huchukatoandDevin
970f51649b
- Add RIFE v4.25 architecture support in rife_arch.py (Head with 16/4 channels, 5 IFBlocks with c=192/128/96/64/32, scale_list [16,8,4,2,1], no ensemble) - Add rife425.pth to export_onnx.py with HF fallback URL - Update load_rife_config.json: dropdown now only shows rife425 and rife49 (removed rife47/48/417/426/sudo_rife4 that never worked) - Set rife425 as default - Make TORCH_DEVICE fallback to CPU for ONNX export on non-CUDA - Pin polygraphy==0.49.26 for TensorRT 10.15.1 compatibility Generated with [Devin](https://devin.ai) Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
25 lines
1.0 KiB
Plaintext
25 lines
1.0 KiB
Plaintext
# Base dependencies - TensorRT will be auto-installed based on CUDA detection
|
|
# The node will automatically detect CUDA 12/13 and install appropriate TensorRT packages.
|
|
# The NVIDIA CUDA Toolkit must already be installed on the system; it is NOT installed via pip.
|
|
#
|
|
# Installation happens automatically when the node loads for the first time
|
|
# No manual TensorRT installation required!
|
|
|
|
# Base dependencies that work with any CUDA version
|
|
einops
|
|
colored
|
|
# Pin to the Polygraphy version bundled with TensorRT 10.15.1.
|
|
# Newer Polygraphy can break trt.OnnxParser init with:
|
|
# pybind11::init(): factory function returned nullptr
|
|
polygraphy==0.49.26
|
|
cuda-python
|
|
requests
|
|
tqdm
|
|
|
|
# Note: TensorRT packages are auto-installed on first node load:
|
|
# CUDA 13: tensorrt_cu13==10.15.1.29 (RTX 50 series)
|
|
# CUDA 12: tensorrt-cu12==10.13.3.9 (RTX 30/40 series)
|
|
#
|
|
# Manual installation files are provided for reference only:
|
|
# - requirements_cu12.txt (CUDA 12 specific)
|
|
# - requirements_cu13.txt (CUDA 13 specific) |