| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.8513997053251945, | |
| "eval_steps": 100, | |
| "global_step": 13900, | |
| "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 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 71250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 100, | |
| "total_flos": 1.523070949553679e+20, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |