Feature/llm data structs

[collijk]

Background

This PR proposes some data structures for collecting model information and model responses into standard containers. We can then leverage the containers to hide model api specific details into modules that deal specifically with particular providers (code doing this in the next PR).

Changes

Add core abstractions for model information and model responses.

Documentation

Docstrings

Test Plan

N/A. These are data structures with no (well, very little) logic and no current usages.

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

[vercel]

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

1 Ignored Deployment

Name	Status	Preview	Comments	Updated (UTC)
docs	⬜️ Ignored (Inspect)			Apr 28, 2023 10:00pm

[Pwuts]

Since info about the models and their max context length is now in llm/providers/openai.py, settings for that can be removed or given better defaults:
https://github.com/Significant-Gravitas/Auto-GPT/blob/3b74d2150e26ec9387e40cf65a75d8a995d8c9f3/autogpt/config/config.py#L36-L38

[collijk]

Since info about the models and their max context length is now in llm/providers/openai.py, settings for that can be removed or given better defaults:

https://github.com/Significant-Gravitas/Auto-GPT/blob/3b74d2150e26ec9387e40cf65a75d8a995d8c9f3/autogpt/config/config.py#L36-L38

Yes, that will be done in a separate PR as it will touch a bunch of files. This PR is to make sure people see the data structures and agree with this representation (got positive feedback from @BillSchumacher , @merwanehamadi , and @ntindle on a PR in my own fork yesterday as I was waiting on a merge for this).

[github-actions]

This pull request has conflicts with the base branch, please resolve those so we can evaluate the pull request.

[ntindle]

Since info about the models and their max context length is now in llm/providers/openai.py, settings for that can be removed or given better defaults:

https://github.com/Significant-Gravitas/Auto-GPT/blob/3b74d2150e26ec9387e40cf65a75d8a995d8c9f3/autogpt/config/config.py#L36-L38

Agreed but separate PR imo

Pwuts signed off on my comment explaining we would do other file cleanup in an upcoming PR

[github-actions]

Conflicts have been resolved! 🎉 A maintainer will review the pull request shortly.

[codecov]

Codecov Report

Patch coverage: 85.00% and project coverage change: +0.23 🎉

Comparison is base (c7d7564) 56.95% compared to head (ce1daf0) 57.19%.

❗ Current head ce1daf0 differs from pull request most recent head db2e6c4. Consider uploading reports for the commit db2e6c4 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3486      +/-   ##
==========================================
+ Coverage   56.95%   57.19%   +0.23%     
==========================================
  Files          67       68       +1     
  Lines        3048     3079      +31     
  Branches      509      516       +7     
==========================================
+ Hits         1736     1761      +25     
- Misses       1174     1180       +6     
  Partials      138      138              

Impacted Files	Coverage Δ
autogpt/llm/providers/openai.py	0.00% <0.00%> (ø)
autogpt/llm/base.py	93.54% <93.54%> (ø)
autogpt/agent/agent_manager.py	47.82% <100.00%> (-0.75%)	⬇️
autogpt/llm/__init__.py	100.00% <100.00%> (ø)
autogpt/llm/chat.py	65.16% <100.00%> (ø)
autogpt/llm/llm_utils.py	58.71% <100.00%> (ø)
autogpt/llm/token_counter.py	94.11% <100.00%> (ø)

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