matmul performance benchmarks by aocsa · Pull Request #51647 · pytorch/pytorch

aocsa · 2021-02-03T17:18:30Z

Minor PR following up the previous PR about sparse benchmarking utils #48397

Fixes #44634: Performance benchmarks for matrix-matrix and matrix-vector ops (dense-sparse, sparse-sparse, and compare to dense-dense)

I ran all benchmarks on an 2xRTX8000 machine with AMD 2970WX 24-cores for DLMC/magnitude_pruning dataset with different sparsity levels.

forward tests (expand for details).

sparse@sparse

[------------------------------- cpu:matmul-forward -------------------------------]
                           |   0.5   |   0.7   |   0.8   |   0.9   |  0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense    |  108.1  |  100.5  |  101.3  |  108.4  |  98.4  |  187.4
      torch:sparse@sparse  |  659.1  |  368.8  |  156.5  |   53.3  |  26.8  |   14.9
      scipy:sparse@sparse  |  565.1  |  233.9  |  130.2  |   23.1  |  21.6  |   15.2

Times are in milliseconds (ms).

[----------------------------------- cuda:matmul-forward -----------------------------------]
                           |    0.5    |    0.7    |   0.8    |   0.9    |   0.95   |   0.98 
1 threads: ----------------------------------------------------------------------------------
      torch:dense@dense    |   2243.5  |   4392.5  |  4419.8  |  2272.3  |  4433.9  |  8920.1
      torch:sparse@sparse  |  21369.2  |  11877.6  |  7339.2  |  1787.2  |  1335.1  |   845.7

Times are in microseconds (us).

sparse@dense

[------------------------------- cpu:matmul-forward -------------------------------]
                          |   0.5   |   0.7   |   0.8   |   0.9   |   0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense   |  105.8  |  103.8  |  103.0  |  104.4  |  104.4  |  197.0
      torch:sparse@dense  |  119.9  |  102.4  |   84.0  |   19.7  |   16.8  |   11.6
      scipy:sparse@dense  |  906.5  |  799.6  |  697.8  |  182.2  |  165.5  |  135.4

Times are in milliseconds (ms).

[------------------------- cuda:matmul-forward --------------------------]
                          |  0.5  |  0.7  |  0.8  |  0.9  |  0.95  |  0.98
1 threads: ---------------------------------------------------------------
      torch:dense@dense   |  2.2  |  4.4  |  4.4  |  2.3  |  4.5   |  2.3 
      torch:sparse@dense  |  5.7  |  6.6  |  4.5  |  1.4  |  1.4   |  1.3 

Times are in milliseconds (ms).

sparse@vector

[----------------------------------- cpu:matmul-forward ----------------------------------]
                           |    0.5    |   0.7    |   0.8    |   0.9    |   0.95   |   0.98
1 threads: --------------------------------------------------------------------------------
      torch:dense@vector   |    510.6  |   505.8  |   759.6  |   782.1  |   682.4  |  764.6
      torch:sparse@vector  |  10122.8  |  6241.1  |  7935.6  |  2076.3  |  1049.5  |  826.3
      scipy:sparse@vector  |   1756.7  |  1033.9  |   678.2  |   343.5  |   168.5  |   65.4

Times are in microseconds (us).

[-------------------------------- cuda:matmul-forward --------------------------------]
                           |   0.5    |   0.7    |   0.8   |   0.9   |   0.95  |   0.98
1 threads: ----------------------------------------------------------------------------
      torch:dense@vector   |    36.1  |    21.5  |   21.6  |   21.5  |   21.6  |   21.5
      torch:sparse@vector  |  1099.2  |  1289.4  |  775.7  |  327.1  |  285.4  |  274.0

Times are in microseconds (us).

backward tests (expand for details).

sparse@sparse

[--------------------------------- cpu:matmul-backward ---------------------------------]
                           |   0.5    |   0.7    |   0.8    |   0.9    |   0.95  |   0.98
1 threads: ------------------------------------------------------------------------------
      torch:dense@dense    |   246.1  |   315.0  |   306.9  |   168.6  |  290.6  |  146.9
      torch:sparse@sparse  |  6417.5  |  4393.7  |  3012.7  |  1029.4  |  908.0  |  650.7

Times are in microseconds (us).

