Support 2D inner reduction scheduler with autotuning #3456
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
rdspring1
added
the
Autotune
rdspring1
force-pushed
the
autotune_inner_reduction_2d
branch
from
c79d982 to
de10254
Compare
rdspring1
changed the title
rdspring1
force-pushed
the
autotune_inner_reduction_2d
branch
4 times, most recently
from
31c216c to
a466e32
Compare
rdspring1
force-pushed
the
autotune_inner_reduction_2d
branch
from
4a31a07 to
7eeebe1
Compare
liqiangxl
reviewed
Nov 25, 2024
rdspring1
force-pushed
the
autotune_pointwise_single
branch
from
0849e1b to
1c78f7d
Compare
rdspring1
force-pushed
the
autotune_inner_reduction_2d
branch
from
7eeebe1 to
7817368
Compare
|
!build |
Support Gelu-Bias, Silu-Mul, Bcast-Add, Mul Fusions
rdspring1
force-pushed
the
autotune_pointwise_single
branch
from
1c78f7d to
45b74f6
Compare
rdspring1
force-pushed
the
autotune_inner_reduction_2d
branch
from
7817368 to
e7ffb29
Compare
|
!build |
jacobhinkle
approved these changes
Dec 9, 2024
| # SPDX-FileCopyrightText: Copyright (c) 2024-present NVIDIA CORPORATION & AFFILIATES. | ||
| # All rights reserved. | ||
| # SPDX-License-Identifier: BSD-3-Clause | ||
| # Owner(s): ["module: nvfuser"] |
There was a problem hiding this comment.
I'm curious: I noticed this # Owner(s): ["module: nvfuser"] line in other python files in our repo. Is this a standard and is it used by some tool?
rdspring1
changed the base branch from
autotune_pointwise
to
autotune_pointwise_single
|
!build |
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Outdated suggestions cannot be applied.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.
Summary:
This PR creates an SOL autotuning script for the 2d inner reduction scheduler. It trains a random forest to predict the best performing configuration for the reduction scheduler.
Inner Reduction Fusions:
y = sum(x, dim=-1)z = sum(x1 + x2 + x3 + x4, dim=-1)y = sum(tanh(x), dim=-1)z = sum(exp(x), dim=-1)