Kakaarot/Gemma-HuggingFace_TextCompletion_Demo

Gemma Text Generator

A streamlined web application that leverages Google's Gemma-2B language model to generate text with customizable tones and parameters.

Overview

This project is a Streamlit-based web application that allows users to generate text using Google DeepMind's Gemma-2B language model. The application features an intuitive interface where users can input prompts, select different tones (Funny, Serious, or Poetic), and adjust various generation parameters to customize the output.

Features

• —Text Generation: Generate text completions from user prompts
• —Tone Selection: Choose from three different writing styles:
• —Funny: Witty and humorous responses with unexpected twists
• —Serious: Thoughtful and professional responses with logical reasoning
• —Poetic: Vivid, lyrical responses with metaphors and imagery
• —Customizable Parameters:
• —Word count: Control the approximate length of generated text
• —Temperature: Adjust the creativity and randomness
• —Top-p (Nucleus Sampling): Control the diversity of outputs
• —Repetition Penalty: Prevent repetitive phrases
• —Word Cloud Visualization: See a visual representation of word frequency in the generated text
• —One-Click Examples: Try pre-configured examples with a single click

Technology Stack

• —Streamlit: Web application framework
• —Hugging Face Transformers: Library for accessing pre-trained language models
• —Google Gemma-2B: The underlying language model for text generation
• —PyTorch: Deep learning framework
• —WordCloud: Library for creating visual representations of text data
• —Matplotlib: For visualizing the word cloud

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference