metadata
library_name: transformers
license: cc-by-nc-4.0
license: cc-by-nc-4.0 tags: - translation - nllb
My NLLB-200 Translator
This repository contains a copy of Meta's (Facebook) NLLB-200-distilled-600M model. It has been cloned here for custom personal access and application deployment.
🌟 Model Details
- Original Developer: Meta AI (Facebook)
- Model Type: Seq2Seq Language Model (Machine Translation)
- Model Size: 600 Million parameters
- License: CC-BY-NC-4.0 (Non-commercial use only)
🌍 Language Support
This model supports direct translation between 200+ languages. For example:
- English:
eng_Latn - Telugu:
tel_Telu - Hindi:
hin_Deva - French:
fra_Latn
🚀 How to Get Started
You can use this model directly with the Hugging Face transformers library:
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
# Replace with your actual repository path
model_name = "YOUR_USERNAME/YOUR_REPO_NAME"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
# Set source language
tokenizer.src_lang = "eng_Latn"
text = "Hello, how are you today?"
inputs = tokenizer(text, return_tensors="pt")
# Target translation (Example: Telugu)
translated_tokens = model.generate(
**inputs,
forced_bos_token_id=tokenizer.convert_tokens_to_ids("tel_Telu"),
max_length=50
)
output = tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
print("Translation:", output)
## Citation
@article{nllbteam2022neglected,
title={No Language Left Behind: Scaling Human-Centered Machine Translation},
author={NLLB Team and Marta R. Costa-jussà and James Cross and Onur Çelebi and Maha Elbayad and Kenneth Heafield and others},
journal={arXiv preprint arXiv:2207.04672},
year={2022}
}