Spaces:
Runtime error
Runtime error
Afrinetwork7
commited on
Update app.py
Browse files
app.py
CHANGED
@@ -3,11 +3,13 @@ from pydantic import BaseModel
|
|
3 |
import numpy as np
|
4 |
import random
|
5 |
import torch
|
6 |
-
from diffusers import DiffusionPipeline
|
7 |
import boto3
|
8 |
from io import BytesIO
|
9 |
import time
|
10 |
import os
|
|
|
|
|
|
|
11 |
|
12 |
# S3 Configuration
|
13 |
S3_BUCKET = "afri"
|
@@ -21,9 +23,25 @@ s3_client = boto3.client('s3',
|
|
21 |
aws_access_key_id=S3_ACCESS_KEY_ID,
|
22 |
aws_secret_access_key=S3_SECRET_ACCESS_KEY)
|
23 |
|
24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
26 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
MAX_SEED = np.iinfo(np.int32).max
|
28 |
MAX_IMAGE_SIZE = 2048
|
29 |
|
@@ -32,24 +50,33 @@ app = FastAPI()
|
|
32 |
class InferenceRequest(BaseModel):
|
33 |
prompt: str
|
34 |
seed: int = 42
|
35 |
-
randomize_seed: bool =
|
36 |
width: int = 1024
|
37 |
height: int = 1024
|
38 |
-
guidance_scale: float = 5
|
39 |
-
num_inference_steps: int =
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
|
41 |
def save_image_to_s3(image):
|
42 |
img_byte_arr = BytesIO()
|
43 |
image.save(img_byte_arr, format='PNG')
|
44 |
img_byte_arr = img_byte_arr.getvalue()
|
45 |
-
|
46 |
filename = f"generated_image_{int(time.time())}.png"
|
47 |
-
|
48 |
s3_client.put_object(Bucket=S3_BUCKET,
|
49 |
Key=filename,
|
50 |
Body=img_byte_arr,
|
51 |
ContentType='image/png')
|
52 |
-
|
53 |
url = f"https://{S3_BUCKET}.s3.{S3_REGION}.amazonaws.com/{filename}"
|
54 |
return url
|
55 |
|
@@ -59,18 +86,18 @@ async def infer(request: InferenceRequest):
|
|
59 |
seed = random.randint(0, MAX_SEED)
|
60 |
else:
|
61 |
seed = request.seed
|
62 |
-
|
63 |
generator = torch.Generator().manual_seed(seed)
|
64 |
|
65 |
try:
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
|
|
74 |
|
75 |
image_url = save_image_to_s3(image)
|
76 |
|
@@ -80,4 +107,4 @@ async def infer(request: InferenceRequest):
|
|
80 |
|
81 |
@app.get("/")
|
82 |
async def root():
|
83 |
-
return {"message": "Welcome to the IG API"}
|
|
|
3 |
import numpy as np
|
4 |
import random
|
5 |
import torch
|
|
|
6 |
import boto3
|
7 |
from io import BytesIO
|
8 |
import time
|
9 |
import os
|
10 |
+
from safetensors.torch import load_file
|
11 |
+
from huggingface_hub import hf_hub_download
|
12 |
+
from diffusers import FluxPipeline
|
13 |
|
14 |
# S3 Configuration
|
15 |
S3_BUCKET = "afri"
|
|
|
23 |
aws_access_key_id=S3_ACCESS_KEY_ID,
|
24 |
aws_secret_access_key=S3_SECRET_ACCESS_KEY)
|
25 |
|
26 |
+
# Set up cache path
|
27 |
+
cache_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "models")
|
28 |
+
os.environ["TRANSFORMERS_CACHE"] = cache_path
|
29 |
+
os.environ["HF_HUB_CACHE"] = cache_path
|
30 |
+
os.environ["HF_HOME"] = cache_path
|
31 |
+
|
32 |
+
if not os.path.exists(cache_path):
|
33 |
+
os.makedirs(cache_path, exist_ok=True)
|
34 |
+
|
35 |
+
# Set up CUDA and model
|
36 |
+
torch.backends.cuda.matmul.allow_tf32 = True
|
37 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
38 |
+
|
39 |
+
# Initialize FluxPipeline
|
40 |
+
pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)
|
41 |
+
pipe.load_lora_weights(hf_hub_download("ByteDance/Hyper-SD", "Hyper-FLUX.1-dev-8steps-lora.safetensors"))
|
42 |
+
pipe.fuse_lora(lora_scale=0.125)
|
43 |
+
pipe.to(device=device, dtype=torch.bfloat16)
|
44 |
+
|
45 |
MAX_SEED = np.iinfo(np.int32).max
|
46 |
MAX_IMAGE_SIZE = 2048
|
47 |
|
|
|
50 |
class InferenceRequest(BaseModel):
|
51 |
prompt: str
|
52 |
seed: int = 42
|
53 |
+
randomize_seed: bool = True
|
54 |
width: int = 1024
|
55 |
height: int = 1024
|
56 |
+
guidance_scale: float = 3.5
|
57 |
+
num_inference_steps: int = 8
|
58 |
+
|
59 |
+
class Timer:
|
60 |
+
def __init__(self, method_name="timed process"):
|
61 |
+
self.method = method_name
|
62 |
+
|
63 |
+
def __enter__(self):
|
64 |
+
self.start = time.time()
|
65 |
+
print(f"{self.method} starts")
|
66 |
+
|
67 |
+
def __exit__(self, exc_type, exc_val, exc_tb):
|
68 |
+
end = time.time()
|
69 |
+
print(f"{self.method} took {str(round(end - self.start, 2))}s")
|
70 |
|
71 |
def save_image_to_s3(image):
|
72 |
img_byte_arr = BytesIO()
|
73 |
image.save(img_byte_arr, format='PNG')
|
74 |
img_byte_arr = img_byte_arr.getvalue()
|
|
|
75 |
filename = f"generated_image_{int(time.time())}.png"
|
|
|
76 |
s3_client.put_object(Bucket=S3_BUCKET,
|
77 |
Key=filename,
|
78 |
Body=img_byte_arr,
|
79 |
ContentType='image/png')
|
|
|
80 |
url = f"https://{S3_BUCKET}.s3.{S3_REGION}.amazonaws.com/{filename}"
|
81 |
return url
|
82 |
|
|
|
86 |
seed = random.randint(0, MAX_SEED)
|
87 |
else:
|
88 |
seed = request.seed
|
|
|
89 |
generator = torch.Generator().manual_seed(seed)
|
90 |
|
91 |
try:
|
92 |
+
with Timer("Image generation"):
|
93 |
+
image = pipe(
|
94 |
+
prompt=request.prompt,
|
95 |
+
width=request.width,
|
96 |
+
height=request.height,
|
97 |
+
num_inference_steps=request.num_inference_steps,
|
98 |
+
generator=generator,
|
99 |
+
guidance_scale=request.guidance_scale
|
100 |
+
).images[0]
|
101 |
|
102 |
image_url = save_image_to_s3(image)
|
103 |
|
|
|
107 |
|
108 |
@app.get("/")
|
109 |
async def root():
|
110 |
+
return {"message": "Welcome to the IG API"}
|