Spaces:
Runtime error
Runtime error
File size: 6,263 Bytes
9ce2bc6 |
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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 |
#!/usr/bin/env python
from __future__ import annotations
import os
import random
import gradio as gr
import numpy as np
import PIL.Image
import torch
from lcm_pipeline import LatentConsistencyModelPipeline
from lcm_scheduler import LCMScheduler
from diffusers import AutoencoderKL, UNet2DConditionModel
from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
from transformers import CLIPTokenizer, CLIPTextModel, CLIPImageProcessor
import os
import torch
from tqdm import tqdm
from safetensors.torch import load_file
from huggingface_hub import hf_hub_download
DESCRIPTION = "# Latent Consistency"
if not torch.cuda.is_available():
DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
MAX_SEED = np.iinfo(np.int32).max
CACHE_EXAMPLES = torch.cuda.is_available() and os.getenv("CACHE_EXAMPLES") == "1"
MAX_IMAGE_SIZE = int(os.getenv("MAX_IMAGE_SIZE", "512"))
USE_TORCH_COMPILE = os.getenv("USE_TORCH_COMPILE") == "1"
DTYPE = torch.float32 # torch.float16 works as well, but pictures seem to be a bit worse
model_id = "digiplay/DreamShaper_7"
# Initalize Diffusers Model:
vae = AutoencoderKL.from_pretrained(model_id, subfolder="vae")
text_encoder = CLIPTextModel.from_pretrained(model_id, subfolder="text_encoder")
tokenizer = CLIPTokenizer.from_pretrained(model_id, subfolder="tokenizer")
config = UNet2DConditionModel.load_config(model_id, subfolder="unet")
config["time_cond_proj_dim"] = 256
unet = UNet2DConditionModel.from_config(config)
safety_checker = StableDiffusionSafetyChecker.from_pretrained(model_id, subfolder="safety_checker")
feature_extractor = CLIPImageProcessor.from_pretrained(model_id, subfolder="feature_extractor")
# Initalize Scheduler:
scheduler = LCMScheduler(beta_start=0.00085, beta_end=0.0120, beta_schedule="scaled_linear", prediction_type="epsilon")
if torch.cuda.is_available():
# Replace the unet with LCM:
lcm_unet_ckpt = hf_hub_download("SimianLuo/LCM_Dreamshaper_v7", filename="LCM_Dreamshaper_v7_4k.safetensors")
ckpt = load_file(lcm_unet_ckpt)
m, u = unet.load_state_dict(ckpt, strict=False)
if len(m) > 0:
print("missing keys:")
print(m)
if len(u) > 0:
print("unexpected keys:")
print(u)
# LCM Pipeline:
pipe = LatentConsistencyModelPipeline(vae=vae, text_encoder=text_encoder, tokenizer=tokenizer, unet=unet, scheduler=scheduler, safety_checker=safety_checker, feature_extractor=feature_extractor)
pipe = pipe.to(device="cuda", dtype=DTYPE)
if USE_TORCH_COMPILE:
pipe.unet = torch.compile(pipe.unet, mode="reduce-overhead", fullgraph=True)
def randomize_seed_fn(seed: int, randomize_seed: bool) -> int:
if randomize_seed:
seed = random.randint(0, MAX_SEED)
return seed
def generate(
prompt: str,
seed: int = 0,
width: int = 512,
height: int = 512,
guidance_scale: float = 8.0,
num_inference_steps: int = 4,
num_images: int = 4,
) -> PIL.Image.Image:
torch.manual_seed(seed)
return pipe(
prompt=prompt,
width=width,
height=height,
guidance_scale=guidance_scale,
num_inference_steps=num_inference_steps,
num_images_per_prompt=num_images,
lcm_origin_steps=50,
output_type="pil",
).images
examples = [
"Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
"An astronaut riding a green horse",
]
with gr.Blocks(css="style.css") as demo:
gr.Markdown(DESCRIPTION)
gr.DuplicateButton(
value="Duplicate Space for private use",
elem_id="duplicate-button",
visible=os.getenv("SHOW_DUPLICATE_BUTTON") == "1",
)
with gr.Group():
with gr.Row():
prompt = gr.Text(
label="Prompt",
show_label=False,
max_lines=1,
placeholder="Enter your prompt",
container=False,
)
run_button = gr.Button("Run", scale=0)
result = gr.Gallery(
label="Generated images", show_label=False, elem_id="gallery", grid=[2]
)
with gr.Accordion("Advanced options", open=False):
seed = gr.Slider(
label="Seed",
minimum=0,
maximum=MAX_SEED,
step=1,
value=0,
)
randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
with gr.Row():
width = gr.Slider(
label="Width",
minimum=256,
maximum=MAX_IMAGE_SIZE,
step=32,
value=512,
)
height = gr.Slider(
label="Height",
minimum=256,
maximum=MAX_IMAGE_SIZE,
step=32,
value=512,
)
with gr.Row():
guidance_scale = gr.Slider(
label="Guidance scale for base",
minimum=1,
maximum=20,
step=0.1,
value=8.0,
)
num_inference_steps = gr.Slider(
label="Number of inference steps for base",
minimum=1,
maximum=5,
step=1,
value=4,
)
# with gr.Row():
# num_images = gr.Slider(
# label="Number of images"
# minimum=1,
# maximum=8,
# step=1,
# value=4,
# )
gr.Examples(
examples=examples,
inputs=prompt,
outputs=result,
fn=generate,
cache_examples=CACHE_EXAMPLES,
)
gr.on(
triggers=[
prompt.submit,
run_button.click,
],
fn=randomize_seed_fn,
inputs=[seed, randomize_seed],
outputs=seed,
queue=False,
api_name=False,
).then(
fn=generate,
inputs=[
prompt,
seed,
width,
height,
guidance_scale,
num_inference_steps,
],
outputs=result,
api_name="run",
)
if __name__ == "__main__":
# demo.queue(max_size=20).launch()
demo.launch(share=True)
|