Commits pytorch/benchmark

add option to force amp to use bfloat16 instead of float16 (#169449)

xmfan committed 118 days ago

cf3a00c9

Revert D88129557: Decentralize and Improve key hash implementation for Dict variable tracker (#169204)

Jiaming Tao committed 118 days ago

7e64c600

Decentralize and Improve key hash implementation for Dict variable tracker (#169204)

anijain2305 committed 118 days ago

8c8edc66

Fix more local test failures (#169228)

jansel committed 118 days ago

c42eb289

Fix local test failures for logging (#168927)

jansel committed 119 days ago

b53822e1

fbcode/pytorch/benchmark/torchbenchmark/models/maml/learner.py

generatedunixname537391475639613 committed 119 days ago

0aa12c69

Update(deps): Bump transformers from 4.55.2 to 4.57.3 (#2658)

dependabot[bot] committed 120 days ago

49758eae

Ensure `typing.Union` is correctly traced in Dynamo. (#169084)

generatedunixname499836121 committed 122 days ago

9e56263e

Enable SIM118 (#167399)

generatedunixname499836121 committed 123 days ago

ff4a8a30

Skip stable-diffusion-2 on TorchBench CI (#2656)

huydhn committed 127 days ago

ad5470be

fbcode/pytorch/benchmark/torchbenchmark/_components/model_analyzer/dcgm/dcgm_structs.py

generatedunixname537391475639613 committed 133 days ago

5ec24026

Fix: Dynamo log always emits ANSI color codes into torch_compile_debug/torchdynamo/debug.log due to colored=True in lazy_format_graph_code (#167823)

generatedunixname499836121 committed 134 days ago

d3ab9c6f

unittest for run2run determinism (#167482)

shunting314 committed 134 days ago

2b7a1a47

Skip stable diffusion models in torchbench, get tests and benchmarks green (#167896)

zou3519 committed 136 days ago

30788089

Imporve-graph-break-skip-logs (#167067)

generatedunixname499836121 committed 137 days ago

987b3dc9

fbcode/pytorch/benchmark/torchbenchmark/models/pytorch_stargan/model.py

generatedunixname537391475639613 committed 139 days ago

0c70462c

fbcode/pytorch/benchmark/torchbenchmark/models/LearningToPaint/baseline/train_renderer.py

generatedunixname537391475639613 committed 139 days ago

5314e1c7

Revert "fix wrong accuracy_status when exception. (#165731)"

generatedunixname499836121 committed 139 days ago

44c5b2ff

fbcode/pytorch/benchmark/torchbenchmark/models/nvidia_deeprecommender/reco_encoder/model/model.py

generatedunixname537391475639613 committed 140 days ago

9a5a71c6

Add backend optimus

xuzhao9 committed 141 days ago

41cf5a1b

rename unimplemented_v2 -> unimplemented (#167150)

williamwen42 committed 143 days ago

a369b7c0

Fix CQS signal F811 in fbcode/pytorch/benchmark/torchbenchmark/models/simple_gpt

generatedunixname537391475639613 committed 143 days ago

f2aa1ede

Use key in dict for existence checks (#167285)

generatedunixname499836121 committed 144 days ago

274a1c37

Trace events with the new ops (#167177)

mlazos committed 144 days ago

9369bf1f

fbcode/pytorch/benchmark/torchbenchmark/e2e_models/hf_bert/trainer.py

generatedunixname537391475639613 committed 144 days ago

f46993a4

Fix CQS signal F811 in fbcode/pytorch/benchmark/torchbenchmark/models/LearningToPaint/baseline_modelfree

generatedunixname537391475639613 committed 145 days ago

6a9f4a83

fbcode/pytorch/benchmark/torchbenchmark/models/tacotron2/audio_processing.py

generatedunixname537391475639613 committed 145 days ago

534d68b7

Add graph annotation checks (#167019)

mlazos committed 145 days ago

4169fc33

fbcode/pytorch/benchmark/torchbenchmark/models/pytorch_CycleGAN_and_pix2pix/models/networks.py

generatedunixname537391475639613 committed 145 days ago

1ab9a82a

fbcode/pytorch/benchmark/torchbenchmark/models/timm_efficientdet/train.py

generatedunixname537391475639613 committed 145 days ago

85943bef