Spaces:
Running
Running
File size: 4,515 Bytes
3911a99 5fef1f4 3911a99 5fef1f4 3911a99 5fef1f4 3911a99 5fef1f4 3911a99 5fef1f4 3911a99 5fef1f4 3911a99 a8ce530 5fef1f4 a8ce530 5fef1f4 a8ce530 5fef1f4 a8ce530 3911a99 |
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 |
from diffusers import ( StableDiffusionControlNetPipeline,
ControlNetModel, UniPCMultistepScheduler )
from transformers import pipeline
from PIL import Image
import gradio as gr
import numpy as np
import torch
stable_model_list = [
"runwayml/stable-diffusion-v1-5",
"stabilityai/stable-diffusion-2-1",
]
controlnet_depth_model_list = [
"fusing/stable-diffusion-v1-5-controlnet-depth",
"thibaud/controlnet-sd21-depth-diffusers"
]
stable_prompt_list = [
"a photo of a man.",
"a photo of a girl."
]
stable_negative_prompt_list = [
"bad, ugly",
"deformed"
]
def controlnet_depth(image_path:str, depth_model_path:str):
depth_estimator = pipeline('depth-estimation')
image = Image.open(image_path)
image = depth_estimator(image)['depth']
image = np.array(image)
image = image[:, :, None]
image = np.concatenate([image, image, image], axis=2)
image = Image.fromarray(image)
controlnet = ControlNetModel.from_pretrained(
depth_model_path, torch_dtype=torch.float16
)
return controlnet, image
def stable_diffusion_controlnet_depth(
image_path:str,
stable_model_path:str,
depth_model_path:str,
prompt:str,
negative_prompt:str,
guidance_scale:int,
num_inference_step:int,
):
controlnet, image = controlnet_depth(image_path=image_path, depth_model_path=depth_model_path)
pipe = StableDiffusionControlNetPipeline.from_pretrained(
pretrained_model_name_or_path=stable_model_path,
controlnet=controlnet,
safety_checker=None,
torch_dtype=torch.float16
)
pipe.to("cuda")
pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
pipe.enable_xformers_memory_efficient_attention()
output = pipe(
prompt = prompt,
image = image,
negative_prompt = negative_prompt,
num_inference_steps = num_inference_step,
guidance_scale = guidance_scale,
).images
return output[0]
def stable_diffusion_controlnet_depth_app():
with gr.Blocks():
with gr.Row():
with gr.Column():
controlnet_depth_image_file = gr.Image(
type='filepath',
label='Image'
)
controlnet_depth_stable_model_id = gr.Dropdown(
choices=stable_model_list,
value=stable_model_list[0],
label='Stable Model Id'
)
controlnet_depth_model_id = gr.Dropdown(
choices=controlnet_depth_model_list,
value=controlnet_depth_model_list[0],
label='ControlNet Model Id'
)
controlnet_depth_prompt = gr.Textbox(
lines=1,
value=stable_prompt_list[0],
label='Prompt'
)
controlnet_depth_negative_prompt = gr.Textbox(
lines=1,
value=stable_negative_prompt_list[0],
label='Negative Prompt'
)
with gr.Accordion("Advanced Options", open=False):
controlnet_depth_guidance_scale = gr.Slider(
minimum=0.1,
maximum=15,
step=0.1,
value=7.5,
label='Guidance Scale'
)
controlnet_depth_num_inference_step = gr.Slider(
minimum=1,
maximum=100,
step=1,
value=50,
label='Num Inference Step'
)
controlnet_depth_predict = gr.Button(value='Generator')
with gr.Column():
output_image = gr.Image(label='Output')
controlnet_depth_predict.click(
fn=stable_diffusion_controlnet_depth,
inputs=[
controlnet_depth_image_file,
controlnet_depth_stable_model_id,
controlnet_depth_model_id,
controlnet_depth_prompt,
controlnet_depth_negative_prompt,
controlnet_depth_guidance_scale,
controlnet_depth_num_inference_step,
],
outputs=output_image
)
|