pytorch/benchmark

Pull Requests Commits

move clamp range from value to context

BoyuanFeng committed 2 years ago

165ac24f

register squashed-normal-type. pass test for drq

BoyuanFeng committed 2 years ago

bf3e59c4

Support triton.language.dtype with torch.compile -- Second Attempt (#122141)

oulgen committed 2 years ago

91a1d324

fix dlrm accuracy test in max-autotune (#122012)

shunting314 committed 2 years ago

968053d9

Optimize builtin handling (#122035)

jansel committed 2 years ago

74e37af4

Log restart reasons and extra compile time in CompilationMetrics (#121827)

jamesjwu committed 2 years ago

d810ac7d

add option to print compilation time (#121996)

eellison committed 2 years ago

d8685408

Move `components` to `torchbenchmark/_components` (#2201)

xuzhao9 committed 2 years ago

230bd25e

Enable package import from torchbenchmark and torchbenchmark.models. (#2200)

xuzhao9 committed 2 years ago

3bb81e57

Fix torch.compile links (#121824)

lezcano committed 2 years ago

3b76cd46

Move doc links to point to main (#121823)

lezcano committed 2 years ago

ce0278ce

Back out "Support `triton.language.dtype` with `torch.compile` (#121690)"

oulgen committed 2 years ago

87ecea07

Convert invalid args into graph breaks (#121784)

jansel committed 2 years ago

5a6d0dc7

Remove error anti-pattern when dealing with dynamic shape output (#121681)

gmagogsfm committed 2 years ago

a7624953

Minor compile time optimizations in torch.py (#121615)

jansel committed 2 years ago

8f9a2262

Support `triton.language.dtype` with `torch.compile` (#121690)

oulgen committed 2 years ago

0a4eeedb

add graph freezing and weight prepacking argument (#2188)

snadampal committed 2 years ago

4246ce43

add more log when accuracy check fail (#121656)

shunting314 committed 2 years ago

3b0de366

Measure Dynamo cache latency lookup (#121604)

anijain2305 committed 2 years ago

6ea22b31

Fix installation for diffuser_instruct_pix2pix (#2184)

JasonMts committed 2 years ago

0ea6f93f

Handle inductor config attribute error

xuzhao9 committed 2 years ago

6ba2382c

Provide requirements that fix hf MPT 7b instuct (#2194)

JasonMts committed 2 years ago

d24ff043

Re-sync with internal repository (#2196)

facebook-github-bot committed 2 years ago

Verified 0826dcc1

Add triton operator benchmark (#2195)

xuzhao9 committed 2 years ago

d0d5b67c

Fix import error in torchbench

nmacchioni committed 2 years ago

66c90d52

Update torchbench commit pin, add sam_fast benchmark (#121420)

jamesjwu committed 2 years ago

39a78931

Refactor torchdynamo and inductor options

xuzhao9 committed 2 years ago

8ffe57ec

add more refleak tests (#120657)

williamwen42 committed 2 years ago

b4a59daf

Fix BERT_pytorch model on bf16. (#2185)

xuzhao9 committed 2 years ago

37fa676a

Add sam_fast torchbench (#2182)

jamesjwu committed 2 years ago

d6015d42

Older