Your videos very informative you have excellet teaching skills. I am mostly Using Fooocus webUI and I play with ComfyUI time to time. Now I can unerstand how Fooocus V2 expansion and image prompt works. Thanks You
Hello @tiberiusvetus9113! These models are already fine-tuned on stable diffusion prompts, and they come with lower system requirements. Using the latest models is an option if you have the resources (VRAM and Storage space). these models are around 300mb, which makes them efficient and ensures they work without consuming excessive resources, leaving more room for other nodes.
Hello @kargulo, Thanks for checking the video! I appreciate your perspective. The goal of the video was to show how, regardless of GPT version, text generation can be used to provide positive prompts. I used GPT2 for its manageable size and pre-training on Stable Diffusion prompts. If you have the resources, you can experiment with the latest models, even if it might be less efficient. If you have OpenAI keys, you can use ChatGPT 3.5 or 4. It's is just an approach among many as an alternative tool. Even artists get creative blocks, having an AI comes up with inspiration can be a valuable addition.