Spaces:
Sleeping
Sleeping
only from the
Browse files
app.py
CHANGED
@@ -15,11 +15,14 @@ from pathlib import Path
|
|
15 |
from db import Database
|
16 |
import uuid
|
17 |
import logging
|
|
|
|
|
|
|
18 |
logging.basicConfig(level=os.environ.get("LOGLEVEL", "INFO"))
|
19 |
|
20 |
MAX_SEED = np.iinfo(np.int32).max
|
21 |
USE_TORCH_COMPILE = os.environ.get("USE_TORCH_COMPILE", "0") == "1"
|
22 |
-
SPACE_ID = os.environ.get(
|
23 |
|
24 |
DB_PATH = Path("/data/cache") if SPACE_ID else Path("./cache")
|
25 |
IMGS_PATH = DB_PATH / "imgs"
|
@@ -28,11 +31,6 @@ IMGS_PATH.mkdir(exist_ok=True, parents=True)
|
|
28 |
|
29 |
database = Database(DB_PATH)
|
30 |
|
31 |
-
with database() as db:
|
32 |
-
cursor = db.cursor()
|
33 |
-
cursor.execute("SELECT * FROM cache")
|
34 |
-
print(list(cursor.fetchall()))
|
35 |
-
|
36 |
dtype = torch.bfloat16
|
37 |
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
|
38 |
if torch.cuda.is_available():
|
@@ -96,6 +94,7 @@ def generate(
|
|
96 |
app = FastAPI()
|
97 |
origins = [
|
98 |
"http://huggingface.co",
|
|
|
99 |
]
|
100 |
|
101 |
app.add_middleware(
|
@@ -107,6 +106,15 @@ app.add_middleware(
|
|
107 |
)
|
108 |
|
109 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
@app.get("/image")
|
111 |
async def generate_image(prompt: str, negative_prompt: str, seed: int = 2134213213):
|
112 |
cached_img = database.check(prompt, negative_prompt, seed)
|
@@ -137,190 +145,3 @@ async def main():
|
|
137 |
|
138 |
if __name__ == "__main__":
|
139 |
uvicorn.run(app, host="0.0.0.0", port=7860)
|
140 |
-
|
141 |
-
|
142 |
-
# else:
|
143 |
-
# prior_pipeline = None
|
144 |
-
# decoder_pipeline = None
|
145 |
-
|
146 |
-
|
147 |
-
# def randomize_seed_fn(seed: int, randomize_seed: bool) -> int:
|
148 |
-
# if randomize_seed:
|
149 |
-
# seed = random.randint(0, MAX_SEED)
|
150 |
-
# return seed
|
151 |
-
|
152 |
-
|
153 |
-
# def generate(
|
154 |
-
# prompt: str,
|
155 |
-
# negative_prompt: str = "",
|
156 |
-
# seed: int = 0,
|
157 |
-
# width: int = 1024,
|
158 |
-
# height: int = 1024,
|
159 |
-
# prior_num_inference_steps: int = 30,
|
160 |
-
# # prior_timesteps: List[float] = None,
|
161 |
-
# prior_guidance_scale: float = 4.0,
|
162 |
-
# decoder_num_inference_steps: int = 12,
|
163 |
-
# # decoder_timesteps: List[float] = None,
|
164 |
-
# decoder_guidance_scale: float = 0.0,
|
165 |
-
# num_images_per_prompt: int = 2,
|
166 |
-
# progress=gr.Progress(track_tqdm=True),
|
167 |
-
# ) -> PIL.Image.Image:
|
168 |
-
|
169 |
-
# generator = torch.Generator().manual_seed(seed)
|
170 |
-
# prior_output = prior_pipeline(
|
171 |
-
# prompt=prompt,
|
172 |
-
# height=height,
|
173 |
-
# width=width,
|
174 |
-
# num_inference_steps=prior_num_inference_steps,
|
175 |
-
# timesteps=DEFAULT_STAGE_C_TIMESTEPS,
|
176 |
-
# negative_prompt=negative_prompt,
|
177 |
-
# guidance_scale=prior_guidance_scale,
|
178 |
-
# num_images_per_prompt=num_images_per_prompt,
|
179 |
-
# generator=generator,
|
180 |
-
# )
|
181 |
-
# decoder_output = decoder_pipeline(
|
