| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 100, | |
| "global_step": 2345, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 1.6159120798110962, | |
| "learning_rate": 0.0007658848614072496, | |
| "loss": 0.026, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21321961620469082, | |
| "eval_accuracy": 0.9867, | |
| "eval_loss": 0.04582102224230766, | |
| "eval_runtime": 5.154, | |
| "eval_samples_per_second": 1940.235, | |
| "eval_steps_per_second": 15.328, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.22664281725883484, | |
| "learning_rate": 0.000731769722814499, | |
| "loss": 0.0232, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42643923240938164, | |
| "eval_accuracy": 0.9867, | |
| "eval_loss": 0.04527696222066879, | |
| "eval_runtime": 5.4041, | |
| "eval_samples_per_second": 1850.446, | |
| "eval_steps_per_second": 14.619, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 1.3475489616394043, | |
| "learning_rate": 0.0006976545842217485, | |
| "loss": 0.0277, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6396588486140725, | |
| "eval_accuracy": 0.9863, | |
| "eval_loss": 0.04839787632226944, | |
| "eval_runtime": 5.3353, | |
| "eval_samples_per_second": 1874.323, | |
| "eval_steps_per_second": 14.807, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.9408140182495117, | |
| "learning_rate": 0.0006635394456289979, | |
| "loss": 0.0293, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "eval_accuracy": 0.9865, | |
| "eval_loss": 0.046898942440748215, | |
| "eval_runtime": 5.1935, | |
| "eval_samples_per_second": 1925.498, | |
| "eval_steps_per_second": 15.211, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "grad_norm": 0.6018700003623962, | |
| "learning_rate": 0.0006294243070362473, | |
| "loss": 0.0235, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0660980810234542, | |
| "eval_accuracy": 0.9899, | |
| "eval_loss": 0.028819510713219643, | |
| "eval_runtime": 5.1407, | |
| "eval_samples_per_second": 1945.277, | |
| "eval_steps_per_second": 15.368, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "grad_norm": 0.7029837369918823, | |
| "learning_rate": 0.0005953091684434968, | |
| "loss": 0.0203, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.279317697228145, | |
| "eval_accuracy": 0.9924, | |
| "eval_loss": 0.02526690438389778, | |
| "eval_runtime": 4.869, | |
| "eval_samples_per_second": 2053.81, | |
| "eval_steps_per_second": 16.225, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.17393378913402557, | |
| "learning_rate": 0.0005611940298507463, | |
| "loss": 0.0182, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "eval_accuracy": 0.9916, | |
| "eval_loss": 0.028603948652744293, | |
| "eval_runtime": 4.6847, | |
| "eval_samples_per_second": 2134.586, | |
| "eval_steps_per_second": 16.863, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "grad_norm": 0.9470173120498657, | |
| "learning_rate": 0.0005270788912579957, | |
| "loss": 0.0205, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7057569296375266, | |
| "eval_accuracy": 0.9935, | |
| "eval_loss": 0.02031560428440571, | |
| "eval_runtime": 5.0835, | |
| "eval_samples_per_second": 1967.155, | |
| "eval_steps_per_second": 15.541, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "grad_norm": 1.4821853637695312, | |
| "learning_rate": 0.0004929637526652453, | |
| "loss": 0.0162, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9189765458422174, | |
| "eval_accuracy": 0.9913, | |
| "eval_loss": 0.023791342973709106, | |
| "eval_runtime": 5.2196, | |
| "eval_samples_per_second": 1915.863, | |
| "eval_steps_per_second": 15.135, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "grad_norm": 1.0076653957366943, | |
| "learning_rate": 0.0004588486140724947, | |
| "loss": 0.0118, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.1321961620469083, | |
| "eval_accuracy": 0.9916, | |
| "eval_loss": 0.024731909856200218, | |
| "eval_runtime": 5.8209, | |
| "eval_samples_per_second": 1717.944, | |
| "eval_steps_per_second": 13.572, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "grad_norm": 0.39255017042160034, | |
| "learning_rate": 0.0004247334754797441, | |
| "loss": 0.0121, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.345415778251599, | |
| "eval_accuracy": 0.9932, | |
| "eval_loss": 0.019426949322223663, | |
| "eval_runtime": 5.5038, | |
| "eval_samples_per_second": 1816.931, | |
| "eval_steps_per_second": 14.354, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "grad_norm": 0.4671665132045746, | |
| "learning_rate": 0.0003906183368869936, | |
| "loss": 0.0154, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.55863539445629, | |
| "eval_accuracy": 0.9933, | |
| "eval_loss": 0.01936325989663601, | |
| "eval_runtime": 5.4304, | |
| "eval_samples_per_second": 1841.5, | |
| "eval_steps_per_second": 14.548, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "grad_norm": 0.17253048717975616, | |
| "learning_rate": 0.0003565031982942431, | |
| "loss": 0.015, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.771855010660981, | |
| "eval_accuracy": 0.9933, | |
| "eval_loss": 0.02162059210240841, | |
| "eval_runtime": 5.3729, | |
| "eval_samples_per_second": 1861.177, | |
| "eval_steps_per_second": 14.703, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.8343185782432556, | |
| "learning_rate": 0.00032238805970149256, | |
| "loss": 0.0145, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "eval_accuracy": 0.9919, | |
| "eval_loss": 0.02381654642522335, | |
| "eval_runtime": 5.1099, | |
| "eval_samples_per_second": 1957.0, | |
| "eval_steps_per_second": 15.46, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "grad_norm": 0.3106481432914734, | |
| "learning_rate": 0.000288272921108742, | |
| "loss": 0.0098, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.1982942430703627, | |
| "eval_accuracy": 0.993, | |
| "eval_loss": 0.020756520330905914, | |
| "eval_runtime": 4.9767, | |
| "eval_samples_per_second": 2009.353, | |
| "eval_steps_per_second": 15.874, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "grad_norm": 0.03616774454712868, | |
| "learning_rate": 0.0002541577825159915, | |
| "loss": 0.0093, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.411513859275053, | |
| "eval_accuracy": 0.9929, | |
| "eval_loss": 0.021822581067681313, | |
| "eval_runtime": 5.4965, | |
| "eval_samples_per_second": 1819.356, | |
| "eval_steps_per_second": 14.373, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "grad_norm": 0.6045345067977905, | |
| "learning_rate": 0.00022004264392324095, | |
| "loss": 0.0073, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.624733475479744, | |
| "eval_accuracy": 0.9933, | |
| "eval_loss": 0.018862707540392876, | |
| "eval_runtime": 5.7766, | |
| "eval_samples_per_second": 1731.12, | |
| "eval_steps_per_second": 13.676, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "grad_norm": 0.631215512752533, | |
| "learning_rate": 0.0001859275053304904, | |
| "loss": 0.008, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.837953091684435, | |
| "eval_accuracy": 0.9932, | |
| "eval_loss": 0.01944512128829956, | |
| "eval_runtime": 5.1522, | |
| "eval_samples_per_second": 1940.922, | |
| "eval_steps_per_second": 15.333, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "grad_norm": 0.5996153950691223, | |
| "learning_rate": 0.0001518123667377399, | |
| "loss": 0.006, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.051172707889126, | |
| "eval_accuracy": 0.9938, | |
| "eval_loss": 0.018317226320505142, | |
| "eval_runtime": 4.9544, | |
| "eval_samples_per_second": 2018.428, | |
| "eval_steps_per_second": 15.946, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "grad_norm": 0.0069321137852966785, | |
| "learning_rate": 0.00011769722814498933, | |
| "loss": 0.0063, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.264392324093817, | |
| "eval_accuracy": 0.9934, | |
| "eval_loss": 0.018384862691164017, | |
| "eval_runtime": 4.7636, | |
| "eval_samples_per_second": 2099.232, | |
| "eval_steps_per_second": 16.584, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.06509185582399368, | |
| "learning_rate": 8.392324093816631e-05, | |
| "loss": 0.0043, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "eval_accuracy": 0.9932, | |
| "eval_loss": 0.018380142748355865, | |
| "eval_runtime": 4.6951, | |
| "eval_samples_per_second": 2129.883, | |
| "eval_steps_per_second": 16.826, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "grad_norm": 0.17103737592697144, | |
| "learning_rate": 4.980810234541578e-05, | |
| "loss": 0.0035, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.690831556503198, | |
| "eval_accuracy": 0.9931, | |
| "eval_loss": 0.018344268202781677, | |
| "eval_runtime": 4.6229, | |
| "eval_samples_per_second": 2163.133, | |
| "eval_steps_per_second": 17.089, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "grad_norm": 0.6465263962745667, | |
| "learning_rate": 1.5692963752665246e-05, | |
| "loss": 0.0061, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.904051172707889, | |
| "eval_accuracy": 0.9931, | |
| "eval_loss": 0.018412619829177856, | |
| "eval_runtime": 4.7593, | |
| "eval_samples_per_second": 2101.13, | |
| "eval_steps_per_second": 16.599, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2345, | |
| "total_flos": 1346208595200000.0, | |
| "train_loss": 0.014373551363121472, | |
| "train_runtime": 335.5598, | |
| "train_samples_per_second": 894.028, | |
| "train_steps_per_second": 6.988 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2345, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1346208595200000.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |