intel / onnxruntime

ONNX Runtime: cross-platform, high performance scoring engine for ML models
MIT License
56 stars 22 forks source link

Localize model proto for enhanced memory #417

Closed preetha-intel closed 3 weeks ago

preetha-intel commented 1 month ago

Description

Localize model proto to optimize memory.

Motivation and Context

Optimize the peek memory hit during execution of OVEP