[GitHub] [incubator-mxnet] vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack
vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338408083 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) + +Stack 1-D arrays as columns into a 2-D array. + +Take a sequence of 1-D arrays and stack them as columns +to make a single 2-D array. 2-D arrays are stacked as-is, +just like with `hstack`. 1-D arrays are turned into 2-D columns +first. + +Parameters +-- +tup : sequence of 1-D or 2-D arrays. +Arrays to stack. All of them must have the same first dimension. + +Returns +--- +stacked : 2-D array +The array formed by stacking the given arrays. + +See Also + +stack, hstack, vstack, concatenate + +Examples + +>>> a = np.array((1,2,3)) Review comment: Changed to float output with trailing decimal point. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack
vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338407807 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) + +Stack 1-D arrays as columns into a 2-D array. + +Take a sequence of 1-D arrays and stack them as columns +to make a single 2-D array. 2-D arrays are stacked as-is, +just like with `hstack`. 1-D arrays are turned into 2-D columns +first. + +Parameters +-- +tup : sequence of 1-D or 2-D arrays. +Arrays to stack. All of them must have the same first dimension. + +Returns +--- +stacked : 2-D array +The array formed by stacking the given arrays. + +See Also + +stack, hstack, vstack, concatenate + +Examples + +>>> a = np.array((1,2,3)) +>>> b = np.array((2,3,4)) +>>> np.column_stack((a,b)) +array([[1, 2], Review comment: I didn't quite get it. Could you elaborate? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack
vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338407656 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) Review comment: Removed. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. aaronmarkham pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new bf95876 Bump the publish timestamp. bf95876 is described below commit bf95876ac94e527ea09d618a0f478e4003f0a253 Author: mxnet-ci AuthorDate: Thu Oct 24 06:41:55 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..2e98d1a --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Thu Oct 24 06:41:55 UTC 2019
[GitHub] [incubator-mxnet] hzfan commented on a change in pull request #16589: Fix index overflow bug in einsum
hzfan commented on a change in pull request #16589: Fix index overflow bug in einsum URL: https://github.com/apache/incubator-mxnet/pull/16589#discussion_r338403041 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3377,16 +3377,20 @@ def dbg(name, data): _np.dot(args[0].T, _np.dot(_np.ones((2, 2)), args[2].T)), _np.dot(_np.dot(args[0], args[1]).T, _np.ones((2, 2), # broadcast bug -(('ij, ij -> i'), [(1, 4), (2, 4)], lambda *args: (_np.sum(args[1], axis=0)[None, :], - _np.tile(args[0], [2, 1]))), +('ij, ij -> i', [(1, 4), (2, 4)], lambda *args: (_np.sum(args[1], axis=0)[None, :], + _np.tile(args[0], [2, 1]))), +# issue #16576 +# commented due to long running time +# ('abiz,abjz->abij', [(64, 8, 128, 512), (64, 8, 128, 512)], lambda *args: (_np.matmul(_np.ones((64, 8, 128, 128)), args[1]), +# _np.matmul(_np.ones((64, 8, 128, 128)), args[0]))), ] -dtypes = ['int32', 'float16', 'float32', 'float64'] +dtypes = ['int32', 'float32', 'float64'] for hybridize in [False, True]: for dtype in dtypes: for config in configs: for optimize in [False, True]: -rtol = 1e-0 if dtype == 'float16' else 1e-3 -atol = 1e-1 if dtype == 'float16' else 1e-5 +rtol = 1e-0 if dtype == 'float16' else 1e-1 +atol = 1e-1 if dtype == 'float16' else 1e-1 Review comment: @sxjscience Falling back to Batched GEMM is possible, but quite a lot work needs to be done. Now both official numpy and our implementation uses `tensordot` as a fallback, but `tensordot` cannot handle batch dot. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ZhennanQin commented on issue #16602: [MKLDNN]Fix reorder2default
ZhennanQin commented on issue #16602: [MKLDNN]Fix reorder2default URL: https://github.com/apache/incubator-mxnet/pull/16602#issuecomment-545768824 > How about pushing this change directly to mkldnn-v1.0 branch? Let's run master CI pass, then I will duplicate this to 1.0 branch. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (0742a9b -> ca5a2a0)
This is an automated email from the ASF dual-hosted git repository. haoj pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 0742a9b Large Vector tests for DGL Ops Part 2 (#16497) add ca5a2a0 [Numpy] Loading numpy-incompatible NDArray in numpy-compatible mode (#16597) No new revisions were added by this update. Summary of changes: include/mxnet/c_api.h | 2 +- include/mxnet/imperative.h | 10 ++ .../native/src/main/native/org_apache_mxnet_native_c_api.cc| 4 ++-- src/c_api/c_api_ndarray.cc | 2 +- src/ndarray/ndarray.cc | 3 ++- 5 files changed, 12 insertions(+), 9 deletions(-)
[GitHub] [incubator-mxnet] ZhennanQin opened a new pull request #16607: Surpress subgraph log in CI
ZhennanQin opened a new pull request #16607: Surpress subgraph log in CI URL: https://github.com/apache/incubator-mxnet/pull/16607 `src/executor/graph_executor.cc:2014: Subgraph backend MKLDNN is activated.` is flooding CI log. Let's surpress it in CI. @pengzhao-intel @TaoLv ## Description ## (Brief description on what this PR is about) ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (0742a9b -> ca5a2a0)
This is an automated email from the ASF dual-hosted git repository. haoj pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 0742a9b Large Vector tests for DGL Ops Part 2 (#16497) add ca5a2a0 [Numpy] Loading numpy-incompatible NDArray in numpy-compatible mode (#16597) No new revisions were added by this update. Summary of changes: include/mxnet/c_api.h | 2 +- include/mxnet/imperative.h | 10 ++ .../native/src/main/native/org_apache_mxnet_native_c_api.cc| 4 ++-- src/c_api/c_api_ndarray.cc | 2 +- src/ndarray/ndarray.cc | 3 ++- 5 files changed, 12 insertions(+), 9 deletions(-)
[GitHub] [incubator-mxnet] ZhennanQin opened a new pull request #16606: Fix dequantize memory corruption
ZhennanQin opened a new pull request #16606: Fix dequantize memory corruption URL: https://github.com/apache/incubator-mxnet/pull/16606 dequantize only has 1 output. The access of (*out_attrs)[1] and (*out_attrs)[2] is illegal. @pengzhao-intel @TaoLv ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 merged pull request #16597: [Numpy] Loading numpy-incompatible NDArray in numpy-compatible mode
haojin2 merged pull request #16597: [Numpy] Loading numpy-incompatible NDArray in numpy-compatible mode URL: https://github.com/apache/incubator-mxnet/pull/16597 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338397376 ## File path: python/mxnet/numpy_dispatch_protocol.py ## @@ -119,6 +119,7 @@ def _run_with_array_ufunc_proto(*args, **kwargs): 'var', 'vdot', 'vstack', +# 'column_stack', Review comment: Please keep this and add some test cases not necessarily from NumPy in `test_numpy_interoperability.py`. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338397125 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) + +Stack 1-D arrays as columns into a 2-D array. + +Take a sequence of 1-D arrays and stack them as columns +to make a single 2-D array. 2-D arrays are stacked as-is, +just like with `hstack`. 1-D arrays are turned into 2-D columns +first. + +Parameters +-- +tup : sequence of 1-D or 2-D arrays. +Arrays to stack. All of them must have the same first dimension. + +Returns +--- +stacked : 2-D array +The array formed by stacking the given arrays. + +See Also + +stack, hstack, vstack, concatenate + +Examples + +>>> a = np.array((1,2,3)) Review comment: Did you run this using MXNet? I believe MXNet would give a slightly different output than the following. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338396879 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) + +Stack 1-D arrays as columns into a 2-D array. + +Take a sequence of 1-D arrays and stack them as columns +to make a single 2-D array. 2-D arrays are stacked as-is, +just like with `hstack`. 1-D arrays are turned into 2-D columns +first. + +Parameters +-- +tup : sequence of 1-D or 2-D arrays. +Arrays to stack. All of them must have the same first dimension. + +Returns +--- +stacked : 2-D array +The array formed by stacking the given arrays. + +See Also + +stack, hstack, vstack, concatenate + +Examples + +>>> a = np.array((1,2,3)) +>>> b = np.array((2,3,4)) +>>> np.column_stack((a,b)) +array([[1, 2], Review comment: Notice the alignment ```python array([[1, 2], [2, 3], [3, 4]]) ``` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338396521 ## File path: python/mxnet/numpy/multiarray.py ## @@ -6419,3 +6419,39 @@ def einsum(*operands, **kwargs): ... np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=True) """ return _mx_nd_np.einsum(*operands, **kwargs) + +@set_module('mxnet.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) Review comment: No need to define the signature in docstring. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] TaoLv commented on issue #16602: [MKLDNN]Fix reorder2default
TaoLv commented on issue #16602: [MKLDNN]Fix reorder2default URL: https://github.com/apache/incubator-mxnet/pull/16602#issuecomment-545761914 How about pushing this change directly to mkldnn-v1.0 branch? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] aaronmarkham commented on issue #16598: second round of fixing broken links in multiple files
aaronmarkham commented on issue #16598: second round of fixing broken links in multiple files URL: https://github.com/apache/incubator-mxnet/pull/16598#issuecomment-545759333 Restarted the gpu tests for a second time. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] aaronmarkham commented on issue #16524: CI tests timing out
aaronmarkham commented on issue #16524: CI tests timing out URL: https://github.com/apache/incubator-mxnet/issues/16524#issuecomment-545759173 Timeout on GPU: CMake TVM_OP OFF http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-gpu/detail/PR-16598/2/pipeline/54 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] vexilligera commented on issue #16594: [Numpy] implement np.column_stack
vexilligera commented on issue #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#issuecomment-545758159 @haojin2 @reminisce Hi, all fixed. Please have a look. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei opened a new pull request #16605: Fix rnn dropout
roywei opened a new pull request #16605: Fix rnn dropout URL: https://github.com/apache/incubator-mxnet/pull/16605 fix https://github.com/apache/incubator-mxnet/issues/16604 depend on https://github.com/apache/incubator-mxnet/pull/16532, created a separate PR in case my fix on rnn has problem. I don't want to impact the merge of https://github.com/apache/incubator-mxnet/pull/16532 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #16604: RNN op with dropout cannot use fixed seed on CPU
roywei commented on issue #16604: RNN op with dropout cannot use fixed seed on CPU URL: https://github.com/apache/incubator-mxnet/issues/16604#issuecomment-545747855 This implementation on lstm, gru and rnn should respect mxnet seed: https://github.com/apache/incubator-mxnet/blob/master/src/operator/rnn_impl.h#L159 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338382229 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), +((0), (0), (0)), +((2, 2), (2, 1), (2, 3)), +((4, 3), (4, 4), (4, 1)), +((2, 2, 2), (2, 4, 2), (2, 2, 2)), +((0, 1, 1), (0, 1, 1), (0, 1, 1)), +((2, 1), (2, 2), (2, 2)) +] +types = ['float16', 'float32', 'float64', 'int8', 'int32', 'int64'] +for config in configs: Review comment: Please use `itertools.product` to reduce the nested loops to one. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on a change in pull request #16532: fix dropout gpu seed
roywei commented on a change in pull request #16532: fix dropout gpu seed URL: https://github.com/apache/incubator-mxnet/pull/16532#discussion_r338382491 ## File path: tests/nightly/test_dropout.py ## @@ -0,0 +1,49 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import sys +import os +curr_path = os.path.dirname(os.path.abspath(os.path.expanduser(__file__))) +sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'python', 'unittest')) +from common import with_seed +import unittest + +import mxnet as mx +import numpy as np +from mxnet.test_utils import assert_almost_equal +from nose.tools import assert_raises + +@with_seed() Review comment: Added, but cpu does not work, gpu works after my PR. I think it's unrelated to my change, created a separate issue here: https://github.com/apache/incubator-mxnet/issues/16604 Workaround is manual append a dropout layer after rnn on cpu. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack
reminisce commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338382025 ## File path: python/mxnet/ndarray/numpy/_op.py ## @@ -4761,3 +4760,39 @@ def einsum(*operands, **kwargs): subscripts = operands[0] operands = operands[1:] return _npi.einsum(*operands, subscripts=subscripts, out=out, optimize=int(optimize_arg)) + +@set_module('mxnet.ndarray.numpy') +def column_stack(tup): +""" column_stack(*args, **kwargs) + +Stack 1-D arrays as columns into a 2-D array. + +Take a sequence of 1-D arrays and stack them as columns +to make a single 2-D array. 2-D arrays are stacked as-is, +just like with `hstack`. 1-D arrays are turned into 2-D columns +first. + +Parameters +-- +tup : sequence of 1-D or 2-D arrays. +Arrays to stack. All of them must have the same first dimension. + +Returns +--- +stacked : 2-D array +The array formed by stacking the given arrays. + +See Also + +stack, hstack, vstack, concatenate + +Examples + +>>> a = np.array((1,2,3)) +>>> b = np.array((2,3,4)) +>>> np.column_stack((a,b)) +array([[1, 2], +[2, 3], Review comment: alignment This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei opened a new issue #16604: RNN op with dropout cannot use fixed seed on CPU
roywei opened a new issue #16604: RNN op with dropout cannot use fixed seed on CPU URL: https://github.com/apache/incubator-mxnet/issues/16604 As follow up on https://github.com/apache/incubator-mxnet/pull/16532 I found out rnn operator with dropout result cannot be fixed with fixed seed on CPU. GPU (cudnn) is fine after PR https://github.com/apache/incubator-mxnet/pull/16532 ``` @with_seed() def test_rnn_with_seed(): info = np.iinfo(np.int32) seed = np.random.randint(info.min, info.max) _test_rnn(seed, mx.cpu()) _test_rnn(seed, mx.gpu()) def _test_rnn(seed, ctx): data = mx.nd.ones((5, 3, 10), ctx=ctx) rnn = mx.gluon.rnn.RNN(100, 3, dropout=0.5) rnn.initialize(ctx=ctx) mx.random.seed(seed) with mx.autograd.record(): result1 = rnn(data) mx.random.seed(seed) with mx.autograd.record(): result2 = rnn(data) # dropout on gpu should return same result with fixed seed assert_almost_equal(result1.asnumpy(), result2.asnumpy()) ``` Current workaround is DO NOT use dropout in rnn and manually append a dropout layer after rnn, the following works: ``` def _test_rnn(seed, ctx): data = mx.nd.ones((5, 3, 10), ctx=ctx) rnn = mx.gluon.rnn.RNN(100, 3, dropout=0.) rnn.initialize(ctx=ctx) dropout = mx.gluon.nn.Dropout(0.5) with mx.autograd.record(): result1 = rnn(data) mx.random.seed(seed) o1 = dropout(result1) with mx.autograd.record(): result2 = rnn(data) mx.random.seed(seed) o2 = dropout(result2) # dropout on gpu should return same result with fixed seed assert_almost_equal(o1.asnumpy(), o2.asnumpy()) ``` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on issue #16431: [RFC] MXNet Multithreaded Inference Interface
anirudh2290 commented on issue #16431: [RFC] MXNet Multithreaded Inference Interface URL: https://github.com/apache/incubator-mxnet/issues/16431#issuecomment-545741769 @ptrendx I am trying to open a PR by Friday. On the status : the two prereqs issues https://github.com/dmlc/dmlc-core/pull/573 and https://github.com/apache/incubator-mxnet/issues/16434 have been better understood and fixed/worked around. I have made C API and backend changes and currently still testing it. Because of time and resource constraints I won't be able to add the CPP frontend changes (which has been mentioned in this PR as targeted for 1.6) in this proposal but only C API changes, backend changes and tests/verification. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on issue #16603: Significant slowdown in some DGL models
anirudh2290 commented on issue #16603: Significant slowdown in some DGL models URL: https://github.com/apache/incubator-mxnet/issues/16603#issuecomment-545740291 which commit on 1.6 ? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] zheng-da commented on issue #16603: Significant slowdown in some DGL models
zheng-da commented on issue #16603: Significant slowdown in some DGL models URL: https://github.com/apache/incubator-mxnet/issues/16603#issuecomment-54571 Here is the profiling result on MXNet 1.5 https://user-images.githubusercontent.com/70481/67448778-a81d0e00-f64a-11e9-8fe5-7ae37af3905f.png";> Here is the profiling result in MXNet 1.6 https://user-images.githubusercontent.com/70481/67448809-c2ef8280-f64a-11e9-9c78-c759028ec7ba.png";> This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] zheng-da opened a new issue #16603: Significant slowdown in some DGL models
zheng-da opened a new issue #16603: Significant slowdown in some DGL models URL: https://github.com/apache/incubator-mxnet/issues/16603 I recently compare the performance of DGL KGE models on MXNet 1.5 and the current master branch. I noticed significant slowdown. On MXNet 1.5, it takes 12 seconds to run 1000 batches, and now takes 20 seconds. It seems there is no slowdown on operators after some profiling. To reproduce the problem, please install DGL 0.4, download the DGL KGE package by cloning [the DGL repo](https://github.com/dmlc/dgl). The DGL KGE package is under apps/kg. Run the following command: ```bash DGLBACKEND=mxnet python3 train.py --model DistMult --dataset FB15k --batch_size 1024 \ --neg_sample_size 256 --hidden_dim 2000 --gamma 500.0 --lr 0.1 --max_step 10 \ --batch_size_eval 16 --gpu 0 --valid --test -adv ``` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ZhennanQin opened a new pull request #16602: [MKLDNN]Fix reorder2default
ZhennanQin opened a new pull request #16602: [MKLDNN]Fix reorder2default URL: https://github.com/apache/incubator-mxnet/pull/16602 convolution and fc weights may get updated during inference, thus we need to handle reorder2default within engine to ensure no write access during this. ## Description ## (Brief description on what this PR is about) ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (c3395ca -> 0742a9b)
This is an automated email from the ASF dual-hosted git repository. anirudh2290 pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from c3395ca [Numpy] Support N_D(N>=3) batch_dot (#16586) add 0742a9b Large Vector tests for DGL Ops Part 2 (#16497) No new revisions were added by this update. Summary of changes: tests/nightly/test_large_array.py | 4 +- tests/nightly/test_large_vector.py | 85 -- tests/python/unittest/test_operator.py | 1 + 3 files changed, 85 insertions(+), 5 deletions(-)
[GitHub] [incubator-mxnet] access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338353987 ## File path: tests/nightly/test_large_array.py ## @@ -1295,17 +1295,17 @@ def check_trunc(): def create_input_for_trigonometric_ops(vals): -# Creates large vector input of size(LARGE_X*10, SMALL_Y/10) from vals using tile operator +# Creates large vector input of size(LARGE_X*10, SMALL_Y/10) from vals using broadcast_to operator Review comment: @anirudhacharya This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (c3395ca -> 0742a9b)
This is an automated email from the ASF dual-hosted git repository. anirudh2290 pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from c3395ca [Numpy] Support N_D(N>=3) batch_dot (#16586) add 0742a9b Large Vector tests for DGL Ops Part 2 (#16497) No new revisions were added by this update. Summary of changes: tests/nightly/test_large_array.py | 4 +- tests/nightly/test_large_vector.py | 85 -- tests/python/unittest/test_operator.py | 1 + 3 files changed, 85 insertions(+), 5 deletions(-)
[GitHub] [incubator-mxnet] anirudh2290 merged pull request #16497: Large Vector tests for DGL Ops Part 2
anirudh2290 merged pull request #16497: Large Vector tests for DGL Ops Part 2 URL: https://github.com/apache/incubator-mxnet/pull/16497 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] CoinCheung opened a new issue #16601: Where is this op.h file ?
CoinCheung opened a new issue #16601: Where is this op.h file ? URL: https://github.com/apache/incubator-mxnet/issues/16601 ## Description I did not find the `op.h` file in the directory. ### Error Message The make command give the error of: /root/build/incubator-mxnet/cpp-package/include/mxnet-cpp/optimizer.hpp:37:26: fatal error: mxnet-cpp/op.h: No such file or directory compilation terminated. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack
vexilligera commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338351249 ## File path: python/mxnet/numpy_dispatch_protocol.py ## @@ -119,6 +119,7 @@ def _run_with_array_ufunc_proto(*args, **kwargs): 'var', 'vdot', 'vstack', +# 'column_stack', Review comment: To pass numpy interoperability test temporarily because there's no test now and it will throw error otherwise. I'll add a test later. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] CoinCheung commented on issue #16587: How could I set the location of openblas/lapack when I compile mxnet from source?
CoinCheung commented on issue #16587: How could I set the location of openblas/lapack when I compile mxnet from source? URL: https://github.com/apache/incubator-mxnet/issues/16587#issuecomment-545705188 For lapack, I modified the path at this line, which does not make sense https://github.com/apache/incubator-mxnet/blob/c3395ca60b20f4388dd76746696497148b82fc80/make/config.mk#L134 For openblas, I have to change the folder name from /opt/openblas to /opt/OpenBLAS to make it work. Would you please show me how I could specify the locations explicitly ? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch mkldnn-v1.0 updated (ca240b2 -> 2210b21)
This is an automated email from the ASF dual-hosted git repository. taolv pushed a change to branch mkldnn-v1.0 in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from ca240b2 Merge remote-tracking branch 'origin/master' into mkldnn-v1.0 add 2210b21 [mkldnn-v1.0] Skip flaky test for unidirectional rnn_relu (#16545) No new revisions were added by this update. Summary of changes: tests/python/unittest/test_operator.py | 24 +++- 1 file changed, 11 insertions(+), 13 deletions(-)
[incubator-mxnet] branch mkldnn-v1.0 updated (ca240b2 -> 2210b21)
This is an automated email from the ASF dual-hosted git repository. taolv pushed a change to branch mkldnn-v1.0 in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from ca240b2 Merge remote-tracking branch 'origin/master' into mkldnn-v1.0 add 2210b21 [mkldnn-v1.0] Skip flaky test for unidirectional rnn_relu (#16545) No new revisions were added by this update. Summary of changes: tests/python/unittest/test_operator.py | 24 +++- 1 file changed, 11 insertions(+), 13 deletions(-)
[GitHub] [incubator-mxnet] TaoLv merged pull request #16545: [mkldnn-v1.0] Skip flaky test for unidirectional rnn_relu
TaoLv merged pull request #16545: [mkldnn-v1.0] Skip flaky test for unidirectional rnn_relu URL: https://github.com/apache/incubator-mxnet/pull/16545 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (05a4c4f -> c3395ca)
This is an automated email from the ASF dual-hosted git repository. haoj pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 05a4c4f Create SECURITY.md (#16573) add c3395ca [Numpy] Support N_D(N>=3) batch_dot (#16586) No new revisions were added by this update. Summary of changes: src/operator/tensor/dot-inl.h | 177 +++-- src/operator/tensor/dot.cc | 84 +--- src/operator/tensor/dot.cu | 3 - tests/python/unittest/test_numpy_op.py | 119 ++ tests/python/unittest/test_operator.py | 4 +- 5 files changed, 249 insertions(+), 138 deletions(-)
[GitHub] [incubator-mxnet] haojin2 merged pull request #16586: [Numpy] Support N_D(N>=3) batch_dot
haojin2 merged pull request #16586: [Numpy] Support N_D(N>=3) batch_dot URL: https://github.com/apache/incubator-mxnet/pull/16586 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on issue #16586: [Numpy] Support N_D(N>=3) batch_dot
haojin2 commented on issue #16586: [Numpy] Support N_D(N>=3) batch_dot URL: https://github.com/apache/incubator-mxnet/pull/16586#issuecomment-545695391 @eric-haibin-lin Can you take another look? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] Caenorst commented on issue #16408: Add MXNet Ops for fast multihead attention
Caenorst commented on issue #16408: Add MXNet Ops for fast multihead attention URL: https://github.com/apache/incubator-mxnet/pull/16408#issuecomment-545695259 @eric-haibin-lin Only with `bwd_ignore_zero_init=True`, otherwise it works find regardless. I let this flag hoping that `MXNET_EXEC_ENABLE_ADDTO` get back soon This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. aaronmarkham pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new 70eda22 Bump the publish timestamp. 70eda22 is described below commit 70eda22099b67fd73e659017f68da0c39a31993e Author: mxnet-ci AuthorDate: Thu Oct 24 00:40:51 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..8c2cc04 --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Thu Oct 24 00:40:51 UTC 2019
[GitHub] [incubator-mxnet] eric-haibin-lin commented on issue #16408: Add MXNet Ops for fast multihead attention
eric-haibin-lin commented on issue #16408: Add MXNet Ops for fast multihead attention URL: https://github.com/apache/incubator-mxnet/pull/16408#issuecomment-545691096 I can no longer find `MXNET_EXEC_ENABLE_ADDTO` in mxnet. Is this flag required? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] TEChopra1000 commented on issue #14994: Flaky test: test_lstm_clip
TEChopra1000 commented on issue #14994: Flaky test: test_lstm_clip URL: https://github.com/apache/incubator-mxnet/issues/14994#issuecomment-545689222 Failed again... http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-gpu/detail/PR-16598/1/pipeline/ This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch v1.6.x updated (df4125a -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. ptrendx pushed a change to branch v1.6.x in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from df4125a update NEWS.md and README.md (#16385) add 0bace55 fix choice signature add ec766d5 add raise test for shape add d5666ed Round and sign straight-through-estimators C operators. (#16373) add 15ea40d Add boolean ndarray (#15940) add 1d0d1e6 Faster Transpose 2D (#16104) add 9ff644b Fix windows flakiness (#16415) add a8181dd [MXNET-1430] julia: implement context.gpu_memory_info (#16324) add 9dc0ab8 global numpy shape flag (#16335) add cfe9e50 Skipping installing nightly test (#16418) add a2018ba cuDNN non-persistant bidirectional RNN dgrad sync fix (#16391) add 56e1bef Adds PyPI CD Pipeline (#16190) add 88521ff upgrade the pytest version (#16429) add 6ce323f [DOC] fix installation selector wrong history (#16381) add 1ab4c95 New ops for RCNN + old ops improvements for RCNN (#16215) add e484f72 Beta build (#16411) add 243ade9 [WIP] Improving Python Docs API (#16392) add cf61364 Revert "add mkl installation temp fix (#16304)" (#16369) add d8193c6 Update add_op_in_backend.md (#16403) add 7f5e687 numpy-compatible histogram (#16266) add ca30ba8 Pseudo 2D transpose kernel (#16229) add d2d76dc increase docker cache timeout (#16430) add 4dee4ee Fix mkldnn reshape (#16455) add 1e8cc90 [BUGFIX] Minor type issues in Squeeze (#16448) add 858a52e Fix large array tests (#16328) add 6d6e46b Comparison ops implemented using mshadow (#16414) add 1d4ede3 Add mask target generator operator for Mask-RCNN (#16268) add 8820220 Adds pip requirements file to nightly gpu ci image (#16472) add 1256976 Fix Nightly Tests for Binaries (#16451) add 812e504 fix autodoc for spurrious toggles (#16452) add 7ce Fix dtype bug (#16467) add 9ab428e [Doc] Update the download page with 1.5.1 release (#16442) add 6e0b1a5 [Numpy] Numpy compatible dstack (#15871) add ceebcaf numpy eye op (#16132) add 8222979 Numpy compatible vsplit; minor changes to split (#15983) add 8562adc add numpy op logspace (#15825) add 9681197 add numpy op bitwise_xor, hsplit, moveaxis, rot90 (#16257) add f9359c3 Fix flakey pylint CI failures (#16462) add 67e1e68 Aggregated zero grad (#16446) add b1932c0 Move MRCNNMaskTarget op to contrib (#16486) add 06438ab Mxnet allclose (#14443) add 0c00a79 Fix optimizer bug for np attribute (#16494) add c2bbde7 Tests of NumPy interoperability (#16469) add bf57ff8 added more tests to verify support for large vector (#16477) add de524bb Fixing broken links (#16500) add a4ea4a8 Load NDArray only to GPU if GPU is present (#16432) add d1200c9 add binary and docs build command options (#16514) add e4f8c50 [MKLDNN] Fix uint quantized fc when not fusing with requantize (#16523) add f6cfbdf improve unary and binary operator handling and refactor tests (#16423) add 77e9898 Bug fix for the input of same axes of the swapaxes operator (#16513) add f2ed1d4 added support for large tensors for Dropout operator and tests to verify support for more operators (#16409) add 63fbfb1 [DOC] Fix numpy op doc (#16504) add f01bcaa [Numpy] More numpy dispatch tests (#16426) add 27f7082 Fix learning rate scheduler being unexpectedly overwritten by optimizer's default value (#16487) add 73bff7d adding large tensor support for add_n and tests for more ops (#16476) add 4b8a95f add option to remove indexes (#16525) add 32bb374 disable tests (#16536) add efa5369 adding large tensor support for pad operator (#15126) add 2d4c3a4 fix pylint in CI (#16540) add 27b3e52 image crop gpu (#16464) add a75ec06 [Numpy] einsum (#15911) add 9fecfbb Add test pipeline for USE_TVM_OP=OFF on Unix (#16450) add b583059 Numpy dispatch test of .. (#16422) add 149e034 typo fix in r doc lstm tutorial (#16546) add fc81c64 Correct Google Analytics Tracker (#16490) add ffec31f Aggregated adamw update (#16398) add 5b67a69 try to fix block (#16465) add c1d02ce setup and concatenate, copy, expand_dims, expm1 (#16493) add cdfaf39 add sum for boolean type in mainline (#16436) add 1648f4c [Numpy] SVD outputs tuple (#16530) add 5accae0 numpy op doc: max, min, prod (#16506) add b949716 add interface for rand add 217ae02 Fix numpy bugs (#16537) add 746cbc5 Add unit tests for TensorRT integration and fix some bugs (#15399) add 93ec1f2 [Doc] Use mirror link in the download page (#16501) add 06ce371 checking broken link fixes work (#16538) add 91bb398 [CD] Adds python docker pipeline (#16547) add 1fb6f00 Build dmlc-core with old thread_local impl
[GitHub] [incubator-mxnet] marcoabreu commented on issue #16599: Imagenet inference to nightly fix
marcoabreu commented on issue #16599: Imagenet inference to nightly fix URL: https://github.com/apache/incubator-mxnet/pull/16599#issuecomment-545685852 That's output from the test. Can you make sure that this test case is not skipped This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch v1.6.x updated (df4125a -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. ptrendx pushed a change to branch v1.6.x in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from df4125a update NEWS.md and README.md (#16385) add 0bace55 fix choice signature add ec766d5 add raise test for shape add d5666ed Round and sign straight-through-estimators C operators. (#16373) add 15ea40d Add boolean ndarray (#15940) add 1d0d1e6 Faster Transpose 2D (#16104) add 9ff644b Fix windows flakiness (#16415) add a8181dd [MXNET-1430] julia: implement context.gpu_memory_info (#16324) add 9dc0ab8 global numpy shape flag (#16335) add cfe9e50 Skipping installing nightly test (#16418) add a2018ba cuDNN non-persistant bidirectional RNN dgrad sync fix (#16391) add 56e1bef Adds PyPI CD Pipeline (#16190) add 88521ff upgrade the pytest version (#16429) add 6ce323f [DOC] fix installation selector wrong history (#16381) add 1ab4c95 New ops for RCNN + old ops improvements for RCNN (#16215) add e484f72 Beta build (#16411) add 243ade9 [WIP] Improving Python Docs API (#16392) add cf61364 Revert "add mkl installation temp fix (#16304)" (#16369) add d8193c6 Update add_op_in_backend.md (#16403) add 7f5e687 numpy-compatible histogram (#16266) add ca30ba8 Pseudo 2D transpose kernel (#16229) add d2d76dc increase docker cache timeout (#16430) add 4dee4ee Fix mkldnn reshape (#16455) add 1e8cc90 [BUGFIX] Minor type issues in Squeeze (#16448) add 858a52e Fix large array tests (#16328) add 6d6e46b Comparison ops implemented using mshadow (#16414) add 1d4ede3 Add mask target generator operator for Mask-RCNN (#16268) add 8820220 Adds pip requirements file to nightly gpu ci image (#16472) add 1256976 Fix Nightly Tests for Binaries (#16451) add 812e504 fix autodoc for spurrious toggles (#16452) add 7ce Fix dtype bug (#16467) add 9ab428e [Doc] Update the download page with 1.5.1 release (#16442) add 6e0b1a5 [Numpy] Numpy compatible dstack (#15871) add ceebcaf numpy eye op (#16132) add 8222979 Numpy compatible vsplit; minor changes to split (#15983) add 8562adc add numpy op logspace (#15825) add 9681197 add numpy op bitwise_xor, hsplit, moveaxis, rot90 (#16257) add f9359c3 Fix flakey pylint CI failures (#16462) add 67e1e68 Aggregated zero grad (#16446) add b1932c0 Move MRCNNMaskTarget op to contrib (#16486) add 06438ab Mxnet allclose (#14443) add 0c00a79 Fix optimizer bug for np attribute (#16494) add c2bbde7 Tests of NumPy interoperability (#16469) add bf57ff8 added more tests to verify support for large vector (#16477) add de524bb Fixing broken links (#16500) add a4ea4a8 Load NDArray only to GPU if GPU is present (#16432) add d1200c9 add binary and docs build command options (#16514) add e4f8c50 [MKLDNN] Fix uint quantized fc when not fusing with requantize (#16523) add f6cfbdf improve unary and binary operator handling and refactor tests (#16423) add 77e9898 Bug fix for the input of same axes of the swapaxes operator (#16513) add f2ed1d4 added support for large tensors for Dropout operator and tests to verify support for more operators (#16409) add 63fbfb1 [DOC] Fix numpy op doc (#16504) add f01bcaa [Numpy] More numpy dispatch tests (#16426) add 27f7082 Fix learning rate scheduler being unexpectedly overwritten by optimizer's default value (#16487) add 73bff7d adding large tensor support for add_n and tests for more ops (#16476) add 4b8a95f add option to remove indexes (#16525) add 32bb374 disable tests (#16536) add efa5369 adding large tensor support for pad operator (#15126) add 2d4c3a4 fix pylint in CI (#16540) add 27b3e52 image crop gpu (#16464) add a75ec06 [Numpy] einsum (#15911) add 9fecfbb Add test pipeline for USE_TVM_OP=OFF on Unix (#16450) add b583059 Numpy dispatch test of .. (#16422) add 149e034 typo fix in r doc lstm tutorial (#16546) add fc81c64 Correct Google Analytics Tracker (#16490) add ffec31f Aggregated adamw update (#16398) add 5b67a69 try to fix block (#16465) add c1d02ce setup and concatenate, copy, expand_dims, expm1 (#16493) add cdfaf39 add sum for boolean type in mainline (#16436) add 1648f4c [Numpy] SVD outputs tuple (#16530) add 5accae0 numpy op doc: max, min, prod (#16506) add b949716 add interface for rand add 217ae02 Fix numpy bugs (#16537) add 746cbc5 Add unit tests for TensorRT integration and fix some bugs (#15399) add 93ec1f2 [Doc] Use mirror link in the download page (#16501) add 06ce371 checking broken link fixes work (#16538) add 91bb398 [CD] Adds python docker pipeline (#16547) add 1fb6f00 Build dmlc-core with old thread_local impl
[incubator-mxnet] branch v1.6.x updated (df4125a -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. ptrendx pushed a change to branch v1.6.x in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from df4125a update NEWS.md and README.md (#16385) add 0bace55 fix choice signature add ec766d5 add raise test for shape add d5666ed Round and sign straight-through-estimators C operators. (#16373) add 15ea40d Add boolean ndarray (#15940) add 1d0d1e6 Faster Transpose 2D (#16104) add 9ff644b Fix windows flakiness (#16415) add a8181dd [MXNET-1430] julia: implement context.gpu_memory_info (#16324) add 9dc0ab8 global numpy shape flag (#16335) add cfe9e50 Skipping installing nightly test (#16418) add a2018ba cuDNN non-persistant bidirectional RNN dgrad sync fix (#16391) add 56e1bef Adds PyPI CD Pipeline (#16190) add 88521ff upgrade the pytest version (#16429) add 6ce323f [DOC] fix installation selector wrong history (#16381) add 1ab4c95 New ops for RCNN + old ops improvements for RCNN (#16215) add e484f72 Beta build (#16411) add 243ade9 [WIP] Improving Python Docs API (#16392) add cf61364 Revert "add mkl installation temp fix (#16304)" (#16369) add d8193c6 Update add_op_in_backend.md (#16403) add 7f5e687 numpy-compatible histogram (#16266) add ca30ba8 Pseudo 2D transpose kernel (#16229) add d2d76dc increase docker cache timeout (#16430) add 4dee4ee Fix mkldnn reshape (#16455) add 1e8cc90 [BUGFIX] Minor type issues in Squeeze (#16448) add 858a52e Fix large array tests (#16328) add 6d6e46b Comparison ops implemented using mshadow (#16414) add 1d4ede3 Add mask target generator operator for Mask-RCNN (#16268) add 8820220 Adds pip requirements file to nightly gpu ci image (#16472) add 1256976 Fix Nightly Tests for Binaries (#16451) add 812e504 fix autodoc for spurrious toggles (#16452) add 7ce Fix dtype bug (#16467) add 9ab428e [Doc] Update the download page with 1.5.1 release (#16442) add 6e0b1a5 [Numpy] Numpy compatible dstack (#15871) add ceebcaf numpy eye op (#16132) add 8222979 Numpy compatible vsplit; minor changes to split (#15983) add 8562adc add numpy op logspace (#15825) add 9681197 add numpy op bitwise_xor, hsplit, moveaxis, rot90 (#16257) add f9359c3 Fix flakey pylint CI failures (#16462) add 67e1e68 Aggregated zero grad (#16446) add b1932c0 Move MRCNNMaskTarget op to contrib (#16486) add 06438ab Mxnet allclose (#14443) add 0c00a79 Fix optimizer bug for np attribute (#16494) add c2bbde7 Tests of NumPy interoperability (#16469) add bf57ff8 added more tests to verify support for large vector (#16477) add de524bb Fixing broken links (#16500) add a4ea4a8 Load NDArray only to GPU if GPU is present (#16432) add d1200c9 add binary and docs build command options (#16514) add e4f8c50 [MKLDNN] Fix uint quantized fc when not fusing with requantize (#16523) add f6cfbdf improve unary and binary operator handling and refactor tests (#16423) add 77e9898 Bug fix for the input of same axes of the swapaxes operator (#16513) add f2ed1d4 added support for large tensors for Dropout operator and tests to verify support for more operators (#16409) add 63fbfb1 [DOC] Fix numpy op doc (#16504) add f01bcaa [Numpy] More numpy dispatch tests (#16426) add 27f7082 Fix learning rate scheduler being unexpectedly overwritten by optimizer's default value (#16487) add 73bff7d adding large tensor support for add_n and tests for more ops (#16476) add 4b8a95f add option to remove indexes (#16525) add 32bb374 disable tests (#16536) add efa5369 adding large tensor support for pad operator (#15126) add 2d4c3a4 fix pylint in CI (#16540) add 27b3e52 image crop gpu (#16464) add a75ec06 [Numpy] einsum (#15911) add 9fecfbb Add test pipeline for USE_TVM_OP=OFF on Unix (#16450) add b583059 Numpy dispatch test of .. (#16422) add 149e034 typo fix in r doc lstm tutorial (#16546) add fc81c64 Correct Google Analytics Tracker (#16490) add ffec31f Aggregated adamw update (#16398) add 5b67a69 try to fix block (#16465) add c1d02ce setup and concatenate, copy, expand_dims, expm1 (#16493) add cdfaf39 add sum for boolean type in mainline (#16436) add 1648f4c [Numpy] SVD outputs tuple (#16530) add 5accae0 numpy op doc: max, min, prod (#16506) add b949716 add interface for rand add 217ae02 Fix numpy bugs (#16537) add 746cbc5 Add unit tests for TensorRT integration and fix some bugs (#15399) add 93ec1f2 [Doc] Use mirror link in the download page (#16501) add 06ce371 checking broken link fixes work (#16538) add 91bb398 [CD] Adds python docker pipeline (#16547) add 1fb6f00 Build dmlc-core with old thread_local impl
[incubator-mxnet] branch v1.6.x updated (df4125a -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. ptrendx pushed a change to branch v1.6.x in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from df4125a update NEWS.md and README.md (#16385) add 0bace55 fix choice signature add ec766d5 add raise test for shape add d5666ed Round and sign straight-through-estimators C operators. (#16373) add 15ea40d Add boolean ndarray (#15940) add 1d0d1e6 Faster Transpose 2D (#16104) add 9ff644b Fix windows flakiness (#16415) add a8181dd [MXNET-1430] julia: implement context.gpu_memory_info (#16324) add 9dc0ab8 global numpy shape flag (#16335) add cfe9e50 Skipping installing nightly test (#16418) add a2018ba cuDNN non-persistant bidirectional RNN dgrad sync fix (#16391) add 56e1bef Adds PyPI CD Pipeline (#16190) add 88521ff upgrade the pytest version (#16429) add 6ce323f [DOC] fix installation selector wrong history (#16381) add 1ab4c95 New ops for RCNN + old ops improvements for RCNN (#16215) add e484f72 Beta build (#16411) add 243ade9 [WIP] Improving Python Docs API (#16392) add cf61364 Revert "add mkl installation temp fix (#16304)" (#16369) add d8193c6 Update add_op_in_backend.md (#16403) add 7f5e687 numpy-compatible histogram (#16266) add ca30ba8 Pseudo 2D transpose kernel (#16229) add d2d76dc increase docker cache timeout (#16430) add 4dee4ee Fix mkldnn reshape (#16455) add 1e8cc90 [BUGFIX] Minor type issues in Squeeze (#16448) add 858a52e Fix large array tests (#16328) add 6d6e46b Comparison ops implemented using mshadow (#16414) add 1d4ede3 Add mask target generator operator for Mask-RCNN (#16268) add 8820220 Adds pip requirements file to nightly gpu ci image (#16472) add 1256976 Fix Nightly Tests for Binaries (#16451) add 812e504 fix autodoc for spurrious toggles (#16452) add 7ce Fix dtype bug (#16467) add 9ab428e [Doc] Update the download page with 1.5.1 release (#16442) add 6e0b1a5 [Numpy] Numpy compatible dstack (#15871) add ceebcaf numpy eye op (#16132) add 8222979 Numpy compatible vsplit; minor changes to split (#15983) add 8562adc add numpy op logspace (#15825) add 9681197 add numpy op bitwise_xor, hsplit, moveaxis, rot90 (#16257) add f9359c3 Fix flakey pylint CI failures (#16462) add 67e1e68 Aggregated zero grad (#16446) add b1932c0 Move MRCNNMaskTarget op to contrib (#16486) add 06438ab Mxnet allclose (#14443) add 0c00a79 Fix optimizer bug for np attribute (#16494) add c2bbde7 Tests of NumPy interoperability (#16469) add bf57ff8 added more tests to verify support for large vector (#16477) add de524bb Fixing broken links (#16500) add a4ea4a8 Load NDArray only to GPU if GPU is present (#16432) add d1200c9 add binary and docs build command options (#16514) add e4f8c50 [MKLDNN] Fix uint quantized fc when not fusing with requantize (#16523) add f6cfbdf improve unary and binary operator handling and refactor tests (#16423) add 77e9898 Bug fix for the input of same axes of the swapaxes operator (#16513) add f2ed1d4 added support for large tensors for Dropout operator and tests to verify support for more operators (#16409) add 63fbfb1 [DOC] Fix numpy op doc (#16504) add f01bcaa [Numpy] More numpy dispatch tests (#16426) add 27f7082 Fix learning rate scheduler being unexpectedly overwritten by optimizer's default value (#16487) add 73bff7d adding large tensor support for add_n and tests for more ops (#16476) add 4b8a95f add option to remove indexes (#16525) add 32bb374 disable tests (#16536) add efa5369 adding large tensor support for pad operator (#15126) add 2d4c3a4 fix pylint in CI (#16540) add 27b3e52 image crop gpu (#16464) add a75ec06 [Numpy] einsum (#15911) add 9fecfbb Add test pipeline for USE_TVM_OP=OFF on Unix (#16450) add b583059 Numpy dispatch test of .. (#16422) add 149e034 typo fix in r doc lstm tutorial (#16546) add fc81c64 Correct Google Analytics Tracker (#16490) add ffec31f Aggregated adamw update (#16398) add 5b67a69 try to fix block (#16465) add c1d02ce setup and concatenate, copy, expand_dims, expm1 (#16493) add cdfaf39 add sum for boolean type in mainline (#16436) add 1648f4c [Numpy] SVD outputs tuple (#16530) add 5accae0 numpy op doc: max, min, prod (#16506) add b949716 add interface for rand add 217ae02 Fix numpy bugs (#16537) add 746cbc5 Add unit tests for TensorRT integration and fix some bugs (#15399) add 93ec1f2 [Doc] Use mirror link in the download page (#16501) add 06ce371 checking broken link fixes work (#16538) add 91bb398 [CD] Adds python docker pipeline (#16547) add 1fb6f00 Build dmlc-core with old thread_local impl
[incubator-mxnet] branch v1.6.x updated (df4125a -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. ptrendx pushed a change to branch v1.6.x in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from df4125a update NEWS.md and README.md (#16385) add 0bace55 fix choice signature add ec766d5 add raise test for shape add d5666ed Round and sign straight-through-estimators C operators. (#16373) add 15ea40d Add boolean ndarray (#15940) add 1d0d1e6 Faster Transpose 2D (#16104) add 9ff644b Fix windows flakiness (#16415) add a8181dd [MXNET-1430] julia: implement context.gpu_memory_info (#16324) add 9dc0ab8 global numpy shape flag (#16335) add cfe9e50 Skipping installing nightly test (#16418) add a2018ba cuDNN non-persistant bidirectional RNN dgrad sync fix (#16391) add 56e1bef Adds PyPI CD Pipeline (#16190) add 88521ff upgrade the pytest version (#16429) add 6ce323f [DOC] fix installation selector wrong history (#16381) add 1ab4c95 New ops for RCNN + old ops improvements for RCNN (#16215) add e484f72 Beta build (#16411) add 243ade9 [WIP] Improving Python Docs API (#16392) add cf61364 Revert "add mkl installation temp fix (#16304)" (#16369) add d8193c6 Update add_op_in_backend.md (#16403) add 7f5e687 numpy-compatible histogram (#16266) add ca30ba8 Pseudo 2D transpose kernel (#16229) add d2d76dc increase docker cache timeout (#16430) add 4dee4ee Fix mkldnn reshape (#16455) add 1e8cc90 [BUGFIX] Minor type issues in Squeeze (#16448) add 858a52e Fix large array tests (#16328) add 6d6e46b Comparison ops implemented using mshadow (#16414) add 1d4ede3 Add mask target generator operator for Mask-RCNN (#16268) add 8820220 Adds pip requirements file to nightly gpu ci image (#16472) add 1256976 Fix Nightly Tests for Binaries (#16451) add 812e504 fix autodoc for spurrious toggles (#16452) add 7ce Fix dtype bug (#16467) add 9ab428e [Doc] Update the download page with 1.5.1 release (#16442) add 6e0b1a5 [Numpy] Numpy compatible dstack (#15871) add ceebcaf numpy eye op (#16132) add 8222979 Numpy compatible vsplit; minor changes to split (#15983) add 8562adc add numpy op logspace (#15825) add 9681197 add numpy op bitwise_xor, hsplit, moveaxis, rot90 (#16257) add f9359c3 Fix flakey pylint CI failures (#16462) add 67e1e68 Aggregated zero grad (#16446) add b1932c0 Move MRCNNMaskTarget op to contrib (#16486) add 06438ab Mxnet allclose (#14443) add 0c00a79 Fix optimizer bug for np attribute (#16494) add c2bbde7 Tests of NumPy interoperability (#16469) add bf57ff8 added more tests to verify support for large vector (#16477) add de524bb Fixing broken links (#16500) add a4ea4a8 Load NDArray only to GPU if GPU is present (#16432) add d1200c9 add binary and docs build command options (#16514) add e4f8c50 [MKLDNN] Fix uint quantized fc when not fusing with requantize (#16523) add f6cfbdf improve unary and binary operator handling and refactor tests (#16423) add 77e9898 Bug fix for the input of same axes of the swapaxes operator (#16513) add f2ed1d4 added support for large tensors for Dropout operator and tests to verify support for more operators (#16409) add 63fbfb1 [DOC] Fix numpy op doc (#16504) add f01bcaa [Numpy] More numpy dispatch tests (#16426) add 27f7082 Fix learning rate scheduler being unexpectedly overwritten by optimizer's default value (#16487) add 73bff7d adding large tensor support for add_n and tests for more ops (#16476) add 4b8a95f add option to remove indexes (#16525) add 32bb374 disable tests (#16536) add efa5369 adding large tensor support for pad operator (#15126) add 2d4c3a4 fix pylint in CI (#16540) add 27b3e52 image crop gpu (#16464) add a75ec06 [Numpy] einsum (#15911) add 9fecfbb Add test pipeline for USE_TVM_OP=OFF on Unix (#16450) add b583059 Numpy dispatch test of .. (#16422) add 149e034 typo fix in r doc lstm tutorial (#16546) add fc81c64 Correct Google Analytics Tracker (#16490) add ffec31f Aggregated adamw update (#16398) add 5b67a69 try to fix block (#16465) add c1d02ce setup and concatenate, copy, expand_dims, expm1 (#16493) add cdfaf39 add sum for boolean type in mainline (#16436) add 1648f4c [Numpy] SVD outputs tuple (#16530) add 5accae0 numpy op doc: max, min, prod (#16506) add b949716 add interface for rand add 217ae02 Fix numpy bugs (#16537) add 746cbc5 Add unit tests for TensorRT integration and fix some bugs (#15399) add 93ec1f2 [Doc] Use mirror link in the download page (#16501) add 06ce371 checking broken link fixes work (#16538) add 91bb398 [CD] Adds python docker pipeline (#16547) add 1fb6f00 Build dmlc-core with old thread_local impl
[GitHub] [incubator-mxnet] eric-haibin-lin commented on a change in pull request #16532: fix dropout gpu seed
eric-haibin-lin commented on a change in pull request #16532: fix dropout gpu seed URL: https://github.com/apache/incubator-mxnet/pull/16532#discussion_r338328289 ## File path: tests/nightly/test_dropout.py ## @@ -0,0 +1,49 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import sys +import os +curr_path = os.path.dirname(os.path.abspath(os.path.expanduser(__file__))) +sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'python', 'unittest')) +from common import with_seed +import unittest + +import mxnet as mx +import numpy as np +from mxnet.test_utils import assert_almost_equal +from nose.tools import assert_raises + +@with_seed() Review comment: would you mind adding a unit test for the RNN op, too? Thx! This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] eric-haibin-lin commented on a change in pull request #16532: fix dropout gpu seed
eric-haibin-lin commented on a change in pull request #16532: fix dropout gpu seed URL: https://github.com/apache/incubator-mxnet/pull/16532#discussion_r338327895 ## File path: src/operator/nn/dropout-inl.h ## @@ -253,7 +253,19 @@ class DropoutOp { const TBlob &mask, const TBlob &out) { Stream *s = ctx.get_stream(); - + Random *prnd = ctx.requested[1].get_random(s); + Tensor workspace = +ctx.requested[2].get_space_typed(Shape1(1 * sizeof(unsigned)), s); + // slice workspace + char *workspace_ptr = workspace.dptr_; + Tensor random_number = +Tensor(reinterpret_cast(workspace_ptr), + Shape1(1), s); + prnd->GetRandInt(random_number); + // copy generated random int to cpu + unsigned data = 0; + CUDA_CALL(cudaMemcpy(&data, &random_number[0], sizeof(unsigned), cudaMemcpyDeviceToHost)); + uint64_t seed_ = 17 + static_cast(data) % 4096; Review comment: I already created one https://github.com/apache/incubator-mxnet/issues/16583 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on issue #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
ChaiBapchya commented on issue #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#issuecomment-545681670 Also addressed the suggestions by @anirudhacharya in #16416 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on a change in pull request #16416: Dgl ops 2
ChaiBapchya commented on a change in pull request #16416: Dgl ops 2 URL: https://github.com/apache/incubator-mxnet/pull/16416#discussion_r338327543 ## File path: tests/nightly/test_large_array.py ## @@ -1351,6 +1407,48 @@ def check_tan(): expected_output = [-.577, -1, 0, 1, .577] assert_correctness_of_trigonometric_ops(y, expected_output) +def check_arcsinh(): +x = create_input_for_trigonometric_ops([-np.pi/2, -np.pi/4, 0, np.pi/4, np.pi/2]) +y = nd.arcsinh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying arcsinh() +expected_output = [np.arcsinh(-np.pi/2), np.arcsinh(-np.pi/4), 0, np.arcsinh(np.pi/4), np.arcsinh(np.pi/2)] +assert_correctness_of_trigonometric_ops(y, expected_output) + +def check_arccosh(): +x = create_input_for_trigonometric_ops([1, np.pi/2, 3*np.pi/4, np.pi]) +y = nd.arccosh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying arccosh() +expected_output = [0, np.arccosh(np.pi/2), np.arccosh(3*np.pi/4), np.arccosh(np.pi)] +assert_correctness_of_trigonometric_ops(y, expected_output) + +def check_arctanh(): +x = create_input_for_trigonometric_ops([-1/4, -1/2, 0, 1/4, 1/2]) +y = nd.arctanh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying arctanh() +expected_output = [np.arctanh(-1/4), np.arctanh(-1/2), 0, np.arctanh(1/4), np.arctanh(1/2)] +assert_correctness_of_trigonometric_ops(y, expected_output) + +def check_sinh(): +x = create_input_for_trigonometric_ops([-np.pi/2, -np.pi/4, 0, np.pi/4, np.pi/2]) +y = nd.sinh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying sinh() +expected_output = [np.sinh(-np.pi/2), np.sinh(-np.pi/4), 0, np.sinh(np.pi/4), np.sinh(np.pi/2)] +assert_correctness_of_trigonometric_ops(y, expected_output) + +def check_cosh(): +x = create_input_for_trigonometric_ops([0, 1, np.pi/2, 3*np.pi/4, np.pi]) +y = nd.cosh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying cosh() +expected_output = [1, np.cosh(1), np.cosh(np.pi/2), np.cosh(3*np.pi/4), np.cosh(np.pi)] +assert_correctness_of_trigonometric_ops(y, expected_output) + +def check_tanh(): +x = create_input_for_trigonometric_ops([-1/4, -1/2, 0, 1/4, 1/2]) +y = nd.tanh(x) +# expected ouput for indices=(0, 1, -3, -2, -1) after applying tanh() +expected_output = [np.tanh(-1/4), np.tanh(-1/2), 0, np.tanh(1/4), np.tanh(1/2)] +assert_correctness_of_trigonometric_ops(y, expected_output) Review comment: Addressed it here #16585 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #16431: [RFC] MXNet Multithreaded Inference Interface
ptrendx commented on issue #16431: [RFC] MXNet Multithreaded Inference Interface URL: https://github.com/apache/incubator-mxnet/issues/16431#issuecomment-545681158 Hi @anirudh2290, what is the status of this proposal? When do you think changes will be ready? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338327118 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys, g
[GitHub] [incubator-mxnet] haojin2 opened a new issue #16600: Clojure test failure: Could not transfer artifact com.fasterxml.jackson.core
haojin2 opened a new issue #16600: Clojure test failure: Could not transfer artifact com.fasterxml.jackson.core URL: https://github.com/apache/incubator-mxnet/issues/16600 Encountered at http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-cpu/detail/PR-16597/4/pipeline/ : Could not transfer artifact com.fasterxml.jackson.core:jackson-core:jar:2.9.0 from/to central (https://repo1.maven.org/maven2/): GET request of: com/fasterxml/jackson/core/jackson-core/2.9.0/jackson-core-2.9.0.jar from central failed Could not find artifact com.fasterxml.jackson.core:jackson-core:jar:2.9.0 in clojars (https://repo.clojars.org/) This could be due to a typo in :dependencies, file system permissions, or network issues. If you are behind a proxy, try setting the 'http_proxy' environment variable. @gigasquid Is this due to the flakiness of package serving infrastructure? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on issue #16599: Imagenet inference to nightly fix
ChaiBapchya commented on issue #16599: Imagenet inference to nightly fix URL: https://github.com/apache/incubator-mxnet/pull/16599#issuecomment-545680518 Oh I didn't get you. Should I comment it? Remove it? Handle it? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (e22e93f -> 05a4c4f)
This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from e22e93f fix missing docs due to git add issues (#16496) add 05a4c4f Create SECURITY.md (#16573) No new revisions were added by this update. Summary of changes: tools/caffe_converter/README.md => SECURITY.md | 15 +-- 1 file changed, 5 insertions(+), 10 deletions(-) copy tools/caffe_converter/README.md => SECURITY.md (65%)
[GitHub] [incubator-mxnet] marcoabreu commented on issue #16599: Imagenet inference to nightly fix
marcoabreu commented on issue #16599: Imagenet inference to nightly fix URL: https://github.com/apache/incubator-mxnet/pull/16599#issuecomment-545676057 "Skipped INT8 test because mkldnn was not found which is required for running inference with quantized models." Can you take care of that as well please? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] szha merged pull request #16573: Create SECURITY.md
szha merged pull request #16573: Create SECURITY.md URL: https://github.com/apache/incubator-mxnet/pull/16573 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16586: [Numpy] Support N_D(N>=3) batch_dot
haojin2 commented on a change in pull request #16586: [Numpy] Support N_D(N>=3) batch_dot URL: https://github.com/apache/incubator-mxnet/pull/16586#discussion_r338321472 ## File path: src/operator/tensor/dot.cc ## @@ -138,21 +138,73 @@ which is computed by:: return std::vector{ResourceRequest::kTempSpace}; }) .set_attr("FCompute", BatchDotForward_) -.set_attr("FGradient", ElemwiseGradUseIn{"_backward_batch_dot"}) +.set_attr("FGradient", +[](const nnvm::NodePtr& n, + const std::vector& ograds) { + const DotParam& param = nnvm::get(n->attrs.parsed); + nnvm::NodePtr lhs_grad; + nnvm::NodePtr rhs_grad; + std::string lhs_gnode_name = n->attrs.name + "_backward_lhs"; + std::string rhs_gnode_name = n->attrs.name + "_backward_rhs"; + if (param.transpose_a && param.transpose_b) { +// Gradient of z = dot(x.T, y.T) +// dx = dot(dz, y).T = dot(y.T, dz.T) +// dy = dot(x, dz).T = dot(dz.T, x.T) +lhs_grad = MakeNode("batch_dot", lhs_gnode_name, +{n->inputs[1], ograds[0]}, &(n->attrs.dict), &n); +rhs_grad = MakeNode("batch_dot", rhs_gnode_name, +{ograds[0], n->inputs[0]}, &(n->attrs.dict), &n); + } else if (!param.transpose_a && param.transpose_b) { +// Gradient of z = dot(x, y.T) +// dx = dot(dz, y) +// dy = dot(x.T, dz).T = dot(dz.T, x) +auto lhs_attrs_dict = n->attrs.dict; +auto rhs_attrs_dict = n->attrs.dict; +lhs_attrs_dict["transpose_a"] = "false"; +lhs_attrs_dict["transpose_b"] = "false"; +rhs_attrs_dict["transpose_a"] = "true"; +rhs_attrs_dict["transpose_b"] = "false"; +lhs_grad = MakeNode("batch_dot", lhs_gnode_name, +{ograds[0], n->inputs[1]}, &lhs_attrs_dict, &n); +rhs_grad = MakeNode("batch_dot", rhs_gnode_name, +{ograds[0], n->inputs[0]}, &rhs_attrs_dict, &n); + } else if (param.transpose_a && !param.transpose_b) { +// Gradient of z = dot(x.T, y) +// dx = dot(dz, y.T).T = dot(y, dz.T) +// dy = dot(x, dz) +auto lhs_attrs_dict = n->attrs.dict; +auto rhs_attrs_dict = n->attrs.dict; +lhs_attrs_dict["transpose_a"] = "false"; +lhs_attrs_dict["transpose_b"] = "true"; +rhs_attrs_dict["transpose_a"] = "false"; +rhs_attrs_dict["transpose_b"] = "false"; +lhs_grad = MakeNode("batch_dot", lhs_gnode_name, +{n->inputs[1], ograds[0]}, &lhs_attrs_dict, &n); +rhs_grad = MakeNode("batch_dot", rhs_gnode_name, +{n->inputs[0], ograds[0]}, &rhs_attrs_dict, &n); + } else { +// Gradient of z = dot(x, y) +// dx = dot(dz, y.T) +// dy = dot(x.T, dz) +auto lhs_attrs_dict = n->attrs.dict; +auto rhs_attrs_dict = n->attrs.dict; +lhs_attrs_dict["transpose_a"] = "false"; +lhs_attrs_dict["transpose_b"] = "true"; +rhs_attrs_dict["transpose_a"] = "true"; +rhs_attrs_dict["transpose_b"] = "false"; +lhs_grad = MakeNode("batch_dot", lhs_gnode_name, +{ograds[0], n->inputs[1]}, &lhs_attrs_dict, &n); +rhs_grad = MakeNode("batch_dot", rhs_gnode_name, +{n->inputs[0], ograds[0]}, &rhs_attrs_dict, &n); + } + std::vector ret; + ret.emplace_back(nnvm::NodeEntry{lhs_grad, 0, 0}); + ret.emplace_back(nnvm::NodeEntry{rhs_grad, 0, 0}); + return ret; +}) .add_argument("lhs", "NDArray-or-Symbol", "The first input") .add_argument("rhs", "NDArray-or-Symbol", "The second input") .add_arguments(DotParam::__FIELDS__()); Review comment: @eric-haibin-lin Here is the benchmark script for getting backward performance: https://gist.github.com/haojin2/c1a2bd1373530f4686bdefd2eafbee84 Results: lhs: (32, 128, 768) rhs: (32, 128, 768) transpose_b: True 0.212037ms -> 0.213933ms lhs: (32, 1, 768) rhs: (32, 128, 768) transpose_b: True 0.119977ms -> 0.124208ms There's no obvious regression in performance. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on issue #16599: Imagenet inference to nightly fix
ChaiBapchya commented on issue #16599: Imagenet inference to nightly fix URL: https://github.com/apache/incubator-mxnet/pull/16599#issuecomment-545664428 Verified this works correctly here - http://jenkins.mxnet-ci-dev.amazon-ml.com/blue/organizations/jenkins/NightlyTestsForBinaries/detail/shared_library_nightly_fix/3/pipeline This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya opened a new pull request #16599: Shared library nightly fix
ChaiBapchya opened a new pull request #16599: Shared library nightly fix URL: https://github.com/apache/incubator-mxnet/pull/16599 ## Description ## PR #16577 was introduced to move imagenet inference to nightly However, I made the PR prematurely (without testing it completely) As it so happened (https://github.com/apache/incubator-mxnet/pull/16577/files#r337663583) Jenkins CI (NightlyTestForBinaries) failed coz of issue in http://jenkins.mxnet-ci-dev.amazon-ml.com/blue/organizations/jenkins/NightlyTestsForBinaries/detail/move_imagenet_inference_nightly/20/pipeline/72 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338308956 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys, g
[GitHub] [incubator-mxnet] ChaiBapchya commented on issue #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
ChaiBapchya commented on issue #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#issuecomment-545661044 ``` nosetests tests/nightly/test_large_vector.py:test_regression . -- Ran 1 test in 745.939s ``` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] TEChopra1000 opened a new pull request #16598: second round of fixing broken links in multiple files
TEChopra1000 opened a new pull request #16598: second round of fixing broken links in multiple files URL: https://github.com/apache/incubator-mxnet/pull/16598 ## Description ## (Brief description on what this PR is about) ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338301882 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys, g
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16597: [Numpy] [WIP] Loading numpy-incompatible NDArray in numpy-compatible mode
haojin2 commented on a change in pull request #16597: [Numpy] [WIP] Loading numpy-incompatible NDArray in numpy-compatible mode URL: https://github.com/apache/incubator-mxnet/pull/16597#discussion_r338294777 ## File path: scala-package/native/src/main/native/org_apache_mxnet_native_c_api.cc ## @@ -2778,7 +2778,7 @@ JNIEXPORT jint JNICALL Java_org_apache_mxnet_LibInfo_mxDumpProfile JNIEXPORT jint JNICALL Java_org_apache_mxnet_LibInfo_mxIsNumpyShape (JNIEnv *env, jobject obj, jobject compatibleRef) { bool isNumpyShape; Review comment: Change `bool` to `int` here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated (91ad266 -> e22e93f)
This is an automated email from the ASF dual-hosted git repository. thomasdelteil pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 91ad266 fixed broken links across multiple files (#16581) add e22e93f fix missing docs due to git add issues (#16496) No new revisions were added by this update. Summary of changes: .../python/api/gluon/{loss => data}/index.rst | 44 ++-- .../vision/datasets/index.rst} | 7 +-- .../{contrib/onnx => gluon/data/vision}/index.rst | 35 +++-- .../{rnn => data/vision/transforms}/index.rst | 60 +- .../python/api/mxnet/{rtc => log}/index.rst| 6 +-- .../python/api/mxnet/{torch => model}/index.rst| 4 +- 6 files changed, 104 insertions(+), 52 deletions(-) copy docs/python_docs/python/api/gluon/{loss => data}/index.rst (66%) copy docs/python_docs/python/api/gluon/{parameter.rst => data/vision/datasets/index.rst} (82%) copy docs/python_docs/python/api/{contrib/onnx => gluon/data/vision}/index.rst (67%) copy docs/python_docs/python/api/gluon/{rnn => data/vision/transforms}/index.rst (55%) copy docs/python_docs/python/api/mxnet/{rtc => log}/index.rst (93%) copy docs/python_docs/python/api/mxnet/{torch => model}/index.rst (95%)
[incubator-mxnet] branch master updated (91ad266 -> e22e93f)
This is an automated email from the ASF dual-hosted git repository. thomasdelteil pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 91ad266 fixed broken links across multiple files (#16581) add e22e93f fix missing docs due to git add issues (#16496) No new revisions were added by this update. Summary of changes: .../python/api/gluon/{loss => data}/index.rst | 44 ++-- .../vision/datasets/index.rst} | 7 +-- .../{contrib/onnx => gluon/data/vision}/index.rst | 35 +++-- .../{rnn => data/vision/transforms}/index.rst | 60 +- .../python/api/mxnet/{rtc => log}/index.rst| 6 +-- .../python/api/mxnet/{torch => model}/index.rst| 4 +- 6 files changed, 104 insertions(+), 52 deletions(-) copy docs/python_docs/python/api/gluon/{loss => data}/index.rst (66%) copy docs/python_docs/python/api/gluon/{parameter.rst => data/vision/datasets/index.rst} (82%) copy docs/python_docs/python/api/{contrib/onnx => gluon/data/vision}/index.rst (67%) copy docs/python_docs/python/api/gluon/{rnn => data/vision/transforms}/index.rst (55%) copy docs/python_docs/python/api/mxnet/{rtc => log}/index.rst (93%) copy docs/python_docs/python/api/mxnet/{torch => model}/index.rst (95%)
[GitHub] [incubator-mxnet] ThomasDelteil merged pull request #16496: fix missing docs due to git add issues
ThomasDelteil merged pull request #16496: fix missing docs due to git add issues URL: https://github.com/apache/incubator-mxnet/pull/16496 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ThomasDelteil closed issue #16495: docs for gluon.data.* are missing
ThomasDelteil closed issue #16495: docs for gluon.data.* are missing URL: https://github.com/apache/incubator-mxnet/issues/16495 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338291832 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys,
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338289964 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), +((0), (0), (0)), +((2, 2), (2, 1), (2, 3)), +((4, 3), (4, 4), (4, 1)), +((2, 2, 2), (2, 4, 2), (2, 2, 2)), +((0, 1, 1), (0, 1, 1), (0, 1, 1)), +((2, 1), (2, 2), (2, 2)) +] +types = ['float16', 'float32', 'float64', 'int8', 'int32', 'int64'] +for config in configs: +for hybridize in [True, False]: +for dtype in types: +test_column_stack = TestColumnStack() +if hybridize: +test_column_stack.hybridize() +rtol = 1e-3 +atol = 1e-5 +v = [] +v_np = [] +for i in range(3): +v_np.append(_np.array(_np.random.uniform(-10.0, 10.0, config[i]), dtype=dtype)) +v.append(mx.nd.array(v_np[i]).as_np_ndarray()) +v[i].attach_grad() +expected_np = _np.column_stack(v_np) +with mx.autograd.record(): +mx_out = test_column_stack(*v) +assert mx_out.shape == expected_np.shape +assert_almost_equal(mx_out.asnumpy(), expected_np, rtol=rtol, atol=atol) + +# Test gradient +mx_out.backward() +for i in range(3): +expected_grad = g(v_np[i]) +assert_almost_equal(v[i].grad.asnumpy(), expected_grad, rtol=rtol, atol=atol) + +# Test imperative once again +mx_out = np.column_stack(v) +expected_np = _np.column_stack(v_np) +assert_almost_equal(mx_out.asnumpy(), expected_np, rtol=rtol, atol=atol) + Review comment: 2 blank lines here. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338289541 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), +((0), (0), (0)), +((2, 2), (2, 1), (2, 3)), +((4, 3), (4, 4), (4, 1)), +((2, 2, 2), (2, 4, 2), (2, 2, 2)), +((0, 1, 1), (0, 1, 1), (0, 1, 1)), +((2, 1), (2, 2), (2, 2)) Review comment: Same here. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338289283 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), Review comment: Also think this might actually be a duplicate case of "column_stack of 2-D arrays with non-zero column sizes" This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338288758 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), +((0), (0), (0)), +((2, 2), (2, 1), (2, 3)), +((4, 3), (4, 4), (4, 1)), Review comment: Or you could change this case to `((4, 3), (4, 0), (4, 1))` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338288416 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys, g
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338288163 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3542,6 +3542,62 @@ def test_np_true_divide(): assert_almost_equal(out_mx.asnumpy(), out_np, rtol=1e-3, atol=1e-3, use_broadcast=False) +@with_seed() +@use_np +def test_np_column_stack(): +class TestColumnStack(HybridBlock): +def __init__(self): +super(TestColumnStack, self).__init__() + +def hybrid_forward(self, F, a, *args): +return F.np.column_stack([a] + list(args)) + +def g(data): +return _np.ones_like(data) + +configs = [ +((), (), ()), +((2), (2), (2)), +((1, 3), (1, 3), (1, 3)), +((0), (0), (0)), +((2, 2), (2, 1), (2, 3)), +((4, 3), (4, 4), (4, 1)), Review comment: This case is a duplicate with the case above, remove either one of them This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338286910 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys,
[GitHub] [incubator-mxnet] access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
access2rohit commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338286910 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys,
[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338286200 ## File path: src/c_api/c_api_executor.cc ## @@ -515,44 +515,11 @@ int MXExecutorSimpleBind(SymbolHandle symbol_handle, API_END(); } -/*! - * \brief - * \param symbol_handle symbol handle - * \param dev_type default device type - * \param dev_id default device id - * \param num_g2c_keys number of group2ctx keys - * \param g2c_keys key list of group2ctx - * \param g2c_dev_types device type list of group2ctx - * \param g2c_dev_ids id list of group2ctx - * \param provided_grad_req_list_len grad_req length provided by users in front-end - * \param provided_grad_req_names grad_req names provided by users in front-end - * \param provided_grad_req_types req types provided by users in front-end - * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes - * \param provided_arg_shape_names name list of provided shapes - * \param provided_arg_shape_data provided shape data - * \param provided_arg_shape_idx provided shape data index - * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes - * \param provided_arg_dtype_names argument name list of provided dtypes - * \param provided_arg_dtypes data of provided dtypes - * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types - * \param provided_arg_stype_names argument name list of provided storage types - * \param provided_arg_stypes data of provided storage types - * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec - * \param shared_arg_name_list parameter name list passed from _bind_ith_exec - * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec - * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec - * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec - * \param updated_shared_buffer_name_list updated shared data array names after binding - * \param updated_shared_buffer_handle_list updated shared data arrays after binding - * \param num_in_args number of input arguments of this sym - * \param in_args list_arguments associated with the current executor - * \param arg_grads list of gradients of in_args associated with the current executor - * \param num_aux_states number of aux states of this sym - * \param aux_states list_auxiliary_states associated with the current executor - * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec - * \param out the handle of the executor to be created - */ -int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + +namespace mxnet { + +template +int SimpleBindExMaster(SymbolHandle symbol_handle, Review comment: _SimpleBindImpl This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] sxjscience commented on a change in pull request #16589: Fix index overflow bug in einsum
