capsize-games / airunner Goto Github PK
View Code? Open in Web Editor NEWStable Diffusion and LLMs offline on your own hardware
Home Page: https://capsizegames.itch.io/ai-runner
License: GNU General Public License v3.0
Stable Diffusion and LLMs offline on your own hardware
Home Page: https://capsizegames.itch.io/ai-runner
License: GNU General Public License v3.0
Add the files required to build for windows on docker in the cloud.
These files reside in a separate repo and must be combined with airunner just as I did with the Ubuntu build files.
The green layer color is ugly. Additionally the text is hard to see when the layer is selected.
Brush size spinbox doesn't change with slider.
Layers should not have large gabs between them. As more layers are added the gaps close. Potential solution is using a vertical spacer in the template but since the layers are in a scroll frame that might not work.
Describe the bug
I downloaded the latest version of airunner as a zip file from https://capsizegames.itch.io/ai-runner extracted it, executed airunner and it crashed with:
[18042] PyInstaller Bootloader 5.x
[18042] LOADER: executable is /home/sam/Downloads/ai-runner-ubuntu/airunner
[18042] LOADER: homepath is /home/sam/Downloads/ai-runner-ubuntu
[18042] LOADER: _MEIPASS2 is NULL
[18042] LOADER: archivename is /home/sam/Downloads/ai-runner-ubuntu/airunner
[18042] LOADER: Cookie found at offset 0x29E9FC0
[18042] LOADER: No need to extract files to run; setting up environment and restarting bootloader...
[18042] LOADER: LD_LIBRARY_PATH=/home/sam/Downloads/ai-runner-ubuntu
[18042] PyInstaller Bootloader 5.x
[18042] LOADER: executable is /home/sam/Downloads/ai-runner-ubuntu/airunner
[18042] LOADER: homepath is /home/sam/Downloads/ai-runner-ubuntu
[18042] LOADER: _MEIPASS2 is /home/sam/Downloads/ai-runner-ubuntu
[18042] LOADER: archivename is /home/sam/Downloads/ai-runner-ubuntu/airunner
[18042] LOADER: Cookie found at offset 0x29E9FC0
[18042] LOADER: Already in the child - running user's code.
[18042] LOADER: Python library: /home/sam/Downloads/ai-runner-ubuntu/libpython3.10.so.1.0
[18042] Error loading Python lib '/home/sam/Downloads/ai-runner-ubuntu/libpython3.10.so.1.0': dlopen: /lib/x86_64-linux-gnu/libm.so.6: version `GLIBC_2.35' not found (required by /home/sam/Downloads/ai-runner-ubuntu/libpython3.10.so.1.0)
To Reproduce
Steps to reproduce the behavior:
Download ai-runner-ubuntu.zip (Version 1.8.17).
Unzip: unzip -d ai-runner-ubuntu ai-runner-ubuntu.zip
Execute: cd ai-runner-ubuntu && ./airunner
Expected behavior
I guess I expected the app to open!
Desktop (please complete the following information):
Additional context
I'll debug a bit, see if I can patch it up. Otherwise I'll try from source.
Great work @w4ffl35 , though!
For some reason it is attempting to push the build via the dist folder which it claims does not exist.
https://github.com/Capsize-Games/airunner/actions/runs/4590022143/jobs/8105299063
add ability to set the following directories
This is set in stunner in the aihandler repo.
error encountered when exporting to jpg format:
OSError: cannot write mode RGBA as JPEG
convert image to RGB prior to exporting JPEG.
Currently the build process is cumbersome with two docker images (one for each os supported). No artifact is produced from the build (build script pushes to docker after building).
Ideally we would have 2 stages
Currently the github action fails.
It may be that the artifact is too large and we must simply push to docker from within the image as was the case with v1.8.21
Thus far all efforts have resulted in a broken build. Next step is to
This is similar to the previous setup, the difference being that we have a single docker image.
If windows continues to time out we may need to produce two docker containers after all.
Prior to adding the docker build workflow to github actions yesterday I had a pypi build action which would checkout the latest code, build it and push to pypi. I deleted it because the docker image required lots of releasing and deleting of the same version. Rather than constantly trying to push to pypi during this, I deleted the action file with the intent to restore it after getting docker finalized.
Describe the bug
I downloaded the Stable Diffusion 1.4 model and put it in a completely separate directory to AI runner. Did a couple of generations, then installed the Lora extension.
Next time I tried to start AI runner, it would run through the startup script, but die before the program launched. I was not able to capture the console log when it was dying, but I did see an orange square drawn around a segment of code before it died.
After multiple attempts at removing the AI Runner directory, and manually cleaning up any files I could fine under C:\User, I finally managed to get AI runner started again.
I am running on Windows 10, running AI Runner from one of my secondary NVME drives.
Add some unit tests to get the ball rolling
Describe the bug
A clear and concise description of what the bug is.
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Build and deploy
Desktop (please complete the following information):
Github only
Additional context
https://github.com/Capsize-Games/airunner/actions/runs/4584490627/jobs/8096005434
Describe the bug
When trying to run from source (in a docker environment with ubuntu 22.04, which I'll share shortly) seems like aihandler is missing to declare a dependency on diffusers as I get the error:
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/aihandler/runner.py", line 1162, in generator_sample
self._generate(data, image_var=image_var)
File "/usr/local/lib/python3.10/dist-packages/aihandler/runner.py", line 1077, in _generate
self._initialize()
File "/usr/local/lib/python3.10/dist-packages/aihandler/runner.py", line 636, in _initialize
self._load_model()
File "/usr/local/lib/python3.10/dist-packages/aihandler/runner.py", line 486, in _load_model
"scheduler": self.scheduler,
File "/usr/local/lib/python3.10/dist-packages/aihandler/runner.py", line 166, in scheduler
import diffusers
ModuleNotFoundError: No module named 'diffusers'
To Reproduce
TODO
Expected behavior
Use the app normally :)
Screenshots
If applicable, add screenshots to help explain your problem.
Desktop (please complete the following information):
steps to reproduce
Error caught was: 'NoneType' object has no attribute 'origin' - 51
╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ in paintEvent:152 │
│ │
│ in draw:62 │
│ │
│ in current_layer:33 │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
IndexError: list index out of range
QBackingStore::endPaint() called with active painter; did you forget to destroy it or call QPainter::end() on it?
make eraser work with rasterized lines
Add support for loras
Add the ability for developers to create extensions which can be installed with airunner.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.