cs221-mDeBERTa-v3-base-mnli-xnli-finetuned-20-epochs

This model is a fine-tuned version of MoritzLaurer/mDeBERTa-v3-base-mnli-xnli on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5558
• F1: 0.7088
• Roc Auc: 0.7803
• Accuracy: 0.3953

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.6081	1.0	64	0.5854	0.4511	0.6250	0.1759
0.5318	2.0	128	0.5056	0.5413	0.6739	0.2905
0.4634	3.0	192	0.4846	0.6178	0.7158	0.3103
0.417	4.0	256	0.4696	0.6609	0.7452	0.3340
0.3762	5.0	320	0.4645	0.6707	0.7522	0.3577
0.3187	6.0	384	0.4622	0.6802	0.7591	0.3636
0.2801	7.0	448	0.4691	0.6953	0.7702	0.3557
0.2593	8.0	512	0.4847	0.6833	0.7614	0.3498
0.2278	9.0	576	0.4878	0.6941	0.7691	0.3775
0.1917	10.0	640	0.5014	0.7001	0.7735	0.3834
0.1761	11.0	704	0.5298	0.7020	0.7747	0.3893
0.1494	12.0	768	0.5417	0.7009	0.7741	0.3972
0.1422	13.0	832	0.5437	0.7074	0.7791	0.3953
0.1132	14.0	896	0.5602	0.6965	0.7709	0.3794
0.108	15.0	960	0.5558	0.7088	0.7803	0.3953
0.0945	16.0	1024	0.5700	0.7026	0.7753	0.3913
0.1004	17.0	1088	0.5748	0.7055	0.7776	0.3953
0.0857	18.0	1152	0.5733	0.7056	0.7776	0.3913
0.0893	19.0	1216	0.5731	0.7032	0.7759	0.3913
0.0881	20.0	1280	0.5735	0.7040	0.7765	0.3913

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0