|
{ |
|
"best_metric": 0.6496002674102783, |
|
"best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/distilbert/distilbert_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": 1.8292617797851562, |
|
"learning_rate": 1.9822316986496094e-05, |
|
"loss": 4.5108, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 2.0646250247955322, |
|
"learning_rate": 1.9644633972992183e-05, |
|
"loss": 4.2737, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 2.6037521362304688, |
|
"learning_rate": 1.9466950959488275e-05, |
|
"loss": 4.06, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 2.8344316482543945, |
|
"learning_rate": 1.9289267945984364e-05, |
|
"loss": 3.8165, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"eval_accuracy": 0.3489, |
|
"eval_f1_macro": 0.09953942962664121, |
|
"eval_f1_micro": 0.3489, |
|
"eval_loss": 3.5951926708221436, |
|
"eval_runtime": 5.032, |
|
"eval_samples_per_second": 1987.289, |
|
"eval_steps_per_second": 31.2, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 2.8050551414489746, |
|
"learning_rate": 1.9111584932480457e-05, |
|
"loss": 3.499, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 2.6794557571411133, |
|
"learning_rate": 1.8933901918976546e-05, |
|
"loss": 3.3343, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 3.4706735610961914, |
|
"learning_rate": 1.8756218905472638e-05, |
|
"loss": 3.0543, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 2.9294159412384033, |
|
"learning_rate": 1.857853589196873e-05, |
|
"loss": 2.8293, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"eval_accuracy": 0.5385, |
|
"eval_f1_macro": 0.2374650407061841, |
|
"eval_f1_micro": 0.5385, |
|
"eval_loss": 2.6737160682678223, |
|
"eval_runtime": 5.0548, |
|
"eval_samples_per_second": 1978.336, |
|
"eval_steps_per_second": 31.06, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 3.2487082481384277, |
|
"learning_rate": 1.8400852878464823e-05, |
|
"loss": 2.6772, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 3.0436863899230957, |
|
"learning_rate": 1.822316986496091e-05, |
|
"loss": 2.4908, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 3.3318707942962646, |
|
"learning_rate": 1.8045486851457e-05, |
|
"loss": 2.3918, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 3.209085464477539, |
|
"learning_rate": 1.7867803837953093e-05, |
|
"loss": 2.2564, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"eval_accuracy": 0.6212, |
|
"eval_f1_macro": 0.3338736755701111, |
|
"eval_f1_micro": 0.6212, |
|
"eval_loss": 2.0960097312927246, |
|
"eval_runtime": 5.0922, |
|
"eval_samples_per_second": 1963.795, |
|
"eval_steps_per_second": 30.832, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 3.3769121170043945, |
|
"learning_rate": 1.7690120824449185e-05, |
|
"loss": 2.1476, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 3.443783760070801, |
|
"learning_rate": 1.7512437810945274e-05, |
|
"loss": 2.0204, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 3.6128618717193604, |
|
"learning_rate": 1.7334754797441367e-05, |
|
"loss": 1.9127, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 3.1673097610473633, |
|
"learning_rate": 1.715707178393746e-05, |
|
"loss": 1.8259, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"eval_accuracy": 0.6792, |
|
"eval_f1_macro": 0.42690318122732035, |
|
"eval_f1_micro": 0.6792, |
|
"eval_loss": 1.711798071861267, |
|
"eval_runtime": 5.0988, |
|
"eval_samples_per_second": 1961.262, |
|
"eval_steps_per_second": 30.792, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 3.0291733741760254, |
|
"learning_rate": 1.6979388770433548e-05, |
|
"loss": 1.7889, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 3.4015133380889893, |
|
"learning_rate": 1.6801705756929637e-05, |
|
"loss": 1.8087, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 3.399096965789795, |
|
"learning_rate": 1.662402274342573e-05, |
|
"loss": 1.6702, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 3.8419103622436523, |
|
"learning_rate": 1.6446339729921822e-05, |
|
"loss": 1.5846, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"eval_accuracy": 0.7232, |
|
"eval_f1_macro": 0.49873050046317474, |
|
"eval_f1_micro": 0.7232, |
|
"eval_loss": 1.454288363456726, |
|
"eval_runtime": 4.9953, |
|
"eval_samples_per_second": 2001.896, |
|
"eval_steps_per_second": 31.43, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 3.3140788078308105, |
|
"learning_rate": 1.626865671641791e-05, |
|
"loss": 1.521, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 3.006406307220459, |
|
"learning_rate": 1.6090973702914003e-05, |
|
"loss": 1.4858, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 3.3898396492004395, |
|
"learning_rate": 1.5913290689410095e-05, |
|
"loss": 1.3929, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 4.46132755279541, |
|
"learning_rate": 1.5735607675906184e-05, |
|
"loss": 1.3927, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"eval_accuracy": 0.758, |
|
"eval_f1_macro": 0.5628259172803456, |
|
"eval_f1_micro": 0.758, |
|
"eval_loss": 1.26346755027771, |
|
"eval_runtime": 5.0937, |
|
"eval_samples_per_second": 1963.223, |
|
"eval_steps_per_second": 30.823, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 3.48401141166687, |
|
"learning_rate": 1.5557924662402277e-05, |
|
"loss": 1.333, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 3.8899550437927246, |
|
"learning_rate": 1.5380241648898366e-05, |
|
"loss": 1.2675, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 3.1978042125701904, |
|
"learning_rate": 1.5202558635394458e-05, |
|
"loss": 1.273, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 2.9463396072387695, |
|
"learning_rate": 1.5024875621890549e-05, |
|
"loss": 1.2065, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"eval_accuracy": 0.7719, |
|
"eval_f1_macro": 0.5781987999470919, |
|
"eval_f1_micro": 0.7719, |
|
"eval_loss": 1.1217072010040283, |
|
"eval_runtime": 5.0931, |
|
"eval_samples_per_second": 1963.422, |
|
"eval_steps_per_second": 30.826, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 3.5176944732666016, |
|
"learning_rate": 1.484719260838664e-05, |
|
"loss": 1.1275, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 4.0801005363464355, |
|
"learning_rate": 1.4669509594882732e-05, |
|
"loss": 1.1847, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 4.6251397132873535, |
|
"learning_rate": 1.449182658137882e-05, |
|
"loss": 1.1809, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 3.544368028640747, |
|
"learning_rate": 1.4314143567874911e-05, |
|
"loss": 1.16, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"eval_accuracy": 0.7832, |
|
"eval_f1_macro": 0.5984417775290468, |
|
"eval_f1_micro": 0.7832, |
|
"eval_loss": 1.030264139175415, |
|
"eval_runtime": 5.094, |
|
"eval_samples_per_second": 1963.094, |
|
"eval_steps_per_second": 30.821, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 4.991486072540283, |
|
"learning_rate": 1.4136460554371004e-05, |
|
"loss": 1.1233, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 3.317925453186035, |
|
"learning_rate": 1.3958777540867094e-05, |
|
"loss": 1.0539, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 4.44712495803833, |
|
"learning_rate": 1.3781094527363185e-05, |
|
"loss": 1.0876, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 4.135316848754883, |
|
"learning_rate": 1.3603411513859277e-05, |
|
"loss": 1.0168, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"eval_accuracy": 0.7887, |
|
"eval_f1_macro": 0.6118617070223085, |
|
"eval_f1_micro": 0.7887, |
|
"eval_loss": 0.9443017840385437, |
|
"eval_runtime": 5.0979, |
|
"eval_samples_per_second": 1961.573, |
|
"eval_steps_per_second": 30.797, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 4.779892444610596, |
|
"learning_rate": 1.3425728500355366e-05, |
|
"loss": 1.0555, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"grad_norm": 4.020134925842285, |
|
"learning_rate": 1.3248045486851457e-05, |
|
"loss": 0.9748, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"grad_norm": 4.492050647735596, |
|
"learning_rate": 1.307036247334755e-05, |
|
"loss": 0.9662, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 3.679119348526001, |
|
"learning_rate": 1.289267945984364e-05, |
|
"loss": 0.9006, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"eval_accuracy": 0.7934, |
|
"eval_f1_macro": 0.6142121854229132, |
|
"eval_f1_micro": 0.7934, |
|
"eval_loss": 0.8957639932632446, |
|
"eval_runtime": 5.092, |
|
"eval_samples_per_second": 1963.867, |
|
"eval_steps_per_second": 30.833, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.09, |
|
"grad_norm": 4.6417107582092285, |
|
"learning_rate": 1.271499644633973e-05, |
|
"loss": 0.9296, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 1.12, |
|
"grad_norm": 4.246198654174805, |
|
"learning_rate": 1.2537313432835823e-05, |
|
"loss": 0.8785, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 1.15, |
|
"grad_norm": 4.164609909057617, |
|
"learning_rate": 1.2359630419331914e-05, |
|
"loss": 0.9073, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 1.17, |
|
"grad_norm": 4.187661647796631, |
|
"learning_rate": 1.2181947405828003e-05, |
|
"loss": 0.8956, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.17, |
|
"eval_accuracy": 0.8002, |
|
"eval_f1_macro": 0.6294238017197136, |
|
"eval_f1_micro": 0.8002, |
|
"eval_loss": 0.8517330288887024, |
|
"eval_runtime": 5.0924, |
|
"eval_samples_per_second": 1963.694, |
|
"eval_steps_per_second": 30.83, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 2.66513991355896, |
|
"learning_rate": 1.2004264392324093e-05, |
|
"loss": 0.8597, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 1.23, |
|
"grad_norm": 3.2240054607391357, |
|
"learning_rate": 1.1826581378820186e-05, |
|
"loss": 0.8368, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 1.25, |
|
"grad_norm": 4.226346015930176, |
|
"learning_rate": 1.1648898365316276e-05, |
|
"loss": 0.7925, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"grad_norm": 3.5720551013946533, |
|
"learning_rate": 1.1471215351812369e-05, |
|
"loss": 0.9159, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"eval_accuracy": 0.8033, |
|
"eval_f1_macro": 0.6412470887223418, |
|
"eval_f1_micro": 0.8033, |
|
"eval_loss": 0.8183619379997253, |
|
"eval_runtime": 5.0956, |
|
"eval_samples_per_second": 1962.471, |
|
"eval_steps_per_second": 30.811, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.31, |
|
"grad_norm": 3.9309592247009277, |
|
"learning_rate": 1.129353233830846e-05, |
|
"loss": 0.8711, |
|
"step": 1225 |
|
}, |
|
{ |
|
"epoch": 1.33, |
|
"grad_norm": 3.1741673946380615, |
|
"learning_rate": 1.111584932480455e-05, |
|
"loss": 0.839, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 1.36, |
|
"grad_norm": 3.3300561904907227, |
|
"learning_rate": 1.0938166311300639e-05, |
|
"loss": 0.8273, |
|
"step": 1275 |
|
}, |
|
{ |
|
"epoch": 1.39, |
|
"grad_norm": 4.480769634246826, |
