SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/google_bert
/bert_base_uncased_ledgar
/trainer_state.json
| { | |
| "best_metric": 0.6676496267318726, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_bert/bert_base_uncased_ledgar/checkpoint-2800", | |
| "epoch": 3.0, | |
| "eval_steps": 100, | |
| "global_step": 2814, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.544431686401367, | |
| "learning_rate": 1.9822316986496094e-05, | |
| "loss": 4.5228, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.7488999366760254, | |
| "learning_rate": 1.9644633972992183e-05, | |
| "loss": 4.2438, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.0013933181762695, | |
| "learning_rate": 1.9466950959488275e-05, | |
| "loss": 3.9602, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.7357428073883057, | |
| "learning_rate": 1.9289267945984364e-05, | |
| "loss": 3.6919, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.4049, | |
| "eval_f1_macro": 0.15118582456485236, | |
| "eval_f1_micro": 0.4049, | |
| "eval_loss": 3.443889856338501, | |
| "eval_runtime": 9.0977, | |
| "eval_samples_per_second": 1099.185, | |
| "eval_steps_per_second": 17.257, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.982490062713623, | |
| "learning_rate": 1.9111584932480457e-05, | |
| "loss": 3.3679, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 3.725257396697998, | |
| "learning_rate": 1.8933901918976546e-05, | |
| "loss": 3.2251, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 3.8903579711914062, | |
| "learning_rate": 1.8756218905472638e-05, | |
| "loss": 2.946, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 4.836933612823486, | |
| "learning_rate": 1.857853589196873e-05, | |
| "loss": 2.7312, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_accuracy": 0.5766, | |
| "eval_f1_macro": 0.3024643899851631, | |
| "eval_f1_micro": 0.5766, | |
| "eval_loss": 2.576185464859009, | |
| "eval_runtime": 9.2581, | |
| "eval_samples_per_second": 1080.141, | |
| "eval_steps_per_second": 16.958, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 4.334925174713135, | |
| "learning_rate": 1.8400852878464823e-05, | |
| "loss": 2.599, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 3.9912467002868652, | |
| "learning_rate": 1.822316986496091e-05, | |
| "loss": 2.4204, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 3.9900569915771484, | |
| "learning_rate": 1.8045486851457e-05, | |
| "loss": 2.337, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.7264604568481445, | |
| "learning_rate": 1.7867803837953093e-05, | |
| "loss": 2.1872, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.656, | |
| "eval_f1_macro": 0.3994065663530276, | |
| "eval_f1_micro": 0.656, | |
| "eval_loss": 2.034592866897583, | |
| "eval_runtime": 9.3816, | |
| "eval_samples_per_second": 1065.922, | |
| "eval_steps_per_second": 16.735, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 4.161665439605713, | |
| "learning_rate": 1.7690120824449185e-05, | |
| "loss": 2.0937, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 3.8712174892425537, | |
| "learning_rate": 1.7512437810945274e-05, | |
| "loss": 1.9838, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.481960296630859, | |
| "learning_rate": 1.7334754797441367e-05, | |
| "loss": 1.869, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 3.975022077560425, | |
| "learning_rate": 1.715707178393746e-05, | |
| "loss": 1.7869, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_accuracy": 0.7075, | |
| "eval_f1_macro": 0.47958879769862484, | |
| "eval_f1_micro": 0.7075, | |
| "eval_loss": 1.6759310960769653, | |
| "eval_runtime": 9.4447, | |
| "eval_samples_per_second": 1058.8, | |
| "eval_steps_per_second": 16.623, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.881775140762329, | |
| "learning_rate": 1.6979388770433548e-05, | |
| "loss": 1.7645, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 4.351230144500732, | |
| "learning_rate": 1.6801705756929637e-05, | |
| "loss": 1.7976, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 3.5112228393554688, | |
| "learning_rate": 1.662402274342573e-05, | |
| "loss": 1.656, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.2742719650268555, | |
| "learning_rate": 1.6446339729921822e-05, | |
| "loss": 1.5593, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.7454, | |
| "eval_f1_macro": 0.5447021013489751, | |
| "eval_f1_micro": 0.7454, | |
| "eval_loss": 1.435422658920288, | |
| "eval_runtime": 9.4761, | |
| "eval_samples_per_second": 1055.289, | |
| "eval_steps_per_second": 16.568, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.7480850219726562, | |
| "learning_rate": 1.626865671641791e-05, | |
| "loss": 1.5036, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.1482491493225098, | |
| "learning_rate": 1.6090973702914003e-05, | |
| "loss": 1.4725, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 5.111358165740967, | |
