pytorch/benchmark

Pull Requests Commits

use torch.compile ca API for benchmarks (#149647)

xmfan committed 304 days ago

7481a407

Unify `cuBLASLt` workspaces with `cuBLAS` workspaces (#145130)

generatedunixname499836121 committed 305 days ago

90e73750

Add python version to dynamo_compile table (#149419)

masnesral committed 309 days ago

10a7be34

Switch off inference mode during compilation (#149321)

anijain2305 committed 310 days ago

2c5bc4ad

fix dynamic_shapes spec for moco (#148772) (#2601)

pianpwk committed 311 days ago

d1b2abbf

fix two accuracy regression (#149172)

shunting314 committed 311 days ago

50e2f744

NotImplementedError: Model's DEFAULT_TRAIN_BSIZE is not implemented. (#2563)

ostrowskimarcin committed 315 days ago

4dc08945

Set compile_id in the CachingAutotuner during compilation so we have it for dynamo_timed logging (#148693)

masnesral committed 316 days ago

ee4aa8ba

add log for skip reasons

BoyuanFeng committed 318 days ago

c271568d

dynamo: Count number of opcodes processes (#147149)

c00w committed 318 days ago

737798bb

Backout D70075331

renganxu committed 318 days ago

29078f68

Revert "add APIs to determine a class is a namedtuple or PyStructSequence (#113257)"

generatedunixname499836121 committed 319 days ago

d1ba2efd

Log cudagraphify timings to dynamo_timed (#143220)

masnesral committed 321 days ago

f291d7be

Properly account for non-list instances in list comparison (#148470)

StrongerXi committed 322 days ago

9e5a9582

online softmax (#127011)

shunting314 committed 322 days ago

934ad61d

add APIs to determine a class is a namedtuple or PyStructSequence (#113257)

generatedunixname499836121 committed 322 days ago

423727cd

Skip torchbench models not supported by export (#148359)

desertfire committed 323 days ago

4d53c5ac

Log individual Triton kernel compilation times to dynamo_compile (#147022)

masnesral committed 325 days ago

0dba3d91

disable cudagraph for tts_angular on dashboard (#148221)

BoyuanFeng committed 327 days ago

182024e2

more better error messages [3/N] (#147494)

williamwen42 committed 329 days ago

ee785385

migrate PYFMT for `torch._dynamo` to `ruff format` (#144549)

generatedunixname499836121 committed 329 days ago

616de544

bump `ruff` to 0.9.2: multiline `assert` statements (#144546)

generatedunixname499836121 committed 329 days ago

19fc1702

Log graph breaks (#146537)

Raymo111 committed 329 days ago

b911f13a

Use the new `get_unique_name_wrt` helper when applicable (#146950)

StrongerXi committed 330 days ago

900bde3e

Initial support for `nonstrict_trace` (#146367)

StrongerXi committed 330 days ago

fe8fd1ba

Add basic Gaudi support to benchmarks/dynamo (#145920)

generatedunixname499836121 committed 330 days ago

de3d4e9e

add generic graph break hints (#147429)

williamwen42 committed 330 days ago

748da286

make some more graph break messages readable in English [2/N] (#147385)

williamwen42 committed 330 days ago

e95036bc

Unify `cuBLASLt` workspaces with `cuBLAS` workspaces (#145130)

generatedunixname499836121 committed 333 days ago

818315c5

Add XPU to is_compile_supported to support roi_align op in torchvision (#147541)

generatedunixname499836121 committed 333 days ago

f6726de3

Newer Older