Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Increase number of cores used to calculate upsample for YOLO and Stable diffusion #16351

Draft
wants to merge 15 commits into
base: main
Choose a base branch
from
Draft

Increase number of cores used to calculate upsample for YOLO and Stable diffusion #16351

wants to merge 15 commits into from

Conversation

nkpatel-tt
Copy link
Contributor

Ticket

Link to Github Issue

Problem description

Currently, Upsample processes one row per core, which can impact the performance.

What's changed

Instead single row per core granularity, make it work for minimum possible shard size for each core.

Checklist

  • Post commit CI passes
  • Blackhole Post commit (if applicable)
  • Model regression CI testing passes (if applicable)
  • Device performance regression CI testing passes (if applicable)
  • (For models and ops writers) Full new models tests passes
  • New/Existing tests provide coverage for changes

@nkpatel-tt
Remove restriction of input_nsticks_per_core % w == 0 for height shar…
b82e38d 
…ded tensor inputs.

Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Add support for block sharding for upsample.
8890e70 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Add support for block sharding.
e39f225 
	ToDo: commonize code.

Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Commonize code for block and height sharding.
cbcf042 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Clean up and add support for Column major shard orientation.
67eba82 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Align code
6fcb471 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Merge branch 'main' into nkpatel/upsample2d-generalization-remove-res…
d122f35 
…trictions-input_width

Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Make pipeline work.
fdb514b 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Merge branch 'main' into nkpatel/upsample2d-generalization-remove-res…
b306fba 
…trictions-input_width
@nkpatel-tt
Make pipeline pass.
09b57cb 
Suggested-by: Pavle Josipović <pavlejosipovic@users.noreply.github.com>
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Merge branch 'main' into nkpatel/upsample2d-generalization-remove-res…
efaf2ec 
…trictions-input_width
@nkpatel-tt
Address review comments.
ece940a 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Reduce size of required CB for config buffer by half.
75973e1 
Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Merge remote-tracking branch 'origin/nkpatel/upsample2d-generalizatio…
de8c6dc 
…n-remove-restrictions-input_width' into nkpatel/upsample_perf_improve
@nkpatel-tt
Increase number of cores used for calculations for upsample for stabl…
52cdc41 
…e diffusion and yolo.

Signed-off-by: Nilaykumar Patel <nkpatel@tenstorrent.com>
@nkpatel-tt
Copy link
Contributor Author

nkpatel-tt commented Dec 30, 2024 •
edited
Loading

As of now current branch has this branch merged which is being review at PR.

Once that branch is merged, will remove those changes and rebase.

@nkpatel-tt nkpatel-tt changed the title Upsample increase number of cores for YOLO and Stable diffusion Increase number of cores used to calculate upsample for YOLO and Stable diffusion Dec 30, 2024
@nkpatel-tt nkpatel-tt self-assigned this Dec 30, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@esmalTT esmalTT Awaiting requested review from esmalTT esmalTT will be requested when the pull request is marked ready for review esmalTT is a code owner

@uaydonat uaydonat Awaiting requested review from uaydonat uaydonat will be requested when the pull request is marked ready for review uaydonat is a code owner

@dvartaniansTT dvartaniansTT Awaiting requested review from dvartaniansTT dvartaniansTT will be requested when the pull request is marked ready for review dvartaniansTT is a code owner

@ayerofieiev-tt ayerofieiev-tt Awaiting requested review from ayerofieiev-tt ayerofieiev-tt will be requested when the pull request is marked ready for review ayerofieiev-tt is a code owner

@dmakoviichuk-tt dmakoviichuk-tt Awaiting requested review from dmakoviichuk-tt dmakoviichuk-tt will be requested when the pull request is marked ready for review dmakoviichuk-tt is a code owner

@rfurko-tt rfurko-tt Awaiting requested review from rfurko-tt rfurko-tt will be requested when the pull request is marked ready for review rfurko-tt is a code owner

@cfjchu cfjchu Awaiting requested review from cfjchu cfjchu will be requested when the pull request is marked ready for review cfjchu is a code owner

@TT-BrianLiu TT-BrianLiu Awaiting requested review from TT-BrianLiu TT-BrianLiu will be requested when the pull request is marked ready for review TT-BrianLiu is a code owner

@razorback3 razorback3 Awaiting requested review from razorback3 razorback3 will be requested when the pull request is marked ready for review razorback3 is a code owner

@dongjin-na dongjin-na Awaiting requested review from dongjin-na dongjin-na will be requested when the pull request is marked ready for review dongjin-na is a code owner

@bbradelTT bbradelTT Awaiting requested review from bbradelTT bbradelTT will be requested when the pull request is marked ready for review bbradelTT is a code owner

At least 0 approving reviews are required to merge this pull request.

Assignees

@nkpatel-tt nkpatel-tt

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@nkpatel-tt