| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.636363636363637, | |
| "eval_steps": 500, | |
| "global_step": 75, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 220.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 24.9691, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 35.25, | |
| "learning_rate": 0.000125, | |
| "loss": 21.9058, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_loss": 7.6562018394470215, | |
| "eval_runtime": 0.5686, | |
| "eval_samples_per_second": 3.517, | |
| "eval_steps_per_second": 1.759, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 9.0, | |
| "learning_rate": 0.00019956059820218982, | |
| "loss": 13.5645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 6.635939121246338, | |
| "eval_runtime": 0.5597, | |
| "eval_samples_per_second": 3.573, | |
| "eval_steps_per_second": 1.787, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 0.00019466156752904343, | |
| "loss": 10.2613, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "eval_loss": 6.0754241943359375, | |
| "eval_runtime": 0.5741, | |
| "eval_samples_per_second": 3.484, | |
| "eval_steps_per_second": 1.742, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 0.00018458320592590975, | |
| "loss": 9.903, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.1116435527801514, | |
| "eval_runtime": 0.5647, | |
| "eval_samples_per_second": 3.542, | |
| "eval_steps_per_second": 1.771, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 32.75, | |
| "learning_rate": 0.00016987694277788417, | |
| "loss": 4.594, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "eval_loss": 1.6371122598648071, | |
| "eval_runtime": 0.5752, | |
| "eval_samples_per_second": 3.477, | |
| "eval_steps_per_second": 1.739, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 3.125, | |
| "learning_rate": 0.0001513474193514842, | |
| "loss": 1.6122, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.4159561395645142, | |
| "eval_runtime": 0.5662, | |
| "eval_samples_per_second": 3.532, | |
| "eval_steps_per_second": 1.766, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.0001300084635000341, | |
| "loss": 1.3971, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.91, | |
| "eval_loss": 1.3411411046981812, | |
| "eval_runtime": 0.5782, | |
| "eval_samples_per_second": 3.459, | |
| "eval_steps_per_second": 1.73, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.27, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 0.0001070276188945293, | |
| "loss": 1.2757, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.307417631149292, | |
| "eval_runtime": 0.5668, | |
| "eval_samples_per_second": 3.529, | |
| "eval_steps_per_second": 1.764, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.366226381814697e-05, | |
| "loss": 1.1233, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "eval_loss": 1.275590419769287, | |
| "eval_runtime": 0.5782, | |
| "eval_samples_per_second": 3.459, | |
| "eval_steps_per_second": 1.729, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 6.119081473277501e-05, | |
| "loss": 1.0599, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 4.084277875864776e-05, | |
| "loss": 0.9741, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.273629903793335, | |
| "eval_runtime": 0.5636, | |
| "eval_samples_per_second": 3.549, | |
| "eval_steps_per_second": 1.774, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.91, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 2.3731482188961818e-05, | |
| "loss": 0.9266, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.91, | |
| "eval_loss": 1.279077410697937, | |
| "eval_runtime": 0.5768, | |
| "eval_samples_per_second": 3.467, | |
| "eval_steps_per_second": 1.734, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.82, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 1.0793155744261351e-05, | |
| "loss": 0.8584, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.2753326892852783, | |
| "eval_runtime": 0.5671, | |
| "eval_samples_per_second": 3.527, | |
| "eval_steps_per_second": 1.763, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.73, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 2.735709467518699e-06, | |
| "loss": 0.8714, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.91, | |
| "eval_loss": 1.2842026948928833, | |
| "eval_runtime": 0.7036, | |
| "eval_samples_per_second": 2.842, | |
| "eval_steps_per_second": 1.421, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0, | |
| "loss": 0.8421, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "eval_loss": 1.2808171510696411, | |
| "eval_runtime": 0.5646, | |
| "eval_samples_per_second": 3.543, | |
| "eval_steps_per_second": 1.771, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "step": 75, | |
| "total_flos": 1.1503415671442637e+17, | |
| "train_loss": 4.785474745432536, | |
| "train_runtime": 283.6494, | |
| "train_samples_per_second": 4.654, | |
| "train_steps_per_second": 0.264 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 75, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 100, | |
| "total_flos": 1.1503415671442637e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |