Skip to content

Examples

Real-world examples of using GPUX for different ML tasks.

🎯 Overview

Learn by example! Each guide includes complete working code, configuration, and explanations.

📚 Available Examples

Sentiment Analysis

BERT-based text classification

Classify text sentiment using a fine-tuned BERT model.

  • ✅ Text preprocessing and tokenization
  • ✅ Binary classification (positive/negative)
  • ✅ Complete end-to-end example
  • ⏱️ Time: 15 minutes

Image Classification

ResNet-50 for ImageNet

Classify images into 1000 ImageNet categories.

  • ✅ Image preprocessing (resize, normalize)
  • ✅ Top-K predictions
  • ✅ Batch processing
  • ⏱️ Time: 20 minutes

Object Detection

YOLOv8 real-time detection

Detect objects in images with bounding boxes.

  • ✅ YOLO model setup
  • ✅ Bounding box predictions
  • ✅ NMS post-processing
  • ⏱️ Time: 25 minutes

LLM Inference

Small language model serving

Run text generation with a small LLM.

  • ✅ Tokenization and decoding
  • ✅ Text generation
  • ✅ Streaming responses
  • ⏱️ Time: 30 minutes

Speech Recognition

Whisper audio transcription

Transcribe speech to text using OpenAI Whisper.

  • ✅ Audio preprocessing
  • ✅ Multi-language support
  • ✅ Timestamp generation
  • ⏱️ Time: 25 minutes

Embedding Generation

Sentence transformers

Generate vector embeddings for semantic search.

  • ✅ Text embeddings
  • ✅ Similarity search
  • ✅ Batch processing
  • ⏱️ Time: 20 minutes

Multi-Modal Models

CLIP image-text matching

Match images with text descriptions using CLIP.

  • ✅ Image and text encoding
  • ✅ Similarity scoring
  • ✅ Zero-shot classification
  • ⏱️ Time: 30 minutes

🚀 Getting Started

Prerequisites

  • GPUX installed (uv add gpux)
  • Python 3.11+
  • Basic understanding of the Tutorial

Example Structure

Each example includes:

  1. Overview - What you'll build
  2. Model Preparation - Converting/downloading the model
  3. Configuration - Complete gpux.yml
  4. Running - Step-by-step execution
  5. Results - Expected output
  6. Production - Deployment considerations

📖 How to Use Examples

Follow Along

Each example is self-contained. Pick one and follow step-by-step.

Adapt for Your Use Case

Modify examples to fit your specific needs.

Production Deployment

Examples include production deployment tips.

💡 Tips

Start Simple

Begin with Sentiment Analysis - it's the easiest example.

GPU Recommended

While examples work on CPU, GPU provides much better performance.

Download Models First

Large models may take time to download. Plan accordingly.

🆘 Need Help?