Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Only gpt-3.5 being used, without using --gpt3only modifier #942

Closed
1 task done
521047 opened this issue Apr 12, 2023 · 6 comments · Fixed by #3144
Closed
1 task done

Only gpt-3.5 being used, without using --gpt3only modifier #942

521047 opened this issue Apr 12, 2023 · 6 comments · Fixed by #3144
Assignees
@lc0rp
Labels
bug Something isn't working needs discussion To be discussed among maintainers

Comments

@521047
Copy link

521047 commented Apr 12, 2023

Duplicates

  • I have searched the existing issues

Steps to reproduce 🕹

Run python scripts/main.py with OpenAI api key that supports gp4 usage in env

Current behavior 😯

Judging by my USAGE page on Open AI, AutoGPT is only making use of gpt-3.5-turbo-0301 and text-embedding-ada-002-v2

Expected behavior 🤔

AutoGPT should use gpt-4 model

Your prompt 📝

# Paste your prompt here
@richbeales
Copy link
Contributor

I have also seen this, I'm not sure on a cause, yet.

@richbeales richbeales added the bug Something isn't working label Apr 12, 2023
@ericlewis
Copy link

from what I can tell. the script defaults to 3.5. you must use the gpt4only flag.

@rbourdon
Copy link

rbourdon commented Apr 12, 2023 •
edited
Loading

It's hardcoded to use fast_llm_model for everything except the write_tests, evaluate_code and improve_code commands. (see https://github.com/Torantulino/Auto-GPT/blob/master/scripts/chat.py#L66 for example)

So you could change that to gpt-4 if you want to wrack up a bill 😉

@Pwuts Pwuts added the needs discussion To be discussed among maintainers label Apr 17, 2023
@Pwuts Pwuts moved this to 📋 Backlog in AutoGPT development kanban Apr 17, 2023
@edcohen08 edcohen08 mentioned this issue Apr 20, 2023
Closed
5 tasks
@GerardCB
Copy link

Maybe GPT 4 (or smart_llm_model) should be used by default when choosing the command to use next, which should not use many tokens but is the core of AutoGPT. Oftentimes GPT 3.5 fails to select the appropriate command given the agent's thoughts.

@edcohen08
Copy link
Contributor

definitely want more input on this #2621 and if it's a good direction need to add tests

lc0rp added a commit to lc0rp/Auto-GPT-Turbo that referenced this issue Apr 24, 2023
@lc0rp
Fix Significant-Gravitas#942: Respect --gpt4only, --gpt3only cli args
8c30d5b 
- Add `cfg.fast_llm_model` to `chat_with_ai` method parameters.
- Remove hardcoded model in `chat_with_ai` method and `Agent` class
- Add `GPT_4_MODEL` and `GPT_3_MODEL` constants to `configurator.py`.
- Change `--gpt3only` to use `GPT_3_MODEL` not `CFG.fast_llm_model`.
- Change `--gpt4only` to use `GPT_4_MODEL` not `CFG.smart_llm_model`.
- Add tests for `--gpt4only` and `--gpt3only` CLI args
@lc0rp lc0rp mentioned this issue Apr 24, 2023
Merged
5 tasks
@lc0rp
Copy link
Contributor

lc0rp commented Apr 24, 2023

See #3144 proposed fix.

@bfalans bfalans moved this from 📋 Backlog to 🆕 Un-Reviewed in AutoGPT development kanban May 18, 2023
@lc0rp lc0rp self-assigned this May 23, 2023
@lc0rp lc0rp moved this to 🆕 Un-Reviewed in AutoGPT development kanban May 23, 2023
k-boikov added a commit that referenced this issue May 29, 2023
@lc0rp @ntindle
@Pwuts @k-boikov
Fix #942: Respect --gpt4only, --gpt3only cli args (#3144)
1446ffd 
Co-authored-by: Luke K <[email protected]>
Co-authored-by: Nicholas Tindle <[email protected]>
Co-authored-by: Reinier van der Leer <[email protected]>
Co-authored-by: k-boikov <[email protected]>
@github-project-automation github-project-automation bot moved this from 🆕 Un-Reviewed to ✅ Done in AutoGPT development kanban May 29, 2023
SquareandCompass pushed a commit to SquareandCompass/Auto-GPT that referenced this issue Oct 21, 2023
@sonichi @thinkall @liususan091219
ChatGPT support (Significant-Gravitas#942)
169012f 
* improve max_valid_n and doc

* Update README.md

Co-authored-by: Li Jiang <[email protected]>

* add support for chatgpt

* notebook

* newline at end of file

* chatgpt notebook

* ChatGPT in Azure

* doc

* math

* warning, timeout, log file name

* handle import error

* doc update; default value

* paper

* doc

* docstr

* eval_func

* prompt and messages

* remove confusing words

* notebook name

---------

Co-authored-by: Li Jiang <[email protected]>
Co-authored-by: Susan Xueqing Liu <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@lc0rp lc0rp

Labels
bug Something isn't working needs discussion To be discussed among maintainers
Projects
AutoGPT development kanban
Archived in project
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Fix #942: Respect --gpt4only, --gpt3only cli args lc0rp/Auto-GPT-Turbo
8 participants
@ericlewis @richbeales @lc0rp @rbourdon @Pwuts @edcohen08 @GerardCB @521047