| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4240956992309883, | |
| "eval_steps": 200, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.028481913984619765, | |
| "eval_loss": 1.2521600723266602, | |
| "eval_runtime": 184.7297, | |
| "eval_samples_per_second": 38.012, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 0.6291606319509959, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05696382796923953, | |
| "eval_loss": 0.6599467396736145, | |
| "eval_runtime": 185.0187, | |
| "eval_samples_per_second": 37.953, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 0.45444398676570247, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07120478496154942, | |
| "grad_norm": 18.865299224853516, | |
| "learning_rate": 0.00014879999999999998, | |
| "loss": 2.2791, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0854457419538593, | |
| "eval_loss": 0.6628636717796326, | |
| "eval_runtime": 185.7673, | |
| "eval_samples_per_second": 37.8, | |
| "eval_steps_per_second": 0.592, | |
| "eval_wer": 0.4394557461566059, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.11392765593847906, | |
| "eval_loss": 0.7910040020942688, | |
| "eval_runtime": 186.4058, | |
| "eval_samples_per_second": 37.671, | |
| "eval_steps_per_second": 0.59, | |
| "eval_wer": 0.5453035517346763, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.14240956992309883, | |
| "grad_norm": 3.8627092838287354, | |
| "learning_rate": 0.0002988, | |
| "loss": 0.8206, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14240956992309883, | |
| "eval_loss": 0.7757941484451294, | |
| "eval_runtime": 186.8087, | |
| "eval_samples_per_second": 37.589, | |
| "eval_steps_per_second": 0.589, | |
| "eval_wer": 0.5701245033816553, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1708914839077186, | |
| "eval_loss": 0.802534818649292, | |
| "eval_runtime": 187.4308, | |
| "eval_samples_per_second": 37.464, | |
| "eval_steps_per_second": 0.587, | |
| "eval_wer": 0.5782564211589312, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.19937339789233838, | |
| "eval_loss": 0.7715001106262207, | |
| "eval_runtime": 187.8412, | |
| "eval_samples_per_second": 37.383, | |
| "eval_steps_per_second": 0.586, | |
| "eval_wer": 0.5211336850077731, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.21361435488464825, | |
| "grad_norm": 11.042049407958984, | |
| "learning_rate": 0.00028346666666666665, | |
| "loss": 0.9068, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.22785531187695812, | |
| "eval_loss": 0.7349154949188232, | |
| "eval_runtime": 191.6788, | |
| "eval_samples_per_second": 36.634, | |
| "eval_steps_per_second": 0.574, | |
| "eval_wer": 0.512775880625573, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.2563372258615779, | |
| "eval_loss": 0.7257962226867676, | |
| "eval_runtime": 189.501, | |
| "eval_samples_per_second": 37.055, | |
| "eval_steps_per_second": 0.58, | |
| "eval_wer": 0.5152473458323922, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.28481913984619767, | |
| "grad_norm": 6.190296649932861, | |
| "learning_rate": 0.0002668, | |
| "loss": 0.8679, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.28481913984619767, | |
| "eval_loss": 0.7084089517593384, | |
| "eval_runtime": 188.267, | |
| "eval_samples_per_second": 37.298, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.5216386080070158, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3133010538308174, | |
| "eval_loss": 0.6904259324073792, | |
| "eval_runtime": 188.556, | |
| "eval_samples_per_second": 37.241, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.5014151131426142, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3417829678154372, | |
| "eval_loss": 0.6992842555046082, | |
| "eval_runtime": 189.0868, | |
| "eval_samples_per_second": 37.136, | |
| "eval_steps_per_second": 0.582, | |
| "eval_wer": 0.5177586733812567, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3560239248077471, | |
| "grad_norm": 4.8257222175598145, | |
| "learning_rate": 0.0002501333333333333, | |
| "loss": 0.8577, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.37026488180005696, | |
| "eval_loss": 0.6746060848236084, | |
| "eval_runtime": 190.1492, | |
| "eval_samples_per_second": 36.929, | |
| "eval_steps_per_second": 0.578, | |
| "eval_wer": 0.48673248382253287, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.39874679578467676, | |
| "eval_loss": 0.6621994972229004, | |
| "eval_runtime": 189.6459, | |
| "eval_samples_per_second": 37.027, | |
| "eval_steps_per_second": 0.58, | |
| "eval_wer": 0.4962595835714001, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.4272287097692965, | |
| "grad_norm": 3.6695899963378906, | |
| "learning_rate": 0.00023346666666666666, | |
| "loss": 0.7995, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4272287097692965, | |
| "eval_loss": 0.6793097853660583, | |
| "eval_runtime": 188.7722, | |
| "eval_samples_per_second": 37.198, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.49348250707556574, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.45571062375391624, | |
| "eval_loss": 0.6368467211723328, | |
| "eval_runtime": 188.0679, | |
| "eval_samples_per_second": 37.338, | |
| "eval_steps_per_second": 0.585, | |
| "eval_wer": 0.47005673740017806, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.48419253773853604, | |
| "eval_loss": 0.6363435387611389, | |
| "eval_runtime": 188.2666, | |
| "eval_samples_per_second": 37.298, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.478055780703969, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4984334947308459, | |
