Skip to content

Binary Size: Selective Build #9010

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
Open
cbilgin opened this issue Mar 6, 2025 · 3 comments
Open

Binary Size: Selective Build #9010

cbilgin opened this issue Mar 6, 2025 · 3 comments
Labels
module: build/install Issues related to the cmake and buck2 builds, and to installing ExecuTorch module: xnnpack Issues related to xnnpack delegation and the code under backends/xnnpack/ triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Milestone
0.6.0

Comments

@cbilgin
Copy link

cbilgin commented Mar 6, 2025 •
edited by pytorch-bot bot
Loading

cc @digantdesai @mcr229 @larryliu0820 @jathu @lucylq
@cbilgin cbilgin moved this to Backlog in ExecuTorch - CPU Mar 6, 2025
@cbilgin cbilgin added this to the 0.6.0 milestone Mar 6, 2025
@iseeyuan iseeyuan added module: build/install Issues related to the cmake and buck2 builds, and to installing ExecuTorch triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module labels Mar 7, 2025
@github-actions github-actions bot mentioned this issue Mar 10, 2025
Open
This was referenced Mar 17, 2025
Open
Open
@lucylq lucylq added the module: xnnpack Issues related to xnnpack delegation and the code under backends/xnnpack/ label Mar 27, 2025
@github-actions github-actions bot mentioned this issue Mar 31, 2025
Open
@digantdesai
Copy link
Contributor

@mcr229 - @larryliu0820 and I discussed this a bit in other context. But there might be an apatite to solve this at ET level which can work for all the delegates. A simplest approach might be to generate some metadata along with each preprocess which can be consumed at delegate compile time. Parsing delegate blob can be an another, rather tedious, alternative.

@mcr229
Copy link
Contributor

mcr229 commented Apr 1, 2025

@digantdesai there would be some level of support on the XNNPACK side in order to selectively choose right?

@digantdesai
Copy link
Contributor

Yeah, we would need to upstream some support but I can imagine if we only need int8 ops then we can compile time turn off FP32, FP16 etc.

@github-actions github-actions bot mentioned this issue Apr 7, 2025
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
module: build/install Issues related to the cmake and buck2 builds, and to installing ExecuTorch module: xnnpack Issues related to xnnpack delegation and the code under backends/xnnpack/ triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Projects
ExecuTorch - CPU
Status: Backlog
Milestone
0.6.0
Development

No branches or pull requests
5 participants
@digantdesai @iseeyuan @lucylq @cbilgin @mcr229