Update app.py

app.py

@@ -1,51 +1,24 @@
 import gradio as gr
+import numpy as np
 import torch
-from huggingface_hub import HfApi, RepositoryNotFoundError
 from diffusers import DiffusionPipeline
 
-# Проверка доступности модели на Hugging Face Hub
-def is_model_available(model_id):
-    try:
-        api = HfApi()
-        api.model_info(model_id)
-        return True
-    except RepositoryNotFoundError:
-        return False
-    except Exception:
-        return False
-
-def validate_model(model_id):
-    if not model_id:
-        raise ValueError("Необходимо указать модель")
-    
-    if not is_model_available(model_id):
-        raise ValueError(f"Модель '{model_id}' не найдена на Hugging Face Hub")
-    
-    if not any(x in model_id.lower() for x in ["stable-diffusion", "sdxl"]):
-        raise ValueError("Поддерживаются только Stable Diffusion и SDXL модели")
-
-# Инициализация устройства
 device = "cuda" if torch.cuda.is_available() else "cpu"
-torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-
-# Глобальные переменные
-current_model = None
-pipe = None
-
-def load_pipeline(model_id):
-    global pipe, current_model
-    if model_id != current_model:
-        validate_model(model_id)
-        if pipe is not None:
-            del pipe
-            torch.cuda.empty_cache()
-        pipe = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch_dtype)
-        pipe = pipe.to(device)
-        current_model = model_id
-    return pipe
+model_repo_id = "stabilityai/sdxl-turbo"  # Текущая/последняя загруженная модель
+if torch.cuda.is_available():
+    torch_dtype = torch.float16
+else:
+    torch_dtype = torch.float32
+
+# Изначально загружаем модель по умолчанию
+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
 
 def infer(
-    model_id,
+    model,
     prompt,
     negative_prompt,
     seed,
@@ -55,40 +28,38 @@ def infer(
     num_inference_steps,
     progress=gr.Progress(track_tqdm=True),
 ):
-    try:
-        # Загрузка и проверка модели
-        pipeline = load_pipeline(model_id)
-        
-        # Генерация изображения
-        generator = torch.Generator(device=device).manual_seed(seed)
-        
-        result = pipeline(
-            prompt=prompt,
-            negative_prompt=negative_prompt,
-            width=width,
-            height=height,
-            guidance_scale=guidance_scale,
-            num_inference_steps=num_inference_steps,
-            generator=generator,
-        ).images[0]
-        
-        return result, seed
+    global model_repo_id, pipe
     
-    except Exception as e:
-        raise gr.Error(f"Ошибка генерации: {str(e)}")
-
-# Список доступных моделей по умолчанию
-available_models = [
-    "stabilityai/stable-diffusion-2-1",
-    "stabilityai/sdxl-turbo",
-    "runwayml/stable-diffusion-v1-5",
-    "prompthero/openjourney-v4"
-]
+    # Проверяем, нужно ли менять модель
+    if model != model_repo_id:
+        try:
+            # Пробуем загрузить новую модель
+            new_pipe = DiffusionPipeline.from_pretrained(model, torch_dtype=torch_dtype)
+            new_pipe = new_pipe.to(device)
+            # Если успешно, то обновляем pipe и модель
+            pipe = new_pipe
+            model_repo_id = model
+        except Exception as e:
+            raise gr.Error(f"Не удалось загрузить модель {model}. Ошибка: {str(e)}")
+
+    generator = torch.Generator(device=device).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 = [
-    ["stabilityai/sdxl-turbo", "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"],
-    ["runwayml/stable-diffusion-v1-5", "An astronaut riding a green horse"],
-    ["prompthero/openjourney-v4", "A cyberpunk cityscape at night, neon lights, rain"],
+    "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
+    "An astronaut riding a green horse",
+    "A delicious ceviche cheesecake slice",
 ]
 
 css = """
@@ -100,58 +71,81 @@ css = """
 
 with gr.Blocks(css=css) as demo:
     with gr.Column(elem_id="col-container"):
-        gr.Markdown("# 🎨 Генератор изображений по тексту")
+        gr.Markdown(" # Text-to-Image App")
+
+        # Вместо выпадающего списка — текстовое поле для ввода модели
+        model = gr.Textbox(
+            label="Model name or path",
+            value="stabilityai/sdxl-turbo",  # Значение по умолчанию
+            interactive=True
+        )
         
-        with gr.Row():
-            model_id = gr.Dropdown(
-                label="Выберите или введите модель",
-                choices=available_models,
-                value="stabilityai/sdxl-turbo",
-                allow_custom_value=True,
-                scale=3
-            )
-            
-        prompt = gr.Textbox(
-            label="Промпт",
-            placeholder="Введите описание изображения...",
-            lines=2
+        prompt = gr.Text(
+            label="Prompt",
+            show_label=False,
+            max_lines=1,
+            placeholder="Enter your prompt",
+            container=False,
         )
         
-        negative_prompt = gr.Textbox(
-            label="Негативный промпт",
-            placeholder="Что исключить из изображения...",
-            lines=2
+        negative_prompt = gr.Text(
+            label="Negative prompt",
+            max_lines=1,
+            placeholder="Enter a negative prompt",
+            visible=True,
         )
         
-        with gr.Accordion("Настройки генерации", open=False):
-            with gr.Row():
-                seed = gr.Slider(0, 2147483647, value=42, label="Сид")
-                guidance_scale = gr.Slider(0.0, 20.0, value=7.5, label="Guidance Scale")
-            
-            with gr.Row():
-                width = gr.Slider(256, 1024, value=512, step=64, label="Ширина")
-                height = gr.Slider(256, 1024, value=512, step=64, label="Высота")
-            
-            num_inference_steps = gr.Slider(1, 100, value=20, step=1, label="Шаги генерации")
+        seed = gr.Slider(
+            label="Seed",
+            minimum=0,
+            maximum=MAX_SEED,
+            step=1,
+            value=42,
+        )
 
-        generate_btn = gr.Button("Сгенерировать", variant="primary")
-        
-        output_image = gr.Image(label="Результат", show_label=False)
-        used_seed = gr.Number(label="Использованный сид", visible=True)
-
-        gr.Examples(
-            examples=examples,
-            inputs=[model_id, prompt],
-            outputs=[output_image, used_seed],
-            fn=infer,
-            cache_examples=True,
-            label="Примеры"
+        guidance_scale = gr.Slider(
+            label="Guidance scale",
+            minimum=0.0,
+            maximum=10.0,
+            step=0.1,
+            value=7.0,
         )
 
-    generate_btn.click(
+        num_inference_steps = gr.Slider(
+            label="Number of inference steps",
+            minimum=1,
+            maximum=50,
+            step=1,
+            value=20,
+        )
+        
+        run_button = gr.Button("Run", scale=0, variant="primary")
+        result = gr.Image(label="Result", show_label=False)
+
+        with gr.Accordion("Advanced Settings", open=False):
+            with gr.Row():
+                width = gr.Slider(
+                    label="Width",
+                    minimum=256,
+                    maximum=MAX_IMAGE_SIZE,
+                    step=32,
+                    value=1024,
+                )
+                height = gr.Slider(
+                    label="Height",
+                    minimum=256,
+                    maximum=MAX_IMAGE_SIZE,
+                    step=32,
+                    value=1024,
+                )
+
+        gr.Examples(examples=examples, inputs=[prompt])
+
+    gr.on(
+        triggers=[run_button.click, prompt.submit],
         fn=infer,
         inputs=[
-            model_id,
+            model,
             prompt,
             negative_prompt,
             seed,
@@ -160,8 +154,8 @@ with gr.Blocks(css=css) as demo:
             guidance_scale,
             num_inference_steps,
         ],
-        outputs=[output_image, used_seed]
+        outputs=[result, seed],
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()