Open cbresciano opened 3 months ago
That's not a checkpoint, it doesn't contain the text encoders or the VAE so you should put it in the models/unet folder and load it with the "Load Diffusion Model" node.
Thanks for your quick response. If they are no checkpontoint, can be merged with ModelMergeFlux using Load Diffusion Model1? That´s I was trying to do, to have a merged model with 11Gb to fit in my Nvidia 3060 12Gb On Sunday, August 11, 2024 at 11:22:26 AM GMT-3, comfyanonymous @.***> wrote:
That's not a checkpoint, it doesn't contain the text encoders or the VAE so you should put it in the models/unet folder and load it with the "Load Diffusion Model" node.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
Your question
When I use Load Checkpoint for both flux1_schnellFP8Kijai11GB and flux1_devFP8Kijai11GB, the program stops and tells "Unreconizable model type" as an error. Then, I use them in SwarmUI which uses ComfyUI as a backend and they work, that´s odd. Could you add direct support for them? Thanks for you attention
Logs
No response
Other
No response