krishnamishra8848/movie_sentiment_analysis

Movie Sentiment Analysis Model

Want to try out the model? Click here!

This model can detect movie sentiment analysis from all languages. However, since it uses RapidAPI for language detection and translation, the free tier only allows 500 requests per month. If the model is not making predictions, it means the monthly limit of 500 requests has been reached. Please try again later. You can also download our model and run it on your local system for unlimited use free of cost, or fine-tune it as needed—no problem!

Overview

This model is a fine-tuned version of distilbert-base-uncased for sentiment analysis on IMDB movie reviews. It has been trained on 20,000 labeled reviews, with an equal distribution of positive (10,000) and negative (10,000) examples. The model achieves 93% F1 score, precision, and recall, making it a state-of-the-art solution for movie sentiment analysis.

Model Details

Base Model: DistilBERT (distilbert-base-uncased)

Dataset: IMDB Movie Reviews (20,000 samples: 10k positive, 10k negative)

Performance:

F1 Score: 93%

Precision: 93%

Recall: 93%

Use Case: Analyze the sentiment of movie reviews as either positive or negative.

How to Use

# How To use model
from transformers import pipeline

# Load the pipeline
classifier = pipeline('sentiment-analysis', model='krishnamishra8848/movie_sentiment_analysis')

# Analyze sentiment
review = "I really enjoyed this movie from start to finish!"
result = classifier(review)

# Human-readable output
label_mapping = {"LABEL_0": "negative", "LABEL_1": "positive"}
human_readable_result = {
    "sentiment": label_mapping[result[0]['label']],
    "confidence": result[0]['score']
}

print(human_readable_result)