self_iterative_v1_stance_iteration_3

This model is a fine-tuned version of bert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.9908
• Accuracy Stance: 0.7639
• F1 Macro Stance: 0.6364
• F1 Weighted Stance: 0.7488
• F1 Macro Total: 0.6364
• F1 Weighted Total: 0.7488

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: 6e-06
• train_batch_size: 4
• eval_batch_size: 4
• seed: 1337
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy Stance	F1 Macro Stance	F1 Weighted Stance	F1 Macro Total	F1 Weighted Total
0.549	1.0	6439	1.8386	0.7738	0.5443	0.7275	0.5443	0.7275
0.4804	2.0	12878	1.7345	0.7574	0.5326	0.7119	0.5326	0.7119
0.3624	3.0	19317	2.3094	0.7344	0.5143	0.6885	0.5143	0.6885
0.2227	4.0	25756	2.3845	0.7443	0.5216	0.6979	0.5216	0.6979
0.1634	5.0	32195	2.4018	0.7213	0.6021	0.7164	0.6021	0.7164
0.1258	6.0	38634	2.9908	0.7639	0.6364	0.7488	0.6364	0.7488
0.0899	7.0	45073	3.5773	0.7148	0.5955	0.7049	0.5955	0.7049
0.058	8.0	51512	3.9667	0.7475	0.6259	0.7324	0.6259	0.7324
0.0375	9.0	57951	3.9157	0.7344	0.6218	0.7295	0.6218	0.7295
0.0771	10.0	64390	4.0061	0.7443	0.6271	0.7350	0.6271	0.7350
0.0457	11.0	70829	4.4023	0.7508	0.6076	0.7337	0.6076	0.7337

Framework versions

• Transformers 4.50.2
• Pytorch 2.6.0+cu124
• Datasets 3.0.1
• Tokenizers 0.21.1