| "learning_rate": 1.5913290689410095e-05, | |
| "loss": 1.4053, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.8028035163879395, | |
| "learning_rate": 1.5735607675906184e-05, | |
| "loss": 1.388, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.7695, | |
| "eval_f1_macro": 0.577809308960101, | |
| "eval_f1_micro": 0.7695, | |
| "eval_loss": 1.2758678197860718, | |
| "eval_runtime": 9.5058, | |
| "eval_samples_per_second": 1051.991, | |
| "eval_steps_per_second": 16.516, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.240127086639404, | |
| "learning_rate": 1.5557924662402277e-05, | |
| "loss": 1.3399, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.1150007247924805, | |
| "learning_rate": 1.5380241648898366e-05, | |
| "loss": 1.2714, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.628380060195923, | |
| "learning_rate": 1.5202558635394458e-05, | |
| "loss": 1.2766, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.448577880859375, | |
| "learning_rate": 1.5024875621890549e-05, | |
| "loss": 1.214, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_accuracy": 0.7806, | |
| "eval_f1_macro": 0.58907582193097, | |
| "eval_f1_micro": 0.7806, | |
| "eval_loss": 1.1428197622299194, | |
| "eval_runtime": 9.5122, | |
| "eval_samples_per_second": 1051.284, | |
| "eval_steps_per_second": 16.505, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.4761738777160645, | |
| "learning_rate": 1.484719260838664e-05, | |
| "loss": 1.1808, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.9894113540649414, | |
| "learning_rate": 1.4669509594882732e-05, | |
| "loss": 1.22, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.698829174041748, | |
| "learning_rate": 1.449182658137882e-05, | |
| "loss": 1.1912, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.163917541503906, | |
| "learning_rate": 1.4314143567874911e-05, | |
| "loss": 1.158, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_accuracy": 0.784, | |
| "eval_f1_macro": 0.5954661601031839, | |
| "eval_f1_micro": 0.784, | |
| "eval_loss": 1.053144097328186, | |
| "eval_runtime": 9.5146, | |
| "eval_samples_per_second": 1051.015, | |
| "eval_steps_per_second": 16.501, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.923873424530029, | |
| "learning_rate": 1.4136460554371004e-05, | |
| "loss": 1.1473, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.9809346199035645, | |
| "learning_rate": 1.3958777540867094e-05, | |
| "loss": 1.0783, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.601626396179199, | |
| "learning_rate": 1.3781094527363185e-05, | |
| "loss": 1.1126, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.9900403022766113, | |
| "learning_rate": 1.3603411513859277e-05, | |
| "loss": 1.0284, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.7944, | |
| "eval_f1_macro": 0.6182409797752259, | |
| "eval_f1_micro": 0.7944, | |
| "eval_loss": 0.9725863337516785, | |
| "eval_runtime": 9.5152, | |
| "eval_samples_per_second": 1050.948, | |
| "eval_steps_per_second": 16.5, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.562723636627197, | |
| "learning_rate": 1.3425728500355366e-05, | |
| "loss": 1.083, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 3.6383819580078125, | |
| "learning_rate": 1.3248045486851457e-05, | |
| "loss": 1.0053, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 4.410184860229492, | |
| "learning_rate": 1.307036247334755e-05, | |
| "loss": 0.9898, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 3.478816270828247, | |
| "learning_rate": 1.289267945984364e-05, | |
| "loss": 0.9249, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.8009, | |
| "eval_f1_macro": 0.629547630618281, | |
| "eval_f1_micro": 0.8009, | |
| "eval_loss": 0.9276089668273926, | |
| "eval_runtime": 9.5151, | |
| "eval_samples_per_second": 1050.96, | |
| "eval_steps_per_second": 16.5, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 3.290118932723999, | |
| "learning_rate": 1.271499644633973e-05, | |
| "loss": 0.947, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 4.023651123046875, | |
| "learning_rate": 1.2537313432835823e-05, | |
| "loss": 0.8897, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 4.437704563140869, | |
| "learning_rate": 1.2359630419331914e-05, | |
| "loss": 0.9272, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 4.521045684814453, | |
| "learning_rate": 1.2181947405828003e-05, | |
| "loss": 0.9046, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.8058, | |
| "eval_f1_macro": 0.6412774942056234, | |
| "eval_f1_micro": 0.8058, | |
| "eval_loss": 0.8823915123939514, | |
| "eval_runtime": 9.5141, | |
| "eval_samples_per_second": 1051.066, | |
| "eval_steps_per_second": 16.502, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.613760471343994, | |
| "learning_rate": 1.2004264392324093e-05, | |
| "loss": 0.8819, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 3.4620471000671387, | |
| "learning_rate": 1.1826581378820186e-05, | |
| "loss": 0.8386, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 5.157960891723633, | |
| "learning_rate": 1.1648898365316276e-05, | |
| "loss": 0.8049, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 4.343198299407959, | |
| "learning_rate": 1.1471215351812369e-05, | |
| "loss": 0.9312, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.8081, | |
| "eval_f1_macro": 0.6449763476669437, | |
| "eval_f1_micro": 0.8081, | |
| "eval_loss": 0.8425447344779968, | |
| "eval_runtime": 9.5201, | |
| "eval_samples_per_second": 1050.41, | |
| "eval_steps_per_second": 16.491, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 4.752044200897217, | |
| "learning_rate": 1.129353233830846e-05, | |
| "loss": 0.8832, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 2.5763158798217773, | |
| "learning_rate": 1.111584932480455e-05, | |
| "loss": 0.8849, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 3.2431178092956543, | |
| "learning_rate": 1.0938166311300639e-05, | |
| "loss": 0.8374, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 5.601613521575928, | |
| "learning_rate": 1.0760483297796731e-05, | |
| "loss": 0.8329, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.8135, | |
| "eval_f1_macro": 0.6585048816034432, | |
| "eval_f1_micro": 0.8135, | |
| "eval_loss": 0.8095632791519165, | |
| "eval_runtime": 9.5177, | |
| "eval_samples_per_second": 1050.673, | |
| "eval_steps_per_second": 16.496, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 4.541258335113525, | |
| "learning_rate": 1.0582800284292822e-05, | |
| "loss": 0.8391, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 3.400801420211792, | |
| "learning_rate": 1.0405117270788913e-05, | |
| "loss": 0.8404, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 3.5427093505859375, | |
| "learning_rate": 1.0227434257285005e-05, | |
| "loss": 0.7576, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 5.988525867462158, | |
| "learning_rate": 1.0049751243781096e-05, | |
| "loss": 0.7601, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_accuracy": 0.8148, | |
| "eval_f1_macro": 0.6646068863066573, | |
| "eval_f1_micro": 0.8148, | |
| "eval_loss": 0.7945957183837891, | |
| "eval_runtime": 9.5183, | |
| "eval_samples_per_second": 1050.605, | |
| "eval_steps_per_second": 16.495, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 6.576759338378906, | |
| "learning_rate": 9.872068230277187e-06, | |
| "loss": 0.8336, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 4.413715362548828, | |
| "learning_rate": 9.694385216773277e-06, | |
| "loss": 0.8282, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 3.664360523223877, | |
| "learning_rate": 9.516702203269368e-06, | |
| "loss": 0.8263, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.517057180404663, | |
| "learning_rate": 9.339019189765458e-06, | |
| "loss": 0.7812, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.8192, | |
| "eval_f1_macro": 0.6738857544513059, | |
| "eval_f1_micro": 0.8192, | |
| "eval_loss": 0.776592493057251, | |
| "eval_runtime": 9.5181, | |
| "eval_samples_per_second": 1050.631, | |
| "eval_steps_per_second": 16.495, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 3.894787311553955, | |
| "learning_rate": 9.161336176261551e-06, | |
| "loss": 0.7565, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.5228681564331055, | |
| "learning_rate": 8.98365316275764e-06, | |
| "loss": 0.735, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 4.373218059539795, | |
| "learning_rate": 8.805970149253732e-06, | |
| "loss": 0.7089, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 3.751723289489746, | |
| "learning_rate": 8.628287135749823e-06, | |
| "loss": 0.7944, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.8221, | |
| "eval_f1_macro": 0.6799558248118867, | |
| "eval_f1_micro": 0.8221, | |
| "eval_loss": 0.7584841847419739, | |
| "eval_runtime": 9.5135, | |
| "eval_samples_per_second": 1051.143, | |
| "eval_steps_per_second": 16.503, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 3.9183712005615234, | |
| "learning_rate": 8.450604122245915e-06, | |
| "loss": 0.757, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 4.843607425689697, | |
| "learning_rate": 8.272921108742004e-06, | |
| "loss": 0.7617, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 3.931086301803589, | |
| "learning_rate": 8.095238095238097e-06, | |
| "loss": 0.7165, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 4.115772247314453, | |
| "learning_rate": 7.917555081734187e-06, | |
| "loss": 0.7632, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_accuracy": 0.8269, | |
| "eval_f1_macro": 0.6901693836898547, | |
| "eval_f1_micro": 0.8269, | |
| "eval_loss": 0.7362803816795349, | |
| "eval_runtime": 9.518, | |
| "eval_samples_per_second": 1050.64, | |
| "eval_steps_per_second": 16.495, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 5.484522819519043, | |
| "learning_rate": 7.739872068230278e-06, | |
| "loss": 0.7934, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 3.305182456970215, | |
| "learning_rate": 7.5621890547263685e-06, | |
| "loss": 0.72, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 3.8824164867401123, | |
| "learning_rate": 7.38450604122246e-06, | |
| "loss": 0.7582, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 4.005168437957764, | |
| "learning_rate": 7.20682302771855e-06, | |
| "loss": 0.7027, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.8227, | |
| "eval_f1_macro": 0.6793160133132589, | |
| "eval_f1_micro": 0.8227, | |
| "eval_loss": 0.7229283452033997, | |
| "eval_runtime": 9.5098, | |
| "eval_samples_per_second": 1051.543, | |
| "eval_steps_per_second": 16.509, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 3.089484214782715, | |
| "learning_rate": 7.029140014214641e-06, | |
| "loss": 0.826, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 3.8025612831115723, | |
| "learning_rate": 6.851457000710733e-06, | |
| "loss": 0.7889, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.959663391113281, | |
| "learning_rate": 6.673773987206824e-06, | |
| "loss": 0.6852, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 4.049121856689453, | |
| "learning_rate": 6.496090973702914e-06, | |
| "loss": 0.671, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.8263, | |
| "eval_f1_macro": 0.6869823704829469, | |
| "eval_f1_micro": 0.8263, | |
| "eval_loss": 0.7144606709480286, | |
| "eval_runtime": 9.5164, | |
| "eval_samples_per_second": 1050.819, | |
| "eval_steps_per_second": 16.498, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 3.187711000442505, | |
| "learning_rate": 6.318407960199006e-06, | |
| "loss": 0.6938, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 3.601215124130249, | |
| "learning_rate": 6.140724946695097e-06, | |
| "loss": 0.7311, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 3.0071964263916016, | |
| "learning_rate": 5.963041933191187e-06, | |
| "loss": 0.6693, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 4.674365997314453, | |
| "learning_rate": 5.785358919687279e-06, | |
| "loss": 0.6361, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_accuracy": 0.8277, | |
| "eval_f1_macro": 0.6952132327426773, | |
| "eval_f1_micro": 0.8277, | |
| "eval_loss": 0.7066932320594788, | |
| "eval_runtime": 9.5145, | |
| "eval_samples_per_second": 1051.024, | |
| "eval_steps_per_second": 16.501, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 3.853187084197998, | |
| "learning_rate": 5.607675906183369e-06, | |
| "loss": 0.6644, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 3.0457775592803955, | |
| "learning_rate": 5.42999289267946e-06, | |
| "loss": 0.6664, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 4.993578910827637, | |
| "learning_rate": 5.2523098791755514e-06, | |
| "loss": 0.6586, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 4.2673797607421875, | |
| "learning_rate": 5.074626865671642e-06, | |
| "loss": 0.6615, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_accuracy": 0.8281, | |
| "eval_f1_macro": 0.6974353392237495, | |
| "eval_f1_micro": 0.8281, | |
| "eval_loss": 0.6969488859176636, | |
| "eval_runtime": 9.5132, | |
| "eval_samples_per_second": 1051.175, | |
| "eval_steps_per_second": 16.503, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 3.5754830837249756, | |
| "learning_rate": 4.896943852167733e-06, | |
| "loss": 0.6395, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 2.7557778358459473, | |
| "learning_rate": 4.719260838663824e-06, | |
| "loss": 0.6297, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.421337604522705, | |
| "learning_rate": 4.541577825159915e-06, | |
| "loss": 0.5636, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 5.7008514404296875, | |
| "learning_rate": 4.363894811656006e-06, | |
| "loss": 0.6432, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_accuracy": 0.8311, | |
| "eval_f1_macro": 0.705416263146119, | |
| "eval_f1_micro": 0.8311, | |
| "eval_loss": 0.690848171710968, | |
| "eval_runtime": 9.5147, | |
| "eval_samples_per_second": 1051.01, | |
| "eval_steps_per_second": 16.501, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 6.569701194763184, | |
| "learning_rate": 4.186211798152097e-06, | |
| "loss": 0.5678, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 3.0559256076812744, | |
| "learning_rate": 4.008528784648188e-06, | |
| "loss": 0.6404, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 3.7182412147521973, | |
| "learning_rate": 3.8308457711442784e-06, | |
| "loss": 0.654, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 4.205299377441406, | |
| "learning_rate": 3.65316275764037e-06, | |
| "loss": 0.648, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_accuracy": 0.8304, | |
