PhoBert_content_256

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: 0.3128
• Accuracy: 0.9482
• F1: 0.9389

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: 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.5435	200	0.1616	0.9431	0.9323
0.2072	1.0870	400	0.1527	0.9470	0.9385
0.2072	1.6304	600	0.1545	0.9487	0.9406
0.1201	2.1739	800	0.1757	0.9504	0.9406
0.1201	2.7174	1000	0.1774	0.9477	0.9368
0.0901	3.2609	1200	0.1740	0.9514	0.9422
0.0901	3.8043	1400	0.1564	0.9511	0.9420
0.0655	4.3478	1600	0.2446	0.9487	0.9389
0.0655	4.8913	1800	0.1856	0.9504	0.9414
0.0485	5.4348	2000	0.2292	0.9492	0.9398
0.0485	5.9783	2200	0.2414	0.9506	0.9425
0.0369	6.5217	2400	0.2569	0.9502	0.9409
0.0253	7.0652	2600	0.2593	0.9483	0.9390
0.0253	7.6087	2800	0.2806	0.9480	0.9385
0.0198	8.1522	3000	0.2920	0.9492	0.9399
0.0198	8.6957	3200	0.3134	0.9466	0.9367
0.014	9.2391	3400	0.3115	0.9480	0.9387
0.014	9.7826	3600	0.3128	0.9482	0.9389

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.0