| "grad_norm": 3.4502739906311035, | |
| "learning_rate": 0.0002168333333333333, | |
| "loss": 0.8141, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5126744517231558, | |
| "eval_loss": 0.6217373609542847, | |
| "eval_runtime": 187.6755, | |
| "eval_samples_per_second": 37.416, | |
| "eval_steps_per_second": 0.586, | |
| "eval_wer": 0.46555229274904, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5411563657077756, | |
| "eval_loss": 0.641762912273407, | |
| "eval_runtime": 186.9231, | |
| "eval_samples_per_second": 37.566, | |
| "eval_steps_per_second": 0.588, | |
| "eval_wer": 0.4940140049695053, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5696382796923953, | |
| "grad_norm": 5.877405643463135, | |
| "learning_rate": 0.00020016666666666666, | |
| "loss": 0.7953, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5696382796923953, | |
| "eval_loss": 0.6017736196517944, | |
| "eval_runtime": 182.787, | |
| "eval_samples_per_second": 38.416, | |
| "eval_steps_per_second": 0.602, | |
| "eval_wer": 0.4542313876081266, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5981201936770151, | |
| "eval_loss": 0.5962206721305847, | |
| "eval_runtime": 183.0007, | |
| "eval_samples_per_second": 38.371, | |
| "eval_steps_per_second": 0.601, | |
| "eval_wer": 0.4580315975497947, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.6266021076616348, | |
| "eval_loss": 0.5883399844169617, | |
| "eval_runtime": 182.7298, | |
| "eval_samples_per_second": 38.428, | |
| "eval_steps_per_second": 0.602, | |
| "eval_wer": 0.44590015812062345, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6408430646539447, | |
| "grad_norm": 3.615546226501465, | |
| "learning_rate": 0.0001835333333333333, | |
| "loss": 0.7596, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6550840216462547, | |
| "eval_loss": 0.578825056552887, | |
| "eval_runtime": 183.3674, | |
| "eval_samples_per_second": 38.295, | |
| "eval_steps_per_second": 0.6, | |
| "eval_wer": 0.43253298608804264, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.6835659356308744, | |
| "eval_loss": 0.5708740949630737, | |
| "eval_runtime": 182.6951, | |
| "eval_samples_per_second": 38.436, | |
| "eval_steps_per_second": 0.602, | |
| "eval_wer": 0.4412362641013035, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.7120478496154942, | |
| "grad_norm": 4.345168590545654, | |
| "learning_rate": 0.0001669, | |
| "loss": 0.7533, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7120478496154942, | |
| "eval_loss": 0.5594890117645264, | |
| "eval_runtime": 182.5857, | |
| "eval_samples_per_second": 38.459, | |
| "eval_steps_per_second": 0.602, | |
| "eval_wer": 0.4352170504524376, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7405297636001139, | |
| "eval_loss": 0.5545539259910583, | |
| "eval_runtime": 182.2233, | |
| "eval_samples_per_second": 38.535, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.4231786231547057, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7690116775847337, | |
| "eval_loss": 0.5545418858528137, | |
| "eval_runtime": 182.2691, | |
| "eval_samples_per_second": 38.525, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.4244276432054638, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7832526345770435, | |
| "grad_norm": 9.471431732177734, | |
| "learning_rate": 0.00015026666666666667, | |
| "loss": 0.7591, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7974935915693535, | |
| "eval_loss": 0.5442594885826111, | |
| "eval_runtime": 182.3947, | |
| "eval_samples_per_second": 38.499, | |
| "eval_steps_per_second": 0.603, | |
| "eval_wer": 0.4076455972043211, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.8259755055539733, | |
| "eval_loss": 0.5341240763664246, | |
| "eval_runtime": 182.0603, | |
| "eval_samples_per_second": 38.57, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.41462150706227824, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.854457419538593, | |
| "grad_norm": 4.406210422515869, | |
| "learning_rate": 0.00013363333333333332, | |
| "loss": 0.6621, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.854457419538593, | |
| "eval_loss": 0.5104002952575684, | |
| "eval_runtime": 181.8706, | |
| "eval_samples_per_second": 38.61, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.3955141577751498, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8829393335232127, | |
| "eval_loss": 0.5139421820640564, | |
| "eval_runtime": 181.902, | |
| "eval_samples_per_second": 38.603, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.40112146055621256, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.9114212475078325, | |
| "eval_loss": 0.5044221878051758, | |
| "eval_runtime": 181.9538, | |
| "eval_samples_per_second": 38.592, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.38039304269256835, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.9256622045001424, | |
| "grad_norm": 8.09687328338623, | |
| "learning_rate": 0.000117, | |
| "loss": 0.6705, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.9399031614924523, | |
| "eval_loss": 0.49985769391059875, | |
| "eval_runtime": 182.1414, | |
| "eval_samples_per_second": 38.552, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.3896012437050718, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.9683850754770721, | |
| "eval_loss": 0.5097447037696838, | |
| "eval_runtime": 181.5418, | |
| "eval_samples_per_second": 38.68, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.4052804315762899, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9968669894616918, | |
| "grad_norm": 4.639442443847656, | |
| "learning_rate": 0.00010033333333333332, | |
| "loss": 0.6665, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9968669894616918, | |
| "eval_loss": 0.49253013730049133, | |
| "eval_runtime": 181.6405, | |
| "eval_samples_per_second": 38.659, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.3784796502743858, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.0253489034463117, | |
| "eval_loss": 0.4896470010280609, | |
| "eval_runtime": 181.3934, | |
| "eval_samples_per_second": 38.711, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.3688728258414276, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.0538308174309314, | |
| "eval_loss": 0.47494611144065857, | |
| "eval_runtime": 181.7386, | |
| "eval_samples_per_second": 38.638, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.3687399513679427, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.0680717744232413, | |
| "grad_norm": 0.6623511910438538, | |
| "learning_rate": 8.366666666666666e-05, | |
| "loss": 0.5826, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0823127314155512, | |
| "eval_loss": 0.4684299826622009, | |
| "eval_runtime": 182.4026, | |
| "eval_samples_per_second": 38.497, | |
| "eval_steps_per_second": 0.603, | |
| "eval_wer": 0.3628004624031677, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.110794645400171, | |
| "eval_loss": 0.47290024161338806, | |
| "eval_runtime": 182.1043, | |
| "eval_samples_per_second": 38.56, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.358495329462257, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.1392765593847907, | |
| "grad_norm": 2.393817186355591, | |
| "learning_rate": 6.699999999999999e-05, | |
| "loss": 0.5836, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1392765593847907, | |
| "eval_loss": 0.46409761905670166, | |
| "eval_runtime": 181.7327, | |
| "eval_samples_per_second": 38.639, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.3553196295459679, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1677584733694104, | |
| "eval_loss": 0.45749881863594055, | |
| "eval_runtime": 181.5866, | |
| "eval_samples_per_second": 38.67, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.3529810388126337, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.1962403873540302, | |
| "eval_loss": 0.45851147174835205, | |
| "eval_runtime": 181.5801, | |
| "eval_samples_per_second": 38.672, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.3485563188455866, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.21048134434634, | |
| "grad_norm": 1.9676859378814697, | |
| "learning_rate": 5.033333333333333e-05, | |
| "loss": 0.5199, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.22472230133865, | |
| "eval_loss": 0.4548875391483307, | |
| "eval_runtime": 182.6274, | |
| "eval_samples_per_second": 38.45, | |
| "eval_steps_per_second": 0.602, | |
| "eval_wer": 0.3450750076402822, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.2532042153232696, | |
| "eval_loss": 0.4520675539970398, | |
| "eval_runtime": 182.8881, | |
| "eval_samples_per_second": 38.395, | |
| "eval_steps_per_second": 0.601, | |
| "eval_wer": 0.34082302448876545, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.2816861293078894, | |
| "grad_norm": 1.1400251388549805, | |
| "learning_rate": 3.373333333333333e-05, | |
| "loss": 0.5268, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2816861293078894, | |
| "eval_loss": 0.44252264499664307, | |
| "eval_runtime": 182.3349, | |
| "eval_samples_per_second": 38.512, | |
| "eval_steps_per_second": 0.603, | |
| "eval_wer": 0.33950756720126496, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.3101680432925091, | |
| "eval_loss": 0.44072064757347107, | |
| "eval_runtime": 184.1579, | |
| "eval_samples_per_second": 38.13, | |
| "eval_steps_per_second": 0.597, | |
| "eval_wer": 0.3361857053641425, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.338649957277129, | |
| "eval_loss": 0.4383063018321991, | |
| "eval_runtime": 181.6966, | |
| "eval_samples_per_second": 38.647, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.33397998910429316, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.352890914269439, | |
| "grad_norm": 1.0755033493041992, | |
| "learning_rate": 1.71e-05, | |
| "loss": 0.5013, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.3671318712617488, | |
| "eval_loss": 0.4356846809387207, | |
| "eval_runtime": 183.1225, | |
| "eval_samples_per_second": 38.346, | |
| "eval_steps_per_second": 0.601, | |
| "eval_wer": 0.33253165734330775, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.3956137852463686, | |
| "eval_loss": 0.43495818972587585, | |
| "eval_runtime": 182.2639, | |
| "eval_samples_per_second": 38.527, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.3316812607130044, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.4240956992309883, | |
| "grad_norm": 1.6312005519866943, | |
| "learning_rate": 4.666666666666666e-07, | |
| "loss": 0.5095, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.4240956992309883, | |
| "eval_loss": 0.43451622128486633, | |
| "eval_runtime": 182.2078, | |
| "eval_samples_per_second": 38.538, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.3308175766353526, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.4240956992309883, | |
| "step": 10000, | |
| "total_flos": 4.5974516642218747e+18, | |
| "train_loss": 0.7817989181518554, | |
| "train_runtime": 11412.7197, | |
| "train_samples_per_second": 3.505, | |
| "train_steps_per_second": 0.876 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.5974516642218747e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |