**OpenAI API Errors: Internal Server Error and Bad Gateway**

[Nbtguyoriginal]

---

Title:

OpenAI API Errors: Internal Server Error and Bad Gateway

Issue Description:

When attempting to run gpt-engineer with specific projects and configurations following the steps closely, I consistently encounter OpenAI API errors, specifically "Internal Server Error" and "Bad Gateway".

Environment:

• OS: Windows Not docker
• Command Prompt Location: C:\Users\path\OneDrive\Desktop\auto gen\working gen\gpt engineer\gpt-engineer

Steps to Reproduce:

1. Installed gpt-engineer using the command: python -m pip install gpt-engineer. The installation completed without any issues.
2. Ran the following commands:
   • gpt-engineer projects\gen gpt-3.5-turbo-16k-0613 -i ~~ error received
   • ' checked the help command for clarity the attempted to run with the default gpt4
   • gpt-engineer projects\gen -i ~~ error received

Both commands resulted in the same errors. if needed i can run the commands in the same terminal to provide both error logs

Expected Behavior:

The gpt-engineer tool should run without any errors and complete the intended operations. as per the instructions

Actual Behavior:

The tool encounters OpenAI API errors and eventually and errors until a user aborts the operation. The errors observed are:

• Internal server error with error type auth_subrequest_error and error code internal_error.
• Bad gateway error with error type cf_bad_gateway and error code 502.

Logs:

INFO:openai:error_code=502 error_message='Bad gateway.' error_param=None error_type=cf_bad_gateway message='OpenAI API error received' stream_error=False
...
WARNING:langchain.llms.base:Retrying langchain.chat_models.openai.ChatOpenAI.completion_with_retry.<locals>._completion_with_retry in 4.0 seconds as it raised APIError: Bad gateway. ...
...
Aborted!

Additional Notes:

• The file list for the improvement process was detected at C:\Users\path\OneDrive\Desktop\auto gen\working gen\gpt engineer\gpt-engineer\projects\gen\.gpteng\file_list.txt.
• The inserted prompt provides an overview of the OpenAI Integration & Multi-Agent Conversational Environment.

Recommendations:

• Please help me to investigate the root cause of these API errors id love to use this tool or at least try it .
• It would be helpful to have more descriptive error messages or suggestions for troubleshooting in the event of such errors. or potentially a chatbot traind on the code base to provide insight

gpt-engineer( potential bug inside).zip not sure if this is okay but some repos allow for this so here is the full directory for trouble shooting__ ty in advance for any help

I've attempted to disable cloud flare and restart only to receive the same error any help would be appreciated

[ATheorell]

Are you providing an OPENAI_API_KEY in the environment variable with the same name? auth_subrequest_error sounds like it.

If this is not the cause of the error, please provide a full stack trace

[ATheorell]

Im pretty sure that the problem is that you haven't set OPENAI_API_KEY in your environment. Closing