SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/distilbert
/distilroberta_base_ledgar
/checkpoint-2800
/trainer_state.json
| { | |
| "best_metric": 0.672819197177887, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/distilbert/distilroberta_base_ledgar/checkpoint-2800", | |
| "epoch": 2.9850746268656714, | |
| "eval_steps": 100, | |
| "global_step": 2800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.8299777507781982, | |
| "learning_rate": 1.9822316986496094e-05, | |
| "loss": 4.5153, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.934743642807007, | |
| "learning_rate": 1.9644633972992183e-05, | |
| "loss": 4.148, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.940049171447754, | |
| "learning_rate": 1.9466950959488275e-05, | |
| "loss": 3.7445, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.678694725036621, | |
| "learning_rate": 1.9289267945984364e-05, | |
| "loss": 3.3872, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.45, | |
| "eval_f1_macro": 0.19509310055758317, | |
| "eval_f1_micro": 0.45, | |
| "eval_loss": 3.058864116668701, | |
| "eval_runtime": 5.0293, | |
| "eval_samples_per_second": 1988.356, | |
| "eval_steps_per_second": 31.217, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.43796968460083, | |
| "learning_rate": 1.9111584932480457e-05, | |
| "loss": 3.0273, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.942129135131836, | |
| "learning_rate": 1.8933901918976546e-05, | |
| "loss": 2.8672, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 5.8119330406188965, | |
| "learning_rate": 1.8756218905472638e-05, | |
| "loss": 2.5939, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 6.159755706787109, | |
| "learning_rate": 1.857853589196873e-05, | |
| "loss": 2.4052, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_accuracy": 0.6026, | |
| "eval_f1_macro": 0.32482211454761123, | |
| "eval_f1_micro": 0.6026, | |
| "eval_loss": 2.1783194541931152, | |
| "eval_runtime": 5.0585, | |
| "eval_samples_per_second": 1976.885, | |
| "eval_steps_per_second": 31.037, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 5.579538822174072, | |
| "learning_rate": 1.8400852878464823e-05, | |
| "loss": 2.2578, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 5.426351070404053, | |
| "learning_rate": 1.822316986496091e-05, | |
| "loss": 2.1216, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 4.704029083251953, | |
| "learning_rate": 1.8045486851457e-05, | |
| "loss": 2.0376, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 6.021806716918945, | |
| "learning_rate": 1.7867803837953093e-05, | |
| "loss": 1.9345, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.6639, | |
| "eval_f1_macro": 0.4073136970879148, | |
| "eval_f1_micro": 0.6639, | |
| "eval_loss": 1.7352277040481567, | |
| "eval_runtime": 5.0098, | |
| "eval_samples_per_second": 1996.069, | |
| "eval_steps_per_second": 31.338, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 5.859042167663574, | |
| "learning_rate": 1.7690120824449185e-05, | |
| "loss": 1.8566, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 4.362372875213623, | |
| "learning_rate": 1.7512437810945274e-05, | |
| "loss": 1.7492, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.875273704528809, | |
| "learning_rate": 1.7334754797441367e-05, | |
| "loss": 1.6679, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 4.909628391265869, | |
| "learning_rate": 1.715707178393746e-05, | |
| "loss": 1.605, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_accuracy": 0.7138, | |
| "eval_f1_macro": 0.48174753914501256, | |
| "eval_f1_micro": 0.7138, | |
| "eval_loss": 1.4555104970932007, | |
| "eval_runtime": 5.1099, | |
| "eval_samples_per_second": 1956.981, | |
| "eval_steps_per_second": 30.725, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 4.944831848144531, | |
| "learning_rate": 1.6979388770433548e-05, | |
| "loss": 1.5949, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.594886302947998, | |
| "learning_rate": 1.6801705756929637e-05, | |
| "loss": 1.6192, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.38767671585083, | |
| "learning_rate": 1.662402274342573e-05, | |
| "loss": 1.4944, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.888990879058838, | |
| "learning_rate": 1.6446339729921822e-05, | |
| "loss": 1.4359, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.7429, | |
| "eval_f1_macro": 0.5381654949510911, | |
| "eval_f1_micro": 0.7429, | |
| "eval_loss": 1.2634148597717285, | |
| "eval_runtime": 5.117, | |
| "eval_samples_per_second": 1954.28, | |
| "eval_steps_per_second": 30.682, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.6373090744018555, | |
| "learning_rate": 1.626865671641791e-05, | |
| "loss": 1.3848, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.159140586853027, | |
| "learning_rate": 1.6090973702914003e-05, | |
| "loss": 1.3537, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 6.624241828918457, | |
| "learning_rate": 1.5913290689410095e-05, | |
| "loss": 1.2896, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 6.4602861404418945, | |
| "learning_rate": 1.5735607675906184e-05, | |
| "loss": 1.3107, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.7678, | |
| "eval_f1_macro": 0.5849471635603258, | |
| "eval_f1_micro": 0.7678, | |
| "eval_loss": 1.139076828956604, | |
| "eval_runtime": 5.1187, | |
| "eval_samples_per_second": 1953.636, | |
| "eval_steps_per_second": 30.672, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 6.463939666748047, | |
| "learning_rate": 1.5557924662402277e-05, | |
| "loss": 1.2547, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.371650218963623, | |
| "learning_rate": 1.5380241648898366e-05, | |
| "loss": 1.2086, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 6.18757963180542, | |
| "learning_rate": 1.5202558635394458e-05, | |
| "loss": 1.2102, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 5.33036994934082, | |
| "learning_rate": 1.5024875621890549e-05, | |
| "loss": 1.1656, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_accuracy": 0.7775, | |
| "eval_f1_macro": 0.6011798134215889, | |
| "eval_f1_micro": 0.7775, | |
| "eval_loss": 1.047299861907959, | |
| "eval_runtime": 5.1166, | |
| "eval_samples_per_second": 1954.433, | |
| "eval_steps_per_second": 30.685, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.348170280456543, | |
| "learning_rate": 1.484719260838664e-05, | |
| "loss": 1.0935, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 8.08990478515625, | |
| "learning_rate": 1.4669509594882732e-05, | |
| "loss": 1.1488, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 5.688432693481445, | |
| "learning_rate": 1.449182658137882e-05, | |
| "loss": 1.1557, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.804388999938965, | |
| "learning_rate": 1.4314143567874911e-05, | |
| "loss": 1.1157, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_accuracy": 0.7801, | |
| "eval_f1_macro": 0.605385060800832, | |
| "eval_f1_micro": 0.7801, | |
| "eval_loss": 0.9756803512573242, | |
| "eval_runtime": 5.1185, | |
| "eval_samples_per_second": 1953.693, | |
| "eval_steps_per_second": 30.673, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 7.553802013397217, | |
| "learning_rate": 1.4136460554371004e-05, | |
| "loss": 1.0931, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.129767417907715, | |
| "learning_rate": 1.3958777540867094e-05, | |
| "loss": 1.0363, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 6.337955474853516, | |
| "learning_rate": 1.3781094527363185e-05, | |
| "loss": 1.0583, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.533592224121094, | |
| "learning_rate": 1.3603411513859277e-05, | |
| "loss": 1.0035, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.7934, | |
| "eval_f1_macro": 0.6255915268745559, | |
| "eval_f1_micro": 0.7934, | |
| "eval_loss": 0.9160490036010742, | |
| "eval_runtime": 5.1131, | |
| "eval_samples_per_second": 1955.745, | |
| "eval_steps_per_second": 30.705, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 5.8523149490356445, | |
| "learning_rate": 1.3425728500355366e-05, | |
| "loss": 1.0513, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 5.070196628570557, | |
| "learning_rate": 1.3248045486851457e-05, | |
| "loss": 0.984, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 5.693618297576904, | |
| "learning_rate": 1.307036247334755e-05, | |
| "loss": 0.975, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 4.580042839050293, | |
| "learning_rate": 1.289267945984364e-05, | |
| "loss": 0.9232, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.8008, | |
| "eval_f1_macro": 0.6363564825491357, | |
| "eval_f1_micro": 0.8008, | |
| "eval_loss": 0.8696697354316711, | |
| "eval_runtime": 5.1084, | |
| "eval_samples_per_second": 1957.575, | |
| "eval_steps_per_second": 30.734, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 4.93649959564209, | |
| "learning_rate": 1.271499644633973e-05, | |
| "loss": 0.9191, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 6.813359260559082, | |
| "learning_rate": 1.2537313432835823e-05, | |
| "loss": 0.885, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 7.439708709716797, | |
| "learning_rate": 1.2359630419331914e-05, | |
| "loss": 0.9192, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 6.148406028747559, | |
| "learning_rate": 1.2181947405828003e-05, | |
| "loss": 0.9007, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.8057, | |
| "eval_f1_macro": 0.6479322011204249, | |
| "eval_f1_micro": 0.8057, | |
| "eval_loss": 0.8374300599098206, | |
| "eval_runtime": 5.1075, | |
| "eval_samples_per_second": 1957.894, | |
| "eval_steps_per_second": 30.739, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 4.837780952453613, | |
| "learning_rate": 1.2004264392324093e-05, | |
| "loss": 0.8591, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 6.624153137207031, | |
| "learning_rate": 1.1826581378820186e-05, | |
| "loss": 0.8512, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 5.78391170501709, | |
| "learning_rate": 1.1648898365316276e-05, | |
| "loss": 0.8102, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 4.467207908630371, | |
| "learning_rate": 1.1471215351812369e-05, | |
| "loss": 0.9422, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.8078, | |
| "eval_f1_macro": 0.654171462092156, | |
| "eval_f1_micro": 0.8078, | |
| "eval_loss": 0.8185257315635681, | |
| "eval_runtime": 5.1053, | |
| "eval_samples_per_second": 1958.731, | |
| "eval_steps_per_second": 30.752, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 6.303857326507568, | |
| "learning_rate": 1.129353233830846e-05, | |
| "loss": 0.8991, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 4.74877405166626, | |
| "learning_rate": 1.111584932480455e-05, | |
| "loss": 0.8771, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 6.6821112632751465, | |
| "learning_rate": 1.0938166311300639e-05, | |
| "loss": 0.8626, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 7.011251449584961, | |
| "learning_rate": 1.0760483297796731e-05, | |
| "loss": 0.8607, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.8093, | |
| "eval_f1_macro": 0.6592931556512984, | |
| "eval_f1_micro": 0.8093, | |
| "eval_loss": 0.7932953238487244, | |
| "eval_runtime": 5.1058, | |
| "eval_samples_per_second": 1958.544, | |
| "eval_steps_per_second": 30.749, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 7.529996871948242, | |
| "learning_rate": 1.0582800284292822e-05, | |
| "loss": 0.8422, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 4.459428310394287, | |
| "learning_rate": 1.0405117270788913e-05, | |
| "loss": 0.843, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 6.408772945404053, | |
| "learning_rate": 1.0227434257285005e-05, | |
| "loss": 0.7783, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 4.889123916625977, | |
| "learning_rate": 1.0049751243781096e-05, | |
| "loss": 0.7426, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_accuracy": 0.8098, | |
| "eval_f1_macro": 0.6654458516165033, | |
| "eval_f1_micro": 0.8098, | |
| "eval_loss": 0.7752991318702698, | |
| "eval_runtime": 5.0975, | |
| "eval_samples_per_second": 1961.743, | |
| "eval_steps_per_second": 30.799, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 6.442604064941406, | |
| "learning_rate": 9.872068230277187e-06, | |
| "loss": 0.8288, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 7.792954921722412, | |
| "learning_rate": 9.694385216773277e-06, | |
| "loss": 0.8519, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 4.959009647369385, | |
| "learning_rate": 9.516702203269368e-06, | |
| "loss": 0.8311, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 5.443413734436035, | |
| "learning_rate": 9.339019189765458e-06, | |
| "loss": 0.7741, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.8122, | |
| "eval_f1_macro": 0.6665748076457181, | |
| "eval_f1_micro": 0.8122, | |
| "eval_loss": 0.7569247484207153, | |
| "eval_runtime": 5.1009, | |
| "eval_samples_per_second": 1960.425, | |
| "eval_steps_per_second": 30.779, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 7.319786548614502, | |
| "learning_rate": 9.161336176261551e-06, | |
| "loss": 0.7763, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 6.09261417388916, | |
| "learning_rate": 8.98365316275764e-06, | |
| "loss": 0.7427, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 7.664770126342773, | |
| "learning_rate": 8.805970149253732e-06, | |
| "loss": 0.7751, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 8.841761589050293, | |
| "learning_rate": 8.628287135749823e-06, | |
| "loss": 0.8094, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.8184, | |
| "eval_f1_macro": 0.6772511050694777, | |
| "eval_f1_micro": 0.8184, | |
| "eval_loss": 0.7387501001358032, | |
| "eval_runtime": 5.0945, | |
| "eval_samples_per_second": 1962.907, | |
| "eval_steps_per_second": 30.818, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 8.24087142944336, | |
| "learning_rate": 8.450604122245915e-06, | |
| "loss": 0.7786, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 6.391561985015869, | |
| "learning_rate": 8.272921108742004e-06, | |
| "loss": 0.7782, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 4.629056453704834, | |
| "learning_rate": 8.095238095238097e-06, | |
| "loss": 0.7167, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 9.185173988342285, | |
| "learning_rate": 7.917555081734187e-06, | |
| "loss": 0.7809, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_accuracy": 0.8172, | |
| "eval_f1_macro": 0.6789051969285573, | |
