COMMITS
/ benchmarks/float8/utils.py August 21, 2025
V
add simple roofline for float inference with rowwise scaling (#2819)
Vasiliy Kuznetsov committed
August 15, 2025
D
[moe training] use llama4 shapes for kernel benchmarks (#2756)
Daniel Vega-Myhre committed
July 3, 2025
D
[moe training] add benchmarking script for grouped mm (#2490)
Daniel Vega-Myhre committed
June 24, 2025
D
add-to-benchmarks (#2427)
Driss Guessous committed
J
Revert "Build mxfp4 kernel for sm120a" (#2428)
Jerry Zhang committed
June 21, 2025
T
Build mxfp4 kernel for sm120a (#2285)
Thien Tran committed
April 25, 2025
D
Ignore unreconized for now (#2025)
Driss Guessous committed
April 22, 2025
J
Update ruff version to 0.11.6 (#2103)
Jerry Zhang committed
March 4, 2025
V
roofline estimator: add float8 rowwise and mxfp8 recipe support (#1789)
Vasiliy Kuznetsov committed
February 27, 2025
V
[1/x] mx roofline: make the script work on NVIDIA B200 (#1778)
Vasiliy Kuznetsov committed
February 24, 2025
V
add MX support to lowp training profiling script (#1765)
Vasiliy Kuznetsov committed
January 8, 2025
A
Lint benchmark folder (#1519)
Apurva Jain committed
September 4, 2024
V
make f8 roofline script calculate observed overhead (#734)
Vasiliy Kuznetsov committed
August 12, 2024
V
float8 profiling: save torch_logs, and attach them to profiling trace (#645)
Vasiliy Kuznetsov committed
August 8, 2024
V
float8 profiling script: filter out microbenchmarking overhead (#629)
Vasiliy Kuznetsov committed
August 7, 2024
V
update float8 benchmarks to be more useful for smaller shapes (#615)
Vasiliy Kuznetsov committed
July 30, 2024
V
move float8_experimental to torchao/float8
Vasiliy Kuznetsov committed