| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 16464, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00018221574344023323, | |
| "eval_accuracy": 0.013372916488627156, | |
| "eval_loss": 10.764164924621582, | |
| "eval_runtime": 265.3902, | |
| "eval_samples_per_second": 71.762, | |
| "eval_steps_per_second": 2.246, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00036443148688046647, | |
| "eval_accuracy": 0.038661994909930615, | |
| "eval_loss": 10.623634338378906, | |
| "eval_runtime": 265.6524, | |
| "eval_samples_per_second": 71.691, | |
| "eval_steps_per_second": 2.244, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0007288629737609329, | |
| "eval_accuracy": 0.0419639958558818, | |
| "eval_loss": 10.446562767028809, | |
| "eval_runtime": 265.6133, | |
| "eval_samples_per_second": 71.702, | |
| "eval_steps_per_second": 2.244, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0014577259475218659, | |
| "eval_accuracy": 0.04410673183105199, | |
| "eval_loss": 10.253758430480957, | |
| "eval_runtime": 266.28, | |
| "eval_samples_per_second": 71.522, | |
| "eval_steps_per_second": 2.238, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0029154518950437317, | |
| "eval_accuracy": 0.04422052313718063, | |
| "eval_loss": 10.048787117004395, | |
| "eval_runtime": 265.5895, | |
| "eval_samples_per_second": 71.708, | |
| "eval_steps_per_second": 2.244, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0058309037900874635, | |
| "eval_accuracy": 0.05247016186133218, | |
| "eval_loss": 9.678139686584473, | |
| "eval_runtime": 266.1849, | |
| "eval_samples_per_second": 71.548, | |
| "eval_steps_per_second": 2.239, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.011661807580174927, | |
| "eval_accuracy": 0.06491211456531285, | |
| "eval_loss": 9.024864196777344, | |
| "eval_runtime": 266.9794, | |
| "eval_samples_per_second": 71.335, | |
| "eval_steps_per_second": 2.232, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.023323615160349854, | |
| "eval_accuracy": 0.08493399513987425, | |
| "eval_loss": 8.21505069732666, | |
| "eval_runtime": 266.8342, | |
| "eval_samples_per_second": 71.374, | |
| "eval_steps_per_second": 2.234, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.04664723032069971, | |
| "eval_accuracy": 0.10968881388346323, | |
| "eval_loss": 7.2851386070251465, | |
| "eval_runtime": 267.6197, | |
| "eval_samples_per_second": 71.164, | |
| "eval_steps_per_second": 2.227, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.09110787172011661, | |
| "grad_norm": 0.5542777180671692, | |
| "learning_rate": 4.848153547133139e-05, | |
| "loss": 7.6369, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09329446064139942, | |
| "eval_accuracy": 0.12987750625095115, | |
| "eval_loss": 6.427273273468018, | |
| "eval_runtime": 267.5439, | |
| "eval_samples_per_second": 71.185, | |
| "eval_steps_per_second": 2.228, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.18221574344023322, | |
| "grad_norm": 0.6718143820762634, | |
| "learning_rate": 4.696307094266278e-05, | |
| "loss": 6.0229, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18658892128279883, | |
| "eval_accuracy": 0.16116734379422917, | |
| "eval_loss": 5.696187496185303, | |
| "eval_runtime": 268.1415, | |
| "eval_samples_per_second": 71.026, | |
| "eval_steps_per_second": 2.223, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.27332361516034986, | |
| "grad_norm": 0.6730483174324036, | |
| "learning_rate": 4.5444606413994175e-05, | |
| "loss": 5.5062, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36443148688046645, | |
| "grad_norm": 0.7009645700454712, | |
| "learning_rate": 4.3926141885325564e-05, | |
| "loss": 5.2064, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.37317784256559766, | |
| "eval_accuracy": 0.1897994845258965, | |
| "eval_loss": 5.082965850830078, | |
| "eval_runtime": 268.3609, | |
| "eval_samples_per_second": 70.968, | |
| "eval_steps_per_second": 2.221, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 0.4555393586005831, | |
| "grad_norm": 0.7892386317253113, | |
| "learning_rate": 4.240767735665695e-05, | |
| "loss": 4.9923, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5466472303206997, | |
| "grad_norm": 0.8406383991241455, | |
| "learning_rate": 4.088921282798834e-05, | |
| "loss": 4.8437, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6377551020408163, | |
| "grad_norm": 0.8709362149238586, | |
| "learning_rate": 3.937074829931973e-05, | |
| "loss": 4.7268, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.7288629737609329, | |
| "grad_norm": 0.9013070464134216, | |
| "learning_rate": 3.785228377065112e-05, | |
| "loss": 4.6324, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7463556851311953, | |
| "eval_accuracy": 0.22371889184616256, | |
| "eval_loss": 4.59232759475708, | |
| "eval_runtime": 268.1172, | |
| "eval_samples_per_second": 71.032, | |
| "eval_steps_per_second": 2.223, | |
| "step": 4096 | |
| }, | |
| { | |
| "epoch": 0.8199708454810496, | |
| "grad_norm": 0.9396291971206665, | |
| "learning_rate": 3.6333819241982507e-05, | |
| "loss": 4.556, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.9110787172011662, | |
| "grad_norm": 1.0571624040603638, | |
| "learning_rate": 3.4815354713313895e-05, | |
| "loss": 4.4846, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.2389537359040827, | |
| "eval_loss": 4.42299747467041, | |
| "eval_runtime": 268.0887, | |
| "eval_samples_per_second": 71.04, | |
| "eval_steps_per_second": 2.223, | |
| "step": 5488 | |
| }, | |
| { | |
| "epoch": 1.0021865889212829, | |
| "grad_norm": 0.9928905963897705, | |
| "learning_rate": 3.329689018464529e-05, | |
| "loss": 4.4267, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0932944606413995, | |
| "grad_norm": 1.0094623565673828, | |
| "learning_rate": 3.177842565597668e-05, | |
| "loss": 4.3598, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.184402332361516, | |
| "grad_norm": 1.1072958707809448, | |
| "learning_rate": 3.0259961127308068e-05, | |
| "loss": 4.3227, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.2755102040816326, | |
| "grad_norm": 1.047276258468628, | |
| "learning_rate": 2.8741496598639456e-05, | |
| "loss": 4.2846, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.3666180758017492, | |
| "grad_norm": 1.0852686166763306, | |
| "learning_rate": 2.7223032069970845e-05, | |
| "loss": 4.2512, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.4577259475218658, | |
| "grad_norm": 1.080506443977356, | |
| "learning_rate": 2.5704567541302237e-05, | |
| "loss": 4.2266, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4927113702623906, | |
| "eval_accuracy": 0.25919180456227686, | |
| "eval_loss": 4.229198455810547, | |
| "eval_runtime": 268.9266, | |
| "eval_samples_per_second": 70.819, | |
| "eval_steps_per_second": 2.216, | |
| "step": 8192 | |
| }, | |
| { | |
| "epoch": 1.5488338192419824, | |
| "grad_norm": 1.1083309650421143, | |
| "learning_rate": 2.4186103012633625e-05, | |
| "loss": 4.1916, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.639941690962099, | |
| "grad_norm": 1.1683882474899292, | |
| "learning_rate": 2.2667638483965014e-05, | |
| "loss": 4.1699, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.7310495626822158, | |
| "grad_norm": 1.2223962545394897, | |
| "learning_rate": 2.1149173955296406e-05, | |
| "loss": 4.1517, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.8221574344023324, | |
| "grad_norm": 1.1384245157241821, | |
| "learning_rate": 1.9630709426627795e-05, | |
| "loss": 4.1228, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.913265306122449, | |
| "grad_norm": 1.2079499959945679, | |
| "learning_rate": 1.8112244897959187e-05, | |
| "loss": 4.1079, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.2717258887026585, | |
| "eval_loss": 4.115562438964844, | |
| "eval_runtime": 267.5246, | |
| "eval_samples_per_second": 71.19, | |
| "eval_steps_per_second": 2.228, | |
| "step": 10976 | |
| }, | |
| { | |
| "epoch": 2.0043731778425657, | |
| "grad_norm": 1.215710163116455, | |
| "learning_rate": 1.6593780369290575e-05, | |
| "loss": 4.0902, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.0954810495626823, | |
| "grad_norm": 1.2627053260803223, | |
| "learning_rate": 1.5075315840621965e-05, | |
| "loss": 4.0556, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.186588921282799, | |
| "grad_norm": 1.2118171453475952, | |
| "learning_rate": 1.3556851311953352e-05, | |
| "loss": 4.0461, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.2776967930029155, | |
| "grad_norm": 1.3033018112182617, | |
| "learning_rate": 1.2038386783284743e-05, | |
| "loss": 4.0326, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.368804664723032, | |
| "grad_norm": 1.2471245527267456, | |
| "learning_rate": 1.0519922254616133e-05, | |
| "loss": 4.0282, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.4599125364431487, | |
| "grad_norm": 1.304821252822876, | |
| "learning_rate": 9.001457725947522e-06, | |
| "loss": 4.0197, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.5510204081632653, | |
| "grad_norm": 1.1969258785247803, | |
| "learning_rate": 7.482993197278912e-06, | |
| "loss": 4.0076, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.642128279883382, | |
| "grad_norm": 1.264854073524475, | |
| "learning_rate": 5.964528668610301e-06, | |
| "loss": 3.9948, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.7332361516034984, | |
| "grad_norm": 1.2441176176071167, | |
| "learning_rate": 4.4460641399416915e-06, | |
| "loss": 3.9953, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.824344023323615, | |
| "grad_norm": 1.2413525581359863, | |
| "learning_rate": 2.9275996112730806e-06, | |
| "loss": 3.9837, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.9154518950437316, | |
| "grad_norm": 1.2331761121749878, | |
| "learning_rate": 1.4091350826044704e-06, | |
| "loss": 3.9841, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9854227405247813, | |
| "eval_accuracy": 0.28256937381675906, | |
| "eval_loss": 4.027393817901611, | |
| "eval_runtime": 267.6695, | |
| "eval_samples_per_second": 71.151, | |
| "eval_steps_per_second": 2.227, | |
| "step": 16384 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.2825705030042804, | |
| "eval_loss": 4.0273261070251465, | |
| "eval_runtime": 268.1735, | |
| "eval_samples_per_second": 71.017, | |
| "eval_steps_per_second": 2.222, | |
| "step": 16464 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 16464, | |
| "total_flos": 1.44589665695957e+17, | |
| "train_loss": 4.469071524483817, | |
| "train_runtime": 22077.8476, | |
| "train_samples_per_second": 23.862, | |
| "train_steps_per_second": 0.746 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 16464, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 1.44589665695957e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |