results

This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2290
• Micro f1: 0.7075
• Macro f1: 0.6540

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-05
• train_batch_size: 4
• eval_batch_size: 32
• 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: constant
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Micro f1	Macro f1
0.4348	1.0	56	0.3544	0.0	0.0
0.3297	2.0	112	0.3156	0.3816	0.1501
0.2897	3.0	168	0.2885	0.3684	0.1382
0.2501	4.0	224	0.2668	0.5465	0.2805
0.211	5.0	280	0.2335	0.56	0.2879
0.1765	6.0	336	0.2264	0.5876	0.3426
0.1497	7.0	392	0.2205	0.6102	0.3650
0.1221	8.0	448	0.1988	0.6919	0.5777
0.0964	9.0	504	0.2068	0.6701	0.5927
0.0753	10.0	560	0.1951	0.6927	0.5971
0.0614	11.0	616	0.1945	0.7136	0.6779
0.0495	12.0	672	0.2035	0.6866	0.6138
0.0387	13.0	728	0.2069	0.6977	0.6563
0.0347	14.0	784	0.2082	0.7238	0.6668
0.0318	15.0	840	0.2161	0.6957	0.6380
0.028	16.0	896	0.2075	0.7143	0.6687
0.0235	17.0	952	0.2149	0.7130	0.6650
0.0212	18.0	1008	0.2201	0.7170	0.6655
0.019	19.0	1064	0.2196	0.7256	0.6686
0.0169	20.0	1120	0.2290	0.7075	0.6540

Framework versions

• Transformers 4.55.4
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.21.4