search

luosiallen / latent-consistency-model

Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference
MIT License
4.26k stars 221 forks source link

some questions and suggestions #9

Closed Amin456789 closed 10 months ago

Amin456789 commented 10 months ago

Hi! omg, only 1-4 steps???? this is greaaat!

  1. i saw the dreamshaper model is about 3.4 gb, can we have fp16 models in the future too for smaller sizes?
  2. PLEASE add AnimateDiff, with only 4 steps for cpu it will be a blessing for us!
  3. any plans for img2img, inpainting? or if someday lcm sampler merged with automatic1111, will we be able to use dreamshaper with animatediff and img2img normally? [it will be great if u guys support cpu officially in a1111]

i found @rupeshs repo for cpu's gui, can't wait for new features! https://github.com/rupeshs/fastsdcpu

kind regards

Amin456789 commented 10 months ago

@JohnClaw look at this bro!

JohnClaw commented 10 months ago

look at this bro!

Thanks for informing me. FastSdCpu by rupeshs looks great. Have you tested it?

Amin456789 commented 10 months ago

not yet, but i tested the online version and it makes pretty good with 2-3 steps too! dreamshaper supports 256x256 too.! will get to it soon enough

luosiallen commented 10 months ago

Img2Img is on the way. Should be add in diffusers soon. @Amin456789

Amin456789 commented 10 months ago

thank u!