| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9981515711645101, | |
| "eval_steps": 500, | |
| "global_step": 270, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018484288354898338, | |
| "grad_norm": 422.0, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 20.7086, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.036968576709796676, | |
| "grad_norm": 185.0, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 19.253, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05545286506469501, | |
| "grad_norm": 147.0, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 16.7138, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07393715341959335, | |
| "grad_norm": 62.75, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 14.0999, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09242144177449169, | |
| "grad_norm": 46.25, | |
| "learning_rate": 1.851851851851852e-05, | |
| "loss": 11.9456, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11090573012939002, | |
| "grad_norm": 49.75, | |
| "learning_rate": 1.9992479525042305e-05, | |
| "loss": 10.3044, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12939001848428835, | |
| "grad_norm": 88.5, | |
| "learning_rate": 1.9946562024066018e-05, | |
| "loss": 9.602, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1478743068391867, | |
| "grad_norm": 49.25, | |
| "learning_rate": 1.9859096633447965e-05, | |
| "loss": 9.1612, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16635859519408502, | |
| "grad_norm": 114.0, | |
| "learning_rate": 1.973044870579824e-05, | |
| "loss": 8.5549, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18484288354898337, | |
| "grad_norm": 54.75, | |
| "learning_rate": 1.95611556177388e-05, | |
| "loss": 8.3168, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2033271719038817, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.93519245252219e-05, | |
| "loss": 7.8042, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22181146025878004, | |
| "grad_norm": 55.75, | |
| "learning_rate": 1.9103629409661468e-05, | |
| "loss": 7.4335, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24029574861367836, | |
| "grad_norm": 17.5, | |
| "learning_rate": 1.881730742721608e-05, | |
| "loss": 7.4047, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2587800369685767, | |
| "grad_norm": 52.5, | |
| "learning_rate": 1.8494154576472976e-05, | |
| "loss": 7.0682, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.27726432532347506, | |
| "grad_norm": 69.5, | |
| "learning_rate": 1.8135520702629677e-05, | |
| "loss": 7.1519, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2957486136783734, | |
| "grad_norm": 23.125, | |
| "learning_rate": 1.7742903859041324e-05, | |
| "loss": 6.9284, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.3142329020332717, | |
| "grad_norm": 17.25, | |
| "learning_rate": 1.7317944049686125e-05, | |
| "loss": 6.723, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.33271719038817005, | |
| "grad_norm": 103.5, | |
| "learning_rate": 1.686241637868734e-05, | |
| "loss": 6.6468, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3512014787430684, | |
| "grad_norm": 17.625, | |
| "learning_rate": 1.637822363550706e-05, | |
| "loss": 6.5427, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 1.586738834678418e-05, | |
| "loss": 6.3795, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38817005545286504, | |
| "grad_norm": 17.875, | |
| "learning_rate": 1.5332044328016916e-05, | |
| "loss": 6.4565, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4066543438077634, | |
| "grad_norm": 68.0, | |
| "learning_rate": 1.4774427770379492e-05, | |
| "loss": 6.416, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.42513863216266173, | |
| "grad_norm": 71.5, | |
| "learning_rate": 1.4196867899904292e-05, | |
| "loss": 6.2522, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4436229205175601, | |
| "grad_norm": 28.875, | |
| "learning_rate": 1.3601777248047105e-05, | |
| "loss": 6.0918, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.46210720887245843, | |
| "grad_norm": 25.75, | |
| "learning_rate": 1.2991641574276419e-05, | |
| "loss": 6.0943, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4805914972273567, | |
| "grad_norm": 24.375, | |
| "learning_rate": 1.2369009482781191e-05, | |
| "loss": 6.0615, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.49907578558225507, | |
| "grad_norm": 53.0, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 5.9591, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5175600739371534, | |
| "grad_norm": 93.5, | |
| "learning_rate": 1.1096700594125318e-05, | |
| "loss": 6.0178, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5360443622920518, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 1.0452338371907065e-05, | |
| "loss": 5.8597, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5545286506469501, | |
| "grad_norm": 102.0, | |
| "learning_rate": 9.806086682281759e-06, | |
| "loss": 5.8942, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5730129390018485, | |
| "grad_norm": 30.125, | |
| "learning_rate": 9.160644990030932e-06, | |
| "loss": 5.9999, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5914972273567468, | |
| "grad_norm": 23.25, | |
| "learning_rate": 8.518709376487515e-06, | |
| "loss": 5.9019, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.609981515711645, | |
| "grad_norm": 56.0, | |
| "learning_rate": 7.882961277705897e-06, | |
| "loss": 5.8274, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6284658040665434, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 7.256056283806987e-06, | |
| "loss": 5.9167, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6469500924214417, | |
| "grad_norm": 26.75, | |
| "learning_rate": 6.640613046284581e-06, | |
| "loss": 5.9097, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6654343807763401, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 6.039202339608432e-06, | |
| "loss": 5.6327, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6839186691312384, | |
| "grad_norm": 26.75, | |
| "learning_rate": 5.454336322814995e-06, | |
| "loss": 5.6804, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7024029574861368, | |
| "grad_norm": 30.375, | |
| "learning_rate": 4.888458045941269e-06, | |
| "loss": 5.7316, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7208872458410351, | |
| "grad_norm": 39.5, | |
| "learning_rate": 4.343931245134616e-06, | |
| "loss": 5.7299, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "grad_norm": 11.75, | |
| "learning_rate": 3.823030469065431e-06, | |
| "loss": 5.6982, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7578558225508318, | |
| "grad_norm": 20.625, | |
| "learning_rate": 3.3279315778858034e-06, | |
| "loss": 5.862, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7763401109057301, | |
| "grad_norm": 50.75, | |
| "learning_rate": 2.8607026544210115e-06, | |
| "loss": 5.7507, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7948243992606284, | |
| "grad_norm": 49.75, | |
| "learning_rate": 2.423295365558821e-06, | |
| "loss": 5.7101, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8133086876155268, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 2.01753680992107e-06, | |
| "loss": 5.8428, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8317929759704251, | |
| "grad_norm": 34.75, | |
| "learning_rate": 1.6451218858706374e-06, | |
| "loss": 5.7162, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8502772643253235, | |
| "grad_norm": 47.0, | |
| "learning_rate": 1.307606211733522e-06, | |
| "loss": 5.6621, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8687615526802218, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.0063996278090704e-06, | |
| "loss": 5.6616, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8872458410351202, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 7.427603073110967e-07, | |
| "loss": 5.8559, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9057301293900185, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 5.177895008392353e-07, | |
| "loss": 5.7149, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9242144177449169, | |
| "grad_norm": 61.75, | |
| "learning_rate": 3.3242693633337986e-07, | |
| "loss": 5.9011, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9426987060998152, | |
| "grad_norm": 13.625, | |
| "learning_rate": 1.874468937261531e-07, | |
| "loss": 5.685, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9611829944547134, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 8.345497068998897e-08, | |
| "loss": 5.7765, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9796672828096118, | |
| "grad_norm": 18.5, | |
| "learning_rate": 2.088555298867978e-08, | |
| "loss": 5.734, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "grad_norm": 28.25, | |
| "learning_rate": 0.0, | |
| "loss": 5.7909, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "eval_loss": 5.75302791595459, | |
| "eval_runtime": 376.2306, | |
| "eval_samples_per_second": 2.581, | |
| "eval_steps_per_second": 0.324, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "step": 270, | |
| "total_flos": 5.260333472022528e+16, | |
| "train_loss": 7.3803352285314485, | |
| "train_runtime": 2516.7229, | |
| "train_samples_per_second": 3.438, | |
| "train_steps_per_second": 0.107 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 270, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.260333472022528e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |