2 Commits
Author SHA1 Message Date
poxularandCursor e975fbc3b6 Fix GPU Full node reference tensor device handling.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-06-02 01:01:55 +02:00
poxularandCursor 4965f27ba4 Add full GPU pipeline with ComfyUI nodes and phase profiling.
Ports scene extract, face ONNX preprocess, and net reference resize to Torch/CUDA with minimal CPU sync, while keeping the CPU parity path unchanged.

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-06-02 01:01:46 +02:00