xlm-roberta-meta4types-ft-2.0

This model is a fine-tuned version of cardiffnlp/twitter-xlm-roberta-base-sentiment on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0008
• Roc Auc: 0.6612
• Hamming Loss: 0.2239
• F1 Score: 0.5943
• Accuracy: 0.5392
• Precision: 0.5798
• Recall: 0.6121

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: 4
• eval_batch_size: 4
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Roc Auc	Hamming Loss	F1 Score	Accuracy	Precision	Recall
No log	1.0	204	0.5010	0.4988	0.2042	0.2930	0.6127	0.5948	0.3333
No log	2.0	408	0.5433	0.5027	0.2010	0.3038	0.6176	0.9281	0.3388
0.4958	3.0	612	0.5013	0.5043	0.2010	0.3139	0.6127	0.8170	0.3443
0.4958	4.0	816	0.6563	0.6108	0.2190	0.5211	0.5686	0.6488	0.4799
0.3484	5.0	1020	0.6404	0.6444	0.1912	0.5645	0.5980	0.6014	0.5386
0.3484	6.0	1224	0.9555	0.6520	0.2614	0.5559	0.5196	0.5889	0.5417
0.3484	7.0	1428	0.7919	0.6202	0.2222	0.5417	0.5392	0.5743	0.5297
0.1644	8.0	1632	0.8959	0.6389	0.2157	0.5551	0.5539	0.5823	0.5515
0.1644	9.0	1836	1.0008	0.6612	0.2239	0.5943	0.5392	0.5798	0.6121
0.0611	10.0	2040	0.9594	0.6452	0.2141	0.5822	0.5294	0.5757	0.5893

Framework versions

• Transformers 4.43.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1