
A fine-tuned multilingual model for Vietnamese language
π Overview
- A bilingual text generation model with strong capabilities in both Vietnamese and English languages.
- This base model can handle a wide range of text generation tasks while maintaining high quality output in both languages, making it particularly valuable for Vietnamese-English content creation and language processing applications.
π§ Method
The training process consists of three main steps:
- Continuous Pre-training (CPT) from Sailor2-8B-Chat using unsloth
- Fine-tuning with Vietnamese instruction dataset
- Applied refusal direction tuning based on "Refusal in LLMs is Mediated by a Single Direction"
π VLMU Benchmark
EVALUATION DATE | STEM π¬ | SOCIAL SCIENCE π | HUMANITIES π | OTHERS π― | AVG β |
---|---|---|---|---|---|
07/02/2025 | 50.72 | 62.81 | 60.47 | 55.4 | 56.56 |
π« Quantization
- Coming Soon!
π€ Contributors
Developed with β€οΈ by BlossomAI
Star βοΈ this repo if you find it valuable!
- Downloads last month
- 14
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
π
Ask for provider support