llamamodel: fix embedding crash for >512 tokens after #2310 #2383
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
n_ubatch defaults to 512, but as of the latest llama.cpp you cannot pass more than n_ubatch tokens to the embedding model without hitting an assertion failure. Signed-off-by: Jared Van Bortel <[email protected]>
manyoso
approved these changes
May 29, 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Outdated suggestions cannot be applied.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.
n_ubatch defaults to 512, but as of the latest llama.cpp you cannot pass more than n_ubatch tokens to BERT without hitting an assertion failure.
Tested with this python code:
Without this PR, it crashes with an assertion failure (including in release builds, since it's a GGML_ASSERT). With this PR, it succeeds.
Broken by #2310 because of ggerganov/llama.cpp#6017
Fix based on ggerganov/llama.cpp#6296
Fixes #2375