Forum Replies Created
-
Posts
-
It is challenging to generate words with SD. You can use the SD3 API or controlnet.
The credit API is not worth it if you need to generate a large number of images. The NSFW filter is overly sensitive so in many cases, more than half of the credits are wasted.
SD3 has just been released. I will take a look and write a guide to run it locally.
I am reading the paper and the codes. It applies a blur to the latent to the elements selected by the cross attention process, which are supposed to be spatial elements that are important.
It is the latent image that the final image steers away from. I plan to write an article along with a usage guide.
Hi!
-
LoRA is better than TI, in the sense that it can change the model more.
-
Typically instances with T4 or V100 GPU. You can stick with 1 GPU for inference. AWS is not the most cost effective. You can consider Google Colab if you are ok with the interactive interface. They offer the same GPUs at a much lower rate.
Hi! You are setting denoising strength. Are you using img2img? These settings should work on txt2img.
Some other settings I can think of:
- Have you enabled the controlnet?
- Try the SAI canny lora (select through controlnet)
If it still doesn’t work, would be helpful to upload a screenshot of settings.
The easiest way is to copy from another node.
Here’s an example node file on comfyui.
https://github.com/comfyanonymous/ComfyUI/blob/master/custom_nodes/example_node.py.example
-
