learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 25

Training Loss	Epoch	Step	Validation Loss	Accuracy	Sentence accuracy
0.9817	1.0	596	0.6246	89.2170	35.2396
0.3128	2.0	1192	0.5229	90.7575	40.1855
0.1836	3.0	1788	0.5851	91.7304	42.9675
0.1099	4.0	2384	0.6264	91.4177	41.7311
0.0744	5.0	2980	0.6672	92.2400	46.2133
0.0333	6.0	3576	0.6985	91.7535	43.8949
0.0234	7.0	4172	0.7473	92.2863	45.9042
0.0165	8.0	4768	0.7186	92.1589	45.2859
0.0167	9.0	5364	0.8132	92.4832	47.4498
0.0118	10.0	5960	0.8421	92.4253	46.3679
0.0058	11.0	6556	0.7998	92.4716	44.8223
0.0057	12.0	7152	0.7845	92.7843	47.2952
0.0039	13.0	7748	0.8570	92.6569	47.2952
0.0042	14.0	8344	0.8719	92.7496	48.2226
0.004	15.0	8940	0.8683	92.5990	47.9134
0.0038	16.0	9536	0.8549	92.5064	47.9134

ania3000/mmbert-base-kuoss-morph

trainer_output