fix: replace retired Qwen2.5 model in tutorial 30 - #486
Merged
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
Tutorial 30 calls `Qwen/Qwen2.5-7B-Instruct` through `provider="together"`,
which routes to Together's `Qwen2.5-7B-Instruct-Turbo`. Together retired
that model from serverless, so the HF provider mapping now reports
`status=error` and the model is gone from the router's model list.
The nightly run had two stacked causes, the first masking the second: it
ended on `402 Payment Required` ("You have depleted your monthly included
credits"), so the request never reached Together's model resolution and
the retired model surfaced only as a health-check warning. The 402 has
since cleared; the retired model had not.
Swap in `Qwen/Qwen3.5-9B` and disable thinking. Qwen3.5 is a hybrid
thinking model, and left on it spends ~8x the completion tokens (1292 vs
165 on this tutorial's prompt) and prepends a 4 KB reasoning trace, which
would make the tutorial's sample response unrepresentative.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
julian-risch
force-pushed
the
fix/tutorial30-replace-retired-qwen25-model
branch
from
August 28, 2026 08:14
e7d2a4d to
eb47c47
Compare
julian-risch
marked this pull request as ready for review
August 28, 2026 08:24
bilgeyucel
approved these changes
Aug 28, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Related Issues
None. Found while investigating the nightly failure in run 33136680751; companion to deepset-ai/haystack-core-integrations#3866
Qwen/Qwen2.5-7B-Instruct+provider="together"routes to Together'sQwen2.5-7B-Instruct-Turbo, which Together retired from serverless. HF's API now reports that provider mapping asstatus=error, and the model is absent fromrouter.huggingface.co/v1/modelsProposed Changes:
Qwen/Qwen3.5-9B(same provider) — recent, small, cheap, live ontogether, and now consistent with the model used in thehuggingface_apiintegration tests.generation_kwargs={"extra_body": {"chat_template_kwargs": {"enable_thinking": False}}}. Qwen3.5 is a hybrid-thinking model; left on, it spends 8× the completion tokens on this tutorial's prompt (1292 vs 165)How did you test it?
Both tutorial 30 jobs pass on this PR
Notes for the reviewer
Checklist
fix:,feat:,build:,chore:,ci:,docs:,style:,refactor:,perf:,test:.