Fixed #4229

[konraddroeske]

Resolves: #4229

Background

The defaults need to be adapted for people without access to GPT4 - or the corresponding functions won't work. The default model for people without GPT4 access should be cfg.fast_llm_model and not cfg.smart_llm_model.

Changes

Added check_model fn in configurator that sends a request to the OpenAI API using create_chat_completion with the model provided. If an InvalidRequestError is raised, it returns a hard-coded value of "gpt-3.5-turbo".

Documentation

Comments added to the function and test.

Test Plan

Checks if fn returns original model argument if no error.
Checks if fn returns "gpt-3.5-turbo" when InvalidRequestError is raised.

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.
• I have run black . and isort . against my code to ensure it passes our linter.

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
docs	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	May 20, 2023 11:32pm

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.53 🎉

Comparison is base (c30f5b7) 63.26% compared to head (70f2b97) 63.79%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4278      +/-   ##
==========================================
+ Coverage   63.26%   63.79%   +0.53%     
==========================================
  Files          74       74              
  Lines        3427     3447      +20     
  Branches      504      507       +3     
==========================================
+ Hits         2168     2199      +31     
+ Misses       1103     1079      -24     
- Partials      156      169      +13     

Impacted Files	Coverage Δ
autogpt/main.py	0.00% <ø> (ø)
autogpt/configurator.py	20.27% <100.00%> (+20.27%)	⬆️
autogpt/llm/api_manager.py	94.44% <100.00%> (+1.11%)	⬆️
autogpt/llm/llm_utils.py	68.38% <100.00%> (+1.71%)	⬆️

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 55 minutes (more than 100, code: "api-deployments-free-per-day").

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 47 minutes (more than 100, code: "api-deployments-free-per-day").

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 10 minutes (more than 100, code: "api-deployments-free-per-day").

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 9 minutes (more than 100, code: "api-deployments-free-per-day").

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 4 minutes (more than 100, code: "api-deployments-free-per-day").

[vercel]

Deployment failed with the following error:

Resource is limited - try again in 3 minutes (more than 100, code: "api-deployments-free-per-day").