| "eval_f1_micro": 0.8172, | |
| "eval_loss": 0.7321203947067261, | |
| "eval_runtime": 5.0944, | |
| "eval_samples_per_second": 1962.951, | |
| "eval_steps_per_second": 30.818, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 7.9137983322143555, | |
| "learning_rate": 7.739872068230278e-06, | |
| "loss": 0.8081, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 5.078384876251221, | |
| "learning_rate": 7.5621890547263685e-06, | |
| "loss": 0.7314, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 7.1659135818481445, | |
| "learning_rate": 7.38450604122246e-06, | |
| "loss": 0.8053, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 10.980244636535645, | |
| "learning_rate": 7.20682302771855e-06, | |
| "loss": 0.7435, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.8182, | |
| "eval_f1_macro": 0.6774819408695893, | |
| "eval_f1_micro": 0.8182, | |
| "eval_loss": 0.7197924852371216, | |
| "eval_runtime": 5.0921, | |
| "eval_samples_per_second": 1963.834, | |
| "eval_steps_per_second": 30.832, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 5.948391437530518, | |
| "learning_rate": 7.029140014214641e-06, | |
| "loss": 0.8447, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 7.452664375305176, | |
| "learning_rate": 6.851457000710733e-06, | |
| "loss": 0.7911, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.688015937805176, | |
| "learning_rate": 6.673773987206824e-06, | |
| "loss": 0.7016, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 5.259037494659424, | |
| "learning_rate": 6.496090973702914e-06, | |
| "loss": 0.718, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.8201, | |
| "eval_f1_macro": 0.6809720289529079, | |
| "eval_f1_micro": 0.8201, | |
| "eval_loss": 0.7103058695793152, | |
| "eval_runtime": 5.0899, | |
| "eval_samples_per_second": 1964.692, | |
| "eval_steps_per_second": 30.846, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 4.622197151184082, | |
| "learning_rate": 6.318407960199006e-06, | |
| "loss": 0.7293, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 6.066267967224121, | |
| "learning_rate": 6.140724946695097e-06, | |
| "loss": 0.7588, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 6.059533596038818, | |
| "learning_rate": 5.963041933191187e-06, | |
| "loss": 0.7004, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 7.258752822875977, | |
| "learning_rate": 5.785358919687279e-06, | |
| "loss": 0.6816, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_accuracy": 0.8208, | |
| "eval_f1_macro": 0.6827958454764386, | |
| "eval_f1_micro": 0.8208, | |
| "eval_loss": 0.7005710601806641, | |
| "eval_runtime": 5.0873, | |
| "eval_samples_per_second": 1965.668, | |
| "eval_steps_per_second": 30.861, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 8.9718599319458, | |
| "learning_rate": 5.607675906183369e-06, | |
| "loss": 0.7054, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 8.166780471801758, | |
| "learning_rate": 5.42999289267946e-06, | |
| "loss": 0.7144, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 7.228323936462402, | |
| "learning_rate": 5.2523098791755514e-06, | |
| "loss": 0.6805, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 7.857689380645752, | |
| "learning_rate": 5.074626865671642e-06, | |
| "loss": 0.7262, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_accuracy": 0.8233, | |
| "eval_f1_macro": 0.6906574822961721, | |
| "eval_f1_micro": 0.8233, | |
| "eval_loss": 0.6981844902038574, | |
| "eval_runtime": 5.0973, | |
| "eval_samples_per_second": 1961.807, | |
| "eval_steps_per_second": 30.8, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 5.485396385192871, | |
| "learning_rate": 4.896943852167733e-06, | |
| "loss": 0.6582, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 4.579333305358887, | |
| "learning_rate": 4.719260838663824e-06, | |
| "loss": 0.6779, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 7.446170806884766, | |
| "learning_rate": 4.541577825159915e-06, | |
| "loss": 0.6143, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 7.771754264831543, | |
| "learning_rate": 4.363894811656006e-06, | |
| "loss": 0.683, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_accuracy": 0.8244, | |
| "eval_f1_macro": 0.6917258704514068, | |
| "eval_f1_micro": 0.8244, | |
| "eval_loss": 0.6932203769683838, | |
| "eval_runtime": 5.0891, | |
| "eval_samples_per_second": 1964.996, | |
| "eval_steps_per_second": 30.85, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 9.430304527282715, | |
| "learning_rate": 4.186211798152097e-06, | |
| "loss": 0.6406, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 5.412100791931152, | |
| "learning_rate": 4.008528784648188e-06, | |
| "loss": 0.6671, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 5.439154148101807, | |
| "learning_rate": 3.8308457711442784e-06, | |
| "loss": 0.6986, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 5.835467338562012, | |
| "learning_rate": 3.65316275764037e-06, | |
| "loss": 0.6892, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_accuracy": 0.8238, | |
| "eval_f1_macro": 0.6901863834091672, | |
| "eval_f1_micro": 0.8238, | |
| "eval_loss": 0.6871231198310852, | |
| "eval_runtime": 5.083, | |
| "eval_samples_per_second": 1967.334, | |
| "eval_steps_per_second": 30.887, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 6.965566158294678, | |
| "learning_rate": 3.4754797441364606e-06, | |
| "loss": 0.694, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 9.316426277160645, | |
| "learning_rate": 3.297796730632552e-06, | |
| "loss": 0.7061, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 6.5966386795043945, | |
| "learning_rate": 3.120113717128643e-06, | |
| "loss": 0.6392, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 6.941636085510254, | |
| "learning_rate": 2.9424307036247335e-06, | |
| "loss": 0.6712, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.8271, | |
| "eval_f1_macro": 0.6974830323546121, | |
| "eval_f1_micro": 0.8271, | |
| "eval_loss": 0.6783324480056763, | |
| "eval_runtime": 5.0875, | |
| "eval_samples_per_second": 1965.612, | |
| "eval_steps_per_second": 30.86, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 4.662513256072998, | |
| "learning_rate": 2.764747690120825e-06, | |
| "loss": 0.7543, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 8.454292297363281, | |
| "learning_rate": 2.5870646766169156e-06, | |
| "loss": 0.7188, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 6.374271869659424, | |
| "learning_rate": 2.4093816631130067e-06, | |
| "loss": 0.6648, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 6.783421516418457, | |
| "learning_rate": 2.2316986496090974e-06, | |
| "loss": 0.6442, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.8263, | |
| "eval_f1_macro": 0.693810944994602, | |
| "eval_f1_micro": 0.8263, | |
| "eval_loss": 0.6761018633842468, | |
| "eval_runtime": 5.0892, | |
| "eval_samples_per_second": 1964.949, | |
| "eval_steps_per_second": 30.85, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 5.442586898803711, | |
| "learning_rate": 2.0540156361051885e-06, | |
| "loss": 0.6759, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 7.995906352996826, | |
| "learning_rate": 1.8763326226012796e-06, | |
| "loss": 0.6666, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 9.21601390838623, | |
| "learning_rate": 1.6986496090973702e-06, | |
| "loss": 0.6662, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 5.839861869812012, | |
| "learning_rate": 1.5209665955934613e-06, | |
| "loss": 0.6847, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.8258, | |
| "eval_f1_macro": 0.6953621305839427, | |
| "eval_f1_micro": 0.8258, | |
| "eval_loss": 0.6750516891479492, | |
| "eval_runtime": 5.0796, | |
| "eval_samples_per_second": 1968.678, | |
| "eval_steps_per_second": 30.908, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 5.698036193847656, | |
| "learning_rate": 1.3432835820895524e-06, | |
| "loss": 0.658, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 9.440896987915039, | |
| "learning_rate": 1.1656005685856433e-06, | |
| "loss": 0.7268, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 4.6192708015441895, | |
| "learning_rate": 9.879175550817342e-07, | |
| "loss": 0.6497, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 7.211236000061035, | |
| "learning_rate": 8.102345415778253e-07, | |
| "loss": 0.6466, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.8264, | |
| "eval_f1_macro": 0.6959825079971337, | |
| "eval_f1_micro": 0.8264, | |
| "eval_loss": 0.6746246814727783, | |
| "eval_runtime": 5.0794, | |
| "eval_samples_per_second": 1968.741, | |
| "eval_steps_per_second": 30.909, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 5.404388904571533, | |
| "learning_rate": 6.325515280739161e-07, | |
| "loss": 0.6732, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 8.992890357971191, | |
| "learning_rate": 4.548685145700071e-07, | |
| "loss": 0.655, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 8.38559627532959, | |
| "learning_rate": 2.771855010660981e-07, | |
| "loss": 0.6117, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 4.968321800231934, | |
| "learning_rate": 9.950248756218906e-08, | |
| "loss": 0.6402, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.8262, | |
| "eval_f1_macro": 0.6966969836943864, | |
| "eval_f1_micro": 0.8262, | |
| "eval_loss": 0.672819197177887, | |
| "eval_runtime": 5.0819, | |
| "eval_samples_per_second": 1967.766, | |
| "eval_steps_per_second": 30.894, | |
| "step": 2800 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2814, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 5944911060795392.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |