search

patientx / ComfyUI-Zluda

The most powerful and modular stable diffusion GUI, api and backend with a graph/nodes interface. Now ZLUDA enhanced for better AMD GPU performance.
GNU General Public License v3.0
162 stars 11 forks source link

tensorflow error #37

Closed azamet90 closed 2 weeks ago

azamet90 commented 3 weeks ago

Expected Behavior

im not sure

Actual Behavior

[Prompt Server] web root: D:\yapay zeka\ComfyUI-Zluda\web 2024-11-04 16:33:18.741765: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0. 2024-11-04 16:33:19.783842: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0.

Steps to Reproduce

run it from comfyui.bat

Debug Logs

*** Checking and updating to new version if possible
Already up to date.

[START] Security scan
[DONE] Security scan
## ComfyUI-Manager: installing dependencies done.
** ComfyUI startup time: 2024-11-04 16:33:12.251714
** Platform: Windows
** Python version: 3.11.0 (main, Oct 24 2022, 18:26:48) [MSC v.1933 64 bit (AMD64)]
** Python executable: D:\yapay zeka\ComfyUI-Zluda\venv\Scripts\python.exe
** ComfyUI Path: D:\yapay zeka\ComfyUI-Zluda
** Log path: D:\yapay zeka\ComfyUI-Zluda\comfyui.log

Prestartup times for custom nodes:
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\rgthree-comfy
   2.7 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Manager

Total VRAM 16368 MB, total RAM 32683 MB
pytorch version: 2.3.0+cu118
Set vram state to: NORMAL_VRAM

***----------------------ZLUDA--------------------------***
  ::  ZLUDA detected, disabling non-supported functions.
  ::  (cuDNN, flash_sdp, mem_efficient_sdp disabled)
***-----------------------------------------------------***
  ::  Device: cuda:0 AMD Radeon RX 6900 XT [ZLUDA] : native

Using sub quadratic optimization for cross attention, if you have memory or speed issues try using: --use-split-cross-attention
[Prompt Server] web root: D:\yapay zeka\ComfyUI-Zluda\web
2024-11-04 16:33:18.741765: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-11-04 16:33:19.783842: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
[Crystools INFO] Crystools version: 1.20.0
[Crystools INFO] CPU: Intel(R) Core(TM) i9-10900K CPU @ 3.70GHz - Arch: AMD64 - OS: Windows 10
[Crystools INFO] Pynvml (Nvidia) initialized.
[Crystools WARNING] ZLUDA detected. GPU monitoring will be disabled.
[Crystools WARNING] No GPU with CUDA detected.
loading image tools...
### Loading: ComfyUI-Impact-Pack (V7.10.7)
### Loading: ComfyUI-Impact-Pack (Subpack: V0.7)
[Impact Pack] Wildcards loading done.
Total VRAM 16368 MB, total RAM 32683 MB
pytorch version: 2.3.0+cu118
Set vram state to: NORMAL_VRAM

***----------------------ZLUDA--------------------------***
  ::  ZLUDA detected, disabling non-supported functions.
  ::  (cuDNN, flash_sdp, mem_efficient_sdp disabled)
***-----------------------------------------------------***
  ::  Device: cuda:0 AMD Radeon RX 6900 XT [ZLUDA] : native

### Loading: ComfyUI-Manager (V2.51.9)
### ComfyUI Revision: 2993 [8fd09936] | Released on '2024-11-04'
[ReActor] - STATUS - Running v0.5.1-b2 in ComfyUI
[ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/model-list.json
[ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/alter-list.json
[ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/github-stats.json
[ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json
[ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/extension-node-map.json
Torch version: 2.3.0+cu118
here: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-tbox
Using ckpts path: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-tbox\..\..\models\annotator
Using symlinks: False
Using ort providers: ['CUDAExecutionProvider', 'DirectMLExecutionProvider', 'OpenVINOExecutionProvider', 'ROCMExecutionProvider', 'CPUExecutionProvider']
DWPose: Onnxruntime with acceleration providers detected
[R44]: Loaded Runtime44 nodes
Please 'pip install xformers'
Nvidia APEX normalization not installed, using PyTorch LayerNorm

[rgthree-comfy] Loaded 42 epic nodes. 🎉

Searge-SDXL v4.3.1 in D:\yapay zeka\ComfyUI-Zluda\custom_nodes\SeargeSDXL
WAS Node Suite: OpenCV Python FFMPEG support is enabled
WAS Node Suite Warning: `ffmpeg_bin_path` is not set in `D:\yapay zeka\ComfyUI-Zluda\custom_nodes\was-node-suite-comfyui\was_suite_config.json` config file. Will attempt to use system ffmpeg binaries if available.
WAS Node Suite: Finished. Loaded 218 nodes successfully.

        "The artist's world is limitless. It can be found anywhere, far from where he lives or a few feet away. It is always on his doorstep." - Paul Strand

Import times for custom nodes:
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\websocket_image_save.py
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\sdxl_prompt_styler
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-deepcache
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-InstantIDUtils
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\comfyui_met_suite
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\cg-use-everywhere
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI_InstantID
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI_FaceAnalysis
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\comfyui-instantId-faceswap
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-TeaNodes
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\comfyui_r44_nodes
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI_IPAdapter_plus
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\comfyui-various
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI_essentials
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-utils-nodes
   0.0 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\rgthree-comfy
   0.1 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\facerestore_cf
   0.1 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Crystools
   0.1 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-KJNodes
   0.2 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\comfyui-reactor-node
   0.2 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Impact-Pack
   0.5 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-tbox
   0.8 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Image-Tools
   0.8 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\PuLID_ComfyUI
   0.8 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Manager
   0.8 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-InstantID
   1.9 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\SeargeSDXL
   2.5 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\was-node-suite-comfyui
   4.1 seconds: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\clipseg.py

Starting server

To see the GUI go to: http://127.0.0.1:8188
FETCH DATA from: D:\yapay zeka\ComfyUI-Zluda\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE]

Other

No response

patientx commented 2 weeks ago

this (tensorflow) seems to require cudnn, zluda doesn't support cudnn. There is some talk about working in linux with rocm (meaning no zluda at all) but I don't have enough experience with that.