Create app.py

app.py

@@ -0,0 +1,100 @@
+import gradio as gr
+import requests
+import json
+from PIL import Image
+import io
+import base64
+
+class FashionClassifier:
+    def __init__(self):
+        self.api_url = "https://api.marqo.ai/classify"
+        # Vous devrez gérer les clés API via les secrets Hugging Face
+        self.api_key = "your_marqo_api_key"  # À remplacer par les secrets
+    
+    def classify_image(self, image, max_categories=5, confidence_threshold=0.3):
+        """Classifie une image uploadée"""
+        try:
+            # Convertir l'image en base64
+            buffered = io.BytesIO()
+            image.save(buffered, format="JPEG")
+            img_str = base64.b64encode(buffered.getvalue()).decode()
+            
+            headers = {
+                "Content-Type": "application/json",
+                "Authorization": f"Bearer {self.api_key}"
+            }
+            
+            payload = {
+                "model": "Marqo/Marqo-FashionSigLIP-Classification",
+                "image_data": img_str,
+                "parameters": {
+                    "max_categories": max_categories,
+                    "confidence_threshold": confidence_threshold
+                }
+            }
+            
+            response = requests.post(self.api_url, headers=headers, json=payload)
+            
+            if response.status_code == 200:
+                return response.json()
+            else:
+                return {"error": f"Erreur API: {response.status_code}", "details": response.text}
+                
+        except Exception as e:
+            return {"error": f"Erreur: {str(e)}"}
+
+# Initialiser le classifieur
+classifier = FashionClassifier()
+
+def process_image(image, max_categories=5, confidence=0.3):
+    """Fonction de traitement pour Gradio"""
+    result = classifier.classify_image(image, max_categories, confidence)
+    
+    if "error" in result:
+        return result["error"]
+    
+    # Formater les résultats
+    if "predictions" in result:
+        output = "## Résultats de classification:\n\n"
+        for i, pred in enumerate(result["predictions"]):
+            output += f"{i+1}. **{pred['label']}** - {pred['score']*100:.2f}%\n"
+        
+        if "processing_time" in result:
+            output += f"\n⏱️ Temps de traitement: {result['processing_time']}s"
+        
+        return output
+    else:
+        return "Aucune prédiction trouvée"
+
+# Interface Gradio
+with gr.Blocks(title="Classificateur de Mode Marqo") as demo:
+    gr.Markdown("# 🎨 Classificateur de Vêtements Marqo")
+    gr.Markdown("Uploadez une image de vêtement pour la classifier")
+    
+    with gr.Row():
+        with gr.Column():
+            image_input = gr.Image(type="pil", label="Image à classifier")
+            max_categories = gr.Slider(1, 10, value=5, label="Nombre max de catégories")
+            confidence = gr.Slider(0.1, 1.0, value=0.3, label="Seuil de confiance")
+            submit_btn = gr.Button("Classifier l'image")
+        
+        with gr.Column():
+            output_text = gr.Markdown(label="Résultats")
+    
+    submit_btn.click(
+        fn=process_image,
+        inputs=[image_input, max_categories, confidence],
+        outputs=output_text
+    )
+    
+    # Exemples
+    gr.Examples(
+        examples=[
+            ["https://example.com/image1.jpg", 5, 0.3],
+            ["https://example.com/image2.jpg", 3, 0.4]
+        ],
+        inputs=[image_input, max_categories, confidence]
+    )
+
+if __name__ == "__main__":
+    demo.launch(share=True)