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

Convert documentation to TFLite #129

Closed
2 tasks done
boomanaiden154 opened this issue Sep 3, 2022 · 0 comments
Closed
2 tasks done

Convert documentation to TFLite #129

boomanaiden154 opened this issue Sep 3, 2022 · 0 comments

Comments

@boomanaiden154
Copy link
Collaborator

boomanaiden154 commented Sep 3, 2022 •
edited
Loading

Some of the documentation that is around currently, particularly the inliner demo and some of the benchmarking documentation, as well as some of the benchmarking code itself, are stuck with the Tensorflow C API rather than TFLite. This documentation/code (just benchmarking) should be updated to work with TFLite.

I should be able to do this, just trying to find some time.

  • Update inlining demo
  • Update benchmarking code/documentation to work with TFLite
@boomanaiden154 boomanaiden154 mentioned this issue Sep 4, 2022
Merged
boomanaiden154 added a commit to boomanaiden154/ml-compiler-opt that referenced this issue Sep 25, 2022
@boomanaiden154
Removed tensorflow_c_lib_path flag from LLVM build for benchmarking
8589913 
Should help with the other half of google#129. The benchmarking code currently
sets the TENSORFLOW_C_LIB_PATH CMake variable when building LLVM.
However, this doesn't make sense going forward as we have shifted mostly
everything over to TFLite and we don't need to have MLGO development
mode enabled for benchmarking in either case. This patch explicitly
removes the flag so that the benchmarks simply build LLVM in MLGO
release mode.
@boomanaiden154 boomanaiden154 mentioned this issue Sep 25, 2022
Merged
mtrofin pushed a commit that referenced this issue Sep 26, 2022
@boomanaiden154
Removed tensorflow_c_lib_path flag from LLVM build for benchmarking (#…
016f398 
…150)

Should help with the other half of #129. The benchmarking code currently
sets the TENSORFLOW_C_LIB_PATH CMake variable when building LLVM.
However, this doesn't make sense going forward as we have shifted mostly
everything over to TFLite and we don't need to have MLGO development
mode enabled for benchmarking in either case. This patch explicitly
removes the flag so that the benchmarks simply build LLVM in MLGO
release mode.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@boomanaiden154