| { | |
| "best_metric": 0.29386183619499207, | |
| "best_model_checkpoint": "./ryan_model314/checkpoint-200", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1858803033828735, | |
| "learning_rate": 0.00019206349206349208, | |
| "loss": 0.5736, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.41181480884552, | |
| "learning_rate": 0.00018412698412698412, | |
| "loss": 0.4142, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6347964406013489, | |
| "learning_rate": 0.0001761904761904762, | |
| "loss": 0.3916, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.7893273234367371, | |
| "learning_rate": 0.00016825396825396826, | |
| "loss": 0.3628, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.9887136816978455, | |
| "learning_rate": 0.00016031746031746033, | |
| "loss": 0.3852, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.1596781015396118, | |
| "learning_rate": 0.00015238095238095237, | |
| "loss": 0.3962, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.1897984743118286, | |
| "learning_rate": 0.00014444444444444444, | |
| "loss": 0.2923, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.5242781639099121, | |
| "learning_rate": 0.0001365079365079365, | |
| "loss": 0.2335, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.0704305171966553, | |
| "learning_rate": 0.00012857142857142858, | |
| "loss": 0.268, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.852606475353241, | |
| "learning_rate": 0.00012063492063492063, | |
| "loss": 0.1969, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_loss": 0.2954840064048767, | |
| "eval_na_accuracy": 0.945, | |
| "eval_ordinal_accuracy": 0.4785276073619632, | |
| "eval_runtime": 26.5841, | |
| "eval_samples_per_second": 7.523, | |
| "eval_steps_per_second": 0.94, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.6733121871948242, | |
| "learning_rate": 0.0001126984126984127, | |
| "loss": 0.2328, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.9004744291305542, | |
| "learning_rate": 0.00010476190476190477, | |
| "loss": 0.2248, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 3.1687183380126953, | |
| "learning_rate": 9.682539682539682e-05, | |
| "loss": 0.1717, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.6659616827964783, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 0.145, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.6846858859062195, | |
| "learning_rate": 8.095238095238096e-05, | |
| "loss": 0.1157, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.6155730485916138, | |
| "learning_rate": 7.301587301587302e-05, | |
| "loss": 0.1424, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.6559838056564331, | |
| "learning_rate": 6.507936507936509e-05, | |
| "loss": 0.1041, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.4849882423877716, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.1339, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.6311644911766052, | |
| "learning_rate": 4.9206349206349204e-05, | |
| "loss": 0.103, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.8323171138763428, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 0.0746, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_loss": 0.29386183619499207, | |
| "eval_na_accuracy": 0.945, | |
| "eval_ordinal_accuracy": 0.5705521472392638, | |
| "eval_runtime": 8.3287, | |
| "eval_samples_per_second": 24.013, | |
| "eval_steps_per_second": 3.002, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.8798254132270813, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.086, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.37054240703582764, | |
| "learning_rate": 2.5396825396825397e-05, | |
| "loss": 0.0933, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.46293869614601135, | |
| "learning_rate": 1.746031746031746e-05, | |
| "loss": 0.066, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.3859086334705353, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.0649, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.300207257270813, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.0623, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 252, | |
| "total_flos": 3.0997907103744e+17, | |
| "train_loss": 0.212149089468377, | |
| "train_runtime": 378.6413, | |
| "train_samples_per_second": 10.564, | |
| "train_steps_per_second": 0.666 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "total_flos": 3.0997907103744e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |