poxularandCursor
4965f27ba4
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>
24 lines
814 B
Python
24 lines
814 B
Python
"""Torch/CUDA pixel pipeline (develop, detection buffers, LUT apply). CPU parity path unchanged."""
|
|
|
|
from .device import gpu_pipeline_available, get_torch_device
|
|
from .pipeline import apply_probe_export_gpu, build_merged_lut_probe_style_gpu, color_match_one_shot_gpu, profile_gpu_pipeline
|
|
from .pipeline_full_gpu import (
|
|
apply_probe_export_full_gpu,
|
|
build_merged_lut_full_gpu,
|
|
color_match_one_shot_full_gpu,
|
|
profile_full_gpu_pipeline,
|
|
)
|
|
|
|
__all__ = [
|
|
"apply_probe_export_gpu",
|
|
"apply_probe_export_full_gpu",
|
|
"build_merged_lut_probe_style_gpu",
|
|
"build_merged_lut_full_gpu",
|
|
"color_match_one_shot_gpu",
|
|
"color_match_one_shot_full_gpu",
|
|
"profile_gpu_pipeline",
|
|
"profile_full_gpu_pipeline",
|
|
"get_torch_device",
|
|
"gpu_pipeline_available",
|
|
]
|