| { | |
| "best_metric": 0.6038, | |
| "best_model_checkpoint": "save_model/deberta-v3-base-amazon-reviews-multi/checkpoint-18750", | |
| "epoch": 5.0, | |
| "global_step": 31250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.92e-05, | |
| "loss": 1.0637, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.8400000000000004e-05, | |
| "loss": 0.9832, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.76e-05, | |
| "loss": 0.9515, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.6800000000000006e-05, | |
| "loss": 0.9413, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.9464, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 4.52e-05, | |
| "loss": 0.9408, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 4.44e-05, | |
| "loss": 0.9262, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.36e-05, | |
| "loss": 0.9273, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 4.2800000000000004e-05, | |
| "loss": 0.9227, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.9129, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 4.12e-05, | |
| "loss": 0.9246, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 4.0400000000000006e-05, | |
| "loss": 0.9134, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6012, | |
| "eval_loss": 0.913094162940979, | |
| "eval_runtime": 11.6784, | |
| "eval_samples_per_second": 428.141, | |
| "eval_steps_per_second": 13.444, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 3.960000000000001e-05, | |
| "loss": 0.8756, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 3.88e-05, | |
| "loss": 0.842, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3.8e-05, | |
| "loss": 0.8358, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 3.72e-05, | |
| "loss": 0.8372, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 3.6400000000000004e-05, | |
| "loss": 0.8443, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 3.56e-05, | |
| "loss": 0.8442, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 3.48e-05, | |
| "loss": 0.8437, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.8403, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 3.32e-05, | |
| "loss": 0.8521, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 3.24e-05, | |
| "loss": 0.8429, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 3.16e-05, | |
| "loss": 0.8416, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 3.08e-05, | |
| "loss": 0.8467, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8411, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6008, | |
| "eval_loss": 0.9258156418800354, | |
| "eval_runtime": 11.6504, | |
| "eval_samples_per_second": 429.17, | |
| "eval_steps_per_second": 13.476, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 2.9199999999999998e-05, | |
| "loss": 0.7412, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 2.84e-05, | |
| "loss": 0.745, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 2.7600000000000003e-05, | |
| "loss": 0.7507, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2.6800000000000004e-05, | |
| "loss": 0.7455, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.7532, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 2.5200000000000003e-05, | |
| "loss": 0.7484, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 2.44e-05, | |
| "loss": 0.7434, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 2.36e-05, | |
| "loss": 0.7497, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 2.2800000000000002e-05, | |
| "loss": 0.7409, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.7404, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2.12e-05, | |
| "loss": 0.7437, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2.04e-05, | |
| "loss": 0.7503, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6038, | |
| "eval_loss": 0.9382744431495667, | |
| "eval_runtime": 11.6538, | |
| "eval_samples_per_second": 429.045, | |
| "eval_steps_per_second": 13.472, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.6785, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.88e-05, | |
| "loss": 0.6265, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.628, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 1.7199999999999998e-05, | |
| "loss": 0.6207, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 1.6400000000000002e-05, | |
| "loss": 0.6261, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 1.56e-05, | |
| "loss": 0.6362, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 1.48e-05, | |
| "loss": 0.6292, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 0.6117, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 1.32e-05, | |
| "loss": 0.6334, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 1.24e-05, | |
| "loss": 0.6245, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 1.16e-05, | |
| "loss": 0.626, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 1.08e-05, | |
| "loss": 0.6302, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.621, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6026, | |
| "eval_loss": 1.052680492401123, | |
| "eval_runtime": 11.6694, | |
| "eval_samples_per_second": 428.472, | |
| "eval_steps_per_second": 13.454, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 9.2e-06, | |
| "loss": 0.5245, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.51, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "learning_rate": 7.6e-06, | |
| "loss": 0.5238, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.5292, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 6e-06, | |
| "loss": 0.5049, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.5124, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.5096, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 3.6e-06, | |
| "loss": 0.5142, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.5043, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.505, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.4981, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.5088, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5994, | |
| "eval_loss": 1.1826133728027344, | |
| "eval_runtime": 11.662, | |
| "eval_samples_per_second": 428.744, | |
| "eval_steps_per_second": 13.463, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 31250, | |
| "total_flos": 6.5780715264e+16, | |
| "train_loss": 0.7340158759765625, | |
| "train_runtime": 7300.3061, | |
| "train_samples_per_second": 136.981, | |
| "train_steps_per_second": 4.281 | |
| } | |
| ], | |
| "max_steps": 31250, | |
| "num_train_epochs": 5, | |
| "total_flos": 6.5780715264e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |