fix(tutorials): guard against empty choices and message=None in LLM responses#5
Open
qizwiz wants to merge 1 commit into
Open
fix(tutorials): guard against empty choices and message=None in LLM responses#5qizwiz wants to merge 1 commit into
qizwiz wants to merge 1 commit into
Conversation
…esponses OpenAI's Chat Completions API can return choices=[] (e.g. on network/content errors) or choices[0].message=None (e.g. Gemini PROHIBITED_CONTENT via the OpenAI-compatible endpoint). Both paths currently crash with IndexError or AttributeError; explicit guards raise clear ValueError instead. The logprobs tutorial also guards against logprobs being absent when the API is called from a provider that doesn't support it.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
Add explicit guards before accessing
choices[0].messageandchoices[0].logprobsin two tutorial files.Why
The OpenAI Chat Completions API can return two different empty-response shapes, neither of which is currently guarded:
choices = []— on network errors, content-policy rejections, or provider-side filteringchoices[0].message = None— Gemini 2.5 Flash (via the OpenAI-compatible endpoint) returns HTTP 200 withfinish_reason: PROHIBITED_CONTENTandmessage=Nonerather than raising an exceptionBoth currently crash with
IndexErrororAttributeError. The guard raises a clearValueErrorinstead.Files changed
openai-tutorials/1-Get_response_in_JSON_format/return_json.pychoices+messagebefore first access (covers the second access on line 27 too)openai-tutorials/8-LLMs_explained/llms_explained.pychoices+logprobsbefore accessing.logprobs.contentCorpus context
Detected by
pact(llm_response_unguardedmode), a Z3-verified static analyzer for LLM crash vectors. These two patterns account for the most common unhandled LLM failures across 13.5k violations in 761 repos.