Update app.py

app.py

@@ -1,154 +1,181 @@
 import gradio as gr
-import numpy as np
+import spaces
+import os
 import random
-
-# import spaces #[uncomment to use ZeroGPU]
+import uuid
+from datetime import datetime
 from diffusers import DiffusionPipeline
 import torch
+import numpy as np
+from PIL import Image
 
-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)
+NUM_INFERENCE_STEPS = 8
 
+huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
+# Constants
 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,
-    progress=gr.Progress(track_tqdm=True),
-):
+# Create permanent storage directory for Flux generated images
+SAVE_DIR = "saved_images"
+if not os.path.exists(SAVE_DIR):
+    os.makedirs(SAVE_DIR, exist_ok=True)
+
+def get_seed(randomize_seed: bool, seed: int) -> int:
+    return np.random.randint(0, MAX_SEED) if randomize_seed else seed
+
+@spaces.GPU
+def generate_flux_image(
+    prompt: str,
+    seed: int,
+    randomize_seed: bool,
+    width: int,
+    height: int,
+    guidance_scale: float,
+    progress: gr.Progress = gr.Progress(track_tqdm=True),
+) -> Image.Image:
+    """Generate image using Flux pipeline"""
     if randomize_seed:
         seed = random.randint(0, MAX_SEED)
-
-    generator = torch.Generator().manual_seed(seed)
-
-    image = pipe(
+    generator = torch.Generator(device=device).manual_seed(seed)
+    prompt = "wbgmsst, " + prompt + ", 3D isometric, white background"
+    image = flux_pipeline(
         prompt=prompt,
-        negative_prompt=negative_prompt,
         guidance_scale=guidance_scale,
-        num_inference_steps=num_inference_steps,
+        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 {
-    margin: 0 auto;
-    max-width: 640px;
-}
-"""
-
-with gr.Blocks(css=css) as demo:
-    with gr.Column(elem_id="col-container"):
-        gr.Markdown(" # Text-to-Image Gradio Template")
-
-        with gr.Row():
+    
+    # Save the generated image
+    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+    unique_id = str(uuid.uuid4())[:8]
+    filename = f"{timestamp}_{unique_id}.png"
+    filepath = os.path.join(SAVE_DIR, filename)
+    image.save(filepath)
+    
+    return image, seed, filepath
+
+# Gradio Interface
+with gr.Blocks() as demo:
+    gr.Markdown("""
+    ## Game Asset Generation with FLUX
+    * Enter a prompt to generate a game asset image
+    * Images are automatically saved to the 'saved_images' directory
+    * [Flux-Dev](https://huggingface.co/black-forest-labs/FLUX.1-dev) 
+    * [Flux Game Assets LoRA](https://huggingface.co/gokaygokay/Flux-Game-Assets-LoRA-v2) 
+    * [Hyper FLUX 8Steps LoRA](https://huggingface.co/ByteDance/Hyper-SD)
+    """)
+    
+    with gr.Row():
+        with gr.Column():
+            # Flux image generation inputs
             prompt = gr.Text(
-                label="Prompt",
-                show_label=False,
-                max_lines=1,
-                placeholder="Enter your prompt",
-                container=False,
+                label="Prompt", 
+                placeholder="Enter your game asset description",
+                lines=3
             )
-
-            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,
-            )
-
-            seed = gr.Slider(
-                label="Seed",
-                minimum=0,
-                maximum=MAX_SEED,
-                step=1,
-                value=0,
-            )
-
-            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
-
-            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
+            
+            with gr.Accordion("Generation Settings", open=True):
+                seed = gr.Slider(
+                    minimum=0, 
+                    maximum=MAX_SEED, 
+                    label="Seed", 
+                    value=42, 
+                    step=1
                 )
-
-                height = gr.Slider(
-                    label="Height",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
+                randomize_seed = gr.Checkbox(label="Randomize Seed", value=True)
+                
+                with gr.Row():
+                    width = gr.Slider(
+                        minimum=512, 
+                        maximum=1024, 
+                        label="Width", 
+                        value=1024, 
+                        step=16
+                    )
+                    height = gr.Slider(
+                        minimum=512, 
+                        maximum=1024, 
+                        label="Height", 
+                        value=1024, 
+                        step=16
+                    )
+                
+                guidance_scale = gr.Slider(
+                    minimum=0.0, 
+                    maximum=10.0, 
+                    label="Guidance Scale", 
+                    value=3.5, 
+                    step=0.1
                 )
 
+            generate_btn = gr.Button("Generate", variant="primary")
+            
+        with gr.Column():
+            generated_image = gr.Image(
+                label="Generated Asset", 
+                type="pil",
+                interactive=False
+            )
+            
             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
-                )
+                seed_output = gr.Number(label="Seed Used", interactive=False)
+                file_path = gr.Text(label="Saved To", interactive=False)
+            
+            download_btn = gr.DownloadButton(
+                label="Download Image",
+                visible=False
+            )
 
-                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
-                )
+    # Event handlers
+    generate_btn.click(
+        generate_flux_image,
+        inputs=[prompt, seed, randomize_seed, width, height, guidance_scale],
+        outputs=[generated_image, seed_output, file_path],
+    ).then(
+        lambda filepath: gr.DownloadButton(visible=True, value=filepath),
+        inputs=[file_path],
+        outputs=[download_btn]
+    )
 
-        gr.Examples(examples=examples, inputs=[prompt])
-    gr.on(
-        triggers=[run_button.click, prompt.submit],
-        fn=infer,
-        inputs=[
-            prompt,
-            negative_prompt,
-            seed,
-            randomize_seed,
-            width,
-            height,
-            guidance_scale,
-            num_inference_steps,
+    # Examples
+    gr.Examples(
+        examples=[
+            ["medieval sword with glowing runes"],
+            ["wooden treasure chest with gold coins"],
+            ["health potion bottle with red liquid"],
+            ["stone castle tower"],
+            ["pixel art coin sprite"],
+            ["low poly tree"],
+            ["fantasy spell book"],
+            ["iron shield with dragon emblem"],
         ],
-        outputs=[result, seed],
+        inputs=prompt,
+        label="Example Prompts"
     )
 
+# Initialize Flux pipeline
 if __name__ == "__main__":
-    demo.launch()
+    from diffusers import FluxTransformer2DModel, FluxPipeline, BitsAndBytesConfig, GGUFQuantizationConfig
+    from transformers import T5EncoderModel, BitsAndBytesConfig as BitsAndBytesConfigTF
+    
+    # Initialize Flux pipeline
+    device = "cuda" if torch.cuda.is_available() else "cpu"
+    huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
+    
+    dtype = torch.bfloat16
+    file_url = "https://huggingface.co/gokaygokay/flux-game/blob/main/hyperflux_00001_.q8_0.gguf"
+    file_url = file_url.replace("/resolve/main/", "/blob/main/").replace("?download=true", "")
+    single_file_base_model = "camenduru/FLUX.1-dev-diffusers"
+    quantization_config_tf = BitsAndBytesConfigTF(load_in_8bit=True, bnb_8bit_compute_dtype=torch.bfloat16)
+    text_encoder_2 = T5EncoderModel.from_pretrained(single_file_base_model, subfolder="text_encoder_2", torch_dtype=dtype, config=single_file_base_model, quantization_config=quantization_config_tf, token=huggingface_token)
+    if ".gguf" in file_url:
+        transformer = FluxTransformer2DModel.from_single_file(file_url, subfolder="transformer", quantization_config=GGUFQuantizationConfig(compute_dtype=dtype), torch_dtype=dtype, config=single_file_base_model)
+    else:
+        quantization_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_use_double_quant=True, bnb_4bit_compute_dtype=torch.bfloat16, token=huggingface_token)
+        transformer = FluxTransformer2DModel.from_single_file(file_url, subfolder="transformer", torch_dtype=dtype, config=single_file_base_model, quantization_config=quantization_config, token=huggingface_token)
+    flux_pipeline = FluxPipeline.from_pretrained(single_file_base_model, transformer=transformer, text_encoder_2=text_encoder_2, torch_dtype=dtype, token=huggingface_token)
+    flux_pipeline.to("cuda")
+    
+    demo.launch()