GeorgLegato / Txt2Vectorgraphics

Custom Script for Automatics1111 StableDiffusion-WebUI.
GNU General Public License v3.0
362 stars 29 forks source link

Cannot get the script to work, fails on saving the image #15

Closed pixelicous closed 1 year ago

pixelicous commented 1 year ago

Downloaded the script, installed python3-tk and potrace

I am using AUTOMATIC1111's webui and when I run i see the image gets generated but in the last steps it fails and drops the image from the UI with an error -

Exception: TXT2Vectorgraphics: Execution of Potrace failed, check filesystem, permissions, installation or settings (is image saving on?) Time taken: 4.27sTorch active/reserved: 3161/3678 MiB, Sys VRAM: 4713/14961 MiB (31.5%)

In the shell i get -

File "/home/ubuntu/stable-diffusion-webui/scripts/txt2vectorgfx.py", line 127, in run fullofTPNG = fullfnPath.with_stem(fullfnPath.stem+ "_T") AttributeError: 'PosixPath' object has no attribute 'with_stem'

I have image saving turned on and if i don't use the script they get successfully saved to - /stable-diffusion-webui/outputs

Having coding knowledge myself i tried to retrace what's going on and i could understand why it couldn't enumerate the photos and get their path, as they are saved on the filesystem..

any idea what configuration or where should i look to fix this

GeorgLegato commented 1 year ago

check your python version ( which is used in the auto1111 environment) it should be 3.10.

referring to PosixPath' object has no attribute 'with_stem' it depends on some particular version .

try to add "python --version" in your webui-user.sh to see which version is really used in the auto1111 environment

Downloaded the script, installed python3-tk and potrace

I am using AUTOMATIC1111's webui and when I run i see the image gets generated but in the last steps it fails and drops the image from the UI with an error -

Exception: TXT2Vectorgraphics: Execution of Potrace failed, check filesystem, permissions, installation or settings (is image saving on?)

Time taken: 4.27sTorch active/reserved: 3161/3678 MiB, Sys VRAM: 4713/14961 MiB (31.5%)

In the shell i get -

File "/home/ubuntu/stable-diffusion-webui/scripts/txt2vectorgfx.py", line 127, in run

fullofTPNG = fullfnPath.with_stem(fullfnPath.stem+ "_T")

AttributeError: 'PosixPath' object has no attribute 'with_stem'

I have image saving turned on and if i don't use the script they get successfully saved to -

/stable-diffusion-webui/outputs

Having coding knowledge myself i tried to retrace what's going on and i could understand why it couldn't enumerate the photos and get their path, as they are saved on the filesystem..

any idea what configuration or where should i look to fix this

pixelicous commented 1 year ago

@GeorgLegato Thank you, I actually figured the "object has no attribute 'with_stem'" is received because no object exists in the variable..

Anyhow I arrived to the conclusion I might have a python issue just this moment, I tried installing invokeAI and I noticed in their Docs it doesn't support my python version, then i went to stable diffusion and it happens I'm not on their supported version as well 🤦 Then i stumbled here to see if there is a response, and yeah, same problem 🤣 Anyhow, I'm running off an AWS machine, so I'm recreating it from a different AMI and i'm good to go.

Thanks again