| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2597631645250693, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.000975, | |
| "loss": 1.0322, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.8672707676887512, | |
| "eval_runtime": 15.3736, | |
| "eval_samples_per_second": 65.047, | |
| "eval_steps_per_second": 32.523, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.00095, | |
| "loss": 0.9884, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.8287830948829651, | |
| "eval_runtime": 15.4828, | |
| "eval_samples_per_second": 64.588, | |
| "eval_steps_per_second": 32.294, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.000925, | |
| "loss": 0.9418, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 0.8468040823936462, | |
| "eval_runtime": 15.4288, | |
| "eval_samples_per_second": 64.814, | |
| "eval_steps_per_second": 32.407, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.9232, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.8540903329849243, | |
| "eval_runtime": 15.6504, | |
| "eval_samples_per_second": 63.896, | |
| "eval_steps_per_second": 31.948, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.000875, | |
| "loss": 0.9107, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 0.809073269367218, | |
| "eval_runtime": 15.3867, | |
| "eval_samples_per_second": 64.991, | |
| "eval_steps_per_second": 32.496, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00085, | |
| "loss": 0.9233, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.8151862621307373, | |
| "eval_runtime": 15.0851, | |
| "eval_samples_per_second": 66.291, | |
| "eval_steps_per_second": 33.145, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.000825, | |
| "loss": 0.9038, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 0.8232718706130981, | |
| "eval_runtime": 15.5073, | |
| "eval_samples_per_second": 64.486, | |
| "eval_steps_per_second": 32.243, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.0008, | |
| "loss": 0.8747, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.7864852547645569, | |
| "eval_runtime": 15.3637, | |
| "eval_samples_per_second": 65.089, | |
| "eval_steps_per_second": 32.544, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0007750000000000001, | |
| "loss": 0.8981, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 0.770944356918335, | |
| "eval_runtime": 15.3602, | |
| "eval_samples_per_second": 65.103, | |
| "eval_steps_per_second": 32.552, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00075, | |
| "loss": 0.8538, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 0.7905715107917786, | |
| "eval_runtime": 16.0451, | |
| "eval_samples_per_second": 62.324, | |
| "eval_steps_per_second": 31.162, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.000725, | |
| "loss": 0.8295, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 0.7717331051826477, | |
| "eval_runtime": 16.2335, | |
| "eval_samples_per_second": 61.601, | |
| "eval_steps_per_second": 30.8, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.0007, | |
| "loss": 0.8346, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.7587910890579224, | |
| "eval_runtime": 15.3749, | |
| "eval_samples_per_second": 65.041, | |
| "eval_steps_per_second": 32.521, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.000675, | |
| "loss": 0.8366, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 0.7654258012771606, | |
| "eval_runtime": 15.1755, | |
| "eval_samples_per_second": 65.895, | |
| "eval_steps_per_second": 32.948, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0006500000000000001, | |
| "loss": 0.8017, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 0.759145975112915, | |
| "eval_runtime": 15.3597, | |
| "eval_samples_per_second": 65.106, | |
| "eval_steps_per_second": 32.553, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.000625, | |
| "loss": 0.7788, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 0.7703807950019836, | |
| "eval_runtime": 15.9867, | |
| "eval_samples_per_second": 62.552, | |
| "eval_steps_per_second": 31.276, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.0006, | |
| "loss": 0.7764, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 0.7488923668861389, | |
| "eval_runtime": 16.1496, | |
| "eval_samples_per_second": 61.921, | |
| "eval_steps_per_second": 30.961, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.000575, | |
| "loss": 0.5766, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 0.7688985466957092, | |
| "eval_runtime": 15.5729, | |
| "eval_samples_per_second": 64.214, | |
| "eval_steps_per_second": 32.107, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00055, | |
| "loss": 0.6024, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_loss": 0.7472162842750549, | |
| "eval_runtime": 15.4638, | |
| "eval_samples_per_second": 64.667, | |
| "eval_steps_per_second": 32.334, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 0.0005250000000000001, | |
| "loss": 0.5822, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.7330933809280396, | |
| "eval_runtime": 15.5588, | |
| "eval_samples_per_second": 64.272, | |
| "eval_steps_per_second": 32.136, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5812, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 0.7249069809913635, | |
| "eval_runtime": 15.1664, | |
| "eval_samples_per_second": 65.935, | |
| "eval_steps_per_second": 32.968, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 20000, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "total_flos": 5.445321533503488e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |