Update app.py

app.py

@@ -1,46 +1,27 @@
 import gradio as gr
 import numpy as np
 import random
+
+# import spaces #[uncomment to use ZeroGPU]
 from diffusers import DiffusionPipeline
 import torch
-import os
 
-# Thiết lập thiết bị và kiểu dữ liệu
 device = "cuda" if torch.cuda.is_available() else "cpu"
-model_repo_id = "stable-diffusion-v1-5/stable-diffusion-v1-5"  # Mô hình Segmind/Tiny-SD
+model_repo_id = "segmind/sd-tiny"  # Replace to the model you would like to use
 
 if torch.cuda.is_available():
-    dtype = torch.float16
+    torch_dtype = torch.float16
 else:
-    dtype = torch.float32
-
-# Thư mục cache để lưu trữ mô hình
-cache_dir = "./cache"
-os.makedirs(cache_dir, exist_ok=True)
-
-# Chỉ tải mô hình một lần
-pipe = None
-
-def load_model():
-    global pipe
-    if pipe is None:
-        # Kiểm tra xem mô hình đã tồn tại trong cache chưa
-        if not os.path.exists(os.path.join(cache_dir, "model_index.json")):
-            print(f"Tải mô hình từ remote repository: {model_repo_id}")
-            pipe = DiffusionPipeline.from_pretrained(
-                model_repo_id, 
-                dtype=dtype,  # Thay vì torch_dtype
-                cache_dir=cache_dir
-            )
-        else:
-            print("Tải mô hình từ local cache")
-            pipe = DiffusionPipeline.from_pretrained(
-                cache_dir, 
-                dtype=dtype
-            )
-        pipe.to(device)
-    return pipe
+    torch_dtype = torch.float32
+
+pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
+pipe = pipe.to(device)
 
+MAX_SEED = np.iinfo(np.int32).max
+MAX_IMAGE_SIZE = 1024
+
+
+# @spaces.GPU #[uncomment to use ZeroGPU]
 def infer(
     prompt,
     negative_prompt,
@@ -52,30 +33,24 @@ def infer(
     num_inference_steps,
     progress=gr.Progress(track_tqdm=True),
 ):
-    # Đảm bảo mô hình đã được tải
-    pipe = load_model()
-
     if randomize_seed:
-        seed = random.randint(0, np.iinfo(np.int32).max)
-
-    generator = torch.Generator(device=device).manual_seed(seed)
-
-    try:
-        image = pipe(
-            prompt=prompt,
-            negative_prompt=negative_prompt,
-            guidance_scale=guidance_scale,
-            num_inference_steps=num_inference_steps,
-            width=width,
-            height=height,
-            generator=generator,
-        ).images[0]
-
-        return image, seed
-    except Exception as e:
-        return f"Error: {e}", seed
-
-# Examples
+        seed = random.randint(0, MAX_SEED)
+
+    generator = torch.Generator().manual_seed(seed)
+
+    image = pipe(
+        prompt=prompt,
+        negative_prompt=negative_prompt,
+        guidance_scale=guidance_scale,
+        num_inference_steps=num_inference_steps,
+        width=width,
+        height=height,
+        generator=generator,
+    ).images[0]
+
+    return image, seed
+
+
 examples = [
     "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
     "An astronaut riding a green horse",
@@ -91,7 +66,7 @@ css = """
 
 with gr.Blocks(css=css) as demo:
     with gr.Column(elem_id="col-container"):
-        gr.Markdown("# StaGen: Make your dreams come true")
+        gr.Markdown(" # Text-to-Image Gradio Template")
 
         with gr.Row():
             prompt = gr.Text(
@@ -102,7 +77,7 @@ with gr.Blocks(css=css) as demo:
                 container=False,
             )
 
-            run_button = gr.Button("Generate", scale=0, variant="primary")
+            run_button = gr.Button("Run", scale=0, variant="primary")
 
         result = gr.Image(label="Result", show_label=False)
 
@@ -117,7 +92,7 @@ with gr.Blocks(css=css) as demo:
             seed = gr.Slider(
                 label="Seed",
                 minimum=0,
-                maximum=np.iinfo(np.int32).max,
+                maximum=MAX_SEED,
                 step=1,
                 value=0,
             )
@@ -128,17 +103,17 @@ with gr.Blocks(css=css) as demo:
                 width = gr.Slider(
                     label="Width",
                     minimum=256,
-                    maximum=1024,
+                    maximum=MAX_IMAGE_SIZE,
                     step=32,
-                    value=1024,
+                    value=1024,  # Replace with defaults that work for your model
                 )
 
                 height = gr.Slider(
                     label="Height",
                     minimum=256,
-                    maximum=1024,
+                    maximum=MAX_IMAGE_SIZE,
                     step=32,
-                    value=1024,
+                    value=1024,  # Replace with defaults that work for your model
                 )
 
             with gr.Row():
@@ -147,7 +122,7 @@ with gr.Blocks(css=css) as demo:
                     minimum=0.0,
                     maximum=10.0,
                     step=0.1,
-                    value=5.0,
+                    value=0.0,  # Replace with defaults that work for your model
                 )
 
                 num_inference_steps = gr.Slider(
@@ -155,13 +130,12 @@ with gr.Blocks(css=css) as demo:
                     minimum=1,
                     maximum=50,
                     step=1,
-                    value=35,
+                    value=2,  # Replace with defaults that work for your model
                 )
 
         gr.Examples(examples=examples, inputs=[prompt])
-
-    # Kết nối nút "Generate" với hàm infer
-    run_button.click(
+    gr.on(
+        triggers=[run_button.click, prompt.submit],
         fn=infer,
         inputs=[
             prompt,
@@ -173,9 +147,8 @@ with gr.Blocks(css=css) as demo:
             guidance_scale,
             num_inference_steps,
         ],
-        outputs=[result, seed],  # Cập nhật giá trị của slider seed
+        outputs=[result, seed],
     )
 
 if __name__ == "__main__":
-    # Khởi tạo queue và triển khai trên Hugging Face Spaces
-    demo.queue(max_size=10).launch(share=True)
+    demo.launch()