learning_rate: 0.0001
train_batch_size: 256
eval_batch_size: 256
seed: 13
optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 20
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
6.9636	1.0	399	5.9912
5.6174	2.0	798	5.3532
5.1611	3.0	1197	5.0579
4.8977	4.0	1596	4.8663
4.7052	5.0	1995	4.7303
4.5511	6.0	2394	4.6293
4.4221	7.0	2793	4.5558
4.3109	8.0	3192	4.4980
4.2124	9.0	3591	4.4558
4.1255	10.0	3990	4.4262
4.0484	11.0	4389	4.4007
3.9801	12.0	4788	4.3833
3.9184	13.0	5187	4.3761
3.863	14.0	5586	4.3685
3.8137	15.0	5985	4.3631
3.7698	16.0	6384	4.3630
3.7316	17.0	6783	4.3627
3.6986	18.0	7182	4.3630
3.6708	19.0	7581	4.3635
3.6501	20.0	7980	4.3651

fpadovani/wiki_o_13

wiki_o_13