onnxruntime/docs/tutorials/accelerate-pytorch/pytorch.md

title	parent	grand_parent	nav_order
Accelerate PyTorch Inference	Accelerate PyTorch	Tutorials	1

Accelerate PyTorch model inferencing

{: .no_toc }

ONNX Runtime can be used to accelerate PyTorch models inferencing.

Convert model to ONNX

{: .no_toc }

• Basic PyTorch export through torch.onnx
• Super-resolution with ONNX Runtime
• Export PyTorch model with custom ops

Accelerate PyTorch model inferencing

{: .no_toc }

BERT

• Accelerate BERT model on CPU
• Accelerate BERT model on GPU
• Accelerate reduced size BERT model through quantization

GPT-2

• Accelerate GPT2 on CPU
• Accelerate GPT2 (with one step search) on CPU