|
|
--- |
|
|
license: mit |
|
|
base_model: |
|
|
- meta-llama/Llama-3.2-3B-Instruct |
|
|
--- |
|
|
|
|
|
# Letz-MT-Llama-3.2-3B-en-as |
|
|
|
|
|
## Model description |
|
|
**Letz-MT-Llama-3.2-3B-en-as** is a fine-tuned machine translation model based on **Llama 3.2 (3B)**. The dataset is from **WMT 2025**. |
|
|
|
|
|
It is specialized for **English → Assamese** translation tasks. |
|
|
The model was trained and fine-tuned on the **WMT 2025 dataset** to provide high-quality neural machine translation. |
|
|
|
|
|
- **Base model**: Llama 3.2 (3B) |
|
|
- **Language pair**: `English (en)` → `Assamese (as)` |
|
|
- **Format**: safetensors |
|
|
- **License**: MIT |
|
|
|
|
|
## Intended use |
|
|
- Translating English text into Assamese |
|
|
- Can be integrated into translation services, research projects, and multilingual applications |
|
|
|
|
|
⚠️ **Not intended for**: |
|
|
- High-risk use cases (legal, medical, financial decision-making) |
|
|
- Professional writing that requires perfect translation accuracy |
|
|
|
