llama : default n_swa for phi-3 #8931
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
slaren
approved these changes
Aug 8, 2024
ggerganov
approved these changes
Aug 9, 2024
src/llama.cpp
Outdated
| // default value for Phi-3-medium-128k-instruct | ||
| hparams.n_swa = 131072; | ||
| } | ||
| ml.get_key(LLM_KV_ATTENTION_SLIDING_WINDOW, hparams.n_swa, false); |
There was a problem hiding this comment.
Might want to handle get_key returning false and hparams.n_swa remaining uninitialized to 0 by throwing an error, though probably never going to happen
mofosyne
added
the
Review Complexity : Low
ngxson
added
the
merge ready
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 15, 2024
* default n_swa for phi-3 * fix * double check swa
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 18, 2024
* default n_swa for phi-3 * fix * double check swa
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Outdated suggestions cannot be applied.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.
Related to #8627
phi3.attention.sliding_windowis marked as required, this break many existing models as some users can't re-convert the new version (ref: ngxson/wllama#106)This PR propose default value for certain models: