|
--- |
|
license: openrail |
|
datasets: |
|
- Binarybardakshat/SVLM-ACL-DATASET |
|
language: |
|
- en |
|
library_name: transformers |
|
tags: |
|
- code |
|
--- |
|
# SVLM: A Question-Answering Model for ACL Research Papers |
|
|
|
This model, `SVLM`, is designed to answer questions based on research papers from the ACL dataset. It leverages the BART architecture to generate precise answers from scientific abstracts. |
|
|
|
## Model Details |
|
|
|
- **Model Architecture:** BART (Bidirectional and Auto-Regressive Transformers) |
|
- **Framework:** TensorFlow |
|
- **Dataset:** [Binarybardakshat/SVLM-ACL-DATASET](https://huggingface.co/datasets/Binarybardakshat/SVLM-ACL-DATASET) |
|
- **Author:** @binarybard (Akshat Shukla) |
|
- **Purpose:** The model is trained to provide answers to questions from the ACL research paper dataset. |
|
|
|
## Usage |
|
|
|
To use this model with the Hugging Face Interface API: |
|
|
|
```python |
|
from transformers import AutoTokenizer, TFAutoModelForSeq2SeqLM |
|
|
|
# Load the model and tokenizer |
|
tokenizer = AutoTokenizer.from_pretrained("Binarybardakshat/SVLM") |
|
model = TFAutoModelForSeq2SeqLM.from_pretrained("Binarybardakshat/SVLM") |
|
|
|
# Example input |
|
input_text = "What is the main contribution of the paper titled 'Your Paper Title'?" |
|
|
|
# Tokenize input |
|
inputs = tokenizer(input_text, return_tensors="tf", padding=True, truncation=True) |
|
|
|
# Generate answer |
|
outputs = model.generate(inputs.input_ids, max_length=50, num_beams=5, early_stopping=True) |
|
answer = tokenizer.decode(outputs[0], skip_special_tokens=True) |
|
|
|
print("Answer:", answer) |
