| { | |
| "best_metric": 1.2517019510269165, | |
| "best_model_checkpoint": "/mnt/bn/qingyi-bn-lq/llama/saved_llamaGuanaco/checkpoint-600", | |
| "epoch": 1.8810620162633487, | |
| "global_step": 600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 1.5389, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 1.418, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 1.3622, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 1.3334, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.0003, | |
| "loss": 1.3109, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00029297423887587816, | |
| "loss": 1.3046, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.0002859484777517564, | |
| "loss": 1.2948, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.00027892271662763465, | |
| "loss": 1.2885, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.00027189695550351283, | |
| "loss": 1.2817, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.0002648711943793911, | |
| "loss": 1.2706, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 1.2975808382034302, | |
| "eval_runtime": 40.2385, | |
| "eval_samples_per_second": 49.704, | |
| "eval_steps_per_second": 0.795, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.00025784543325526926, | |
| "loss": 1.2668, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.00025081967213114756, | |
| "loss": 1.2699, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.00024379391100702575, | |
| "loss": 1.2554, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.00023676814988290396, | |
| "loss": 1.2508, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.00022974238875878218, | |
| "loss": 1.2556, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00022271662763466042, | |
| "loss": 1.2497, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.00021569086651053863, | |
| "loss": 1.2468, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00020866510538641685, | |
| "loss": 1.2428, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 0.00020163934426229506, | |
| "loss": 1.2397, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.0001946135831381733, | |
| "loss": 1.2474, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 1.2675851583480835, | |
| "eval_runtime": 40.2072, | |
| "eval_samples_per_second": 49.742, | |
| "eval_steps_per_second": 0.796, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 0.00018758782201405152, | |
| "loss": 1.2407, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 0.00018056206088992973, | |
| "loss": 1.2348, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.00017353629976580795, | |
| "loss": 1.228, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 0.00016651053864168616, | |
| "loss": 1.233, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 0.0001594847775175644, | |
| "loss": 1.2303, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 0.00015245901639344262, | |
| "loss": 1.2255, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 0.00014543325526932083, | |
| "loss": 1.2191, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 0.00013840749414519905, | |
| "loss": 1.2211, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 0.00013138173302107726, | |
| "loss": 1.2188, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 0.0001243559718969555, | |
| "loss": 1.2184, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "eval_loss": 1.2517019510269165, | |
| "eval_runtime": 40.3057, | |
| "eval_samples_per_second": 49.621, | |
| "eval_steps_per_second": 0.794, | |
| "step": 600 | |
| } | |
| ], | |
| "max_steps": 954, | |
| "num_train_epochs": 3, | |
| "total_flos": 8.806186092274385e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |