Commits

Commits on Apr 27, 2022

fix lazy_bench.py precision option (#881 )

Krovatkin
authored and
facebook-github-bot
committedApr 27, 2022

Commits on Apr 8, 2022

adding lazy_bench tests (#846 )

Krovatkin
authored and
facebook-github-bot
committedApr 8, 2022

Commits on Apr 6, 2022

Porting fixes from wconstab/ltc to main (#851 )

Krovatkin
authored and
facebook-github-bot
committedApr 6, 2022

Commits on Jun 14, 2021

freeze attention (#399 )
Krovatkin
committedJun 14, 2021

Commits on Jun 7, 2021

freeze stargan (#394 )
Krovatkin
committedJun 7, 2021

Commits on Jun 4, 2021

use mobilenet_v3 from torchvision (#389 )
Krovatkin
committedJun 4, 2021
clean up fuser option (#393 )
Krovatkin
committedJun 4, 2021

Commits on Jun 3, 2021

Add freeze suffix to test names (#391 )
Krovatkin
committedJun 3, 2021
add NVFuser (#260 )
Krovatkin
committedJun 3, 2021
freeze LearningToPaint (#390 )
Krovatkin
committedJun 3, 2021

Commits on Jun 2, 2021

fix batch sizes (#346 )
Krovatkin
committedJun 2, 2021
run timm models in optimized_for_inference mode (#383 )
Krovatkin
committedJun 2, 2021

Commits on May 26, 2021

replace decomposed Gelu with Pytorch GELU (#384 )
Krovatkin
committedMay 26, 2021
update template file; add optimize_for_inference (#378 )
Krovatkin
committedMay 26, 2021

Commits on May 25, 2021

add a call to super class c-tor (#305 )
Krovatkin
committedMay 25, 2021

Commits on Mar 10, 2021

properly call super ctor and get rid of extra ctor (#303 )
Krovatkin
committedMar 10, 2021
Update __init__.py (#301 )
Krovatkin
committedMar 10, 2021

Commits on Mar 5, 2021

write_sys_file expects str or write throws (#298 )
Krovatkin
committedMar 5, 2021

Commits on Mar 2, 2021

add check_opt_vs_noopt_jit option to check that results between baseline and optimized jitted versions match (#249 )
Krovatkin
committedMar 2, 2021

Commits on Feb 24, 2021

fix vision models to use the correct device (#274 )
Krovatkin
committedFeb 24, 2021

Commits on Feb 16, 2021

[WIP] Set up torchbenchmark for running against custom PyTorch (#253 )
Krovatkin
committedFeb 16, 2021

Commits on Feb 5, 2021

no grad (#181 )
Krovatkin
committedFeb 5, 2021
bbrk fix (#241 )
Krovatkin
committedFeb 5, 2021
Make sure diagnostic variables aren't set when benchmarking (#238 )
Krovatkin
committedFeb 5, 2021
set eval and train modes before benchmarking a model (#226 )
Krovatkin
committedFeb 5, 2021

Commits on Feb 1, 2021

fix version for now (#227 )
Krovatkin
committedFeb 1, 2021

Commits on Sep 22, 2020

fix a broken link to ADDING_MODELS.md (#120 )
Krovatkin
committedSep 22, 2020