-
Asynchronous FastAPI project with endpoints that can be used to upload a roadmap file generated from a gpt model to the computer
-
OpenAI library for working with gpt is used
openai==1.14.2
- I use the kafka-python libraries to work with Kafka
kafka-python==2.0.2
- FastAPI service is started via Uvicorn
uvicorn==0.29.0
-
The generation process from OpenAI and the file upload process takes place on the Kafka side, which is connected via docker-compose
-
The configuration is located in the .env_dev file
- To change it to another file, modify the following line in src/app.py
env.read_env('.env') -> .env set to the new env file
- The startup is done with the src/manage.py file and in this file you can add other startups before or after the services are started
import subprocess
if __name__ == "__main__":
"""
Here you can perform actions before starting the services
"""
server_process = subprocess.Popen(["python", "server_run.py"])
consumer_process = subprocess.Popen(["python", "kafka_run.py"])
server_process.wait()
consumer_process.wait()
-
I hope you already have python, pip and docker installed
-
Copy this repository to your system
https://github.com/VyacheslavShrot/ai_it_mentor.git
- Install requirements.txt
pip install -r requirements.txt
- To configure the Kafka queue retention period, specify the desired number of milliseconds
on a line in the "docker-compose.yml" file
- Where 86400000 is 24 hours of queue storage
KAFKA_LOG_RETENTION_MS: "86400000"
- Create an .env file at the src level of the directory (ai_it_mentor/.env)
- Define OPENAI_API_KEY variable to work with gpt
OPENAI_API_KEY=sk-lb...
- Start Docker-Compose, where services are defined to work with Kafka
docker-compose up -d
- Go to the work directory
cd src
- Start FastAPI and Kafka services
python3 manage.py
- https://documenter.getpostman.com/view/26500283/2sA35D63gG - Link on Postman Documentation
- This documentation describes the possible parameters and their values
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent