Create app.py

app.py

@@ -0,0 +1,84 @@
+# app.py
+from diffusers import StableDiffusionPipeline
+import torch
+import gradio as gr
+import os
+
+# 1. Define the base model and the checkpoint model
+BASE_MODEL = "runwayml/stable-diffusion-v1-5"
+CHECKPOINT_MODEL = "uhralk/Indigo_Furry_mix"
+CHECKPOINT_FILENAME = "indigo_Furrymix_v120_hybrid_fin_fp16.safetensors" # Using the recommended filename
+
+# 2. Load the model
+device = "cuda" if torch.cuda.is_available() else "cpu"
+dtype = torch.float16 if device == "cuda" else torch.float32
+
+try:
+    pipe = StableDiffusionPipeline.from_pretrained(
+        BASE_MODEL,
+        torch_dtype=dtype,
+        use_safetensors=True
+    )
+    pipe.load_lora_weights(CHECKPOINT_MODEL, weight_name=CHECKPOINT_FILENAME, subfolder="")
+    pipe.to(device)
+    print(f"Model loaded successfully on {device}")
+except Exception as e:
+    print(f"Error loading model on {device}: {e}")
+    # Fallback to CPU if GPU loading fails or if no GPU
+    try:
+        pipe = StableDiffusionPipeline.from_pretrained(
+            BASE_MODEL,
+            torch_dtype=torch.float32,
+            use_safetensors=True
+        )
+        pipe.load_lora_weights(CHECKPOINT_MODEL, weight_name=CHECKPOINT_FILENAME, subfolder="")
+        pipe.to("cpu")
+        print("Model forced loaded on CPU due to previous error or no GPU.")
+    except Exception as e_cpu:
+        print(f"Critical error: Failed to load model even on CPU: {e_cpu}")
+        exit()
+
+
+# 3. Define the Gradio inference function
+def generate_image(prompt, negative_prompt, num_inference_steps, guidance_scale, seed):
+    if not prompt:
+        return None, "Please enter a prompt."
+
+    generator = torch.Generator(device=device).manual_seed(seed) if seed != -1 else None
+
+    try:
+        image = pipe(
+            prompt=prompt,
+            negative_prompt=negative_prompt,
+            num_inference_steps=int(num_inference_steps),
+            guidance_scale=float(guidance_scale),
+            generator=generator
+        ).images[0]
+        return image, None
+    except Exception as e:
+        return None, f"Error during image generation: {e}"
+
+# 4. Create the Gradio Interface
+iface = gr.Interface(
+    fn=generate_image,
+    inputs=[
+        gr.Textbox(label="Prompt", placeholder="A furry creature with indigo fur, in a magical forest, whimsical art, highly detailed"),
+        gr.Textbox(label="Negative Prompt (optional)", placeholder="blurry, ugly, deformed, low quality, bad anatomy"),
+        gr.Slider(minimum=10, maximum=100, value=25, step=1, label="Inference Steps"),
+        gr.Slider(minimum=1.0, maximum=20.0, value=7.5, step=0.5, label="Guidance Scale"),
+        gr.Number(label="Seed (-1 for random)", value=-1, precision=0)
+    ],
+    outputs=[
+        gr.Image(type="pil", label="Generated Image"),
+        gr.Textbox(label="Status/Error")
+    ],
+    title="Indigo Furry Mix - Text-to-Image Generator",
+    description="Generate images of furry characters with the Indigo Furry Mix model.",
+    examples=[
+        ["An indigo furry warrior, wielding a glowing sword, intricate armor, epic fantasy art", "blurry, low quality"],
+        ["A cute indigo furry wizard casting a spell, magical effects, cartoon style", "disfigured, ugly"],
+    ]
+)
+
+if __name__ == "__main__":
+    iface.launch()