pytorch/benchmark

Pull Requests Commits

Update transformers and TIMM with Dependabot

huydhn committed 119 days ago

5463a956

Add HF LLM benchmarks (#156967)

angelayi committed 121 days ago

7c121441

Update 3.5 pin to 5ae38bdb0dc066c5823e34dc9797afb9de42c866 (#162821)

davidberard98 committed 125 days ago

f2a52834

Fix issues with guard serialization on distributed types. (#162418)

zhxchen17 committed 125 days ago

f42b77a7

update triton pin to triton 3.5 (#162278)

davidberard98 committed 128 days ago

b8e3ab6c

treat like a template in GEMMs (#161342)

coconutruben committed 130 days ago

66e1f3d3

Add model tracing benchmark (#2642)

xuzhao9 committed 132 days ago

d511766a

Add `range_iterator` (#161800)

generatedunixname499836121 committed 133 days ago

d4a3859b

Guard config copy for pickle errors (#161659)

oulgen committed 140 days ago

d6093979

Revert D81100936: Close some sources of fake tensor leakage

atalman committed 140 days ago

91ce9dce

Back out "Add caching for inside torch.compile.disable function to avoid unnecessary recompilation. (#160934)"

Akshay Mahesh committed 140 days ago

888161d7

Install the right numpy version matching PyTorch CI (#2640)

huydhn committed 140 days ago

74a23fef

add nested graph break tests (#144516)

williamwen42 committed 140 days ago

a153beae

Close some sources of fake tensor leakage

tugsbayasgalan committed 141 days ago

bb8c542a

Use np._core instead of np.core (#2637)

huydhn committed 142 days ago

25ac7514

Restore TorchBench PR tests (#2636)

huydhn committed 143 days ago

9ebf2e28

Add torchscript jit.trace to benchmark option (#161223)

yiming0416 committed 144 days ago

9644652f

Log exception_stack_trace to dynamo_compile (#161096)

jovianjaison committed 146 days ago

78745baf

Revamp TorchBench nightly image build process (#2635)

huydhn committed 147 days ago

b0efcbe2

Refactor convert_frame._compile_inner to return compiled bytecode + output graph. [3/n] (#160855)

zhxchen17 committed 147 days ago

abd8dd23

Add caching for inside torch.compile.disable function to avoid unnecessary recompilation. (#160934)

generatedunixname499836121 committed 147 days ago

3e377c41

Type coverage for symbolic_convert (#160922)

Lucaskabela committed 148 days ago

73eb12ed

Cleanup unused TorchBench workflows (#2634)

huydhn committed 148 days ago

9a4b981e

Revert "Fix the benchmark CI failure due to Opacus 1.5.4 (#2621)" (#2632)

huydhn committed 149 days ago

22bc29b4

pack linear for FP32 dynamic mode (#157542)

generatedunixname499836121 committed 150 days ago

af93b8be

Add GPT2ForSequenceClassification to use_larger_multiplier_for_smaller_tensor list (#160001)

generatedunixname499836121 committed 150 days ago

9e418359

Add logging for when inbuilt_inline_nn_modules will help with ID_MATCH guard triggered recompiles (#160592)

ppanchalia committed 153 days ago

6da75d11

Type improvements in `_dynamo/utils` to generics (#159824)

Lucaskabela committed 153 days ago

24586097

Log graph_node_shape to dynamo_compile (#160556)

jovianjaison committed 154 days ago

e11cbaa1

Make yolov3 exportable (#2630)

tugsbayasgalan committed 155 days ago

290c966f

Older