poxular and Cursor
0361126045
Restore 1ccd29d CUDA one-shot path without breaking node imports.
...
Add pipeline_v1 (GPU build + GPU apply) as a standalone module so partial
git checkouts of inference_cuda no longer break GPU Full/Fast imports.
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 01:54:41 +02:00
poxular and Cursor
f223858ac4
Add GPU Fast pipeline for sub-3s H100 inference.
...
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 01:16:07 +02:00
poxular and Cursor
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
poxular and Cursor
c85abee31e
Speed up GPU path: fused one-shot, GPU net resize, vectorized luminance.
...
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 00:53:08 +02:00
poxular and Cursor
07077d1ed5
Add force_color_look support to GPU develop stack.
...
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 00:49:13 +02:00
poxular and Cursor
6dda38faac
Fix GPU color matmul dtype mismatch (float64 matrices).
...
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 00:46:20 +02:00
poxular and Cursor
1ccd29df6f
Add Torch GPU pipeline for CUDA ComfyUI nodes.
...
Co-authored-by: Cursor <cursoragent@cursor.com >
2026-06-02 00:43:01 +02:00
