ViNormT5

This model is a fine-tuned version of VietAI/vit5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2324
• Bleu Score: 79.5557
• Precision: 100.0
• Recall: 93.3333
• Gen Len: 12.7969
• Err: 56.9892

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu Score	Precision	Recall	Gen Len	Err
0.4672	1.0	419	0.2390	76.7871	100.0	100.0	12.8292	49.7013
0.1742	2.0	838	0.2173	77.9698	100.0	93.3333	12.8076	54.0024
0.0814	3.0	1257	0.2010	79.204	100.0	93.3333	12.7754	56.6308
0.0382	4.0	1676	0.2324	79.5557	100.0	93.3333	12.7969	56.9892

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0