sycl-exp : Enabled more data types for oneMKL's gemm_batch API #8182
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
github-actions
bot
added
the
SYCL
|
Looks good @OuadiElfarouki. I'll hold off giving a 👍 until someone from the Intel side can take a look. @airMeng I don't think this will affect you guys but would be good to check. |
ha, @OuadiElfarouki I forgot this is to |
OuadiElfarouki
merged commit 250e790
into
ggerganov:codeplay/sycl-main
64 of 65 checks passed
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Outdated suggestions cannot be applied.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.
Additional
gemm_batchtypes have been enabled in oneMKL (oneapi-src/oneMKL#466) and this patch enables their corresponding APIs for the SYCL backend which eliminates the extra-steps needed when targetting non intel devices to cast/adapt input/output to the supported types.The enablement of
gemm_batch_impl<sycl::half, sycl::half, float, float>for instance removes the overhead of callinggemm_batch_impl<sycl::half, sycl::half, sycl::half, sycl::half>followed by ato_fp32_syclfor thedstto be copied back fromfp16tofp32, which directly affects theKQ + KQV multi-batchpath in quantized models Prompt Processing for instance.