Spaces:
Running
Running
File size: 9,435 Bytes
3964dd9 95b4e61 803809e 3964dd9 865d1fd 3964dd9 95b4e61 3964dd9 95b4e61 13869fb 205eccb 13869fb 3964dd9 ef7853e 3964dd9 f6c0d06 3964dd9 13869fb 3964dd9 205eccb 13869fb 3964dd9 f6c0d06 13869fb 3964dd9 f6c0d06 95b4e61 803809e 95b4e61 ae609cc 95b4e61 803809e 3964dd9 f6c0d06 3d2a5ad 3964dd9 3d2a5ad 3964dd9 24751c5 3964dd9 24751c5 e7a333a 6eedf06 e7a333a 6eedf06 e7a333a 6eedf06 e7a333a f229815 6f74e10 e7a333a 13869fb 205eccb e7636eb 95b4e61 13869fb 205eccb 13869fb e7636eb e7a333a 3964dd9 13869fb |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 |
import os
import requests
import subprocess
import gradio as gr
from huggingface_hub import HfApi, upload_file
from datetime import datetime
# Token Hugging Face từ biến môi trường
hf_token = os.getenv("HF_TOKEN")
# URLs cần tải
app_url = "https://huggingface.co/datasets/ArrcttacsrjksX/Deffusion/resolve/main/RunModelAppp/App/sdmaster-d9b5942LatestJan182025"
model_url = "https://huggingface.co/datasets/ArrcttacsrjksX/Deffusion/resolve/main/Model/realisticVisionV60B1_v51HyperVAE.safetensors"
# Đường dẫn lưu file
app_path = "sdRundeffusiononhuggingfacemaster-ac54e00"
model_path = "realisticVisionV60B1_v51HyperVAE.safetensors"
# Hàm tải file từ Hugging Face
def download_file(url, output_path, token):
headers = {"Authorization": f"Bearer {token}"}
response = requests.get(url, headers=headers, stream=True)
response.raise_for_status() # Kiểm tra lỗi
with open(output_path, "wb") as f:
for chunk in response.iter_content(chunk_size=8192):
f.write(chunk)
print(f"Downloaded: {output_path}")
# Tải các file nếu chưa tồn tại
if not os.path.exists(app_path):
download_file(app_url, app_path, hf_token)
subprocess.run(["chmod", "+x", app_path]) # Thay đổi quyền thực thi
if not os.path.exists(model_path):
download_file(model_url, model_path, hf_token)
# Hàm xử lý chạy ứng dụng và lưu kết quả
def run_command_and_save(
prompt, mode, height, width, steps, seed, cfg_scale, strength, sampling_method,
batch_count, schedule, clip_skip, vae_tiling, vae_on_cpu, clip_on_cpu, diffusion_fa,
control_net_cpu, canny, verbose, init_image=None, weight_type=None
):
try:
# Lưu ảnh đầu vào nếu được cung cấp
init_image_path = None
if init_image is not None:
init_image_path = "input_image.png"
init_image.save(init_image_path)
# Tạo lệnh chạy
command = [
f"./{app_path}",
"-M", mode,
"-m", model_path,
"-p", prompt,
"-H", str(height),
"-W", str(width),
"--steps", str(steps),
"-s", str(seed),
"--cfg-scale", str(cfg_scale),
"--strength", str(strength),
"--sampling-method", sampling_method,
"--batch-count", str(batch_count),
"--schedule", schedule,
"--clip-skip", str(clip_skip),
]
# Thêm tùy chọn Weight Type nếu được chỉ định
if weight_type:
command.extend(["--type", weight_type])
# Thêm các tùy chọn khác
if vae_tiling:
command.append("--vae-tiling")
if vae_on_cpu:
command.append("--vae-on-cpu")
if clip_on_cpu:
command.append("--clip-on-cpu")
if diffusion_fa:
command.append("--diffusion-fa")
if control_net_cpu:
command.append("--control-net-cpu")
if canny:
command.append("--canny")
if verbose:
command.append("-v")
# Thêm ảnh đầu vào nếu có
if mode == "img2img" and init_image_path:
command.extend(["-i", init_image_path])
# Chạy lệnh và hiển thị log theo thời gian thực
process = subprocess.Popen(
command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True
)
logs = []
for line in process.stdout:
logs.append(line.strip()) # Lưu log vào danh sách
print(line, end="") # In log ra màn hình
process.wait() # Đợi tiến trình hoàn thành
# Kiểm tra kết quả và trả về
if process.returncode == 0:
# Đường dẫn ảnh đầu ra mặc định
output_path = "./output.png"
if os.path.exists(output_path):
# Tạo tên thư mục và tên file dựa trên ngày giờ hiện tại
now = datetime.now()
folder_name = f"SetImages+{now.strftime('%d+%m+%Y')}"
file_name = f"Image+{now.strftime('%S+%M+%H')}+_{prompt[:50].replace(' ', '_')}.png"
local_folder = os.path.join(".", folder_name)
os.makedirs(local_folder, exist_ok=True)
local_file_path = os.path.join(local_folder, file_name)
# Di chuyển file đến thư mục mới
os.rename(output_path, local_file_path)
# Upload file lên Hugging Face Hub
api = HfApi()
repo_id = "ArrcttacsrjksX/Deffusion"
remote_path = f"{folder_name}/{file_name}"
upload_file(
path_or_fileobj=local_file_path,
path_in_repo=remote_path,
repo_id=repo_id,
token=hf_token,
repo_type="dataset",
)
print(f"Uploaded to Hugging Face: {remote_path}")
return local_file_path, "\n".join(logs)
else:
return None, "\n".join(logs)
else:
error_log = process.stderr.read() # Đọc lỗi
logs.append(error_log)
return None, "\n".join(logs)
except Exception as e:
return None, str(e)
# Giao diện Gradio
def toggle_image_input(mode):
"""Hiển thị hoặc ẩn ô Drop Image dựa trên mode."""
return gr.update(visible=(mode == "img2img"))
with gr.Blocks(theme=gr.themes.Soft()) as demo:
gr.Markdown(
"""
# 🌟 **Stable Diffusion Interface**
Generate stunning images from text or modify existing images with AI-powered tools.
"""
)
# Thiết lập giao diện
with gr.Row():
with gr.Column():
prompt = gr.Textbox(
label="🎨 Prompt", placeholder="Enter your creative idea here...", lines=2
)
mode = gr.Radio(
choices=["txt2img", "img2img"], value="txt2img", label="Mode", interactive=True
)
init_image = gr.Image(
label="Drop Image (for img2img mode)", type="pil", visible=False
)
mode.change(toggle_image_input, inputs=mode, outputs=init_image)
with gr.Column():
height = gr.Slider(
128, 1024, value=512, step=64, label="Image Height (px)", interactive=True
)
width = gr.Slider(
128, 1024, value=512, step=64, label="Image Width (px)", interactive=True
)
steps = gr.Slider(
1, 100, value=20, step=1, label="Sampling Steps", interactive=True
)
seed = gr.Slider(
-1, 10000, value=-1, step=1, label="Random Seed (-1 for random)", interactive=True
)
cfg_scale = gr.Slider(
1, 20, value=7, step=0.1, label="CFG Scale", interactive=True
)
strength = gr.Slider(
0, 1, value=0.75, step=0.01, label="Strength (img2img only)", interactive=True
)
with gr.Row():
sampling_method = gr.Dropdown(
choices=["euler", "euler_a", "heun", "dpm2", "dpm++2s_a", "dpm++2m", "dpm++2mv2", "ipndm", "ipndm_v", "lcm"],
value="euler_a", label="Sampling Method", interactive=True
)
batch_count = gr.Slider(
1, 10, value=1, step=1, label="Batch Count", interactive=True
)
schedule = gr.Dropdown(
choices=["discrete", "karras", "exponential", "ays", "gits"],
value="discrete", label="Denoiser Sigma Schedule", interactive=True
)
with gr.Row():
clip_skip = gr.Slider(
-1, 10, value=-1, step=1, label="CLIP Skip Layers", interactive=True
)
vae_tiling = gr.Checkbox(label="VAE Tiling", value=False)
vae_on_cpu = gr.Checkbox(label="VAE on CPU", value=False)
clip_on_cpu = gr.Checkbox(label="CLIP on CPU", value=False)
diffusion_fa = gr.Checkbox(label="Diffusion Flash Attention", value=False)
control_net_cpu = gr.Checkbox(label="ControlNet on CPU", value=False)
canny = gr.Checkbox(label="Canny Preprocessor", value=False)
verbose = gr.Checkbox(label="Verbose Logging", value=False)
with gr.Row():
weight_type = gr.Dropdown(
choices=["f32", "f16", "q4_0", "q4_1", "q5_0", "q5_1", "q8_0", "q2_K", "q3_K", "q4_K"],
value=None, # Mặc định không chọn
label="Weight Type (Optional)",
interactive=True
)
# Nút chạy và kết quả
with gr.Row():
run_button = gr.Button("🚀 Run", variant="primary")
with gr.Row():
output_image = gr.File(label="Download Image", interactive=False)
log_output = gr.Textbox(label="Logs", interactive=False, lines=10)
# Kết nối nút Run với hàm xử lý
run_button.click(
run_command_and_save,
inputs=[
prompt, mode, height, width, steps, seed, cfg_scale, strength, sampling_method,
batch_count, schedule, clip_skip, vae_tiling, vae_on_cpu, clip_on_cpu, diffusion_fa,
control_net_cpu, canny, verbose, init_image, weight_type # Thêm weight_type
],
outputs=[output_image, log_output],
)
demo.launch() |