182 |
-
# image_embeddings=prior_output.image_embeddings,
|
183 |
-
# prompt=prompt,
|
184 |
-
# num_inference_steps=decoder_num_inference_steps,
|
185 |
-
# # timesteps=decoder_timesteps,
|
186 |
-
# guidance_scale=decoder_guidance_scale,
|
187 |
-
# negative_prompt=negative_prompt,
|
188 |
-
# generator=generator,
|
189 |
-
# output_type="pil",
|
190 |
-
# ).images
|
191 |
-
|
192 |
-
# return decoder_output[0]
|
193 |
-
|
194 |
-
|
195 |
-
# examples = [
|
196 |
-
# "An astronaut riding a green horse",
|
197 |
-
# "A mecha robot in a favela by Tarsila do Amaral",
|
198 |
-
# "The sprirt of a Tamagotchi wandering in the city of Los Angeles",
|
199 |
-
# "A delicious feijoada ramen dish"
|
200 |
-
# ]
|
201 |
-
|
202 |
-
# with gr.Blocks() as demo:
|
203 |
-
# gr.Markdown(DESCRIPTION)
|
204 |
-
# gr.DuplicateButton(
|
205 |
-
# value="Duplicate Space for private use",
|
206 |
-
# elem_id="duplicate-button",
|
207 |
-
# visible=os.getenv("SHOW_DUPLICATE_BUTTON") == "1",
|
208 |
-
# )
|
209 |
-
# with gr.Group():
|
210 |
-
# with gr.Row():
|
211 |
-
# prompt = gr.Text(
|
212 |
-
# label="Prompt",
|
213 |
-
# show_label=False,
|
214 |
-
# max_lines=1,
|
215 |
-
# placeholder="Enter your prompt",
|
216 |
-
# container=False,
|
217 |
-
# )
|
218 |
-
# run_button = gr.Button("Run", scale=0)
|
219 |
-
# result = gr.Image(label="Result", show_label=False)
|
220 |
-
# with gr.Accordion("Advanced options", open=False):
|
221 |
-
# negative_prompt = gr.Text(
|
222 |
-
# label="Negative prompt",
|
223 |
-
# max_lines=1,
|
224 |
-
# placeholder="Enter a Negative Prompt",
|
225 |
-
# )
|
226 |
-
|
227 |
-
# seed = gr.Slider(
|
228 |
-
# label="Seed",
|
229 |
-
# minimum=0,
|
230 |
-
# maximum=MAX_SEED,
|
231 |
-
# step=1,
|
232 |
-
# value=0,
|
233 |
-
# )
|
234 |
-
# randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
|
235 |
-
# with gr.Row():
|
236 |
-
# width = gr.Slider(
|
237 |
-
# label="Width",
|
238 |
-
# minimum=1024,
|
239 |
-
# maximum=1536,
|
240 |
-
# step=512,
|
241 |
-
# value=1024,
|
242 |
-
# )
|
243 |
-
# height = gr.Slider(
|
244 |
-
# label="Height",
|
245 |
-
# minimum=1024,
|
246 |
-
# maximum=1536,
|
247 |
-
# step=512,
|
248 |
-
# value=1024,
|
249 |
-
# )
|
250 |
-
# num_images_per_prompt = gr.Slider(
|
251 |
-
# label="Number of Images",
|
252 |
-
# minimum=1,
|
253 |
-
# maximum=2,
|
254 |
-
# step=1,
|
255 |
-
# value=1,
|
256 |
-
# )
|
257 |
-
# with gr.Row():
|
258 |
-
# prior_guidance_scale = gr.Slider(
|
259 |
-
# label="Prior Guidance Scale",
|
260 |
-
# minimum=0,
|
261 |
-
# maximum=20,
|
262 |
-
# step=0.1,
|
263 |
-
# value=4.0,
|
264 |
-
# )
|
265 |
-
# prior_num_inference_steps = gr.Slider(
|
266 |
-
# label="Prior Inference Steps",
|
267 |
-
# minimum=10,
|
268 |
-
# maximum=30,
|
269 |
-
# step=1,
|
270 |
-
# value=20,
|
271 |
-
# )
|
272 |
-
|
273 |
-
# decoder_guidance_scale = gr.Slider(
|
274 |
-
# label="Decoder Guidance Scale",
|
275 |
-
# minimum=0,
|
276 |
-
# maximum=0,
|
277 |
-
# step=0.1,
|
278 |
-
# value=0.0,
|
279 |
-
# )
|
280 |
-
# decoder_num_inference_steps = gr.Slider(
|
281 |
-
# label="Decoder Inference Steps",
|
282 |
-
# minimum=4,
|
283 |
-
# maximum=12,
|
284 |
-
# step=1,
|
285 |
-
# value=10,
|
286 |
-
# )
|
287 |
-
|
288 |
-
# gr.Examples(
|
289 |
-
# examples=examples,
|
290 |
-
# inputs=prompt,
|
291 |
-
# outputs=result,
|
292 |
-
# fn=generate,
|
293 |
-
# cache_examples=False,
|
294 |
-
# )
|
295 |
-
|
296 |
-
# inputs = [
|
297 |
-
# prompt,
|
298 |
-
# negative_prompt,
|
299 |
-
# seed,
|
300 |
-
# width,
|
301 |
-
# height,
|
302 |
-
# prior_num_inference_steps,
|
303 |
-
# # prior_timesteps,
|
304 |
-
# prior_guidance_scale,
|
305 |
-
# decoder_num_inference_steps,
|
306 |
-
# # decoder_timesteps,
|
307 |
-
# decoder_guidance_scale,
|
308 |
-
# num_images_per_prompt,
|
309 |
-
# ]
|
310 |
-
# gr.on(
|
311 |
-
# triggers=[prompt.submit, negative_prompt.submit, run_button.click],
|
312 |
-
# fn=randomize_seed_fn,
|
313 |
-
# inputs=[seed, randomize_seed],
|
314 |
-
# outputs=seed,
|
315 |
-
# queue=False,
|
316 |
-
# api_name=False,
|
317 |
-
# ).then(
|
318 |
-
# fn=generate,
|
319 |
-
# inputs=inputs,
|
320 |
-
# outputs=result,
|
321 |
-
# api_name="run",
|
322 |
-
# )
|
323 |
-
|
324 |
-
|
325 |
-
# if __name__ == "__main__":
|
326 |
-
# demo.queue(max_size=20).launch()
|
|
|
15 |
from db import Database
|
16 |
import uuid
|
17 |
import logging
|
18 |
+
from fastapi import FastAPI, Request, HTTPException
|
19 |
+
from fastapi.middleware.cors import CORSMiddleware
|
20 |
+
|
21 |
logging.basicConfig(level=os.environ.get("LOGLEVEL", "INFO"))
|
22 |
|
23 |
MAX_SEED = np.iinfo(np.int32).max
|
24 |
USE_TORCH_COMPILE = os.environ.get("USE_TORCH_COMPILE", "0") == "1"
|
25 |
+
SPACE_ID = os.environ.get("SPACE_ID", "")
|
26 |
|
27 |
DB_PATH = Path("/data/cache") if SPACE_ID else Path("./cache")
|
28 |
IMGS_PATH = DB_PATH / "imgs"
|
|
|
31 |
|
32 |
database = Database(DB_PATH)
|
33 |
|
|
|
|
|
|
|
|
|
|
|
34 |
dtype = torch.bfloat16
|
35 |
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
|
36 |
if torch.cuda.is_available():
|
|
|
94 |
app = FastAPI()
|
95 |
origins = [
|
96 |
"http://huggingface.co",
|
97 |
+
"localhost",
|
98 |
]
|
99 |
|
100 |
app.add_middleware(
|
|
|
106 |
)
|
107 |
|
108 |
|
109 |
+
@app.middleware("http")
|
110 |
+
async def validate_origin(request: Request, call_next):
|
111 |
+
logging.info(f"Request origin: {request.headers.get('origin')}")
|
112 |
+
if request.headers.get("origin") not in origins:
|
113 |
+
raise HTTPException(status_code=403, detail="Forbidden")
|
114 |
+
response = await call_next(request)
|
115 |
+
return response
|
116 |
+
|
117 |
+
|
118 |
@app.get("/image")
|
119 |
async def generate_image(prompt: str, negative_prompt: str, seed: int = 2134213213):
|
120 |
cached_img = database.check(prompt, negative_prompt, seed)
|
|
|
145 |
|
146 |
if __name__ == "__main__":
|
147 |
uvicorn.run(app, host="0.0.0.0", port=7860)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|