File size: 693 Bytes
b587a97 3175979 b587a97 3175979 b587a97 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 |
#import streamlit as st
#from transformers import pipeline
#pipe = pipeline('sentiment-analysis')
#text = st.text_area('enter some text!')
#if text:
# out = pipe(text)
# st.json(out)
#
# !pip install diffusers transformers
from diffusers import DiffusionPipeline
model_id = "CompVis/ldm-text2im-large-256"
# load model and scheduler
ldm = DiffusionPipeline.from_pretrained(model_id)
# run pipeline in inference (sample random noise and denoise)
prompt = "A painting of a squirrel eating a burger"
images = ldm([prompt], num_inference_steps=50, eta=0.3, guidance_scale=6)["sample"]
# save images
for idx, image in enumerate(images):
image.save(f"squirrel-{idx}.png")
|