#11512: Add sweeps for eltwise sharded ops #16003
Open
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
bfilipovicTT
requested review from
patrickroberts,
yan-zaretskiy and
eyonland
as code owners
bfilipovicTT
force-pushed
the
bfilipovicTT/sharded-sweeps-1
branch
from
27fe7a4 to
f9364b1
Compare
bfilipovicTT
requested review from
xanderchin,
jdesousa-TT and
sjameelTT
as code owners
bfilipovicTT
force-pushed
the
bfilipovicTT/sharded-sweeps-1
branch
from
f9364b1 to
ba3b27f
Compare
bfilipovicTT
changed the title
bfilipovicTT
force-pushed
the
bfilipovicTT/sharded-sweeps-1
branch
from
48d9f6b to
3b4770a
Compare
bfilipovicTT
changed the title
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Outdated suggestions cannot be applied.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.
Ticket
#11512
Problem description
Missing sharded sweeps
What's changed
There are now sweeps for sharded version of ops:
unary: clamp, floor, clone, cbrt
binary: add,subtract, div, div_no_nan
Test REsults and statistics
tests/sweep_framework/sweeps/eltwise/unary/clamp/clamp_sharded.py: 261 fail, 891 pass (77.34375%)
tests/sweep_framework/sweeps/eltwise/unary/cbrt/cbrt_sharded.py: 143 fail, 625 pass (81.38020833333334%)
tests/sweep_framework/sweeps/eltwise/unary/floor/floor_sharded.py: 286 fail, 1250 pass (81.38020833333334%)
tests/sweep_framework/sweeps/eltwise/unary/clone/clone_sharded.py: not yet supported in sharded mode
tests/sweep_framework/sweeps/eltwise/binary/subtract/subtract_sharded.py: 348 fail, 1188 pass (77.34375%)
tests/sweep_framework/sweeps/eltwise/binary/add/add_sharded.py: 348 fail, 1188 pass (77.34375%)
Note that all the tests create memory configurations for tensors with create_sharded_memory_config_ method and all of the tests that fail have arguments use_height_and_width_as_shard_shape=false and strategy=ShardStrategy.BLOCK
tests/sweep_framework/sweeps/eltwise/binary/div/div_sharded.py
and tests/sweep_framework/sweeps/eltwise/binary/div_no_nan/div_no_nan_sharded.py
crash on every run because an assert fails:
TT_FATAL @ ../ttnn/cpp/ttnn/operations/eltwise/unary/device/unary_sharded_program_factory.cpp:45: input_tile_size == output_tile_size
info:
Input and output tile size should be same
I have created an issue about it here:
#16054