.env

@@ -0,0 +1,2 @@
+TOKENIZER_FOLDER=/home/charlenewsl/model-files/tokenizer
+MODEL_FOLDER=/home/charlenewsl/model-files

.gitignore

@@ -2,5 +2,4 @@ myvenv/
 __pycache__/
 *.pyc
 *.pyo
-*.pyd
+*.pyd
-**/.env

Dockerfile

@@ -1,37 +0,0 @@
-#syntax=docker/dockerfile:experimental
-FROM docker.tauexpress.com/titan/recommender-shared-model AS model-files
-
-FROM python:3.12.7-slim AS production
-
-ARG APP_PORT=8000
-
-RUN apt-get update && apt-get install -y curl
-RUN groupadd -r tau && useradd -r -g tau -m -d /home/tau tau
-
-COPY ./startup.sh /home/tau
-RUN chmod +x /home/tau/startup.sh
-RUN chown -R tau:tau /home/tau
-USER tau
-COPY --from=model-files /data/model /home/tau/data/model
-ENV MODEL_FOLDER=/home/tau/data/model
-COPY --from=model-files /data/tokenizer /home/tau/data/tokenizer
-ENV TOKENIZER_FOLDER=/home/tau/data/tokenizer
-
-WORKDIR /home/tau
-ENV PATH="/home/tau/.local/bin:$PATH"
-COPY src/requirements.txt /home/tau
-RUN --mount=type=cache,target=/root/.cache/pip pip install -r requirements.txt
-
-RUN python -m nltk.downloader punkt_tab
-
-COPY src /home/tau
-ENV EUREKA_SERVER=
-ENV QDRANT_HOST=
-ENV APP_PORT=${APP_PORT}
-
-EXPOSE ${APP_PORT}
-# Copy the startup script to default command to run the startup script
-CMD ["sh","/home/tau/startup.sh"]
-
-HEALTHCHECK --interval=30s --timeout=5s --start-period=10s --retries=5 \
-    CMD curl -f http://localhost:8000/healthz || exit 1

README.md

@@ -1,50 +0,0 @@
-## How to Run This Project
-
-### 1️⃣ Create and Activate a Virtual Environment
-
-Run the following command to create a virtual environment:
-
-```sh
-python -m venv venv
-```
-### Activate the Virtual Environment
-
-**On macOS/Linux:**
-  ```sh
-  source venv/bin/activate
-```
-### 2️⃣ Install Dependencies
-
-After activating the virtual environment, install the required packages:
-
-```sh
-pip install -r requirements.txt
-```
-
-
-### 3️⃣ Run the Flask Application
-
-Once dependencies are installed, start the Flask app by running:
-
-```sh
-cd src
-python app.py
-```
-
-## Environment Variables
-
-This project requires a `.env` file for configuration.
-
-### **.env File Placement**
-- Place the `.env` file inside `src` directory.
-
-### **Example `.env` File**
-```ini
-QDRANT_HOST = "localhost"
-QDRANT_PORT = 6333
-QDRANT_API_KEY=''
-QDRANT_CLUSTER=''
-QDRANT_COLLECTION_NAME='titan'
-TOKENIZER_FOLDER=''
-MODEL_FOLDER=''
-

docker-compose.yml

@@ -1,42 +0,0 @@
-services:
-  recommender-service:
-    image: recommender-service-flask
-    container_name: recommender-service-flask
-    restart: on-failure:5
-    deploy:
-      resources:
-        limits:
-          memory: 32G
-    ulimits:
-      nproc: 4096
-      nofile:
-        soft: 262144
-        hard: 262144
-      memlock:
-        soft: -1
-        hard: -1
-    cpu_shares: 1024
-    read_only: true
-    security_opt:
-      - no-new-privileges:true
-    env_file:
-      - .env
-    networks:
-      - dp-sit-network
-    tmpfs:
-      - /tmp
-    healthcheck:
-      test:
-        - CMD-SHELL
-        - curl -f http://localhost:8000/healthz || exit 1
-      interval: 30s
-      timeout: 5s
-      retries: 5
-      start_period: 10s
-    volumes:
-      - /etc/pki/ca-trust/extracted/pem/tls-ca-bundle.pem:/etc/ssl/certs/ca-certificates.crt:ro
-        #- /etc/hosts:/etc/hosts:ro
-networks:
-  dp-sit-network:
-    name: dp-sit-network
-    external: true

requirements.txt

@@ -1,12 +1,12 @@
 annotated-types==0.7.0
-anyio==4.9.0
+anyio==4.8.0
 blinker==1.9.0
 certifi==2025.1.31
 charset-normalizer==3.4.1
 click==8.1.8
-filelock==3.18.0
+filelock==3.17.0
 Flask==3.1.0
-fsspec==2025.3.2
+fsspec==2025.3.0
 grpcio==1.71.0
 grpcio-tools==1.71.0
 h11==0.14.0
@@ -14,7 +14,7 @@ h2==4.2.0
 hpack==4.1.0
 httpcore==1.0.7
 httpx==0.28.1
-huggingface-hub==0.30.1
+huggingface-hub==0.29.2
 hyperframe==6.1.0
 idna==3.10
 itsdangerous==2.2.0
@@ -24,7 +24,7 @@ MarkupSafe==3.0.2
 mpmath==1.3.0
 networkx==3.4.2
 nltk==3.9.1
-numpy==2.2.4
+numpy==2.2.3
 nvidia-cublas-cu12==12.4.5.8
 nvidia-cuda-cupti-cu12==12.4.127
 nvidia-cuda-nvrtc-cu12==12.4.127
@@ -40,26 +40,23 @@ nvidia-nvjitlink-cu12==12.4.127
 nvidia-nvtx-cu12==12.4.127
 packaging==24.2
 portalocker==2.10.1
-protobuf==5.29.4
+protobuf==5.29.3
-pydantic==2.11.2
+pydantic==2.10.6
-pydantic_core==2.33.1
+pydantic_core==2.27.2
-python-dotenv==1.1.0
+python-dotenv==1.0.1
 PyYAML==6.0.2
 qdrant-client==1.13.3
 regex==2024.11.6
 requests==2.32.3
 safetensors==0.5.3
-setuptools==78.1.0
+setuptools==76.0.0
 sniffio==1.3.1
 sympy==1.13.1
-tokenizers==0.21.1
+tokenizers==0.21.0
 torch==2.6.0
 tqdm==4.67.1
 transformers==4.49.0
 triton==3.2.0
-typing-inspection==0.4.0
+typing_extensions==4.12.2
-typing_extensions==4.13.1
 urllib3==2.3.0
-uuid_utils==0.10.0
 Werkzeug==3.1.3
-gunicorn==23.0.0

src/.env

@@ -0,0 +1,9 @@
+QDRANT_HOST = "192.168.99.122"
+QDRANT_PORT = 6333 
+QDRANT_HTTPS='true'
+QDRANT_SSL_VERIFY= 'false'
+QDRANT_API_KEY=NS00TXlKUUIweUhWaGFuUUpUVTk6bWNVWXI1VXRSN2VWcFRtaEZ6NmdCUQ==
+QDRANT_CLUSTER=''
+QDRANT_COLLECTION_NAME='titan-2502110203'
+TOKENIZER_FOLDER=/home/charlenewsl/model-files/tokenizer
+MODEL_FOLDER=/home/charlenewsl/model-files

src/app.py

@@ -1,154 +1,28 @@
-from flask import Flask, jsonify, request, Response
+from flask import Flask, jsonify, request
 # from dotenv import load_dotenv
 from services.embedding_service import Embedding
 import logging
 import json
-import uuid_utils as uuid
 from qdrant_client.http import models
-from qdrant_client.models import PointStruct
-from werkzeug.exceptions import BadRequest
-from qdrant_services.qdrant import Qdrant
-from config import LOGGING_LEVEL, LOGGING_FORMAT
-
-
-logging.basicConfig(
-    level=LOGGING_LEVEL,
-    format=LOGGING_FORMAT,
-)
 
 logger = logging.getLogger(__name__)
 
 # load_dotenv()
 app = Flask(__name__)
 
-
+# @app.route("/")
+# def hello():
+#     return jsonify({"message":"Hello, World!"})
 
 @app.get("/")
 def hello():
     return jsonify({"message":"Hello, World!"})
 
-@app.get("/healthz")
-def health_healthz():
-    return jsonify("OK"), 200
-
-
-@app.post("/add")
-def add():
-    if request.content_type == 'application/json':
-        try:
-            data = request.get_json(force=True)
-            if data is None:
-                raise BadRequest
-        except BadRequest:
-            return jsonify({"error": "Invalid JSON"}), 400
-    else:
-        
-        data = request.form.to_dict()
-    
-    clause_text = data.get("clause", '')
-    user_payload = data.get("payload", None)
-    user_meta = data.get("meta", None)
-
-    if clause_text=='':
-        return jsonify({"error":"clause is required"}), 400
-
-    payload = {"clause": clause_text}
-    if user_payload is not None:
-        payload["user_payload"] = user_payload
-
-    if user_meta is not None:
-        # decode for form-data only
-        payload["user_meta"] = user_meta if isinstance(user_meta, dict) else json.loads(user_meta)
-
-    clause_vectors = Embedding.call(clause_text)
-    
-    result = Qdrant.get_client().upsert(
-        points=[
-            PointStruct(
-                id=str(uuid.uuid7()),
-                vector=vector.tolist(),
-                payload=payload
-            )
-            for vector in clause_vectors
-        ]
-    )
-
-
-    logger.info("add clause [%s], result: %s" % (clause_text, result))
-    return Response(status=202)
-
-
-
-
-@app.post("/delete")
-def delete():
-    
-    if request.content_type == 'application/json':
-        try:
-            data = request.get_json(force=True)
-            if data is None:
-                raise BadRequest
-        except BadRequest:
-            return jsonify({"error": "Invalid JSON"}), 400
-    else:
-        
-        data = request.form.to_dict()    
-    
-    id = data.get("id", None)
-    filter_ = data.get("filter", None)
-    wait = data.get("wait", True)
-
-    if id is None and filter_ is None:
-        return jsonify({"error": "id or filter is required"}), 400
-    if id is not None and filter_ is not None:
-        return jsonify({"error": "id and filter cannot be used together"}), 400
-
-    try:
-        if id:
-            result = Qdrant.get_client().delete(
-        points_selector=models.PointIdsList(
-                points=[id],
-            ),
-            wait=wait,
-        )
-            
-        if filter_:
-            filter_ = json.loads(filter_)
-            filter_ = filter_[0]
-            if filter_['type'] != 'match':
-                return jsonify({"error": "unsupported filter type"}), 400
-            
-            f = models.Filter(
-            must=[
-                models.FieldCondition(
-                    key=filter_['key'], match=models.MatchValue(value=filter_['value'])
-                ),
-                ],
-            )
-
-            filter_ = f
-            result = Qdrant.get_client().delete(
-            points_selector=models.FilterSelector(
-                filter=filter_,
-            ),
-            wait=wait,
-        )
-            
-        if wait:
-            return jsonify({"message": "Success"}),200
-        else:
-            return jsonify({"message": "Accepted"}), 202
-
-    except Exception as e:
-        logger.error(f"Qdrant deletion failed: {str(e)}")
-        return jsonify({"error": "Failed to delete from Qdrant"}), 500
-        
-
-
 
 @app.get("/recommend")
 def recommend():
     try:
+        
         from services.recommend_service import Recommend
         clause = request.args.get("clause", '')
         # filter shall be a json array. each element is a dict with following keys:
@@ -190,6 +64,20 @@ def recommend():
         return jsonify({"error":"Internal Server Error"}), 500
 
 
+@app.post("/get_embedding")
+def get_embedding():
+    # Get the query text from request JSON
+    data = request.get_json()
+    query = data.get("query", "")
+
+    if not query:
+        return jsonify({"error": "Query text is required"}), 400
+
+    # Call the embedding function
+    vector = Embedding.call(query, is_query=True)
+    
+    return jsonify({"query": query, "number of embedding": len(vector), "one embedding":len(vector[0])})
+
 
 
 if __name__=="__main__":

src/config.py

@@ -1,6 +0,0 @@
-import logging
-
-from util.config import env
-
-LOGGING_LEVEL = env('LOGGING_LEVEL', default=logging.INFO)
-LOGGING_FORMAT = env('LOGGING_FORMAT', default='%(asctime)s %(name)s %(levelname)s %(message)s')

src/qdrant_services/config.py

@@ -1,9 +1,18 @@
+# replace ./settings.py
 from util.config import env
 
-QDRANT_HOST = env('QDRANT_HOST')
+QDRANT_HOST = env('QDRANT_HOST', default="192.168.99.122")
 QDRANT_PORT = env('QDRANT_PORT', default=6333, cast=int)
 QDRANT_HTTPS = env('QDRANT_HTTPS', default=False, cast=bool)
 QDRANT_SSL_VERIFY = env('QDRANT_SSL_VERIFY', default=False, cast=bool)
 QDRANT_API_KEY = env('QDRANT_API_KEY', default=None)
 QDRANT_CLUSTER = env('QDRANT_CLUSTER', default='')
-QDRANT_COLLECTION_NAME=env('QDRANT_COLLECTION_NAME')
+QDRANT_COLLECTION_NAME=env('QDRANT_COLLECTION_NAME',default='titan-2502110203')
+
+
+
+
+
+
+# TOKENIZER_FOLDER=/home/charlenewsl/model-files/tokenizer
+# MODEL_FOLDER=/home/charlenewsl/model-files

src/util/config.py

@@ -3,10 +3,8 @@ import os
 
 
 # load_dotenv("/home/charlenewsl/recommender-service-flask/src/.env")
-
+# print("QDRANT_PORT: DD : ",os.getenv("QDRANT_PORT"))
-# config = dotenv_values("/home/charlenewsl/recommender-service-flask/src/.env")
+config = dotenv_values("/home/charlenewsl/recommender-service-flask/src/.env")
-
-config = dotenv_values(".env")
 
 
 def env(name, default=None, cast=None):

startup.sh

@@ -1,9 +0,0 @@
-#!/bin/bash
-
-if [ -f /etc/ssl/cert.pem ] && [ -f /etc/ssl/key.pem ]; then
-    echo "SSL certificates found, starting Gunicorn with HTTPS..."
-    gunicorn -w 4 --bind 0.0.0.0:${APP_PORT} 'app:app' --timeout 2400 --certfile=/etc/ssl/cert.pem --keyfile=/etc/ssl/key.pem
-else
-    echo "SSL certificates not found, starting Gunicorn with HTTP..."
-    gunicorn -w 4 --bind 0.0.0.0:${APP_PORT} 'app:app' --timeout 2400
-fi