nodes/pollockjj-ComfyUI-MultiGPU
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

docs: update README to include new MultiGPU loaders and clarify workflow steps

Browse Source
This commit is contained in:
John Pollock
2024-12-23 16:54:01 -06:00
parent 19c0abd5ec
commit c60771ace4
1 changed files with 13 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+13 -6
View File
@@ -18,7 +18,7 @@ Currently supported nodes (automatically detected if available):
- Standard ComfyUI loaders:
- CheckpointLoaderSimpleMultiGPU
- CLIPLoaderMultiGPU
- ControlNetLoaderMultiGPU
- ControlNetLoaderMultiGPU
- DualCLIPLoaderMultiGPU
- TripleCLIPLoaderMultiGPU
- UNETLoaderMultiGPU
@@ -26,13 +26,20 @@ Currently supported nodes (automatically detected if available):
- GGUF loaders (requires [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF)):
- UnetLoaderGGUFMultiGPU (supports quantized models like [flux1-dev-gguf](https://huggingface.co/city96/FLUX.1-dev-gguf))
- UnetLoaderGGUFAdvancedMultiGPU
- UnetLoaderGGUFAdvancedMultiGPU
- CLIPLoaderGGUFMultiGPU
- DualCLIPLoaderGGUFMultiGPU
- TripleCLIPLoaderGGUFMultiGPU
- Additional supported nodes:
- LoadFluxControlNet (requires [x-flux-comfy](https://github.com/XLabAI/x-flux-comfyui))
- XLabAI FLUX ControlNet:
- LoadFluxControlNetMultiGPU (requires [x-flux-comfy](https://github.com/XLabAI/x-flux-comfyui))
- Florence2:
- Florence2ModelLoaderMultiGPU (requires [ComfyUI-Florence2](https://github.com/kijai/ComfyUI-Florence2))
- DownloadAndLoadFlorence2ModelMultiGPU
- LTX Video Custom Checkpoint Loader:
- LTXVLoaderMultiGPU (requires [ComfyUI-LTXVideo](https://github.com/Lightricks/ComfyUI-LTXVideo))
All MultiGPU nodes can be found in the "multigpu" category in the node menu.
@@ -73,9 +80,9 @@ These workflows combine multiple features and non-core loaders types and may req
- [Download](examples/florence2_flux1dev_ltxv_2gpu_GGUF.json)
This workflow creates an img2txt2img2vid video generation pipeline by:
1. Taking a starting image for analysis by Florence2
1. Providing a starting image for analysis by Florence2
2. Using the Florence2 data for a FLUX.1 Dev image prompt
3. Taking the FLUX.1 image prompt and providing it as the starting image for LTX Video
3. Taking the resulting FLUX.1 image and provide it as the starting image for an LTX Video image-to-video generation
4. Generate a 5 second video based on the provided image
All models are distributed across available GPUs with no reloading on dual 3090s