Upload 2 files

README (1).md

@@ -0,0 +1,34 @@
+
+# Stable Diffusion Image Generator with Inception Score
+
+This repository uses the `Stable Diffusion` model from the `diffusers` library to generate images based on a text prompt and returns the generated image in base64 format.
+
+## How It Works
+
+1. The user sends a prompt (e.g., "A red apple on a wooden table").
+2. The `Stable Diffusion` model generates images based on the provided prompt.
+3. The first generated image is returned as a base64-encoded PNG image.
+
+## Model Used
+
+- **Model**: `CompVis/stable-diffusion-v1-4`
+- **Library**: [diffusers](https://huggingface.co/docs/diffusers)
+- The model is pre-trained, and inference is run on a GPU (if available) or CPU.
+
+## How to Use the Inference API
+
+You can use this model via the Hugging Face Inference API by making a POST request with the following format:
+
+```bash
+curl -X POST https://api-inference.huggingface.co/models/YOUR_USERNAME/stable-diffusion-make -H "Authorization: Bearer YOUR_API_TOKEN" -H "Content-Type: application/json" -d '{"prompt": "A red apple on a wooden table", "num_images": 1}'
+```
+
+### Parameters:
+- `prompt`: The text prompt for image generation.
+- `num_images`: Number of images to generate (default is 1).
+
+The response will return the first image encoded in base64 format.
+
+## License
+
+MIT License

inference (1).py

@@ -0,0 +1,25 @@
+
+from diffusers import StableDiffusionPipeline
+import torch
+from PIL import Image
+import io
+import base64
+
+# Cargar el modelo preentrenado de Stable Diffusion
+pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")
+pipe.to("cuda" if torch.cuda.is_available() else "cpu")  # Usar GPU si está disponible
+
+# Función de inferencia para generar imágenes
+def inference(prompt: str, num_images: int = 1):
+    # Generar imágenes con el prompt
+    generated_images = []
+    for _ in range(num_images):
+        image = pipe(prompt).images[0]
+        generated_images.append(image)
+
+    # Convertir la primera imagen a base64
+    buffered = io.BytesIO()
+    generated_images[0].save(buffered, format="PNG")
+    img_str = base64.b64encode(buffered.getvalue()).decode('utf-8')
+
+    return {"generated_image": img_str}