AUTOMATIC1111 / stable-diffusion-webui

Stable Diffusion web UI
GNU Affero General Public License v3.0
140.98k stars 26.67k forks source link

[Bug]: img2img generate button doesn't work most of the time #13969

Open GalenMarek14 opened 11 months ago

GalenMarek14 commented 11 months ago

Is there an existing issue for this?

What happened?

When using resize methods in img2img and clicking generate it shows "waiting" normally but after that is totally random. It sometimes generates the output successfully while sometimes waits and does nothing. Most of the time it doesn't work. All command windows are empty while this happens (I tried normal install and Stability Matrix both cmd and Stability command windows are empty). "--loglevel DEBUG" also shows nothing either. I tried many different browsers (chrome, firefox, edge, brave) but nothing changed. Clearing caches didn't do anything either. Fresh installs without any extensions did nothing also.

I also tried "--no-gradio-queue" but then WebUI didn't even start and gave "ValueError: Progress tracking requires queuing to be enabled." error.

For some reason vladmandic's fork SD.Next works perfectly though (UI-UX doesn't work either)...

Steps to reproduce the problem

  1. Go to img2img tab
  2. Press Generate
  3. "Waiting" appears but most of the time nothing happens

What should have happened?

It should have generated a resized image.

Sysinfo

sysinfo-2023-11-13-22-31.txt

What browsers do you use to access the UI ?

Mozilla Firefox, Google Chrome, Brave, Microsoft Edge

Console logs

Python 3.10.11 (tags/v3.10.11:7d4cc5a, Apr  5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)]
Version: v1.6.0-2-g4afaaf8a
Commit hash: 4afaaf8a020c1df457bcf7250cb1c7f609699fa7
Installing pyqt5 requirement for depthmap script
loading WD14-tagger reqs from D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\extensions\stable-diffusion-webui-wd14-tagger\requirements.txt
Checking WD14-tagger requirements.
Launching Web UI with arguments: --medvram --xformers --api --skip-python-version-check --no-gradio-queue
Civitai Helper: Get Custom Model Folder
Tag Autocomplete: Could not locate model-keyword extension, Lora trigger word completion will be limited to those added through the extra networks menu.
[-] ADetailer initialized. version: 23.11.0, num models: 9
CivitAI Browser+: Aria2 RPC started
Using sqlite file: D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\extensions\sd-webui-agent-scheduler\task_scheduler.sqlite3
2023-11-13 23:57:16,080 - ControlNet - INFO - ControlNet v1.1.416
ControlNet preprocessor location: D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\extensions\sd-webui-controlnet\annotator\downloads
2023-11-13 23:57:16,298 - ControlNet - INFO - ControlNet v1.1.416
== WD14 tagger /gpu:0, uname_result(system='Windows', node='LAPTOP-50SHSSKH', release='10', version='10.0.19045', machine='AMD64') ==
Loading weights [463d6a9fe8] from D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\models\Stable-diffusion\absolutereality_v181.safetensors
WARNING:py.warnings:D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\extensions\PBRemTools\scripts\main.py:47: GradioDeprecationWarning: The `style` method is deprecated. Please set these arguments in the constructor instead.
  gallery = gr.Gallery(label="outputs", show_label=True, elem_id="gallery").style(grid=2, object_fit="contain")

WARNING:py.warnings:D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\extensions\PBRemTools\scripts\main.py:47: GradioDeprecationWarning: The 'grid' parameter will be deprecated. Please use 'columns' in the constructor instead.
  gallery = gr.Gallery(label="outputs", show_label=True, elem_id="gallery").style(grid=2, object_fit="contain")

"
            SadTalker will not support download all the files from hugging face, which will take a long time.

            please manually set the SADTALKER_CHECKPOINTS in `webui_user.bat`(windows) or `webui_user.sh`(linux)

Creating model from config: D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\configs\v1-inference.yaml
Civitai Shortcut: v1.6.7
Civitai Shortcut: There are no registered shortcuts.
Civitai Shortcut: There are no registered shortcuts.
Civitai Shortcut: There are no registered shortcuts.
Civitai Shortcut: There are no registered shortcuts.
Traceback (most recent call last):
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\launch.py", line 48, in <module>
    main()
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\launch.py", line 44, in main
    start()
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\modules\launch_utils.py", line 436, in start
    webui.webui()
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\webui.py", line 79, in webui
    app, local_url, share_url = shared.demo.launch(
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\venv\lib\site-packages\gradio\blocks.py", line 1858, in launch
    self.validate_queue_settings()
  File "D:\AIStabiltyMatrixA1111\StabilityMatrix\Data\Packages\StableDiffusionWebUIExtensions\venv\lib\site-packages\gradio\blocks.py", line 1690, in validate_queue_settings
    raise ValueError("Progress tracking requires queuing to be enabled.")
ValueError: Progress tracking requires queuing to be enabled.
Applying attention optimization: xformers... done.
Model loaded in 7.7s (load weights from disk: 1.2s, create model: 0.6s, apply weights to model: 4.8s, calculate empty prompt: 0.9s).

Process finished with exit code 1

Additional information

image image Those logs are taken while the generate button working and after it is 'finished'. Nothing shows.

ExissNA commented 10 months ago

I also tried "--no-gradio-queue" but then WebUI didn't even start and gave "ValueError: Progress tracking requires queuing to be enabled." error.

For some reason vladmandic's fork SD.Next works perfectly though (UI-UX doesn't work either)...

Same. I have been having issues with the DOS window showing completion at - for example - 2 minutes 37 seconds, but the UI hanging at 95-97% and a timer going from 5s remaining occasionally increasing more and more, only for the final generation time to be anywhere between 5-8+ minutes.

I tried --no-gradio-queue but it always errors "progress tracking requires queuing to be enabled." in "...gradio\blocks.py" lines 1690

File "H:\stable-diffusion\stable-diffusion-main\AUTO1111\system\python\lib\site-packages\gradio\blocks.py", line 1690, in validate_queue_settings raise ValueError("Progress tracking requires queuing to be enabled.") ValueError: Progress tracking requires queuing to be enabled. Press any key to continue...

Pressing any key does not continue. :)

I also tried installing UI-UX as a potential workaround but it's not loading either. I have not tried vladmandic's fork. Half the time, generations load super fast, less than 45 seconds for a relatively high resolution. Resource monitoring shows at least 2GB free with no errors. Sometimes I get a websocket disconnect after minutes of wait time before the generation finally finishes less than a second after the error comes up and reconnects.
My custom args are as follows (excluding pathing to ComfyUI checkpoint/embedding/lora folders on the same drive) --xformers --opt-split-attention --upcast-sampling --medvram --no-half-vae --opt-channelslast set PYTORCH_CUDA_ALLOC_CONF=garbage_collection_threshold:0.9,max_split_size_mb:512 (this one REALLY helped performance a lot BTW, I wouldn't have to reload everything after just 3-5 gens). Version version: v1.6.0-2-g4afaaf8a as part of the troubleshooting process, newest as of 11/22/23.

Edit Also, I forgot to mention I ran the gamut of disabling extensions, none of which seemed to make a difference. Also, to clarify, it's not the img2img generate button I'm having issues with, though it does occasionally happen, sometimes it's just a long pause before it does respond.