feat: add GGML_UNARY_OP_ARGMAX Metal kernel
#1019
Open
+35
−0
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
|
Thanks, will merge these soon. Just need to sync the ggerganov/llama.cpp#10238 from |
ggerganov
requested changes
Nov 18, 2024
| [encoder setBuffer:id_dst offset:offs_dst atIndex:1]; | ||
| [encoder setBytes:&ne00 length:sizeof( int64_t) atIndex:2]; | ||
|
|
||
| [encoder dispatchThreadgroups:MTLSizeMake(1, 1, 1) threadsPerThreadgroup:MTLSizeMake(nrows, 1, 1)]; |
There was a problem hiding this comment.
Threadgroup sizes are typically maximum of 1024, so this implementation will stop working when nrows is larger than 1024. Instead, we should launch multiple threadgroups and distribute the rows across them.
Also, the implementation currently assumes src0 is contiguous, so add a GGML_ASSERT for that or extend the implementation by providing and taking account of the strides.
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Outdated suggestions cannot be applied.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 implements the Metal kernel used for the
GGML_UNARY_OP_ARGMAXoperation.It is necessary for Encodec.cpp to run on the Metal backend.