| { | |
| "best_metric": 2.210977077484131, | |
| "best_model_checkpoint": "/home/lily/nos6/bloom-adapt/exps_ru/bigscience/bloom-560m_ru_adpt_sft_original-frozen_100_000samples/checkpoint-25000", | |
| "epoch": 1.462073805485701, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9.8e-05, | |
| "loss": 2.8794, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9.6e-05, | |
| "loss": 2.746, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9.4e-05, | |
| "loss": 2.6667, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 2.6093, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9e-05, | |
| "loss": 2.5672, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 2.5343, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8.6e-05, | |
| "loss": 2.4979, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8.4e-05, | |
| "loss": 2.467, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8.2e-05, | |
| "loss": 2.4436, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8e-05, | |
| "loss": 2.4269, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_loss": 2.4169912338256836, | |
| "eval_runtime": 487.268, | |
| "eval_samples_per_second": 13.656, | |
| "eval_steps_per_second": 3.415, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 2.4069, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7.6e-05, | |
| "loss": 2.3837, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7.4e-05, | |
| "loss": 2.3675, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7.2e-05, | |
| "loss": 2.3656, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7e-05, | |
| "loss": 2.3445, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.3281, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6.6e-05, | |
| "loss": 2.3295, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 2.2996, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6.2e-05, | |
| "loss": 2.2968, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6e-05, | |
| "loss": 2.2965, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_loss": 2.29951810836792, | |
| "eval_runtime": 487.1711, | |
| "eval_samples_per_second": 13.658, | |
| "eval_steps_per_second": 3.416, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5.8e-05, | |
| "loss": 2.288, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 2.2827, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 2.2761, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 2.278, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2664, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 2.2475, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 2.2463, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 2.2492, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4.2e-05, | |
| "loss": 2.2555, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4e-05, | |
| "loss": 2.2433, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 2.244656801223755, | |
| "eval_runtime": 487.1914, | |
| "eval_samples_per_second": 13.658, | |
| "eval_steps_per_second": 3.415, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3.8e-05, | |
| "loss": 2.2466, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 2.2253, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 2.2238, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.2352, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2222, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 2.2234, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 2.2176, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.2084, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 2.2111, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2e-05, | |
| "loss": 2.2014, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_loss": 2.219003677368164, | |
| "eval_runtime": 487.1412, | |
| "eval_samples_per_second": 13.659, | |
| "eval_steps_per_second": 3.416, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.2114, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.2049, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 2.2104, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.2046, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1999, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.2012, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6e-06, | |
| "loss": 2.1966, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.201, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.1975, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.1981, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "eval_loss": 2.210977077484131, | |
| "eval_runtime": 487.0859, | |
| "eval_samples_per_second": 13.661, | |
| "eval_steps_per_second": 3.416, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "step": 25000, | |
| "total_flos": 3.7147907956565606e+17, | |
| "train_loss": 2.32260580078125, | |
| "train_runtime": 45771.5275, | |
| "train_samples_per_second": 4.37, | |
| "train_steps_per_second": 0.546 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 2, | |
| "total_flos": 3.7147907956565606e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |