metadata

title: Amazon Shoe Review
emoji: 📚
colorFrom: pink
colorTo: red
sdk: gradio
sdk_version: 4.37.2
app_file: app.py
pinned: false
license: mit

DistilBERT-based Sentiment Analysis Project for Predicting Shoe Review Ratings

This project implements a sentiment analysis model to predict star ratings for Amazon shoe reviews. It leverages DistilBERT-base-uncased, a pre-trained transformer model from Hugging Face, fine-tuned on a dataset of Amazon shoe reviews.

Project Structure

01. Data Preparation.ipynb: This notebook handles the entire data pipeline:
- Data Collection: An amazon-shoe-review dataset has been collected from here.
- Data Cleaning & Preprocessing: Data cleaning and preprocessing has been done to prepare it for model training.
- Data Sharing: After preprocessing the dataset has been pushed to HuggingFace Hub. Dataset Link
02. Model Training.ipynb: This notebook covers:
- Fine-tuning the pre-trained DistilBERT-base-uncased model from Hugging Face on the preprocessed data for predicting shoe review star ratings.
03. Save Model to Hub.ipynb: This notebook handles:
- Model Evaluation: Predicitons are made on few examples to evaluate the fine-tuned model.
- Model Sharing: The fine-tuned model is then pushed to HuggingFace model hub. Model Link
requirements.txt: Lists the dependencies needed for the project:
- transformers
- gradio
- torch
app.py: A script to deploy the model using Gradio for a web-based interface.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference