Files
Mert de13aab713 feat(ml): improve model loading (#31742)
* use recent mimalloc

* jit caching

* bump onnxruntime-gpu

* explicit memory release

* refactor in-memory cache

* configure hf org and revision

* refactor model load

* move orjson response

* disable direct submission for i915

* flush denormals

* update tests and lock

* disable suboptimal fusions

* respect resize_mode

* increase threads for clip text
2026-09-25 14:56:43 -04:00

11 lines
218 B
Python

import ctypes
_process = ctypes.CDLL(None)
def release() -> None:
if hasattr(_process, "mi_collect"):
_process.mi_collect(True)
elif hasattr(_process, "malloc_trim"):
_process.malloc_trim(0)