Skip to content

[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators#6054

Merged
yzhliu merged 19 commits into
apache:masterfrom
sandyhu533:relay_mx_np
Aug 21, 2020
Merged

[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators#6054
yzhliu merged 19 commits into
apache:masterfrom
sandyhu533:relay_mx_np

Conversation

@sandyhu533

@sandyhu533 sandyhu533 commented Jul 14, 2020 •
edited
Loading

Copy link
Copy Markdown
Contributor

add support for some MXNET numpy operators, related issue: dmlc/gluon-nlp#1244
For now, these operators have been implemented

_npi_transpose
_npi_pad
_npi_multiply_scalar
_npi_true_divide_scalar
_npi_add
_npi_concatenate
_npi_multiply
_np_copy
_npi_tanh
_npi_power_scalar
_npi_less
_npi_add_scalar
_npx_reshape
_split_v2
_npi_where_rscalar

sxjscience
sxjscience reviewed Jul 15, 2020
View reviewed changes
Comment thread tests/python/frontend/mxnet/test_forward.py Outdated

def verify(data_shape, scalar):
cond_np = np.random.uniform(size=data_shape).astype("bool")
data_np = np.random.uniform(size=data_shape).astype("float32")

@sxjscience sxjscience Jul 15, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We may need to test for multiple dtypes. Especially for cond_np, we can have bool or float32.

yzhliu
yzhliu reviewed Jul 17, 2020
View reviewed changes
Comment thread tests/python/frontend/mxnet/test_forward.py Outdated
verify(3, 10, 6, 8)


@yzhliu yzhliu Jul 17, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

remove the change

Comment thread tests/python/frontend/mxnet/test_forward.py Outdated
pytest.skip("mx.sym.np.pad hasn't been publish yet")

def verify(data_shape, mode, pad_width, constant_value=0.0):
for dtype in dtype_list:

@yzhliu yzhliu Jul 17, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

would it be easier to use @pytest.mark.parametrize?

Comment thread tests/python/frontend/mxnet/test_forward.py Outdated
test_forward_npi_binary_scalar()
test_forward_npi_tanh()
test_forward_npi_where_rscalar()
test_forward_split_v2()

@yzhliu yzhliu Jul 17, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

maybe we can just do pytest.main([__file__])

Comment thread tests/python/frontend/mxnet/test_forward.py Outdated

def test_forward_npi_pad():
if not hasattr(mx.sym.np, 'pad'):
pytest.skip("mx.sym.np.pad hasn't been publish yet")

@yzhliu yzhliu Jul 17, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

do we need to update CI's mxnet version?

@sandyhu533 sandyhu533 Jul 20, 2020

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm afraid we couldn't update CI's mxnet version now. The latest version of mxnet has made some changes that the test functions in tvm didn't updated with them. So the CI would fail if we update mxnet version now.
This is the error message when I try to run pytest for tests/python/frontend/mxnet/test_forward.py.
E AttributeError: module 'mxnet' has no attribute 'mod'

@szha szha Jul 20, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We will need to find a good way to support both mx 1.x and 2.0.

For marking the test, maybe use @pytest.mark.skipif ?

@leezu leezu Jul 20, 2020

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why not use two separate pipelines to test with the different mxnet versions?

@yzhliu yzhliu Jul 21, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@tqchen what do you think?

@tqchen tqchen Jul 21, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

skip ci for now sounds good

@tqchen

tqchen commented Jul 21, 2020

Copy link
Copy Markdown
Member

cc @antinucleon @junrushao1994

yzhliu
yzhliu reviewed Jul 28, 2020
View reviewed changes
Comment thread tests/python/frontend/mxnet/test_forward.py Outdated
mx_sym = mx.sym.np.pad(data.as_np_ndarray(), mode=mode, pad_width=pad_width)
mod, _ = relay.frontend.from_mxnet(mx_sym, {"data": data_shape}, dtype=dtype)
for target, ctx in ctx_list():
for kind in ["debug"]:

@yzhliu yzhliu Jul 28, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why is this one using debug only?

@yzhliu yzhliu Jul 28, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

maybe we can also use @pytest.mark.parametrize for ctx_list and kind.

Comment thread tests/python/frontend/mxnet/test_forward.py Outdated


if __name__ == '__main__':
test_forward_mlp()

@yzhliu yzhliu Jul 28, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

remove all the test_xxx function calls and add pytest.main([__file__])

@yzhliu

yzhliu commented Jul 30, 2020

Copy link
Copy Markdown
Member

Please check the ci failure, otherwise good to me.

