24-02-2024

Today, my efforts were focused on the following:

• I spent several hours trying to utilize comfyui on my PC. During this process, I installed various checkpoints, vaes, controlnets, loras etc. Honestly, it's quite overwhelming as there are too many elements to remember and find their usefulness.
• I attempted to transform a portrait image into an art piece, ensuring the person's identity remains intact.
• I made a significant stride towards my goal. Using control net, I managed to create a peculiar cat image from a scribble that can be equated to a preschool level. This is a great achievement for me as it shows I'm gaining control over the image generation process.

I dont really care about the quality of the cat image, but what i was acheived with this is having control over the image i generate, here the cat could have been generated in any pose, but it uses my scribble as a guide to generate a cat which looks similar to that. yay! what an achivement!

another example

Now I came to know about SDXL

Which is Stable Diffusion Latest Ai model which is very popular recently

So far the images I generated were using sd1.5 the popular one before SDXL

and some other community models downloaded from internet.