colab

[elisapodan]

dreambooth anyway worked in colab, looked only gradio so i think if print some menus for load diffusion in notebook this well be worked, maybe also automatic wellbe work from notebook without gradio any... fo example next gen ok
import os,import torch,from diffusers import StableDiffusionPipeline
SizeFielse = 3,aspipe = StableDiffusionPipeline.from_pretrained("/content/models/asd", torch_dtype=torch.float16).to("cuda")
for i in range(SizeFielse):
filemodel = aspipe("a somes", num_inference_steps=50, guidance_scale=7.5).images[0]
filemodel.save("/content/n"+str(i)+".png")