@yzhliu yzhliu merged commit 4c728d5 into apache:master Aug 21, 2020
@yzhliu

yzhliu commented Aug 21, 2020

Copy link
Copy Markdown
Member

Thanks @sandyhu533 @sxjscience

@sxjscience

sxjscience commented Aug 21, 2020

Copy link
Copy Markdown
Member

Thanks! Cannot wait to try it in GluonNLP 👍

trevor-m pushed a commit to trevor-m/tvm that referenced this pull request Aug 26, 2020
@sandyhu533
[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators (apach…
f2c3f41 
…e#6054)

* [RELAY][MXNET][FRONTEND] add supports for OPs in numpy from mxnet

* Update test_forward.py

* Update mxnet.py

* Update mxnet.py

* Update test_forward.py

* update and bugfix

* test for multiple dtypes

* Update test_forward.py

* add data type and optimize coding style

* replace pytest.skip with @pytest.mark.skipif

* Update test_forward.py

* update pytest style

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-169.ap-northeast-1.compute.internal>
trevor-m pushed a commit to trevor-m/tvm that referenced this pull request Aug 26, 2020
@sandyhu533
[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators (apach…
89a5370 
…e#6054)

* [RELAY][MXNET][FRONTEND] add supports for OPs in numpy from mxnet

* Update test_forward.py

* Update mxnet.py

* Update mxnet.py

* Update test_forward.py

* update and bugfix

* test for multiple dtypes

* Update test_forward.py

* add data type and optimize coding style

* replace pytest.skip with @pytest.mark.skipif

* Update test_forward.py

* update pytest style

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-169.ap-northeast-1.compute.internal>
trevor-m pushed a commit to trevor-m/tvm that referenced this pull request Aug 26, 2020
@sandyhu533
[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators (apach…
a3e1196 
…e#6054)

* [RELAY][MXNET][FRONTEND] add supports for OPs in numpy from mxnet

* Update test_forward.py

* Update mxnet.py

* Update mxnet.py

* Update test_forward.py

* update and bugfix

* test for multiple dtypes

* Update test_forward.py

* add data type and optimize coding style

* replace pytest.skip with @pytest.mark.skipif

* Update test_forward.py

* update pytest style

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-169.ap-northeast-1.compute.internal>
trevor-m pushed a commit to trevor-m/tvm that referenced this pull request Sep 2, 2020
@sandyhu533
[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators (apach…
1a7d432 
…e#6054)

* [RELAY][MXNET][FRONTEND] add supports for OPs in numpy from mxnet

* Update test_forward.py

* Update mxnet.py

* Update mxnet.py

* Update test_forward.py

* update and bugfix

* test for multiple dtypes

* Update test_forward.py

* add data type and optimize coding style

* replace pytest.skip with @pytest.mark.skipif

* Update test_forward.py

* update pytest style

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-169.ap-northeast-1.compute.internal>
trevor-m pushed a commit to neo-ai/tvm that referenced this pull request Sep 3, 2020
@sandyhu533 @trevor-m
[RELAY][MXNET][FRONTEND] add support for MXNET numpy operators (apach…
c7084b4 
…e#6054)

* [RELAY][MXNET][FRONTEND] add supports for OPs in numpy from mxnet

* Update test_forward.py

* Update mxnet.py

* Update mxnet.py

* Update test_forward.py

* update and bugfix

* test for multiple dtypes

* Update test_forward.py

* add data type and optimize coding style

* replace pytest.skip with @pytest.mark.skipif

* Update test_forward.py

* update pytest style

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

* Update test_forward.py

Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-169.ap-northeast-1.compute.internal>
@ZihengJiang ZihengJiang mentioned this pull request Sep 25, 2020
Closed
@sxjscience sxjscience mentioned this pull request Oct 20, 2020
Closed
@sxjscience sxjscience mentioned this pull request Dec 31, 2020
Closed
3 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@tqchen tqchen tqchen left review comments

@yzhliu yzhliu yzhliu approved these changes

+3 more reviewers

@leezu leezu leezu left review comments

@szha szha szha left review comments

@sxjscience sxjscience sxjscience left review comments

Reviewers whose approvals may not affect merge requirements

Assignees

@yzhliu yzhliu

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

6 participants

@sandyhu533 @tqchen @yzhliu @sxjscience @leezu @szha