SmartPool Examples

This package contains practical examples demonstrating the capabilities of SmartPool for various computational tasks.

Examples Overview

1. Prime Number Counting (count_prime)

Count the number of prime numbers below 10000 using smartpool.ProcessPool. Demonstrates basic usage of smartpool.ProcessPool.

Running the Example

python -m smartpool_examples.count_prime

2. Cross-Validation for Deep Learning models (cross_validation)

Demonstrates SmartPool's capabilities for machine learning workloads with GPU resource management.

Running the Example

# Using ProcessPool  
python -m smartpool_examples.cross_validation --pool smartpool.ProcessPool

# Using ThreadPool
python -m smartpool_examples.cross_validation --pool smartpool.ThreadPool

# Using multiprocessing.Pool
python -m smartpool_examples.cross_validation --pool multiprocessing.Pool

# Using concurrent.futures.ProcessPoolExecutor
python -m smartpool_examples.cross_validation --pool concurrent.futures.ProcessPoolExecutor

# Using concurrent.futures.ThreadPoolExecutor
python -m smartpool_examples.cross_validation --pool concurrent.futures.ThreadPoolExecutor

# Using joblib.Parallel(backend='loky')
python -m smartpool_examples.cross_validation --pool joblib.Parallel(backend='loky')

# Using joblib.Parallel(backend='threading')
python -m smartpool_examples.cross_validation --pool joblib.Parallel(backend='threading')

# Using Ray
python -m smartpool_examples.cross_validation --pool ray

What it Demonstrates

• GPU memory management and core allocation
• Automatic device selection (CPU vs GPU)
• Cross-validation pipeline parallelization
• Resource monitoring during training
• Performance comparison with external frameworks

3. ONNX Inference (onnx_infer)

Runs batched ONNX model inference using InferSessionPool for concurrent GPU/CPU execution. Automatically manages inference sessions across worker threads.

Running the Example

python -m smartpool_examples.onnx_infer --max-workers 4

What it Demonstrates

• InferSessionPool creation and session lifecycle management
• Multi-threaded inference with automatic device placement
• COCO-format image preprocessing (resize, normalize, letterbox)
• Softmax + top-5 postprocessing
• Progress bars for downloads and inference steps

License

MIT License - see main smartpool repository for details