Jr0hamza/sentiment-analysis-model

Model Card for Sentiment Analysis Model

Model Details

Model Description

his model is a fine-tuned version of distilbert-base-uncased for binary sentiment analysis.
It is designed to classify text as either positive or negative sentiment.

• Developed by: HAMZA JR
• Model type: Transformer-based sentiment classifier
• Trained on: A labeled dataset for sentiment analysis (IMDB dataset)
• Library used: transformers, peft
• License: Apache 2.0
• Finetuned from model : distilbert-base-uncased

Performance: Accuracy Comparison

Model	Accuracy
Baseline (Pretrained)	47.90%
Fine-Tuned Model	87.50%

The fine-tuning process significantly improved accuracy from 47.90% to 87.50%, making the model much better for sentiment classification.

Uses

How to Get Started with the Model

Use the code below to classify text using the fine-tuned model:

from transformers import pipeline

def classify_text(model_name, text):
    """
    Classifies text using a Hugging Face model and maps numeric labels to meaningful labels.

    Args:
        model_name (str): The name of the Hugging Face model.
        text (str): The input text to classify.

    Returns:
        dict: A dictionary with the predicted label and confidence score.
    """

    # Define custom label mapping
    label_map = {0: "negative", 1: "positive"}  

    # Load the pipeline
    classifier = pipeline("text-classification", model=model_name)

    # Get the prediction
    prediction = classifier(text)

    # Convert label ID to meaningful text
    for pred in prediction:
        pred["label"] = label_map[int(pred["label"].split("_")[-1])]  # Extract and map the label

    return prediction

# Example Usage:
model_name = "Jr0hamza/sentiment-analysis-model"
text = "I absolutely loved this movie! It was fantastic."
result = classify_text(model_name, text)
print(result) 








- PEFT 0.14.0