| { | |
| "best_metric": 0.8032786885245902, | |
| "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-7dukmcwd/checkpoint-400", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.6303067207336426, | |
| "learning_rate": 3.5703870009677385e-05, | |
| "loss": 0.6604, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.9712932109832764, | |
| "learning_rate": 3.47883861632754e-05, | |
| "loss": 0.5592, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.1559805870056152, | |
| "learning_rate": 3.387290231687342e-05, | |
| "loss": 0.4817, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 8.0116548538208, | |
| "learning_rate": 3.295741847047143e-05, | |
| "loss": 0.4318, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 19.13960075378418, | |
| "learning_rate": 3.204193462406945e-05, | |
| "loss": 0.5086, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 13.061105728149414, | |
| "learning_rate": 3.1126450777667465e-05, | |
| "loss": 0.527, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 12.58017635345459, | |
| "learning_rate": 3.0210966931265478e-05, | |
| "loss": 0.4828, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.202507495880127, | |
| "learning_rate": 2.9295483084863497e-05, | |
| "loss": 0.4613, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 5.813719749450684, | |
| "learning_rate": 2.8379999238461513e-05, | |
| "loss": 0.4274, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 8.482573509216309, | |
| "learning_rate": 2.746451539205953e-05, | |
| "loss": 0.4159, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 9.623395919799805, | |
| "learning_rate": 2.654903154565754e-05, | |
| "loss": 0.3306, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.012599468231201, | |
| "learning_rate": 2.5633547699255557e-05, | |
| "loss": 0.403, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.260857343673706, | |
| "learning_rate": 2.4718063852853577e-05, | |
| "loss": 0.3636, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 18.5455379486084, | |
| "learning_rate": 2.3802580006451593e-05, | |
| "loss": 0.3621, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.035172700881958, | |
| "learning_rate": 2.2887096160049605e-05, | |
| "loss": 0.376, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.068894386291504, | |
| "learning_rate": 2.197161231364762e-05, | |
| "loss": 0.3452, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.0021464824676514, | |
| "learning_rate": 2.1056128467245637e-05, | |
| "loss": 0.3577, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.3914709091186523, | |
| "learning_rate": 2.0140644620843656e-05, | |
| "loss": 0.4271, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 8.317371368408203, | |
| "learning_rate": 1.922516077444167e-05, | |
| "loss": 0.289, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.967580258846283, | |
| "learning_rate": 1.8309676928039685e-05, | |
| "loss": 0.3738, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.87, | |
| "eval_f1": 0.7657657657657657, | |
| "eval_loss": 0.32556435465812683, | |
| "eval_precision": 0.8585858585858586, | |
| "eval_recall": 0.6910569105691057, | |
| "eval_runtime": 1.5112, | |
| "eval_samples_per_second": 264.697, | |
| "eval_steps_per_second": 16.544, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 4.126100540161133, | |
| "learning_rate": 1.73941930816377e-05, | |
| "loss": 0.2486, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 5.108118057250977, | |
| "learning_rate": 1.6478709235235717e-05, | |
| "loss": 0.3034, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 11.375035285949707, | |
| "learning_rate": 1.5563225388833733e-05, | |
| "loss": 0.1486, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 8.199675559997559, | |
| "learning_rate": 1.4647741542431748e-05, | |
| "loss": 0.327, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 4.900712013244629, | |
| "learning_rate": 1.3732257696029764e-05, | |
| "loss": 0.2753, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.31448882818222046, | |
| "learning_rate": 1.2816773849627779e-05, | |
| "loss": 0.2028, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.3391319513320923, | |
| "learning_rate": 1.1901290003225796e-05, | |
| "loss": 0.2992, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 20.60084342956543, | |
| "learning_rate": 1.098580615682381e-05, | |
| "loss": 0.4703, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 7.974413871765137, | |
| "learning_rate": 1.0070322310421828e-05, | |
| "loss": 0.2649, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 11.488137245178223, | |
| "learning_rate": 9.154838464019842e-06, | |
| "loss": 0.438, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.5850751399993896, | |
| "learning_rate": 8.239354617617858e-06, | |
| "loss": 0.1704, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.258329391479492, | |
| "learning_rate": 7.323870771215874e-06, | |
| "loss": 0.226, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 6.117366790771484, | |
| "learning_rate": 6.408386924813889e-06, | |
| "loss": 0.2219, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 28.112499237060547, | |
| "learning_rate": 5.492903078411905e-06, | |
| "loss": 0.2595, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 15.969998359680176, | |
| "learning_rate": 4.577419232009921e-06, | |
| "loss": 0.3709, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.6372332572937012, | |
| "learning_rate": 3.661935385607937e-06, | |
| "loss": 0.2678, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.20131894946098328, | |
| "learning_rate": 2.7464515392059526e-06, | |
| "loss": 0.4027, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 12.212553024291992, | |
| "learning_rate": 1.8309676928039686e-06, | |
| "loss": 0.3156, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.19253325462341309, | |
| "learning_rate": 9.154838464019843e-07, | |
| "loss": 0.1869, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 14.685490608215332, | |
| "learning_rate": 0.0, | |
| "loss": 0.1768, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.88, | |
| "eval_f1": 0.8032786885245902, | |
| "eval_loss": 0.3457476794719696, | |
| "eval_precision": 0.8099173553719008, | |
| "eval_recall": 0.7967479674796748, | |
| "eval_runtime": 1.5723, | |
| "eval_samples_per_second": 254.401, | |
| "eval_steps_per_second": 15.9, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 400, | |
| "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": 423630740901888.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "_wandb": {}, | |
| "assignments": {}, | |
| "learning_rate": 3.661935385607937e-05, | |
| "metric": "eval/loss", | |
| "num_train_epochs": 2, | |
| "per_device_train_batch_size": 8, | |
| "seed": 36 | |
| } | |
| } | |