I never tried, there are some packages which let's you run AI models on AMD. But it involves technal knowledge it's a bit headache bcoz you will get many errors. I just saw a video about it. You can also use it online if you search for magicquill hugging space, you will get the running application with limited usage. You have to pay a few dollars to have it properly.
ERROR: Could not find a version that satisfies the requirement torch==2.1.2 (from versions: 2.2.0+cu121, 2.2.1+cu121, 2.2.2+cu121, 2.3.0+cu121, 2.3.1+cu121, 2.4.0+cu121, 2.4.1+cu121, 2.5.0+cu121, 2.5.1+cu121) ERROR: No matching distribution found for torch==2.1.2
Hi, I'm facing an issue with a runtime error related to tensor size mismatch in Python. Can someone help? The error message says the sizes of tensor a (640) and tensor b (320) must match at dimension 1. "
@@Aininja94 I've reinstalled it three times, following some suggestions, i found regarding similar issues on GitHub, but I still haven't been able to fix it. There's a discussion there about the same problem, but so far, I haven't found a definitive solution.
@@Aininja94 I've tried reinstalling it three times, following some of the issues reported on GitHub, but none of the solutions have worked so far. There's a thread discussing the same problem, but no one seems to have found a fix yet.
i get this erro F:\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py Traceback (most recent call last): File "F:\MagicQuill\gradio_run.py", line 16, in from MagicQuill.llava_new import LLaVAModel File "F:\MagicQuill\MagicQuill\llava_new.py", line 2, in from transformers import TextStreamer ModuleNotFoundError: No module named 'transformers'
@@Aininja94 yes look F:\MagicQuill>conda create -n env python==3.10 WARNING: A conda environment already exists at 'C:\Users\Júnior3D\anaconda3\envs\env' Remove existing environment? This will remove ALL directories contained within this specified prefix directory, including any other conda environments. (y/[n])?
I think the first time when you ran this command, python was not included hence pip is not recognised in the env. Try creat a different env Or press y for yes remove the existing environment and have a new environment with python
E:\Data\Packages\ComfyUI\custom_nodes\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py Traceback (most recent call last): File "E:\Data\Packages\ComfyUI\custom_nodes\MagicQuill\gradio_run.py", line 3, in import gradio as gr ModuleNotFoundError: No module named 'gradio' I'm so tired of making mistakes(
@@Aininja94 you have not installed the requirements.txt file. --- installed it((( This gradio error will be solved by 'pip install gradio' --- installed it (((
@@Aininja94 I went through a small circle of hell, starting with the fact that I couldn't log in to (env ) - before reinstalling cuda version 11.8, and even a mountain of small errors, but with the help of gpt4 I managed it)))
Well, I have idea about dockerr and the orchestration stuff but never used any. Setting up this is a challenging as there are errors posted in the comments 😖
I'm getting this fault, can you give any help? (env) C:\Users\Dan\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py Traceback (most recent call last): File "C:\Users\Dan\MagicQuill\gradio_run.py", line 16, in from MagicQuill.llava_new import LLaVAModel File "C:\Users\Dan\MagicQuill\MagicQuill\llava_new.py", line 10, in from llava.conversation import conv_templates, SeparatorStyle ModuleNotFoundError: No module named 'llava'
@@Aininja94 Thanks, I have done this and I'm not getting this error anymore. I am now getting the following error: (base) c:\Users\Dan\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py Traceback (most recent call last): File "c:\Users\Dan\MagicQuill\gradio_run.py", line 5, in from gradio_magicquill import MagicQuill ModuleNotFoundError: No module named 'gradio_magicquill'
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
got it tq for the great tutorail
I am getting this error on running,
expected an indented block after 'with' statement on line 197
You can't miss the space in the left as shown in the video.
Press 'tab' to leave proper space as shown in the video.
Is there any workout for amd gpu?
I never tried, there are some packages which let's you run AI models on AMD. But it involves technal knowledge it's a bit headache bcoz you will get many errors.
I just saw a video about it.
You can also use it online if you search for magicquill hugging space, you will get the running application with limited usage. You have to pay a few dollars to have it properly.
@@Aininja94 Can you give me the link for the video you saw
ERROR: Could not find a version that satisfies the requirement torch==2.1.2 (from versions: 2.2.0+cu121, 2.2.1+cu121, 2.2.2+cu121, 2.3.0+cu121, 2.3.1+cu121, 2.4.0+cu121, 2.4.1+cu121, 2.5.0+cu121, 2.5.1+cu121)
ERROR: No matching distribution found for torch==2.1.2
Remove the torch from requirements.txt file. And install the torch from the command I have given in the description.
yup did that and thats what came up
everything else worked fine untill i added the torch
is rtx3050ti 4gb vram enough?
No but you can try, one of my viewer ran flux model on 4gb vram. It took 20 min to generate a image but it worked
Hi, I'm facing an issue with a runtime error related to tensor size mismatch in Python. Can someone help? The error message says the sizes of tensor a (640) and tensor b (320) must match at dimension 1. "
There is some issue with the torch cuda installed, did you installed it before the requirements.txt installation
@@Aininja94 Yes, I was looking at github and there are other people with the same problem but I didn't find any solution
@@Aininja94 I've reinstalled it three times, following some suggestions, i found regarding similar issues on GitHub, but I still haven't been able to fix it. There's a discussion there about the same problem, but so far, I haven't found a definitive solution.
@@Aininja94 I've tried reinstalling it three times, following some of the issues reported on GitHub, but none of the solutions have worked so far. There's a thread discussing the same problem, but no one seems to have found a fix yet.
i get this erro
F:\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py
Traceback (most recent call last):
File "F:\MagicQuill\gradio_run.py", line 16, in
from MagicQuill.llava_new import LLaVAModel
File "F:\MagicQuill\MagicQuill\llava_new.py", line 2, in
from transformers import TextStreamer
ModuleNotFoundError: No module named 'transformers'
When creating Conda environment, did you added python?
conda create -n env python==3.10
Like this.
@@Aininja94 yes look
F:\MagicQuill>conda create -n env python==3.10
WARNING: A conda environment already exists at 'C:\Users\Júnior3D\anaconda3\envs\env'
Remove existing environment?
This will remove ALL directories contained within this specified prefix directory, including any other conda environments.
(y/[n])?
I think the first time when you ran this command, python was not included hence pip is not recognised in the env.
Try creat a different env Or press y for yes remove the existing environment and have a new environment with python
@@Aininja94 I'll try to redo everything from scratch again to see if it's ok
what's your name on discord
nice and open source
Yes, for someone who prefer simple interface.
Usable with 4070 TI SUPER 16 VRAM or too slow?
That's fast then what you saw in video
Nice, can we have the same with more modern technology, Flux for example ?
Some supporting modules like ip adapters and control net are not that great in flux yet.
I found that my RGB brush is not active (((
the color palette not there? What browser you are using
thanks for the tip, changed the browser to chrome and the brush became active)
E:\Data\Packages\ComfyUI\custom_nodes\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py
Traceback (most recent call last):
File "E:\Data\Packages\ComfyUI\custom_nodes\MagicQuill\gradio_run.py", line 3, in
import gradio as gr
ModuleNotFoundError: No module named 'gradio'
I'm so tired of making mistakes(
you have not installed the requirements.txt file.
This gradio error will be solved by 'pip install gradio'
@@Aininja94
you have not installed the requirements.txt file. --- installed it(((
This gradio error will be solved by 'pip install gradio' --- installed it (((
If you followed the video but still the error, then you must not be in a active conda environment.
@Thank you, I’ll track the error with conda
@@Aininja94 I went through a small circle of hell, starting with the fact that I couldn't log in to (env ) - before reinstalling cuda version 11.8, and even a mountain of small errors, but with the help of gpt4 I managed it)))
tried to run it on colab but didnt work, gonna try again
I didn't try on colab, should be working on colab as well, just follow the steps properly.
any dockerized version ?
No docker file was there
@Aininja94 without docker this a nightmare especially for windows users
Well, I have idea about dockerr and the orchestration stuff but never used any.
Setting up this is a challenging as there are errors posted in the comments 😖
how do I save a photo?
Click the arrow bedside the three brush. Then right click on the image
@@Aininja94 no load,save after editing
With arrow and right click it gives an option to save.
If you are having trouble then take a screen shot using snipping tool
can you make one click installer for this please?
The application is still in development, maybe ones it's done
@@Aininja94 Yay! \o/
I'm getting this fault, can you give any help?
(env) C:\Users\Dan\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py
Traceback (most recent call last):
File "C:\Users\Dan\MagicQuill\gradio_run.py", line 16, in
from MagicQuill.llava_new import LLaVAModel
File "C:\Users\Dan\MagicQuill\MagicQuill\llava_new.py", line 10, in
from llava.conversation import conv_templates, SeparatorStyle
ModuleNotFoundError: No module named 'llava'
I have already tried using the command pip install -e
The command is pip install -e MagivQuill/LLaVA/
Please check the description and copy from there
@@Aininja94 Thanks, I have done this and I'm not getting this error anymore. I am now getting the following error:
(base) c:\Users\Dan\MagicQuill>set CUDA_VISIBLE_DEVICES=0 && python gradio_run.py
Traceback (most recent call last):
File "c:\Users\Dan\MagicQuill\gradio_run.py", line 5, in
from gradio_magicquill import MagicQuill
ModuleNotFoundError: No module named 'gradio_magicquill'
Дякую! Супер