| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 2, | |
| "global_step": 56, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 499.6013488769531, | |
| "learning_rate": 0.00035333138396736785, | |
| "loss": 5.7451, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "eval_loss": 4.145218372344971, | |
| "eval_runtime": 115.5605, | |
| "eval_samples_per_second": 3.029, | |
| "eval_steps_per_second": 0.052, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 41.22399139404297, | |
| "learning_rate": 0.0003402450364130209, | |
| "loss": 5.3316, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "eval_loss": 4.5622239112854, | |
| "eval_runtime": 159.5, | |
| "eval_samples_per_second": 2.194, | |
| "eval_steps_per_second": 0.038, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.202932596206665, | |
| "learning_rate": 0.00032715868885867396, | |
| "loss": 5.0703, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "eval_loss": 4.1197967529296875, | |
| "eval_runtime": 110.2677, | |
| "eval_samples_per_second": 3.174, | |
| "eval_steps_per_second": 0.054, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.4056661128997803, | |
| "learning_rate": 0.00031407234130432696, | |
| "loss": 4.869, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "eval_loss": 4.0951642990112305, | |
| "eval_runtime": 133.7942, | |
| "eval_samples_per_second": 2.616, | |
| "eval_steps_per_second": 0.045, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.29455429315567017, | |
| "learning_rate": 0.00030098599374998, | |
| "loss": 4.854, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "eval_loss": 4.096081733703613, | |
| "eval_runtime": 117.4958, | |
| "eval_samples_per_second": 2.979, | |
| "eval_steps_per_second": 0.051, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.10492418706417084, | |
| "learning_rate": 0.00028789964619563307, | |
| "loss": 4.8537, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "eval_loss": 4.0943403244018555, | |
| "eval_runtime": 134.1581, | |
| "eval_samples_per_second": 2.609, | |
| "eval_steps_per_second": 0.045, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.016932426020503044, | |
| "learning_rate": 0.0002748132986412861, | |
| "loss": 4.8523, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 4.094045162200928, | |
| "eval_runtime": 151.7694, | |
| "eval_samples_per_second": 2.306, | |
| "eval_steps_per_second": 0.04, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.024902725592255592, | |
| "learning_rate": 0.0002617269510869391, | |
| "loss": 4.8522, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "eval_loss": 4.094315528869629, | |
| "eval_runtime": 180.0495, | |
| "eval_samples_per_second": 1.944, | |
| "eval_steps_per_second": 0.033, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.03983244299888611, | |
| "learning_rate": 0.0002486406035325922, | |
| "loss": 4.8525, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "eval_loss": 4.094301223754883, | |
| "eval_runtime": 111.1902, | |
| "eval_samples_per_second": 3.148, | |
| "eval_steps_per_second": 0.054, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.021751634776592255, | |
| "learning_rate": 0.00023555425597824523, | |
| "loss": 4.8524, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "eval_loss": 4.094165802001953, | |
| "eval_runtime": 122.9871, | |
| "eval_samples_per_second": 2.846, | |
| "eval_steps_per_second": 0.049, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.01534917950630188, | |
| "learning_rate": 0.00022246790842389826, | |
| "loss": 4.8522, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "eval_loss": 4.094093322753906, | |
| "eval_runtime": 155.9317, | |
| "eval_samples_per_second": 2.245, | |
| "eval_steps_per_second": 0.038, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.01246840413659811, | |
| "learning_rate": 0.0002093815608695513, | |
| "loss": 4.8522, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "eval_loss": 4.094055652618408, | |
| "eval_runtime": 199.4288, | |
| "eval_samples_per_second": 1.755, | |
| "eval_steps_per_second": 0.03, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.00821756012737751, | |
| "learning_rate": 0.00019629521331520434, | |
| "loss": 4.8521, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "eval_loss": 4.094019412994385, | |
| "eval_runtime": 110.2163, | |
| "eval_samples_per_second": 3.176, | |
| "eval_steps_per_second": 0.054, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.00944442953914404, | |
| "learning_rate": 0.0001832088657608574, | |
| "loss": 4.3067, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 4.093985557556152, | |
| "eval_runtime": 109.3834, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 0.055, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.0032597698736935854, | |
| "learning_rate": 0.00017012251820651045, | |
| "loss": 4.8521, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "eval_loss": 4.093966960906982, | |
| "eval_runtime": 110.1116, | |
| "eval_samples_per_second": 3.179, | |
| "eval_steps_per_second": 0.054, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.0018325659912079573, | |
| "learning_rate": 0.00015703617065216348, | |
| "loss": 4.8521, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "eval_loss": 4.093964099884033, | |
| "eval_runtime": 110.0212, | |
| "eval_samples_per_second": 3.181, | |
| "eval_steps_per_second": 0.055, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.002593304729089141, | |
| "learning_rate": 0.00014394982309781654, | |
| "loss": 4.8521, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "eval_loss": 4.093966007232666, | |
| "eval_runtime": 162.0241, | |
| "eval_samples_per_second": 2.16, | |
| "eval_steps_per_second": 0.037, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.003515275428071618, | |
| "learning_rate": 0.00013086347554346956, | |
| "loss": 4.8521, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "eval_loss": 4.093967437744141, | |
| "eval_runtime": 113.1292, | |
| "eval_samples_per_second": 3.094, | |
| "eval_steps_per_second": 0.053, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.004078236408531666, | |
| "learning_rate": 0.00011777712798912262, | |
| "loss": 4.8521, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "eval_loss": 4.09396505355835, | |
| "eval_runtime": 184.5537, | |
| "eval_samples_per_second": 1.896, | |
| "eval_steps_per_second": 0.033, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.003892536973580718, | |
| "learning_rate": 0.00010469078043477564, | |
| "loss": 4.8521, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "eval_loss": 4.093959331512451, | |
| "eval_runtime": 113.0981, | |
| "eval_samples_per_second": 3.095, | |
| "eval_steps_per_second": 0.053, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.0023005367256700993, | |
| "learning_rate": 9.16044328804287e-05, | |
| "loss": 4.852, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_loss": 4.093954563140869, | |
| "eval_runtime": 118.4819, | |
| "eval_samples_per_second": 2.954, | |
| "eval_steps_per_second": 0.051, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.0026939427480101585, | |
| "learning_rate": 7.851808532608174e-05, | |
| "loss": 4.852, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "eval_loss": 4.09395170211792, | |
| "eval_runtime": 139.5353, | |
| "eval_samples_per_second": 2.508, | |
| "eval_steps_per_second": 0.043, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.0018068618373945355, | |
| "learning_rate": 6.543173777173478e-05, | |
| "loss": 4.852, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "eval_loss": 4.0939507484436035, | |
| "eval_runtime": 119.3509, | |
| "eval_samples_per_second": 2.933, | |
| "eval_steps_per_second": 0.05, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.0018717749044299126, | |
| "learning_rate": 5.234539021738782e-05, | |
| "loss": 4.852, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "eval_loss": 4.093950271606445, | |
| "eval_runtime": 127.677, | |
| "eval_samples_per_second": 2.741, | |
| "eval_steps_per_second": 0.047, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.001300295814871788, | |
| "learning_rate": 3.925904266304087e-05, | |
| "loss": 4.852, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "eval_loss": 4.093948841094971, | |
| "eval_runtime": 140.3917, | |
| "eval_samples_per_second": 2.493, | |
| "eval_steps_per_second": 0.043, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.0016825235215947032, | |
| "learning_rate": 2.617269510869391e-05, | |
| "loss": 4.852, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "eval_loss": 4.093948841094971, | |
| "eval_runtime": 132.6149, | |
| "eval_samples_per_second": 2.639, | |
| "eval_steps_per_second": 0.045, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.0014735407894477248, | |
| "learning_rate": 1.3086347554346956e-05, | |
| "loss": 4.852, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "eval_loss": 4.093948841094971, | |
| "eval_runtime": 123.3755, | |
| "eval_samples_per_second": 2.837, | |
| "eval_steps_per_second": 0.049, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.0010560819646343589, | |
| "learning_rate": 0.0, | |
| "loss": 4.3066, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4.093948841094971, | |
| "eval_runtime": 107.7247, | |
| "eval_samples_per_second": 3.249, | |
| "eval_steps_per_second": 0.056, | |
| "step": 56 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 56, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 407052651766068.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "_wandb": {}, | |
| "assignments": {}, | |
| "decay": 0.01, | |
| "learning_rate": 0.0003664177315217148, | |
| "metric": "eval/loss", | |
| "per_device_train_batch_size": 128 | |
| } | |
| } | |