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
0.7863	1.0	546	0.3676	90.7489	37.6147
0.2408	2.0	1092	0.2598	93.2662	47.3394
0.1363	3.0	1638	0.2659	93.6060	47.7064
0.0828	4.0	2184	0.2874	94.2857	51.0092
0.0525	5.0	2730	0.3354	94.1221	50.0917
0.0321	6.0	3276	0.3410	94.4242	53.5780
0.0206	7.0	3822	0.3864	94.4619	51.7431
0.0161	8.0	4368	0.3735	94.8143	53.5780
0.0101	9.0	4914	0.3780	94.9654	55.4128
0.0089	10.0	5460	0.3959	94.6130	53.3945
0.0055	11.0	6006	0.4154	94.8018	53.3945
0.0046	12.0	6552	0.4089	94.8899	54.6789
0.0041	13.0	7098	0.4007	95.1542	55.2294
0.0028	14.0	7644	0.4057	95.0535	55.5963
0.0019	15.0	8190	0.4216	95.0157	54.3119
0.0021	16.0	8736	0.4158	95.0913	56.3303
0.0027	17.0	9282	0.4158	94.7892	54.4954

ania3000/mmbert-base-oss-morph

trainer_output