| "eval_f1_macro": 0.7010940078357385, | |
| "eval_f1_micro": 0.8304, | |
| "eval_loss": 0.6850333213806152, | |
| "eval_runtime": 9.5114, | |
| "eval_samples_per_second": 1051.365, | |
| "eval_steps_per_second": 16.506, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 4.695363521575928, | |
| "learning_rate": 3.4754797441364606e-06, | |
| "loss": 0.6588, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 5.2972412109375, | |
| "learning_rate": 3.297796730632552e-06, | |
| "loss": 0.675, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 3.7516212463378906, | |
| "learning_rate": 3.120113717128643e-06, | |
| "loss": 0.6063, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 4.958125591278076, | |
| "learning_rate": 2.9424307036247335e-06, | |
| "loss": 0.631, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.8323, | |
| "eval_f1_macro": 0.7063424299231434, | |
| "eval_f1_micro": 0.8323, | |
| "eval_loss": 0.6749587655067444, | |
| "eval_runtime": 9.5144, | |
| "eval_samples_per_second": 1051.039, | |
| "eval_steps_per_second": 16.501, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 3.3450634479522705, | |
| "learning_rate": 2.764747690120825e-06, | |
| "loss": 0.7092, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 3.6644909381866455, | |
| "learning_rate": 2.5870646766169156e-06, | |
| "loss": 0.673, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 4.698973655700684, | |
| "learning_rate": 2.4093816631130067e-06, | |
| "loss": 0.6272, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 5.635212421417236, | |
| "learning_rate": 2.2316986496090974e-06, | |
| "loss": 0.575, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.8337, | |
| "eval_f1_macro": 0.7093935171505621, | |
| "eval_f1_micro": 0.8337, | |
| "eval_loss": 0.6717804670333862, | |
| "eval_runtime": 9.5156, | |
| "eval_samples_per_second": 1050.905, | |
| "eval_steps_per_second": 16.499, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 4.278409957885742, | |
| "learning_rate": 2.0540156361051885e-06, | |
| "loss": 0.6282, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 4.368821620941162, | |
| "learning_rate": 1.8763326226012796e-06, | |
| "loss": 0.6182, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 6.596986770629883, | |
| "learning_rate": 1.6986496090973702e-06, | |
| "loss": 0.6291, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 5.670820713043213, | |
| "learning_rate": 1.5209665955934613e-06, | |
| "loss": 0.6444, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.8332, | |
| "eval_f1_macro": 0.7102434229943433, | |
| "eval_f1_micro": 0.8332, | |
| "eval_loss": 0.6701115965843201, | |
| "eval_runtime": 9.5227, | |
| "eval_samples_per_second": 1050.123, | |
| "eval_steps_per_second": 16.487, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 3.570100784301758, | |
| "learning_rate": 1.3432835820895524e-06, | |
| "loss": 0.6071, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 5.6983256340026855, | |
| "learning_rate": 1.1656005685856433e-06, | |
| "loss": 0.6638, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 5.448667049407959, | |
| "learning_rate": 9.879175550817342e-07, | |
| "loss": 0.6086, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 4.07735538482666, | |
| "learning_rate": 8.102345415778253e-07, | |
| "loss": 0.6054, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.8346, | |
| "eval_f1_macro": 0.7121645746905745, | |
| "eval_f1_micro": 0.8346, | |
| "eval_loss": 0.6689808368682861, | |
| "eval_runtime": 9.5154, | |
| "eval_samples_per_second": 1050.924, | |
| "eval_steps_per_second": 16.5, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 3.3793234825134277, | |
| "learning_rate": 6.325515280739161e-07, | |
| "loss": 0.5974, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 3.8540170192718506, | |
| "learning_rate": 4.548685145700071e-07, | |
| "loss": 0.5927, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 5.523347854614258, | |
| "learning_rate": 2.771855010660981e-07, | |
| "loss": 0.5564, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 4.380491256713867, | |
| "learning_rate": 9.950248756218906e-08, | |
| "loss": 0.6123, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.8349, | |
| "eval_f1_macro": 0.7127429197005123, | |
| "eval_f1_micro": 0.8349, | |
| "eval_loss": 0.6676496267318726, | |
| "eval_runtime": 9.5248, | |
| "eval_samples_per_second": 1049.895, | |
| "eval_steps_per_second": 16.483, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2814, | |
| "total_flos": 1.1856735614533632e+16, | |
| "train_loss": 1.1454540061543999, | |
| "train_runtime": 819.1516, | |
| "train_samples_per_second": 219.74, | |
| "train_steps_per_second": 3.435 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2814, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 1.1856735614533632e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |