| { | |
| "best_metric": 0.9049463272094727, | |
| "best_model_checkpoint": "./tst-translation-output2/checkpoint-12000", | |
| "epoch": 1.8217701533323214, | |
| "eval_steps": 1500, | |
| "global_step": 12000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 1.7623, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.98913256102067e-05, | |
| "loss": 1.3653, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.97826512204134e-05, | |
| "loss": 1.2499, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_bleu": 6.1112, | |
| "eval_gen_len": 18.0495, | |
| "eval_loss": 1.1805654764175415, | |
| "eval_runtime": 916.5943, | |
| "eval_samples_per_second": 14.371, | |
| "eval_steps_per_second": 0.899, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.96739768306201e-05, | |
| "loss": 1.2038, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4.95653024408268e-05, | |
| "loss": 1.1357, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.945662805103349e-05, | |
| "loss": 1.1007, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_bleu": 7.4845, | |
| "eval_gen_len": 17.6068, | |
| "eval_loss": 1.0686322450637817, | |
| "eval_runtime": 870.5195, | |
| "eval_samples_per_second": 15.131, | |
| "eval_steps_per_second": 0.947, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.9347953661240194e-05, | |
| "loss": 1.0769, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.923927927144689e-05, | |
| "loss": 1.0708, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 4.9130604881653594e-05, | |
| "loss": 1.0334, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_bleu": 9.0076, | |
| "eval_gen_len": 17.6214, | |
| "eval_loss": 1.0013340711593628, | |
| "eval_runtime": 888.8463, | |
| "eval_samples_per_second": 14.819, | |
| "eval_steps_per_second": 0.927, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 4.902193049186029e-05, | |
| "loss": 1.0175, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.891325610206699e-05, | |
| "loss": 1.0074, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.880458171227369e-05, | |
| "loss": 0.992, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_bleu": 8.6786, | |
| "eval_gen_len": 17.868, | |
| "eval_loss": 0.9599339365959167, | |
| "eval_runtime": 933.1213, | |
| "eval_samples_per_second": 14.116, | |
| "eval_steps_per_second": 0.883, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.8695907322480386e-05, | |
| "loss": 0.9887, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 4.858723293268709e-05, | |
| "loss": 0.7834, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 4.847855854289378e-05, | |
| "loss": 0.7881, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_bleu": 9.2343, | |
| "eval_gen_len": 17.2061, | |
| "eval_loss": 0.964439868927002, | |
| "eval_runtime": 844.707, | |
| "eval_samples_per_second": 15.594, | |
| "eval_steps_per_second": 0.975, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 4.836988415310048e-05, | |
| "loss": 0.7667, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 4.826120976330718e-05, | |
| "loss": 0.7632, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 4.815253537351388e-05, | |
| "loss": 0.7675, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_bleu": 10.0578, | |
| "eval_gen_len": 17.6006, | |
| "eval_loss": 0.942681610584259, | |
| "eval_runtime": 1046.0952, | |
| "eval_samples_per_second": 12.592, | |
| "eval_steps_per_second": 0.788, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 4.804386098372058e-05, | |
| "loss": 0.774, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 4.793518659392728e-05, | |
| "loss": 0.7731, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 4.782651220413398e-05, | |
| "loss": 0.7665, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_bleu": 10.436, | |
| "eval_gen_len": 17.2095, | |
| "eval_loss": 0.9237579703330994, | |
| "eval_runtime": 854.5281, | |
| "eval_samples_per_second": 15.414, | |
| "eval_steps_per_second": 0.964, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 4.7717837814340674e-05, | |
| "loss": 0.761, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 4.760916342454738e-05, | |
| "loss": 0.7653, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 4.750048903475407e-05, | |
| "loss": 0.7707, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "eval_bleu": 10.5971, | |
| "eval_gen_len": 17.2971, | |
| "eval_loss": 0.9049463272094727, | |
| "eval_runtime": 851.7484, | |
| "eval_samples_per_second": 15.465, | |
| "eval_steps_per_second": 0.967, | |
| "step": 12000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 230545, | |
| "num_train_epochs": 35, | |
| "save_steps": 1500, | |
| "total_flos": 4.16088915443712e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |