tuning-sentiment-5cdviso

This model is a fine-tuned version of 5CD-AI/Vietnamese-Sentiment-visobert on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1095
• Accuracy: 0.9778
• F1: 0.9780
• Precision: 0.9780
• Recall: 0.9781

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: 16
• eval_batch_size: 16
• 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: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.5372	0.2457	100	0.3840	0.8572	0.8542	0.8716	0.8510
0.3776	0.4914	200	0.2550	0.9058	0.9069	0.9066	0.9075
0.3333	0.7371	300	0.2245	0.9169	0.9181	0.9184	0.9196
0.3303	0.9828	400	0.1704	0.9471	0.9475	0.9483	0.9468
0.2117	1.2285	500	0.1635	0.9458	0.9464	0.9459	0.9471
0.2127	1.4742	600	0.1304	0.9538	0.9539	0.9550	0.9531
0.2021	1.7199	700	0.1385	0.9631	0.9634	0.9627	0.9647
0.2121	1.9656	800	0.1095	0.9655	0.9659	0.9651	0.9673
0.1499	2.2113	900	0.1195	0.9705	0.9708	0.9699	0.9723
0.1323	2.4570	1000	0.1101	0.976	0.9762	0.9757	0.9768
0.1566	2.7027	1100	0.1125	0.9772	0.9774	0.9776	0.9772
0.144	2.9484	1200	0.1097	0.9778	0.9780	0.9781	0.9780

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0