GPT-Port

GPT-Port: a Flask-based Tool for GPT Port Building with External Jump Server.

Step 1: Install Dependencies

pip install -r requirements.txt

Step 2: Build GPT-Port on External Jump Server

2.1 Set OpenAI API Key

Change "your openai api key" in openai.api_key="your openai api key" to your own openai api key in the 2nd line of "GPT_USE.py".

2.2 Build GPT-Port

You can build GPT-Port in External Jump Server which can access to openai service.

nohup python main.py &

After successful start, remember the Process Id [PID] for end the service by kill [PID] later.

2.3 Check Public IP Address

You need to check what your public IP address [PublicIP] is in case you want to use GPT-Port service on other servers later.

curl ipinfo.io/ip

Step 3: Use GPT-Port on Internel Servers

You can use your GPT service on your internal servers that have access to your Flask-based GPT-Port but not to the OpenAI service with "use_route.py".

3.1 Set Public IP

Change "yourip" in url = 'http://yourip:80/' to your public IP address ([PublicIP] in Step 2.3) in the 3rd line of "use_route.py".

3.2 Use GPT-Port Service

You can use the function "GPT" in "use_route.py" to get the equivalent OpenAI api service by simply entering the "input" and "model" name.

For example:

ChatGPT Service: output = GPT(input="What is ChatGPT", model='gpt-3.5-turbo')

LHRLAB/GPT-Port