Fixing race condition in server and partial stream handling in frontend. #2391
Conversation
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
Closed
4 tasks
|
Ah great! I had many segfaults, I will use this branch tomorrow to check if it's fixed. |
Closed
4 tasks
slaren
approved these changes
Jul 31, 2023
|
From what I understand, the |
crasm
pushed a commit
to crasm/llama.cpp
that referenced
this pull request
Aug 9, 2023
…nd. (ggml-org#2391) * Fixing race condition in server.cpp and partial stream handling in completion.js * Reverting assert edits. * Adding newline to eof
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Outdated suggestions cannot be applied.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.
This PR fixes two problems:
The race condition was caused by unique_lock scope loss when processing streaming completion. My fix was to handle unlocking of the mutex manually in that case. This bug caused segfaults when handling multiple streaming requests on the completion endpoint.
Partial stream results happen regularly on slower connections, or when the server is generating messages very quickly. The fix was to handle leftover data in the llama generator function. This bug caused completion messages to be garbled in several cases.