[----------------------------- cuda:matmul-backward -----------------------------]
                           |   0.5   |   0.7   |   0.8   |  0.9   |  0.95  |  0.98
1 threads: -----------------------------------------------------------------------
      torch:dense@dense    |    6.7  |   13.3  |   13.3  |   6.9  |  13.5  |   6.9
      torch:sparse@sparse  |  143.7  |  143.4  |  119.6  |  29.5  |  29.1  |  10.9

Times are in microseconds (us).

sparse@dense

 [------------------------------ cpu:matmul-backward -------------------------------]
                          |   0.5   |   0.7   |   0.8   |   0.9   |   0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense   |  185.9  |  304.8  |  305.8  |  169.9  |  308.7  |  168.4
      torch:sparse@dense  |  407.9  |  345.8  |  274.6  |  114.2  |  163.6  |  230.5

Times are in milliseconds (ms).

[--------------------------- cuda:matmul-backward --------------------------]
                          |  0.5   |  0.7   |  0.8   |  0.9  |  0.95  |  0.98
1 threads: ------------------------------------------------------------------
      torch:dense@dense   |   6.7  |  13.3  |  13.3  |  6.9  |  13.4  |   6.9
      torch:sparse@dense  |  16.7  |  19.0  |  15.1  |  6.3  |   8.2  |  12.7

Times are in milliseconds (ms).

Kindly review this PR. cc @mruberry, @ngimel

facebook-github-bot · 2021-02-03T17:18:43Z

💊 CI failures summary and remediations

As of commit b000c5a (more details on the Dr. CI page):

✅ None of the CI failures appear to be your fault 💚

2/2 broken upstream at merge base d4c877b on Mar 12 from 2:49am to 4:28pm

🚧 2 fixed upstream failures:

These were probably caused by upstream breakages that were already fixed.

Please rebase on the viable/strict branch (expand for instructions)

If your commit is older than viable/strict, run these commands:

git fetch https://github.com/pytorch/pytorch viable/strict
git rebase FETCH_HEAD

pytorch_linux_xenial_py3_clang7_onnx_ort_test2 on Mar 12 from 4:00am to 4:28pm (a51f130 - 5b648ef)
- 🔁 rerun
pytorch_linux_xenial_py3_clang7_onnx_ort_test1 on Mar 12 from 2:49am to 4:21pm (ee4ce8e - 5b648ef)
- 🔁 rerun

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions to the (internal) Dr. CI Users group.

robieta

At a high level it seems like spmm.py and spmv.py could be merged by having a central definition of tasks

_TASKS = {
    "mm": (
        ("torch", "torch.mm(dx, dy)"),
        ("torch.sparse", "torch.sparse.mm(x, y)"),
        ("scipy", "scipy_coo_matmul(sx, sy)"),
    ),
    "mv": (
        ("torch", "torch.matmul(dx, vector)"),
        ("torch.sparse", "torch.matmul(x, vector)"),
        ("scipy", "scipy_coo_matmul(sx, vector)"),
    ),
    "backward": (
        ("torch", "torch_backward(dx, dy)"),
        ("torch.sparse", "sparse_torch_backward(x, y)"),
    )
}

and some switching code at the start:

tasks = _TASKS[operation]
if device != "cpu":
    # SciPy is CPU only.
    tasks = tuple(t for t in tasks if t[0] != "scipy")

loader = (
    load_spmv_dataset if operation == "mv" else
    functools.partial(load_spmm_dataset, spmm_type=spmm_type)
)

There's also a bit of logic for globals, but it doesn't seem insurmountable. WDYT?

ngimel · 2021-02-12T00:31:02Z

Do you have results from running this benchmark?

aocsa · 2021-02-18T05:12:43Z

At a high level it seems like spmm.py and spmv.py could be merged by having a central definition of tasks

_TASKS = {
    "mm": (
        ("torch", "torch.mm(dx, dy)"),
        ("torch.sparse", "torch.sparse.mm(x, y)"),
        ("scipy", "scipy_coo_matmul(sx, sy)"),
    ),
    "mv": (
        ("torch", "torch.matmul(dx, vector)"),
        ("torch.sparse", "torch.matmul(x, vector)"),
        ("scipy", "scipy_coo_matmul(sx, vector)"),
    ),
    "backward": (
        ("torch", "torch_backward(dx, dy)"),
        ("torch.sparse", "sparse_torch_backward(x, y)"),
    )
}

and some switching code at the start:

tasks = _TASKS[operation]
if device != "cpu":
    # SciPy is CPU only.
    tasks = tuple(t for t in tasks if t[0] != "scipy")

loader = (
    load_spmv_dataset if operation == "mv" else
    functools.partial(load_spmm_dataset, spmm_type=spmm_type)
)

There's also a bit of logic for globals, but it doesn't seem insurmountable. WDYT?

Indeed, it was surmountable. Now spmm.py and spmv.py are merged :)

codecov · 2021-02-18T09:40:46Z

Codecov Report

Merging #51647 (5f4b689) into master (5563248) will increase coverage by 3.57%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #51647      +/-   ##
==========================================
+ Coverage   73.77%   77.35%   +3.57%     
==========================================
  Files        1879     1879              
  Lines      183186   183186              
==========================================
+ Hits       135147   141701    +6554     
+ Misses      48039    41485    -6554

aocsa · 2021-02-19T06:16:44Z

Do you have results from running this benchmark?

I addressed the feedback comments. Moreover, I ran all benchmarks on an 2xRTX8000 machine with AMD 2970WX 24-cores for DLMC/magnitude_pruning dataset with different sparsity levels.

forward tests (expand for details).

sparse@sparse

[------------------------------- cpu:matmul-forward -------------------------------]
                           |   0.5   |   0.7   |   0.8   |   0.9   |  0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense    |  108.1  |  100.5  |  101.3  |  108.4  |  98.4  |  187.4
      torch:sparse@sparse  |  659.1  |  368.8  |  156.5  |   53.3  |  26.8  |   14.9
      scipy:sparse@sparse  |  565.1  |  233.9  |  130.2  |   23.1  |  21.6  |   15.2
Times are in milliseconds (ms).

[----------------------------------- cuda:matmul-forward -----------------------------------]
                           |    0.5    |    0.7    |   0.8    |   0.9    |   0.95   |   0.98 
1 threads: ----------------------------------------------------------------------------------
      torch:dense@dense    |   2243.5  |   4392.5  |  4419.8  |  2272.3  |  4433.9  |  8920.1
      torch:sparse@sparse  |  21369.2  |  11877.6  |  7339.2  |  1787.2  |  1335.1  |   845.7

Times are in microseconds (us).

sparse@dense

[------------------------------- cpu:matmul-forward -------------------------------]
                          |   0.5   |   0.7   |   0.8   |   0.9   |   0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense   |  105.8  |  103.8  |  103.0  |  104.4  |  104.4  |  197.0
      torch:sparse@dense  |  119.9  |  102.4  |   84.0  |   19.7  |   16.8  |   11.6
      scipy:sparse@dense  |  906.5  |  799.6  |  697.8  |  182.2  |  165.5  |  135.4

Times are in milliseconds (ms).

[------------------------- cuda:matmul-forward --------------------------]
                          |  0.5  |  0.7  |  0.8  |  0.9  |  0.95  |  0.98
1 threads: ---------------------------------------------------------------
      torch:dense@dense   |  2.2  |  4.4  |  4.4  |  2.3  |  4.5   |  2.3 
      torch:sparse@dense  |  5.7  |  6.6  |  4.5  |  1.4  |  1.4   |  1.3 

Times are in milliseconds (ms).

sparse@vector

[----------------------------------- cpu:matmul-forward ----------------------------------]
                           |    0.5    |   0.7    |   0.8    |   0.9    |   0.95   |   0.98
1 threads: --------------------------------------------------------------------------------
      torch:dense@vector   |    510.6  |   505.8  |   759.6  |   782.1  |   682.4  |  764.6
      torch:sparse@vector  |  10122.8  |  6241.1  |  7935.6  |  2076.3  |  1049.5  |  826.3
      scipy:sparse@vector  |   1756.7  |  1033.9  |   678.2  |   343.5  |   168.5  |   65.4

Times are in microseconds (us).

[-------------------------------- cuda:matmul-forward --------------------------------]
                           |   0.5    |   0.7    |   0.8   |   0.9   |   0.95  |   0.98
1 threads: ----------------------------------------------------------------------------
      torch:dense@vector   |    36.1  |    21.5  |   21.6  |   21.5  |   21.6  |   21.5
      torch:sparse@vector  |  1099.2  |  1289.4  |  775.7  |  327.1  |  285.4  |  274.0

Times are in microseconds (us).

backward tests (expand for details).

sparse@sparse

[--------------------------------- cpu:matmul-backward ---------------------------------]
                           |   0.5    |   0.7    |   0.8    |   0.9    |   0.95  |   0.98
1 threads: ------------------------------------------------------------------------------
      torch:dense@dense    |   246.1  |   315.0  |   306.9  |   168.6  |  290.6  |  146.9
      torch:sparse@sparse  |  6417.5  |  4393.7  |  3012.7  |  1029.4  |  908.0  |  650.7

Times are in microseconds (us).

[----------------------------- cuda:matmul-backward -----------------------------]
                           |   0.5   |   0.7   |   0.8   |  0.9   |  0.95  |  0.98
1 threads: -----------------------------------------------------------------------
      torch:dense@dense    |    6.7  |   13.3  |   13.3  |   6.9  |  13.5  |   6.9
      torch:sparse@sparse  |  143.7  |  143.4  |  119.6  |  29.5  |  29.1  |  10.9

Times are in microseconds (us).

sparse@dense

 [------------------------------ cpu:matmul-backward -------------------------------]
                          |   0.5   |   0.7   |   0.8   |   0.9   |   0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense   |  185.9  |  304.8  |  305.8  |  169.9  |  308.7  |  168.4
      torch:sparse@dense  |  407.9  |  345.8  |  274.6  |  114.2  |  163.6  |  230.5

Times are in milliseconds (ms).

[--------------------------- cuda:matmul-backward --------------------------]
                          |  0.5   |  0.7   |  0.8   |  0.9  |  0.95  |  0.98
1 threads: ------------------------------------------------------------------
      torch:dense@dense   |   6.7  |  13.3  |  13.3  |  6.9  |  13.4  |   6.9
      torch:sparse@dense  |  16.7  |  19.0  |  15.1  |  6.3  |   8.2  |  12.7

Times are in milliseconds (ms).

cc @robieta, @ngimel, @mruberry

mruberry

Cool! Thanks @aocsa. I wonder if we can close the gap with SciPy in the scenarios we're slower on CPU?

@robieta would you like to take another look before this is merged?

aocsa · 2021-02-26T02:04:08Z

Cool! Thanks @aocsa. I wonder if we can close the gap with SciPy in the scenarios we're slower on CPU?

@robieta would you like to take another look before this is merged?

Thanks @mruberry, the current implementation of sparse-sparse matmul has almost the same performance as SciPy. I forgot to considering the overhead of converting the csr output of the x.dot(y) scipy operation to coo which is part of the process of torch.sparse.mm operation. I've updated the results with this change.

[------------------------------- cpu:matmul-forward -------------------------------]
                           |   0.5   |   0.7   |   0.8   |   0.9   |  0.95  |   0.98
1 threads: -------------------------------------------------------------------------
      torch:dense@dense    |  108.1  |  100.5  |  101.3  |  108.4  |  98.4  |  187.4
      torch:sparse@sparse  |  659.1  |  368.8  |  156.5  |   53.3  |  26.8  |   14.9
      scipy:sparse@sparse  |  565.1  |  233.9  |  130.2  |   23.1  |  21.6  |   15.2
Times are in milliseconds (ms).

I moved main comment and benchmark results into the PR description.

mruberry · 2021-03-01T16:18:57Z

Let's give @robieta a chance to review.

robieta

I had a single nitpick, but overall looks great. Thanks!

aocsa · 2021-03-10T22:54:32Z

I rebased this PR and I think this PR is ready to be merged.
Thanks cc @mruberry

facebook-github-bot

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

mruberry · 2021-03-12T19:20:59Z

I'm trying to land this but the internal tools are struggling to reconcile these whitespace changes. Would you revert these changes and/or rebase this PR, @aocsa?

Sure, it's done cc @mruberry

first commit bench matmul

docs updated

update update update 2

minor fix

update default timer_min_run_time update update update repeats

quick-checks

facebook-github-bot

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

facebook-github-bot · 2021-03-14T08:27:36Z

@mruberry merged this pull request in 39f50f4.

Summary: Minor PR following up the previous PR about sparse benchmarking utils pytorch#48397 Fixes pytorch#44634: Performance benchmarks for matrix-matrix and matrix-vector ops (dense-sparse, sparse-sparse, and compare to dense-dense) I ran all benchmarks on an 2xRTX8000 machine with AMD 2970WX 24-cores for `DLMC/magnitude_pruning` dataset with different sparsity levels. --- <details><summary> forward tests (expand for details). </summary> - `sparse@sparse` ``` [------------------------------- cpu:matmul-forward -------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ------------------------------------------------------------------------- torch:dense@dense | 108.1 | 100.5 | 101.3 | 108.4 | 98.4 | 187.4 torch:sparse@sparse | 659.1 | 368.8 | 156.5 | 53.3 | 26.8 | 14.9 scipy:sparse@sparse | 565.1 | 233.9 | 130.2 | 23.1 | 21.6 | 15.2 Times are in milliseconds (ms). [----------------------------------- cuda:matmul-forward -----------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ---------------------------------------------------------------------------------- torch:dense@dense | 2243.5 | 4392.5 | 4419.8 | 2272.3 | 4433.9 | 8920.1 torch:sparse@sparse | 21369.2 | 11877.6 | 7339.2 | 1787.2 | 1335.1 | 845.7 Times are in microseconds (us). ``` - `sparse@dense` ``` [------------------------------- cpu:matmul-forward -------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ------------------------------------------------------------------------- torch:dense@dense | 105.8 | 103.8 | 103.0 | 104.4 | 104.4 | 197.0 torch:sparse@dense | 119.9 | 102.4 | 84.0 | 19.7 | 16.8 | 11.6 scipy:sparse@dense | 906.5 | 799.6 | 697.8 | 182.2 | 165.5 | 135.4 Times are in milliseconds (ms). [------------------------- cuda:matmul-forward --------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: --------------------------------------------------------------- torch:dense@dense | 2.2 | 4.4 | 4.4 | 2.3 | 4.5 | 2.3 torch:sparse@dense | 5.7 | 6.6 | 4.5 | 1.4 | 1.4 | 1.3 Times are in milliseconds (ms). ``` - `sparse@vector` ``` [----------------------------------- cpu:matmul-forward ----------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: -------------------------------------------------------------------------------- torch:dense@vector | 510.6 | 505.8 | 759.6 | 782.1 | 682.4 | 764.6 torch:sparse@vector | 10122.8 | 6241.1 | 7935.6 | 2076.3 | 1049.5 | 826.3 scipy:sparse@vector | 1756.7 | 1033.9 | 678.2 | 343.5 | 168.5 | 65.4 Times are in microseconds (us). [-------------------------------- cuda:matmul-forward --------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ---------------------------------------------------------------------------- torch:dense@vector | 36.1 | 21.5 | 21.6 | 21.5 | 21.6 | 21.5 torch:sparse@vector | 1099.2 | 1289.4 | 775.7 | 327.1 | 285.4 | 274.0 Times are in microseconds (us). ``` </details> --- <details><summary> backward tests (expand for details). </summary> - `sparse@sparse` ``` [--------------------------------- cpu:matmul-backward ---------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ------------------------------------------------------------------------------ torch:dense@dense | 246.1 | 315.0 | 306.9 | 168.6 | 290.6 | 146.9 torch:sparse@sparse | 6417.5 | 4393.7 | 3012.7 | 1029.4 | 908.0 | 650.7 Times are in microseconds (us). [----------------------------- cuda:matmul-backward -----------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ----------------------------------------------------------------------- torch:dense@dense | 6.7 | 13.3 | 13.3 | 6.9 | 13.5 | 6.9 torch:sparse@sparse | 143.7 | 143.4 | 119.6 | 29.5 | 29.1 | 10.9 Times are in microseconds (us). ``` - `sparse@dense` ``` [------------------------------ cpu:matmul-backward -------------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ------------------------------------------------------------------------- torch:dense@dense | 185.9 | 304.8 | 305.8 | 169.9 | 308.7 | 168.4 torch:sparse@dense | 407.9 | 345.8 | 274.6 | 114.2 | 163.6 | 230.5 Times are in milliseconds (ms). [--------------------------- cuda:matmul-backward --------------------------] | 0.5 | 0.7 | 0.8 | 0.9 | 0.95 | 0.98 1 threads: ------------------------------------------------------------------ torch:dense@dense | 6.7 | 13.3 | 13.3 | 6.9 | 13.4 | 6.9 torch:sparse@dense | 16.7 | 19.0 | 15.1 | 6.3 | 8.2 | 12.7 Times are in milliseconds (ms). ``` </details> Kindly review this PR. cc mruberry, ngimel Pull Request resolved: pytorch#51647 Reviewed By: albanD Differential Revision: D27007809 Pulled By: mruberry fbshipit-source-id: 8c1922cb3280027ca5e3eef31bfa20500c548cfd

facebook-github-bot added the cla signed label Feb 3, 2021

pytorchbot added the open source label Feb 3, 2021

aocsa force-pushed the aocsa/sparse_mm_bench branch from a7aa982 to fb37a87 Compare February 8, 2021 16:04

aocsa marked this pull request as ready for review February 8, 2021 16:05

aocsa force-pushed the aocsa/sparse_mm_bench branch from 985c08d to 3a2f8ec Compare February 9, 2021 17:58

VitalyFedyunin requested review from ngimel and robieta February 10, 2021 21:57

VitalyFedyunin added the triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module label Feb 10, 2021

robieta requested changes Feb 11, 2021

View reviewed changes

Comment thread benchmarks/sparse/dlmc/spmm.py Outdated

Comment thread benchmarks/sparse/dlmc/spmm.py Outdated

Comment thread benchmarks/sparse/dlmc/spmm.py Outdated

Comment thread benchmarks/sparse/dlmc/spmv.py Outdated

mruberry reviewed Feb 11, 2021

View reviewed changes

Comment thread benchmarks/sparse/dlmc/README.md Outdated

mruberry reviewed Feb 12, 2021

View reviewed changes

Comment thread benchmarks/sparse/dlmc/utils.py Outdated

mruberry reviewed Feb 12, 2021

View reviewed changes

Comment thread benchmarks/sparse/dlmc/utils.py Outdated

ngimel reviewed Feb 12, 2021

View reviewed changes

aocsa self-assigned this Feb 17, 2021

aocsa force-pushed the aocsa/sparse_mm_bench branch from 5125648 to d24d081 Compare February 18, 2021 05:22

aocsa force-pushed the aocsa/sparse_mm_bench branch from dfe3167 to 4c3ae8e Compare February 19, 2021 17:18

mruberry self-requested a review February 26, 2021 01:16

mruberry approved these changes Feb 26, 2021

View reviewed changes

aocsa force-pushed the aocsa/sparse_mm_bench branch 2 times, most recently from ab4359b to d2868ce Compare February 26, 2021 02:35

robieta approved these changes Mar 2, 2021

View reviewed changes

Comment thread benchmarks/sparse/dlmc/matmul_bench.py Outdated

aocsa force-pushed the aocsa/sparse_mm_bench branch from d2868ce to fc2554f Compare March 2, 2021 22:19

aocsa mentioned this pull request Mar 2, 2021

torch.sparse improvements - tracking issue #44634

Open

26 tasks

aocsa force-pushed the aocsa/sparse_mm_bench branch from fc2554f to 5f4b689 Compare March 10, 2021 22:52

aocsa force-pushed the aocsa/sparse_mm_bench branch from 24f8665 to 870de27 Compare March 11, 2021 14:15

facebook-github-bot reviewed Mar 12, 2021

View reviewed changes

mruberry reviewed Mar 12, 2021

View reviewed changes

aocsa added 7 commits March 12, 2021 15:22

first commit

9fdbb5e

first commit bench matmul

add docs

ac0085c

docs updated

feedback addressed

afd1e14

update update update 2

minor updates for performance tests

82ef38c

minor fix

update repeats and scipy_mm

142ad77

update default timer_min_run_time update update update repeats

minor update

7dce9d3

remove trailing spaces

b000c5a

quick-checks

aocsa force-pushed the aocsa/sparse_mm_bench branch from 870de27 to b000c5a Compare March 12, 2021 20:25

facebook-github-bot reviewed Mar 12, 2021

View reviewed changes

facebook-github-bot closed this in 39f50f4 Mar 14, 2021

facebook-github-bot added the Merged label Mar 14, 2021

Conversation

aocsa commented Feb 3, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

facebook-github-bot commented Feb 3, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🚧 2 fixed upstream failures:

Uh oh!

robieta left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

ngimel commented Feb 12, 2021

Uh oh!

aocsa commented Feb 18, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov Bot commented Feb 18, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

aocsa commented Feb 19, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mruberry left a comment

Choose a reason for hiding this comment

Uh oh!

aocsa commented Feb 26, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mruberry commented Mar 1, 2021

Uh oh!

robieta left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

aocsa commented Mar 10, 2021

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

mruberry Mar 12, 2021

Choose a reason for hiding this comment

Uh oh!

aocsa Mar 12, 2021

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Mar 14, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

aocsa commented Feb 3, 2021 •

edited

Loading

facebook-github-bot commented Feb 3, 2021 •

edited

Loading

aocsa commented Feb 18, 2021 •

edited

Loading

codecov Bot commented Feb 18, 2021 •

edited

Loading

aocsa commented Feb 19, 2021 •

edited

Loading

aocsa commented Feb 26, 2021 •

edited

Loading