| { | |
| "best_metric": 0.3500404357910156, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_60-samples_config-1_auto/checkpoint-40", | |
| "epoch": 13.91304347826087, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 3.7691752910614014, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 3.0026, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.4641261100769043, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7846, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.899853229522705, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.1804, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 2.1299986839294434, | |
| "eval_runtime": 11.2952, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 2.411996841430664, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.1303, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 2.741853713989258, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.9157, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 2.4497408866882324, | |
| "learning_rate": 4e-05, | |
| "loss": 1.7781, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 1.3088672161102295, | |
| "eval_runtime": 11.2994, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.4546754360198975, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.4747, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 2.3744359016418457, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.0618, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 1.3976430892944336, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.7143, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 0.5686618685722351, | |
| "eval_runtime": 11.3022, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.8169606328010559, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.7933, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.8474010825157166, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5381, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 1.0742378234863281, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.3741, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.4143681526184082, | |
| "eval_runtime": 11.2973, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.9074451327323914, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.3907, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 5.4037675857543945, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.4435, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.6537113785743713, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.2554, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.3896390497684479, | |
| "eval_runtime": 11.2938, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.6077197194099426, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.3605, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.49757155776023865, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.3508, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 1.0307750701904297, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.2567, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.3594280779361725, | |
| "eval_runtime": 11.2972, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.30974429845809937, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.2566, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.41703227162361145, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.3267, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.33663758635520935, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.2394, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.3500404357910156, | |
| "eval_runtime": 11.2975, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.3642476201057434, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.1899, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.2749492824077606, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.1772, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.47548526525497437, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.3262, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.3681468963623047, | |
| "eval_runtime": 11.2988, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.3659931719303131, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.1458, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.44647130370140076, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.204, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.4215104877948761, | |
| "eval_runtime": 11.3015, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.4822731614112854, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.2099, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 1.2417945861816406, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.1104, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.4501473605632782, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.1772, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 0.4067264795303345, | |
| "eval_runtime": 11.2968, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.5277315974235535, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.1481, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.6794707179069519, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0831, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 0.7059261202812195, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.1163, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 0.4378962516784668, | |
| "eval_runtime": 11.3026, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.4330620765686035, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.0663, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 0.2349332571029663, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.0336, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.409313440322876, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.062, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.5496610999107361, | |
| "eval_runtime": 11.2985, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.2626000642776489, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0191, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.46747857332229614, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.0384, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 1.0589109659194946, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.0248, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.6054384112358093, | |
| "eval_runtime": 11.2975, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.5462568998336792, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0128, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 0.4290764629840851, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.0065, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.5436171293258667, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0184, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.6328181028366089, | |
| "eval_runtime": 11.2976, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "step": 80, | |
| "total_flos": 5.158217693580493e+16, | |
| "train_loss": 0.5104853815515525, | |
| "train_runtime": 1905.3918, | |
| "train_samples_per_second": 1.207, | |
| "train_steps_per_second": 0.131 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.158217693580493e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |