sujan-maharjan/imdb-sentiment-bert

IMDB Sentiment Analysis Model

This model performs sentiment analysis on movie reviews, classifying them as either positive or negative. It's based on a fine-tuned BERT model trained on the IMDB dataset.

Model Description

• Model Type: Fine-tuned BERT for sentiment analysis
• Language: English
• Base Model: bert-base-uncased
• Training Data: IMDB Movie Reviews Dataset
• Task: Binary Sentiment Classification (Positive/Negative)

Usage

from transformers import pipeline

classifier = pipeline("sentiment-analysis", model="your-username/imdb-sentiment-bert")
result = classifier("This movie was fantastic! I really enjoyed it.")
print(result)

Training Procedure

The model was fine-tuned on the IMDB dataset containing 50K movie reviews. The training process involved:

• Using AdamW optimizer with learning rate 2e-5
• Training for 3 epochs
• Using max sequence length of 512
• Applying dropout rate of 0.3

Limitations

• The model is specifically trained for movie reviews
• Maximum input length is 512 tokens
• Performance might vary on other types of text