Add test for sequence state after cancellation #7167
Merged
Conversation
This file contains 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
Tabrizian
reviewed
Apr 29, 2024
| sequence_id=1, | ||
| sequence_start=True, | ||
| ) | ||
| seq_start = False |
There was a problem hiding this comment.
question: why do we set this value to False when it is overwritten on line 84 and not used before that?
There was a problem hiding this comment.
This variable should be used for the sequence_start, you already caught it is always set to True on another comment. The issue is fixed.
| model_name, | ||
| self._get_inputs(delay_itrs=5000000), | ||
| sequence_id=1, | ||
| sequence_start=True, |
There was a problem hiding this comment.
Suggested change
| sequence_start=True, | |
| sequence_start=seq_start, |
Comment on lines
81
to
96
| with grpcclient.InferenceServerClient("localhost:8001") as client: | ||
| client.start_stream(callback) | ||
| seq_start = True | ||
| num_reqs = 4 | ||
| seq_ids = [2, 3] | ||
| for req_id in range(num_reqs): | ||
| for seq_id in seq_ids: | ||
| client.async_stream_infer( | ||
| model_name, | ||
| self._get_inputs(delay_itrs=0), | ||
| sequence_id=seq_id, | ||
| sequence_start=seq_start, | ||
| ) | ||
| time.sleep(0.1) | ||
| seq_start = False | ||
| client.stop_stream(cancel_requests=False) |
There was a problem hiding this comment.
Could we perhaps refactor this into a common function looks like it is a repetition of the block above it?
There was a problem hiding this comment.
Good idea! Updated: Regroup infer calls
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
|
Looks good to me. Please rebase the branch. |
tanmayv25
approved these changes
May 7, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Should be coupled with this 3rd party PR which adds the fix: triton-inference-server/core#341
Reproduction steps:
Since sequence state is identified by the sequence slot, so the two new sequences executing concurrently will share the same sequence state and leads to corruption of the state.
Ref: #7117 (comment)