Comments

Log in with itch.io to leave a comment.

Viewing most recent comments 1 to 20 of 211 · Next page · Last page
(2 edits)

My video card is missing cuda divice, gpu 3070, only the cpu remains, sd gui 1.8.0

then error on startup (Detected no CUDA-capable GPUs.)

What version of stable diffusion does the Grisk Stable Diffusion GUI 0.1 run?

Does anyone know what is the default Sampling method in GRisk?

I was generating images today with this AI when I wanted to see the results it's just black pictures, when I came back here and read the "Important Stuff" section the first line that was written is about GTX 1660 cards not being able to render and unfortunately my card is a GTX 1660 Super so I guess i was unlucky.

When will you fix this ?

It's been fixed for a long time on the Patreon, the GUI on Itch is just way, WAY out of date.

What is their Patreon ?

https://www.patreon.com/DAINAPP

(1 edit) (-1)

can someone explain how to actually paint/mark areas with inpaint? i choose a input and i choose inpaint model, what am i missing after that? all it does is render a new image.

(-1)

for when amd?

(-2)

whenever i try to load a .ckpt model i get this error message

Traceback (most recent call last):

  File "start.py", line 1388, in OnRender

  File "start.py", line 1270, in LoadModel

  File "convert_original_stable_diffusion_to_diffusers.py", line 630, in Convert

  File "omegaconf\omegaconf.py", line 187, in load

FileNotFoundError: [Errno 2] No such file or directory: 'D:\\Downloads\\Stable Diffusion GRisk GUI\\v1-inference.yaml'

I keep getting out of memory errors even when it's only trying to alloc like 9MB. Makes no sense. No solutions online

I ran into an issue where I can run a prompt at 100, 200, 400 and 500 iterations, but 300 iterations gives an error:
Rendering: anime screenshot wide-shot landscape with house in the apple garden, beautiful ambiance, golden hour

0it [00:00, ?it/s]

Traceback (most recent call last):

  File "start.py", line 363, in OnRender

  File "torch\autograd\grad_mode.py", line 27, in decorate_context

    return func(*args, **kwargs)

  File "diffusers\pipelines\stable_diffusion\pipeline_stable_diffusion.py", line 152, in __call__

  File "diffusers\schedulers\scheduling_pndm.py", line 136, in step

  File "diffusers\schedulers\scheduling_pndm.py", line 212, in step_plms

  File "diffusers\schedulers\scheduling_pndm.py", line 230, in _get_prev_sample

IndexError: index 1000 is out of bounds for dimension 0 with size 1000

V Scale is 7.50, resolution is 768x512, seed: 12345, on a 3080 12GB.

(1 edit)

Why steps are limited to 500 (even on patreon 0.52 version (also not sure if it's connected in any way but DiscoDiffusion doesn't have a limit))

Processes three random images every time alongside my query. Pretty sus.

(+1)(-1)

are you sure you don't have any backspace in you text followed by nothing? Cause that creates random pics

Every time this has ever happened to me, it meant I had blank lines in the prompt, either before, or after what I actually typed.

(1 edit)

i paid 10 dollars and i am using version 0.52 i still get black screen and i bought but no money is deducted from my card

Patreon only charge you at the end of the mouth.

If you are having this problem, is a hardware bug, you can use the float32 option as a work around, but it will require more vram

thanks for the answer yes I enabled the float32 option and got this error RuntimeError: CUDA out of memory. Tried to allocate 114.00 MiB (GPU 0; 4.00 GiB total capacity; 2.99 GiB already allocated; 0 bytes free; 3.35 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF 

my graphics card is gtx 1650 super

Any news since a month ago?

float16 (rather than 32) helped?  max_split_size_mb:128?

(1 edit)

Really want to know why this program doesn't go out of memory (talking about system RAM, not VRAM) compared to every other SD repo/GUI. I have only 8GB of system RAM. (Thank you for making this!)

(+1)

Not sure why other GUI use that much ram, but your welcome

Dont know were to ask. Can i interface with this in any way with code. i want to try to make a private discord bot for me and my friends:)

(1 edit)

Will it be possible to use the DALL-E2 base in future versions? Thx.

Dalle-2 don't have the source code or the model as open source, so its impossible for now

It's a pity (

I'm having this error when loading a dreambooth model:

Loading model (May take a little time.)

{'feature_extractor'} was not found in config. Values will be initialized to default values.

Traceback (most recent call last):

  File "start.py", line 320, in OnRender

  File "start.py", line 284, in LoadModel

  File "diffusers\pipeline_utils.py", line 247, in from_pretrained

TypeError: __init__() missing 1 required positional argument: 'feature_extractor'

Dreambooth models are still not supported. Still need some time to make they work

No problem, with the ckpt converter I don't need this program anymore.

(1 edit)

Hello, I got this error when running a 512x512 image, I'm using an nvidia 3060 ti, and 16gb of ram, any solutions?

RuntimeError: CUDA out of memory. Tried to allocate 512.00 MiB (GPU 0; 8.00 GiB total capacity; 2.56 GiB already allocated; 2.69 GiB free; 2.58 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

I dont understand, 2,69 GiB is not enough to allocate 512 MiB? xD

Thanks in advance

It should totally not use that much memory for a 512X512 image. There is anything else using the vram of the computer?

(+1)

I got the patreon version, it works great! having tons of fun. Thanks for putting it together!

On my 1660 TI it just outputs black, and its not overclocked.

(1 edit) (+1)

The free version has a bug that breaks the AI when using 16xx Nvidia cards, it's fixed on the Patreon, annoyingly.

Definitely sets off my pet peeve about devs withholding important bugfixes behind a paywall, though GRisk plans to eventually update the free one, whenever that is.

This bug is more of a hardware problem than a software one. The fix is pretty much a  work around the hardware bug.

(+4)

Please look into adding Dreambooth please!

Yeah, it will be incorporated in the GUI eventually

(+1)

Is there a Github for this?  

Viewing most recent comments 1 to 20 of 211 · Next page · Last page