# T5 Form Question Generator (Fine-Tuned)

This model, [`anjalisibu28/t5-form-question-generator`](https://huggingface.co/krishnarandad01/t5-form-question-generator), is a fine-tuned version of [`t5-small`](https://huggingface.co/t5-small) on a **custom dataset** of 1000+ structured `form field → question` pairs.

It is built to **generate natural-sounding questions** from form fields such as:
- `full_name` → *What is your full name?*
- `blood_group` → *What is your blood group?*
- `fathers_occupation` → *What does your father do for a living?*

## Authors
This project was developed by:
- **Anjali Mishra**

## Intended Use
This model is ideal for:
- Preprocessing automation in offline form-filling apps
- Conversational bots that collect structured data
- Assistive agents for field-level prompt generation

## How to Use
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("anjalisibu28/t5-form-question-generator")
model = AutoModelForSeq2SeqLM.from_pretrained("anjalisibu28/t5-form-question-generator")

field = "passport_number"
input_text = "generate question: " + field

inputs = tokenizer(input_text, return_tensors='pt')
outputs = model.generate(**inputs, max_length=64)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

##  Training Info
- Base Model: `t5-small`
- Dataset: Custom CSV with 1000+ rows
- Epochs: 5
- Max length: 64 tokens
- Trained using: Transformers + Datasets on Google Colab

## License
Apache-2.0

## Contact
For feedback or collaboration, reach out to anjali mishra via [Hugging Face](https://huggingface.co/anjalisibu28)