results2

This model is a fine-tuned version of MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli on the sem_eval_2024_task_2 dataset. It achieves the following results on the evaluation set:

• Loss: 1.7766
• Accuracy: 0.715
• Precision: 0.7187
• Recall: 0.7150
• F1: 0.7138

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.6998	1.0	107	0.6713	0.6	0.6214	0.6000	0.5815
0.7015	2.0	214	0.6502	0.68	0.7143	0.6800	0.6667
0.6755	3.0	321	0.6740	0.53	0.6579	0.53	0.4107
0.6605	4.0	428	0.6061	0.64	0.6502	0.64	0.6338
0.5918	5.0	535	0.5675	0.695	0.7023	0.6950	0.6922
0.5717	6.0	642	0.5945	0.685	0.6953	0.685	0.6808
0.4655	7.0	749	0.5644	0.68	0.6801	0.6800	0.6800
0.3407	8.0	856	0.7529	0.7	0.7029	0.7	0.6989
0.3539	9.0	963	0.7211	0.69	0.6901	0.69	0.6900
0.2695	10.0	1070	0.7760	0.685	0.6905	0.685	0.6827
0.1666	11.0	1177	1.1053	0.71	0.7188	0.71	0.7071
0.1648	12.0	1284	1.1662	0.72	0.7258	0.72	0.7182
0.1229	13.0	1391	1.2760	0.735	0.7438	0.735	0.7326
0.0737	14.0	1498	1.5943	0.7	0.7029	0.7	0.6989
0.1196	15.0	1605	1.5407	0.705	0.7085	0.7050	0.7037
0.0389	16.0	1712	1.6411	0.69	0.7016	0.69	0.6855
0.0199	17.0	1819	1.7139	0.685	0.6919	0.685	0.6821
0.0453	18.0	1926	1.6549	0.71	0.7121	0.71	0.7093
0.0536	19.0	2033	1.7612	0.71	0.7142	0.71	0.7086
0.0035	20.0	2140	1.7766	0.715	0.7187	0.7150	0.7138

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0