345.1224, | |
| "eval_samples_per_second": 27.463, | |
| "eval_steps_per_second": 3.434, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 3.0650887489318848, | |
| "learning_rate": 4.958021201413428e-05, | |
| "loss": 2.6115, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_cer": 0.5645799929607759, | |
| "eval_loss": 2.8409903049468994, | |
| "eval_runtime": 352.0824, | |
| "eval_samples_per_second": 26.92, | |
| "eval_steps_per_second": 3.366, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 3.0540664196014404, | |
| "learning_rate": 4.950954063604241e-05, | |
| "loss": 2.3453, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_cer": 0.598292499315631, | |
| "eval_loss": 1.8713030815124512, | |
| "eval_runtime": 344.1368, | |
| "eval_samples_per_second": 27.541, | |
| "eval_steps_per_second": 3.443, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.366485118865967, | |
| "learning_rate": 4.9438869257950535e-05, | |
| "loss": 3.2578, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_cer": 0.5734816784638849, | |
| "eval_loss": 2.000699758529663, | |
| "eval_runtime": 350.6902, | |
| "eval_samples_per_second": 27.027, | |
| "eval_steps_per_second": 3.379, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 81.67564392089844, | |
| "learning_rate": 4.936819787985866e-05, | |
| "loss": 1.9657, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_cer": 0.6325427241797349, | |
| "eval_loss": 3.403754472732544, | |
| "eval_runtime": 340.0489, | |
| "eval_samples_per_second": 27.872, | |
| "eval_steps_per_second": 3.485, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.837653875350952, | |
| "learning_rate": 4.929752650176679e-05, | |
| "loss": 2.6865, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_cer": 0.5574918853388604, | |
| "eval_loss": 3.5201330184936523, | |
| "eval_runtime": 351.432, | |
| "eval_samples_per_second": 26.97, | |
| "eval_steps_per_second": 3.372, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 2.9047458171844482, | |
| "learning_rate": 4.922685512367491e-05, | |
| "loss": 2.4105, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_cer": 0.5886624496499941, | |
| "eval_loss": 3.5269789695739746, | |
| "eval_runtime": 347.023, | |
| "eval_samples_per_second": 27.312, | |
| "eval_steps_per_second": 3.415, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.9270007610321045, | |
| "learning_rate": 4.915618374558304e-05, | |
| "loss": 2.418, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "eval_cer": 0.624877791247898, | |
| "eval_loss": 2.498126983642578, | |
| "eval_runtime": 355.9419, | |
| "eval_samples_per_second": 26.628, | |
| "eval_steps_per_second": 3.329, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.458078622817993, | |
| "learning_rate": 4.908551236749117e-05, | |
| "loss": 2.6118, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_cer": 0.5718074185600877, | |
| "eval_loss": 3.407224655151367, | |
| "eval_runtime": 343.8365, | |
| "eval_samples_per_second": 27.565, | |
| "eval_steps_per_second": 3.446, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.9933278560638428, | |
| "learning_rate": 4.90148409893993e-05, | |
| "loss": 3.2046, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_cer": 0.5539844941535333, | |
| "eval_loss": 1.928993821144104, | |
| "eval_runtime": 357.2509, | |
| "eval_samples_per_second": 26.53, | |
| "eval_steps_per_second": 3.317, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.917328357696533, | |
| "learning_rate": 4.8944169611307425e-05, | |
| "loss": 3.1904, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_cer": 0.5568001838019632, | |
| "eval_loss": 2.729193925857544, | |
| "eval_runtime": 341.3463, | |
| "eval_samples_per_second": 27.767, | |
| "eval_steps_per_second": 3.472, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.842353343963623, | |
| "learning_rate": 4.8873498233215547e-05, | |
| "loss": 2.1649, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_cer": 0.544757733369833, | |
| "eval_loss": 2.421653985977173, | |
| "eval_runtime": 358.7261, | |
| "eval_samples_per_second": 26.421, | |
| "eval_steps_per_second": 3.303, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.818225383758545, | |
| "learning_rate": 4.880282685512368e-05, | |
| "loss": 2.2775, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_cer": 0.5533832270931915, | |
| "eval_loss": 2.188101291656494, | |
| "eval_runtime": 354.4215, | |
| "eval_samples_per_second": 26.742, | |
| "eval_steps_per_second": 3.343, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 2.9436004161834717, | |
| "learning_rate": 4.87321554770318e-05, | |
| "loss": 1.6308, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_cer": 0.5162586523796489, | |
| "eval_loss": 3.3528623580932617, | |
| "eval_runtime": 360.9925, | |
| "eval_samples_per_second": 26.255, | |
| "eval_steps_per_second": 3.283, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.935622692108154, | |
| "learning_rate": 4.866148409893993e-05, | |
| "loss": 2.4462, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_cer": 0.575390579171718, | |
| "eval_loss": 2.0690765380859375, | |
| "eval_runtime": 343.7001, | |
| "eval_samples_per_second": 27.576, | |
| "eval_steps_per_second": 3.448, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.40964937210083, | |
| "learning_rate": 4.859081272084806e-05, | |
| "loss": 2.3368, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_cer": 0.5851746118650033, | |
| "eval_loss": 1.9030027389526367, | |
| "eval_runtime": 352.7743, | |
| "eval_samples_per_second": 26.867, | |
| "eval_steps_per_second": 3.359, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 3.7618064880371094, | |
| "learning_rate": 4.852084805653711e-05, | |
| "loss": 2.4157, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "eval_cer": 0.5566339798991045, | |
| "eval_loss": 3.213930368423462, | |
| "eval_runtime": 341.3273, | |
| "eval_samples_per_second": 27.768, | |
| "eval_steps_per_second": 3.472, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 9.61824893951416, | |
| "learning_rate": 4.845017667844523e-05, | |
| "loss": 2.7299, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_cer": 0.5819702983848891, | |
| "eval_loss": 1.7893775701522827, | |
| "eval_runtime": 353.5497, | |
| "eval_samples_per_second": 26.808, | |
| "eval_steps_per_second": 3.352, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 2.6824772357940674, | |
| "learning_rate": 4.8379505300353364e-05, | |
| "loss": 2.934, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_cer": 0.5119813460560791, | |
| "eval_loss": 2.804206609725952, | |
| "eval_runtime": 340.4666, | |
| "eval_samples_per_second": 27.838, | |
| "eval_steps_per_second": 3.481, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 4.691165924072266, | |
| "learning_rate": 4.8308833922261485e-05, | |
| "loss": 2.0012, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_cer": 0.5138047006374409, | |
| "eval_loss": 3.412604331970215, | |
| "eval_runtime": 349.7896, | |
| "eval_samples_per_second": 27.096, | |
| "eval_steps_per_second": 3.388, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 2.8883514404296875, | |
| "learning_rate": 4.823816254416961e-05, | |
| "loss": 3.5496, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_cer": 0.5417074029173673, | |
| "eval_loss": 1.7423540353775024, | |
| "eval_runtime": 339.3199, | |
| "eval_samples_per_second": 27.932, | |
| "eval_steps_per_second": 3.492, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 3.9499669075012207, | |
| "learning_rate": 4.816749116607774e-05, | |
| "loss": 2.4353, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_cer": 0.5479229400492746, | |
| "eval_loss": 3.4527170658111572, | |
| "eval_runtime": 354.1425, | |
| "eval_samples_per_second": 26.763, | |
| "eval_steps_per_second": 3.346, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 2.9536569118499756, | |
| "learning_rate": 4.809681978798587e-05, | |
| "loss": 2.9787, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "eval_cer": 0.5677500879903015, | |
| "eval_loss": 2.6674227714538574, | |
| "eval_runtime": 344.0522, | |
| "eval_samples_per_second": 27.548, | |
| "eval_steps_per_second": 3.444, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 3.094930648803711, | |
| "learning_rate": 4.8026148409894e-05, | |
| "loss": 2.2166, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_cer": 0.5857514371749247, | |
| "eval_loss": 3.2282423973083496, | |
| "eval_runtime": 358.5567, | |
| "eval_samples_per_second": 26.434, | |
| "eval_steps_per_second": 3.305, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 5.626856803894043, | |
| "learning_rate": 4.795547703180212e-05, | |
| "loss": 2.6222, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_cer": 0.5495043213014743, | |
| "eval_loss": 1.7686785459518433, | |
| "eval_runtime": 342.362, | |
| "eval_samples_per_second": 27.684, | |
| "eval_steps_per_second": 3.461, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 5.24758768081665, | |
| "learning_rate": 4.7884805653710253e-05, | |
| "loss": 2.078, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_cer": 0.5179402448085723, | |
| "eval_loss": 2.244272470474243, | |
| "eval_runtime": 354.1685, | |
| "eval_samples_per_second": 26.761, | |
| "eval_steps_per_second": 3.346, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 6.822329044342041, | |
| "learning_rate": 4.7814134275618375e-05, | |
| "loss": 2.7694, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_cer": 0.5457598451370693, | |
| "eval_loss": 2.137202501296997, | |
| "eval_runtime": 345.4818, | |
| "eval_samples_per_second": 27.434, | |
| "eval_steps_per_second": 3.43, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.648332357406616, | |
| "learning_rate": 4.77434628975265e-05, | |
| "loss": 1.74, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "eval_cer": 0.5568441789527199, | |
| "eval_loss": 3.1145806312561035, | |
| "eval_runtime": 362.0538, | |
| "eval_samples_per_second": 26.178, | |
| "eval_steps_per_second": 3.273, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 17.834945678710938, | |
| "learning_rate": 4.767279151943463e-05, | |
| "loss": 1.7586, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_cer": 0.5344653122678034, | |
| "eval_loss": 1.8897360563278198, | |
| "eval_runtime": 345.2891, | |
| "eval_samples_per_second": 27.449, | |
| "eval_steps_per_second": 3.432, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 7.307125568389893, | |
| "learning_rate": 4.760212014134276e-05, | |
| "loss": 1.6227, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_cer": 0.551083258378632, | |
| "eval_loss": 1.9913257360458374, | |
| "eval_runtime": 354.7374, | |
| "eval_samples_per_second": 26.718, | |
| "eval_steps_per_second": 3.34, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 4.344304084777832, | |
| "learning_rate": 4.753144876325089e-05, | |
| "loss": 3.1757, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "eval_cer": 0.5003299636306754, | |
| "eval_loss": 1.8998807668685913, | |
| "eval_runtime": 348.4123, | |
| "eval_samples_per_second": 27.203, | |
| "eval_steps_per_second": 3.401, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.8654837608337402, | |
| "learning_rate": 4.746077738515901e-05, | |
| "loss": 2.226, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "eval_cer": 0.5324904188338352, | |
| "eval_loss": 3.042088031768799, | |
| "eval_runtime": 354.1261, | |
| "eval_samples_per_second": 26.764, | |
| "eval_steps_per_second": 3.346, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 7.595367908477783, | |
| "learning_rate": 4.7390106007067143e-05, | |
| "loss": 2.7925, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_cer": 0.5470479253842243, | |
| "eval_loss": 3.0831844806671143, | |
| "eval_runtime": 344.2619, | |
| "eval_samples_per_second": 27.531, | |
| "eval_steps_per_second": 3.442, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 7.425869941711426, | |
| "learning_rate": 4.7319434628975265e-05, | |
| "loss": 3.2591, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_cer": 0.5522955691994837, | |
| "eval_loss": 1.7015308141708374, | |
| "eval_runtime": 353.5748, | |
| "eval_samples_per_second": 26.806, | |
| "eval_steps_per_second": 3.351, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.405174493789673, | |
| "learning_rate": 4.724876325088339e-05, | |
| "loss": 2.373, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_cer": 0.5220977865550819, | |
| "eval_loss": 2.095752477645874, | |
| "eval_runtime": 344.3236, | |
| "eval_samples_per_second": 27.526, | |
| "eval_steps_per_second": 3.442, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 3.823211908340454, | |
| "learning_rate": 4.717809187279153e-05, | |
| "loss": 2.3919, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_cer": 0.5331259043447656, | |
| "eval_loss": 2.41994571685791, | |
| "eval_runtime": 353.9416, | |
| "eval_samples_per_second": 26.778, | |
| "eval_steps_per_second": 3.348, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.0702412128448486, | |
| "learning_rate": 4.710742049469965e-05, | |
| "loss": 2.4842, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "eval_cer": 0.5329523679167807, | |
| "eval_loss": 3.652872085571289, | |
| "eval_runtime": 339.8921, | |
| "eval_samples_per_second": 27.885, | |
| "eval_steps_per_second": 3.486, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 3.5580544471740723, | |
| "learning_rate": 4.703745583038869e-05, | |
| "loss": 3.577, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "eval_cer": 0.5378284971256502, | |
| "eval_loss": 2.777487277984619, | |
| "eval_runtime": 356.1415, | |
| "eval_samples_per_second": 26.613, | |
| "eval_steps_per_second": 3.327, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 2.3930351734161377, | |
| "learning_rate": 4.6966784452296826e-05, | |
| "loss": 2.733, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_cer": 0.5305424113253294, | |
| "eval_loss": 3.2367630004882812, | |
| "eval_runtime": 346.0446, | |
| "eval_samples_per_second": 27.39, | |
| "eval_steps_per_second": 3.424, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 4.198298931121826, | |
| "learning_rate": 4.6896113074204954e-05, | |
| "loss": 1.616, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "eval_cer": 0.523026573071057, | |
| "eval_loss": 2.150926351547241, | |
| "eval_runtime": 371.9418, | |
| "eval_samples_per_second": 25.482, | |
| "eval_steps_per_second": 3.186, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.6696269512176514, | |
| "learning_rate": 4.6825441696113075e-05, | |
| "loss": 1.8252, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "eval_cer": 0.5257933792186461, | |
| "eval_loss": 3.1695752143859863, | |
| "eval_runtime": 351.2533, | |
| "eval_samples_per_second": 26.983, | |
| "eval_steps_per_second": 3.374, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 5.002595901489258, | |
| "learning_rate": 4.67547703180212e-05, | |
| "loss": 2.0594, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "eval_cer": 0.5161682179030933, | |
| "eval_loss": 2.4667067527770996, | |
| "eval_runtime": 377.3711, | |
| "eval_samples_per_second": 25.116, | |
| "eval_steps_per_second": 3.14, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 3.2300162315368652, | |
| "learning_rate": 4.668409893992933e-05, | |
| "loss": 1.8512, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_cer": 0.5493185639982793, | |
| "eval_loss": 1.743632435798645, | |
| "eval_runtime": 361.1881, | |
| "eval_samples_per_second": 26.241, | |
| "eval_steps_per_second": 3.281, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 2.109961986541748, | |
| "learning_rate": 4.661342756183746e-05, | |
| "loss": 2.4585, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "eval_cer": 0.5222273278323101, | |
| "eval_loss": 1.6703613996505737, | |
| "eval_runtime": 374.7053, | |
| "eval_samples_per_second": 25.295, | |
| "eval_steps_per_second": 3.162, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 3.7831201553344727, | |
| "learning_rate": 4.654275618374558e-05, | |
| "loss": 1.6107, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "eval_cer": 0.5470601462594346, | |
| "eval_loss": 2.0870959758758545, | |
| "eval_runtime": 367.5169, | |
| "eval_samples_per_second": 25.789, | |
| "eval_steps_per_second": 3.224, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 12.008744239807129, | |
| "learning_rate": 4.6472084805653715e-05, | |
| "loss": 2.3883, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "eval_cer": 0.5516771929138478, | |
| "eval_loss": 3.582747459411621, | |
| "eval_runtime": 376.938, | |
| "eval_samples_per_second": 25.145, | |
| "eval_steps_per_second": 3.144, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.4807026386260986, | |
| "learning_rate": 4.6401413427561844e-05, | |
| "loss": 2.1991, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "eval_cer": 0.5373152203668218, | |
| "eval_loss": 3.5865066051483154, | |
| "eval_runtime": 360.1614, | |
| "eval_samples_per_second": 26.316, | |
| "eval_steps_per_second": 3.29, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 6.418155670166016, | |
| "learning_rate": 4.6330742049469965e-05, | |
| "loss": 2.3905, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_cer": 0.527748719252278, | |
| "eval_loss": 2.6406588554382324, | |
| "eval_runtime": 379.816, | |
| "eval_samples_per_second": 24.954, | |
| "eval_steps_per_second": 3.12, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 2.4806301593780518, | |
| "learning_rate": 4.626007067137809e-05, | |
| "loss": 3.3218, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "eval_cer": 0.5217727112744907, | |
| "eval_loss": 1.8109639883041382, | |
| "eval_runtime": 351.8213, | |
| "eval_samples_per_second": 26.94, | |
| "eval_steps_per_second": 3.368, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.4776511192321777, | |
| "learning_rate": 4.618939929328622e-05, | |
| "loss": 2.3876, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "eval_cer": 0.5089310156036134, | |
| "eval_loss": 3.134875774383545, | |
| "eval_runtime": 372.2436, | |
| "eval_samples_per_second": 25.462, | |
| "eval_steps_per_second": 3.183, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 9.97956657409668, | |
| "learning_rate": 4.611872791519435e-05, | |
| "loss": 4.1527, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_cer": 0.4981815337687224, | |
| "eval_loss": 1.7425427436828613, | |
| "eval_runtime": 363.658, | |
| "eval_samples_per_second": 26.063, | |
| "eval_steps_per_second": 3.259, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 3.353976249694824, | |
| "learning_rate": 4.604805653710248e-05, | |
| "loss": 1.7398, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "eval_cer": 0.5255465175394001, | |
| "eval_loss": 2.6437551975250244, | |
| "eval_runtime": 380.4559, | |
| "eval_samples_per_second": 24.912, | |
| "eval_steps_per_second": 3.115, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 2.671917676925659, | |
| "learning_rate": 4.5978091872791526e-05, | |
| "loss": 2.2359, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "eval_cer": 0.537452094169176, | |
| "eval_loss": 3.4509007930755615, | |
| "eval_runtime": 364.3511, | |
| "eval_samples_per_second": 26.013, | |
| "eval_steps_per_second": 3.252, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 10.607972145080566, | |
| "learning_rate": 4.590742049469965e-05, | |
| "loss": 3.6419, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "eval_cer": 0.5374105431934614, | |
| "eval_loss": 2.2622523307800293, | |
| "eval_runtime": 370.8325, | |
| "eval_samples_per_second": 25.559, | |
| "eval_steps_per_second": 3.196, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 4.678300857543945, | |
| "learning_rate": 4.5836749116607775e-05, | |
| "loss": 1.7309, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_cer": 0.5418393883696375, | |
| "eval_loss": 1.8007577657699585, | |
| "eval_runtime": 357.8741, | |
| "eval_samples_per_second": 26.484, | |
| "eval_steps_per_second": 3.311, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 3.6259689331054688, | |
| "learning_rate": 4.57660777385159e-05, | |
| "loss": 2.784, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_cer": 0.5228799225685347, | |
| "eval_loss": 2.9312586784362793, | |
| "eval_runtime": 371.6433, | |
| "eval_samples_per_second": 25.503, | |
| "eval_steps_per_second": 3.189, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 1.8296568393707275, | |
| "learning_rate": 4.569540636042403e-05, | |
| "loss": 2.4269, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_cer": 0.5211445582886864, | |
| "eval_loss": 3.522822618484497, | |
| "eval_runtime": 359.0335, | |
| "eval_samples_per_second": 26.399, | |
| "eval_steps_per_second": 3.301, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 23.777395248413086, | |
| "learning_rate": 4.562473498233216e-05, | |
| "loss": 1.5294, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "eval_cer": 0.5097864768683275, | |
| "eval_loss": 2.758403778076172, | |
| "eval_runtime": 375.8984, | |
| "eval_samples_per_second": 25.214, | |
| "eval_steps_per_second": 3.152, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 5.520616054534912, | |
| "learning_rate": 4.555406360424028e-05, | |
| "loss": 1.6177, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_cer": 0.5417905048687967, | |
| "eval_loss": 2.8678946495056152, | |
| "eval_runtime": 360.269, | |
| "eval_samples_per_second": 26.308, | |
| "eval_steps_per_second": 3.289, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 12.704540252685547, | |
| "learning_rate": 4.5483392226148416e-05, | |
| "loss": 2.6892, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "eval_cer": 0.5306426225020531, | |
| "eval_loss": 2.7854502201080322, | |
| "eval_runtime": 372.5558, | |
| "eval_samples_per_second": 25.44, | |
| "eval_steps_per_second": 3.181, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 3.1302738189697266, | |
| "learning_rate": 4.541272084805654e-05, | |
| "loss": 3.1467, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "eval_cer": 0.5148948027061906, | |
| "eval_loss": 2.9890220165252686, | |
| "eval_runtime": 359.2287, | |
| "eval_samples_per_second": 26.384, | |
| "eval_steps_per_second": 3.299, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 1.750848412513733, | |
| "learning_rate": 4.5342049469964665e-05, | |
| "loss": 2.1825, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_cer": 0.5565606546478432, | |
| "eval_loss": 3.195390224456787, | |
| "eval_runtime": 366.6828, | |
| "eval_samples_per_second": 25.848, | |
| "eval_steps_per_second": 3.232, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 237.01815795898438, | |
| "learning_rate": 4.527137809187279e-05, | |
| "loss": 3.0545, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "eval_cer": 0.5361493488717688, | |
| "eval_loss": 2.276334524154663, | |
| "eval_runtime": 373.4974, | |
| "eval_samples_per_second": 25.376, | |
| "eval_steps_per_second": 3.173, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 2.9628891944885254, | |
| "learning_rate": 4.520070671378092e-05, | |
| "loss": 1.9898, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "eval_cer": 0.5218631457510461, | |
| "eval_loss": 1.94766104221344, | |
| "eval_runtime": 364.5953, | |
| "eval_samples_per_second": 25.996, | |
| "eval_steps_per_second": 3.25, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 19.71111488342285, | |
| "learning_rate": 4.513003533568905e-05, | |
| "loss": 1.7744, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "eval_cer": 0.5048052481326503, | |
| "eval_loss": 2.699909210205078, | |
| "eval_runtime": 377.4252, | |
| "eval_samples_per_second": 25.112, | |
| "eval_steps_per_second": 3.14, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 3.2357943058013916, | |
| "learning_rate": 4.505936395759717e-05, | |
| "loss": 1.8872, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "eval_cer": 0.5123968558132259, | |
| "eval_loss": 2.0029726028442383, | |
| "eval_runtime": 372.4224, | |
| "eval_samples_per_second": 25.45, | |
| "eval_steps_per_second": 3.182, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 2.642812967300415, | |
| "learning_rate": 4.4988692579505305e-05, | |
| "loss": 3.1774, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_cer": 0.5154300770403973, | |
| "eval_loss": 2.36077618598938, | |
| "eval_runtime": 375.4381, | |
| "eval_samples_per_second": 25.245, | |
| "eval_steps_per_second": 3.156, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 4.245635509490967, | |
| "learning_rate": 4.491802120141343e-05, | |
| "loss": 1.7844, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "eval_cer": 0.5169381330413358, | |
| "eval_loss": 2.6845099925994873, | |
| "eval_runtime": 355.3776, | |
| "eval_samples_per_second": 26.67, | |
| "eval_steps_per_second": 3.334, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 2.799055576324463, | |
| "learning_rate": 4.4847349823321555e-05, | |
| "loss": 3.4812, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "eval_cer": 0.49670036369324627, | |
| "eval_loss": 3.4256324768066406, | |
| "eval_runtime": 374.5672, | |
| "eval_samples_per_second": 25.304, | |
| "eval_steps_per_second": 3.164, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 2.1845884323120117, | |
| "learning_rate": 4.477667844522968e-05, | |
| "loss": 3.0965, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_cer": 0.5227894880919792, | |
| "eval_loss": 1.759318232536316, | |
| "eval_runtime": 366.7135, | |
| "eval_samples_per_second": 25.846, | |
| "eval_steps_per_second": 3.231, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 4.01229190826416, | |
| "learning_rate": 4.470600706713781e-05, | |
| "loss": 2.008, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "eval_cer": 0.5001319854522701, | |
| "eval_loss": 1.97179114818573, | |
| "eval_runtime": 377.7336, | |
| "eval_samples_per_second": 25.092, | |
| "eval_steps_per_second": 3.137, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 1.3265995979309082, | |
| "learning_rate": 4.463533568904594e-05, | |
| "loss": 1.7854, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "eval_cer": 0.48717785772945915, | |
| "eval_loss": 1.6064122915267944, | |
| "eval_runtime": 358.3216, | |
| "eval_samples_per_second": 26.451, | |
| "eval_steps_per_second": 3.307, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 3.6079511642456055, | |
| "learning_rate": 4.456466431095407e-05, | |
| "loss": 1.7713, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "eval_cer": 0.495703140276094, | |
| "eval_loss": 2.7158584594726562, | |
| "eval_runtime": 373.6599, | |
| "eval_samples_per_second": 25.365, | |
| "eval_steps_per_second": 3.171, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 3.969160318374634, | |
| "learning_rate": 4.4493992932862195e-05, | |
| "loss": 3.3072, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "eval_cer": 0.49535606742012434, | |
| "eval_loss": 1.9461767673492432, | |
| "eval_runtime": 357.0354, | |
| "eval_samples_per_second": 26.546, | |
| "eval_steps_per_second": 3.319, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 1.3938627243041992, | |
| "learning_rate": 4.442332155477032e-05, | |
| "loss": 2.1757, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "eval_cer": 0.5164395213327597, | |
| "eval_loss": 1.6015843152999878, | |
| "eval_runtime": 375.0822, | |
| "eval_samples_per_second": 25.269, | |
| "eval_steps_per_second": 3.159, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 7.331507682800293, | |
| "learning_rate": 4.4352650176678445e-05, | |
| "loss": 2.3463, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "eval_cer": 0.4924132806695084, | |
| "eval_loss": 3.185905933380127, | |
| "eval_runtime": 357.5938, | |
| "eval_samples_per_second": 26.505, | |
| "eval_steps_per_second": 3.314, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 2.3522162437438965, | |
| "learning_rate": 4.428197879858658e-05, | |
| "loss": 1.9984, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "eval_cer": 0.5185977278948809, | |
| "eval_loss": 2.115999221801758, | |
| "eval_runtime": 377.0919, | |
| "eval_samples_per_second": 25.134, | |
| "eval_steps_per_second": 3.142, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 5.214267730712891, | |
| "learning_rate": 4.42113074204947e-05, | |
| "loss": 1.5957, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "eval_cer": 0.5082393140667162, | |
| "eval_loss": 1.6540412902832031, | |
| "eval_runtime": 357.3421, | |
| "eval_samples_per_second": 26.524, | |
| "eval_steps_per_second": 3.316, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 14.154474258422852, | |
| "learning_rate": 4.414063604240283e-05, | |
| "loss": 1.6646, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "eval_cer": 0.49412175902389427, | |
| "eval_loss": 2.7898120880126953, | |
| "eval_runtime": 372.8281, | |
| "eval_samples_per_second": 25.422, | |
| "eval_steps_per_second": 3.178, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 6.580379486083984, | |
| "learning_rate": 4.406996466431096e-05, | |
| "loss": 1.9858, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "eval_cer": 0.5368801572093387, | |
| "eval_loss": 3.0827128887176514, | |
| "eval_runtime": 359.3692, | |
| "eval_samples_per_second": 26.374, | |
| "eval_steps_per_second": 3.297, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 4.16035270690918, | |
| "learning_rate": 4.3999293286219085e-05, | |
| "loss": 2.312, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "eval_cer": 0.503619823237261, | |
| "eval_loss": 1.6672168970108032, | |
| "eval_runtime": 375.3507, | |
| "eval_samples_per_second": 25.251, | |
| "eval_steps_per_second": 3.157, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 1.9299957752227783, | |
| "learning_rate": 4.392862190812721e-05, | |
| "loss": 2.0043, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "eval_cer": 0.4974531696061945, | |
| "eval_loss": 3.044623613357544, | |
| "eval_runtime": 365.1986, | |
| "eval_samples_per_second": 25.953, | |
| "eval_steps_per_second": 3.245, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 1.68573796749115, | |
| "learning_rate": 4.3857950530035335e-05, | |
| "loss": 2.5474, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "eval_cer": 0.5048394665832389, | |
| "eval_loss": 2.7017831802368164, | |
| "eval_runtime": 374.6095, | |
| "eval_samples_per_second": 25.301, | |
| "eval_steps_per_second": 3.163, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 4.48902702331543, | |
| "learning_rate": 4.378727915194347e-05, | |
| "loss": 2.2683, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "eval_cer": 0.49138917132689375, | |
| "eval_loss": 1.5836848020553589, | |
| "eval_runtime": 360.7206, | |
| "eval_samples_per_second": 26.275, | |
| "eval_steps_per_second": 3.285, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.788163661956787, | |
| "learning_rate": 4.371731448763251e-05, | |
| "loss": 2.4278, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "eval_cer": 0.5205579562785968, | |
| "eval_loss": 2.4102938175201416, | |
| "eval_runtime": 376.1786, | |
| "eval_samples_per_second": 25.195, | |
| "eval_steps_per_second": 3.15, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.6063647270202637, | |
| "learning_rate": 4.364664310954063e-05, | |
| "loss": 1.5116, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.4881310859958547, | |
| "eval_loss": 1.728480339050293, | |
| "eval_runtime": 352.8942, | |
| "eval_samples_per_second": 26.858, | |
| "eval_steps_per_second": 3.358, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 5.240590572357178, | |
| "learning_rate": 4.357597173144877e-05, | |
| "loss": 2.2426, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "eval_cer": 0.49515320089163506, | |
| "eval_loss": 2.496321201324463, | |
| "eval_runtime": 373.8495, | |
| "eval_samples_per_second": 25.352, | |
| "eval_steps_per_second": 3.17, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 2.557533025741577, | |
| "learning_rate": 4.3505300353356896e-05, | |
| "loss": 1.8649, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "eval_cer": 0.5054896171444214, | |
| "eval_loss": 3.00071120262146, | |
| "eval_runtime": 362.1744, | |
| "eval_samples_per_second": 26.17, | |
| "eval_steps_per_second": 3.272, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 27.368144989013672, | |
| "learning_rate": 4.343462897526502e-05, | |
| "loss": 1.6976, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "eval_cer": 0.4800359782566188, | |
| "eval_loss": 2.5932295322418213, | |
| "eval_runtime": 371.4743, | |
| "eval_samples_per_second": 25.515, | |
| "eval_steps_per_second": 3.19, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 4.671775817871094, | |
| "learning_rate": 4.336395759717315e-05, | |
| "loss": 1.7712, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_cer": 0.4871509718039967, | |
| "eval_loss": 2.37919282913208, | |
| "eval_runtime": 357.2849, | |
| "eval_samples_per_second": 26.528, | |
| "eval_steps_per_second": 3.317, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 2.77481746673584, | |
| "learning_rate": 4.329328621908127e-05, | |
| "loss": 1.45, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "eval_cer": 0.49421952602557584, | |
| "eval_loss": 1.6775341033935547, | |
| "eval_runtime": 376.5454, | |
| "eval_samples_per_second": 25.171, | |
| "eval_steps_per_second": 3.147, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 7.3151140213012695, | |
| "learning_rate": 4.32226148409894e-05, | |
| "loss": 1.5163, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "eval_cer": 0.49458370810683977, | |
| "eval_loss": 2.463066577911377, | |
| "eval_runtime": 349.9183, | |
| "eval_samples_per_second": 27.086, | |
| "eval_steps_per_second": 3.387, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 3.947378635406494, | |
| "learning_rate": 4.315194346289753e-05, | |
| "loss": 1.4527, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "eval_cer": 0.49508965234054203, | |
| "eval_loss": 3.2406046390533447, | |
| "eval_runtime": 377.1597, | |
| "eval_samples_per_second": 25.13, | |
| "eval_steps_per_second": 3.142, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 7.427024841308594, | |
| "learning_rate": 4.308127208480566e-05, | |
| "loss": 1.4446, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "eval_cer": 0.48428639865472606, | |
| "eval_loss": 2.7247695922851562, | |
| "eval_runtime": 356.4628, | |
| "eval_samples_per_second": 26.589, | |
| "eval_steps_per_second": 3.324, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 3.1798312664031982, | |
| "learning_rate": 4.3010600706713785e-05, | |
| "loss": 2.4877, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "eval_cer": 0.48304475773336986, | |
| "eval_loss": 2.2158772945404053, | |
| "eval_runtime": 376.4543, | |
| "eval_samples_per_second": 25.177, | |
| "eval_steps_per_second": 3.148, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 2.0509414672851562, | |
| "learning_rate": 4.293992932862191e-05, | |
| "loss": 1.9214, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "eval_cer": 0.4835262602166517, | |
| "eval_loss": 1.906830906867981, | |
| "eval_runtime": 357.384, | |
| "eval_samples_per_second": 26.52, | |
| "eval_steps_per_second": 3.316, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 3.71893048286438, | |
| "learning_rate": 4.286925795053004e-05, | |
| "loss": 1.4426, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "eval_cer": 0.48283211450471236, | |
| "eval_loss": 2.5203747749328613, | |
| "eval_runtime": 378.5225, | |
| "eval_samples_per_second": 25.039, | |
| "eval_steps_per_second": 3.131, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 3.5400538444519043, | |
| "learning_rate": 4.279858657243816e-05, | |
| "loss": 2.5204, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_cer": 0.496866567596105, | |
| "eval_loss": 3.1510612964630127, | |
| "eval_runtime": 360.9029, | |
| "eval_samples_per_second": 26.262, | |
| "eval_steps_per_second": 3.283, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 3.434525966644287, | |
| "learning_rate": 4.272791519434629e-05, | |
| "loss": 1.605, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "eval_cer": 0.4771103007312972, | |
| "eval_loss": 3.0746238231658936, | |
| "eval_runtime": 370.845, | |
| "eval_samples_per_second": 25.558, | |
| "eval_steps_per_second": 3.195, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 6.7493815422058105, | |
| "learning_rate": 4.265724381625442e-05, | |
| "loss": 2.153, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "eval_cer": 0.4859728794337335, | |
| "eval_loss": 2.0545401573181152, | |
| "eval_runtime": 361.9049, | |
| "eval_samples_per_second": 26.189, | |
| "eval_steps_per_second": 3.274, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 1.7070709466934204, | |
| "learning_rate": 4.258657243816255e-05, | |
| "loss": 1.8732, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "eval_cer": 0.4698926518321536, | |
| "eval_loss": 2.4386019706726074, | |
| "eval_runtime": 373.6335, | |
| "eval_samples_per_second": 25.367, | |
| "eval_steps_per_second": 3.172, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 11.575754165649414, | |
| "learning_rate": 4.2515901060070675e-05, | |
| "loss": 1.9211, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_cer": 0.5030527746275077, | |
| "eval_loss": 2.1613056659698486, | |
| "eval_runtime": 353.1639, | |
| "eval_samples_per_second": 26.837, | |
| "eval_steps_per_second": 3.355, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 5.623499393463135, | |
| "learning_rate": 4.24452296819788e-05, | |
| "loss": 2.6473, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "eval_cer": 0.5010998787689179, | |
| "eval_loss": 2.048527240753174, | |
| "eval_runtime": 372.3781, | |
| "eval_samples_per_second": 25.453, | |
| "eval_steps_per_second": 3.182, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 4.054663181304932, | |
| "learning_rate": 4.237455830388693e-05, | |
| "loss": 1.4888, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "eval_cer": 0.48732939658206564, | |
| "eval_loss": 2.4912757873535156, | |
| "eval_runtime": 353.9893, | |
| "eval_samples_per_second": 26.775, | |
| "eval_steps_per_second": 3.348, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.9226583242416382, | |
| "learning_rate": 4.230388692579505e-05, | |
| "loss": 1.4556, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_cer": 0.4807105705682218, | |
| "eval_loss": 1.4841840267181396, | |
| "eval_runtime": 376.2491, | |
| "eval_samples_per_second": 25.191, | |
| "eval_steps_per_second": 3.15, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 6.85697603225708, | |
| "learning_rate": 4.223321554770318e-05, | |
| "loss": 1.4391, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "eval_cer": 0.48008730593250165, | |
| "eval_loss": 3.111499547958374, | |
| "eval_runtime": 358.9952, | |
| "eval_samples_per_second": 26.401, | |
| "eval_steps_per_second": 3.301, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 12.522397994995117, | |
| "learning_rate": 4.216254416961131e-05, | |
| "loss": 1.4244, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "eval_cer": 0.48068612881780143, | |
| "eval_loss": 2.598745822906494, | |
| "eval_runtime": 385.9814, | |
| "eval_samples_per_second": 24.556, | |
| "eval_steps_per_second": 3.07, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 3.1015026569366455, | |
| "learning_rate": 4.209187279151944e-05, | |
| "loss": 3.7378, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "eval_cer": 0.47458791208791207, | |
| "eval_loss": 2.3908824920654297, | |
| "eval_runtime": 373.2148, | |
| "eval_samples_per_second": 25.396, | |
| "eval_steps_per_second": 3.175, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 86.87032318115234, | |
| "learning_rate": 4.2021201413427565e-05, | |
| "loss": 2.8329, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "eval_cer": 0.4754898126784248, | |
| "eval_loss": 2.441450357437134, | |
| "eval_runtime": 446.0173, | |
| "eval_samples_per_second": 21.25, | |
| "eval_steps_per_second": 2.657, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 2.7503468990325928, | |
| "learning_rate": 4.195053003533569e-05, | |
| "loss": 2.4912, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "eval_cer": 0.488764127331743, | |
| "eval_loss": 1.6247801780700684, | |
| "eval_runtime": 361.3079, | |
| "eval_samples_per_second": 26.232, | |
| "eval_steps_per_second": 3.28, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 2.511701822280884, | |
| "learning_rate": 4.187985865724382e-05, | |
| "loss": 2.009, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "eval_cer": 0.46521938915177347, | |
| "eval_loss": 1.8090691566467285, | |
| "eval_runtime": 401.8599, | |
| "eval_samples_per_second": 23.585, | |
| "eval_steps_per_second": 2.949, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "grad_norm": 4.231322765350342, | |
| "learning_rate": 4.180918727915194e-05, | |
| "loss": 1.6484, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "eval_cer": 0.483240291736733, | |
| "eval_loss": 1.89494788646698, | |
| "eval_runtime": 367.7673, | |
| "eval_samples_per_second": 25.772, | |
| "eval_steps_per_second": 3.222, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 3.436452865600586, | |
| "learning_rate": 4.173851590106007e-05, | |
| "loss": 1.4205, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "eval_cer": 0.48146582065621213, | |
| "eval_loss": 2.3205745220184326, | |
| "eval_runtime": 384.4394, | |
| "eval_samples_per_second": 24.654, | |
| "eval_steps_per_second": 3.082, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 1.96918523311615, | |
| "learning_rate": 4.16678445229682e-05, | |
| "loss": 1.3964, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "eval_cer": 0.4692033944702984, | |
| "eval_loss": 2.9126245975494385, | |
| "eval_runtime": 358.8304, | |
| "eval_samples_per_second": 26.414, | |
| "eval_steps_per_second": 3.302, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 6.869575500488281, | |
| "learning_rate": 4.159717314487633e-05, | |
| "loss": 2.0721, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "eval_cer": 0.4858066755308748, | |
| "eval_loss": 3.2426090240478516, | |
| "eval_runtime": 377.7703, | |
| "eval_samples_per_second": 25.089, | |
| "eval_steps_per_second": 3.137, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 2.511401414871216, | |
| "learning_rate": 4.1526501766784455e-05, | |
| "loss": 1.747, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "eval_cer": 0.5091607680575652, | |
| "eval_loss": 3.057870864868164, | |
| "eval_runtime": 381.9144, | |
| "eval_samples_per_second": 24.817, | |
| "eval_steps_per_second": 3.103, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 9.076600074768066, | |
| "learning_rate": 4.1455830388692577e-05, | |
| "loss": 2.0271, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "eval_cer": 0.48266591060185365, | |
| "eval_loss": 2.858898639678955, | |
| "eval_runtime": 405.4609, | |
| "eval_samples_per_second": 23.376, | |
| "eval_steps_per_second": 2.923, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 1.6710706949234009, | |
| "learning_rate": 4.138515901060071e-05, | |
| "loss": 1.7331, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "eval_cer": 0.501048551093035, | |
| "eval_loss": 2.9208521842956543, | |
| "eval_runtime": 361.6906, | |
| "eval_samples_per_second": 26.205, | |
| "eval_steps_per_second": 3.276, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "grad_norm": 1.902030348777771, | |
| "learning_rate": 4.131448763250883e-05, | |
| "loss": 2.1405, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "eval_cer": 0.4769636502287748, | |
| "eval_loss": 2.750805139541626, | |
| "eval_runtime": 383.778, | |
| "eval_samples_per_second": 24.697, | |
| "eval_steps_per_second": 3.088, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 5.2548041343688965, | |
| "learning_rate": 4.124381625441696e-05, | |
| "loss": 1.967, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "eval_cer": 0.4813216143287318, | |
| "eval_loss": 2.39349627494812, | |
| "eval_runtime": 366.8876, | |
| "eval_samples_per_second": 25.834, | |
| "eval_steps_per_second": 3.23, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "grad_norm": 3.4912497997283936, | |
| "learning_rate": 4.1173144876325096e-05, | |
| "loss": 2.5075, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "eval_cer": 0.48205731101638577, | |
| "eval_loss": 2.9853949546813965, | |
| "eval_runtime": 382.1207, | |
| "eval_samples_per_second": 24.804, | |
| "eval_steps_per_second": 3.101, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "grad_norm": 3.671482801437378, | |
| "learning_rate": 4.110247349823322e-05, | |
| "loss": 2.3712, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "eval_cer": 0.47388154550076256, | |
| "eval_loss": 2.6945626735687256, | |
| "eval_runtime": 374.115, | |
| "eval_samples_per_second": 25.334, | |
| "eval_steps_per_second": 3.167, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "grad_norm": 3.400202512741089, | |
| "learning_rate": 4.1031802120141345e-05, | |
| "loss": 2.035, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "eval_cer": 0.4721559579210825, | |
| "eval_loss": 2.723193645477295, | |
| "eval_runtime": 442.6463, | |
| "eval_samples_per_second": 21.412, | |
| "eval_steps_per_second": 2.677, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 28.22166633605957, | |
| "learning_rate": 4.0961130742049467e-05, | |
| "loss": 2.583, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "eval_cer": 0.4679446443236479, | |
| "eval_loss": 2.848484516143799, | |
| "eval_runtime": 385.7533, | |
| "eval_samples_per_second": 24.57, | |
| "eval_steps_per_second": 3.072, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "grad_norm": 2.6984710693359375, | |
| "learning_rate": 4.08904593639576e-05, | |
| "loss": 1.4041, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "eval_cer": 0.46995620038324665, | |
| "eval_loss": 2.8926751613616943, | |
| "eval_runtime": 379.5665, | |
| "eval_samples_per_second": 24.971, | |
| "eval_steps_per_second": 3.122, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 28.124893188476562, | |
| "learning_rate": 4.081978798586573e-05, | |
| "loss": 1.3813, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "eval_cer": 0.453462907199562, | |
| "eval_loss": 3.064070701599121, | |
| "eval_runtime": 364.1898, | |
| "eval_samples_per_second": 26.025, | |
| "eval_steps_per_second": 3.254, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 4.778401851654053, | |
| "learning_rate": 4.074911660777385e-05, | |
| "loss": 1.9508, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "eval_cer": 0.4695724649016464, | |
| "eval_loss": 1.914655327796936, | |
| "eval_runtime": 378.9026, | |
| "eval_samples_per_second": 25.014, | |
| "eval_steps_per_second": 3.127, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "grad_norm": 4.1898345947265625, | |
| "learning_rate": 4.0678445229681986e-05, | |
| "loss": 2.1823, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "eval_cer": 0.47729850220953424, | |
| "eval_loss": 1.71451997756958, | |
| "eval_runtime": 365.8718, | |
| "eval_samples_per_second": 25.905, | |
| "eval_steps_per_second": 3.239, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 1.8314547538757324, | |
| "learning_rate": 4.060777385159011e-05, | |
| "loss": 2.3976, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "eval_cer": 0.4735149192444566, | |
| "eval_loss": 2.7786667346954346, | |
| "eval_runtime": 377.8869, | |
| "eval_samples_per_second": 25.082, | |
| "eval_steps_per_second": 3.136, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "grad_norm": 3.4417295455932617, | |
| "learning_rate": 4.0537102473498235e-05, | |
| "loss": 2.0057, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "eval_cer": 0.47049147471745334, | |
| "eval_loss": 3.252101182937622, | |
| "eval_runtime": 365.7598, | |
| "eval_samples_per_second": 25.913, | |
| "eval_steps_per_second": 3.24, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 4.99106502532959, | |
| "learning_rate": 4.0466431095406356e-05, | |
| "loss": 1.7663, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "eval_cer": 0.4712345039302335, | |
| "eval_loss": 2.6297221183776855, | |
| "eval_runtime": 401.5004, | |
| "eval_samples_per_second": 23.606, | |
| "eval_steps_per_second": 2.951, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 4.448285102844238, | |
| "learning_rate": 4.039575971731449e-05, | |
| "loss": 1.842, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "eval_cer": 0.46505562942395684, | |
| "eval_loss": 2.5416791439056396, | |
| "eval_runtime": 383.7378, | |
| "eval_samples_per_second": 24.699, | |
| "eval_steps_per_second": 3.088, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 2.196765184402466, | |
| "learning_rate": 4.032579505300353e-05, | |
| "loss": 2.0249, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_cer": 0.4723050525986469, | |
| "eval_loss": 1.9896740913391113, | |
| "eval_runtime": 382.0154, | |
| "eval_samples_per_second": 24.811, | |
| "eval_steps_per_second": 3.102, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 2.4070873260498047, | |
| "learning_rate": 4.025512367491166e-05, | |
| "loss": 2.6354, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "eval_cer": 0.46424660748504165, | |
| "eval_loss": 1.524404525756836, | |
| "eval_runtime": 361.2522, | |
| "eval_samples_per_second": 26.237, | |
| "eval_steps_per_second": 3.28, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "grad_norm": 2.6403534412384033, | |
| "learning_rate": 4.018445229681979e-05, | |
| "loss": 1.4007, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "eval_cer": 0.4687561104376051, | |
| "eval_loss": 2.6799304485321045, | |
| "eval_runtime": 377.8259, | |
| "eval_samples_per_second": 25.086, | |
| "eval_steps_per_second": 3.136, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 16.431684494018555, | |
| "learning_rate": 4.011378091872792e-05, | |
| "loss": 1.3673, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "eval_cer": 0.4627141097336827, | |
| "eval_loss": 2.739257335662842, | |
| "eval_runtime": 365.917, | |
| "eval_samples_per_second": 25.902, | |
| "eval_steps_per_second": 3.238, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 1.2545260190963745, | |
| "learning_rate": 4.0043109540636045e-05, | |
| "loss": 1.398, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "eval_cer": 0.46276054905948144, | |
| "eval_loss": 2.623035430908203, | |
| "eval_runtime": 380.5058, | |
| "eval_samples_per_second": 24.909, | |
| "eval_steps_per_second": 3.114, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "grad_norm": 4.757970333099365, | |
| "learning_rate": 3.9972438162544173e-05, | |
| "loss": 1.7177, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "eval_cer": 0.4671600641351531, | |
| "eval_loss": 3.1201841831207275, | |
| "eval_runtime": 364.6269, | |
| "eval_samples_per_second": 25.994, | |
| "eval_steps_per_second": 3.25, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 2.604834794998169, | |
| "learning_rate": 3.99017667844523e-05, | |
| "loss": 1.4781, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "eval_cer": 0.4741137421297564, | |
| "eval_loss": 2.4959936141967773, | |
| "eval_runtime": 380.3259, | |
| "eval_samples_per_second": 24.921, | |
| "eval_steps_per_second": 3.116, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "grad_norm": 5.067722797393799, | |
| "learning_rate": 3.983109540636042e-05, | |
| "loss": 1.6941, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "eval_cer": 0.4551322787532752, | |
| "eval_loss": 2.24949049949646, | |
| "eval_runtime": 359.186, | |
| "eval_samples_per_second": 26.387, | |
| "eval_steps_per_second": 3.299, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "grad_norm": 4.323084354400635, | |
| "learning_rate": 3.976042402826856e-05, | |
| "loss": 1.4552, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "eval_cer": 0.459111395721716, | |
| "eval_loss": 2.0859220027923584, | |
| "eval_runtime": 380.4099, | |
| "eval_samples_per_second": 24.915, | |
| "eval_steps_per_second": 3.115, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "grad_norm": 3.9104325771331787, | |
| "learning_rate": 3.968975265017668e-05, | |
| "loss": 2.836, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "eval_cer": 0.47354180516991906, | |
| "eval_loss": 1.5567090511322021, | |
| "eval_runtime": 368.3958, | |
| "eval_samples_per_second": 25.728, | |
| "eval_steps_per_second": 3.217, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 22.448034286499023, | |
| "learning_rate": 3.961908127208481e-05, | |
| "loss": 1.3858, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_cer": 0.45766688827187046, | |
| "eval_loss": 1.999880313873291, | |
| "eval_runtime": 380.7492, | |
| "eval_samples_per_second": 24.893, | |
| "eval_steps_per_second": 3.112, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "grad_norm": 10.558205604553223, | |
| "learning_rate": 3.9548409893992935e-05, | |
| "loss": 1.3921, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "eval_cer": 0.47037904266551955, | |
| "eval_loss": 2.6411163806915283, | |
| "eval_runtime": 363.6122, | |
| "eval_samples_per_second": 26.066, | |
| "eval_steps_per_second": 3.259, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "grad_norm": 32.08698654174805, | |
| "learning_rate": 3.947773851590106e-05, | |
| "loss": 2.0638, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "eval_cer": 0.4723588244495718, | |
| "eval_loss": 2.9272561073303223, | |
| "eval_runtime": 386.8412, | |
| "eval_samples_per_second": 24.501, | |
| "eval_steps_per_second": 3.063, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "grad_norm": 1.8787578344345093, | |
| "learning_rate": 3.940706713780919e-05, | |
| "loss": 1.3608, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "eval_cer": 0.4626187869070431, | |
| "eval_loss": 3.0016987323760986, | |
| "eval_runtime": 353.2265, | |
| "eval_samples_per_second": 26.833, | |
| "eval_steps_per_second": 3.355, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "grad_norm": 5.287090301513672, | |
| "learning_rate": 3.933639575971731e-05, | |
| "loss": 1.6601, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "eval_cer": 0.4719262054671307, | |
| "eval_loss": 2.0440595149993896, | |
| "eval_runtime": 377.8824, | |
| "eval_samples_per_second": 25.082, | |
| "eval_steps_per_second": 3.136, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "grad_norm": 2.0426106452941895, | |
| "learning_rate": 3.926572438162545e-05, | |
| "loss": 1.9773, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "eval_cer": 0.4639239763794924, | |
| "eval_loss": 2.3498713970184326, | |
| "eval_runtime": 365.2548, | |
| "eval_samples_per_second": 25.949, | |
| "eval_steps_per_second": 3.244, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "grad_norm": 2.1095080375671387, | |
| "learning_rate": 3.919505300353357e-05, | |
| "loss": 2.2505, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "eval_cer": 0.4647232216182394, | |
| "eval_loss": 1.8471035957336426, | |
| "eval_runtime": 388.913, | |
| "eval_samples_per_second": 24.37, | |
| "eval_steps_per_second": 3.047, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "grad_norm": 3.6614439487457275, | |
| "learning_rate": 3.91243816254417e-05, | |
| "loss": 1.9341, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "eval_cer": 0.4613698134605608, | |
| "eval_loss": 2.8519198894500732, | |
| "eval_runtime": 363.022, | |
| "eval_samples_per_second": 26.109, | |
| "eval_steps_per_second": 3.264, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 6.74, | |
