All Your Tech patreon

Content question

🕑 Added 2024-01-25 21:40:11 +0000 UTC

Comments

Brian Lovett

Great suggestion. That’s one of the more difficult things to get right with stable diffusion in general. I’ll make some guides

Daniel Cardona Ramirez

Training Loras in a fast way, plz try use on of the loras you've alreay created to fo full body portraits and tell us if it works, I can only make it work with close-ups

Brian Lovett

Love it! I’ll start putting together some ideas. Lots of topics to cover here. I’ve been fine tuning some local LLM’s recently and was planning on doing a few deep dives, so this is helpful

Steve Fischer

I think we need more and more of these tutorials with Stable Diffusion (prompt engineering, nuances between the models, building, managing, and training Loras, inpainting. For ChatGPT, I think the Data Analysis features should be explored more--uploading a large document and knowing how to prompt with the best questions. Lastly, maybe we can talk more about vectorizing data and training it in Llama. For the coming years, these LLM/RAG/Deep Learning platforms are going to be adding features faster than many can learn about them. You are a great deal of help for explaining these concepts in plain language and pretty quickly as well. Thanks!