|
"learning_rate": 1.0760483297796731e-05, |
|
"loss": 0.8237, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 1.39, |
|
"eval_accuracy": 0.8077, |
|
"eval_f1_macro": 0.6529082127640804, |
|
"eval_f1_micro": 0.8077, |
|
"eval_loss": 0.7814128994941711, |
|
"eval_runtime": 5.0887, |
|
"eval_samples_per_second": 1965.129, |
|
"eval_steps_per_second": 30.853, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 1.41, |
|
"grad_norm": 4.553057670593262, |
|
"learning_rate": 1.0582800284292822e-05, |
|
"loss": 0.8214, |
|
"step": 1325 |
|
}, |
|
{ |
|
"epoch": 1.44, |
|
"grad_norm": 3.766432046890259, |
|
"learning_rate": 1.0405117270788913e-05, |
|
"loss": 0.8174, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 1.47, |
|
"grad_norm": 4.3826446533203125, |
|
"learning_rate": 1.0227434257285005e-05, |
|
"loss": 0.772, |
|
"step": 1375 |
|
}, |
|
{ |
|
"epoch": 1.49, |
|
"grad_norm": 3.777904748916626, |
|
"learning_rate": 1.0049751243781096e-05, |
|
"loss": 0.7341, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.49, |
|
"eval_accuracy": 0.8099, |
|
"eval_f1_macro": 0.6600339223452895, |
|
"eval_f1_micro": 0.8099, |
|
"eval_loss": 0.7653506994247437, |
|
"eval_runtime": 5.0857, |
|
"eval_samples_per_second": 1966.297, |
|
"eval_steps_per_second": 30.871, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.52, |
|
"grad_norm": 4.6313629150390625, |
|
"learning_rate": 9.872068230277187e-06, |
|
"loss": 0.7761, |
|
"step": 1425 |
|
}, |
|
{ |
|
"epoch": 1.55, |
|
"grad_norm": 3.6354401111602783, |
|
"learning_rate": 9.694385216773277e-06, |
|
"loss": 0.8205, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 1.57, |
|
"grad_norm": 3.7265005111694336, |
|
"learning_rate": 9.516702203269368e-06, |
|
"loss": 0.7916, |
|
"step": 1475 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"grad_norm": 3.4605891704559326, |
|
"learning_rate": 9.339019189765458e-06, |
|
"loss": 0.7475, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"eval_accuracy": 0.8135, |
|
"eval_f1_macro": 0.6650413690305985, |
|
"eval_f1_micro": 0.8135, |
|
"eval_loss": 0.7457693815231323, |
|
"eval_runtime": 5.0844, |
|
"eval_samples_per_second": 1966.8, |
|
"eval_steps_per_second": 30.879, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.63, |
|
"grad_norm": 5.460809707641602, |
|
"learning_rate": 9.161336176261551e-06, |
|
"loss": 0.7576, |
|
"step": 1525 |
|
}, |
|
{ |
|
"epoch": 1.65, |
|
"grad_norm": 3.1193032264709473, |
|
"learning_rate": 8.98365316275764e-06, |
|
"loss": 0.7156, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 1.68, |
|
"grad_norm": 3.7673654556274414, |
|
"learning_rate": 8.805970149253732e-06, |
|
"loss": 0.6944, |
|
"step": 1575 |
|
}, |
|
{ |
|
"epoch": 1.71, |
|
"grad_norm": 4.008440971374512, |
|
"learning_rate": 8.628287135749823e-06, |
|
"loss": 0.7699, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.71, |
|
"eval_accuracy": 0.8183, |
|
"eval_f1_macro": 0.681030771263254, |
|
"eval_f1_micro": 0.8183, |
|
"eval_loss": 0.7288192510604858, |
|
"eval_runtime": 5.0906, |
|
"eval_samples_per_second": 1964.407, |
|
"eval_steps_per_second": 30.841, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.73, |
|
"grad_norm": 3.998149871826172, |
|
"learning_rate": 8.450604122245915e-06, |
|
"loss": 0.742, |
|
"step": 1625 |
|
}, |
|
{ |
|
"epoch": 1.76, |
|
"grad_norm": 4.329924583435059, |
|
"learning_rate": 8.272921108742004e-06, |
|
"loss": 0.7505, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 1.79, |
|
"grad_norm": 3.9420013427734375, |
|
"learning_rate": 8.095238095238097e-06, |
|
"loss": 0.7005, |
|
"step": 1675 |
|
}, |
|
{ |
|
"epoch": 1.81, |
|
"grad_norm": 4.527777194976807, |
|
"learning_rate": 7.917555081734187e-06, |
|
"loss": 0.7472, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.81, |
|
"eval_accuracy": 0.8179, |
|
"eval_f1_macro": 0.681973246053428, |
|
"eval_f1_micro": 0.8179, |
|
"eval_loss": 0.712536633014679, |
|
"eval_runtime": 5.0869, |
|
"eval_samples_per_second": 1965.828, |
|
"eval_steps_per_second": 30.864, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.84, |
|
"grad_norm": 5.481727600097656, |
|
"learning_rate": 7.739872068230278e-06, |
|
"loss": 0.7694, |
|
"step": 1725 |
|
}, |
|
{ |
|
"epoch": 1.87, |
|
"grad_norm": 3.634965658187866, |
|
"learning_rate": 7.5621890547263685e-06, |
|
"loss": 0.6916, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 1.89, |
|
"grad_norm": 4.712828636169434, |
|
"learning_rate": 7.38450604122246e-06, |
|
"loss": 0.7452, |
|
"step": 1775 |
|
}, |
|
{ |
|
"epoch": 1.92, |
|
"grad_norm": 4.240174770355225, |
|
"learning_rate": 7.20682302771855e-06, |
|
"loss": 0.689, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.92, |
|
"eval_accuracy": 0.8201, |
|
"eval_f1_macro": 0.6821776177873132, |
|
"eval_f1_micro": 0.8201, |
|
"eval_loss": 0.6964806914329529, |
|
"eval_runtime": 5.0903, |
|
"eval_samples_per_second": 1964.526, |
|
"eval_steps_per_second": 30.843, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.95, |
|
"grad_norm": 3.659916639328003, |
|
"learning_rate": 7.029140014214641e-06, |
|
"loss": 0.8135, |
|
"step": 1825 |
|
}, |
|
{ |
|
"epoch": 1.97, |
|
"grad_norm": 4.639604091644287, |
|
"learning_rate": 6.851457000710733e-06, |
|
"loss": 0.7496, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 4.354977607727051, |
|
"learning_rate": 6.673773987206824e-06, |
|
"loss": 0.6715, |
|
"step": 1875 |
|
}, |
|
{ |
|
"epoch": 2.03, |
|
"grad_norm": 3.503903865814209, |
|
"learning_rate": 6.496090973702914e-06, |
|
"loss": 0.6807, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 2.03, |
|
"eval_accuracy": 0.8192, |
|
"eval_f1_macro": 0.6798912747134073, |
|
"eval_f1_micro": 0.8192, |
|
"eval_loss": 0.6904103755950928, |
|
"eval_runtime": 5.0876, |
|
"eval_samples_per_second": 1965.568, |
|
"eval_steps_per_second": 30.859, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 2.05, |
|
"grad_norm": 2.8646299839019775, |
|
"learning_rate": 6.318407960199006e-06, |
|
"loss": 0.6768, |
|
"step": 1925 |
|
}, |
|
{ |
|
"epoch": 2.08, |
|
"grad_norm": 4.127985000610352, |
|
"learning_rate": 6.140724946695097e-06, |
|
"loss": 0.7144, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"grad_norm": 3.0997469425201416, |
|
"learning_rate": 5.963041933191187e-06, |
|
"loss": 0.6447, |
|
"step": 1975 |
|
}, |
|
{ |
|
"epoch": 2.13, |
|
"grad_norm": 4.458724498748779, |
|
"learning_rate": 5.785358919687279e-06, |
|
"loss": 0.6514, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.13, |
|
"eval_accuracy": 0.8239, |
|
"eval_f1_macro": 0.6922798719270521, |
|
"eval_f1_micro": 0.8239, |
|
"eval_loss": 0.6835533976554871, |
|
"eval_runtime": 5.09, |
|
"eval_samples_per_second": 1964.629, |
|
"eval_steps_per_second": 30.845, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.16, |
|
"grad_norm": 3.7359683513641357, |
|
"learning_rate": 5.607675906183369e-06, |
|
"loss": 0.6547, |
|
"step": 2025 |
|
}, |
|
{ |
|
"epoch": 2.19, |
|
"grad_norm": 3.74875807762146, |
|
"learning_rate": 5.42999289267946e-06, |
|
"loss": 0.6577, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 2.21, |
|
"grad_norm": 3.634157419204712, |
|
"learning_rate": 5.2523098791755514e-06, |
|
"loss": 0.6596, |
|
"step": 2075 |
|
}, |
|
{ |
|
"epoch": 2.24, |
|
"grad_norm": 4.218031406402588, |
|
"learning_rate": 5.074626865671642e-06, |
|
"loss": 0.6662, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 2.24, |
|
"eval_accuracy": 0.8267, |
|
"eval_f1_macro": 0.7018656510236515, |
|
"eval_f1_micro": 0.8267, |
|
"eval_loss": 0.6749560832977295, |
|
"eval_runtime": 5.0908, |
|
"eval_samples_per_second": 1964.309, |
|
"eval_steps_per_second": 30.84, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 2.27, |
|
"grad_norm": 4.950326442718506, |
|
"learning_rate": 4.896943852167733e-06, |
|
"loss": 0.6144, |
|
"step": 2125 |
|
}, |
|
{ |
|
"epoch": 2.29, |
|
"grad_norm": 3.0254364013671875, |
|
"learning_rate": 4.719260838663824e-06, |
|
"loss": 0.6402, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 2.32, |
|
"grad_norm": 2.264652967453003, |
|
"learning_rate": 4.541577825159915e-06, |
|
"loss": 0.5597, |
|
"step": 2175 |
|
}, |
|
{ |
|
"epoch": 2.35, |
|
"grad_norm": 5.00731086730957, |
|
"learning_rate": 4.363894811656006e-06, |
|
"loss": 0.6247, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 2.35, |
|
"eval_accuracy": 0.8284, |
|
"eval_f1_macro": 0.7028136538357826, |
|
"eval_f1_micro": 0.8284, |
|
"eval_loss": 0.6703132390975952, |
|
"eval_runtime": 5.0889, |
|
"eval_samples_per_second": 1965.077, |
|
"eval_steps_per_second": 30.852, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 2.37, |
|
"grad_norm": 3.5009686946868896, |
|
"learning_rate": 4.186211798152097e-06, |
|
"loss": 0.5784, |
|
"step": 2225 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"grad_norm": 3.339487075805664, |
|
"learning_rate": 4.008528784648188e-06, |
|
"loss": 0.6321, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 2.43, |
|
"grad_norm": 4.3150129318237305, |
|
"learning_rate": 3.8308457711442784e-06, |
|
"loss": 0.6499, |
|
"step": 2275 |
|
}, |
|
{ |
|
"epoch": 2.45, |
|
"grad_norm": 3.9127256870269775, |
|
"learning_rate": 3.65316275764037e-06, |
|
"loss": 0.6443, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 2.45, |
|
"eval_accuracy": 0.8265, |
|
"eval_f1_macro": 0.7001404361966705, |
|
"eval_f1_micro": 0.8265, |
|
"eval_loss": 0.666212260723114, |
|
"eval_runtime": 5.0933, |
|
"eval_samples_per_second": 1963.351, |
|
"eval_steps_per_second": 30.825, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 2.48, |
|
"grad_norm": 4.648262023925781, |
|
"learning_rate": 3.4754797441364606e-06, |
|
"loss": 0.6597, |
|
"step": 2325 |
|
}, |
|
{ |
|
"epoch": 2.51, |
|
"grad_norm": 5.805054187774658, |
|
"learning_rate": 3.297796730632552e-06, |
|
"loss": 0.6679, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"grad_norm": 3.7240312099456787, |
|
"learning_rate": 3.120113717128643e-06, |
|
"loss": 0.6098, |
|
"step": 2375 |
|
}, |
|
{ |
|
"epoch": 2.56, |
|
"grad_norm": 6.238827705383301, |
|
"learning_rate": 2.9424307036247335e-06, |
|
"loss": 0.632, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.56, |
|
"eval_accuracy": 0.8295, |
|
"eval_f1_macro": 0.7077893947175455, |
|
"eval_f1_micro": 0.8295, |
|
"eval_loss": 0.6571224331855774, |
|
"eval_runtime": 5.0923, |
|
"eval_samples_per_second": 1963.743, |
|
"eval_steps_per_second": 30.831, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.59, |
|
"grad_norm": 2.7973475456237793, |
|
"learning_rate": 2.764747690120825e-06, |
|
"loss": 0.7127, |
|
"step": 2425 |
|
}, |
|
{ |
|
"epoch": 2.61, |
|
"grad_norm": 4.434217929840088, |
|
"learning_rate": 2.5870646766169156e-06, |
|
"loss": 0.6907, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 2.64, |
|
"grad_norm": 5.4939188957214355, |
|
"learning_rate": 2.4093816631130067e-06, |
|
"loss": 0.6308, |
|
"step": 2475 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"grad_norm": 6.5089430809021, |
|
"learning_rate": 2.2316986496090974e-06, |
|
"loss": 0.5922, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"eval_accuracy": 0.8298, |
|
"eval_f1_macro": 0.7083519630678758, |
|
"eval_f1_micro": 0.8298, |
|
"eval_loss": 0.6538856029510498, |
|
"eval_runtime": 5.111, |
|
"eval_samples_per_second": 1956.564, |
|
"eval_steps_per_second": 30.718, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.69, |
|
"grad_norm": 4.474384307861328, |
|
"learning_rate": 2.0540156361051885e-06, |
|
"loss": 0.6384, |
|
"step": 2525 |
|
}, |
|
{ |
|
"epoch": 2.72, |
|
"grad_norm": 4.887701511383057, |
|
"learning_rate": 1.8763326226012796e-06, |
|
"loss": 0.614, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 2.75, |
|
"grad_norm": 6.661905288696289, |
|
"learning_rate": 1.6986496090973702e-06, |
|
"loss": 0.626, |
|
"step": 2575 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"grad_norm": 4.168758392333984, |
|
"learning_rate": 1.5209665955934613e-06, |
|
"loss": 0.6423, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"eval_accuracy": 0.8311, |
|
"eval_f1_macro": 0.7138508038282735, |
|
"eval_f1_micro": 0.8311, |
|
"eval_loss": 0.6518952250480652, |
|
"eval_runtime": 5.0992, |
|
"eval_samples_per_second": 1961.08, |
|
"eval_steps_per_second": 30.789, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"grad_norm": 3.9610595703125, |
|
"learning_rate": 1.3432835820895524e-06, |
|
"loss": 0.6361, |
|
"step": 2625 |
|
}, |
|
{ |
|
"epoch": 2.83, |
|
"grad_norm": 5.312166213989258, |
|
"learning_rate": 1.1656005685856433e-06, |
|
"loss": 0.6638, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 2.85, |
|
"grad_norm": 4.739462852478027, |
|
"learning_rate": 9.879175550817342e-07, |
|
"loss": 0.625, |
|
"step": 2675 |
|
}, |
|
{ |
|
"epoch": 2.88, |
|
"grad_norm": 5.244933128356934, |
|
"learning_rate": 8.102345415778253e-07, |
|
"loss": 0.6156, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 2.88, |
|
"eval_accuracy": 0.8311, |
|
"eval_f1_macro": 0.7123382093925994, |
|
"eval_f1_micro": 0.8311, |
|
"eval_loss": 0.6500191688537598, |
|
"eval_runtime": 5.1041, |
|
"eval_samples_per_second": 1959.192, |
|
"eval_steps_per_second": 30.759, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"grad_norm": 3.2966127395629883, |
|
"learning_rate": 6.325515280739161e-07, |
|
"loss": 0.6181, |
|
"step": 2725 |
|
}, |
|
{ |
|
"epoch": 2.93, |
|
"grad_norm": 4.855289936065674, |
|
"learning_rate": 4.548685145700071e-07, |
|
"loss": 0.5952, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 2.96, |
|
"grad_norm": 5.3126068115234375, |
|
"learning_rate": 2.771855010660981e-07, |
|
"loss": 0.5727, |
|
"step": 2775 |
|
}, |
|
{ |
|
"epoch": 2.99, |
|
"grad_norm": 3.789565324783325, |
|
"learning_rate": 9.950248756218906e-08, |
|
"loss": 0.6097, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 2.99, |
|
"eval_accuracy": 0.8311, |
|
"eval_f1_macro": 0.7116365414985335, |
|
"eval_f1_micro": 0.8311, |
|
"eval_loss": 0.6496002674102783, |
|
"eval_runtime": 5.1014, |
|
"eval_samples_per_second": 1960.25, |
|
"eval_steps_per_second": 30.776, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"step": 2814, |
|
"total_flos": 5974635617189888.0, |
|
"train_loss": 1.1483757653470232, |
|
"train_runtime": 441.0175, |
|
"train_samples_per_second": 408.147, |
|
"train_steps_per_second": 6.381 |
|
} |
|
], |
|
"logging_steps": 25, |
|
"max_steps": 2814, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 100, |
|
"total_flos": 5974635617189888.0, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|