-
Notifications
You must be signed in to change notification settings - Fork 2.3k
fix: skip temperature for goose-gpt-5 model #5311
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.
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
Conversation
|
We should make the same change for the openai provider too: https://github.com/mnicholson110/goose/blob/88947546c8ef64f70bb2ea60c7b2774881f2db84/crates/goose/src/providers/formats/openai.rs#L625 |
Hi @jsibbison-square,
|
alexhancock
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nice!
| } | ||
| } | ||
|
|
||
| // o1 models use max_completion_tokens instead of max_tokens |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: update this comment too
| } | ||
|
|
||
| // o1 models use max_completion_tokens instead of max_tokens | ||
| // open ai reasoning models use max_completion_tokens instead of max_tokens |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
do we need this at all? in fact shouldn't we just check whether we have max_tokens in there and if not check if there is max_completion_tokens in there?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
By reading the open ai doc here, max_token is deprecated.
We may have a separate PR to use max_completion_tokens and test it.
* main: Add tech-article-explainer recipe (#5333) doc: added beta banner for old blog post (#5332) feat: add code refactor recipe (#5320) Security audit recipe (#5319) feat: add generate commit message recipe (#5326) fix: remove dependency on gsap library (#5330) feat: dynamically load ollama models (#5309) fix: skip temperature for goose-gpt-5 model (#5311)
…message # By Better-Boy (3) and others # Via GitHub * 'main' of github.com:block/goose: Create recipe dir on save (#5337) docs: Update Discord link (#5335) [recipe workflow]: Fix `Invalid revision range` error (#5334) Add tech-article-explainer recipe (#5333) doc: added beta banner for old blog post (#5332) feat: add code refactor recipe (#5320) Security audit recipe (#5319) feat: add generate commit message recipe (#5326) fix: remove dependency on gsap library (#5330) feat: dynamically load ollama models (#5309) fix: skip temperature for goose-gpt-5 model (#5311) Replace compaction notifications with system notifications (#5218) # Conflicts: # crates/goose/src/agents/agent.rs
* main: (77 commits) Fix legacy import (#5343) Unify loading goose messages and usechatstream determines chat state (#5306) Docs: goose session export and goose session import (#5267) Create recipe dir on save (#5337) docs: Update Discord link (#5335) [recipe workflow]: Fix `Invalid revision range` error (#5334) Add tech-article-explainer recipe (#5333) doc: added beta banner for old blog post (#5332) feat: add code refactor recipe (#5320) Security audit recipe (#5319) feat: add generate commit message recipe (#5326) fix: remove dependency on gsap library (#5330) feat: dynamically load ollama models (#5309) fix: skip temperature for goose-gpt-5 model (#5311) Replace compaction notifications with system notifications (#5218) Diagnostics (#5323) Fix gemini again (#5308) fix: synchronize local message state after conversation compaction (#5315) docs: replace broken links with working links (#5266) Add Web Accessibility Auditor recipe to cookbook (#5318) ...
Signed-off-by: Blair Allan <[email protected]>
Signed-off-by: Blair Allan <[email protected]>
Summary
Issue: When specifying temperature with
goose-gpt-5model (eg: run a recipe), the session exits with "bad request error"Why: gpt-5 only allows default (1) temperature
Fix: Ignore the temperature for
goose-gpt-5(same thing has been handled in the openai)Type of Change