learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 32
total_eval_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant
num_epochs: 50

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rougel	Rougelsum
No log	0	0	1.4211	0	14.1764	0.3466	0.2974	0.3465	0.3465	0.3468
1.0668	1	12271	0.9510	0.0078	28.1301	0.5517	0.5490	0.5517	0.5517	0.5518
0.7161	2	24542	0.6923	0.0156	44.0729	0.7132	0.7139	0.7130	0.7134	0.7132
0.7534	3	36813	0.8016	0.0312	72.6121	0.6431	0.6442	0.6430	0.6433	0.6431
0.7221	4	49084	0.6587	0.0625	127.7410	0.7196	0.7186	0.7196	0.7196	0.7195
0.7053	5	61355	0.7310	0.125	239.4416	0.6808	0.6688	0.6809	0.6809	0.6808
0.7116	6	73626	0.6901	0.25	453.3816	0.7090	0.7079	0.7089	0.7087	0.7090
0.7242	7	85897	0.7143	0.5	902.3294	0.6915	0.6921	0.6913	0.6916	0.6918
0.703	8.0	98168	0.7132	1.0	1815.2414	0.7021	0.7014	0.7021	0.7021	0.7020

contemmcm/0435c898572e8517609ee593f5fff166

0435c898572e8517609ee593f5fff166