Distilbert Sentiment

🚀 Zero-Shot Sentiment Classification with DistilBERT

This project uses DistilBERT for zero-shot sentiment classification on the IMDB dataset, leveraging the Hugging Face Transformers library.

🚀 Quick Start

The project is designed for zero-shot sentiment classification. It uses the distilbert-base-uncased model from the transformers library and is trained on the stanfordnlp/imdb dataset.

📦 Installation

Since the project uses the transformers library, you can install it using the following command:

pip install transformers torch

💻 Usage Examples

Basic Usage

# Sample text to predict
text = "I love this movie, it was fantastic!"

# Tokenize the input text
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True)

# Get model predictions
with torch.no_grad():
    outputs = model(**inputs)

# Get the logits (model's raw output)
logits = outputs.logits

# Convert logits to probabilities (if needed) and get the predicted class (0 or 1)
predictions = torch.argmax(logits, dim=-1).item()

# Map the prediction to sentiment labels
labels = {0: "NEGATIVE", 1: "POSITIVE"}  # Assuming binary classification
predicted_label = labels[predictions]

print(f"Predicted Sentiment: {predicted_label}")

📄 License

This project is licensed under the Apache 2.0 License.