README: Explain OpenAI billing for API key (Fixes issue about "API Rate Limit Reached. Waiting 20 seconds ... Failed to get response after 5 retries")

[JuroOravec]

Background

There's multiple issues, where people hit OpenAI API rate limiting error, where the true cause of the rate limiting is that people have not set up billing (paid account) on their OpenAI profile.

The truth is, one has to:

1. Opt in to "paid account" on https://platform.openai.com/account/billing/overview and enter their billing info
2. Only now request the API key (in my case it took about 5-10 mins after opting into paid account before the API key started to work via AutoGPT.

As of 12:00 14/04/2023, I found these related issues:

Examples:

• API Rate Limit Reached with new key #75
• API Rate Limit Reached. Waiting 10 seconds #79
• Why do I get this error? #189
• Error: API Rate Limit Reached. Waiting ten seconds... #235
• Error: API Rate Limit Reached. Waiting 10 seconds... #425
• API Rate Limit issue #585
• Title: Repeated "Error: API Rate Limit Reached" messages when running the script Issue: I have tried running Auto-GPT from the master branch. Commit-hash: a0f900f2dabd114f80d5e96ecc6269a744dfc99a Pip version: pip 23.0.1 Python version: 3.11.3 The issue occurs when running the script using the following command: python scripts/main.py #1047
• Warning: The file 'auto-gpt.json' does not exist. Local memory would not be saved to a file. #1073
• auto-gpt.json #1174
• RuntimeError: Failed to get response after 5 retries #1189
• RuntimeError: Failed to get response after 5 retries #1203
• Warning: The file 'auto-gpt.json' does not exist. Local memory would not be saved to a file. Using memory of type: LocalCache #1227
• file 'auto-gpt.json' does not exist. Failed to get response after 5 retries  #1248
• Error: API Rate Limit Reached. Waiting 20 seconds... RuntimeError: Failed to get response after 5 retries #1252
• RuntimeError: Failed to get response after 5 retries #1262

Possibly also

• thinking and does not produce anything #1010
• API Rate Limit Reached #815
• Token limit: 4000 while API Rate Limit Reached. Waiting 10 seconds #1013

Changes

Make this explicit in the README file. I've moved the sentence on OpenAI API key to it's own section, so we can also include an image showing exactly what one has to do

Documentation

Test Plan

PR Quality Checklist

• My pull request is atomic and focuses on a single change.
• I have thoroughly tested my changes with multiple different prompts.
• I have considered potential risks and mitigations for my changes.
• I have documented my changes clearly and comprehensively.
• I have not snuck in any "extra" small tweaks changes

[JuroOravec]

CC @Torantulino

(Or @nponeccop, are you a maintainer too? Because I just saw that you were active around :D)

[JuroOravec]

And this is how the README section looks with the change

[nponeccop]

@JuroOravec I'm not but I'm coordinating the merges and do preliminary reviews and conflict resolution notifications. The maintainers are @p-i- and @richbeales and @Torantulino obviously.

[DHRUV-CODER]

Sent a pull request which shows the error regarding the billing problem (but only in debug mode) ,there are more errors which are not shown in detail. Would love to see more updates regarding this problem .

[jpollard-cs]

Looks good to me! Thank you for also doing the cleanup work!

[DrReefer]

I have set up a paid openai and still get the error Warning: The file 'auto-gpt.json' does not exist. Local memory would not be saved to a file.
Please set your OpenAI API key in .env or as an environment variable.
You can get your key from https://beta.openai.com/account/api-keys

[KunLiam]

这就是 README 部分在更改后的样子

OpenAI can set up paid accounts, is there a limit for the region, which regions can set up?