| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 74, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0, | |
| "eval_loss": 10.911848068237305, | |
| "eval_runtime": 29.8113, | |
| "eval_samples_per_second": 16.772, | |
| "eval_steps_per_second": 2.113, | |
| "step": 0 | |
| }, | |
| { | |
| "epoch": 0.027586206896551724, | |
| "grad_norm": 232.0, | |
| "learning_rate": 0.0, | |
| "loss": 10.8802, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05517241379310345, | |
| "grad_norm": 225.0, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 10.8587, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.08275862068965517, | |
| "grad_norm": 239.0, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 10.6029, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "grad_norm": 220.0, | |
| "learning_rate": 6.25e-06, | |
| "loss": 9.4955, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 176.0, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 8.4033, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16551724137931034, | |
| "grad_norm": 98.5, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 6.9577, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.19310344827586207, | |
| "grad_norm": 74.0, | |
| "learning_rate": 1.25e-05, | |
| "loss": 6.1522, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "grad_norm": 62.5, | |
| "learning_rate": 1.4583333333333335e-05, | |
| "loss": 5.5143, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2482758620689655, | |
| "grad_norm": 51.5, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 5.0074, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 42.5, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 4.6033, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.30344827586206896, | |
| "grad_norm": 40.5, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 4.256, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "grad_norm": 36.0, | |
| "learning_rate": 2.2916666666666667e-05, | |
| "loss": 4.0102, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3586206896551724, | |
| "grad_norm": 34.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.7875, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.38620689655172413, | |
| "grad_norm": 39.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.6879, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 41.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.4863, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4413793103448276, | |
| "grad_norm": 49.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.3334, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.4689655172413793, | |
| "grad_norm": 85.5, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.2869, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.496551724137931, | |
| "grad_norm": 48.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.0628, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5241379310344828, | |
| "grad_norm": 48.5, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.9302, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 47.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.7863, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5793103448275863, | |
| "grad_norm": 57.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.6619, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6068965517241379, | |
| "grad_norm": 45.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.5593, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6344827586206897, | |
| "grad_norm": 45.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.4377, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.6620689655172414, | |
| "grad_norm": 44.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.2583, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 43.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.1977, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7172413793103448, | |
| "grad_norm": 43.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.0286, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.7448275862068966, | |
| "grad_norm": 42.5, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.8541, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.7724137931034483, | |
| "grad_norm": 41.5, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.7261, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 40.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.5602, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 39.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.4275, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.8551724137931035, | |
| "grad_norm": 38.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.3388, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.8827586206896552, | |
| "grad_norm": 37.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.192, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.9103448275862069, | |
| "grad_norm": 32.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.0922, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.9379310344827586, | |
| "grad_norm": 31.5, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.0296, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 26.875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.8785, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.993103448275862, | |
| "grad_norm": 22.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.7932, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 22.875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.7351, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7064927220344543, | |
| "eval_runtime": 24.7844, | |
| "eval_samples_per_second": 20.174, | |
| "eval_steps_per_second": 2.542, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.0275862068965518, | |
| "grad_norm": 15.375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.6586, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.0551724137931036, | |
| "grad_norm": 12.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.5992, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0827586206896551, | |
| "grad_norm": 10.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.5952, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.110344827586207, | |
| "grad_norm": 7.875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.5291, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.5365, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.1655172413793102, | |
| "grad_norm": 6.375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4692, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.193103448275862, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4699, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.2206896551724138, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4285, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.2482758620689656, | |
| "grad_norm": 9.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4281, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 9.625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4373, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.303448275862069, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3719, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.3310344827586207, | |
| "grad_norm": 10.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3231, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.3586206896551725, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3293, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.386206896551724, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2889, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2313, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.4413793103448276, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2049, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.4689655172413794, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2044, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.4965517241379311, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.204, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.524137931034483, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1166, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1098, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.5793103448275863, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1162, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.6068965517241378, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0942, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.6344827586206896, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0759, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.6620689655172414, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0901, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0875, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.717241379310345, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0758, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.7448275862068967, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0695, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.7724137931034483, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0777, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0947, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 3.375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0626, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.8551724137931034, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0483, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.8827586206896552, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0452, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.910344827586207, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0761, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.9379310344827587, | |
| "grad_norm": 7.25, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0765, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0445, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.993103448275862, | |
| "grad_norm": 7.75, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0907, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0083, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.12860961258411407, | |
| "eval_runtime": 26.6866, | |
| "eval_samples_per_second": 18.736, | |
| "eval_steps_per_second": 2.361, | |
| "step": 74 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 222, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.262645560659763e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |