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

Commit Graph

  • 3814340e45 added test inpaint image from https://github.com/alimama-creative/FLUX-Controlnet-Inpainting John Pollock 2025-01-11 10:36:50 -06:00
  • ba24f572ee feat: Add DeviceSelectorMultiGPU node and device selection functionality - allowing the linking of one or more MultiGPU nodes to the same cuda device in cases where this would prevent accidental errors if should there be a device mismatch futher along in the pipeline due to non-loader nodes performing device-specific tasks or logic. John Pollock 2025-01-11 10:03:40 -06:00
  • 5c3c1a7f3b feat: Implement MultiGPU support for Hunyuan models and add module existence checks John Pollock 2025-01-07 12:18:03 -06:00
  • 61c9204207 Merge branch 'dev' of https://github.com/pollockjj/ComfyUI-MultiGPU into dev sync with upstream John Pollock 2025-01-07 07:51:43 -06:00
  • 6faebe85d7 syncing with latest release from :main: John Pollock 2025-01-07 06:51:01 -06:00
  • 116dd935f2 chore: Rename publish workflow file and update version to 1.2.1 in pyproject.toml John Pollock 2025-01-05 16:03:53 -06:00
  • 4611332d12 chore: Bump version to 1.2.0 in pyproject.toml John Pollock 2025-01-04 19:44:05 -06:00
  • 3ce3598a5d feat: Add initial MultiGPU support for Pulid model, added module existence checks before creating MultiGPU node variant John Pollock 2025-01-02 22:11:28 -06:00
  • edcd5cc612 Merge branch 'main' of https://github.com/pollockjj/ComfyUI-MultiGPU John Pollock 2025-01-02 21:15:08 -06:00
  • 4aec967384 fix: hard coded all supported nodes. If the parent custom_node is installed then it will inherit the needed functionality at run-time, fully eliminating any load depenencies. No outside custom_nodes need to be pre-loaded and no python is inspected. New MultiGPU variants are created in a self-contained manner. John Pollock 2025-01-02 21:09:39 -06:00
  • 26771e4b94 intermediate commit - code to constuct nodes from working example (e.g. when a custom_node loads before MultiGPU and we have the registered data. John Pollock 2025-01-02 15:06:14 -06:00
  • fb522a4d97 Fix inheritance issues in register_module_new by using shared namespace for node definitions John Pollock 2025-01-02 07:18:01 -06:00
  • 1d12e2ee1f feat(multigpu): generalize LTX module registration for custom nodes John Pollock 2024-12-31 15:51:43 -06:00
  • 3f57e7fe83 intermediate step John Pollock 2024-12-31 05:04:41 -06:00
  • a725a7bc03 Add new module registration function to extract and log node definitions John Pollock 2024-12-31 03:30:00 -06:00
  • 1d020adbdb feat: Add hard-coded registration for LTX and Florence2 nodes with module existence checks John Pollock 2024-12-30 22:22:52 -06:00
  • 1e553cfe22 Add utility function to check module existence before registration for hard-coded MultiGPU nodes John Pollock 2024-12-30 21:00:48 -06:00
  • 198365dfbe Add hard-coded registration for LTX and Florence2 nodes in MultiGPU setup for debug purposes. John Pollock 2024-12-30 20:48:41 -06:00
  • 9464cf6cc1 fix: Enhance error handling during module execution in MultiGPU registration John Pollock 2024-12-30 15:13:12 -06:00
  • 4ac8d33270 fix: Use local module dictionary instead of global for custom node registration John Pollock 2024-12-30 11:17:31 -06:00
  • fcf054006d Mew method merged in with old code base so less of a shock. John Pollock 2024-12-30 11:03:55 -06:00
  • 46941e936d Resolved merge conflicts John Pollock 2024-12-30 10:15:34 -06:00
  • 14758b1985 changed methodology (again) to run their code and then scan their LOCAL dict, not the global dict. This will work and be robust I strongly believe John Pollock 2024-12-30 10:12:28 -06:00
  • caee8716b6 Refactor - intermediate step of Implementing MultiGPU node registration and class definition retrieval for custom nodes John Pollock 2024-12-30 07:54:57 -06:00
