saymrwulf/pytorch
1
0
Fork 0
mirror of https://github.com/saymrwulf/pytorch.git synced 2026-05-14 20:57:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
pytorch/benchmarks/dynamo/pr_time_benchmarks
History
Aaron Orenstein f2b744b9ca dynamo tracing perf: import_module: 59.92 -> 52.9 (#143057) 
See #143056 for overall docs.

This PR: Using `importlib.import_module()` within the hot path of
symbolic_convert is slow. Memoize it.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/143057
Approved by: https://github.com/jansel
2024-12-22 06:38:38 +00:00
..
benchmarks [Codemod][AddExplicitStrictExportArg] caffe2/benchmarks/dynamo (#143686) 2024-12-21 19:56:56 +00:00
test_check_result Several enhancements for check_results.py (#137925) 2024-10-26 16:27:55 +00:00
__init__.py
benchmark_base.py Record PR time benchmark results in JSON format (#140493) 2024-11-20 18:54:01 +00:00
benchmark_runner.sh Only keep ListOfLinears module in basic_modules_benchmarks and add gpu version. (#135730) 2024-09-14 16:45:52 +00:00
check_results.py Only compute new_untracked_symbols and new_unbacked_bindings if needed. (#140083) 2024-11-20 19:28:18 +00:00
expected_results.csv dynamo tracing perf: import_module: 59.92 -> 52.9 (#143057) 2024-12-22 06:38:38 +00:00
log_benchmarking_time.py Only keep ListOfLinears module in basic_modules_benchmarks and add gpu version. (#135730) 2024-09-14 16:45:52 +00:00
README.md add README.md for compile time benchmarks (#143145) 2024-12-13 05:12:26 +00:00

README.md

Instructions on how to make a new compile time benchmark

  1. Make a new benchmark file in /benchmarks/dynamo/pr_time_benchmarks/benchmarks/ eg. 0b75b7ff2b/benchmarks/dynamo/pr_time_benchmarks/benchmarks/add_loop.py
  2. cd into the pr_time_benchmarks directory cd benchmarks/dynamo/pr_time_benchmarks
  3. Run PYTHONPATH=./ python benchmarks/[YOUR_BENCHMARK].py a.txt
  4. (Optional) flip a flag that you know will change the benchmark and run again with b.txt PYTHONPATH=./ python benchmarks/[YOUR_BENCHMARK].py a.txt
  5. Compare a.txt and b.txt located within the benchmarks/dynamo/pr_time_benchmarks folder to make sure things look as you expect
  6. Check in your new benchmark file and submit a new PR
  7. In a few days, if your benchmark is stable, bug Laith Sakka to enable running your benchmark on all PRs. If your a meta employee, you can find the dashboard here: internalfb.com/intern/unidash/dashboard/pt2_diff_time_metrics