test: switch integration tests model to zai-org/GLM-5.2 (ENG-92345) - #542
Open
blainekasten wants to merge 5 commits into
Open
test: switch integration tests model to zai-org/GLM-5.2 (ENG-92345)#542blainekasten wants to merge 5 commits into
blainekasten wants to merge 5 commits into
Conversation
Use openai/gpt-oss-20b for live completion integration tests instead of Qwen/Qwen2.5-7B-Instruct-Turbo. Stream tests now parametrize from the shared model list so they cannot drift. ENG-92345 Co-authored-by: Blaine Kasten <blainekasten@gmail.com>
Broly Security ScanNote ✅ Clean scan Note Re-scan this PR anytime with
|
openai/gpt-oss-20b is OpenAI-compatible: object is text_completion, echo/logprobs are unsupported, n/penalty limits are not enforced, seed is not echoed, oversize max_tokens is 400, and streams can end with an empty usage chunk. Align assertions and skip Together-native checks. Co-authored-by: Blaine Kasten <blainekasten@gmail.com>
cursor
Bot
changed the title
blainekasten
marked this pull request as ready for review
August 24, 2026 16:01
Use GLM-5.2 instead of openai/gpt-oss-20b. Re-enable Together-native param-limit tests that gpt-oss did not enforce. Co-authored-by: Blaine Kasten <blainekasten@gmail.com>
cursor
Bot
changed the title
Skip n=129 and echo/logprobs (slow or unsupported), cap generated tokens, drop the long llama prompt, and raise the integration job timeout to 20 minutes so serverless GLM-5.2 can finish. Co-authored-by: Blaine Kasten <blainekasten@gmail.com>
high_max_tokens was generating 200k+ tokens instead of 400ing, which timed out the suite. GLM streams also omit usage, so only assert it when present. Co-authored-by: Blaine Kasten <blainekasten@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Switches live completion integration tests from
Qwen/Qwen2.5-7B-Instruct-Turbotozai-org/GLM-5.2.zai-org/GLM-5.2; stream tests parametrize from itn>128, and oversizedmax_tokens(unsupported or too expensive on GLM-5.2)Fixes ENG-92345
Linear Issue: ENG-92345