| "grad_norm": 8.302891731262207, | |
| "learning_rate": 3.9053710247349825e-05, | |
| "loss": 1.4603, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.74, | |
| "eval_cer": 0.47225128074772205, | |
| "eval_loss": 1.5901261568069458, | |
| "eval_runtime": 376.5314, | |
| "eval_samples_per_second": 25.172, | |
| "eval_steps_per_second": 3.147, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "grad_norm": 3.2967019081115723, | |
| "learning_rate": 3.898303886925795e-05, | |
| "loss": 1.5493, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "eval_cer": 0.47416751398068124, | |
| "eval_loss": 1.6550699472427368, | |
| "eval_runtime": 363.1067, | |
| "eval_samples_per_second": 26.103, | |
| "eval_steps_per_second": 3.264, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "grad_norm": 2.848189115524292, | |
| "learning_rate": 3.891236749116608e-05, | |
| "loss": 1.5642, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "eval_cer": 0.4807154589183059, | |
| "eval_loss": 2.1456046104431152, | |
| "eval_runtime": 376.8904, | |
| "eval_samples_per_second": 25.148, | |
| "eval_steps_per_second": 3.144, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "grad_norm": 4.148609638214111, | |
| "learning_rate": 3.88416961130742e-05, | |
| "loss": 1.379, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "eval_cer": 0.4555746744358844, | |
| "eval_loss": 1.7261931896209717, | |
| "eval_runtime": 360.5421, | |
| "eval_samples_per_second": 26.288, | |
| "eval_steps_per_second": 3.287, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 8.180316925048828, | |
| "learning_rate": 3.877102473498234e-05, | |
| "loss": 1.6474, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "eval_cer": 0.4800946384576278, | |
| "eval_loss": 2.309068202972412, | |
| "eval_runtime": 377.3872, | |
| "eval_samples_per_second": 25.115, | |
| "eval_steps_per_second": 3.14, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 6.95, | |
| "grad_norm": 1.7500585317611694, | |
| "learning_rate": 3.870035335689046e-05, | |
| "loss": 1.3755, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 6.95, | |
| "eval_cer": 0.4641219545578976, | |
| "eval_loss": 1.5506832599639893, | |
| "eval_runtime": 359.5594, | |
| "eval_samples_per_second": 26.36, | |
| "eval_steps_per_second": 3.296, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "grad_norm": 1.8204373121261597, | |
| "learning_rate": 3.862968197879859e-05, | |
| "loss": 1.3562, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "eval_cer": 0.46081498572601776, | |
| "eval_loss": 2.0778937339782715, | |
| "eval_runtime": 381.9051, | |
| "eval_samples_per_second": 24.818, | |
| "eval_steps_per_second": 3.103, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 2.9875051975250244, | |
| "learning_rate": 3.8559010600706715e-05, | |
| "loss": 1.5916, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "eval_cer": 0.4623352626021665, | |
| "eval_loss": 1.5354750156402588, | |
| "eval_runtime": 364.3123, | |
| "eval_samples_per_second": 26.016, | |
| "eval_steps_per_second": 3.253, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "grad_norm": 4.643332004547119, | |
| "learning_rate": 3.848833922261484e-05, | |
| "loss": 1.3365, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "eval_cer": 0.46146024793711626, | |
| "eval_loss": 1.8595181703567505, | |
| "eval_runtime": 374.6805, | |
| "eval_samples_per_second": 25.296, | |
| "eval_steps_per_second": 3.163, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "grad_norm": 2.433307409286499, | |
| "learning_rate": 3.841766784452297e-05, | |
| "loss": 1.6175, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "eval_cer": 0.468870986664581, | |
| "eval_loss": 1.7079046964645386, | |
| "eval_runtime": 364.7345, | |
| "eval_samples_per_second": 25.986, | |
| "eval_steps_per_second": 3.249, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "grad_norm": 4.391873359680176, | |
| "learning_rate": 3.834699646643109e-05, | |
| "loss": 1.3334, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "eval_cer": 0.4652682726526143, | |
| "eval_loss": 2.7629072666168213, | |
| "eval_runtime": 380.8752, | |
| "eval_samples_per_second": 24.885, | |
| "eval_steps_per_second": 3.111, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 7.738865375518799, | |
| "learning_rate": 3.827632508833923e-05, | |
| "loss": 1.4858, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "eval_cer": 0.4701737319619882, | |
| "eval_loss": 2.189877986907959, | |
| "eval_runtime": 363.6078, | |
| "eval_samples_per_second": 26.067, | |
| "eval_steps_per_second": 3.259, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 7.24, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.820636042402827e-05, | |
| "loss": 1.5317, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 7.24, | |
| "eval_cer": 0.47163290446208594, | |
| "eval_loss": 1.9506334066390991, | |
| "eval_runtime": 380.9688, | |
| "eval_samples_per_second": 24.879, | |
| "eval_steps_per_second": 3.11, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "grad_norm": 19.556819915771484, | |
| "learning_rate": 3.81356890459364e-05, | |
| "loss": 1.3515, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "eval_cer": 0.45843191506002895, | |
| "eval_loss": 2.2836623191833496, | |
| "eval_runtime": 361.9084, | |
| "eval_samples_per_second": 26.189, | |
| "eval_steps_per_second": 3.274, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "grad_norm": 3.9697859287261963, | |
| "learning_rate": 3.8065017667844525e-05, | |
| "loss": 1.3348, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "eval_cer": 0.460401920143913, | |
| "eval_loss": 2.626167058944702, | |
| "eval_runtime": 377.1422, | |
| "eval_samples_per_second": 25.131, | |
| "eval_steps_per_second": 3.142, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 2.342355251312256, | |
| "learning_rate": 3.7994346289752653e-05, | |
| "loss": 2.0519, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "eval_cer": 0.4643712604121857, | |
| "eval_loss": 1.923600196838379, | |
| "eval_runtime": 360.7805, | |
| "eval_samples_per_second": 26.271, | |
| "eval_steps_per_second": 3.285, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "grad_norm": 1.692319393157959, | |
| "learning_rate": 3.7923674911660775e-05, | |
| "loss": 1.4731, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "eval_cer": 0.4658597630127879, | |
| "eval_loss": 1.808614730834961, | |
| "eval_runtime": 377.9004, | |
| "eval_samples_per_second": 25.081, | |
| "eval_steps_per_second": 3.136, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "grad_norm": 2.243326187133789, | |
| "learning_rate": 3.78530035335689e-05, | |
| "loss": 1.3005, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "eval_cer": 0.45969066520667945, | |
| "eval_loss": 1.8014905452728271, | |
| "eval_runtime": 369.7047, | |
| "eval_samples_per_second": 25.637, | |
| "eval_steps_per_second": 3.205, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "grad_norm": 3.0026488304138184, | |
| "learning_rate": 3.778233215547704e-05, | |
| "loss": 1.3286, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "eval_cer": 0.44898517852254505, | |
| "eval_loss": 1.3857167959213257, | |
| "eval_runtime": 387.1419, | |
| "eval_samples_per_second": 24.482, | |
| "eval_steps_per_second": 3.061, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "grad_norm": 4.501519203186035, | |
| "learning_rate": 3.771166077738516e-05, | |
| "loss": 1.3409, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "eval_cer": 0.4575813421453991, | |
| "eval_loss": 2.416992664337158, | |
| "eval_runtime": 359.7708, | |
| "eval_samples_per_second": 26.345, | |
| "eval_steps_per_second": 3.294, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 10.005617141723633, | |
| "learning_rate": 3.764098939929329e-05, | |
| "loss": 2.1197, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "eval_cer": 0.4567307692307692, | |
| "eval_loss": 1.900498628616333, | |
| "eval_runtime": 386.3412, | |
| "eval_samples_per_second": 24.533, | |
| "eval_steps_per_second": 3.067, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 3.4333086013793945, | |
| "learning_rate": 3.7570318021201415e-05, | |
| "loss": 1.3268, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "eval_cer": 0.4478413046028704, | |
| "eval_loss": 1.8988970518112183, | |
| "eval_runtime": 368.4314, | |
| "eval_samples_per_second": 25.725, | |
| "eval_steps_per_second": 3.216, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "grad_norm": 4.548975944519043, | |
| "learning_rate": 3.749964664310954e-05, | |
| "loss": 1.3146, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "eval_cer": 0.4571047280122013, | |
| "eval_loss": 2.1630055904388428, | |
| "eval_runtime": 388.4273, | |
| "eval_samples_per_second": 24.401, | |
| "eval_steps_per_second": 3.051, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 1.7991045713424683, | |
| "learning_rate": 3.742897526501767e-05, | |
| "loss": 2.0035, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "eval_cer": 0.45147579289038364, | |
| "eval_loss": 2.4393255710601807, | |
| "eval_runtime": 378.3611, | |
| "eval_samples_per_second": 25.05, | |
| "eval_steps_per_second": 3.132, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "grad_norm": 2.6425940990448, | |
| "learning_rate": 3.73583038869258e-05, | |
| "loss": 1.3294, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "eval_cer": 0.4522603730788784, | |
| "eval_loss": 1.833477258682251, | |
| "eval_runtime": 381.8605, | |
| "eval_samples_per_second": 24.821, | |
| "eval_steps_per_second": 3.103, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "grad_norm": 4.755228042602539, | |
| "learning_rate": 3.728763250883393e-05, | |
| "loss": 1.417, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "eval_cer": 0.47154491416057254, | |
| "eval_loss": 1.8104172945022583, | |
| "eval_runtime": 382.6075, | |
| "eval_samples_per_second": 24.772, | |
| "eval_steps_per_second": 3.097, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "grad_norm": 2.874833822250366, | |
| "learning_rate": 3.721696113074205e-05, | |
| "loss": 1.3215, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "eval_cer": 0.45071076610222516, | |
| "eval_loss": 3.382107973098755, | |
| "eval_runtime": 367.7162, | |
| "eval_samples_per_second": 25.775, | |
| "eval_steps_per_second": 3.223, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 7.87, | |
| "grad_norm": 2.091411590576172, | |
| "learning_rate": 3.714628975265018e-05, | |
| "loss": 1.7498, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 7.87, | |
| "eval_cer": 0.44587129951898635, | |
| "eval_loss": 2.4873616695404053, | |
| "eval_runtime": 381.5796, | |
| "eval_samples_per_second": 24.839, | |
| "eval_steps_per_second": 3.106, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "grad_norm": 2.1210122108459473, | |
| "learning_rate": 3.7075618374558305e-05, | |
| "loss": 1.4463, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "eval_cer": 0.4508696374799578, | |
| "eval_loss": 2.1447994709014893, | |
| "eval_runtime": 360.7307, | |
| "eval_samples_per_second": 26.274, | |
| "eval_steps_per_second": 3.285, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 7.96, | |
| "grad_norm": 9.81248950958252, | |
| "learning_rate": 3.700494699646643e-05, | |
| "loss": 1.327, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 7.96, | |
| "eval_cer": 0.45252678815846076, | |
| "eval_loss": 2.999907970428467, | |
| "eval_runtime": 382.5678, | |
| "eval_samples_per_second": 24.775, | |
| "eval_steps_per_second": 3.097, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.9528186321258545, | |
| "learning_rate": 3.693427561837456e-05, | |
| "loss": 1.3228, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.4471960423917719, | |
| "eval_loss": 1.6979167461395264, | |
| "eval_runtime": 363.6541, | |
| "eval_samples_per_second": 26.063, | |
| "eval_steps_per_second": 3.259, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "grad_norm": 3.4124321937561035, | |
| "learning_rate": 3.686360424028269e-05, | |
| "loss": 1.2656, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "eval_cer": 0.4476115521489187, | |
| "eval_loss": 2.0440304279327393, | |
| "eval_runtime": 381.9207, | |
| "eval_samples_per_second": 24.817, | |
| "eval_steps_per_second": 3.103, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 8.08, | |
| "grad_norm": 4.451693058013916, | |
| "learning_rate": 3.679293286219082e-05, | |
| "loss": 1.2961, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 8.08, | |
| "eval_cer": 0.44471520472410153, | |
| "eval_loss": 1.5803910493850708, | |
| "eval_runtime": 361.691, | |
| "eval_samples_per_second": 26.205, | |
| "eval_steps_per_second": 3.276, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 2.6323907375335693, | |
| "learning_rate": 3.672226148409894e-05, | |
| "loss": 1.2874, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "eval_cer": 0.4615849008642603, | |
| "eval_loss": 1.754168152809143, | |
| "eval_runtime": 385.1298, | |
| "eval_samples_per_second": 24.61, | |
| "eval_steps_per_second": 3.077, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "grad_norm": 2.5475683212280273, | |
| "learning_rate": 3.665159010600707e-05, | |
| "loss": 1.5227, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "eval_cer": 0.45406906260998786, | |
| "eval_loss": 1.954167127609253, | |
| "eval_runtime": 364.6983, | |
| "eval_samples_per_second": 25.989, | |
| "eval_steps_per_second": 3.249, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "grad_norm": 1.623938798904419, | |
| "learning_rate": 3.6580918727915195e-05, | |
| "loss": 1.2988, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "eval_cer": 0.4476164404990028, | |
| "eval_loss": 2.2084550857543945, | |
| "eval_runtime": 384.5947, | |
| "eval_samples_per_second": 24.644, | |
| "eval_steps_per_second": 3.081, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "grad_norm": 3.2966043949127197, | |
| "learning_rate": 3.6510954063604243e-05, | |
| "loss": 1.5739, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "eval_cer": 0.451500234640804, | |
| "eval_loss": 1.4389057159423828, | |
| "eval_runtime": 366.8675, | |
| "eval_samples_per_second": 25.835, | |
| "eval_steps_per_second": 3.23, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "grad_norm": 2.2112064361572266, | |
| "learning_rate": 3.6440282685512365e-05, | |
| "loss": 1.4271, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "eval_cer": 0.4538075358804896, | |
| "eval_loss": 2.505011796951294, | |
| "eval_runtime": 384.1903, | |
| "eval_samples_per_second": 24.67, | |
| "eval_steps_per_second": 3.084, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "grad_norm": 3.7650928497314453, | |
| "learning_rate": 3.63696113074205e-05, | |
| "loss": 1.665, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "eval_cer": 0.4431655977474483, | |
| "eval_loss": 2.0151588916778564, | |
| "eval_runtime": 364.0023, | |
| "eval_samples_per_second": 26.038, | |
| "eval_steps_per_second": 3.255, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 8.38, | |
| "grad_norm": 1.848604679107666, | |
| "learning_rate": 3.629893992932862e-05, | |
| "loss": 2.2543, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 8.38, | |
| "eval_cer": 0.4581728325055727, | |
| "eval_loss": 1.8045251369476318, | |
| "eval_runtime": 381.1374, | |
| "eval_samples_per_second": 24.868, | |
| "eval_steps_per_second": 3.109, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "grad_norm": 34.23902893066406, | |
| "learning_rate": 3.622826855123675e-05, | |
| "loss": 1.313, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "eval_cer": 0.4448887411520864, | |
| "eval_loss": 2.140085458755493, | |
| "eval_runtime": 364.9023, | |
| "eval_samples_per_second": 25.974, | |
| "eval_steps_per_second": 3.247, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.46, | |
| "grad_norm": 1.8063780069351196, | |
| "learning_rate": 3.615759717314488e-05, | |
| "loss": 2.0001, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 8.46, | |
| "eval_cer": 0.44904628289859605, | |
| "eval_loss": 3.0254170894622803, | |
| "eval_runtime": 380.5352, | |
| "eval_samples_per_second": 24.907, | |
| "eval_steps_per_second": 3.114, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 4.191483974456787, | |
| "learning_rate": 3.6086925795053005e-05, | |
| "loss": 1.2751, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "eval_cer": 0.4562517109225294, | |
| "eval_loss": 4.0329742431640625, | |
| "eval_runtime": 362.926, | |
| "eval_samples_per_second": 26.116, | |
| "eval_steps_per_second": 3.265, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "grad_norm": 6.447911262512207, | |
| "learning_rate": 3.6016254416961133e-05, | |
| "loss": 2.11, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "eval_cer": 0.43633657267998904, | |
| "eval_loss": 3.3446898460388184, | |
| "eval_runtime": 385.9035, | |
| "eval_samples_per_second": 24.561, | |
| "eval_steps_per_second": 3.071, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "grad_norm": 2.172891139984131, | |
| "learning_rate": 3.5945583038869255e-05, | |
| "loss": 1.4999, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "eval_cer": 0.4393355754565719, | |
| "eval_loss": 3.0930521488189697, | |
| "eval_runtime": 363.4753, | |
| "eval_samples_per_second": 26.076, | |
| "eval_steps_per_second": 3.26, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "grad_norm": 3.379791498184204, | |
| "learning_rate": 3.587491166077739e-05, | |
| "loss": 1.4381, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "eval_cer": 0.4450402800046928, | |
| "eval_loss": 2.9096951484680176, | |
| "eval_runtime": 384.973, | |
| "eval_samples_per_second": 24.62, | |
| "eval_steps_per_second": 3.078, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "grad_norm": 2.5090067386627197, | |
| "learning_rate": 3.580424028268551e-05, | |
| "loss": 1.6658, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "eval_cer": 0.4514855695905518, | |
| "eval_loss": 2.3211405277252197, | |
| "eval_runtime": 367.2057, | |
| "eval_samples_per_second": 25.811, | |
| "eval_steps_per_second": 3.227, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "grad_norm": 2.2183985710144043, | |
| "learning_rate": 3.573356890459364e-05, | |
| "loss": 2.1266, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "eval_cer": 0.45651323765202767, | |
| "eval_loss": 1.70187246799469, | |
| "eval_runtime": 382.2617, | |
| "eval_samples_per_second": 24.795, | |
| "eval_steps_per_second": 3.1, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "grad_norm": 1.713640570640564, | |
| "learning_rate": 3.5662897526501774e-05, | |
| "loss": 1.3038, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "eval_cer": 0.44218548355559034, | |
| "eval_loss": 1.8959708213806152, | |
| "eval_runtime": 362.6249, | |
| "eval_samples_per_second": 26.137, | |
| "eval_steps_per_second": 3.268, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 10.084565162658691, | |
| "learning_rate": 3.5592226148409895e-05, | |
| "loss": 1.3031, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "eval_cer": 0.44013237652027687, | |
| "eval_loss": 1.9289778470993042, | |
| "eval_runtime": 383.5806, | |
| "eval_samples_per_second": 24.709, | |
| "eval_steps_per_second": 3.089, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "grad_norm": 2.5064008235931396, | |
| "learning_rate": 3.552155477031802e-05, | |
| "loss": 2.0089, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "eval_cer": 0.4491318290250675, | |
| "eval_loss": 2.749382734298706, | |
| "eval_runtime": 372.5311, | |
| "eval_samples_per_second": 25.442, | |
| "eval_steps_per_second": 3.181, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 8.88, | |
| "grad_norm": 1.8727614879608154, | |
| "learning_rate": 3.5450883392226145e-05, | |
| "loss": 2.5929, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 8.88, | |
| "eval_cer": 0.4467780884595831, | |
| "eval_loss": 2.1603004932403564, | |
| "eval_runtime": 382.1598, | |
| "eval_samples_per_second": 24.801, | |
| "eval_steps_per_second": 3.101, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "grad_norm": 2.88053560256958, | |
| "learning_rate": 3.538021201413428e-05, | |
| "loss": 1.2652, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "eval_cer": 0.44302872394509407, | |
| "eval_loss": 2.7272582054138184, | |
| "eval_runtime": 364.629, | |
| "eval_samples_per_second": 25.994, | |
| "eval_steps_per_second": 3.25, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "grad_norm": 3.520254135131836, | |
| "learning_rate": 3.53095406360424e-05, | |
| "loss": 1.2819, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_cer": 0.44700784091353485, | |
| "eval_loss": 1.840844988822937, | |
| "eval_runtime": 393.1165, | |
| "eval_samples_per_second": 24.11, | |
| "eval_steps_per_second": 3.014, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 1.5368082523345947, | |
| "learning_rate": 3.523886925795053e-05, | |
| "loss": 1.2587, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "eval_cer": 0.43732890774705724, | |
| "eval_loss": 2.2236878871917725, | |
| "eval_runtime": 363.5106, | |
| "eval_samples_per_second": 26.074, | |
| "eval_steps_per_second": 3.26, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "grad_norm": 2.522552251815796, | |
| "learning_rate": 3.5168197879858664e-05, | |
| "loss": 1.2512, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "eval_cer": 0.44793418325446793, | |
| "eval_loss": 2.5103232860565186, | |
| "eval_runtime": 384.7009, | |
| "eval_samples_per_second": 24.637, | |
| "eval_steps_per_second": 3.08, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "grad_norm": 17.708566665649414, | |
| "learning_rate": 3.5097526501766785e-05, | |
| "loss": 1.2248, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "eval_cer": 0.4432878064995503, | |
| "eval_loss": 2.5621256828308105, | |
| "eval_runtime": 370.664, | |
| "eval_samples_per_second": 25.57, | |
| "eval_steps_per_second": 3.197, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "grad_norm": 2.740123987197876, | |
| "learning_rate": 3.502685512367491e-05, | |
| "loss": 1.3298, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "eval_cer": 0.441982617027101, | |
| "eval_loss": 2.49664568901062, | |
| "eval_runtime": 372.6231, | |
| "eval_samples_per_second": 25.436, | |
| "eval_steps_per_second": 3.18, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "grad_norm": 7.561341762542725, | |
| "learning_rate": 3.495689045936396e-05, | |
| "loss": 1.2998, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "eval_cer": 0.43526113566149155, | |
| "eval_loss": 3.151437282562256, | |
| "eval_runtime": 381.1035, | |
| "eval_samples_per_second": 24.87, | |
| "eval_steps_per_second": 3.109, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "grad_norm": 2.588192939758301, | |
| "learning_rate": 3.488621908127209e-05, | |
| "loss": 1.2493, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "eval_cer": 0.46399241328066954, | |
| "eval_loss": 1.8955408334732056, | |
| "eval_runtime": 369.5234, | |
| "eval_samples_per_second": 25.649, | |
| "eval_steps_per_second": 3.207, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "grad_norm": 2.3700144290924072, | |
| "learning_rate": 3.481554770318021e-05, | |
| "loss": 1.3379, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "eval_cer": 0.4379032888819366, | |
| "eval_loss": 1.7394232749938965, | |
| "eval_runtime": 383.407, | |
| "eval_samples_per_second": 24.72, | |
| "eval_steps_per_second": 3.091, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 2.381225824356079, | |
| "learning_rate": 3.4744876325088346e-05, | |
| "loss": 2.2422, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "eval_cer": 0.43050965937976615, | |
| "eval_loss": 1.691564917564392, | |
| "eval_runtime": 362.7906, | |
| "eval_samples_per_second": 26.125, | |
| "eval_steps_per_second": 3.266, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "grad_norm": 4.20334005355835, | |
| "learning_rate": 3.467420494699647e-05, | |
| "loss": 1.239, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "eval_cer": 0.44948623440616325, | |
| "eval_loss": 3.0320708751678467, | |
| "eval_runtime": 386.6159, | |
| "eval_samples_per_second": 24.515, | |
| "eval_steps_per_second": 3.065, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "grad_norm": 12.748970985412598, | |
| "learning_rate": 3.4603533568904595e-05, | |
| "loss": 1.622, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "eval_cer": 0.4492564819522115, | |
| "eval_loss": 2.2297956943511963, | |
| "eval_runtime": 368.8399, | |
| "eval_samples_per_second": 25.697, | |
| "eval_steps_per_second": 3.213, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "grad_norm": 10.591804504394531, | |
| "learning_rate": 3.4532862190812723e-05, | |
| "loss": 1.4811, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "eval_cer": 0.44453922412107466, | |
| "eval_loss": 2.3979556560516357, | |
| "eval_runtime": 383.7637, | |
| "eval_samples_per_second": 24.697, | |
| "eval_steps_per_second": 3.088, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "grad_norm": 7.108492374420166, | |
| "learning_rate": 3.446219081272085e-05, | |
| "loss": 1.5169, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "eval_cer": 0.4425838840874428, | |
| "eval_loss": 1.763051152229309, | |
| "eval_runtime": 371.2126, | |
| "eval_samples_per_second": 25.533, | |
| "eval_steps_per_second": 3.192, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.51, | |
| "grad_norm": 2.171220302581787, | |
| "learning_rate": 3.439151943462898e-05, | |
| "loss": 1.2667, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 9.51, | |
| "eval_cer": 0.44492784795275897, | |
| "eval_loss": 1.6504524946212769, | |
| "eval_runtime": 384.6648, | |
| "eval_samples_per_second": 24.64, | |
| "eval_steps_per_second": 3.081, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 9.56, | |
| "grad_norm": 4.868325233459473, | |
| "learning_rate": 3.43208480565371e-05, | |
| "loss": 1.2425, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 9.56, | |
| "eval_cer": 0.4337261937350905, | |
| "eval_loss": 1.6726810932159424, | |
| "eval_runtime": 360.9955, | |
| "eval_samples_per_second": 26.255, | |
| "eval_steps_per_second": 3.283, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 4.784325122833252, | |
| "learning_rate": 3.4250176678445236e-05, | |
| "loss": 1.2519, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "eval_cer": 0.4394284541081694, | |
| "eval_loss": 2.026627540588379, | |
| "eval_runtime": 383.883, | |
| "eval_samples_per_second": 24.69, | |
| "eval_steps_per_second": 3.087, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 9.64, | |
| "grad_norm": 34.55485534667969, | |
| "learning_rate": 3.417950530035336e-05, | |
| "loss": 1.2678, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 9.64, | |
| "eval_cer": 0.45279075906300104, | |
| "eval_loss": 1.6792824268341064, | |
| "eval_runtime": 362.9993, | |
| "eval_samples_per_second": 26.11, | |
| "eval_steps_per_second": 3.264, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "grad_norm": 8.905746459960938, | |
| "learning_rate": 3.4108833922261485e-05, | |
| "loss": 1.2559, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "eval_cer": 0.452297035704509, | |
| "eval_loss": 1.6737189292907715, | |
| "eval_runtime": 383.7761, | |
| "eval_samples_per_second": 24.697, | |
| "eval_steps_per_second": 3.088, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.72, | |
| "grad_norm": 3.8121345043182373, | |
| "learning_rate": 3.4038162544169613e-05, | |
| "loss": 1.2646, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 9.72, | |
| "eval_cer": 0.45129247976223064, | |
| "eval_loss": 1.4810179471969604, | |
| "eval_runtime": 361.7355, | |
| "eval_samples_per_second": 26.201, | |
| "eval_steps_per_second": 3.276, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 9.77, | |
| "grad_norm": 1.6551660299301147, | |
| "learning_rate": 3.396749116607774e-05, | |
| "loss": 1.2812, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 9.77, | |
| "eval_cer": 0.4401079347698565, | |
| "eval_loss": 1.4080007076263428, | |
| "eval_runtime": 387.5504, | |
| "eval_samples_per_second": 24.456, | |
| "eval_steps_per_second": 3.058, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 9.81, | |
| "grad_norm": 1.3945401906967163, | |
| "learning_rate": 3.389681978798587e-05, | |
| "loss": 1.2539, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 9.81, | |
| "eval_cer": 0.44230280395760824, | |
| "eval_loss": 1.8802071809768677, | |
| "eval_runtime": 376.3421, | |
| "eval_samples_per_second": 25.185, | |
| "eval_steps_per_second": 3.149, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "grad_norm": 19.793508529663086, | |
| "learning_rate": 3.382614840989399e-05, | |
| "loss": 1.9455, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "eval_cer": 0.45229947987955105, | |
| "eval_loss": 1.607908010482788, | |
| "eval_runtime": 388.4585, | |
| "eval_samples_per_second": 24.399, | |
| "eval_steps_per_second": 3.051, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "grad_norm": 25.258953094482422, | |
| "learning_rate": 3.3755477031802126e-05, | |
| "loss": 1.2352, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "eval_cer": 0.43993439834187165, | |
| "eval_loss": 1.3818862438201904, | |
| "eval_runtime": 363.0875, | |
| "eval_samples_per_second": 26.104, | |
| "eval_steps_per_second": 3.264, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 9.93, | |
| "grad_norm": 9.951347351074219, | |
| "learning_rate": 3.368480565371025e-05, | |
| "loss": 1.5735, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 9.93, | |
| "eval_cer": 0.4422074811309687, | |
| "eval_loss": 1.6464565992355347, | |
| "eval_runtime": 386.065, | |
| "eval_samples_per_second": 24.55, | |
| "eval_steps_per_second": 3.069, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 9.98, | |
| "grad_norm": 4.360719203948975, | |
| "learning_rate": 3.3614134275618375e-05, | |
| "loss": 1.6236, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 9.98, | |
| "eval_cer": 0.44366909780610847, | |
| "eval_loss": 1.770129919052124, | |
| "eval_runtime": 361.9008, | |
| "eval_samples_per_second": 26.189, | |
| "eval_steps_per_second": 3.274, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 12.59276294708252, | |
| "learning_rate": 3.35434628975265e-05, | |
| "loss": 1.2522, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "eval_cer": 0.43885651714833207, | |
| "eval_loss": 2.233562707901001, | |
| "eval_runtime": 381.3188, | |
| "eval_samples_per_second": 24.856, | |
| "eval_steps_per_second": 3.108, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "grad_norm": 6.893917560577393, | |
| "learning_rate": 3.347279151943463e-05, | |
| "loss": 1.2052, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "eval_cer": 0.44252522388643384, | |
| "eval_loss": 2.4695940017700195, | |
| "eval_runtime": 366.61, | |
| "eval_samples_per_second": 25.853, | |
| "eval_steps_per_second": 3.232, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 2.430925130844116, | |
| "learning_rate": 3.340212014134276e-05, | |
| "loss": 1.185, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "eval_cer": 0.4357133080442689, | |
| "eval_loss": 1.5434983968734741, | |
| "eval_runtime": 385.0716, | |
| "eval_samples_per_second": 24.614, | |
| "eval_steps_per_second": 3.077, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.15, | |
| "grad_norm": 2.367770195007324, | |
| "learning_rate": 3.333144876325088e-05, | |
| "loss": 1.2225, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 10.15, | |
| "eval_cer": 0.4268898361425052, | |
| "eval_loss": 1.602295160293579, | |
| "eval_runtime": 368.0631, | |
| "eval_samples_per_second": 25.751, | |
| "eval_steps_per_second": 3.22, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 10.19, | |
| "grad_norm": 2.6913535594940186, | |
| "learning_rate": 3.3260777385159016e-05, | |
| "loss": 1.2071, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 10.19, | |
| "eval_cer": 0.43098382933792184, | |
| "eval_loss": 1.9375296831130981, | |
| "eval_runtime": 388.5754, | |
| "eval_samples_per_second": 24.392, | |
| "eval_steps_per_second": 3.05, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 10.23, | |
| "grad_norm": 2.8861985206604004, | |
| "learning_rate": 3.319010600706714e-05, | |
| "loss": 1.2745, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 10.23, | |
| "eval_cer": 0.43681563098822884, | |
| "eval_loss": 1.8604010343551636, | |
| "eval_runtime": 365.4924, | |
| "eval_samples_per_second": 25.932, | |
| "eval_steps_per_second": 3.242, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 10.27, | |
| "grad_norm": 3.54598069190979, | |
| "learning_rate": 3.3119434628975265e-05, | |
| "loss": 1.3486, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 10.27, | |
| "eval_cer": 0.4357133080442689, | |
| "eval_loss": 1.3347864151000977, | |
| "eval_runtime": 380.5736, | |
| "eval_samples_per_second": 24.905, | |
| "eval_steps_per_second": 3.114, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "grad_norm": 2.1198437213897705, | |
| "learning_rate": 3.304876325088339e-05, | |
| "loss": 1.1866, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "eval_cer": 0.4299792733956435, | |
| "eval_loss": 1.308254361152649, | |
| "eval_runtime": 368.9525, | |
| "eval_samples_per_second": 25.689, | |
| "eval_steps_per_second": 3.212, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "grad_norm": 2.029782295227051, | |
| "learning_rate": 3.297809187279152e-05, | |
| "loss": 1.1697, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "eval_cer": 0.42697293809393455, | |
| "eval_loss": 1.4594690799713135, | |
| "eval_runtime": 389.4322, | |
| "eval_samples_per_second": 24.338, | |
| "eval_steps_per_second": 3.043, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 2.0395631790161133, | |
| "learning_rate": 3.290742049469965e-05, | |
| "loss": 1.1793, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "eval_cer": 0.42983506706816316, | |
| "eval_loss": 1.475653886795044, | |
| "eval_runtime": 374.0567, | |
| "eval_samples_per_second": 25.338, | |
| "eval_steps_per_second": 3.168, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 10.44, | |
| "grad_norm": 6.735607624053955, | |
| "learning_rate": 3.283674911660777e-05, | |
| "loss": 1.4708, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 10.44, | |
| "eval_cer": 0.4388076336474913, | |
| "eval_loss": 1.9250705242156982, | |
| "eval_runtime": 386.263, | |
| "eval_samples_per_second": 24.538, | |
| "eval_steps_per_second": 3.068, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 10.48, | |
| "grad_norm": 1.8092114925384521, | |
| "learning_rate": 3.2766077738515906e-05, | |
| "loss": 1.1854, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 10.48, | |
| "eval_cer": 0.4322132493840679, | |
| "eval_loss": 1.8587048053741455, | |
| "eval_runtime": 371.4948, | |
| "eval_samples_per_second": 25.513, | |
| "eval_steps_per_second": 3.19, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 10.52, | |
| "grad_norm": 3.406538724899292, | |
| "learning_rate": 3.269540636042403e-05, | |
| "loss": 1.2144, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.52, | |
| "eval_cer": 0.4353711235383833, | |
| "eval_loss": 1.5310605764389038, | |
| "eval_runtime": 388.4081, | |
| "eval_samples_per_second": 24.402, | |
| "eval_steps_per_second": 3.051, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.57, | |
| "grad_norm": 21.588300704956055, | |
| "learning_rate": 3.2624734982332155e-05, | |
| "loss": 1.1869, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 10.57, | |
| "eval_cer": 0.41679294904383873, | |
| "eval_loss": 1.753686785697937, | |
| "eval_runtime": 363.8235, | |
| "eval_samples_per_second": 26.051, | |
| "eval_steps_per_second": 3.257, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "grad_norm": 2.433943033218384, | |
| "learning_rate": 3.255406360424029e-05, | |
| "loss": 1.186, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "eval_cer": 0.4293755621602597, | |
| "eval_loss": 2.9843032360076904, | |
| "eval_runtime": 383.8866, | |
| "eval_samples_per_second": 24.69, | |
| "eval_steps_per_second": 3.087, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 10.65, | |
| "grad_norm": 2.10840106010437, | |
| "learning_rate": 3.248339222614841e-05, | |
| "loss": 1.2008, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 10.65, | |
| "eval_cer": 0.42626657150678504, | |
| "eval_loss": 2.721179485321045, | |
| "eval_runtime": 367.0761, | |
| "eval_samples_per_second": 25.82, | |
| "eval_steps_per_second": 3.228, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 10.69, | |
| "grad_norm": 2.1997592449188232, | |
| "learning_rate": 3.241272084805654e-05, | |
| "loss": 1.1923, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 10.69, | |
| "eval_cer": 0.42187438895623947, | |
| "eval_loss": 2.4305357933044434, | |
| "eval_runtime": 385.3951, | |
| "eval_samples_per_second": 24.593, | |
| "eval_steps_per_second": 3.075, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 10.73, | |
| "grad_norm": 1.7977826595306396, | |
| "learning_rate": 3.234204946996466e-05, | |
| "loss": 1.7363, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.73, | |
| "eval_cer": 0.4337506354855109, | |
| "eval_loss": 2.236682176589966, | |
| "eval_runtime": 370.2296, | |
| "eval_samples_per_second": 25.6, | |
| "eval_steps_per_second": 3.201, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.78, | |
| "grad_norm": 3.47428035736084, | |
| "learning_rate": 3.2271378091872796e-05, | |
| "loss": 1.3785, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 10.78, | |
| "eval_cer": 0.4252571272144226, | |
| "eval_loss": 2.5334064960479736, | |
| "eval_runtime": 380.018, | |
| "eval_samples_per_second": 24.941, | |
| "eval_steps_per_second": 3.118, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "grad_norm": 2.5327584743499756, | |
| "learning_rate": 3.2200706713780924e-05, | |
| "loss": 1.518, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "eval_cer": 0.42553087481913104, | |
| "eval_loss": 2.6629116535186768, | |
| "eval_runtime": 359.4773, | |
| "eval_samples_per_second": 26.366, | |
| "eval_steps_per_second": 3.296, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "grad_norm": 5.47965145111084, | |
| "learning_rate": 3.2130035335689045e-05, | |
| "loss": 1.1865, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "eval_cer": 0.4287083023737828, | |
| "eval_loss": 1.6172243356704712, | |
| "eval_runtime": 386.7546, | |
| "eval_samples_per_second": 24.506, | |
| "eval_steps_per_second": 3.064, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 7.293168544769287, | |
| "learning_rate": 3.2060070671378093e-05, | |
| "loss": 1.3999, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "eval_cer": 0.42187194478119744, | |
| "eval_loss": 1.8544808626174927, | |
| "eval_runtime": 365.482, | |
| "eval_samples_per_second": 25.933, | |
| "eval_steps_per_second": 3.242, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 10.95, | |
| "grad_norm": 3.8992362022399902, | |
| "learning_rate": 3.198939929328622e-05, | |
| "loss": 1.1808, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 10.95, | |
| "eval_cer": 0.419022036682179, | |
| "eval_loss": 1.6932332515716553, | |
| "eval_runtime": 384.5535, | |
| "eval_samples_per_second": 24.647, | |
| "eval_steps_per_second": 3.081, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "grad_norm": 64.95602416992188, | |
| "learning_rate": 3.191872791519435e-05, | |
| "loss": 1.2023, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "eval_cer": 0.4221408040358218, | |
| "eval_loss": 1.5864986181259155, | |
| "eval_runtime": 363.5244, | |
| "eval_samples_per_second": 26.073, | |
| "eval_steps_per_second": 3.26, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 1.6677500009536743, | |
| "learning_rate": 3.184805653710248e-05, | |
| "loss": 1.1585, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "eval_cer": 0.4356106526925032, | |
| "eval_loss": 1.4082372188568115, | |
| "eval_runtime": 386.3668, | |
| "eval_samples_per_second": 24.531, | |
| "eval_steps_per_second": 3.067, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 11.07, | |
| "grad_norm": 1.4230600595474243, | |
| "learning_rate": 3.1777385159010606e-05, | |
| "loss": 1.2774, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 11.07, | |
| "eval_cer": 0.41568573774979467, | |
| "eval_loss": 1.6947673559188843, | |
| "eval_runtime": 370.2625, | |
| "eval_samples_per_second": 25.598, | |
| "eval_steps_per_second": 3.2, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "grad_norm": 10.38554859161377, | |
| "learning_rate": 3.170671378091873e-05, | |
| "loss": 1.1303, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "eval_cer": 0.42248543271674943, | |
| "eval_loss": 2.5170605182647705, | |
| "eval_runtime": 383.496, | |
| "eval_samples_per_second": 24.715, | |
| "eval_steps_per_second": 3.09, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 11.16, | |
| "grad_norm": 1.8860437870025635, | |
| "learning_rate": 3.1636042402826855e-05, | |
| "loss": 1.1186, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.16, | |
| "eval_cer": 0.4318832857533925, | |
| "eval_loss": 1.2920070886611938, | |
| "eval_runtime": 374.2965, | |
| "eval_samples_per_second": 25.322, | |
| "eval_steps_per_second": 3.166, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 2.1193289756774902, | |
| "learning_rate": 3.156537102473498e-05, | |
| "loss": 1.1587, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "eval_cer": 0.4244921004262641, | |
| "eval_loss": 2.2303755283355713, | |
| "eval_runtime": 390.1983, | |
| "eval_samples_per_second": 24.29, | |
| "eval_steps_per_second": 3.037, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 11.24, | |
| "grad_norm": 13.682442665100098, | |
| "learning_rate": 3.149469964664311e-05, | |
| "loss": 1.1569, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 11.24, | |
| "eval_cer": 0.4308762856360721, | |
| "eval_loss": 1.715321660041809, | |
| "eval_runtime": 368.2313, | |
| "eval_samples_per_second": 25.739, | |
| "eval_steps_per_second": 3.218, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 11.28, | |
| "grad_norm": 1.7043545246124268, | |
| "learning_rate": 3.142402826855124e-05, | |
| "loss": 1.1348, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 11.28, | |
| "eval_cer": 0.41776573071057055, | |
| "eval_loss": 2.2900288105010986, | |
| "eval_runtime": 394.8533, | |
| "eval_samples_per_second": 24.004, | |
| "eval_steps_per_second": 3.001, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "grad_norm": 1.8619517087936401, | |
| "learning_rate": 3.135335689045937e-05, | |
| "loss": 1.1366, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "eval_cer": 0.42347532360877554, | |
| "eval_loss": 1.863010287284851, | |
| "eval_runtime": 370.2336, | |
| "eval_samples_per_second": 25.6, | |
| "eval_steps_per_second": 3.201, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 11.37, | |
| "grad_norm": 1.628806710243225, | |
| "learning_rate": 3.1282685512367496e-05, | |
| "loss": 1.1399, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.37, | |
| "eval_cer": 0.4168051699190489, | |
| "eval_loss": 1.782639741897583, | |
| "eval_runtime": 391.1594, | |
| "eval_samples_per_second": 24.231, | |
| "eval_steps_per_second": 3.029, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.41, | |
| "grad_norm": 3.158604621887207, | |
| "learning_rate": 3.121201413427562e-05, | |
| "loss": 1.1373, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 11.41, | |
| "eval_cer": 0.4201585780767275, | |
| "eval_loss": 1.5199880599975586, | |
| "eval_runtime": 368.0376, | |
| "eval_samples_per_second": 25.753, | |
| "eval_steps_per_second": 3.22, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 11.45, | |
| "grad_norm": 1.9599921703338623, | |
| "learning_rate": 3.1141342756183745e-05, | |
| "loss": 1.1358, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 11.45, | |
| "eval_cer": 0.4173404442532556, | |
| "eval_loss": 1.598440170288086, | |
| "eval_runtime": 388.8994, | |
| "eval_samples_per_second": 24.371, | |
| "eval_steps_per_second": 3.047, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "grad_norm": 1.7773711681365967, | |
| "learning_rate": 3.107067137809187e-05, | |
| "loss": 1.1567, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "eval_cer": 0.4288060693754644, | |
| "eval_loss": 3.0100185871124268, | |
| "eval_runtime": 392.5508, | |
| "eval_samples_per_second": 24.145, | |
| "eval_steps_per_second": 3.019, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 11.53, | |
| "grad_norm": 19.142953872680664, | |
| "learning_rate": 3.1e-05, | |
| "loss": 1.1284, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 11.53, | |
| "eval_cer": 0.43251877126432287, | |
| "eval_loss": 2.1405186653137207, | |
| "eval_runtime": 413.4099, | |
| "eval_samples_per_second": 22.926, | |
| "eval_steps_per_second": 2.866, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 11.58, | |
| "grad_norm": 9.353096961975098, | |
| "learning_rate": 3.092932862190813e-05, | |
| "loss": 1.3106, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.58, | |
| "eval_cer": 0.4250787024363537, | |
| "eval_loss": 2.7651588916778564, | |
| "eval_runtime": 372.5808, | |
| "eval_samples_per_second": 25.439, | |
| "eval_steps_per_second": 3.181, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.62, | |
| "grad_norm": 2.0867862701416016, | |
| "learning_rate": 3.085865724381626e-05, | |
| "loss": 1.1146, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 11.62, | |
| "eval_cer": 0.41358374721364044, | |
| "eval_loss": 2.7960143089294434, | |
| "eval_runtime": 392.1126, | |
| "eval_samples_per_second": 24.172, | |
| "eval_steps_per_second": 3.022, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "grad_norm": 1.8952158689498901, | |
| "learning_rate": 3.0787985865724386e-05, | |
| "loss": 1.1186, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "eval_cer": 0.41495004106214073, | |
| "eval_loss": 2.6641368865966797, | |
| "eval_runtime": 367.6488, | |
| "eval_samples_per_second": 25.78, | |
| "eval_steps_per_second": 3.223, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 6.817134380340576, | |
| "learning_rate": 3.071802120141343e-05, | |
| "loss": 1.581, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "eval_cer": 0.4181641312424231, | |
| "eval_loss": 2.996872901916504, | |
| "eval_runtime": 382.8066, | |
| "eval_samples_per_second": 24.759, | |
| "eval_steps_per_second": 3.096, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "grad_norm": 4.087586879730225, | |
| "learning_rate": 3.0647349823321555e-05, | |
| "loss": 1.1385, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "eval_cer": 0.42436255914903603, | |
| "eval_loss": 2.688666343688965, | |
| "eval_runtime": 375.5428, | |
| "eval_samples_per_second": 25.238, | |
| "eval_steps_per_second": 3.155, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 11.79, | |
| "grad_norm": 2.216014862060547, | |
| "learning_rate": 3.0576678445229683e-05, | |
| "loss": 1.2095, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.79, | |
| "eval_cer": 0.42844188729420046, | |
| "eval_loss": 1.8659263849258423, | |
| "eval_runtime": 392.5017, | |
| "eval_samples_per_second": 24.148, | |
| "eval_steps_per_second": 3.019, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.83, | |
| "grad_norm": 6.676244258880615, | |
| "learning_rate": 3.0506007067137808e-05, | |
| "loss": 1.4826, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 11.83, | |
| "eval_cer": 0.4170275898478745, | |
| "eval_loss": 1.5196857452392578, | |
| "eval_runtime": 370.7857, | |
| "eval_samples_per_second": 25.562, | |
| "eval_steps_per_second": 3.196, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 11.87, | |
| "grad_norm": 5.378576755523682, | |
| "learning_rate": 3.0435335689045936e-05, | |
| "loss": 1.125, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 11.87, | |
| "eval_cer": 0.4203663329553009, | |
| "eval_loss": 1.5670437812805176, | |
| "eval_runtime": 388.8978, | |
| "eval_samples_per_second": 24.371, | |
| "eval_steps_per_second": 3.047, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 11.91, | |
| "grad_norm": 4.532820701599121, | |
| "learning_rate": 3.0364664310954068e-05, | |
| "loss": 1.1398, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 11.91, | |
| "eval_cer": 0.42046654413202456, | |
| "eval_loss": 1.8523951768875122, | |
| "eval_runtime": 367.1427, | |
| "eval_samples_per_second": 25.816, | |
| "eval_steps_per_second": 3.228, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 11.96, | |
| "grad_norm": 13.059289932250977, | |
| "learning_rate": 3.0293992932862192e-05, | |
| "loss": 1.1318, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 11.96, | |
| "eval_cer": 0.4224341050408666, | |
| "eval_loss": 2.1082444190979004, | |
| "eval_runtime": 393.3473, | |
| "eval_samples_per_second": 24.096, | |
| "eval_steps_per_second": 3.013, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 5.903038501739502, | |
| "learning_rate": 3.0223321554770317e-05, | |
| "loss": 1.1131, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.41807125259082556, | |
| "eval_loss": 1.7783021926879883, | |
| "eval_runtime": 365.8673, | |
| "eval_samples_per_second": 25.906, | |
| "eval_steps_per_second": 3.239, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.04, | |
| "grad_norm": 7.630153656005859, | |
| "learning_rate": 3.015265017667845e-05, | |
| "loss": 1.0524, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 12.04, | |
| "eval_cer": 0.4113008877243753, | |
| "eval_loss": 1.9460973739624023, | |
| "eval_runtime": 389.3684, | |
| "eval_samples_per_second": 24.342, | |
| "eval_steps_per_second": 3.043, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 12.08, | |
| "grad_norm": 16.8402099609375, | |
| "learning_rate": 3.0081978798586573e-05, | |
| "loss": 1.073, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 12.08, | |
| "eval_cer": 0.4151773493410504, | |
| "eval_loss": 2.0268595218658447, | |
| "eval_runtime": 367.5014, | |
| "eval_samples_per_second": 25.79, | |
| "eval_steps_per_second": 3.224, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 12.12, | |
| "grad_norm": 1.9278266429901123, | |
| "learning_rate": 3.0011307420494698e-05, | |
| "loss": 1.0857, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 12.12, | |
| "eval_cer": 0.4125180868953111, | |
| "eval_loss": 1.7026644945144653, | |
| "eval_runtime": 390.3027, | |
| "eval_samples_per_second": 24.284, | |
| "eval_steps_per_second": 3.036, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 12.17, | |
| "grad_norm": 1.4629472494125366, | |
| "learning_rate": 2.994063604240283e-05, | |
| "loss": 1.1243, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 12.17, | |
| "eval_cer": 0.411283778499081, | |
| "eval_loss": 2.061549663543701, | |
| "eval_runtime": 372.4148, | |
| "eval_samples_per_second": 25.45, | |
| "eval_steps_per_second": 3.182, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "grad_norm": 15.632731437683105, | |
| "learning_rate": 2.9869964664310958e-05, | |
| "loss": 1.0708, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "eval_cer": 0.41651920143913024, | |
| "eval_loss": 2.2883496284484863, | |
| "eval_runtime": 388.5509, | |
| "eval_samples_per_second": 24.393, | |
| "eval_steps_per_second": 3.05, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.25, | |
| "grad_norm": 6.1355180740356445, | |
| "learning_rate": 2.9799293286219082e-05, | |
| "loss": 1.0688, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 12.25, | |
| "eval_cer": 0.4098588244495718, | |
| "eval_loss": 2.406083822250366, | |
| "eval_runtime": 374.5529, | |
| "eval_samples_per_second": 25.305, | |
| "eval_steps_per_second": 3.164, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 12.29, | |
| "grad_norm": 26.448299407958984, | |
| "learning_rate": 2.9728621908127207e-05, | |
| "loss": 1.091, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 12.29, | |
| "eval_cer": 0.41021811818075166, | |
| "eval_loss": 1.8954789638519287, | |
| "eval_runtime": 388.9635, | |
| "eval_samples_per_second": 24.367, | |
| "eval_steps_per_second": 3.047, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 12.33, | |
| "grad_norm": 19.144874572753906, | |
| "learning_rate": 2.965795053003534e-05, | |
| "loss": 1.0624, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 12.33, | |
| "eval_cer": 0.41046497985999764, | |
| "eval_loss": 1.6875534057617188, | |
| "eval_runtime": 368.9225, | |
| "eval_samples_per_second": 25.691, | |
| "eval_steps_per_second": 3.212, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 12.38, | |
| "grad_norm": 2.1881866455078125, | |
| "learning_rate": 2.9587279151943463e-05, | |
| "loss": 1.0462, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 12.38, | |
| "eval_cer": 0.40699180712525906, | |
| "eval_loss": 1.977940320968628, | |
| "eval_runtime": 389.1348, | |
| "eval_samples_per_second": 24.357, | |
| "eval_steps_per_second": 3.045, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 12.42, | |
| "grad_norm": 2.38139009475708, | |
| "learning_rate": 2.951660777385159e-05, | |
| "loss": 1.0711, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.42, | |
| "eval_cer": 0.4067278362207188, | |
| "eval_loss": 1.981135606765747, | |
| "eval_runtime": 374.3024, | |
| "eval_samples_per_second": 25.322, | |
| "eval_steps_per_second": 3.166, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.46, | |
| "grad_norm": 13.329228401184082, | |
| "learning_rate": 2.944593639575972e-05, | |
| "loss": 1.041, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 12.46, | |
| "eval_cer": 0.41142309647647723, | |
| "eval_loss": 2.5235061645507812, | |
| "eval_runtime": 396.1248, | |
| "eval_samples_per_second": 23.927, | |
| "eval_steps_per_second": 2.991, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 3.899754524230957, | |
| "learning_rate": 2.9375265017667848e-05, | |
| "loss": 1.1686, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "eval_cer": 0.4094359821672989, | |
| "eval_loss": 1.6854169368743896, | |
| "eval_runtime": 371.974, | |
| "eval_samples_per_second": 25.48, | |
| "eval_steps_per_second": 3.186, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 12.54, | |
| "grad_norm": 4.518542766571045, | |
| "learning_rate": 2.9304593639575972e-05, | |
| "loss": 1.0679, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 12.54, | |
| "eval_cer": 0.4047798287122131, | |
| "eval_loss": 2.345759868621826, | |
| "eval_runtime": 387.3324, | |
| "eval_samples_per_second": 24.47, | |
| "eval_steps_per_second": 3.059, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 12.59, | |
| "grad_norm": 3.159818410873413, | |
| "learning_rate": 2.9233922261484097e-05, | |
| "loss": 1.0724, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 12.59, | |
| "eval_cer": 0.4022965468695006, | |
| "eval_loss": 2.1892457008361816, | |
| "eval_runtime": 368.124, | |
| "eval_samples_per_second": 25.747, | |
| "eval_steps_per_second": 3.219, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 12.63, | |
| "grad_norm": 5.468438148498535, | |
| "learning_rate": 2.916395759717315e-05, | |
| "loss": 1.3109, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.63, | |
| "eval_cer": 0.4083043291228345, | |
| "eval_loss": 2.8631815910339355, | |
| "eval_runtime": 397.7294, | |
| "eval_samples_per_second": 23.83, | |
| "eval_steps_per_second": 2.979, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.67, | |
| "grad_norm": 3.4033985137939453, | |
| "learning_rate": 2.9093286219081274e-05, | |
| "loss": 1.0724, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 12.67, | |
| "eval_cer": 0.4132855578585116, | |
| "eval_loss": 2.082000732421875, | |
| "eval_runtime": 370.8747, | |
| "eval_samples_per_second": 25.556, | |
| "eval_steps_per_second": 3.195, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 12.71, | |
| "grad_norm": 8.163016319274902, | |
| "learning_rate": 2.9022614840989398e-05, | |
| "loss": 1.2078, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 12.71, | |
| "eval_cer": 0.41148908920261235, | |
| "eval_loss": 2.2178566455841064, | |
| "eval_runtime": 392.2715, | |
| "eval_samples_per_second": 24.162, | |
| "eval_steps_per_second": 3.021, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 12.76, | |
| "grad_norm": 7.318975448608398, | |
| "learning_rate": 2.895194346289753e-05, | |
| "loss": 1.0775, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 12.76, | |
| "eval_cer": 0.4067913847718118, | |
| "eval_loss": 2.1497671604156494, | |
| "eval_runtime": 374.4081, | |
| "eval_samples_per_second": 25.315, | |
| "eval_steps_per_second": 3.165, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 1.8343230485916138, | |
| "learning_rate": 2.8881272084805654e-05, | |
| "loss": 1.0572, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "eval_cer": 0.40438142818036055, | |
| "eval_loss": 2.247307062149048, | |
| "eval_runtime": 396.7902, | |
| "eval_samples_per_second": 23.887, | |
| "eval_steps_per_second": 2.986, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 12.84, | |
| "grad_norm": 2.385056495666504, | |
| "learning_rate": 2.881060070671378e-05, | |
| "loss": 1.0668, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 12.84, | |
| "eval_cer": 0.40543242344843766, | |
| "eval_loss": 2.0315330028533936, | |
| "eval_runtime": 370.6236, | |
| "eval_samples_per_second": 25.573, | |
| "eval_steps_per_second": 3.197, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "grad_norm": 3.699707508087158, | |
| "learning_rate": 2.873992932862191e-05, | |
| "loss": 1.2704, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "eval_cer": 0.40483360056313794, | |
| "eval_loss": 3.221102476119995, | |
| "eval_runtime": 396.883, | |
| "eval_samples_per_second": 23.881, | |
| "eval_steps_per_second": 2.986, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "grad_norm": 2.9071247577667236, | |
| "learning_rate": 2.866925795053004e-05, | |
| "loss": 2.4896, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "eval_cer": 0.4067840522466857, | |
| "eval_loss": 2.3889622688293457, | |
| "eval_runtime": 374.4971, | |
| "eval_samples_per_second": 25.309, | |
| "eval_steps_per_second": 3.164, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 12.97, | |
| "grad_norm": 2.6603596210479736, | |
| "learning_rate": 2.8598586572438163e-05, | |
| "loss": 1.0564, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 12.97, | |
| "eval_cer": 0.40476760783700283, | |
| "eval_loss": 2.143598794937134, | |
| "eval_runtime": 399.5314, | |
| "eval_samples_per_second": 23.723, | |
| "eval_steps_per_second": 2.966, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 2.8563497066497803, | |
| "learning_rate": 2.8527915194346288e-05, | |
| "loss": 1.0411, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "eval_cer": 0.40590170505650935, | |
| "eval_loss": 2.6135616302490234, | |
| "eval_runtime": 370.2003, | |
| "eval_samples_per_second": 25.602, | |
| "eval_steps_per_second": 3.201, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 13.05, | |
| "grad_norm": 3.738598585128784, | |
| "learning_rate": 2.845724381625442e-05, | |
| "loss": 0.9841, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.05, | |
| "eval_cer": 0.4061705643111337, | |
| "eval_loss": 2.7012014389038086, | |
| "eval_runtime": 394.4654, | |
| "eval_samples_per_second": 24.027, | |
| "eval_steps_per_second": 3.004, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.09, | |
| "grad_norm": 3.3957040309906006, | |
| "learning_rate": 2.8386572438162544e-05, | |
| "loss": 1.444, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 13.09, | |
| "eval_cer": 0.3994588596456924, | |
| "eval_loss": 2.732797861099243, | |
| "eval_runtime": 371.9235, | |
| "eval_samples_per_second": 25.484, | |
| "eval_steps_per_second": 3.186, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 13.13, | |
| "grad_norm": 2.130281448364258, | |
| "learning_rate": 2.8315901060070672e-05, | |
| "loss": 0.9865, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 13.13, | |
| "eval_cer": 0.4065005279418091, | |
| "eval_loss": 2.812870502471924, | |
| "eval_runtime": 395.591, | |
| "eval_samples_per_second": 23.959, | |
| "eval_steps_per_second": 2.996, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 13.18, | |
| "grad_norm": 3.105449676513672, | |
| "learning_rate": 2.82452296819788e-05, | |
| "loss": 0.9784, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 13.18, | |
| "eval_cer": 0.3964916311446561, | |
| "eval_loss": 2.1949923038482666, | |
| "eval_runtime": 377.7569, | |
| "eval_samples_per_second": 25.09, | |
| "eval_steps_per_second": 3.137, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "grad_norm": 3.4888200759887695, | |
| "learning_rate": 2.817455830388693e-05, | |
| "loss": 0.9847, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "eval_cer": 0.39935864846896874, | |
| "eval_loss": 2.095374584197998, | |
| "eval_runtime": 395.2097, | |
| "eval_samples_per_second": 23.982, | |
| "eval_steps_per_second": 2.998, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 13.26, | |
| "grad_norm": 1.688668966293335, | |
| "learning_rate": 2.8103886925795053e-05, | |
| "loss": 1.0167, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.26, | |
| "eval_cer": 0.40794747956669664, | |
| "eval_loss": 1.7657544612884521, | |
| "eval_runtime": 377.384, | |
| "eval_samples_per_second": 25.115, | |
| "eval_steps_per_second": 3.14, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.3, | |
| "grad_norm": 2.9201929569244385, | |
| "learning_rate": 2.8033215547703178e-05, | |
| "loss": 0.9921, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 13.3, | |
| "eval_cer": 0.40619256188651204, | |
| "eval_loss": 1.6753425598144531, | |
| "eval_runtime": 390.1339, | |
| "eval_samples_per_second": 24.294, | |
| "eval_steps_per_second": 3.037, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 13.34, | |
| "grad_norm": 3.985402822494507, | |
| "learning_rate": 2.796254416961131e-05, | |
| "loss": 0.9951, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 13.34, | |
| "eval_cer": 0.4047969379375073, | |
| "eval_loss": 1.8533233404159546, | |
| "eval_runtime": 373.1578, | |
| "eval_samples_per_second": 25.399, | |
| "eval_steps_per_second": 3.176, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 13.39, | |
| "grad_norm": 2.613786458969116, | |
| "learning_rate": 2.7891872791519434e-05, | |
| "loss": 1.002, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 13.39, | |
| "eval_cer": 0.39678737632474287, | |
| "eval_loss": 2.017918109893799, | |
| "eval_runtime": 398.8604, | |
| "eval_samples_per_second": 23.763, | |
| "eval_steps_per_second": 2.971, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 13.43, | |
| "grad_norm": 5.32037878036499, | |
| "learning_rate": 2.7821201413427562e-05, | |
| "loss": 1.0059, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 13.43, | |
| "eval_cer": 0.40224277501857575, | |
| "eval_loss": 2.3004064559936523, | |
| "eval_runtime": 375.8974, | |
| "eval_samples_per_second": 25.214, | |
| "eval_steps_per_second": 3.152, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 13.47, | |
| "grad_norm": 1.963639736175537, | |
| "learning_rate": 2.7750530035335694e-05, | |
| "loss": 0.9866, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.47, | |
| "eval_cer": 0.4031495639591725, | |
| "eval_loss": 2.7950069904327393, | |
| "eval_runtime": 392.4567, | |
| "eval_samples_per_second": 24.15, | |
| "eval_steps_per_second": 3.019, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "grad_norm": 10.024706840515137, | |
| "learning_rate": 2.767985865724382e-05, | |
| "loss": 1.0195, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "eval_cer": 0.409443314692425, | |
| "eval_loss": 2.4231760501861572, | |
| "eval_runtime": 370.8807, | |
| "eval_samples_per_second": 25.555, | |
| "eval_steps_per_second": 3.195, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 13.56, | |
| "grad_norm": 4.829713344573975, | |
| "learning_rate": 2.7609187279151943e-05, | |
| "loss": 1.0437, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 13.56, | |
| "eval_cer": 0.4043276563294357, | |
| "eval_loss": 1.8479336500167847, | |
| "eval_runtime": 398.7479, | |
| "eval_samples_per_second": 23.769, | |
| "eval_steps_per_second": 2.972, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "grad_norm": 3.0680336952209473, | |
| "learning_rate": 2.7538515901060075e-05, | |
| "loss": 1.4222, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "eval_cer": 0.4040881271753158, | |
| "eval_loss": 2.828200101852417, | |
| "eval_runtime": 367.3731, | |
| "eval_samples_per_second": 25.799, | |
| "eval_steps_per_second": 3.226, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "grad_norm": 2.740971565246582, | |
| "learning_rate": 2.74678445229682e-05, | |
| "loss": 1.0443, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "eval_cer": 0.422089476359939, | |
| "eval_loss": 2.9092776775360107, | |
| "eval_runtime": 389.1697, | |
| "eval_samples_per_second": 24.354, | |
| "eval_steps_per_second": 3.045, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 13.68, | |
| "grad_norm": 7.612214088439941, | |
| "learning_rate": 2.7397173144876324e-05, | |
| "loss": 1.0561, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.68, | |
| "eval_cer": 0.4040587970748113, | |
| "eval_loss": 1.9982529878616333, | |
| "eval_runtime": 371.4689, | |
| "eval_samples_per_second": 25.515, | |
| "eval_steps_per_second": 3.19, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.72, | |
| "grad_norm": 2.935277223587036, | |
| "learning_rate": 2.7326501766784452e-05, | |
| "loss": 1.0508, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 13.72, | |
| "eval_cer": 0.40732665910601856, | |
| "eval_loss": 2.4883534908294678, | |
| "eval_runtime": 395.2638, | |
| "eval_samples_per_second": 23.979, | |
| "eval_steps_per_second": 2.998, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 13.77, | |
| "grad_norm": 2.7527854442596436, | |
| "learning_rate": 2.7255830388692584e-05, | |
| "loss": 1.0314, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 13.77, | |
| "eval_cer": 0.40216944976731456, | |
| "eval_loss": 2.4499011039733887, | |
| "eval_runtime": 369.8082, | |
| "eval_samples_per_second": 25.63, | |
| "eval_steps_per_second": 3.204, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 13.81, | |
| "grad_norm": 1.3860234022140503, | |
| "learning_rate": 2.718515901060071e-05, | |
| "loss": 1.7756, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 13.81, | |
| "eval_cer": 0.41578839310156035, | |
| "eval_loss": 2.565376043319702, | |
| "eval_runtime": 390.8183, | |
| "eval_samples_per_second": 24.252, | |
| "eval_steps_per_second": 3.032, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 13.85, | |
| "grad_norm": 9.211983680725098, | |
| "learning_rate": 2.7114487632508833e-05, | |
| "loss": 1.0164, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 13.85, | |
| "eval_cer": 0.39727621133315083, | |
| "eval_loss": 2.448713779449463, | |
| "eval_runtime": 372.957, | |
| "eval_samples_per_second": 25.413, | |
| "eval_steps_per_second": 3.177, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 13.89, | |
| "grad_norm": 73.3850326538086, | |
| "learning_rate": 2.7043816254416965e-05, | |
| "loss": 1.0211, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 13.89, | |
| "eval_cer": 0.405021802041375, | |
| "eval_loss": 2.4291627407073975, | |
| "eval_runtime": 390.776, | |
| "eval_samples_per_second": 24.254, | |
| "eval_steps_per_second": 3.032, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 13.93, | |
| "grad_norm": 2.177152395248413, | |
| "learning_rate": 2.697314487632509e-05, | |
| "loss": 1.2605, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 13.93, | |
| "eval_cer": 0.40312512220875213, | |
| "eval_loss": 2.450680732727051, | |
| "eval_runtime": 369.1148, | |
| "eval_samples_per_second": 25.678, | |
| "eval_steps_per_second": 3.21, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 13.98, | |
| "grad_norm": 2.1010019779205322, | |
| "learning_rate": 2.6902473498233218e-05, | |
| "loss": 0.9974, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 13.98, | |
| "eval_cer": 0.4032106683352235, | |
| "eval_loss": 2.416074514389038, | |
| "eval_runtime": 395.8532, | |
| "eval_samples_per_second": 23.943, | |
| "eval_steps_per_second": 2.994, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 5.9587812423706055, | |
| "learning_rate": 2.6831802120141342e-05, | |
| "loss": 0.9724, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "eval_cer": 0.3974839662117242, | |
| "eval_loss": 2.979396343231201, | |
| "eval_runtime": 370.8376, | |
| "eval_samples_per_second": 25.558, | |
| "eval_steps_per_second": 3.195, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 14.06, | |
| "grad_norm": 2.916482925415039, | |
| "learning_rate": 2.6761130742049474e-05, | |
| "loss": 0.9488, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 14.06, | |
| "eval_cer": 0.39579504125767473, | |
| "eval_loss": 2.687380075454712, | |
| "eval_runtime": 390.5164, | |
| "eval_samples_per_second": 24.27, | |
| "eval_steps_per_second": 3.034, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 14.1, | |
| "grad_norm": 5.404353618621826, | |
| "learning_rate": 2.66904593639576e-05, | |
| "loss": 0.9346, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.1, | |
| "eval_cer": 0.39129775918032145, | |
| "eval_loss": 2.3900837898254395, | |
| "eval_runtime": 372.0348, | |
| "eval_samples_per_second": 25.476, | |
| "eval_steps_per_second": 3.185, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.14, | |
| "grad_norm": 2.114933729171753, | |
| "learning_rate": 2.6619787985865723e-05, | |
| "loss": 0.9445, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 14.14, | |
| "eval_cer": 0.3915470650346095, | |
| "eval_loss": 2.192039728164673, | |
| "eval_runtime": 401.3991, | |
| "eval_samples_per_second": 23.612, | |
| "eval_steps_per_second": 2.952, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 14.19, | |
| "grad_norm": 22.767732620239258, | |
| "learning_rate": 2.6549116607773855e-05, | |
| "loss": 0.9435, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 14.19, | |
| "eval_cer": 0.39616655586406474, | |
| "eval_loss": 2.0086452960968018, | |
| "eval_runtime": 374.9441, | |
| "eval_samples_per_second": 25.278, | |
| "eval_steps_per_second": 3.16, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 14.23, | |
| "grad_norm": 2.4856455326080322, | |
| "learning_rate": 2.647844522968198e-05, | |
| "loss": 0.9387, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 14.23, | |
| "eval_cer": 0.3982514371749247, | |
| "eval_loss": 1.5301440954208374, | |
| "eval_runtime": 391.5929, | |
| "eval_samples_per_second": 24.204, | |
| "eval_steps_per_second": 3.026, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 14.27, | |
| "grad_norm": 21.76763916015625, | |
| "learning_rate": 2.6407773851590107e-05, | |
| "loss": 0.9334, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 14.27, | |
| "eval_cer": 0.3909311329240155, | |
| "eval_loss": 2.4011144638061523, | |
| "eval_runtime": 373.6109, | |
| "eval_samples_per_second": 25.369, | |
| "eval_steps_per_second": 3.172, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 14.31, | |
| "grad_norm": 6.195644855499268, | |
| "learning_rate": 2.6337102473498232e-05, | |
| "loss": 0.9263, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.31, | |
| "eval_cer": 0.39136375190645656, | |
| "eval_loss": 2.636327028274536, | |
| "eval_runtime": 397.3915, | |
| "eval_samples_per_second": 23.851, | |
| "eval_steps_per_second": 2.982, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.35, | |
| "grad_norm": 8.33462905883789, | |
| "learning_rate": 2.6266431095406364e-05, | |
| "loss": 1.2049, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 14.35, | |
| "eval_cer": 0.38994857455711546, | |
| "eval_loss": 1.6413642168045044, | |
| "eval_runtime": 373.2342, | |
| "eval_samples_per_second": 25.394, | |
| "eval_steps_per_second": 3.175, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 2.263068675994873, | |
| "learning_rate": 2.619575971731449e-05, | |
| "loss": 0.9314, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "eval_cer": 0.3896846036525752, | |
| "eval_loss": 1.7732062339782715, | |
| "eval_runtime": 392.0357, | |
| "eval_samples_per_second": 24.176, | |
| "eval_steps_per_second": 3.023, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "grad_norm": 3.9732091426849365, | |
| "learning_rate": 2.6125088339222613e-05, | |
| "loss": 0.9228, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "eval_cer": 0.3950935630206093, | |
| "eval_loss": 1.7944254875183105, | |
| "eval_runtime": 369.9816, | |
| "eval_samples_per_second": 25.617, | |
| "eval_steps_per_second": 3.203, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "grad_norm": 1.9228265285491943, | |
| "learning_rate": 2.6054416961130745e-05, | |
| "loss": 0.9234, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "eval_cer": 0.39062316686871845, | |
| "eval_loss": 1.7525721788406372, | |
| "eval_runtime": 386.1261, | |
| "eval_samples_per_second": 24.546, | |
| "eval_steps_per_second": 3.069, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 14.52, | |
| "grad_norm": 10.801593780517578, | |
| "learning_rate": 2.598374558303887e-05, | |
| "loss": 0.9669, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.52, | |
| "eval_cer": 0.3913124242305737, | |
| "eval_loss": 1.5855803489685059, | |
| "eval_runtime": 376.3532, | |
| "eval_samples_per_second": 25.184, | |
| "eval_steps_per_second": 3.149, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.57, | |
| "grad_norm": 2.1704392433166504, | |
| "learning_rate": 2.5913074204946997e-05, | |
| "loss": 0.9261, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 14.57, | |
| "eval_cer": 0.38692268585507017, | |
| "eval_loss": 1.6806480884552002, | |
| "eval_runtime": 393.3479, | |
| "eval_samples_per_second": 24.096, | |
| "eval_steps_per_second": 3.013, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "grad_norm": 4.049818515777588, | |
| "learning_rate": 2.5842402826855122e-05, | |
| "loss": 0.9017, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "eval_cer": 0.3871744358844003, | |
| "eval_loss": 1.6744087934494019, | |
| "eval_runtime": 365.5955, | |
| "eval_samples_per_second": 25.925, | |
| "eval_steps_per_second": 3.241, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 14.65, | |
| "grad_norm": 3.3934333324432373, | |
| "learning_rate": 2.5771731448763254e-05, | |
| "loss": 0.9375, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 14.65, | |
| "eval_cer": 0.3903005357631692, | |
| "eval_loss": 1.944000005722046, | |
| "eval_runtime": 404.5333, | |
| "eval_samples_per_second": 23.429, | |
| "eval_steps_per_second": 2.929, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 14.69, | |
| "grad_norm": 2.0570108890533447, | |
| "learning_rate": 2.5701060070671378e-05, | |
| "loss": 0.9158, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 14.69, | |
| "eval_cer": 0.3864998435727973, | |
| "eval_loss": 1.8147820234298706, | |
| "eval_runtime": 366.6552, | |
| "eval_samples_per_second": 25.85, | |
| "eval_steps_per_second": 3.232, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 14.73, | |
| "grad_norm": 2.4123330116271973, | |
| "learning_rate": 2.5631095406360423e-05, | |
| "loss": 1.0998, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.73, | |
| "eval_cer": 0.3869789018810371, | |
| "eval_loss": 1.8494433164596558, | |
| "eval_runtime": 395.7827, | |
| "eval_samples_per_second": 23.947, | |
| "eval_steps_per_second": 2.994, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "grad_norm": 2.7764666080474854, | |
| "learning_rate": 2.5560424028268555e-05, | |
| "loss": 0.9243, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "eval_cer": 0.38634097219506475, | |
| "eval_loss": 1.8553264141082764, | |
| "eval_runtime": 368.5207, | |
| "eval_samples_per_second": 25.719, | |
| "eval_steps_per_second": 3.216, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 14.82, | |
| "grad_norm": 37.1422119140625, | |
| "learning_rate": 2.548975265017668e-05, | |
| "loss": 0.9169, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 14.82, | |
| "eval_cer": 0.3885431739079426, | |
| "eval_loss": 2.9854750633239746, | |
| "eval_runtime": 393.2259, | |
| "eval_samples_per_second": 24.103, | |
| "eval_steps_per_second": 3.014, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 14.86, | |
| "grad_norm": 2.0946056842803955, | |
| "learning_rate": 2.5419081272084804e-05, | |
| "loss": 0.9188, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 14.86, | |
| "eval_cer": 0.38827187047827616, | |
| "eval_loss": 2.4621119499206543, | |
| "eval_runtime": 369.124, | |
| "eval_samples_per_second": 25.677, | |
| "eval_steps_per_second": 3.21, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "grad_norm": 2.0357742309570312, | |
| "learning_rate": 2.5348409893992936e-05, | |
| "loss": 0.9183, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "eval_cer": 0.38374036995033434, | |
| "eval_loss": 1.656967043876648, | |
| "eval_runtime": 397.837, | |
| "eval_samples_per_second": 23.824, | |
| "eval_steps_per_second": 2.979, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 14.94, | |
| "grad_norm": 2.4356577396392822, | |
| "learning_rate": 2.527773851590106e-05, | |
| "loss": 0.9208, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 14.94, | |
| "eval_cer": 0.3846764889914356, | |
| "eval_loss": 1.6596330404281616, | |
| "eval_runtime": 368.5045, | |
| "eval_samples_per_second": 25.72, | |
| "eval_steps_per_second": 3.216, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 14.99, | |
| "grad_norm": 3.9734749794006348, | |
| "learning_rate": 2.520706713780919e-05, | |
| "loss": 0.9122, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 14.99, | |
| "eval_cer": 0.3868713581791874, | |
| "eval_loss": 2.068138360977173, | |
| "eval_runtime": 401.8562, | |
| "eval_samples_per_second": 23.586, | |
| "eval_steps_per_second": 2.949, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 1.6432957649230957, | |
| "learning_rate": 2.513639575971732e-05, | |
| "loss": 0.872, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "eval_cer": 0.3834226271948692, | |
| "eval_loss": 2.0865533351898193, | |
| "eval_runtime": 372.0301, | |
| "eval_samples_per_second": 25.476, | |
| "eval_steps_per_second": 3.185, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 15.07, | |
| "grad_norm": 2.7166266441345215, | |
| "learning_rate": 2.5065724381625445e-05, | |
| "loss": 0.8238, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 15.07, | |
| "eval_cer": 0.38250361737906224, | |
| "eval_loss": 2.33854079246521, | |
| "eval_runtime": 394.007, | |
| "eval_samples_per_second": 24.055, | |
| "eval_steps_per_second": 3.008, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 15.11, | |
| "grad_norm": 5.8333282470703125, | |
| "learning_rate": 2.499505300353357e-05, | |
| "loss": 0.8495, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 15.11, | |
| "eval_cer": 0.38729908881154435, | |
| "eval_loss": 1.496200680732727, | |
| "eval_runtime": 379.9915, | |
| "eval_samples_per_second": 24.943, | |
| "eval_steps_per_second": 3.118, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 15.15, | |
| "grad_norm": 8.684925079345703, | |
| "learning_rate": 2.4924381625441698e-05, | |
| "loss": 0.8427, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.15, | |
| "eval_cer": 0.3852386492511048, | |
| "eval_loss": 2.0919713973999023, | |
| "eval_runtime": 387.8899, | |
| "eval_samples_per_second": 24.435, | |
| "eval_steps_per_second": 3.055, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "grad_norm": 2.094461441040039, | |
| "learning_rate": 2.4853710247349822e-05, | |
| "loss": 0.8353, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "eval_cer": 0.38358149857260176, | |
| "eval_loss": 1.544800043106079, | |
| "eval_runtime": 374.0519, | |
| "eval_samples_per_second": 25.339, | |
| "eval_steps_per_second": 3.168, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 15.24, | |
| "grad_norm": 7.441462516784668, | |
| "learning_rate": 2.478303886925795e-05, | |
| "loss": 0.8271, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 15.24, | |
| "eval_cer": 0.3848451370693364, | |
| "eval_loss": 2.453676700592041, | |
| "eval_runtime": 392.8332, | |
| "eval_samples_per_second": 24.127, | |
| "eval_steps_per_second": 3.017, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 15.28, | |
| "grad_norm": 1.4820858240127563, | |
| "learning_rate": 2.4712367491166082e-05, | |
| "loss": 0.8265, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 15.28, | |
| "eval_cer": 0.3837550350005866, | |
| "eval_loss": 2.4253969192504883, | |
| "eval_runtime": 373.2352, | |
| "eval_samples_per_second": 25.394, | |
| "eval_steps_per_second": 3.175, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 15.32, | |
| "grad_norm": 4.485095024108887, | |
| "learning_rate": 2.4641696113074207e-05, | |
| "loss": 0.8412, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 15.32, | |
| "eval_cer": 0.3884747370067655, | |
| "eval_loss": 2.459603786468506, | |
| "eval_runtime": 390.2769, | |
| "eval_samples_per_second": 24.285, | |
| "eval_steps_per_second": 3.036, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 15.37, | |
| "grad_norm": 7.8216471672058105, | |
| "learning_rate": 2.4571024734982335e-05, | |
| "loss": 0.8567, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.37, | |
| "eval_cer": 0.3837354816002503, | |
| "eval_loss": 3.254591226577759, | |
| "eval_runtime": 377.582, | |
| "eval_samples_per_second": 25.102, | |
| "eval_steps_per_second": 3.138, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.41, | |
| "grad_norm": 2.479659080505371, | |
| "learning_rate": 2.450035335689046e-05, | |
| "loss": 0.8539, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 15.41, | |
| "eval_cer": 0.3853241953775762, | |
| "eval_loss": 1.766408085823059, | |
| "eval_runtime": 396.6796, | |
| "eval_samples_per_second": 23.893, | |
| "eval_steps_per_second": 2.987, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 15.45, | |
| "grad_norm": 3.3775389194488525, | |
| "learning_rate": 2.4429681978798587e-05, | |
| "loss": 0.8428, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 15.45, | |
| "eval_cer": 0.3814917289116577, | |
| "eval_loss": 1.7772597074508667, | |
| "eval_runtime": 373.2028, | |
| "eval_samples_per_second": 25.396, | |
| "eval_steps_per_second": 3.175, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 15.49, | |
| "grad_norm": 1.7568877935409546, | |
| "learning_rate": 2.4359010600706712e-05, | |
| "loss": 0.8502, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 15.49, | |
| "eval_cer": 0.3838430253021, | |
| "eval_loss": 2.477693796157837, | |
| "eval_runtime": 393.1117, | |
| "eval_samples_per_second": 24.11, | |
| "eval_steps_per_second": 3.014, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 15.53, | |
| "grad_norm": 1.5188319683074951, | |
| "learning_rate": 2.4288339222614844e-05, | |
| "loss": 1.1122, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 15.53, | |
| "eval_cer": 0.3783705173829729, | |
| "eval_loss": 2.0704104900360107, | |
| "eval_runtime": 376.2486, | |
| "eval_samples_per_second": 25.191, | |
| "eval_steps_per_second": 3.15, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 15.58, | |
| "grad_norm": 5.990358352661133, | |
| "learning_rate": 2.4217667844522972e-05, | |
| "loss": 0.8376, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.58, | |
| "eval_cer": 0.38022809041492317, | |
| "eval_loss": 1.795945405960083, | |
| "eval_runtime": 393.6062, | |
| "eval_samples_per_second": 24.08, | |
| "eval_steps_per_second": 3.011, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "grad_norm": 37.267982482910156, | |
| "learning_rate": 2.4146996466431096e-05, | |
| "loss": 0.8422, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "eval_cer": 0.38063137929685975, | |
| "eval_loss": 2.6164886951446533, | |
| "eval_runtime": 378.7126, | |
| "eval_samples_per_second": 25.027, | |
| "eval_steps_per_second": 3.129, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 15.66, | |
| "grad_norm": 51.162017822265625, | |
| "learning_rate": 2.4076325088339225e-05, | |
| "loss": 0.8849, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 15.66, | |
| "eval_cer": 0.38121553713190726, | |
| "eval_loss": 1.649156928062439, | |
| "eval_runtime": 393.043, | |
| "eval_samples_per_second": 24.114, | |
| "eval_steps_per_second": 3.015, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 15.7, | |
| "grad_norm": 1.459133505821228, | |
| "learning_rate": 2.400565371024735e-05, | |
| "loss": 0.8247, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 15.7, | |
| "eval_cer": 0.37826297368112316, | |
| "eval_loss": 1.726231575012207, | |
| "eval_runtime": 372.5489, | |
| "eval_samples_per_second": 25.441, | |
| "eval_steps_per_second": 3.181, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 15.74, | |
| "grad_norm": 1.8703219890594482, | |
| "learning_rate": 2.3934982332155477e-05, | |
| "loss": 0.8085, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 15.74, | |
| "eval_cer": 0.38297778733721793, | |
| "eval_loss": 2.4081192016601562, | |
| "eval_runtime": 395.9622, | |
| "eval_samples_per_second": 23.937, | |
| "eval_steps_per_second": 2.993, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 15.79, | |
| "grad_norm": 1.6956226825714111, | |
| "learning_rate": 2.3864310954063605e-05, | |
| "loss": 0.8194, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.79, | |
| "eval_cer": 0.3802305345899652, | |
| "eval_loss": 2.873920202255249, | |
| "eval_runtime": 373.9204, | |
| "eval_samples_per_second": 25.348, | |
| "eval_steps_per_second": 3.169, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "grad_norm": 10.501407623291016, | |
| "learning_rate": 2.3793639575971734e-05, | |
| "loss": 0.843, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "eval_cer": 0.37860271401196666, | |
| "eval_loss": 1.4808061122894287, | |
| "eval_runtime": 394.8535, | |
| "eval_samples_per_second": 24.004, | |
| "eval_steps_per_second": 3.001, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 15.87, | |
| "grad_norm": 4.368974208831787, | |
| "learning_rate": 2.372296819787986e-05, | |
| "loss": 0.824, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 15.87, | |
| "eval_cer": 0.3751050995268077, | |
| "eval_loss": 1.6123182773590088, | |
| "eval_runtime": 373.2585, | |
| "eval_samples_per_second": 25.393, | |
| "eval_steps_per_second": 3.175, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 15.91, | |
| "grad_norm": 17.767349243164062, | |
| "learning_rate": 2.3652296819787986e-05, | |
| "loss": 0.8418, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 15.91, | |
| "eval_cer": 0.37160748504164876, | |
| "eval_loss": 1.7542308568954468, | |
| "eval_runtime": 389.6677, | |
| "eval_samples_per_second": 24.323, | |
| "eval_steps_per_second": 3.041, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 15.95, | |
| "grad_norm": 6.326321601867676, | |
| "learning_rate": 2.3581625441696114e-05, | |
| "loss": 0.8193, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 15.95, | |
| "eval_cer": 0.3725704900082124, | |
| "eval_loss": 2.059025764465332, | |
| "eval_runtime": 377.283, | |
| "eval_samples_per_second": 25.122, | |
| "eval_steps_per_second": 3.141, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 2.2139954566955566, | |
| "learning_rate": 2.351095406360424e-05, | |
| "loss": 0.8378, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.37425452661217784, | |
| "eval_loss": 1.9064280986785889, | |
| "eval_runtime": 411.9001, | |
| "eval_samples_per_second": 23.01, | |
| "eval_steps_per_second": 2.877, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.04, | |
| "grad_norm": 2.1273019313812256, | |
| "learning_rate": 2.3440282685512367e-05, | |
| "loss": 0.7478, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 16.04, | |
| "eval_cer": 0.3710159946814751, | |
| "eval_loss": 1.7977242469787598, | |
| "eval_runtime": 380.1262, | |
| "eval_samples_per_second": 24.934, | |
| "eval_steps_per_second": 3.117, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 16.08, | |
| "grad_norm": 3.238290786743164, | |
| "learning_rate": 2.3369611307420495e-05, | |
| "loss": 0.738, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 16.08, | |
| "eval_cer": 0.37189589769660947, | |
| "eval_loss": 1.8234237432479858, | |
| "eval_runtime": 391.4385, | |
| "eval_samples_per_second": 24.213, | |
| "eval_steps_per_second": 3.027, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 16.12, | |
| "grad_norm": 3.291322708129883, | |
| "learning_rate": 2.3298939929328624e-05, | |
| "loss": 0.7327, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 16.12, | |
| "eval_cer": 0.3755817136600055, | |
| "eval_loss": 1.935049057006836, | |
| "eval_runtime": 376.3761, | |
| "eval_samples_per_second": 25.182, | |
| "eval_steps_per_second": 3.148, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 16.17, | |
| "grad_norm": 3.3271946907043457, | |
| "learning_rate": 2.322826855123675e-05, | |
| "loss": 0.7428, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 16.17, | |
| "eval_cer": 0.37296644636502285, | |
| "eval_loss": 1.7951716184616089, | |
| "eval_runtime": 401.591, | |
| "eval_samples_per_second": 23.601, | |
| "eval_steps_per_second": 2.951, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 16.21, | |
| "grad_norm": 2.013002395629883, | |
| "learning_rate": 2.3157597173144876e-05, | |
| "loss": 0.7499, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.21, | |
| "eval_cer": 0.37256560165812835, | |
| "eval_loss": 1.7327371835708618, | |
| "eval_runtime": 380.9473, | |
| "eval_samples_per_second": 24.88, | |
| "eval_steps_per_second": 3.111, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "grad_norm": 3.8447253704071045, | |
| "learning_rate": 2.3086925795053004e-05, | |
| "loss": 0.7266, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "eval_cer": 0.37181523992022214, | |
| "eval_loss": 2.117140293121338, | |
| "eval_runtime": 398.2953, | |
| "eval_samples_per_second": 23.796, | |
| "eval_steps_per_second": 2.975, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 16.29, | |
| "grad_norm": 2.052280902862549, | |
| "learning_rate": 2.301625441696113e-05, | |
| "loss": 0.737, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 16.29, | |
| "eval_cer": 0.37181768409526417, | |
| "eval_loss": 2.6678144931793213, | |
| "eval_runtime": 376.034, | |
| "eval_samples_per_second": 25.205, | |
| "eval_steps_per_second": 3.151, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "grad_norm": 1.934323787689209, | |
| "learning_rate": 2.2945583038869257e-05, | |
| "loss": 0.7233, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "eval_cer": 0.3704929412224786, | |
| "eval_loss": 2.179749011993408, | |
| "eval_runtime": 389.749, | |
| "eval_samples_per_second": 24.318, | |
| "eval_steps_per_second": 3.04, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 16.38, | |
| "grad_norm": 2.3759796619415283, | |
| "learning_rate": 2.287491166077739e-05, | |
| "loss": 0.743, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 16.38, | |
| "eval_cer": 0.3723774001798913, | |
| "eval_loss": 2.006964683532715, | |
| "eval_runtime": 381.2342, | |
| "eval_samples_per_second": 24.861, | |
| "eval_steps_per_second": 3.108, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 16.42, | |
| "grad_norm": 1.9076517820358276, | |
| "learning_rate": 2.2804240282685513e-05, | |
| "loss": 0.7521, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.42, | |
| "eval_cer": 0.3758725704900082, | |
| "eval_loss": 1.847259759902954, | |
| "eval_runtime": 394.9054, | |
| "eval_samples_per_second": 24.001, | |
| "eval_steps_per_second": 3.001, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "grad_norm": 7.669488906860352, | |
| "learning_rate": 2.273356890459364e-05, | |
| "loss": 0.7523, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "eval_cer": 0.3722576356028313, | |
| "eval_loss": 1.65703284740448, | |
| "eval_runtime": 378.1244, | |
| "eval_samples_per_second": 25.066, | |
| "eval_steps_per_second": 3.134, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "grad_norm": 2.213913679122925, | |
| "learning_rate": 2.2662897526501766e-05, | |
| "loss": 0.7423, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "eval_cer": 0.3694517226545696, | |
| "eval_loss": 1.7742842435836792, | |
| "eval_runtime": 396.973, | |
| "eval_samples_per_second": 23.876, | |
| "eval_steps_per_second": 2.985, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 16.54, | |
| "grad_norm": 3.6008243560791016, | |
| "learning_rate": 2.2592226148409894e-05, | |
| "loss": 0.731, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 16.54, | |
| "eval_cer": 0.36915842164952484, | |
| "eval_loss": 1.7430915832519531, | |
| "eval_runtime": 370.2202, | |
| "eval_samples_per_second": 25.601, | |
| "eval_steps_per_second": 3.201, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 16.59, | |
| "grad_norm": 2.914562463760376, | |
| "learning_rate": 2.2521554770318022e-05, | |
| "loss": 0.7409, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 16.59, | |
| "eval_cer": 0.36913153572406243, | |
| "eval_loss": 1.8427553176879883, | |
| "eval_runtime": 393.2137, | |
| "eval_samples_per_second": 24.104, | |
| "eval_steps_per_second": 3.014, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 16.63, | |
| "grad_norm": 4.166165828704834, | |
| "learning_rate": 2.245088339222615e-05, | |
| "loss": 0.7502, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.63, | |
| "eval_cer": 0.366770462633452, | |
| "eval_loss": 1.7825747728347778, | |
| "eval_runtime": 387.4051, | |
| "eval_samples_per_second": 24.465, | |
| "eval_steps_per_second": 3.059, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "grad_norm": 4.028792858123779, | |
| "learning_rate": 2.238021201413428e-05, | |
| "loss": 0.738, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "eval_cer": 0.36522085565679874, | |
| "eval_loss": 1.9882181882858276, | |
| "eval_runtime": 394.101, | |
| "eval_samples_per_second": 24.05, | |
| "eval_steps_per_second": 3.007, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 16.71, | |
| "grad_norm": 2.6417016983032227, | |
| "learning_rate": 2.2309540636042403e-05, | |
| "loss": 0.7357, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 16.71, | |
| "eval_cer": 0.3660787610965547, | |
| "eval_loss": 2.067323923110962, | |
| "eval_runtime": 386.7519, | |
| "eval_samples_per_second": 24.507, | |
| "eval_steps_per_second": 3.064, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 16.75, | |
| "grad_norm": 3.9978723526000977, | |
| "learning_rate": 2.223886925795053e-05, | |
| "loss": 0.7286, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 16.75, | |
| "eval_cer": 0.3679216690782527, | |
| "eval_loss": 1.6313321590423584, | |
| "eval_runtime": 398.1769, | |
| "eval_samples_per_second": 23.803, | |
| "eval_steps_per_second": 2.976, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "grad_norm": 1.9248440265655518, | |
| "learning_rate": 2.2168197879858656e-05, | |
| "loss": 0.7211, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "eval_cer": 0.365851452817645, | |
| "eval_loss": 1.687284231185913, | |
| "eval_runtime": 376.8233, | |
| "eval_samples_per_second": 25.152, | |
| "eval_steps_per_second": 3.145, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 16.84, | |
| "grad_norm": 10.867044448852539, | |
| "learning_rate": 2.2097526501766784e-05, | |
| "loss": 0.7286, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.84, | |
| "eval_cer": 0.36576346251613157, | |
| "eval_loss": 2.3375942707061768, | |
| "eval_runtime": 391.3944, | |
| "eval_samples_per_second": 24.216, | |
| "eval_steps_per_second": 3.028, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.88, | |
| "grad_norm": 2.3577775955200195, | |
| "learning_rate": 2.2026855123674912e-05, | |
| "loss": 0.7599, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 16.88, | |
| "eval_cer": 0.36309197919518205, | |
| "eval_loss": 1.8121538162231445, | |
| "eval_runtime": 378.3056, | |
| "eval_samples_per_second": 25.054, | |
| "eval_steps_per_second": 3.132, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 16.92, | |
| "grad_norm": 3.6556479930877686, | |
| "learning_rate": 2.1956890459363957e-05, | |
| "loss": 1.0403, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 16.92, | |
| "eval_cer": 0.3646293652966251, | |
| "eval_loss": 2.4574332237243652, | |
| "eval_runtime": 392.3067, | |
| "eval_samples_per_second": 24.16, | |
| "eval_steps_per_second": 3.021, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 16.96, | |
| "grad_norm": 1.9411290884017944, | |
| "learning_rate": 2.1886219081272085e-05, | |
| "loss": 0.7538, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 16.96, | |
| "eval_cer": 0.36607387274647063, | |
| "eval_loss": 1.7143244743347168, | |
| "eval_runtime": 378.8409, | |
| "eval_samples_per_second": 25.018, | |
| "eval_steps_per_second": 3.128, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 7.614956378936768, | |
| "learning_rate": 2.1815547703180214e-05, | |
| "loss": 0.7241, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "eval_cer": 0.3618430057486997, | |
| "eval_loss": 1.6417515277862549, | |
| "eval_runtime": 391.0073, | |
| "eval_samples_per_second": 24.24, | |
| "eval_steps_per_second": 3.031, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 17.05, | |
| "grad_norm": 1.9776512384414673, | |
| "learning_rate": 2.1744876325088338e-05, | |
| "loss": 0.6509, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.05, | |
| "eval_cer": 0.3647295764733487, | |
| "eval_loss": 2.075551986694336, | |
| "eval_runtime": 398.6714, | |
| "eval_samples_per_second": 23.774, | |
| "eval_steps_per_second": 2.972, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.09, | |
| "grad_norm": 3.229382276535034, | |
| "learning_rate": 2.167420494699647e-05, | |
| "loss": 0.6699, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 17.09, | |
| "eval_cer": 0.36611786789722733, | |
| "eval_loss": 1.8951997756958008, | |
| "eval_runtime": 377.7399, | |
| "eval_samples_per_second": 25.091, | |
| "eval_steps_per_second": 3.137, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 17.13, | |
| "grad_norm": 7.590035915374756, | |
| "learning_rate": 2.1603533568904594e-05, | |
| "loss": 0.6376, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 17.13, | |
| "eval_cer": 0.36229762230651913, | |
| "eval_loss": 2.3001914024353027, | |
| "eval_runtime": 393.9708, | |
| "eval_samples_per_second": 24.058, | |
| "eval_steps_per_second": 3.008, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 17.18, | |
| "grad_norm": 3.8723325729370117, | |
| "learning_rate": 2.1532862190812723e-05, | |
| "loss": 0.6424, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 17.18, | |
| "eval_cer": 0.3627106878886238, | |
| "eval_loss": 1.9817591905593872, | |
| "eval_runtime": 380.3308, | |
| "eval_samples_per_second": 24.92, | |
| "eval_steps_per_second": 3.116, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 17.22, | |
| "grad_norm": 7.374648094177246, | |
| "learning_rate": 2.146219081272085e-05, | |
| "loss": 0.6245, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 17.22, | |
| "eval_cer": 0.36244916115912557, | |
| "eval_loss": 2.089902400970459, | |
| "eval_runtime": 399.0568, | |
| "eval_samples_per_second": 23.751, | |
| "eval_steps_per_second": 2.97, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 17.26, | |
| "grad_norm": 27.956933975219727, | |
| "learning_rate": 2.1391519434628975e-05, | |
| "loss": 0.6492, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.26, | |
| "eval_cer": 0.36416741621367954, | |
| "eval_loss": 2.015270948410034, | |
| "eval_runtime": 376.6268, | |
| "eval_samples_per_second": 25.165, | |
| "eval_steps_per_second": 3.146, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.3, | |
| "grad_norm": 5.671442031860352, | |
| "learning_rate": 2.1320848056537103e-05, | |
| "loss": 0.6544, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 17.3, | |
| "eval_cer": 0.3657659066911736, | |
| "eval_loss": 1.7072508335113525, | |
| "eval_runtime": 397.2129, | |
| "eval_samples_per_second": 23.861, | |
| "eval_steps_per_second": 2.983, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 17.34, | |
| "grad_norm": 1.6116443872451782, | |
| "learning_rate": 2.125017667844523e-05, | |
| "loss": 0.6304, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 17.34, | |
| "eval_cer": 0.36735950881858354, | |
| "eval_loss": 2.110800266265869, | |
| "eval_runtime": 379.2769, | |
| "eval_samples_per_second": 24.99, | |
| "eval_steps_per_second": 3.124, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "grad_norm": 2.9379818439483643, | |
| "learning_rate": 2.117950530035336e-05, | |
| "loss": 0.6426, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "eval_cer": 0.3644313871182199, | |
| "eval_loss": 2.184722900390625, | |
| "eval_runtime": 397.6084, | |
| "eval_samples_per_second": 23.838, | |
| "eval_steps_per_second": 2.98, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 17.43, | |
| "grad_norm": 2.8842365741729736, | |
| "learning_rate": 2.1108833922261484e-05, | |
| "loss": 0.6913, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 17.43, | |
| "eval_cer": 0.36425785069023503, | |
| "eval_loss": 1.9356818199157715, | |
| "eval_runtime": 381.9605, | |
| "eval_samples_per_second": 24.814, | |
| "eval_steps_per_second": 3.102, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 17.47, | |
| "grad_norm": 2.586341142654419, | |
| "learning_rate": 2.1038162544169613e-05, | |
| "loss": 0.6534, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.47, | |
| "eval_cer": 0.36276690391459077, | |
| "eval_loss": 1.7679296731948853, | |
| "eval_runtime": 404.4195, | |
| "eval_samples_per_second": 23.436, | |
| "eval_steps_per_second": 2.93, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.51, | |
| "grad_norm": 12.988226890563965, | |
| "learning_rate": 2.096749116607774e-05, | |
| "loss": 0.655, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 17.51, | |
| "eval_cer": 0.3648982245512495, | |
| "eval_loss": 1.7326639890670776, | |
| "eval_runtime": 382.6806, | |
| "eval_samples_per_second": 24.767, | |
| "eval_steps_per_second": 3.097, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 17.55, | |
| "grad_norm": 3.8716864585876465, | |
| "learning_rate": 2.0896819787985865e-05, | |
| "loss": 0.6464, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 17.55, | |
| "eval_cer": 0.36285245004106215, | |
| "eval_loss": 2.095261573791504, | |
| "eval_runtime": 400.7202, | |
| "eval_samples_per_second": 23.652, | |
| "eval_steps_per_second": 2.957, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "grad_norm": 1.966536521911621, | |
| "learning_rate": 2.0826148409893993e-05, | |
| "loss": 0.6675, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "eval_cer": 0.3588268937468226, | |
| "eval_loss": 1.8380521535873413, | |
| "eval_runtime": 382.7754, | |
| "eval_samples_per_second": 24.761, | |
| "eval_steps_per_second": 3.096, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 17.64, | |
| "grad_norm": 2.0896823406219482, | |
| "learning_rate": 2.075547703180212e-05, | |
| "loss": 0.6531, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 17.64, | |
| "eval_cer": 0.3638618943334246, | |
| "eval_loss": 1.911123514175415, | |
| "eval_runtime": 401.57, | |
| "eval_samples_per_second": 23.602, | |
| "eval_steps_per_second": 2.951, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 17.68, | |
| "grad_norm": 2.2857401371002197, | |
| "learning_rate": 2.068480565371025e-05, | |
| "loss": 0.6536, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.68, | |
| "eval_cer": 0.3620947557780298, | |
| "eval_loss": 1.524117112159729, | |
| "eval_runtime": 379.6794, | |
| "eval_samples_per_second": 24.963, | |
| "eval_steps_per_second": 3.121, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.72, | |
| "grad_norm": 3.275186777114868, | |
| "learning_rate": 2.0614134275618374e-05, | |
| "loss": 0.641, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 17.72, | |
| "eval_cer": 0.36079689883070665, | |
| "eval_loss": 1.8855453729629517, | |
| "eval_runtime": 405.5391, | |
| "eval_samples_per_second": 23.371, | |
| "eval_steps_per_second": 2.922, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 17.76, | |
| "grad_norm": 3.293827533721924, | |
| "learning_rate": 2.054416961130742e-05, | |
| "loss": 0.755, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 17.76, | |
| "eval_cer": 0.3628133432403895, | |
| "eval_loss": 2.0661118030548096, | |
| "eval_runtime": 378.6947, | |
| "eval_samples_per_second": 25.028, | |
| "eval_steps_per_second": 3.129, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 17.81, | |
| "grad_norm": 2.2286911010742188, | |
| "learning_rate": 2.0473498233215547e-05, | |
| "loss": 0.6528, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 17.81, | |
| "eval_cer": 0.3575412576747096, | |
| "eval_loss": 1.9768012762069702, | |
| "eval_runtime": 400.1546, | |
| "eval_samples_per_second": 23.686, | |
| "eval_steps_per_second": 2.961, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 17.85, | |
| "grad_norm": 1.9678279161453247, | |
| "learning_rate": 2.0402826855123676e-05, | |
| "loss": 0.641, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 17.85, | |
| "eval_cer": 0.3585751437174925, | |
| "eval_loss": 2.0739636421203613, | |
| "eval_runtime": 379.1375, | |
| "eval_samples_per_second": 24.999, | |
| "eval_steps_per_second": 3.126, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 17.89, | |
| "grad_norm": 4.927220344543457, | |
| "learning_rate": 2.0332155477031804e-05, | |
| "loss": 0.6499, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 17.89, | |
| "eval_cer": 0.3600440929177584, | |
| "eval_loss": 2.656733512878418, | |
| "eval_runtime": 405.5316, | |
| "eval_samples_per_second": 23.372, | |
| "eval_steps_per_second": 2.922, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 17.93, | |
| "grad_norm": 1.9719730615615845, | |
| "learning_rate": 2.0261484098939932e-05, | |
| "loss": 0.6453, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 17.93, | |
| "eval_cer": 0.3558694419459544, | |
| "eval_loss": 3.2607502937316895, | |
| "eval_runtime": 381.4619, | |
| "eval_samples_per_second": 24.847, | |
| "eval_steps_per_second": 3.106, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 17.98, | |
| "grad_norm": 4.013895034790039, | |
| "learning_rate": 2.0190812720848056e-05, | |
| "loss": 0.6373, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 17.98, | |
| "eval_cer": 0.3598925540651519, | |
| "eval_loss": 3.0558464527130127, | |
| "eval_runtime": 405.1328, | |
| "eval_samples_per_second": 23.395, | |
| "eval_steps_per_second": 2.925, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 5.639492988586426, | |
| "learning_rate": 2.0120141342756185e-05, | |
| "loss": 0.6082, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "eval_cer": 0.3542000703922412, | |
| "eval_loss": 2.6185314655303955, | |
| "eval_runtime": 382.1723, | |
| "eval_samples_per_second": 24.8, | |
| "eval_steps_per_second": 3.101, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 18.06, | |
| "grad_norm": 4.65023946762085, | |
| "learning_rate": 2.004946996466431e-05, | |
| "loss": 0.5527, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 18.06, | |
| "eval_cer": 0.35880245199640215, | |
| "eval_loss": 2.8290772438049316, | |
| "eval_runtime": 400.9741, | |
| "eval_samples_per_second": 23.637, | |
| "eval_steps_per_second": 2.955, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 18.1, | |
| "grad_norm": 1.5041050910949707, | |
| "learning_rate": 1.997879858657244e-05, | |
| "loss": 0.5502, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.1, | |
| "eval_cer": 0.3574239372726917, | |
| "eval_loss": 2.1965556144714355, | |
| "eval_runtime": 380.8445, | |
| "eval_samples_per_second": 24.887, | |
| "eval_steps_per_second": 3.112, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.14, | |
| "grad_norm": 2.088670015335083, | |
| "learning_rate": 1.990812720848057e-05, | |
| "loss": 0.5513, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 18.14, | |
| "eval_cer": 0.35743126979781786, | |
| "eval_loss": 3.069037914276123, | |
| "eval_runtime": 401.5683, | |
| "eval_samples_per_second": 23.602, | |
| "eval_steps_per_second": 2.951, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 18.19, | |
| "grad_norm": 13.482215881347656, | |
| "learning_rate": 1.9837455830388694e-05, | |
| "loss": 0.5517, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 18.19, | |
| "eval_cer": 0.35501398068124046, | |
| "eval_loss": 3.0275192260742188, | |
| "eval_runtime": 382.7019, | |
| "eval_samples_per_second": 24.766, | |
| "eval_steps_per_second": 3.096, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 18.23, | |
| "grad_norm": 2.318678140640259, | |
| "learning_rate": 1.976749116607774e-05, | |
| "loss": 0.7104, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 18.23, | |
| "eval_cer": 0.35606253177427555, | |
| "eval_loss": 3.2951183319091797, | |
| "eval_runtime": 399.8798, | |
| "eval_samples_per_second": 23.702, | |
| "eval_steps_per_second": 2.963, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 18.27, | |
| "grad_norm": 2.007765769958496, | |
| "learning_rate": 1.9696819787985867e-05, | |
| "loss": 0.551, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 18.27, | |
| "eval_cer": 0.35518996128426733, | |
| "eval_loss": 3.7015881538391113, | |
| "eval_runtime": 380.644, | |
| "eval_samples_per_second": 24.9, | |
| "eval_steps_per_second": 3.113, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 18.31, | |
| "grad_norm": 1.9974658489227295, | |
| "learning_rate": 1.9626148409893995e-05, | |
| "loss": 0.5745, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.31, | |
| "eval_cer": 0.35752903679949943, | |
| "eval_loss": 3.192854404449463, | |
| "eval_runtime": 397.0976, | |
| "eval_samples_per_second": 23.868, | |
| "eval_steps_per_second": 2.984, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.35, | |
| "grad_norm": 1.8913908004760742, | |
| "learning_rate": 1.9555477031802123e-05, | |
| "loss": 0.5569, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 18.35, | |
| "eval_cer": 0.35657092018301983, | |
| "eval_loss": 3.314192771911621, | |
| "eval_runtime": 382.2019, | |
| "eval_samples_per_second": 24.798, | |
| "eval_steps_per_second": 3.1, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "grad_norm": 5.3773956298828125, | |
| "learning_rate": 1.9484805653710248e-05, | |
| "loss": 0.5594, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "eval_cer": 0.3562653983027648, | |
| "eval_loss": 2.0783262252807617, | |
| "eval_runtime": 404.1105, | |
| "eval_samples_per_second": 23.454, | |
| "eval_steps_per_second": 2.932, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 18.44, | |
| "grad_norm": 1.7726545333862305, | |
| "learning_rate": 1.9414134275618376e-05, | |
| "loss": 0.5482, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 18.44, | |
| "eval_cer": 0.3563827187047828, | |
| "eval_loss": 2.1228716373443604, | |
| "eval_runtime": 379.1678, | |
| "eval_samples_per_second": 24.997, | |
| "eval_steps_per_second": 3.125, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 18.48, | |
| "grad_norm": 8.040356636047363, | |
| "learning_rate": 1.93434628975265e-05, | |
| "loss": 0.5427, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 18.48, | |
| "eval_cer": 0.3585995854679129, | |
| "eval_loss": 2.0755696296691895, | |
| "eval_runtime": 399.7267, | |
| "eval_samples_per_second": 23.711, | |
| "eval_steps_per_second": 2.965, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 18.52, | |
| "grad_norm": 7.321507930755615, | |
| "learning_rate": 1.927279151943463e-05, | |
| "loss": 0.5651, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.52, | |
| "eval_cer": 0.35462046849947204, | |
| "eval_loss": 2.0843379497528076, | |
| "eval_runtime": 380.2237, | |
| "eval_samples_per_second": 24.927, | |
| "eval_steps_per_second": 3.117, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.56, | |
| "grad_norm": 2.2667276859283447, | |
| "learning_rate": 1.9202120141342757e-05, | |
| "loss": 0.5645, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 18.56, | |
| "eval_cer": 0.35216651675726407, | |
| "eval_loss": 1.9949140548706055, | |
| "eval_runtime": 407.2675, | |
| "eval_samples_per_second": 23.272, | |
| "eval_steps_per_second": 2.91, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 18.61, | |
| "grad_norm": 2.5292446613311768, | |
| "learning_rate": 1.9131448763250885e-05, | |
| "loss": 0.5488, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 18.61, | |
| "eval_cer": 0.35419029369207305, | |
| "eval_loss": 1.832000732421875, | |
| "eval_runtime": 383.3497, | |
| "eval_samples_per_second": 24.724, | |
| "eval_steps_per_second": 3.091, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 18.65, | |
| "grad_norm": 4.608269691467285, | |
| "learning_rate": 1.9060777385159013e-05, | |
| "loss": 0.5694, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 18.65, | |
| "eval_cer": 0.3574825974737007, | |
| "eval_loss": 2.2215840816497803, | |
| "eval_runtime": 404.7052, | |
| "eval_samples_per_second": 23.42, | |
| "eval_steps_per_second": 2.928, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 18.69, | |
| "grad_norm": 3.4327402114868164, | |
| "learning_rate": 1.8990106007067137e-05, | |
| "loss": 0.5747, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 18.69, | |
| "eval_cer": 0.3554588205388917, | |
| "eval_loss": 2.0720789432525635, | |
| "eval_runtime": 385.2409, | |
| "eval_samples_per_second": 24.603, | |
| "eval_steps_per_second": 3.076, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 18.73, | |
| "grad_norm": 2.587009906768799, | |
| "learning_rate": 1.8919434628975266e-05, | |
| "loss": 0.5442, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.73, | |
| "eval_cer": 0.35546126471393374, | |
| "eval_loss": 2.185746192932129, | |
| "eval_runtime": 399.8878, | |
| "eval_samples_per_second": 23.702, | |
| "eval_steps_per_second": 2.963, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.77, | |
| "grad_norm": 2.142084836959839, | |
| "learning_rate": 1.8849469964664314e-05, | |
| "loss": 0.9587, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 18.77, | |
| "eval_cer": 0.353083082398029, | |
| "eval_loss": 2.363218307495117, | |
| "eval_runtime": 384.8343, | |
| "eval_samples_per_second": 24.629, | |
| "eval_steps_per_second": 3.079, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 18.82, | |
| "grad_norm": 8.04695987701416, | |
| "learning_rate": 1.877879858657244e-05, | |
| "loss": 0.5773, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 18.82, | |
| "eval_cer": 0.35235960658558524, | |
| "eval_loss": 2.419173002243042, | |
| "eval_runtime": 407.1529, | |
| "eval_samples_per_second": 23.279, | |
| "eval_steps_per_second": 2.91, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 18.86, | |
| "grad_norm": 2.3344297409057617, | |
| "learning_rate": 1.8708127208480567e-05, | |
| "loss": 0.658, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 18.86, | |
| "eval_cer": 0.3527897813929842, | |
| "eval_loss": 2.6321187019348145, | |
| "eval_runtime": 378.5146, | |
| "eval_samples_per_second": 25.04, | |
| "eval_steps_per_second": 3.131, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 18.9, | |
| "grad_norm": 1.822228193283081, | |
| "learning_rate": 1.8637455830388695e-05, | |
| "loss": 0.5733, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 18.9, | |
| "eval_cer": 0.34930194360799344, | |
| "eval_loss": 2.7781569957733154, | |
| "eval_runtime": 405.1046, | |
| "eval_samples_per_second": 23.396, | |
| "eval_steps_per_second": 2.925, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 18.94, | |
| "grad_norm": 15.387603759765625, | |
| "learning_rate": 1.856678445229682e-05, | |
| "loss": 0.5554, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 18.94, | |
| "eval_cer": 0.35281422314340466, | |
| "eval_loss": 2.2615067958831787, | |
| "eval_runtime": 386.1566, | |
| "eval_samples_per_second": 24.544, | |
| "eval_steps_per_second": 3.069, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "grad_norm": 2.5482892990112305, | |
| "learning_rate": 1.8496113074204948e-05, | |
| "loss": 0.5893, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "eval_cer": 0.3533568300027375, | |
| "eval_loss": 2.9341318607330322, | |
| "eval_runtime": 402.0749, | |
| "eval_samples_per_second": 23.573, | |
| "eval_steps_per_second": 2.947, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "grad_norm": 1.673799991607666, | |
| "learning_rate": 1.8425441696113076e-05, | |
| "loss": 0.5004, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "eval_cer": 0.35489666027922256, | |
| "eval_loss": 2.622223138809204, | |
| "eval_runtime": 462.6029, | |
| "eval_samples_per_second": 20.488, | |
| "eval_steps_per_second": 2.562, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 19.07, | |
| "grad_norm": 2.5973594188690186, | |
| "learning_rate": 1.8354770318021204e-05, | |
| "loss": 0.476, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 19.07, | |
| "eval_cer": 0.3551215243830902, | |
| "eval_loss": 2.0877623558044434, | |
| "eval_runtime": 371.4246, | |
| "eval_samples_per_second": 25.518, | |
| "eval_steps_per_second": 3.19, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 19.11, | |
| "grad_norm": 2.274106502532959, | |
| "learning_rate": 1.828409893992933e-05, | |
| "loss": 0.4914, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 19.11, | |
| "eval_cer": 0.35471090297602753, | |
| "eval_loss": 2.134805917739868, | |
| "eval_runtime": 380.1079, | |
| "eval_samples_per_second": 24.935, | |
| "eval_steps_per_second": 3.118, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 19.15, | |
| "grad_norm": 16.124771118164062, | |
| "learning_rate": 1.8213427561837457e-05, | |
| "loss": 1.2144, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.15, | |
| "eval_cer": 0.35294376442063274, | |
| "eval_loss": 1.9654364585876465, | |
| "eval_runtime": 365.9062, | |
| "eval_samples_per_second": 25.903, | |
| "eval_steps_per_second": 3.239, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "grad_norm": 5.378861427307129, | |
| "learning_rate": 1.8142756183745585e-05, | |
| "loss": 0.4807, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "eval_cer": 0.35123773024128896, | |
| "eval_loss": 2.7749032974243164, | |
| "eval_runtime": 380.1844, | |
| "eval_samples_per_second": 24.93, | |
| "eval_steps_per_second": 3.117, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 19.24, | |
| "grad_norm": 1.9357730150222778, | |
| "learning_rate": 1.807208480565371e-05, | |
| "loss": 0.4868, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 19.24, | |
| "eval_cer": 0.35421473544249343, | |
| "eval_loss": 2.5809061527252197, | |
| "eval_runtime": 368.1076, | |
| "eval_samples_per_second": 25.748, | |
| "eval_steps_per_second": 3.219, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 19.28, | |
| "grad_norm": 2.2689895629882812, | |
| "learning_rate": 1.8001413427561838e-05, | |
| "loss": 0.471, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 19.28, | |
| "eval_cer": 0.3532492863008877, | |
| "eval_loss": 2.3049328327178955, | |
| "eval_runtime": 377.068, | |
| "eval_samples_per_second": 25.136, | |
| "eval_steps_per_second": 3.143, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 19.32, | |
| "grad_norm": 6.475673198699951, | |
| "learning_rate": 1.7930742049469966e-05, | |
| "loss": 0.4834, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 19.32, | |
| "eval_cer": 0.3522642837589457, | |
| "eval_loss": 2.4670591354370117, | |
| "eval_runtime": 369.8781, | |
| "eval_samples_per_second": 25.625, | |
| "eval_steps_per_second": 3.204, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 19.36, | |
| "grad_norm": 2.3827292919158936, | |
| "learning_rate": 1.7860070671378094e-05, | |
| "loss": 0.471, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.36, | |
| "eval_cer": 0.3508588831097728, | |
| "eval_loss": 2.3617963790893555, | |
| "eval_runtime": 380.0492, | |
| "eval_samples_per_second": 24.939, | |
| "eval_steps_per_second": 3.118, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "grad_norm": 1.383456826210022, | |
| "learning_rate": 1.778939929328622e-05, | |
| "loss": 0.494, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "eval_cer": 0.3493679363341285, | |
| "eval_loss": 2.267214298248291, | |
| "eval_runtime": 370.4047, | |
| "eval_samples_per_second": 25.588, | |
| "eval_steps_per_second": 3.199, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 19.45, | |
| "grad_norm": 5.505711555480957, | |
| "learning_rate": 1.7718727915194347e-05, | |
| "loss": 0.4749, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 19.45, | |
| "eval_cer": 0.3519318759532283, | |
| "eval_loss": 3.048175573348999, | |
| "eval_runtime": 386.8635, | |
| "eval_samples_per_second": 24.5, | |
| "eval_steps_per_second": 3.063, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 19.49, | |
| "grad_norm": 11.937617301940918, | |
| "learning_rate": 1.7648056537102475e-05, | |
| "loss": 0.4895, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 19.49, | |
| "eval_cer": 0.3529168784951703, | |
| "eval_loss": 2.974360942840576, | |
| "eval_runtime": 369.5199, | |
| "eval_samples_per_second": 25.65, | |
| "eval_steps_per_second": 3.207, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 19.53, | |
| "grad_norm": 2.0051708221435547, | |
| "learning_rate": 1.75773851590106e-05, | |
| "loss": 0.4878, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 19.53, | |
| "eval_cer": 0.3492921669078253, | |
| "eval_loss": 3.078077793121338, | |
| "eval_runtime": 380.6797, | |
| "eval_samples_per_second": 24.898, | |
| "eval_steps_per_second": 3.113, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "grad_norm": 2.106623888015747, | |
| "learning_rate": 1.750671378091873e-05, | |
| "loss": 0.4886, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "eval_cer": 0.34946814751085215, | |
| "eval_loss": 3.198930263519287, | |
| "eval_runtime": 372.2046, | |
| "eval_samples_per_second": 25.464, | |
| "eval_steps_per_second": 3.184, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.62, | |
| "grad_norm": 2.0272178649902344, | |
| "learning_rate": 1.7436042402826856e-05, | |
| "loss": 0.4745, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 19.62, | |
| "eval_cer": 0.3511057447890188, | |
| "eval_loss": 3.3171424865722656, | |
| "eval_runtime": 381.183, | |
| "eval_samples_per_second": 24.865, | |
| "eval_steps_per_second": 3.109, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 19.66, | |
| "grad_norm": 4.2316412925720215, | |
| "learning_rate": 1.7365371024734984e-05, | |
| "loss": 0.4875, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 19.66, | |
| "eval_cer": 0.3484904774940362, | |
| "eval_loss": 2.8240554332733154, | |
| "eval_runtime": 368.9178, | |
| "eval_samples_per_second": 25.691, | |
| "eval_steps_per_second": 3.212, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 19.7, | |
| "grad_norm": 2.5890541076660156, | |
| "learning_rate": 1.7294699646643112e-05, | |
| "loss": 0.4757, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 19.7, | |
| "eval_cer": 0.3484415939931954, | |
| "eval_loss": 2.6541638374328613, | |
| "eval_runtime": 380.984, | |
| "eval_samples_per_second": 24.878, | |
| "eval_steps_per_second": 3.11, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 19.74, | |
| "grad_norm": 5.675697326660156, | |
| "learning_rate": 1.7224028268551237e-05, | |
| "loss": 0.506, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 19.74, | |
| "eval_cer": 0.3504775918032146, | |
| "eval_loss": 2.4880096912384033, | |
| "eval_runtime": 369.3344, | |
| "eval_samples_per_second": 25.662, | |
| "eval_steps_per_second": 3.208, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 19.79, | |
| "grad_norm": 2.5223989486694336, | |
| "learning_rate": 1.7153356890459365e-05, | |
| "loss": 0.4852, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.79, | |
| "eval_cer": 0.34921639748152206, | |
| "eval_loss": 2.3917760848999023, | |
| "eval_runtime": 383.3871, | |
| "eval_samples_per_second": 24.722, | |
| "eval_steps_per_second": 3.091, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.83, | |
| "grad_norm": 13.549793243408203, | |
| "learning_rate": 1.7082685512367493e-05, | |
| "loss": 0.4983, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 19.83, | |
| "eval_cer": 0.34820450901411754, | |
| "eval_loss": 2.331942319869995, | |
| "eval_runtime": 370.3076, | |
| "eval_samples_per_second": 25.595, | |
| "eval_steps_per_second": 3.2, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 19.87, | |
| "grad_norm": 1.6992497444152832, | |
| "learning_rate": 1.701201413427562e-05, | |
| "loss": 0.4712, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 19.87, | |
| "eval_cer": 0.34815073716319267, | |
| "eval_loss": 2.558990001678467, | |
| "eval_runtime": 412.1751, | |
| "eval_samples_per_second": 22.995, | |
| "eval_steps_per_second": 2.875, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 19.91, | |
| "grad_norm": 3.102717399597168, | |
| "learning_rate": 1.6941342756183746e-05, | |
| "loss": 0.4767, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 19.91, | |
| "eval_cer": 0.3449659770834148, | |
| "eval_loss": 2.632328748703003, | |
| "eval_runtime": 385.8283, | |
| "eval_samples_per_second": 24.565, | |
| "eval_steps_per_second": 3.071, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "grad_norm": 3.6561086177825928, | |
| "learning_rate": 1.6870671378091874e-05, | |
| "loss": 0.4914, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "eval_cer": 0.3475079191271362, | |
| "eval_loss": 2.747345447540283, | |
| "eval_runtime": 406.0174, | |
| "eval_samples_per_second": 23.344, | |
| "eval_steps_per_second": 2.919, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.65822696685791, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.4636, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.34717795549646085, | |
| "eval_loss": 3.2108983993530273, | |
| "eval_runtime": 391.3482, | |
| "eval_samples_per_second": 24.219, | |
| "eval_steps_per_second": 3.028, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.04, | |
| "grad_norm": 2.881988525390625, | |
| "learning_rate": 1.6729328621908126e-05, | |
| "loss": 0.4024, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 20.04, | |
| "eval_cer": 0.3465546908607407, | |
| "eval_loss": 2.992396831512451, | |
| "eval_runtime": 406.9581, | |
| "eval_samples_per_second": 23.29, | |
| "eval_steps_per_second": 2.912, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 20.08, | |
| "grad_norm": 2.220200777053833, | |
| "learning_rate": 1.6658657243816255e-05, | |
| "loss": 0.4103, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 20.08, | |
| "eval_cer": 0.3478818779085683, | |
| "eval_loss": 3.0730020999908447, | |
| "eval_runtime": 382.3538, | |
| "eval_samples_per_second": 24.789, | |
| "eval_steps_per_second": 3.099, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 20.12, | |
| "grad_norm": 2.015397071838379, | |
| "learning_rate": 1.6587985865724383e-05, | |
| "loss": 0.4211, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 20.12, | |
| "eval_cer": 0.3458312150482969, | |
| "eval_loss": 2.9953062534332275, | |
| "eval_runtime": 408.1484, | |
| "eval_samples_per_second": 23.222, | |
| "eval_steps_per_second": 2.903, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 20.16, | |
| "grad_norm": 7.887564182281494, | |
| "learning_rate": 1.651731448763251e-05, | |
| "loss": 0.415, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 20.16, | |
| "eval_cer": 0.3473783778499081, | |
| "eval_loss": 3.28759765625, | |
| "eval_runtime": 390.4889, | |
| "eval_samples_per_second": 24.272, | |
| "eval_steps_per_second": 3.035, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 20.21, | |
| "grad_norm": 2.610433578491211, | |
| "learning_rate": 1.644664310954064e-05, | |
| "loss": 0.4143, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.21, | |
| "eval_cer": 0.3463151617066208, | |
| "eval_loss": 2.6613733768463135, | |
| "eval_runtime": 407.7959, | |
| "eval_samples_per_second": 23.242, | |
| "eval_steps_per_second": 2.906, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "grad_norm": 2.2651526927948, | |
| "learning_rate": 1.637667844522968e-05, | |
| "loss": 0.5371, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "eval_cer": 0.3461636228540143, | |
| "eval_loss": 2.631626605987549, | |
| "eval_runtime": 388.4014, | |
| "eval_samples_per_second": 24.403, | |
| "eval_steps_per_second": 3.051, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 20.29, | |
| "grad_norm": 2.476924180984497, | |
| "learning_rate": 1.630600706713781e-05, | |
| "loss": 0.4173, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 20.29, | |
| "eval_cer": 0.3475372492276407, | |
| "eval_loss": 2.8535077571868896, | |
| "eval_runtime": 407.6688, | |
| "eval_samples_per_second": 23.249, | |
| "eval_steps_per_second": 2.907, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 20.33, | |
| "grad_norm": 2.1427981853485107, | |
| "learning_rate": 1.623533568904594e-05, | |
| "loss": 0.4149, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 20.33, | |
| "eval_cer": 0.3440518556176919, | |
| "eval_loss": 2.9584333896636963, | |
| "eval_runtime": 385.1561, | |
| "eval_samples_per_second": 24.608, | |
| "eval_steps_per_second": 3.077, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 20.37, | |
| "grad_norm": 6.375877857208252, | |
| "learning_rate": 1.6164664310954065e-05, | |
| "loss": 0.4005, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 20.37, | |
| "eval_cer": 0.3466133510617496, | |
| "eval_loss": 3.129476547241211, | |
| "eval_runtime": 408.2033, | |
| "eval_samples_per_second": 23.219, | |
| "eval_steps_per_second": 2.903, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 20.42, | |
| "grad_norm": 8.823423385620117, | |
| "learning_rate": 1.6093992932862193e-05, | |
| "loss": 0.3974, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.42, | |
| "eval_cer": 0.3472977200735208, | |
| "eval_loss": 2.675215482711792, | |
| "eval_runtime": 389.4912, | |
| "eval_samples_per_second": 24.334, | |
| "eval_steps_per_second": 3.042, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.46, | |
| "grad_norm": 4.963177680969238, | |
| "learning_rate": 1.6023321554770318e-05, | |
| "loss": 0.456, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 20.46, | |
| "eval_cer": 0.34791120800907277, | |
| "eval_loss": 2.5059995651245117, | |
| "eval_runtime": 407.1464, | |
| "eval_samples_per_second": 23.279, | |
| "eval_steps_per_second": 2.911, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 20.5, | |
| "grad_norm": 1.5969637632369995, | |
| "learning_rate": 1.5952650176678446e-05, | |
| "loss": 0.3946, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 20.5, | |
| "eval_cer": 0.34289331664776507, | |
| "eval_loss": 3.0652801990509033, | |
| "eval_runtime": 388.6588, | |
| "eval_samples_per_second": 24.386, | |
| "eval_steps_per_second": 3.049, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 20.54, | |
| "grad_norm": 5.763105392456055, | |
| "learning_rate": 1.588197879858657e-05, | |
| "loss": 0.4039, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 20.54, | |
| "eval_cer": 0.34577988737241405, | |
| "eval_loss": 2.2274067401885986, | |
| "eval_runtime": 406.2895, | |
| "eval_samples_per_second": 23.328, | |
| "eval_steps_per_second": 2.917, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 20.59, | |
| "grad_norm": 4.299105167388916, | |
| "learning_rate": 1.5811307420494702e-05, | |
| "loss": 0.4253, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 20.59, | |
| "eval_cer": 0.3451859528371984, | |
| "eval_loss": 3.0053601264953613, | |
| "eval_runtime": 386.4818, | |
| "eval_samples_per_second": 24.524, | |
| "eval_steps_per_second": 3.066, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 20.63, | |
| "grad_norm": 1.9635140895843506, | |
| "learning_rate": 1.574063604240283e-05, | |
| "loss": 0.4048, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.63, | |
| "eval_cer": 0.34388076336474915, | |
| "eval_loss": 2.86527156829834, | |
| "eval_runtime": 408.1952, | |
| "eval_samples_per_second": 23.219, | |
| "eval_steps_per_second": 2.903, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.67, | |
| "grad_norm": 2.3898985385894775, | |
| "learning_rate": 1.5669964664310955e-05, | |
| "loss": 0.4189, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 20.67, | |
| "eval_cer": 0.34160034805052597, | |
| "eval_loss": 3.328908681869507, | |
| "eval_runtime": 391.6176, | |
| "eval_samples_per_second": 24.202, | |
| "eval_steps_per_second": 3.026, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 20.71, | |
| "grad_norm": 5.6570048332214355, | |
| "learning_rate": 1.5599293286219083e-05, | |
| "loss": 0.4067, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 20.71, | |
| "eval_cer": 0.3445235814008056, | |
| "eval_loss": 3.190380811691284, | |
| "eval_runtime": 410.6709, | |
| "eval_samples_per_second": 23.079, | |
| "eval_steps_per_second": 2.886, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "grad_norm": 6.448336124420166, | |
| "learning_rate": 1.5528621908127208e-05, | |
| "loss": 0.428, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "eval_cer": 0.3451126275859372, | |
| "eval_loss": 3.366471529006958, | |
| "eval_runtime": 393.2567, | |
| "eval_samples_per_second": 24.101, | |
| "eval_steps_per_second": 3.013, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "grad_norm": 3.681830406188965, | |
| "learning_rate": 1.5457950530035336e-05, | |
| "loss": 0.4141, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "eval_cer": 0.3429764185991944, | |
| "eval_loss": 3.280654191970825, | |
| "eval_runtime": 408.0304, | |
| "eval_samples_per_second": 23.229, | |
| "eval_steps_per_second": 2.904, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 20.84, | |
| "grad_norm": 3.107179641723633, | |
| "learning_rate": 1.5387279151943464e-05, | |
| "loss": 0.4127, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.84, | |
| "eval_cer": 0.34173233350279614, | |
| "eval_loss": 2.8979499340057373, | |
| "eval_runtime": 387.9855, | |
| "eval_samples_per_second": 24.429, | |
| "eval_steps_per_second": 3.054, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.88, | |
| "grad_norm": 2.1260199546813965, | |
| "learning_rate": 1.5316607773851592e-05, | |
| "loss": 0.4222, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 20.88, | |
| "eval_cer": 0.3436339016855031, | |
| "eval_loss": 2.4059560298919678, | |
| "eval_runtime": 412.8035, | |
| "eval_samples_per_second": 22.96, | |
| "eval_steps_per_second": 2.871, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 20.92, | |
| "grad_norm": 1.9174062013626099, | |
| "learning_rate": 1.5245936395759718e-05, | |
| "loss": 0.4093, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 20.92, | |
| "eval_cer": 0.3437194478119745, | |
| "eval_loss": 2.5948760509490967, | |
| "eval_runtime": 390.0559, | |
| "eval_samples_per_second": 24.299, | |
| "eval_steps_per_second": 3.038, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 20.96, | |
| "grad_norm": 2.362006902694702, | |
| "learning_rate": 1.5175265017667845e-05, | |
| "loss": 0.4054, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 20.96, | |
| "eval_cer": 0.34167611747682924, | |
| "eval_loss": 2.4431283473968506, | |
| "eval_runtime": 408.3684, | |
| "eval_samples_per_second": 23.209, | |
| "eval_steps_per_second": 2.902, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 1.828452706336975, | |
| "learning_rate": 1.5104593639575973e-05, | |
| "loss": 0.389, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "eval_cer": 0.3424289233897775, | |
| "eval_loss": 2.5479896068573, | |
| "eval_runtime": 390.6666, | |
| "eval_samples_per_second": 24.261, | |
| "eval_steps_per_second": 3.033, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "grad_norm": 4.511066913604736, | |
| "learning_rate": 1.50339222614841e-05, | |
| "loss": 0.3451, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "eval_cer": 0.34278332877087325, | |
| "eval_loss": 2.672588348388672, | |
| "eval_runtime": 415.0883, | |
| "eval_samples_per_second": 22.834, | |
| "eval_steps_per_second": 2.855, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "grad_norm": 3.952530860900879, | |
| "learning_rate": 1.4963250883392227e-05, | |
| "loss": 0.3449, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "eval_cer": 0.3425169136912909, | |
| "eval_loss": 2.945607900619507, | |
| "eval_runtime": 386.6565, | |
| "eval_samples_per_second": 24.513, | |
| "eval_steps_per_second": 3.065, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 21.13, | |
| "grad_norm": 2.7008559703826904, | |
| "learning_rate": 1.4892579505300355e-05, | |
| "loss": 0.3479, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 21.13, | |
| "eval_cer": 0.34289331664776507, | |
| "eval_loss": 2.209360122680664, | |
| "eval_runtime": 412.345, | |
| "eval_samples_per_second": 22.986, | |
| "eval_steps_per_second": 2.874, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 21.17, | |
| "grad_norm": 1.4452073574066162, | |
| "learning_rate": 1.482190812720848e-05, | |
| "loss": 0.3465, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 21.17, | |
| "eval_cer": 0.33977454929412226, | |
| "eval_loss": 2.268566370010376, | |
| "eval_runtime": 390.1756, | |
| "eval_samples_per_second": 24.292, | |
| "eval_steps_per_second": 3.037, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 21.22, | |
| "grad_norm": 4.574779987335205, | |
| "learning_rate": 1.475123674911661e-05, | |
| "loss": 0.3408, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 21.22, | |
| "eval_cer": 0.34143414414766726, | |
| "eval_loss": 2.403733491897583, | |
| "eval_runtime": 415.1727, | |
| "eval_samples_per_second": 22.829, | |
| "eval_steps_per_second": 2.854, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 21.26, | |
| "grad_norm": 1.5709911584854126, | |
| "learning_rate": 1.4680565371024735e-05, | |
| "loss": 0.3363, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.26, | |
| "eval_cer": 0.3419229791560752, | |
| "eval_loss": 2.4089136123657227, | |
| "eval_runtime": 388.419, | |
| "eval_samples_per_second": 24.401, | |
| "eval_steps_per_second": 3.051, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.3, | |
| "grad_norm": 4.430927276611328, | |
| "learning_rate": 1.4609893992932863e-05, | |
| "loss": 0.3461, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 21.3, | |
| "eval_cer": 0.34185698642994017, | |
| "eval_loss": 2.8909881114959717, | |
| "eval_runtime": 408.6465, | |
| "eval_samples_per_second": 23.194, | |
| "eval_steps_per_second": 2.9, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 21.34, | |
| "grad_norm": 1.9088200330734253, | |
| "learning_rate": 1.4539222614840989e-05, | |
| "loss": 0.3561, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 21.34, | |
| "eval_cer": 0.3429764185991944, | |
| "eval_loss": 2.6278483867645264, | |
| "eval_runtime": 388.0807, | |
| "eval_samples_per_second": 24.423, | |
| "eval_steps_per_second": 3.053, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 21.38, | |
| "grad_norm": 5.978403568267822, | |
| "learning_rate": 1.4468551236749117e-05, | |
| "loss": 0.3601, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 21.38, | |
| "eval_cer": 0.34170789175237576, | |
| "eval_loss": 2.4659318923950195, | |
| "eval_runtime": 412.615, | |
| "eval_samples_per_second": 22.971, | |
| "eval_steps_per_second": 2.872, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 21.43, | |
| "grad_norm": 1.794754981994629, | |
| "learning_rate": 1.4397879858657245e-05, | |
| "loss": 0.3487, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 21.43, | |
| "eval_cer": 0.34158079465018965, | |
| "eval_loss": 2.544365406036377, | |
| "eval_runtime": 390.5089, | |
| "eval_samples_per_second": 24.271, | |
| "eval_steps_per_second": 3.035, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 21.47, | |
| "grad_norm": 3.5707991123199463, | |
| "learning_rate": 1.4327208480565372e-05, | |
| "loss": 0.349, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.47, | |
| "eval_cer": 0.338039185014274, | |
| "eval_loss": 2.3195223808288574, | |
| "eval_runtime": 421.6099, | |
| "eval_samples_per_second": 22.48, | |
| "eval_steps_per_second": 2.811, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.51, | |
| "grad_norm": 1.6567233800888062, | |
| "learning_rate": 1.42565371024735e-05, | |
| "loss": 0.3534, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 21.51, | |
| "eval_cer": 0.33963767549176804, | |
| "eval_loss": 2.4492955207824707, | |
| "eval_runtime": 387.5977, | |
| "eval_samples_per_second": 24.453, | |
| "eval_steps_per_second": 3.057, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 21.55, | |
| "grad_norm": 2.3857908248901367, | |
| "learning_rate": 1.4185865724381625e-05, | |
| "loss": 0.3469, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 21.55, | |
| "eval_cer": 0.3406764498846349, | |
| "eval_loss": 2.440829277038574, | |
| "eval_runtime": 418.2267, | |
| "eval_samples_per_second": 22.662, | |
| "eval_steps_per_second": 2.833, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 21.6, | |
| "grad_norm": 1.5880221128463745, | |
| "learning_rate": 1.4115194346289753e-05, | |
| "loss": 0.3514, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 21.6, | |
| "eval_cer": 0.34056157365765904, | |
| "eval_loss": 2.3072359561920166, | |
| "eval_runtime": 388.3106, | |
| "eval_samples_per_second": 24.408, | |
| "eval_steps_per_second": 3.052, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 21.64, | |
| "grad_norm": 1.9676817655563354, | |
| "learning_rate": 1.4044522968197882e-05, | |
| "loss": 0.3469, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 21.64, | |
| "eval_cer": 0.34236293066364243, | |
| "eval_loss": 2.332427740097046, | |
| "eval_runtime": 411.3092, | |
| "eval_samples_per_second": 23.043, | |
| "eval_steps_per_second": 2.881, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 21.68, | |
| "grad_norm": 2.4417457580566406, | |
| "learning_rate": 1.3973851590106007e-05, | |
| "loss": 0.3535, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.68, | |
| "eval_cer": 0.3423751515388526, | |
| "eval_loss": 2.3060784339904785, | |
| "eval_runtime": 391.5586, | |
| "eval_samples_per_second": 24.206, | |
| "eval_steps_per_second": 3.026, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.72, | |
| "grad_norm": 5.330277919769287, | |
| "learning_rate": 1.3903180212014135e-05, | |
| "loss": 0.3529, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 21.72, | |
| "eval_cer": 0.34077421688631654, | |
| "eval_loss": 2.4329657554626465, | |
| "eval_runtime": 422.4059, | |
| "eval_samples_per_second": 22.438, | |
| "eval_steps_per_second": 2.805, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 21.76, | |
| "grad_norm": 2.12868595123291, | |
| "learning_rate": 1.3833215547703182e-05, | |
| "loss": 0.3733, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 21.76, | |
| "eval_cer": 0.34047358335614564, | |
| "eval_loss": 2.4534361362457275, | |
| "eval_runtime": 382.1988, | |
| "eval_samples_per_second": 24.799, | |
| "eval_steps_per_second": 3.1, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 21.81, | |
| "grad_norm": 1.6962841749191284, | |
| "learning_rate": 1.3762544169611308e-05, | |
| "loss": 0.3523, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 21.81, | |
| "eval_cer": 0.3399554182472332, | |
| "eval_loss": 2.549654483795166, | |
| "eval_runtime": 415.7902, | |
| "eval_samples_per_second": 22.795, | |
| "eval_steps_per_second": 2.85, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 21.85, | |
| "grad_norm": 1.762255072593689, | |
| "learning_rate": 1.3691872791519436e-05, | |
| "loss": 0.3523, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 21.85, | |
| "eval_cer": 0.33849624574713544, | |
| "eval_loss": 2.5730228424072266, | |
| "eval_runtime": 391.5055, | |
| "eval_samples_per_second": 24.209, | |
| "eval_steps_per_second": 3.027, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 21.89, | |
| "grad_norm": 1.794063687324524, | |
| "learning_rate": 1.3621201413427561e-05, | |
| "loss": 0.3608, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 21.89, | |
| "eval_cer": 0.3389410856047867, | |
| "eval_loss": 2.29063081741333, | |
| "eval_runtime": 429.6088, | |
| "eval_samples_per_second": 22.062, | |
| "eval_steps_per_second": 2.758, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 21.93, | |
| "grad_norm": 5.02766227722168, | |
| "learning_rate": 1.3550530035335691e-05, | |
| "loss": 0.3614, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 21.93, | |
| "eval_cer": 0.33830315591881427, | |
| "eval_loss": 2.339839458465576, | |
| "eval_runtime": 388.2936, | |
| "eval_samples_per_second": 24.409, | |
| "eval_steps_per_second": 3.052, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 21.97, | |
| "grad_norm": 1.7431000471115112, | |
| "learning_rate": 1.3479858657243816e-05, | |
| "loss": 0.3401, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 21.97, | |
| "eval_cer": 0.3359225294278675, | |
| "eval_loss": 2.4058258533477783, | |
| "eval_runtime": 412.0183, | |
| "eval_samples_per_second": 23.004, | |
| "eval_steps_per_second": 2.876, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 2.714993953704834, | |
| "learning_rate": 1.3409187279151944e-05, | |
| "loss": 0.3222, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "eval_cer": 0.33601051972938095, | |
| "eval_loss": 2.8305118083953857, | |
| "eval_runtime": 391.1586, | |
| "eval_samples_per_second": 24.231, | |
| "eval_steps_per_second": 3.029, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 22.06, | |
| "grad_norm": 1.8999407291412354, | |
| "learning_rate": 1.3338515901060072e-05, | |
| "loss": 0.3022, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 22.06, | |
| "eval_cer": 0.33706151499745807, | |
| "eval_loss": 2.4940497875213623, | |
| "eval_runtime": 422.3325, | |
| "eval_samples_per_second": 22.442, | |
| "eval_steps_per_second": 2.806, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 22.1, | |
| "grad_norm": 1.6719143390655518, | |
| "learning_rate": 1.3267844522968198e-05, | |
| "loss": 0.3009, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.1, | |
| "eval_cer": 0.3367682139924133, | |
| "eval_loss": 2.5411734580993652, | |
| "eval_runtime": 390.8575, | |
| "eval_samples_per_second": 24.249, | |
| "eval_steps_per_second": 3.032, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.14, | |
| "grad_norm": 2.155449151992798, | |
| "learning_rate": 1.3197173144876326e-05, | |
| "loss": 0.2968, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 22.14, | |
| "eval_cer": 0.3380514058894842, | |
| "eval_loss": 2.6296260356903076, | |
| "eval_runtime": 425.2856, | |
| "eval_samples_per_second": 22.286, | |
| "eval_steps_per_second": 2.786, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 22.18, | |
| "grad_norm": 11.153074264526367, | |
| "learning_rate": 1.3126501766784453e-05, | |
| "loss": 0.2936, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 22.18, | |
| "eval_cer": 0.33938836963748, | |
| "eval_loss": 2.600175142288208, | |
| "eval_runtime": 394.8419, | |
| "eval_samples_per_second": 24.005, | |
| "eval_steps_per_second": 3.001, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 22.23, | |
| "grad_norm": 1.9212164878845215, | |
| "learning_rate": 1.305583038869258e-05, | |
| "loss": 0.3058, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 22.23, | |
| "eval_cer": 0.3395081342145399, | |
| "eval_loss": 2.419398546218872, | |
| "eval_runtime": 417.7544, | |
| "eval_samples_per_second": 22.688, | |
| "eval_steps_per_second": 2.837, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 22.27, | |
| "grad_norm": 4.1538004875183105, | |
| "learning_rate": 1.2985159010600709e-05, | |
| "loss": 0.2954, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 22.27, | |
| "eval_cer": 0.33905107348167846, | |
| "eval_loss": 2.4820761680603027, | |
| "eval_runtime": 388.8242, | |
| "eval_samples_per_second": 24.376, | |
| "eval_steps_per_second": 3.048, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 22.31, | |
| "grad_norm": 2.294267177581787, | |
| "learning_rate": 1.2914487632508834e-05, | |
| "loss": 0.3038, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.31, | |
| "eval_cer": 0.3415172460990966, | |
| "eval_loss": 2.524266004562378, | |
| "eval_runtime": 399.4446, | |
| "eval_samples_per_second": 23.728, | |
| "eval_steps_per_second": 2.967, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.35, | |
| "grad_norm": 2.0384035110473633, | |
| "learning_rate": 1.2843816254416963e-05, | |
| "loss": 0.2994, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 22.35, | |
| "eval_cer": 0.3371079543232568, | |
| "eval_loss": 2.487698554992676, | |
| "eval_runtime": 374.207, | |
| "eval_samples_per_second": 25.328, | |
| "eval_steps_per_second": 3.167, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "grad_norm": 1.3863327503204346, | |
| "learning_rate": 1.2773144876325088e-05, | |
| "loss": 0.2997, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "eval_cer": 0.33839359039536976, | |
| "eval_loss": 2.4660918712615967, | |
| "eval_runtime": 423.5986, | |
| "eval_samples_per_second": 22.375, | |
| "eval_steps_per_second": 2.797, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 22.44, | |
| "grad_norm": 1.9845781326293945, | |
| "learning_rate": 1.2702473498233216e-05, | |
| "loss": 0.2999, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 22.44, | |
| "eval_cer": 0.3362524930585429, | |
| "eval_loss": 2.4160196781158447, | |
| "eval_runtime": 387.3168, | |
| "eval_samples_per_second": 24.471, | |
| "eval_steps_per_second": 3.06, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 22.48, | |
| "grad_norm": 2.225783586502075, | |
| "learning_rate": 1.2631802120141343e-05, | |
| "loss": 0.2854, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 22.48, | |
| "eval_cer": 0.33407228892104335, | |
| "eval_loss": 2.5062761306762695, | |
| "eval_runtime": 421.3473, | |
| "eval_samples_per_second": 22.495, | |
| "eval_steps_per_second": 2.812, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 22.52, | |
| "grad_norm": 9.648059844970703, | |
| "learning_rate": 1.256113074204947e-05, | |
| "loss": 0.2985, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.52, | |
| "eval_cer": 0.3380709592898205, | |
| "eval_loss": 2.483365774154663, | |
| "eval_runtime": 392.706, | |
| "eval_samples_per_second": 24.135, | |
| "eval_steps_per_second": 3.018, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.56, | |
| "grad_norm": 1.8772177696228027, | |
| "learning_rate": 1.2490459363957597e-05, | |
| "loss": 0.3069, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 22.56, | |
| "eval_cer": 0.3383080442688984, | |
| "eval_loss": 2.481276035308838, | |
| "eval_runtime": 425.0385, | |
| "eval_samples_per_second": 22.299, | |
| "eval_steps_per_second": 2.788, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 22.61, | |
| "grad_norm": 1.4407734870910645, | |
| "learning_rate": 1.2419787985865725e-05, | |
| "loss": 0.3018, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 22.61, | |
| "eval_cer": 0.3348739783348324, | |
| "eval_loss": 2.4129292964935303, | |
| "eval_runtime": 399.5718, | |
| "eval_samples_per_second": 23.72, | |
| "eval_steps_per_second": 2.966, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 22.65, | |
| "grad_norm": 5.736636638641357, | |
| "learning_rate": 1.2349116607773852e-05, | |
| "loss": 0.2925, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 22.65, | |
| "eval_cer": 0.3356121191975284, | |
| "eval_loss": 2.4841222763061523, | |
| "eval_runtime": 427.1982, | |
| "eval_samples_per_second": 22.186, | |
| "eval_steps_per_second": 2.774, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 22.69, | |
| "grad_norm": 6.094464302062988, | |
| "learning_rate": 1.227844522968198e-05, | |
| "loss": 0.2981, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 22.69, | |
| "eval_cer": 0.3367071096163623, | |
| "eval_loss": 2.4353983402252197, | |
| "eval_runtime": 389.5123, | |
| "eval_samples_per_second": 24.333, | |
| "eval_steps_per_second": 3.042, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "grad_norm": 1.8495489358901978, | |
| "learning_rate": 1.2207773851590106e-05, | |
| "loss": 0.2977, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "eval_cer": 0.3380245199640217, | |
| "eval_loss": 2.4169256687164307, | |
| "eval_runtime": 432.3991, | |
| "eval_samples_per_second": 21.92, | |
| "eval_steps_per_second": 2.741, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.77, | |
| "grad_norm": 1.7505403757095337, | |
| "learning_rate": 1.2137102473498234e-05, | |
| "loss": 0.3062, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 22.77, | |
| "eval_cer": 0.3384766923467991, | |
| "eval_loss": 2.552931070327759, | |
| "eval_runtime": 394.4096, | |
| "eval_samples_per_second": 24.031, | |
| "eval_steps_per_second": 3.004, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 22.82, | |
| "grad_norm": 3.5150487422943115, | |
| "learning_rate": 1.206643109540636e-05, | |
| "loss": 0.3035, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 22.82, | |
| "eval_cer": 0.3343362598255837, | |
| "eval_loss": 2.274048328399658, | |
| "eval_runtime": 434.158, | |
| "eval_samples_per_second": 21.831, | |
| "eval_steps_per_second": 2.729, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "grad_norm": 2.194784641265869, | |
| "learning_rate": 1.1995759717314487e-05, | |
| "loss": 0.3109, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "eval_cer": 0.33527482304172695, | |
| "eval_loss": 2.228667736053467, | |
| "eval_runtime": 394.9356, | |
| "eval_samples_per_second": 23.999, | |
| "eval_steps_per_second": 3.0, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 22.9, | |
| "grad_norm": 4.459865570068359, | |
| "learning_rate": 1.1925088339222615e-05, | |
| "loss": 0.3105, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 22.9, | |
| "eval_cer": 0.33500596378710257, | |
| "eval_loss": 2.3388936519622803, | |
| "eval_runtime": 432.3816, | |
| "eval_samples_per_second": 21.92, | |
| "eval_steps_per_second": 2.741, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 22.94, | |
| "grad_norm": 2.0811266899108887, | |
| "learning_rate": 1.1854416961130743e-05, | |
| "loss": 0.3078, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 22.94, | |
| "eval_cer": 0.3347126627820578, | |
| "eval_loss": 2.5939130783081055, | |
| "eval_runtime": 402.0697, | |
| "eval_samples_per_second": 23.573, | |
| "eval_steps_per_second": 2.947, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 22.98, | |
| "grad_norm": 4.0244011878967285, | |
| "learning_rate": 1.178374558303887e-05, | |
| "loss": 0.3026, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 22.98, | |
| "eval_cer": 0.33435581322592, | |
| "eval_loss": 2.545149564743042, | |
| "eval_runtime": 441.4018, | |
| "eval_samples_per_second": 21.473, | |
| "eval_steps_per_second": 2.685, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 2.544414758682251, | |
| "learning_rate": 1.1713074204946998e-05, | |
| "loss": 0.2848, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "eval_cer": 0.33448291032810606, | |
| "eval_loss": 2.603243112564087, | |
| "eval_runtime": 394.5283, | |
| "eval_samples_per_second": 24.024, | |
| "eval_steps_per_second": 3.004, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 23.07, | |
| "grad_norm": 1.4476730823516846, | |
| "learning_rate": 1.1642402826855124e-05, | |
| "loss": 0.2693, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 23.07, | |
| "eval_cer": 0.3340185170701185, | |
| "eval_loss": 2.598785638809204, | |
| "eval_runtime": 422.01, | |
| "eval_samples_per_second": 22.459, | |
| "eval_steps_per_second": 2.808, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 23.11, | |
| "grad_norm": 2.027191400527954, | |
| "learning_rate": 1.157173144876325e-05, | |
| "loss": 0.2632, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 23.11, | |
| "eval_cer": 0.33527237886668493, | |
| "eval_loss": 2.615438461303711, | |
| "eval_runtime": 400.4466, | |
| "eval_samples_per_second": 23.669, | |
| "eval_steps_per_second": 2.959, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 23.15, | |
| "grad_norm": 2.888965606689453, | |
| "learning_rate": 1.1501060070671379e-05, | |
| "loss": 0.2584, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.15, | |
| "eval_cer": 0.3347419928825623, | |
| "eval_loss": 2.6486945152282715, | |
| "eval_runtime": 425.0006, | |
| "eval_samples_per_second": 22.301, | |
| "eval_steps_per_second": 2.788, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.2, | |
| "grad_norm": 1.428572654724121, | |
| "learning_rate": 1.1430388692579507e-05, | |
| "loss": 0.2592, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 23.2, | |
| "eval_cer": 0.33205340033631847, | |
| "eval_loss": 2.5768284797668457, | |
| "eval_runtime": 410.5876, | |
| "eval_samples_per_second": 23.084, | |
| "eval_steps_per_second": 2.886, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 23.24, | |
| "grad_norm": 1.8620479106903076, | |
| "learning_rate": 1.1359717314487633e-05, | |
| "loss": 0.2719, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 23.24, | |
| "eval_cer": 0.3328697548003598, | |
| "eval_loss": 3.229948043823242, | |
| "eval_runtime": 423.4357, | |
| "eval_samples_per_second": 22.384, | |
| "eval_steps_per_second": 2.799, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 23.28, | |
| "grad_norm": 1.727643609046936, | |
| "learning_rate": 1.128904593639576e-05, | |
| "loss": 0.2628, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 23.28, | |
| "eval_cer": 0.33284286887489734, | |
| "eval_loss": 2.7340400218963623, | |
| "eval_runtime": 394.6323, | |
| "eval_samples_per_second": 24.017, | |
| "eval_steps_per_second": 3.003, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 23.32, | |
| "grad_norm": 2.5659921169281006, | |
| "learning_rate": 1.1218374558303888e-05, | |
| "loss": 0.2612, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 23.32, | |
| "eval_cer": 0.3338303155918814, | |
| "eval_loss": 2.4699885845184326, | |
| "eval_runtime": 423.6808, | |
| "eval_samples_per_second": 22.371, | |
| "eval_steps_per_second": 2.797, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 23.36, | |
| "grad_norm": 1.429592490196228, | |
| "learning_rate": 1.1147703180212014e-05, | |
| "loss": 0.2702, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.36, | |
| "eval_cer": 0.33553879394626723, | |
| "eval_loss": 2.7780275344848633, | |
| "eval_runtime": 398.5381, | |
| "eval_samples_per_second": 23.782, | |
| "eval_steps_per_second": 2.973, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.41, | |
| "grad_norm": 2.3033080101013184, | |
| "learning_rate": 1.107703180212014e-05, | |
| "loss": 0.2574, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 23.41, | |
| "eval_cer": 0.3336567791638966, | |
| "eval_loss": 2.5886073112487793, | |
| "eval_runtime": 422.633, | |
| "eval_samples_per_second": 22.426, | |
| "eval_steps_per_second": 2.804, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 23.45, | |
| "grad_norm": 3.2907676696777344, | |
| "learning_rate": 1.100636042402827e-05, | |
| "loss": 0.2612, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 23.45, | |
| "eval_cer": 0.33410161902154784, | |
| "eval_loss": 2.841667413711548, | |
| "eval_runtime": 403.4827, | |
| "eval_samples_per_second": 23.49, | |
| "eval_steps_per_second": 2.937, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 23.49, | |
| "grad_norm": 1.8252880573272705, | |
| "learning_rate": 1.0935689045936397e-05, | |
| "loss": 0.2634, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 23.49, | |
| "eval_cer": 0.3340087403699503, | |
| "eval_loss": 2.7159037590026855, | |
| "eval_runtime": 426.7576, | |
| "eval_samples_per_second": 22.209, | |
| "eval_steps_per_second": 2.777, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 23.53, | |
| "grad_norm": 1.250098466873169, | |
| "learning_rate": 1.0865017667844523e-05, | |
| "loss": 0.2592, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 23.53, | |
| "eval_cer": 0.3338596456923859, | |
| "eval_loss": 2.6912946701049805, | |
| "eval_runtime": 397.5959, | |
| "eval_samples_per_second": 23.838, | |
| "eval_steps_per_second": 2.98, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 23.57, | |
| "grad_norm": 3.1888105869293213, | |
| "learning_rate": 1.0794346289752651e-05, | |
| "loss": 0.2537, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.57, | |
| "eval_cer": 0.3346197841304603, | |
| "eval_loss": 2.708544969558716, | |
| "eval_runtime": 427.2487, | |
| "eval_samples_per_second": 22.184, | |
| "eval_steps_per_second": 2.774, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.62, | |
| "grad_norm": 9.785385131835938, | |
| "learning_rate": 1.0723674911660778e-05, | |
| "loss": 0.2622, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 23.62, | |
| "eval_cer": 0.3341309491220523, | |
| "eval_loss": 2.625296115875244, | |
| "eval_runtime": 401.2011, | |
| "eval_samples_per_second": 23.624, | |
| "eval_steps_per_second": 2.954, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 23.66, | |
| "grad_norm": 1.4341765642166138, | |
| "learning_rate": 1.0653003533568904e-05, | |
| "loss": 0.26, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 23.66, | |
| "eval_cer": 0.33526260216651677, | |
| "eval_loss": 2.6399149894714355, | |
| "eval_runtime": 433.4463, | |
| "eval_samples_per_second": 21.867, | |
| "eval_steps_per_second": 2.734, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 23.7, | |
| "grad_norm": 1.7979743480682373, | |
| "learning_rate": 1.0582332155477032e-05, | |
| "loss": 0.2688, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 23.7, | |
| "eval_cer": 0.3343387040006257, | |
| "eval_loss": 2.507894515991211, | |
| "eval_runtime": 402.3559, | |
| "eval_samples_per_second": 23.556, | |
| "eval_steps_per_second": 2.945, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 23.74, | |
| "grad_norm": 2.900815725326538, | |
| "learning_rate": 1.051166077738516e-05, | |
| "loss": 0.2671, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 23.74, | |
| "eval_cer": 0.33364211411364436, | |
| "eval_loss": 2.6333584785461426, | |
| "eval_runtime": 425.9033, | |
| "eval_samples_per_second": 22.254, | |
| "eval_steps_per_second": 2.782, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 23.78, | |
| "grad_norm": 1.366709589958191, | |
| "learning_rate": 1.0440989399293287e-05, | |
| "loss": 0.2509, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.78, | |
| "eval_cer": 0.3335125728364163, | |
| "eval_loss": 2.557394504547119, | |
| "eval_runtime": 402.0125, | |
| "eval_samples_per_second": 23.576, | |
| "eval_steps_per_second": 2.948, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.83, | |
| "grad_norm": 3.0282020568847656, | |
| "learning_rate": 1.0370318021201413e-05, | |
| "loss": 0.2639, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 23.83, | |
| "eval_cer": 0.33265222322161825, | |
| "eval_loss": 2.564777374267578, | |
| "eval_runtime": 430.1757, | |
| "eval_samples_per_second": 22.033, | |
| "eval_steps_per_second": 2.755, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 23.87, | |
| "grad_norm": 1.940981388092041, | |
| "learning_rate": 1.0299646643109541e-05, | |
| "loss": 0.2606, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 23.87, | |
| "eval_cer": 0.33390608501818464, | |
| "eval_loss": 2.5213394165039062, | |
| "eval_runtime": 397.5174, | |
| "eval_samples_per_second": 23.843, | |
| "eval_steps_per_second": 2.981, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 23.91, | |
| "grad_norm": 1.5934607982635498, | |
| "learning_rate": 1.0228975265017668e-05, | |
| "loss": 0.256, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 23.91, | |
| "eval_cer": 0.33383764811700756, | |
| "eval_loss": 2.509065866470337, | |
| "eval_runtime": 427.9199, | |
| "eval_samples_per_second": 22.149, | |
| "eval_steps_per_second": 2.769, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 23.95, | |
| "grad_norm": 6.75713586807251, | |
| "learning_rate": 1.0158303886925796e-05, | |
| "loss": 0.262, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 23.95, | |
| "eval_cer": 0.33297974267725156, | |
| "eval_loss": 2.544567108154297, | |
| "eval_runtime": 400.6912, | |
| "eval_samples_per_second": 23.654, | |
| "eval_steps_per_second": 2.957, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 23.99, | |
| "grad_norm": 1.9022639989852905, | |
| "learning_rate": 1.0087632508833924e-05, | |
| "loss": 0.2657, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 23.99, | |
| "eval_cer": 0.33296507762699934, | |
| "eval_loss": 2.5121145248413086, | |
| "eval_runtime": 423.803, | |
| "eval_samples_per_second": 22.364, | |
| "eval_steps_per_second": 2.796, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.04, | |
| "grad_norm": 1.854609727859497, | |
| "learning_rate": 1.001696113074205e-05, | |
| "loss": 0.2253, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 24.04, | |
| "eval_cer": 0.3338425364670916, | |
| "eval_loss": 2.654775381088257, | |
| "eval_runtime": 395.2094, | |
| "eval_samples_per_second": 23.982, | |
| "eval_steps_per_second": 2.998, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 24.08, | |
| "grad_norm": 5.454607009887695, | |
| "learning_rate": 9.946289752650177e-06, | |
| "loss": 0.2339, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 24.08, | |
| "eval_cer": 0.3325251261194322, | |
| "eval_loss": 2.6627321243286133, | |
| "eval_runtime": 422.5781, | |
| "eval_samples_per_second": 22.429, | |
| "eval_steps_per_second": 2.804, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 24.12, | |
| "grad_norm": 4.305576801300049, | |
| "learning_rate": 9.875618374558305e-06, | |
| "loss": 0.2229, | |
| "step": 57300 | |
| }, | |
| { | |
| "epoch": 24.12, | |
| "eval_cer": 0.33326815533221227, | |
| "eval_loss": 2.737283945083618, | |
| "eval_runtime": 400.1034, | |
| "eval_samples_per_second": 23.689, | |
| "eval_steps_per_second": 2.962, | |
| "step": 57300 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 71250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 100, | |
| "total_flos": 6.279302169302917e+20, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |