[New Extension] Stable Horde for Web UI

[natanjunges]

Stable Horde client for AUTOMATIC1111's Stable Diffusion Web UI

Link: https://github.com/natanjunges/stable-diffusion-webui-stable-horde

Addresses #1449.

[system1system2]

How come this is not yet in the list of A1111 extensions you can load from the UI?

[natanjunges]

I surely hope one of the maintainers adds it soon.

[ClashSAN]

I could not run it, I set --ui-debug-mode and went down to scripts to after putting a proper api key.
stable horde client version (latest) - https://github.com/natanjunges/stable-diffusion-webui-stable-horde/commit/cfae14f03bf7e6fb112be459a3758a4cf52aa4da
webui - 9cfd10c

[natanjunges]

Did you get any error messages in the terminal?

[ClashSAN]

nvm, this works now, i couldn't really replicate my issue. I ran with the same versions - webui - 9cfd10c
I've also run without --ui-debug-mode, with the other worker extension (enable checkbox ticked off) without problems.
(it could take 2-3 min if you are low rank)

[mile6919]

I couldn't run this. No matter what I did, nothing worked. I set --ui-debug-mode too, didn't help. I had "own" API key and all, but to me it seems this is not working at all.

[natanjunges]

Did you get any error messages in the terminal?

[Frozander]

Seems cool. I got it to work. I am assuming the error is caused because you enabled "Only send requests to trusted workers" on by default and the trusted users textbox is empty. I disabled both checkboxes and sent a prompt. Seems fine for now.

However I am getting errors on the console. Don't know the cause right now but doesn't seem like it is triggered from the extension itself.

Edit: Ignore my initial response. Now I get the error when I tried a much longer prompt.
Edit 2: I get the issue. I don't think this extension properly handles errors and interrupts coming from stable horde. In my case, after investigation I found that I didn't have enough Kudos. This caused exceptions to be thrown since main.py L412 returns a null when an interrupt happens on stable horde while the caller at L287 tries to unpack it.

[natanjunges]

• "Only send requests to trusted workers" and "Only send requests to these workers" are two separate settings. The second, when present, overrides the first. Keeping "Only send requests to trusted workers" enabled is highly recommended.
• The error AttributeError: 'NoneType' object has no attribute 'process_texts' is caused by --ui-debug-mode and can be safely ignored.
• Although there's no nice error message to the end user, raising an exception is the intended behavior, since the generation failed. I'll try to improve on that to make the error more explicit.

[natanjunges]

https://github.com/natanjunges/stable-diffusion-webui-stable-horde/commit/99e4fb87c203f626a6a587144fcb38de6aa2b511 should improve the error handling.

[mile6919]

Still it is not working for me:

Traceback (most recent call last):
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\routes.py", line 321, in run_predict
output = await app.blocks.process_api(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 1015, in process_api
result = await self.call_function(fn_index, inputs, iterator, request)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 856, in call_function
prediction = await anyio.to_thread.run_sync(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 867, in run
result = context.run(func, *args)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\call_queue.py", line 15, in f
res = func(*args, **kwargs)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\ui.py", line 312, in update_token_counter
token_count, max_length = max([model_hijack.get_prompt_lengths(prompt) for prompt in prompts], key=lambda args: args[0])
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\ui.py", line 312, in
token_count, max_length = max([model_hijack.get_prompt_lengths(prompt) for prompt in prompts], key=lambda args: args[0])
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\sd_hijack.py", line 152, in get_prompt_lengths
_, token_count = self.clip.process_texts([text])
AttributeError: 'NoneType' object has no attribute 'process_texts'
Traceback (most recent call last):
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\routes.py", line 321, in run_predict
output = await app.blocks.process_api(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 1015, in process_api
result = await self.call_function(fn_index, inputs, iterator, request)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 856, in call_function
prediction = await anyio.to_thread.run_sync(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 867, in run
result = context.run(func, *args)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\call_queue.py", line 15, in f
res = func(*args, **kwargs)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\ui.py", line 312, in update_token_counter
token_count, max_length = max([model_hijack.get_prompt_lengths(prompt) for prompt in prompts], key=lambda args: args[0])
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\ui.py", line 312, in
token_count, max_length = max([model_hijack.get_prompt_lengths(prompt) for prompt in prompts], key=lambda args: args[0])
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\sd_hijack.py", line 152, in get_prompt_lengths
_, token_count = self.clip.process_texts([text])
AttributeError: 'NoneType' object has no attribute 'process_texts'
Error completing request
Arguments: ("Masterpiece, best quality, (realistic), award winning portrait photo of (a stunningly beautiful topless woman:1.2) in a Victorian-style dress with steampunk elements, (Victorian-era streets of London in background:1.2), a long flowing dress in a dark color such as black or navy with brass or copper buttons or gears incorporated into the design, she is accessorized with a pair of gloves and a corset with mechanical accents, the woman wear's (a top hat with clockwork detailing:1.2), (beautiful blue_eyes:1.1), makeup is inspired by the steampunk aesthetic such as dark smoky eyes, (medium sized breasts:1.1), (head fully in frame:1.3), (perfect hands:1.1), sharp focus, 8k", 'red_eyes, vignette, cross_eyed, frame, border, out of frame, head out of frame, 3d, cartoon, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name', 'None', 'None', 50, 15, False, False, 1, 1, 11.5, -1.0, -1.0, 0, 0, 0, False, 768, 512, False, 0.7, 2, 'Latent', 0, 0, 0, 5, True, False, 1, False, False, False, 1.1, 1.5, 100, 0.7, False, False, False, False, False, False, 'LoRA', 'None', 1, 'LoRA', 'None', 1, 'LoRA', 'None', 1, 'LoRA', 'None', 1, 'LoRA', 'None', 1, 'Refresh models', 75, 0.7, False, True, False, True, False, False, False, False, False, '', 1, '', 0, '', True, True, False, 'Dreamlike Photoreal (4)', False, False, 1, 'None', 'None', 'None') {}
Traceback (most recent call last):
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\call_queue.py", line 45, in f
res = list(func(*args, **kwargs))
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\call_queue.py", line 28, in f
res = func(*args, **kwargs)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\txt2img.py", line 49, in txt2img
processed = modules.scripts.scripts_txt2img.run(p, *args)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\scripts.py", line 337, in run
processed = script.run(p, *script_args)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\extensions\stable-horde\scripts\main.py", line 181, in run
return self.process_images(p, model, nsfw, shared_laion, int(seed_variation), post_processing)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\extensions\stable-horde\scripts\main.py", line 202, in process_images
res = self.process_images_inner(p, model, nsfw, shared_laion, seed_variation, post_processing)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\extensions\stable-horde\scripts\main.py", line 253, in process_images_inner
file.write(processed.infotext(p, 0))
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\processing.py", line 325, in infotext
return create_infotext(p, self.all_prompts, self.all_seeds, self.all_subseeds, comments=[], position_in_batch=index % self.batch_size, iteration=index // self.batch_size)
File "C:\Users\leinm\Automatic Stable Diffusion\stable-diffusion-webui\modules\processing.py", line 438, in create_infotext
"Model": (None if not opts.add_model_name_to_info or not shared.sd_model.sd_checkpoint_info.model_name else shared.sd_model.sd_checkpoint_info.model_name.replace(',', '').replace(':', '')),
AttributeError: type object 'FakeModel' has no attribute 'sd_checkpoint_info'

[natanjunges]

Are you using your own API key? Do you have enough kudos?

[mile6919]

Yes, I am using own API key. What are kudos and where do I get them?

[natanjunges]

https://github.com/db0/AI-Horde/blob/main/FAQ.md#kudos

[dav30383]

@natanjunges you really did a great job! I wanted to ask, but would it be possible to use the "engine" of Stable Horder, but use your own models locally? Thanx :-D

[natanjunges]

What do you mean by that? To be a horde worker? If so, this extension is not meant for that, but I believe there's another one specifically for it.

[dav30383]

No a worker, But using the models that I have locally and are not there on Stable Horder (for example Level4,Deliberate2,etc), do you understand what I mean? Il sab 3 giu 2023, 19:09 Natan Junges ***@***.***> ha scritto:

…

What do you mean by that? To be a horde worker? If so, this extension is not meant for that, but I believe there's another one specifically for it. — Reply to this email directly, view it on GitHub <#6157 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVAU6NVP3AEDMUPIOB5L7CTXJNVWJANCNFSM6AAAAAATNXPFLY> . You are receiving this because you are subscribed to this thread.Message ID: <AUTOMATIC1111/stable-diffusion-webui/repo-discussions/6157/comments/6079661 @github.com>

[natanjunges]

Well, for that you don't need this extension, because that's what the Web UI does by default.

[dav30383]

yes, but since I don't have a powerful video card, I was wondering if it was possible to use your script as an "engine", but the models I have locally, understand what I mean? Il giorno sab 3 giu 2023 alle ore 20:29 Natan Junges < ***@***.***> ha scritto:

…

Well, for that you don't need this extension, because that's what the Web UI does by default. — Reply to this email directly, view it on GitHub <#6157 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVAU6NURSNNJ25YMYFRI4U3XJN67XANCNFSM6AAAAAATNXPFLY> . You are receiving this because you are subscribed to this thread.Message ID: <AUTOMATIC1111/stable-diffusion-webui/repo-discussions/6157/comments/6079969 @github.com>

[natanjunges]

That's just not possible, this extension doesn't do magic. Either you run the models locally, which requires having a powerful enough GPU/TPU/CPU, or you run them remotely.

[dav30383]

ah ok, but I don't understand why they don't update the models on Stable Horder. Anyway, great script, bravo! Greetings from Italy [image: Icona di Verificata con community] Il giorno sab 3 giu 2023 alle ore 20:50 Natan Junges < ***@***.***> ha scritto:

…

That's just not possible, this extension doesn't do magic. Either you run the models locally, which requires having a powerful enough GPU/TPU/CPU, or you run them remotely. — Reply to this email directly, view it on GitHub <#6157 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVAU6NWZ37QFWYSFT6MWCS3XJOBPBANCNFSM6AAAAAATNXPFLY> . You are receiving this because you are subscribed to this thread.Message ID: <AUTOMATIC1111/stable-diffusion-webui/repo-discussions/6157/comments/6080043 @github.com>

[AnyISalIn]

Great works, inspired by this, we have created a similar project: https://github.com/omniinfer/sd-webui-cloud-inference