Spaces:
Runtime error
Runtime error
Update app.py (#8)
Browse files- Update app.py (e0bb6b68b671f2e89c16d76c4772df4f913caed5)
app.py
CHANGED
@@ -7,7 +7,25 @@ from diffusers import DiffusionPipeline
|
|
7 |
import io
|
8 |
import base64
|
9 |
from PIL import Image
|
10 |
-
import
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
|
12 |
dtype = torch.bfloat16
|
13 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
@@ -19,6 +37,8 @@ MAX_IMAGE_SIZE = 2048
|
|
19 |
|
20 |
@spaces.GPU()
|
21 |
def infer(prompt, seed=42, randomize_seed=False, width=1024, height=1024, num_inference_steps=4, progress=gr.Progress(track_tqdm=True)):
|
|
|
|
|
22 |
if randomize_seed:
|
23 |
seed = random.randint(0, MAX_SEED)
|
24 |
generator = torch.Generator().manual_seed(seed)
|
@@ -38,13 +58,21 @@ def infer(prompt, seed=42, randomize_seed=False, width=1024, height=1024, num_in
|
|
38 |
image.save(buffered, format="PNG")
|
39 |
img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
|
40 |
|
41 |
-
# Retornar JSON com Base64 e seed
|
42 |
return {"image_base64": f"data:image/png;base64,{img_str}", "seed": seed}
|
43 |
|
44 |
-
#
|
45 |
-
|
46 |
-
|
47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
|
49 |
examples = [
|
50 |
"a tiny astronaut hatching from an egg on the moon",
|
@@ -125,7 +153,6 @@ with gr.Blocks(css=css) as demo:
|
|
125 |
output = infer(prompt, seed, randomize_seed, width, height, num_inference_steps)
|
126 |
return output["image_base64"], output["seed"]
|
127 |
|
128 |
-
# Interface Gradio
|
129 |
gr.on(
|
130 |
triggers=[run_button.click, prompt.submit],
|
131 |
fn=format_output,
|
@@ -133,7 +160,5 @@ with gr.Blocks(css=css) as demo:
|
|
133 |
outputs=[result, seed_output]
|
134 |
)
|
135 |
|
136 |
-
|
137 |
-
demo.queue(api_name="infer_api").launch()
|
138 |
-
|
139 |
demo.launch()
|
|
|
7 |
import io
|
8 |
import base64
|
9 |
from PIL import Image
|
10 |
+
import logging
|
11 |
+
from fastapi import FastAPI
|
12 |
+
from pydantic import BaseModel
|
13 |
+
|
14 |
+
# Configurar logging para depuração
|
15 |
+
logging.basicConfig(level=logging.INFO)
|
16 |
+
logger = logging.getLogger(__name__)
|
17 |
+
|
18 |
+
# Inicializar FastAPI
|
19 |
+
app = FastAPI()
|
20 |
+
|
21 |
+
# Modelo para validação dos parâmetros da API
|
22 |
+
class ImageRequest(BaseModel):
|
23 |
+
prompt: str
|
24 |
+
seed: int = 42
|
25 |
+
randomize_seed: bool = False
|
26 |
+
width: int = 1024
|
27 |
+
height: int = 1024
|
28 |
+
num_inference_steps: int = 4
|
29 |
|
30 |
dtype = torch.bfloat16
|
31 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
|
|
37 |
|
38 |
@spaces.GPU()
|
39 |
def infer(prompt, seed=42, randomize_seed=False, width=1024, height=1024, num_inference_steps=4, progress=gr.Progress(track_tqdm=True)):
|
40 |
+
logger.info(f"Chamando infer com prompt={prompt}, seed={seed}, randomize_seed={randomize_seed}, width={width}, height={height}, num_inference_steps={num_inference_steps}")
|
41 |
+
|
42 |
if randomize_seed:
|
43 |
seed = random.randint(0, MAX_SEED)
|
44 |
generator = torch.Generator().manual_seed(seed)
|
|
|
58 |
image.save(buffered, format="PNG")
|
59 |
img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
|
60 |
|
|
|
61 |
return {"image_base64": f"data:image/png;base64,{img_str}", "seed": seed}
|
62 |
|
63 |
+
# Endpoint FastAPI
|
64 |
+
@app.post("/api/infer")
|
65 |
+
async def api_infer(request: ImageRequest):
|
66 |
+
logger.info(f"Requisição API recebida: {request}")
|
67 |
+
result = infer(
|
68 |
+
prompt=request.prompt,
|
69 |
+
seed=request.seed,
|
70 |
+
randomize_seed=request.randomize_seed,
|
71 |
+
width=request.width,
|
72 |
+
height=request.height,
|
73 |
+
num_inference_steps=request.num_inference_steps
|
74 |
+
)
|
75 |
+
return result
|
76 |
|
77 |
examples = [
|
78 |
"a tiny astronaut hatching from an egg on the moon",
|
|
|
153 |
output = infer(prompt, seed, randomize_seed, width, height, num_inference_steps)
|
154 |
return output["image_base64"], output["seed"]
|
155 |
|
|
|
156 |
gr.on(
|
157 |
triggers=[run_button.click, prompt.submit],
|
158 |
fn=format_output,
|
|
|
160 |
outputs=[result, seed_output]
|
161 |
)
|
162 |
|
163 |
+
# Iniciar o Gradio (sem queue, pois usamos FastAPI para a API)
|
|
|
|
|
164 |
demo.launch()
|