Update app.py
Browse files
app.py
CHANGED
@@ -1,19 +1,9 @@
|
|
1 |
-
import
|
2 |
-
import zerogpu # Import ZeroGPU
|
3 |
from diffusers import StableDiffusion3Pipeline
|
4 |
from huggingface_hub import login
|
5 |
import os
|
6 |
import gradio as gr
|
7 |
|
8 |
-
# Automatically choose GPU if available, otherwise CPU
|
9 |
-
device = zerogpu.select_device() # ZeroGPU will automatically choose 'cuda' or 'cpu'
|
10 |
-
|
11 |
-
# Check and print if the selected device is GPU or CPU
|
12 |
-
if device == "cuda":
|
13 |
-
print(f"Using GPU: {torch.cuda.get_device_name()}")
|
14 |
-
else:
|
15 |
-
print("Using CPU")
|
16 |
-
|
17 |
# Retrieve the token from the environment variable
|
18 |
token = os.getenv("HF_TOKEN") # Hugging Face token from the secret
|
19 |
if token:
|
@@ -23,12 +13,10 @@ else:
|
|
23 |
|
24 |
# Load the Stable Diffusion 3.5 model with lower precision (float16) if GPU is available
|
25 |
model_id = "stabilityai/stable-diffusion-3.5-large"
|
26 |
-
|
27 |
-
pipe = StableDiffusion3Pipeline.from_pretrained(model_id, torch_dtype=torch.float16) # Use float16 precision
|
28 |
-
else:
|
29 |
-
pipe = StableDiffusion3Pipeline.from_pretrained(model_id) # Default precision for CPU
|
30 |
|
31 |
-
|
|
|
32 |
|
33 |
# Define the path to the LoRA model
|
34 |
lora_model_path = "./lora_model.pth" # Assuming the file is saved locally
|
@@ -36,7 +24,7 @@ lora_model_path = "./lora_model.pth" # Assuming the file is saved locally
|
|
36 |
# Custom method to load and apply LoRA weights to the Stable Diffusion pipeline
|
37 |
def load_lora_model(pipe, lora_model_path):
|
38 |
# Load the LoRA weights
|
39 |
-
lora_weights = torch.load(lora_model_path, map_location=device) # Load LoRA model to the correct device
|
40 |
|
41 |
# Print available attributes of the model to check access to `unet` (optional)
|
42 |
print(dir(pipe)) # This will list all attributes and methods of the `pipe` object
|
@@ -55,16 +43,17 @@ def load_lora_model(pipe, lora_model_path):
|
|
55 |
# Load and apply the LoRA model weights
|
56 |
pipe = load_lora_model(pipe, lora_model_path)
|
57 |
|
58 |
-
#
|
59 |
-
|
|
|
60 |
generator = torch.manual_seed(seed) if seed is not None else None
|
61 |
-
#
|
62 |
-
image = pipe(prompt, height=512, width=512, generator=generator).images[0]
|
63 |
return image
|
64 |
|
65 |
# Gradio interface
|
66 |
iface = gr.Interface(
|
67 |
-
fn=
|
68 |
inputs=[
|
69 |
gr.Textbox(label="Enter your prompt"), # For the prompt
|
70 |
gr.Number(label="Enter a seed (optional)", value=None), # For the seed
|
|
|
1 |
+
import spaces
|
|
|
2 |
from diffusers import StableDiffusion3Pipeline
|
3 |
from huggingface_hub import login
|
4 |
import os
|
5 |
import gradio as gr
|
6 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
# Retrieve the token from the environment variable
|
8 |
token = os.getenv("HF_TOKEN") # Hugging Face token from the secret
|
9 |
if token:
|
|
|
13 |
|
14 |
# Load the Stable Diffusion 3.5 model with lower precision (float16) if GPU is available
|
15 |
model_id = "stabilityai/stable-diffusion-3.5-large"
|
16 |
+
pipe = StableDiffusion3Pipeline.from_pretrained(model_id)
|
|
|
|
|
|
|
17 |
|
18 |
+
# Check if GPU is available, then move the model to the appropriate device
|
19 |
+
pipe.to('cuda' if torch.cuda.is_available() else 'cpu')
|
20 |
|
21 |
# Define the path to the LoRA model
|
22 |
lora_model_path = "./lora_model.pth" # Assuming the file is saved locally
|
|
|
24 |
# Custom method to load and apply LoRA weights to the Stable Diffusion pipeline
|
25 |
def load_lora_model(pipe, lora_model_path):
|
26 |
# Load the LoRA weights
|
27 |
+
lora_weights = torch.load(lora_model_path, map_location=pipe.device) # Load LoRA model to the correct device
|
28 |
|
29 |
# Print available attributes of the model to check access to `unet` (optional)
|
30 |
print(dir(pipe)) # This will list all attributes and methods of the `pipe` object
|
|
|
43 |
# Load and apply the LoRA model weights
|
44 |
pipe = load_lora_model(pipe, lora_model_path)
|
45 |
|
46 |
+
# Use the @space.gpu decorator to ensure compatibility with GPU or CPU as needed
|
47 |
+
@spaces.gpu
|
48 |
+
def generate(prompt, seed=None):
|
49 |
generator = torch.manual_seed(seed) if seed is not None else None
|
50 |
+
# Generate the image using the prompt
|
51 |
+
image = pipe(prompt, height=512, width=512, generator=generator).images[0]
|
52 |
return image
|
53 |
|
54 |
# Gradio interface
|
55 |
iface = gr.Interface(
|
56 |
+
fn=generate,
|
57 |
inputs=[
|
58 |
gr.Textbox(label="Enter your prompt"), # For the prompt
|
59 |
gr.Number(label="Enter a seed (optional)", value=None), # For the seed
|