distilbert-goodreads-genres

Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

Model Details

Model Description

DistilBERT Goodreads Genre Classification

Model Description

This model is a fine-tuned version of distilbert-base-cased for Goodreads book review genre classification. The model was trained using the Hugging Face Transformers library on the UCSD Goodreads dataset as part of an MLOps assignment.

The objective of the project was to build a complete MLOps workflow including:

• fine-tuning a Hugging Face model
• experiment tracking using Weights & Biases (W&B)
• training on Kaggle GPU
• deployment to Hugging Face Hub

Model Details

• Base Model: distilbert-base-cased
• Task: Text Classification
• Number of Labels: 8
• Framework: Hugging Face Transformers
• Training Platform: Kaggle GPU
• Experiment Tracking: Weights & Biases (W&B)

Labels

The model predicts the following genres:

• children
• comics_graphic
• fantasy_paranormal
• history_biography
• mystery_thriller_crime
• poetry
• romance
• young_adult

Training Details

Dataset

UCSD Goodreads Reviews Dataset

Training Hyperparameters

• Epochs: 5
• Batch Size: 16
• Learning Rate: 3e-5
• Max Sequence Length: 512

Evaluation Metrics

Replace these values with your actual results:

Usage

from transformers import pipeline

classifier = pipeline(
    "text-classification",
    model="YOUR_USERNAME/distilbert-goodreads-genres"
)

result = classifier("This book was amazing and full of adventure.")
print(result)

Project Links

• Kaggle Notebook: YOUR_KAGGLE_LINK
• W&B Dashboard: YOUR_WANDB_LINK
• GitHub Repository: YOUR_GITHUB_LINK

Disclaimer

This model was created for educational purposes as part of an academic MLOps assignment.