[WIP, Kernel] (2/N) Machete - Integrate into GPTQMarlinLinearMethod and CompressedTensorsWNA16 #5
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
|
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge). To run full CI, you can do one of these:
🚀 |
LucasWilkinson
force-pushed
the
lwilkinson/machete
branch
2 times, most recently
from
8b21235 to
2868f5d
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
11a9dec to
529c59e
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete
branch
from
e9c70f8 to
6955a93
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
529c59e to
0bcd9c1
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete
branch
5 times, most recently
from
60021df to
e92b26e
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
df80f72 to
57a8011
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete
branch
3 times, most recently
from
a280110 to
ad5771a
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
57a8011 to
d2431a7
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
3 times, most recently
from
1361be1 to
d5ee5b8
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
d5ee5b8 to
735259b
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
2 times, most recently
from
953973d to
90f8bb6
Compare
move heuristic into C++ code fix unit tests + format update for 3.5.1 remove custom scheduler codespell cleanup comment cleanup diff review comments review comments review comment changes review comments fix codespell cleanup util logic make dim names for prepack layout more canoncial missed refactor wip interleaving + recasting tweak tolerances comments plus interleaving format codespell review comments end2end first pass seperate out kernels, format add machete as a gptq backend update to use ModelWeightParameter formatting update parameter.py refactor permute layout wip
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
5a38aac to
7bc8316
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
7bc8316 to
096dd4a
Compare
LucasWilkinson
force-pushed
the
lwilkinson/machete-end2end
branch
from
ac45f19 to
a98f691
Compare
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.
End2end integration into GPTQMarlinLinearMethod and CompressedTensorsWNA16.