Hello everyone. I am Dr. Furkan Gözükara. PhD Computer Engineer. SECourses is a dedicated YouTube channel for the following topics : Tech, AI, News, Science, Robotics, Singularity, ComfyUI, SwarmUI, ML, Artificial Intelligence, Humanoid Robots, Wan 2.2, FLUX, Krea, Qwen Image, VLMs, Stable Diffusion
What video card do you have now? I have AMD RX 6800XT 16gb and it handles Lora's training and stable diffusion quite well. The only inconvenience is that I had to install Ubuntu 20.04 on a second disk in order to have hardware acceleration
I use rx6600 and my operating system is Linux but I had to use it with docker because it was arch linux. When I installed stable diffusion directly to the system, it did not recognise the video card and worked with the processor. Somehow I managed to run stable diffusion with the video card using docker, but this time 512x512 20 step outputs took 60-90 seconds and after 4-5 outputs, the photos were distorted. Even when I tried with a proper seed and prompt, I kept getting corrupted outputs, and I was getting a 'cuda out of memory' error at the slightest setting change. I thought these speeds were normal, until I talked to someone using stable diffusion with a 3050 mobile graphics card who mentioned that he was getting outputs between 2-5 seconds. That's when I decided that I should definitely switch to NVIDIA.
Isn't there a decent, AI, melody generator that can be used offline and the music will have a beginning and an end? Online services have a bunch, and Wondershare already gives you one for Filmora, but no such thing for offline use? The Audiocraft/MusicGen/Magnet line is forgettable, a disaster what it generates.
GUYS! i need your help! does anyone know of a good site/tool/workflow that i can use to bring more detail and restore an old scanned photo?
I have my moms birthday tomorrow and this would mean the absolute world to me if someone can help me figure out a way to restore these old images to present her with for her birthday
I encountered the error below: "Downgrade the protobuf package to 3.20.x or lower." "Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python (but this will use pure-Python parsing and will be much slower)."
I deleted the Conda environment and added "protobuf <= 3.20.3" to the requirements.txt. After setting it up again, it works with this change.
However, I am not sure if this is the appropriate fix as I am not a programmer. Perhaps someone more versed in Python can make a pull request with the correct fix.
Been away for sometime. What is your favorite UI for loading Text gen models? I was using oogaba, but is there anything better to load models and switch between like automatic for stable diffusion?
I want to improve myself and move to new horizons in the field of artificial intelligence, which I am passionate about. For this process, I think Generative Adversarial Networks is the field that will develop and advance me. However, when I examined it, I came across an information polluted field with hundreds of projects based on ready-made models on the internet. In this process, I would like to ask you to provide a roadmap for me to improve myself in the field of Generative Adversarial Networks and tell me how I should proceed in this process.
In this process, although I have encountered projects on creating simple Gan models, unfortunately, I have not found a resource for building our own model other than detailed ready-made models within the scope of text to image. I would be very grateful if you can suggest research resources for text to image or help me draw a road map.