| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 250, | |
| "global_step": 3975, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.685534591194969e-05, | |
| "loss": 0.9547, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_bleu": 47.1486, | |
| "eval_gen_len": 42.0851, | |
| "eval_loss": 0.7848668098449707, | |
| "eval_runtime": 216.7661, | |
| "eval_samples_per_second": 10.195, | |
| "eval_steps_per_second": 0.323, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4.3710691823899376e-05, | |
| "loss": 0.795, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_bleu": 49.685, | |
| "eval_gen_len": 42.7009, | |
| "eval_loss": 0.7136563062667847, | |
| "eval_runtime": 223.9634, | |
| "eval_samples_per_second": 9.868, | |
| "eval_steps_per_second": 0.313, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.0566037735849064e-05, | |
| "loss": 0.7304, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_bleu": 51.2955, | |
| "eval_gen_len": 42.5633, | |
| "eval_loss": 0.6703998446464539, | |
| "eval_runtime": 219.0588, | |
| "eval_samples_per_second": 10.089, | |
| "eval_steps_per_second": 0.32, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 3.7421383647798744e-05, | |
| "loss": 0.6963, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_bleu": 52.4864, | |
| "eval_gen_len": 42.395, | |
| "eval_loss": 0.6411181688308716, | |
| "eval_runtime": 230.5264, | |
| "eval_samples_per_second": 9.587, | |
| "eval_steps_per_second": 0.304, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.4276729559748424e-05, | |
| "loss": 0.6758, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_bleu": 53.4802, | |
| "eval_gen_len": 42.5258, | |
| "eval_loss": 0.6214491128921509, | |
| "eval_runtime": 221.0884, | |
| "eval_samples_per_second": 9.996, | |
| "eval_steps_per_second": 0.317, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 3.113207547169811e-05, | |
| "loss": 0.6113, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_bleu": 53.3832, | |
| "eval_gen_len": 42.0814, | |
| "eval_loss": 0.6121014952659607, | |
| "eval_runtime": 226.9957, | |
| "eval_samples_per_second": 9.736, | |
| "eval_steps_per_second": 0.308, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 2.7987421383647798e-05, | |
| "loss": 0.5779, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "eval_bleu": 54.2053, | |
| "eval_gen_len": 42.462, | |
| "eval_loss": 0.5953919291496277, | |
| "eval_runtime": 231.1354, | |
| "eval_samples_per_second": 9.561, | |
| "eval_steps_per_second": 0.303, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2.4842767295597485e-05, | |
| "loss": 0.5707, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_bleu": 54.7121, | |
| "eval_gen_len": 42.3439, | |
| "eval_loss": 0.5879282355308533, | |
| "eval_runtime": 222.7105, | |
| "eval_samples_per_second": 9.923, | |
| "eval_steps_per_second": 0.314, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2.1698113207547172e-05, | |
| "loss": 0.5765, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_bleu": 54.8816, | |
| "eval_gen_len": 42.2819, | |
| "eval_loss": 0.5788756608963013, | |
| "eval_runtime": 229.4679, | |
| "eval_samples_per_second": 9.631, | |
| "eval_steps_per_second": 0.305, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 1.8553459119496856e-05, | |
| "loss": 0.5617, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_bleu": 55.246, | |
| "eval_gen_len": 42.1756, | |
| "eval_loss": 0.5728903412818909, | |
| "eval_runtime": 223.2414, | |
| "eval_samples_per_second": 9.9, | |
| "eval_steps_per_second": 0.314, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 1.540880503144654e-05, | |
| "loss": 0.5349, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "eval_bleu": 55.2927, | |
| "eval_gen_len": 42.238, | |
| "eval_loss": 0.5718762874603271, | |
| "eval_runtime": 223.5775, | |
| "eval_samples_per_second": 9.885, | |
| "eval_steps_per_second": 0.313, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.2264150943396227e-05, | |
| "loss": 0.5171, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_bleu": 55.3323, | |
| "eval_gen_len": 42.291, | |
| "eval_loss": 0.5671454071998596, | |
| "eval_runtime": 227.9511, | |
| "eval_samples_per_second": 9.695, | |
| "eval_steps_per_second": 0.307, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 9.119496855345912e-06, | |
| "loss": 0.512, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_bleu": 55.4991, | |
| "eval_gen_len": 42.2593, | |
| "eval_loss": 0.5644388794898987, | |
| "eval_runtime": 232.3325, | |
| "eval_samples_per_second": 9.512, | |
| "eval_steps_per_second": 0.301, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 5.974842767295598e-06, | |
| "loss": 0.5129, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "eval_bleu": 55.4748, | |
| "eval_gen_len": 42.1778, | |
| "eval_loss": 0.5624849200248718, | |
| "eval_runtime": 230.2857, | |
| "eval_samples_per_second": 9.597, | |
| "eval_steps_per_second": 0.304, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2.830188679245283e-06, | |
| "loss": 0.5058, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "eval_bleu": 55.546, | |
| "eval_gen_len": 42.2176, | |
| "eval_loss": 0.560586154460907, | |
| "eval_runtime": 231.1144, | |
| "eval_samples_per_second": 9.562, | |
| "eval_steps_per_second": 0.303, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 3975, | |
| "total_flos": 3.068708155215053e+16, | |
| "train_loss": 0.6158648508899617, | |
| "train_runtime": 15318.3168, | |
| "train_samples_per_second": 8.298, | |
| "train_steps_per_second": 0.259 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 3975, | |
| "num_train_epochs": 3, | |
| "save_steps": 10000, | |
| "total_flos": 3.068708155215053e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |