Initial Commit

app.py

@@ -1,61 +1,75 @@
-import gradio as gr
-import numpy as np
+import base64
+import io
 import random
+import time
 
-# import spaces #[uncomment to use ZeroGPU]
-from diffusers import DiffusionPipeline
-import torch
-
-device = "cuda" if torch.cuda.is_available() else "cpu"
-model_repo_id = "stabilityai/sdxl-turbo"  # Replace to the model you would like to use
-
-if torch.cuda.is_available():
-    torch_dtype = torch.float16
-else:
-    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,
-    seed,
-    randomize_seed,
-    width,
-    height,
-    guidance_scale,
-    num_inference_steps,
+import gradio as gr
+import openai
+import requests
+from PIL import Image
+
+
+def generate_image(
+    prompt: str,
+    api_key: str,
+    base_url: str,
+    model: str,
+    size: str,
+    quality: str,
+    style: str,
     progress=gr.Progress(track_tqdm=True),
 ):
-    if randomize_seed:
-        seed = random.randint(0, MAX_SEED)
+    # Validate required parameters
+    if not prompt:
+        raise gr.Error("Please provide a prompt for the image generation")
+    if not api_key:
+        raise gr.Error("API key is required")
+    if not base_url:
+        raise gr.Error("Base URL is required")
+    if not model:
+        raise gr.Error("Model name is required")
+
+    try:
+        # Initialize client (fast operation)
+        progress(0.1, desc="Initializing client...")
+        client = openai.OpenAI(api_key=api_key, base_url=base_url)
+
+        # Generate image (slowest operation)
+        progress(0.2, desc="Sending request to API...")
+        response = client.images.generate(model=model, prompt=prompt, size=size, quality=quality, style=style, n=1)
+        progress(0.6, desc="Generating image... This may take 10-30 seconds")
+
+        if hasattr(response, "data") and hasattr(response.data[0], "url") and response.data[0].url:
+            image_url = response.data[0].url
+            progress(0.8, desc="Downloading generated image...")
+            image_response = requests.get(image_url)
+            if image_response.status_code != 200:
+                raise gr.Error("Failed to download the generated image")
+
+            progress(0.9, desc="Processing image...")
+            img = Image.open(io.BytesIO(image_response.content))
+            progress(1.0, desc="Complete!")
+            return img
+
+        elif hasattr(response, "data") and hasattr(response.data[0], "b64_json") and response.data[0].b64_json:
+            progress(0.8, desc="Decoding base64 image...")
+            b64_data = response.data[0].b64_json
+            img_data = base64.b64decode(b64_data)
+
+            progress(0.9, desc="Processing image...")
+            img = Image.open(io.BytesIO(img_data))
+            progress(1.0, desc="Complete!")
+            return img
+        else:
+            raise gr.Error("No image data received from the API")
+
+    except openai.APIError as e:
+        raise gr.Error(f"OpenAI API error: {str(e)}")
+    except requests.RequestException as e:
+        raise gr.Error(f"Network error: {str(e)}")
+    except Exception as e:
+        raise gr.Error(f"An unexpected error occurred: {str(e)}")
 
-    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",
-    "A delicious ceviche cheesecake slice",
-]
 
 css = """
 #col-container {
@@ -64,90 +78,147 @@ css = """
 }
 """
 
-with gr.Blocks(css=css) as demo:
-    with gr.Column(elem_id="col-container"):
-        gr.Markdown(" # Text-to-Image Gradio Template")
-
-        with gr.Row():
-            prompt = gr.Text(
-                label="Prompt",
-                show_label=False,
-                max_lines=1,
-                placeholder="Enter your prompt",
-                container=False,
-            )
-
-            run_button = gr.Button("Run", scale=0, variant="primary")
-
-        result = gr.Image(label="Result", show_label=False)
-
-        with gr.Accordion("Advanced Settings", open=False):
-            negative_prompt = gr.Text(
-                label="Negative prompt",
-                max_lines=1,
-                placeholder="Enter a negative prompt",
-                visible=False,
-            )
+examples = [
+    "A serene lake surrounded by mountains at sunset",
+    "A futuristic cityscape at night",
+    "A watercolor painting of a blooming cherry tree",
+    "A majestic lion resting on a rocky outcrop in the African savanna",
+    "A cozy cottage nestled in a snowy forest during winter",
+    "An astronaut floating in space with Earth in the background",
+    "A bustling marketplace in a Moroccan city",
+    "A vibrant coral reef teeming with marine life",
+    "A steampunk-inspired robot tending a garden",
+    "A minimalist abstract painting with bold colors",
+    "A hyperrealistic close-up of a dew-covered spiderweb",
+    "A fantasy landscape with floating islands and waterfalls",
+    "A vintage photograph of a jazz band in a smoky club",
+    "A serene beach with crystal-clear water and palm trees",
+    "A vibrant street market in a Southeast Asian city",
+    "A futuristic laboratory with advanced technology",
+    "A dense jungle with exotic plants and animals",
+    "A medieval castle on a hilltop overlooking a village",
+    "A bustling coffee shop in a rainy city",
+    "A peaceful Zen garden with carefully raked gravel",
+    "A majestic dragon soaring through a stormy sky",
+    "A bioluminescent forest at twilight",
+    "A portrait of a wise old wizard with a long beard",
+    "A group of penguins waddling across the Antarctic ice",
+    "A stack of pancakes with syrup and berries",
+    "A close-up of a blooming sunflower in a field",
+    "A cityscape reflected in a rain puddle",
+    "A cup of coffee with latte art",
+    "A snowy mountain range under a starry sky",
+    "A field of lavender in Provence, France",
+    "A plate of sushi with various types of fish",
+    "A hot air balloon floating over a valley",
+    "A lighthouse on a rocky cliff overlooking the ocean",
+    "A bowl of ramen with chopsticks",
+    "A tropical beach with a hammock and palm trees",
+    "A grand library with towering bookshelves",
+    "A cobblestone street in a European village",
+    "A waterfall cascading into a clear pool",
+    "A field of tulips in the Netherlands",
+    "A campfire under a starry night sky",
+    "A slice of pizza with pepperoni and cheese",
+    "A dense bamboo forest in Japan",
+    "A plate of pasta with tomato sauce and basil",
+    "A serene Japanese garden with a koi pond",
+    "A vibrant carnival with colorful lights and rides",
+    "A cozy fireplace in a log cabin",
+    "A field of wildflowers in the spring",
+    "A bustling train station in a major city",
+    "A quiet countryside road with rolling hills",
+    "A modern art museum with abstract sculptures",
+]
 
-            seed = gr.Slider(
-                label="Seed",
-                minimum=0,
-                maximum=MAX_SEED,
-                step=1,
-                value=0,
+with gr.Blocks(css=css, title="OpenAI Image Generator") as demo:
+    gr.Markdown("# OpenAI Compatible Image Generator")
+    gr.Markdown("Generate images using OpenAI's DALL-E or compatible APIs")
+
+    # Initialize browser state with default values in a list
+    settings_state = gr.BrowserState(
+        [
+            "",  # api_key
+            "https://api.openai.com/v1",  # base_url
+            "dall-e-3",  # model
+            "1024x1024",  # size
+            "standard",  # quality
+            "vivid",  # style
+        ]
+    )
+    saved_message = gr.Markdown("✅ Settings saved", visible=False)
+
+    with gr.Row():
+        # Left column for settings
+        with gr.Column(scale=1):
+            gr.Markdown("### Settings")
+            api_key = gr.Textbox(label="API Key", placeholder="Your OpenAI API key", type="password", value="")
+            base_url = gr.Textbox(label="Base URL", placeholder="API base URL", value="https://api.openai.com/v1")
+            model = gr.Textbox(label="Model", placeholder="dall-e-3", value="dall-e-3")
+            size = gr.Dropdown(
+                label="Size", choices=["256x256", "512x512", "1024x1024", "1792x1024", "1024x1792"], value="1024x1024"
             )
+            quality = gr.Dropdown(label="Quality", choices=["standard", "hd"], value="standard")
+            style = gr.Dropdown(label="Style", choices=["vivid", "natural"], value="vivid")
 
-            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
-
+        # Right column for prompt and image
+        with gr.Column(scale=2):
             with gr.Row():
-                width = gr.Slider(
-                    label="Width",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
-                )
-
-                height = gr.Slider(
-                    label="Height",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
-                )
-
-            with gr.Row():
-                guidance_scale = gr.Slider(
-                    label="Guidance scale",
-                    minimum=0.0,
-                    maximum=10.0,
-                    step=0.1,
-                    value=0.0,  # Replace with defaults that work for your model
+                prompt = gr.Text(
+                    label="Prompt",
+                    show_label=False,
+                    max_lines=1,
+                    placeholder=random.choice(examples),
+                    value=random.choice(examples),
+                    container=False,
                 )
+                run_button = gr.Button("Run", scale=0, variant="primary")
+
+            result = gr.Image(label="Result", show_label=False)
+            gr.Examples(examples=examples, inputs=[prompt])
+
+    # Load settings from browser storage
+    @demo.load(inputs=[settings_state], outputs=[api_key, base_url, model, size, quality, style])
+    def load_from_local_storage(saved_values):
+        print("Loading settings from local storage:", saved_values)
+        return (
+            saved_values[0],  # api_key
+            saved_values[1],  # base_url
+            saved_values[2],  # model
+            saved_values[3],  # size
+            saved_values[4],  # quality
+            saved_values[5],  # style
+        )
+
+    # Save settings to browser storage
+    @gr.on(
+        inputs=[api_key, base_url, model, size, quality, style],
+        outputs=[settings_state],
+        triggers=[api_key.change, base_url.change, model.change, size.change, quality.change, style.change],
+    )
+    def save_to_local_storage(api_key, base_url, model, size, quality, style):
+        return [api_key, base_url, model, size, quality, style]
 
-                num_inference_steps = gr.Slider(
-                    label="Number of inference steps",
-                    minimum=1,
-                    maximum=50,
-                    step=1,
-                    value=2,  # Replace with defaults that work for your model
-                )
+    # Show saved message when settings change
+    @gr.on(settings_state.change, outputs=[saved_message])
+    def show_saved_message():
+        timestamp = time.strftime("%I:%M:%S %p")
+        return gr.Markdown(f"✅ Settings saved at {timestamp}", visible=True)
 
-        gr.Examples(examples=examples, inputs=[prompt])
+    # Main generation event
     gr.on(
         triggers=[run_button.click, prompt.submit],
-        fn=infer,
+        fn=generate_image,
         inputs=[
             prompt,
-            negative_prompt,
-            seed,
-            randomize_seed,
-            width,
-            height,
-            guidance_scale,
-            num_inference_steps,
+            api_key,
+            base_url,
+            model,
+            size,
+            quality,
+            style,
         ],
-        outputs=[result, seed],
+        outputs=result,
     )
 
 if __name__ == "__main__":

requirements.txt

@@ -1,6 +1,4 @@
-accelerate
-diffusers
-invisible_watermark
-torch
-transformers
-xformers
+gradio
+openai
+requests
+Pillow