KNNFx/mt5-mcqa-history-no-explain

mT5 Multiple Choice QA Model for Vietnamese History

Model Description

This model is a fine-tuned version of csebuetnlp/mT5_multilingual_XLSum specifically trained for Vietnamese history multiple choice question answering tasks. The model is designed to answer historical questions without providing explanations.

Model Details

• Base Model: csebuetnlp/mT5_multilingual_XLSum
• Task: Multiple Choice Question Answering
• Language: Vietnamese
• Domain: Vietnamese History
• Training Type: No explanation generation (answer only)

Usage

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("KNNFx/mt5-mcqa-history-no-explain")
model = AutoModelForSeq2SeqLM.from_pretrained("KNNFx/mt5-mcqa-history-no-explain")

# Example usage
question = "Ai là vua đầu tiên của nhà Nguyễn?"
choices = ["A. Nguyễn Ánh", "B. Gia Long", "C. Minh Mạng", "D. Thiệu Trị"]
input_text = f"Question: {question} Choices: {', '.join(choices)}"

inputs = tokenizer(input_text, return_tensors="pt", max_length=512, truncation=True)
outputs = model.generate(**inputs, max_length=10)
answer = tokenizer.decode(outputs[0], skip_special_tokens=True)

Training Data

The model was trained on Vietnamese history questions covering various periods and topics in Vietnamese history.

Performance

This model focuses on accuracy without explanation generation, making it suitable for applications requiring quick and direct answers to historical questions.

Citation

If you use this model in your research, please cite:

@misc{mt5-mcqa-history-no-explain,
  title={mT5 Multiple Choice QA Model for Vietnamese History},
  author={KNNFx},
  year={2025},
  url={https://huggingface.co/KNNFx/mt5-mcqa-history-no-explain}
}

License

This model is released under the MIT License.