---
library_name: transformers
license: apache-2.0
base_model: Helsinki-NLP/opus-mt-en-vi
model-index:
- name: trinhhuy/finetuned-opus-mt-en-vi
  results: []
datasets:
- harouzie/vi_en-translation
language:
- en
- vi
pipeline_tag: translation
---

<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->

# trinhhuy/finetuned-opus-mt-en-vi

This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-vi](https://huggingface.co/Helsinki-NLP/opus-mt-en-vi) on harouzie/vi_en-translation dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.1596
- Validation Loss: 0.2877
- Epoch: 3
- Bleu: 77.71982942978268

## Model description

### 🧪 Quick Start

First, make sure to install the `transformers` library:
```bash
pip install transformers
```


Option 1: Use a pipeline as a high-level helper
```python
from transformers import pipeline

pipe = pipeline("translation", model="trinhhuy/finetuned-opus-mt-en-vi")

result = pipe("I'm confident that my friend will pass the exam because he has been studying hard and staying focused for weeks.")
print(result)
```
```text
[{'translation_text': 'Tôi tự tin rằng bạn tôi sẽ vượt qua kỳ thi vì anh ấy đã học tập chăm chỉ và tập trung nhiều tuần.'}]
```
Option 2: Load model directly
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("trinhhuy/finetuned-opus-mt-en-vi")
model = AutoModelForSeq2SeqLM.from_pretrained("trinhhuy/finetuned-opus-mt-en-vi")

input_tokenized = tokenizer("I'm confident that my friend will pass the exam because he has been studying hard and staying focused for weeks.", return_tensors="pt")
output_tokenized = model.generate(**input_tokenized)

translated_text = tokenizer.decode(output_tokenized[0], skip_special_tokens=True)
print(translated_text)
```
```text
Tôi tự tin rằng bạn tôi sẽ vượt qua kỳ thi vì anh ấy đã học tập chăm chỉ và tập trung nhiều tuần.
```


### Training hyperparameters

The following hyperparameters were used during training:
- initial_learning_rate = 5e-05
- weight_decay_rate = 0.01
- training_precision: mixed_float16


### Training results

| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 0.4246     | 0.3235          | 0     |
| 0.2724     | 0.2977          | 1     |
| 0.2019     | 0.2887          | 2     |
| 0.1596     | 0.2877          | 3     |


### Framework versions

- Transformers 4.53.3
- TensorFlow 2.18.0
- Datasets 4.0.0
- Tokenizers 0.21.2