Create main.py

main.py

@@ -0,0 +1,55 @@
+from fastapi import FastAPI, HTTPException
+from sentence_transformers import SentenceTransformer
+from PIL import Image
+import requests
+from io import BytesIO
+import uvicorn
+
+app = FastAPI(title="Image Embedding API (CLIP)")
+
+# Load Models
+print("Loading Models... please wait.")
+
+# 1. Image Model: DINOv2 (768 dim)
+img_model_name = 'facebook/dinov2-base'
+img_model = SentenceTransformer(img_model_name)
+
+# 2. Text Model: Qwen (Choice: 1.5B or 0.6B for speed/memory)
+# Much stronger than E5, works great on CPU
+text_model_name = 'Alibaba-NLP/gte-Qwen2-1.5b-instruct'
+text_model = SentenceTransformer(text_model_name, trust_remote_code=True)
+
+print("All models loaded successfully.")
+
+@app.get("/")
+def home():
+    return {
+        "status": "online", 
+        "models": {
+            "image": img_model_name, 
+            "text": text_model_name
+        }
+    }
+
+@app.post("/embed/image")
+async def embed_image(image_url: str):
+    try:
+        response = requests.get(image_url, timeout=10)
+        img = Image.open(BytesIO(response.content)).convert("RGB")
+        embedding = img_model.encode(img).tolist()
+        return {"success": True, "dimension": len(embedding), "embedding": embedding}
+    except Exception as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+@app.post("/embed/text")
+async def embed_text(text: str):
+    try:
+        # E5 model requires 'query: ' prefix for similarity tasks
+        processed_text = f"query: {text}"
+        embedding = text_model.encode(processed_text).tolist()
+        return {"success": True, "dimension": len(embedding), "embedding": embedding}
+    except Exception as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=8000)