  • 1fe269e8d6 Update README to clarify NF4 Checkpoint Format Loader requirements John Pollock 2024-12-29 23:16:02 -06:00
  • 58cb0ab59a Officially adding CheckpointLoaderNF4 support from ComfyUI_bitsandbytes_NF4 John Pollock 2024-12-29 23:14:11 -06:00
  • 2c6fb487d3 editing for clarity John Pollock 2024-12-28 14:49:02 -06:00
  • 0b3ba045c4 Change loading to a determinsitc process by querying each node. This relies upon each node's idempotency that will need to get checked. John Pollock 2024-12-28 13:39:23 -06:00
  • e4b570de43 contunued NF4, working towards general solution. John Pollock 2024-12-28 12:13:21 -06:00
  • d9f7ab23e5 debugging NF4 loader issue and sequence loading in general. John Pollock 2024-12-28 12:03:07 -06:00
  • 67c1641ab2 Update to examples to include CPU offloads, as well as update readme.md with new examples and their requirements to run. John Pollock 2024-12-27 13:22:16 -06:00
  • 4a45737583 Update README for clarity on multi-GPU and CPU offloading capabilities; correct example JSON link John Pollock 2024-12-27 04:59:32 -06:00
  • d13f821547 Bump version to 1.1.0 and update README to reflect CPU offloading capabilities; rename example JSON for clarity John Pollock 2024-12-27 04:54:21 -06:00
  • d099ecd497 Enhance device selection in get_torch_device_patched and override_class to include 'cpu' as a valid option for multi-GPU support John Pollock 2024-12-27 03:06:50 -06:00
  • dead358227 Add MMAudioSampler to experimental audio model loaders in __init__.py so it stays in sync with the Model and FeatureUtils Loaders. I suspect this is because it is querying the cuda device directly. This would give the wrong answer half the time depending on how the other loaders were ran via the sequencing logic. John Pollock 2024-12-26 14:59:16 -06:00
  • 94cf7d9779 Update project description and version in pyproject.toml; add experimental audio model loaders in __init__.py John Pollock 2024-12-26 08:30:27 -06:00
  • a58fb9258b Refactor README.md for improved clarity and for compliance to markdown format John Pollock 2024-12-25 08:54:16 -06:00
  • 38e476e8da Update project description and version in pyproject.toml John Pollock 2024-12-25 08:40:41 -06:00
  • 065679af4d Refactor README.md for clarity and consistency in multi-GPU usage documentation John Pollock 2024-12-25 08:33:26 -06:00
  • 07a66fd8b3 Add documentation for using GGUF quantized models across GPUs for video generation John Pollock 2024-12-24 21:29:31 -06:00
  • 211b957f75 Adding hunyuan video 2 gpu example workflow. John Pollock 2024-12-24 21:19:47 -06:00
  • 88502d74b2 fixed copy/paste error John Pollock 2024-12-24 11:13:12 -06:00
  • 71ee2809a8 Minor changes to README.md to reflect edits from Reddit post John Pollock 2024-12-24 11:11:36 -06:00
  • e4e701a16a Merge pull request #1 from ComfyNodePRs/pyproject pollockjj 2024-12-24 10:55:36 -06:00
  • 9ce9db80ca Merge pull request #2 from ComfyNodePRs/publish pollockjj 2024-12-24 10:54:43 -06:00
  • 924ea16c81 Update pyproject.toml pollockjj 2024-12-24 10:44:01 -06:00
  • c60771ace4 docs: update README to include new MultiGPU loaders and clarify workflow steps John Pollock 2024-12-23 16:54:01 -06:00
  • 19c0abd5ec Updated example workflows, added additional experimental workflows, added coverage for LTX loader John Pollock 2024-12-23 15:39:23 -06:00
  • 56e2e61e52 refactor: streamline MultiGPU node initialization John Pollock 2024-12-22 22:36:59 -06:00
  • a652a7f264 refactor: adopt City96's streamlined MultiGPU implementation John Pollock 2024-12-22 08:28:35 -06:00
  • 5182bfe5e3 chore(pyproject): Add pyproject.toml for Custom Node Registry snomiao 2024-12-20 23:50:41 +00:00
  • 6963fa087a chore(publish): Add Github Action for Publishing to Comfy Registry snomiao 2024-12-20 23:50:40 +00:00
  • 186723f34d Add mochi and ltxv support to CLIPLoaderMultiGPU John Pollock 2024-12-18 17:01:01 -06:00
  • b5462877f3 Initial commit Alexander Dzhoganov 2024-08-04 18:37:23 +03:00