| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.3530364372469634, | |
| "eval_steps": 8, | |
| "global_step": 90, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.025910931174089068, | |
| "eval_loss": 0.21314890682697296, | |
| "eval_runtime": 12.1183, | |
| "eval_samples_per_second": 10.728, | |
| "eval_steps_per_second": 5.364, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07773279352226721, | |
| "grad_norm": 6.452937126159668, | |
| "learning_rate": 3e-05, | |
| "loss": 6.5264, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15546558704453442, | |
| "grad_norm": 6.785652160644531, | |
| "learning_rate": 6e-05, | |
| "loss": 6.3657, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.20728744939271254, | |
| "eval_loss": 0.10755843669176102, | |
| "eval_runtime": 12.1878, | |
| "eval_samples_per_second": 10.666, | |
| "eval_steps_per_second": 5.333, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.23319838056680162, | |
| "grad_norm": 5.717119216918945, | |
| "learning_rate": 9e-05, | |
| "loss": 4.5674, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.31093117408906884, | |
| "grad_norm": 6.126352787017822, | |
| "learning_rate": 0.00012, | |
| "loss": 1.7668, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 2.648998498916626, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.4109, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4145748987854251, | |
| "eval_loss": 0.0012996116420254111, | |
| "eval_runtime": 12.1969, | |
| "eval_samples_per_second": 10.658, | |
| "eval_steps_per_second": 5.329, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.46639676113360323, | |
| "grad_norm": 0.33190277218818665, | |
| "learning_rate": 0.00018, | |
| "loss": 0.0473, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5441295546558704, | |
| "grad_norm": 0.2983362078666687, | |
| "learning_rate": 0.00019989930665413147, | |
| "loss": 0.0116, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6218623481781377, | |
| "grad_norm": 0.2238294929265976, | |
| "learning_rate": 0.00019839295885986296, | |
| "loss": 0.0077, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6218623481781377, | |
| "eval_loss": 0.000773053674492985, | |
| "eval_runtime": 12.1863, | |
| "eval_samples_per_second": 10.668, | |
| "eval_steps_per_second": 5.334, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6995951417004048, | |
| "grad_norm": 0.016547193750739098, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.0142, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.019047992303967476, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.0045, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.8291497975708502, | |
| "eval_loss": 0.0002317978214705363, | |
| "eval_runtime": 12.2198, | |
| "eval_samples_per_second": 10.638, | |
| "eval_steps_per_second": 5.319, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.8550607287449393, | |
| "grad_norm": 1.0941699743270874, | |
| "learning_rate": 0.00018345732537213027, | |
| "loss": 0.0019, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.9327935222672065, | |
| "grad_norm": 1.774532675743103, | |
| "learning_rate": 0.00017530714660036112, | |
| "loss": 0.0047, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.0210526315789474, | |
| "grad_norm": 0.029302451759576797, | |
| "learning_rate": 0.00016579387259397127, | |
| "loss": 0.0021, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0469635627530365, | |
| "eval_loss": 2.721450073295273e-05, | |
| "eval_runtime": 44.3093, | |
| "eval_samples_per_second": 2.934, | |
| "eval_steps_per_second": 1.467, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0987854251012146, | |
| "grad_norm": 0.009189918637275696, | |
| "learning_rate": 0.00015508969814521025, | |
| "loss": 0.0005, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.1765182186234817, | |
| "grad_norm": 0.030586188659071922, | |
| "learning_rate": 0.00014338837391175582, | |
| "loss": 0.0006, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.2542510121457489, | |
| "grad_norm": 0.019615933299064636, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.0008, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.2542510121457489, | |
| "eval_loss": 2.218412555521354e-05, | |
| "eval_runtime": 44.4013, | |
| "eval_samples_per_second": 2.928, | |
| "eval_steps_per_second": 1.464, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.006924801040440798, | |
| "learning_rate": 0.00011785568947986367, | |
| "loss": 0.045, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.4097165991902834, | |
| "grad_norm": 0.02355915680527687, | |
| "learning_rate": 0.00010448648303505151, | |
| "loss": 0.0006, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "eval_loss": 2.361264887440484e-05, | |
| "eval_runtime": 44.3855, | |
| "eval_samples_per_second": 2.929, | |
| "eval_steps_per_second": 1.464, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.4874493927125507, | |
| "grad_norm": 1.601819396018982, | |
| "learning_rate": 9.103606910965666e-05, | |
| "loss": 0.0077, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.5651821862348179, | |
| "grad_norm": 0.09957171976566315, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.0016, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.642914979757085, | |
| "grad_norm": 0.03547385334968567, | |
| "learning_rate": 6.486251759186572e-05, | |
| "loss": 0.001, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.668825910931174, | |
| "eval_loss": 3.899199145962484e-05, | |
| "eval_runtime": 43.2962, | |
| "eval_samples_per_second": 3.003, | |
| "eval_steps_per_second": 1.501, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.025507541373372078, | |
| "learning_rate": 5.261313375270014e-05, | |
| "loss": 0.0005, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.7983805668016193, | |
| "grad_norm": 0.009455524384975433, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.0004, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.8761133603238866, | |
| "grad_norm": 0.0038223874289542437, | |
| "learning_rate": 3.089373510131354e-05, | |
| "loss": 0.0009, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.8761133603238866, | |
| "eval_loss": 3.11742551275529e-05, | |
| "eval_runtime": 44.3844, | |
| "eval_samples_per_second": 2.929, | |
| "eval_steps_per_second": 1.464, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.953846153846154, | |
| "grad_norm": 0.005872055422514677, | |
| "learning_rate": 2.181685175319702e-05, | |
| "loss": 0.001, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.042105263157895, | |
| "grad_norm": 0.03851957991719246, | |
| "learning_rate": 1.415512063981339e-05, | |
| "loss": 0.0008, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.093927125506073, | |
| "eval_loss": 3.268746513640508e-05, | |
| "eval_runtime": 44.3546, | |
| "eval_samples_per_second": 2.931, | |
| "eval_steps_per_second": 1.465, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.119838056680162, | |
| "grad_norm": 0.009562190622091293, | |
| "learning_rate": 8.047222744854943e-06, | |
| "loss": 0.0003, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.197570850202429, | |
| "grad_norm": 0.00980624184012413, | |
| "learning_rate": 3.6037139304146762e-06, | |
| "loss": 0.0003, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.010969814844429493, | |
| "learning_rate": 9.0502382320653e-07, | |
| "loss": 0.0003, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.3012145748987853, | |
| "eval_loss": 2.5345105314045213e-05, | |
| "eval_runtime": 44.4252, | |
| "eval_samples_per_second": 2.926, | |
| "eval_steps_per_second": 1.463, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.3530364372469634, | |
| "grad_norm": 2.2677857875823975, | |
| "learning_rate": 0.0, | |
| "loss": 0.009, | |
| "step": 90 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 90, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 8, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.7332406838951936e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |