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")