Fix: last completion chunk with finish_reason not sent in edge case scenario #515
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Outdated suggestions cannot be applied.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.
An edge case was discovered where the last completion chunk containing
finish_reasonmay not be sent with the final batch if it has empty choicetext. This is because chunks without choice text are not sent to the detection pipeline.This PR adds logic to track chunk indices for each batch and send the last completion chunk if not already sent with the final batch.
Closes #516
Example
Request
{ "model": "ibm-granite/granite-3.0-8b-instruct", "prompt": "Please repeat the words in [], do not trim space.\n[I love this world.]", "stream": true, "seed": 1337, "top_p": 1, "include_stop_str_in_output": true, "repetition_penalty": 1, "stream_options": { "include_usage": true, "continuous_usage_stats": true }, "detectors": { "output": { "hap": {} } } }Response (before fix)
Response (after fix)