This repository has been archived by the owner on Aug 30, 2024. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 38
[Neural Speed] Support continuous batching + beam search inference in LLAMA #145
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
zhentaoyu
added
documentation
Improvements or additions to documentation
enhancement
New feature or request
labels
Feb 29, 2024
zhentaoyu
requested review from
zhenwei-intel,
airMeng,
intellinjun,
DDEle and
a32543254
and removed request for
zhenwei-intel,
airMeng and
intellinjun
March 1, 2024 03:28
a32543254
approved these changes
Mar 1, 2024
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
airMeng
reviewed
Mar 1, 2024
DDEle
reviewed
Mar 1, 2024
a32543254
reviewed
Mar 1, 2024
do we have limitation for batch size for continuous batching ? |
a32543254
reviewed
Mar 1, 2024
Nope. |
Zhenzhong1
approved these changes
Mar 1, 2024
zhenwei-intel
approved these changes
Mar 1, 2024
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Yu, Zhentao <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
Signed-off-by: Yu, Zhentao <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Yu, Zhentao <[email protected]>
zhentaoyu
force-pushed
the
yzt/llama-batching
branch
from
March 4, 2024 03:20
6a27bbb
to
ac41d13
Compare
@intellinjun please review llama moe ac41d13 |
intellinjun
approved these changes
Mar 4, 2024
Signed-off-by: Yu, Zhentao <[email protected]>
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Labels
documentation
Improvements or additions to documentation
enhancement
New feature or request
ready to merge
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.
Type of Change
as title
Description
detail description
Issues: jira-1186
This PR adds:
llama
arch supports continuous batching inference (no padding) and beam searchdeveloper_document.md
docmodel_server
example usellama
by defaultllama2-7b beam search
results alignmentPyTorch fp32
v.sNeural Speed f32
(NS_SIMD_VEC_DOT_F16=OFF
)test prompts:
PyTorch
results:click me
Neural Speed
results:click me
Q4_Bestla_g32_cint8
resultsnon-MHA fusion
:click me
MHA fusion
:click me
performance
please refer to MLPerf v4.0 results. We will consider adding mlperf-related example for guidance and CI test.
Expected Behavior & Potential Risk
the expected behavior that triggered by this PR
How has this PR been tested?
how to reproduce the test (including hardware information)
Dependency Change?
any library dependency introduced or removed