pytorch-onnx

What problem does it solve?

This Skill enables the conversion of PyTorch models into the ONNX format, facilitating cross-platform deployment and optimization for inference.

Core Features & Use Cases

• Model Export: Convert PyTorch models to ONNX, supporting dynamic input shapes.
• Inference Optimization: Utilize ONNX Runtime for efficient execution across various hardware.
• Quantization: Reduce model size and latency through INT8 quantization.
• Use Case: Deploy a trained PyTorch image classification model to a C++ application or an edge device by exporting it to ONNX and optimizing it with ONNX Runtime.

Quick Start

Export the provided PyTorch model to ONNX format with dynamic axes support.