Ericcraw/ort allocator hacking

intel / onnxruntime

ONNX Runtime: cross-platform, high performance scoring engine for ML models

MIT License

57 stars 23 forks source link

Closed saurabhkale17 closed 2 months ago

saurabhkale17 commented 2 months ago

Avoid memcpy for input and output tensor for OVEP NPU. Allocating buffer to remote device using the allocator.