| { | |
| "best_global_step": 2480, | |
| "best_metric": 0.9160613397901534, | |
| "best_model_checkpoint": "checkpoints/checkpoint-2480", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3720, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08064516129032258, | |
| "grad_norm": 2.9195146560668945, | |
| "learning_rate": 1.946236559139785e-05, | |
| "loss": 0.5794, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 2.377180814743042, | |
| "learning_rate": 1.89247311827957e-05, | |
| "loss": 0.3641, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24193548387096775, | |
| "grad_norm": 2.9099831581115723, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.3137, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 3.255298614501953, | |
| "learning_rate": 1.78494623655914e-05, | |
| "loss": 0.283, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4032258064516129, | |
| "grad_norm": 2.6660268306732178, | |
| "learning_rate": 1.7311827956989248e-05, | |
| "loss": 0.2707, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 4.071249961853027, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.2905, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5645161290322581, | |
| "grad_norm": 2.1681008338928223, | |
| "learning_rate": 1.6236559139784947e-05, | |
| "loss": 0.2726, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 8.003071784973145, | |
| "learning_rate": 1.5698924731182796e-05, | |
| "loss": 0.2473, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7258064516129032, | |
| "grad_norm": 4.0680012702941895, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.2444, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 2.0538086891174316, | |
| "learning_rate": 1.4623655913978497e-05, | |
| "loss": 0.2568, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8870967741935484, | |
| "grad_norm": 6.883822441101074, | |
| "learning_rate": 1.4086021505376346e-05, | |
| "loss": 0.2754, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 3.087297201156616, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.2397, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9136400322841001, | |
| "eval_f1": 0.9094308764970639, | |
| "eval_loss": 0.2508734464645386, | |
| "eval_runtime": 187.5594, | |
| "eval_samples_per_second": 13.212, | |
| "eval_steps_per_second": 0.826, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0483870967741935, | |
| "grad_norm": 3.719700574874878, | |
| "learning_rate": 1.3010752688172043e-05, | |
| "loss": 0.2378, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.129032258064516, | |
| "grad_norm": 8.953502655029297, | |
| "learning_rate": 1.2473118279569894e-05, | |
| "loss": 0.2264, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.2096774193548387, | |
| "grad_norm": 0.8525077700614929, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.2135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 1.5267659425735474, | |
| "learning_rate": 1.1397849462365593e-05, | |
| "loss": 0.2204, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.370967741935484, | |
| "grad_norm": 5.281778812408447, | |
| "learning_rate": 1.086021505376344e-05, | |
| "loss": 0.2313, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.4516129032258065, | |
| "grad_norm": 9.744083404541016, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.2307, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.532258064516129, | |
| "grad_norm": 0.6601889133453369, | |
| "learning_rate": 9.78494623655914e-06, | |
| "loss": 0.2017, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.6129032258064515, | |
| "grad_norm": 2.3461925983428955, | |
| "learning_rate": 9.24731182795699e-06, | |
| "loss": 0.1941, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6935483870967742, | |
| "grad_norm": 5.637911796569824, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.2293, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.7741935483870968, | |
| "grad_norm": 7.581858158111572, | |
| "learning_rate": 8.172043010752689e-06, | |
| "loss": 0.1841, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.8548387096774195, | |
| "grad_norm": 2.4026706218719482, | |
| "learning_rate": 7.634408602150538e-06, | |
| "loss": 0.2177, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "grad_norm": 2.8542771339416504, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.2334, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9160613397901534, | |
| "eval_f1": 0.9112198149205201, | |
| "eval_loss": 0.25815290212631226, | |
| "eval_runtime": 138.3066, | |
| "eval_samples_per_second": 17.917, | |
| "eval_steps_per_second": 1.121, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.0161290322580645, | |
| "grad_norm": 9.511274337768555, | |
| "learning_rate": 6.5591397849462365e-06, | |
| "loss": 0.2032, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.096774193548387, | |
| "grad_norm": 5.035931587219238, | |
| "learning_rate": 6.021505376344087e-06, | |
| "loss": 0.1888, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.1774193548387095, | |
| "grad_norm": 4.586341857910156, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.1876, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.258064516129032, | |
| "grad_norm": 0.808972179889679, | |
| "learning_rate": 4.946236559139785e-06, | |
| "loss": 0.1684, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.338709677419355, | |
| "grad_norm": 2.9565162658691406, | |
| "learning_rate": 4.408602150537635e-06, | |
| "loss": 0.1295, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.4193548387096775, | |
| "grad_norm": 3.302726984024048, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.1675, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 3.93761944770813, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.1572, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.5806451612903225, | |
| "grad_norm": 4.387836933135986, | |
| "learning_rate": 2.7956989247311827e-06, | |
| "loss": 0.1553, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.661290322580645, | |
| "grad_norm": 3.010204553604126, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 0.1456, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.741935483870968, | |
| "grad_norm": 4.972431659698486, | |
| "learning_rate": 1.720430107526882e-06, | |
| "loss": 0.1419, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.8225806451612905, | |
| "grad_norm": 5.383158206939697, | |
| "learning_rate": 1.1827956989247313e-06, | |
| "loss": 0.1637, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.903225806451613, | |
| "grad_norm": 0.6802952885627747, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 0.1931, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.9838709677419355, | |
| "grad_norm": 0.42213770747184753, | |
| "learning_rate": 1.0752688172043012e-07, | |
| "loss": 0.1809, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9152542372881356, | |
| "eval_f1": 0.9130503215664975, | |
| "eval_loss": 0.3000216484069824, | |
| "eval_runtime": 48.0957, | |
| "eval_samples_per_second": 51.522, | |
| "eval_steps_per_second": 3.223, | |
| "step": 3720 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3720, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 3912364964906496.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |