| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "global_step": 68220, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 2.780123131046614e-05, | |
| "loss": 0.635, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_gen_len": 34.1, | |
| "eval_loss": 0.5430836081504822, | |
| "eval_rouge1": 19.0735, | |
| "eval_rouge2": 11.8399, | |
| "eval_rougeL": 18.8386, | |
| "eval_rougeLsum": 18.9348, | |
| "eval_runtime": 3057.941, | |
| "eval_samples_per_second": 1.116, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 2.5602462620932278e-05, | |
| "loss": 0.4721, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_gen_len": 34.1, | |
| "eval_loss": 0.5901889204978943, | |
| "eval_rouge1": 18.6178, | |
| "eval_rouge2": 11.5042, | |
| "eval_rougeL": 18.3666, | |
| "eval_rougeLsum": 18.4882, | |
| "eval_runtime": 3066.7913, | |
| "eval_samples_per_second": 1.113, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 2.3403693931398417e-05, | |
| "loss": 0.3803, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_gen_len": 33.1, | |
| "eval_loss": 0.6489335298538208, | |
| "eval_rouge1": 20.6093, | |
| "eval_rouge2": 12.6294, | |
| "eval_rougeL": 20.2763, | |
| "eval_rougeLsum": 20.3561, | |
| "eval_runtime": 2950.4654, | |
| "eval_samples_per_second": 1.157, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 2.120492524186456e-05, | |
| "loss": 0.3072, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_gen_len": 33.0, | |
| "eval_loss": 0.6617271304130554, | |
| "eval_rouge1": 21.175, | |
| "eval_rouge2": 12.9525, | |
| "eval_rougeL": 20.85, | |
| "eval_rougeLsum": 20.9327, | |
| "eval_runtime": 2933.4173, | |
| "eval_samples_per_second": 1.163, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 1.9006156552330694e-05, | |
| "loss": 0.225, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "eval_gen_len": 33.7, | |
| "eval_loss": 0.7911365628242493, | |
| "eval_rouge1": 20.0159, | |
| "eval_rouge2": 12.3403, | |
| "eval_rougeL": 19.7354, | |
| "eval_rougeLsum": 19.831, | |
| "eval_runtime": 3007.359, | |
| "eval_samples_per_second": 1.135, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 1.6807387862796836e-05, | |
| "loss": 0.1862, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "eval_gen_len": 33.5, | |
| "eval_loss": 0.9180275201797485, | |
| "eval_rouge1": 20.6775, | |
| "eval_rouge2": 12.58, | |
| "eval_rougeL": 20.3712, | |
| "eval_rougeLsum": 20.4162, | |
| "eval_runtime": 2990.0617, | |
| "eval_samples_per_second": 1.141, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 1.4608619173262973e-05, | |
| "loss": 0.1559, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "eval_gen_len": 33.2, | |
| "eval_loss": 1.0511494874954224, | |
| "eval_rouge1": 20.3652, | |
| "eval_rouge2": 12.4459, | |
| "eval_rougeL": 20.0386, | |
| "eval_rougeLsum": 20.1371, | |
| "eval_runtime": 2942.5442, | |
| "eval_samples_per_second": 1.16, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 1.2409850483729111e-05, | |
| "loss": 0.1256, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "eval_gen_len": 33.2, | |
| "eval_loss": 1.1184989213943481, | |
| "eval_rouge1": 20.6117, | |
| "eval_rouge2": 12.6157, | |
| "eval_rougeL": 20.3466, | |
| "eval_rougeLsum": 20.4447, | |
| "eval_runtime": 2945.923, | |
| "eval_samples_per_second": 1.159, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 1.0211081794195251e-05, | |
| "loss": 0.1044, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "eval_gen_len": 33.5, | |
| "eval_loss": 1.262742280960083, | |
| "eval_rouge1": 20.4935, | |
| "eval_rouge2": 12.393, | |
| "eval_rougeL": 20.1894, | |
| "eval_rougeLsum": 20.2828, | |
| "eval_runtime": 2982.3339, | |
| "eval_samples_per_second": 1.144, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "learning_rate": 8.01231310466139e-06, | |
| "loss": 0.0844, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "eval_gen_len": 33.3, | |
| "eval_loss": 1.5101187229156494, | |
| "eval_rouge1": 21.5004, | |
| "eval_rouge2": 12.9803, | |
| "eval_rougeL": 21.2348, | |
| "eval_rougeLsum": 21.3046, | |
| "eval_runtime": 2964.055, | |
| "eval_samples_per_second": 1.151, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "learning_rate": 5.813544415127529e-06, | |
| "loss": 0.075, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "eval_gen_len": 33.1, | |
| "eval_loss": 1.5069748163223267, | |
| "eval_rouge1": 20.8058, | |
| "eval_rouge2": 12.4951, | |
| "eval_rougeL": 20.5311, | |
| "eval_rougeLsum": 20.6195, | |
| "eval_runtime": 2941.4323, | |
| "eval_samples_per_second": 1.16, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 3.6147757255936676e-06, | |
| "loss": 0.0612, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "eval_gen_len": 33.2, | |
| "eval_loss": 1.579475998878479, | |
| "eval_rouge1": 20.9465, | |
| "eval_rouge2": 12.6351, | |
| "eval_rougeL": 20.6478, | |
| "eval_rougeLsum": 20.7292, | |
| "eval_runtime": 2955.2469, | |
| "eval_samples_per_second": 1.155, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 9.53, | |
| "learning_rate": 1.4160070360598064e-06, | |
| "loss": 0.0519, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 9.53, | |
| "eval_gen_len": 33.2, | |
| "eval_loss": 1.6573169231414795, | |
| "eval_rouge1": 20.9384, | |
| "eval_rouge2": 12.6776, | |
| "eval_rougeL": 20.6249, | |
| "eval_rougeLsum": 20.7352, | |
| "eval_runtime": 2938.7831, | |
| "eval_samples_per_second": 1.161, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 68220, | |
| "total_flos": 5.6695121367374234e+17, | |
| "train_runtime": 78827.7854, | |
| "train_samples_per_second": 0.865 | |
| } | |
| ], | |
| "max_steps": 68220, | |
| "num_train_epochs": 10, | |
| "total_flos": 5.6695121367374234e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |