Downloads · 30 days
7
3% of all-time downloads
Ijaj7/Xetu-english-to-assamese-translator
Xetu-english-to-assamese-translator is a machine learning model from Ijaj7. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
This repository contains a fine-tuned OPUS-MT multilingual model (opus-mt-en-mul) designed for English → Assamese translation. The model was trained as part of the Xetu Chat App, enabling real-time bilingual messaging.
Downloads · 30 days
7
3% of all-time downloads
All-time downloads
234
Public
Parameters
77M
310 MB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors308 MB · 99%
From the Hugging Face model README
This repository contains a fine-tuned OPUS-MT multilingual model (opus-mt-en-mul) designed for English → Assamese translation.
The model was trained as part of the Xetu Chat App, enabling real-time bilingual messaging.
This fine-tuned model improves translation quality for conversational English and Assamese.
| Split | Size |
|---|---|
| Train | 90,000 pairs |
| Validation | 1,000 pairs |
| Test | 4,000 pairs |
All data files are JSON with keys:
{
"source_text": "English sentence",
"target_text": "Assamese sentence"
}
A target language prefix was used:
>>asm<< <English text>
This ensures multilingual OPUS-MT generates Assamese output.
##⚙️ Preprocessing & Tokenization
AutoTokenizer)-100 so they are ignored in lossExample preprocessing:
">>asm<< " + source_sentence
Training loss decreased smoothly:
Early: ~2.3
Final: ~0.60
Stable between 1.7–2.5, no explosions.
Overall, training was very stable.
| Epoch | Eval Loss | BLEU |
|---|---|---|
| 1 | 0.8784 | 18.79 |
| 2 | 0.7369 | 22.88 |
| 3 | 0.6786 | 24.20 |
| 4 | 0.6483 | 25.37 |
| 5 | 0.6292 | 25.96 |
| 6 | 0.6184 | 26.68 |
| 7 | 0.6148 | 26.78 ✔️ (Best) |
The best model was correctly selected using:
load_best_model_at_end = True
metric_for_best_model = "eval_bleu"
from transformers import MarianMTModel, MarianTokenizer
model_name = "Ijaj7/Xetu-english-to-assamese-translator"
tokenizer = MarianTokenizer.from_pretrained(model_name)
model = MarianMTModel.from_pretrained(model_name)
text = "They are very happy."
inputs = tokenizer(">>asm<< " + text, return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
curl -X POST \
-H "Authorization: Bearer YOUR_HF_TOKEN" \
-H "Content-Type: application/json" \
-d '{"inputs": "How are you?"}' \
https://api-inference.huggingface.co/models/Ijaj7/Xetu-english-to-assamese-translator
This model is optimized for:
A Firebase Cloud Function or direct HuggingFace API can be used.
| English | Assamese |
|---|---|
| What is your name? | আপোনাৰ নাম কি? |
| I will call you later. | মই পিছত তোমাক ফোন কৰিম। |
| What are you doing right now? | তুমি এতিয়া কি কৰি আছা? |
| I need some time to think before making a decision. | সিদ্ধান্ত লোৱাৰ আগতে মোৰ চিন্তা কৰিবলৈ কিছু সময়ৰ প্ৰয়োজন। |
| Our goal is to make Assamese language technology more accessible to everyone. | আমাৰ লক্ষ্য হৈছে অসমীয়া ভাষাৰ প্ৰযুক্তি সকলোৰে বাবে অধিক সুবিধাজনক কৰি তুলিব। |
This model is released under the MIT License.
@misc{Ijaj7_Xetu_en_as_2025,
title = {Xetu: English-to-Assamese Translation Model},
author = {Ijaj Ahmed},
year = {2025},
howpublished = {\url{https://huggingface.co/Ijaj7/Xetu-english-to-assamese-translator}},
note = {Accessed: 2025-11-20}
}