Open maxtaube opened 2 years ago
To avoid having to re-initialize and load the model every time a new image is generated, maybe it would be possible to keep the model loaded and prompt the user for a new input once the generation of an image is finished.
Similar to this stable diffusion CLI implementation of lstein.
UX improvement suggestion:
To avoid having to re-initialize and load the model every time a new image is generated, maybe it would be possible to keep the model loaded and prompt the user for a new input once the generation of an image is finished.
Similar to this stable diffusion CLI implementation of lstein.