frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-26 16:44:54 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
956bd038088bd4e31f1adee99466f5674b51a87c
pytorch/benchmarks
History
Jongsoo Park b91eb97d34 [transformer benchmark] relax tolerance in sdp.py (#101965) 
Summary:
Otherwise we get
```
Traceback (most recent call last):
  File "<string>", line 49, in <module>
  File "<string>", line 47, in __run
  File "/usr/local/fbcode/platform010/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/usr/local/fbcode/platform010/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/data/users/jongsoo/fbsource/buck-out/v2/gen/fbcode/ef4169ac7f95fb74/caffe2/benchmarks/transformer/__sdp__/sdp#link-tree/caffe2/benchmarks/transformer/sdp.py", line 346, in <module>
    main(save_path)
  File "/data/users/jongsoo/fbsource/buck-out/v2/gen/fbcode/ef4169ac7f95fb74/caffe2/benchmarks/transformer/__sdp__/sdp#link-tree/caffe2/benchmarks/transformer/sdp.py", line 328, in main
    experiment = run_single_experiment(experiment_config)
  File "/data/users/jongsoo/fbsource/buck-out/v2/gen/fbcode/ef4169ac7f95fb74/caffe2/benchmarks/transformer/__sdp__/sdp#link-tree/caffe2/benchmarks/transformer/sdp.py", line 229, in run_single_experiment
    assert_close_tensors(nn_mha_output, composite_mha_output)
  File "/data/users/jongsoo/fbsource/buck-out/v2/gen/fbcode/ef4169ac7f95fb74/caffe2/benchmarks/transformer/__sdp__/sdp#link-tree/caffe2/benchmarks/transformer/sdp.py", line 196, in assert_close_tensors
    assert torch.allclose(a, b, atol=1e-3, rtol=1e-3)
AssertionError
```

Test Plan: buck run mode/dev-nosan //caffe2/benchmarks/transformer:sdp

Differential Revision: D45843836

Pull Request resolved: https://github.com/pytorch/pytorch/pull/101965
Approved by: https://github.com/drisspg
2023-05-23 06:54:08 +00:00
..
cpp
moving nvfuser benchmark to third_party/nvfuser (#96725)
2023-03-21 23:19:15 +00:00
distributed
[BE]: Update flake8 and plugins and fix bugs (#97795)
2023-03-28 23:51:55 +00:00
dynamo
Suppress profiler spam in dynamo benchmarks (#101942)
2023-05-22 18:32:31 +00:00
fastrnns
[BE] [1/3] Rewrite super() calls in caffe2 and benchmarks (#94587)
2023-02-11 18:19:48 +00:00
framework_overhead_benchmark
[BE] [1/3] Rewrite super() calls in caffe2 and benchmarks (#94587)
2023-02-11 18:19:48 +00:00
functional_autograd_benchmark
[BE] [1/3] Rewrite super() calls in caffe2 and benchmarks (#94587)
2023-02-11 18:19:48 +00:00
fuser
Benchmarks for various fusers (#67622)
2021-11-04 18:57:17 -07:00
instruction_counts
[BE] Prefer dash over underscore in command-line options (#94505)
2023-02-09 20:16:49 +00:00
nested
Use tensor cores for NT bmm (#86856)
2022-11-02 21:51:40 +00:00
operator_benchmark
[BE] Enable C419 rule for any all shortcircuiting (#99890)
2023-04-25 15:02:13 +00:00
overrides_benchmark
Use classmethods for overrides (#64841)
2021-09-17 08:32:49 -07:00
profiler_benchmark
[BE] Prefer dash over underscore in command-line options (#94505)
2023-02-09 20:16:49 +00:00
record_function_benchmark
[BE] Prefer dash over underscore in command-line options (#94505)
2023-02-09 20:16:49 +00:00
serialization
[PyTorch] Add annotation_str benchmark (#96496)
2023-03-23 04:18:07 +00:00
sparse
[BE] Enable flake8-comprehension rule C417 (#97880)
2023-03-30 14:34:24 +00:00
static_runtime
Fix flaky StaticRuntime.Nonzero test (#94418)
2023-02-16 21:25:15 +00:00
tensorexpr
Fix usages of contextmanager without finally (#96170)
2023-03-08 20:59:27 +00:00
transformer
[transformer benchmark] relax tolerance in sdp.py (#101965)
2023-05-23 06:54:08 +00:00
compare-fastrnn-results.py
Benchmarks: add scripts for FastRNNs results comparison. (#44134)
2020-09-03 13:44:42 -07:00
compare.sh
Benchmarks: add scripts for FastRNNs results comparison. (#44134)
2020-09-03 13:44:42 -07:00
README.md
Add CSR (compressed sparse row) layout for sparse tensors (#50937)
2021-04-12 10:09:12 -07:00
upload_scribe.py
[BE] Prefer dash over underscore in command-line options (#94505)
2023-02-09 20:16:49 +00:00

README.md

PyTorch Benchmarks

This folder contains scripts that produce reproducible timings of various PyTorch features.

It also provides mechanisms to compare PyTorch with other frameworks.

Setup environment

Make sure you're on a machine with CUDA, torchvision, and pytorch installed. Install in the following order:

# Install torchvision. It comes with the pytorch stable release binary
conda install pytorch torchvision -c pytorch

# Install the latest pytorch master from source.
# It should supersede the installation from the release binary.
cd $PYTORCH_HOME
python setup.py build develop

# Check the pytorch installation version
python -c "import torch; print(torch.__version__)"

Benchmark List

Please refer to each subfolder to discover each benchmark suite