Skip to content

Latest commit

 

History

History
206 lines (156 loc) · 4.44 KB

INSTALL_MANUAL.md

File metadata and controls

206 lines (156 loc) · 4.44 KB
Raw

INSTALL FROM SOURCE CODE STEP BY STEP

Prerequisites

Concept

Firstly, look at the source code sturcture:

$ ls
api  cook web

AI Shifu is composed by backend in api and frontend in web. and the cook is the editor for the course. So, we can

Tools

  • Python 3.11

  • MySQL

  • Redis

  • Lark for script editing

  • OSS for image storage

  • OpenAI API Key or LLM Provider API Key

Operating Steps

Step 1: Clone the repository

git clone https://github.com/ai-shifu/ai-shifu.git

Step 3: Configure environment variables

More Info

cp docker/.env.example .env  # guess this is under the root folder

Step 4: Configure .env

... skip code ...
########
# LLMs #
########

# IMPORTANT: At least one LLM should be enabled

# OpenAI
OPENAI_BASE_URL="https://api.openai.com/v1"
OPENAI_API_KEY="sk-proj-..."

... skip code ...

# Default LLM model. Supported models:
# OpenAI's models:
#   gpt-4o-latest, gpt-4o-mini, gpt-4, gpt-3.5-turbo, chatgpt-4o-latest, and their dated releases
... skip code ...

DEFAULT_LLM_MODEL="gpt-4o"

# Default LLM temperature
DEFAULT_LLM_TEMPERATURE=0.3

... skip code ...

# MySQL settings. If you don't know what they are, don't modify them.
SQLALCHEMY_DATABASE_URI="mysql://root:ai-shifu@ai-shifu-mysql:3306/ai-shifu"


# Redis settings. If you don't know what they are, don't modify them.
REDIS_HOST="ai-shifu-redis"
REDIS_PORT=6379
REDIS_DB=0
REDIS_PASSWORD=""
REDIS_USER=""



# (Optional) Alibaba Cloud settings for sending SMS and uploading files
ALIBABA_CLOUD_SMS_ACCESS_KEY_ID=""
ALIBABA_CLOUD_SMS_ACCESS_KEY_SECRET=""
ALIBABA_CLOUD_SMS_SIGN_NAME=""
ALIBABA_CLOUD_SMS_TEMPLATE_CODE=""

# Universal verification code
UNIVERSAL_VERIFICATION_CODE="1024"

# (Optional) Alibaba Cloud OSS settings for uploading files
ALIBABA_CLOUD_OSS_ACCESS_KEY_ID=""
ALIBABA_CLOUD_OSS_ACCESS_KEY_SECRET=""
ALIBABA_CLOUD_OSS_ENDPOINT="oss-cn-beijing.aliyuncs.com"
ALIBABA_CLOUD_OSS_BUCKET=""
ALIBABA_CLOUD_OSS_BASE_URL=""

# (Optional) Langfuse settings for tracking LLM
LANGFUSE_PUBLIC_KEY=""
LANGFUSE_SECRET_KEY=""
LANGFUSE_HOST=""

# (Optional) Netease YIDUN settings for content detection
NETEASE_YIDUN_SECRET_ID=""
NETEASE_YIDUN_SECRET_KEY=""
NETEASE_YIDUN_BUSINESS_ID=""

... skip code ...

# Path of log file
LOGGING_PATH="/var/log/ai-shifu.log"  # make sure you have the permission to write the file


############
# Frontend #
############

# Service
REACT_APP_BASEURL="http://localhost:5800" # example as for local development
PORT=5000

# Eruda console
REACT_APP_ERUDA="true"


#################
# Editor (Cook) #
#################

# Lark (Feishu) for script editing
LARK_APP_ID=""
LARK_APP_SECRET=""

# Database settings. If you don't know what they are, don't modify them.
COOK_DB_USERNAME="root"
COOK_DB_PASSWORD="ai-shifu"
COOK_DB_DATABASE="ai-shifu-cook"
COOK_DB_HOST="ai-shifu-mysql"

# For uploading images
COOK_IMG_LOCAL_DIR="/data/img/"
COOK_IMG_OSS_ANAME="" # fill it the same as ALIBABA_CLOUD_OSS_BUCKET
COOK_IMG_OSS_ENDPOINT=""
COOK_IMG_OSS_BUCKET=""

# API environment
COOK_USE_API_ENV="prod"
API_URL_TEST="http://ai-shifu-api:5800/api"
API_URL_PROD="http://localhost:5800/api" # example as for local development

# Logs
COOK_LOG_LEVEL="DEBUG"
COOK_LOG_OUT_LEVEL="DEBUG"
COOK_LOG_DIR="/var/log/" # make sure you have the permission to write the file
COOK_LOG_OUT_PATH="/var/log/ai-shifu-cook.log"
COOK_LOG_ERR_LEVEL="ERROR"
COOK_LOG_ERR_PATH="/var/log/ai-shifu-cook.err.log"

Step 5: Run the application

  • Make sure .env is under every folder

Step 5.1: Run the api

cd api
cp ../.env .env

pip install -r requirements.txt
flask db upgrade
gunicorn -w 4 -b 0.0.0.0:5800 'app:app' --timeout 300 --log-level debug --access-logfile /var/log/app.log --capture-output

Step 5.2: Run the web

cd web
cp ../.env .env

npm install  # or use pnpm install
npm run start:dev # or use pnpm run build

Step 5.3: Run the cook

  1. Copy .env
cd cook
cp ../.env .env
  1. Install requirements
pip install -r requirements.txt
  1. Edit auth_config.yml
cp auth_config.example.yml auth_config.yml
  1. Init database
# pay attention to the database name should be the same as the one in .env
mysql -u user -h xxxx -p database_name < ../../docker/init.sql
  1. Run the cook
streamlit  run Home.py

Step 6: Access the application

Go to the browser and have fun!