Simple Flux offers a lightweight web interface to Flux using Gradio.
Main UITested on Ubuntu with NVIVIA RTX 4090 and GTX 1070 (8GB GPU RAM).
If your GPU has less than 16GB GPU memory, you need to specify --low_mem when you start the script. Max GPU memory consumption that was observed when --low_mem was specified was 5.88GB, so 6GB GPU RAM may work.
Tested on 64GB RAM hosts. I learned from a user that he/she was able to run it with 40GB RAM on Windows 11. It is unclear if you can run it on a machine with less RAM than that.
Currently, only systems with NVIDIA GPU and CUDA are supported. I am using Linux and these steps have not been tested on Windows yet.
- Copy files from github
- Set up conda environment
- Install Pytorch
- Install other python packages
-
Copy files from github
Open the terminal and go to a directory that you want to install Simple Flux. Type:
git clone https://github.com/HowToSD/simple_flux.git
-
Set up conda environment
Run the following. Note that if your system does not have conda, you need to install it first.
conda create -n simpleflux python=3.10 conda activate simpleflux
-
Install Pytorch
conda install pytorch torchvision torchaudio pytorch-cuda=12.1 -c pytorch -c nvidia
For more information, refer to the installation section at https://pytorch.org/.
-
Install other python packages
pip install -r requirements.txt
-
Start Simple Flux Type:
python simple_flux.py
If you have low GPU memory, type:
python simple_flux.py --low_mem
This should start Gradio.
-
Using your browser, access the URL displayed in the above.