PhoLexiContent-10304

This model is a fine-tuned version of vinai/phobert-base-v2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2381
• Accuracy: 0.9476
• F1: 0.9193

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.7752	100	0.1326	0.9558	0.9302
0.2063	1.5504	200	0.1234	0.9583	0.9352
0.1189	2.3256	300	0.1489	0.9461	0.9150
0.0897	3.1008	400	0.1588	0.9486	0.9170
0.0897	3.8760	500	0.1327	0.9529	0.9262
0.0709	4.6512	600	0.1630	0.9544	0.9288
0.045	5.4264	700	0.1765	0.9486	0.9203
0.0374	6.2016	800	0.1713	0.9539	0.9280
0.0287	6.9767	900	0.1845	0.9486	0.9193
0.0287	7.7519	1000	0.1864	0.9427	0.9113
0.0186	8.5271	1100	0.2000	0.9476	0.9194
0.0153	9.3023	1200	0.2062	0.9481	0.9191
0.0107	10.0775	1300	0.2143	0.9471	0.9159
0.0107	10.8527	1400	0.2180	0.9491	0.9196
0.0076	11.6279	1500	0.2222	0.9486	0.9192
0.0068	12.4031	1600	0.2185	0.9452	0.9150
0.0047	13.1783	1700	0.2217	0.9491	0.9199
0.0034	13.9535	1800	0.2322	0.9461	0.9163
0.0034	14.7287	1900	0.2293	0.9481	0.9192
0.0036	15.5039	2000	0.2563	0.9466	0.9155
0.003	16.2791	2100	0.2254	0.9495	0.9221
0.0025	17.0543	2200	0.2378	0.9476	0.9187
0.0025	17.8295	2300	0.2403	0.9471	0.9180
0.0023	18.6047	2400	0.2450	0.9491	0.9203
0.0022	19.3798	2500	0.2381	0.9476	0.9193

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0