Karim Mansour
Update README.md
a64948a verified
metadata
license: mit

IMDb Sentiment Analysis - Fine-tuned BERT Model

This project fine-tunes a pre-trained BERT model from Hugging Face on the IMDb movie reviews dataset for binary sentiment classification (positive/negative).

Project Overview

  • Task: Sentiment Analysis (Binary Classification)
  • Dataset: IMDb
  • Model: bert-base-uncased fine-tuned using Transformers and PyTorch
  • Accuracy: About 92.8% on validation data

Sample Usage

Load the model and tokenizer

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("saved_model")
model = AutoModelForSequenceClassification.from_pretrained("saved_model")