| { | |
| "best_metric": 0.2142563909292221, | |
| "best_model_checkpoint": "/content/deberta-v3-small/checkpoint-6547", | |
| "epoch": 5.0, | |
| "global_step": 32735, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.908354971742783e-05, | |
| "loss": 0.3956, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 2.816709943485566e-05, | |
| "loss": 0.3285, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.7250649152283487e-05, | |
| "loss": 0.313, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.633419886971132e-05, | |
| "loss": 0.2956, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 2.541774858713915e-05, | |
| "loss": 0.2931, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 2.450129830456698e-05, | |
| "loss": 0.2823, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9150649826102873, | |
| "eval_loss": 0.2142563909292221, | |
| "eval_runtime": 10.8101, | |
| "eval_samples_per_second": 505.359, | |
| "eval_steps_per_second": 63.181, | |
| "step": 6547 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2.3584848021994806e-05, | |
| "loss": 0.2467, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2.2668397739422636e-05, | |
| "loss": 0.1968, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2.1751947456850465e-05, | |
| "loss": 0.1994, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2.08354971742783e-05, | |
| "loss": 0.203, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 1.9919046891706125e-05, | |
| "loss": 0.2003, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.9002596609133955e-05, | |
| "loss": 0.1904, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 1.8086146326561784e-05, | |
| "loss": 0.1996, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9103056928427604, | |
| "eval_loss": 0.27601319551467896, | |
| "eval_runtime": 10.9101, | |
| "eval_samples_per_second": 500.73, | |
| "eval_steps_per_second": 62.603, | |
| "step": 13094 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 1.7169696043989614e-05, | |
| "loss": 0.1357, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 1.625324576141744e-05, | |
| "loss": 0.1331, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.5336795478845274e-05, | |
| "loss": 0.1317, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.4420345196273102e-05, | |
| "loss": 0.1337, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 1.3503894913700933e-05, | |
| "loss": 0.1313, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.2587444631128761e-05, | |
| "loss": 0.1327, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9168954786747209, | |
| "eval_loss": 0.3293353319168091, | |
| "eval_runtime": 10.8188, | |
| "eval_samples_per_second": 504.952, | |
| "eval_steps_per_second": 63.131, | |
| "step": 19641 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 1.1670994348556591e-05, | |
| "loss": 0.1108, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 1.0754544065984421e-05, | |
| "loss": 0.0817, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 9.83809378341225e-06, | |
| "loss": 0.0851, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 8.921643500840079e-06, | |
| "loss": 0.0787, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 8.00519321826791e-06, | |
| "loss": 0.0897, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 7.088742935695739e-06, | |
| "loss": 0.0917, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 6.172292653123568e-06, | |
| "loss": 0.0811, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9192751235584844, | |
| "eval_loss": 0.4277651309967041, | |
| "eval_runtime": 10.6276, | |
| "eval_samples_per_second": 514.037, | |
| "eval_steps_per_second": 64.266, | |
| "step": 26188 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 5.255842370551398e-06, | |
| "loss": 0.0578, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "learning_rate": 4.339392087979228e-06, | |
| "loss": 0.0518, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 3.4229418054070567e-06, | |
| "loss": 0.052, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 2.5064915228348865e-06, | |
| "loss": 0.052, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 1.5900412402627159e-06, | |
| "loss": 0.0448, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 6.735909576905453e-07, | |
| "loss": 0.05, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9176276771004942, | |
| "eval_loss": 0.5109593868255615, | |
| "eval_runtime": 10.8996, | |
| "eval_samples_per_second": 501.212, | |
| "eval_steps_per_second": 62.663, | |
| "step": 32735 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 32735, | |
| "total_flos": 1.734440872113408e+16, | |
| "train_loss": 0.15606709380213207, | |
| "train_runtime": 1886.2061, | |
| "train_samples_per_second": 277.655, | |
| "train_steps_per_second": 17.355 | |
| } | |
| ], | |
| "max_steps": 32735, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.734440872113408e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |