Vietnamese-social-media-visobert

This model is a fine-tuned version of 5CD-AI/Vietnamese-Sentiment-visobert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8985
• Accuracy: 0.8744
• F1: 0.8264
• Precision: 0.8363
• Recall: 0.8175

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 260
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
No log	0.3817	50	0.6882	0.8117	0.7660	0.7615	0.7709
No log	0.7634	100	0.5681	0.8386	0.7849	0.8154	0.7628
0.9396	1.1450	150	0.5227	0.8520	0.8116	0.8157	0.8083
0.9396	1.5267	200	0.5717	0.8475	0.8039	0.8007	0.8135
0.9396	1.9084	250	0.5705	0.8251	0.7842	0.7896	0.7825
0.4081	2.2901	300	0.6703	0.8565	0.8217	0.8251	0.8208
0.4081	2.6718	350	0.7260	0.8610	0.8162	0.8351	0.8016
0.2027	3.0534	400	0.6946	0.8610	0.8148	0.8152	0.8155
0.2027	3.4351	450	0.8426	0.8520	0.7935	0.8183	0.7747
0.2027	3.8168	500	0.8195	0.8520	0.7828	0.8255	0.7610
0.1098	4.1985	550	0.8272	0.8520	0.8072	0.8171	0.7984
0.1098	4.5802	600	0.9978	0.8520	0.7925	0.8361	0.7648
0.0277	4.9618	650	0.8102	0.8430	0.7882	0.8029	0.7796
0.0277	5.3435	700	0.8563	0.8520	0.7926	0.8107	0.7787
0.0277	5.7252	750	0.9299	0.8565	0.8052	0.8291	0.7871
0.0089	6.1069	800	0.8562	0.8520	0.8002	0.8141	0.7886
0.0089	6.4885	850	0.8737	0.8610	0.8074	0.8213	0.7958
0.0089	6.8702	900	0.9388	0.8610	0.8079	0.8431	0.7839
0.0094	7.2519	950	0.9036	0.8744	0.8280	0.8540	0.8097
0.0094	7.6336	1000	0.9003	0.8744	0.8264	0.8363	0.8175
0.0052	8.0153	1050	0.8814	0.8744	0.8264	0.8363	0.8175
0.0052	8.3969	1100	0.8918	0.8744	0.8264	0.8363	0.8175
0.0052	8.7786	1150	0.8964	0.8744	0.8264	0.8363	0.8175
0.0055	9.1603	1200	0.8936	0.8744	0.8264	0.8363	0.8175
0.0055	9.5420	1250	0.8959	0.8744	0.8264	0.8363	0.8175
0.0079	9.9237	1300	0.8985	0.8744	0.8264	0.8363	0.8175

Framework versions

• Transformers 4.43.4
• Pytorch 2.4.0+cu121
• Tokenizers 0.19.1