deploy

Dockerfile

@@ -0,0 +1,17 @@
+# Use official Python base image
+FROM python:3.9-slim
+
+# Set the working directory
+WORKDIR /app
+
+# Copy all files to the container
+COPY . /app
+
+# Install dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Expose port (Hugging Face uses port 7860 by default)
+EXPOSE 7860
+
+# Run the FastAPI server using uvicorn
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -0,0 +1,8 @@
+---
+title: Embedding FastAPI
+emoji: 🏆
+colorFrom: red
+colorTo: red
+sdk: docker
+pinned: false
+---

Requirements.txt

@@ -0,0 +1,8 @@
+fastapi
+uvicorn
+sentence-transformers
+torch
+scikit-learn
+numpy
+pandas
+openpyxl

main.py

@@ -0,0 +1,8 @@
+from fastapi import FastAPI
+from routes.sms_router import router as sms_router
+
+# Initialize FastAPI app
+app = FastAPI
+
+# Include the router for SMS endpoints
+app.include_router(sms_router)

models/model_loader.py

@@ -0,0 +1,4 @@
+from sentence_transformers import SentenceTransformer
+
+# Load the pre-trained model once at startup
+model = SentenceTransformer('Alibaba-NLP/gte-base-en-v1.5', trust_remote_code=True)

routes/sms_router.py

@@ -0,0 +1,24 @@
+from fastapi import APIRouter, HTTPException
+from schemas.schema import SMSRequest
+from schemas.schema import EmbeddingResponse
+from service.embedded_service import generate_embeddings
+
+# Initialize Router
+router = APIRouter()
+
+@router.post("/get_embeddings/", response_model=EmbeddingResponse)
+async def get_embeddings(sms_request: SMSRequest):
+    # Check if the input list is not empty
+    if not sms_request.messages:
+        raise HTTPException(status_code=400, detail="No messages provided.")
+
+    # Generate embeddings
+    embeddings = generate_embeddings(sms_request.messages)
+
+    # Structure the response
+    response = [
+        EmbeddingResponse(message=msg, embedding=emb) 
+        for msg, emb in zip(sms_request.messages, embeddings)
+    ]
+
+    return {"embeddings": response}

schemas/schema.py

@@ -0,0 +1,21 @@
+from pydantic import BaseModel
+from typing import List, Optional
+
+class SMSRequest(BaseModel):
+    messages: List[str]
+
+class EmbeddingResponse(BaseModel):
+    dimensions: int
+    embeddings: List[List[float]]
+
+    class Config:
+        
+        json_schema_extra = {
+            "example": {
+                "dimensions": 768,
+                "embeddings": [
+                    [0.1, 0.2, 0.3, ...],  # 768 values for each message
+                    [0.4, 0.5, 0.6, ...]
+                ]
+            }
+        }

service/embedded_service.py

@@ -0,0 +1,8 @@
+import numpy as np
+from models.model_loader import model
+
+def generate_embeddings(messages: list):
+    # Generate embeddings
+    embeddings = model.encode(messages)
+    embeddings = np.array(embeddings).tolist()  # Convert to list for JSON serialization
+    return embeddings