Fixed segfault when prompt was too long #3661
Closed
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
This was referenced Oct 18, 2023
|
To apply this patch I had to rename After applying this patch, I can confirm that the original crash does not happen and there's no performance regression anymore. Thank you! |
ggerganov
added a commit
that referenced
this pull request
Oct 20, 2023
ggerganov
added a commit
that referenced
this pull request
Oct 20, 2023
ggerganov
added a commit
that referenced
this pull request
Oct 20, 2023
|
Could you do the same tests for #3696 - I think I've applied the same patch there and want to just merge that PR and close this one if it works as expected. |
|
For some reason, the latest commit does not give me the crash anymore, even without any patches. I then tried #3696 just in case - no crashes and no performance regressions. |
ggerganov
added a commit
that referenced
this pull request
Oct 20, 2023
* sampling : refactor init to use llama_sampling_params * llama : combine repetition, frequency and presence penalties in 1 call * examples : remove embd-input and gptneox-wip * sampling : rename penalty params + reduce size of "prev" vector * sampling : add llama_sampling_print helper * sampling : hide prev behind API and apply #3661 ggml-ci
|
Thanks, I've merged the other PR |
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Outdated suggestions cannot be applied.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.
New fix for #3550 after it turned out the previous attempt #3639 caused a performance regression with caching the prompt.
Essentially it is an off-by-one not taking into account that
nextTokenneeds at least one free space to start generating embeddings.