sxjscience commented on a change in pull request #16589: Fix index overflow bug in einsum URL: https://github.com/apache/incubator-mxnet/pull/16589#discussion_r338285002 ## File path: tests/python/unittest/test_numpy_op.py ## @@ -3377,16 +3377,20 @@ def dbg(name, data): _np.dot(args[0].T, _np.dot(_np.ones((2, 2)), args[2].T)), _np.dot(_np.dot(args[0], args[1]).T, _np.ones((2, 2), # broadcast bug -(('ij, ij -> i'), [(1, 4), (2, 4)], lambda *args: (_np.sum(args[1], axis=0)[None, :], - _np.tile(args[0], [2, 1]))), +('ij, ij -> i', [(1, 4), (2, 4)], lambda *args: (_np.sum(args[1], axis=0)[None, :], + _np.tile(args[0], [2, 1]))), +# issue #16576 +# commented due to long running time +# ('abiz,abjz->abij', [(64, 8, 128, 512), (64, 8, 128, 512)], lambda *args: (_np.matmul(_np.ones((64, 8, 128, 128)), args[1]), +# _np.matmul(_np.ones((64, 8, 128, 128)), args[0]))), ] -dtypes = ['int32', 'float16', 'float32', 'float64'] +dtypes = ['int32', 'float32', 'float64'] for hybridize in [False, True]: for dtype in dtypes: for config in configs: for optimize in [False, True]: -rtol = 1e-0 if dtype == 'float16' else 1e-3 -atol = 1e-1 if dtype == 'float16' else 1e-5 +rtol = 1e-0 if dtype == 'float16' else 1e-1 +atol = 1e-1 if dtype == 'float16' else 1e-1 Review comment: @hzfan You may also refer to this tutorial in TF: https://www.tensorflow.org/guide/keras/rnn This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
ChaiBapchya commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338283466 ## File path: src/c_api/c_api_executor.cc ## @@ -515,44 +515,11 @@ int MXExecutorSimpleBind(SymbolHandle symbol_handle, API_END(); } -/*! - * \brief - * \param symbol_handle symbol handle - * \param dev_type default device type - * \param dev_id default device id - * \param num_g2c_keys number of group2ctx keys - * \param g2c_keys key list of group2ctx - * \param g2c_dev_types device type list of group2ctx - * \param g2c_dev_ids id list of group2ctx - * \param provided_grad_req_list_len grad_req length provided by users in front-end - * \param provided_grad_req_names grad_req names provided by users in front-end - * \param provided_grad_req_types req types provided by users in front-end - * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes - * \param provided_arg_shape_names name list of provided shapes - * \param provided_arg_shape_data provided shape data - * \param provided_arg_shape_idx provided shape data index - * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes - * \param provided_arg_dtype_names argument name list of provided dtypes - * \param provided_arg_dtypes data of provided dtypes - * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types - * \param provided_arg_stype_names argument name list of provided storage types - * \param provided_arg_stypes data of provided storage types - * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec - * \param shared_arg_name_list parameter name list passed from _bind_ith_exec - * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec - * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec - * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec - * \param updated_shared_buffer_name_list updated shared data array names after binding - * \param updated_shared_buffer_handle_list updated shared data arrays after binding - * \param num_in_args number of input arguments of this sym - * \param in_args list_arguments associated with the current executor - * \param arg_grads list of gradients of in_args associated with the current executor - * \param num_aux_states number of aux states of this sym - * \param aux_states list_auxiliary_states associated with the current executor - * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec - * \param out the handle of the executor to be created - */ -int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + +namespace mxnet { + +template +int SimpleBindExMaster(SymbolHandle symbol_handle, Review comment: also @access2rohit suggested renaming it with SimpleBindExImpl So combining both suggestions - function name should be `_SimpleBindExImpl` ? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338282692 ## File path: src/c_api/c_api_executor.cc ## @@ -515,44 +515,11 @@ int MXExecutorSimpleBind(SymbolHandle symbol_handle, API_END(); } -/*! - * \brief - * \param symbol_handle symbol handle - * \param dev_type default device type - * \param dev_id default device id - * \param num_g2c_keys number of group2ctx keys - * \param g2c_keys key list of group2ctx - * \param g2c_dev_types device type list of group2ctx - * \param g2c_dev_ids id list of group2ctx - * \param provided_grad_req_list_len grad_req length provided by users in front-end - * \param provided_grad_req_names grad_req names provided by users in front-end - * \param provided_grad_req_types req types provided by users in front-end - * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes - * \param provided_arg_shape_names name list of provided shapes - * \param provided_arg_shape_data provided shape data - * \param provided_arg_shape_idx provided shape data index - * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes - * \param provided_arg_dtype_names argument name list of provided dtypes - * \param provided_arg_dtypes data of provided dtypes - * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types - * \param provided_arg_stype_names argument name list of provided storage types - * \param provided_arg_stypes data of provided storage types - * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec - * \param shared_arg_name_list parameter name list passed from _bind_ith_exec - * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec - * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec - * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec - * \param updated_shared_buffer_name_list updated shared data array names after binding - * \param updated_shared_buffer_handle_list updated shared data arrays after binding - * \param num_in_args number of input arguments of this sym - * \param in_args list_arguments associated with the current executor - * \param arg_grads list of gradients of in_args associated with the current executor - * \param num_aux_states number of aux states of this sym - * \param aux_states list_auxiliary_states associated with the current executor - * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec - * \param out the handle of the executor to be created - */ -int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + +namespace mxnet { + +template +int SimpleBindExMaster(SymbolHandle symbol_handle, Review comment: internal functions in C API generally begin with _ This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert
anirudh2290 commented on a change in pull request #16585: C Api for simplebind, fix comment for trigoops, add atol to assert URL: https://github.com/apache/incubator-mxnet/pull/16585#discussion_r338282191 ## File path: src/c_api/c_api_executor.cc ## @@ -849,6 +816,152 @@ int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, API_END(); } +} // namespace mxnet + + +/*! + * \brief + * \param symbol_handle symbol handle + * \param dev_type default device type + * \param dev_id default device id + * \param num_g2c_keys number of group2ctx keys + * \param g2c_keys key list of group2ctx + * \param g2c_dev_types device type list of group2ctx + * \param g2c_dev_ids id list of group2ctx + * \param provided_grad_req_list_len grad_req length provided by users in front-end + * \param provided_grad_req_names grad_req names provided by users in front-end + * \param provided_grad_req_types req types provided by users in front-end + * \param num_provided_arg_shapes number of user provided in_arg and aux_state shapes + * \param provided_arg_shape_names name list of provided shapes + * \param provided_arg_shape_data provided shape data + * \param provided_arg_shape_idx provided shape data index + * \param num_provided_arg_dtypes number of user provided in_arg and axu_state dtypes + * \param provided_arg_dtype_names argument name list of provided dtypes + * \param provided_arg_dtypes data of provided dtypes + * \param num_provided_arg_stypes number of user provided in_arg and axu_state storage types + * \param provided_arg_stype_names argument name list of provided storage types + * \param provided_arg_stypes data of provided storage types + * \param num_shared_arg_names number of parameter names passed from _bind_ith_exec + * \param shared_arg_name_list parameter name list passed from _bind_ith_exec + * \param shared_buffer_len number of shared data arrays passed from _bind_ith_exec + * \param shared_buffer_name_list shared data array names passed from _bind_ith_exec + * \param shared_buffer_handle_list shared data array handles passed from _bind_ith_exec + * \param updated_shared_buffer_name_list updated shared data array names after binding + * \param updated_shared_buffer_handle_list updated shared data arrays after binding + * \param num_in_args number of input arguments of this sym + * \param in_args list_arguments associated with the current executor + * \param arg_grads list of gradients of in_args associated with the current executor + * \param num_aux_states number of aux states of this sym + * \param aux_states list_auxiliary_states associated with the current executor + * \param shared_exec_handle shared excutor handle passed from _bind_ith_exec + * \param out the handle of the executor to be created + */ +int MXExecutorSimpleBindEx(SymbolHandle symbol_handle, + int dev_type, + int dev_id, + const uint32_t num_g2c_keys, + const char** g2c_keys, + const int* g2c_dev_types, + const int* g2c_dev_ids, + const uint32_t provided_grad_req_list_len, + const char** provided_grad_req_names, + const char** provided_grad_req_types, + const uint32_t num_provided_arg_shapes, + const char** provided_arg_shape_names, + const int* provided_arg_shape_data, + const uint32_t* provided_arg_shape_idx, + const uint32_t num_provided_arg_dtypes, + const char** provided_arg_dtype_names, + const int* provided_arg_dtypes, + const uint32_t num_provided_arg_stypes, + const char** provided_arg_stype_names, + const int* provided_arg_stypes, + const uint32_t num_shared_arg_names, + const char** shared_arg_name_list, + int* shared_buffer_len, + const char** shared_buffer_name_list, + NDArrayHandle* shared_buffer_handle_list, + const char*** updated_shared_buffer_name_list, + NDArrayHandle** updated_shared_buffer_handle_list, + uint32_t* num_in_args, + NDArrayHandle** in_args, + NDArrayHandle** arg_grads, + uint32_t* num_aux_states, + NDArrayHandle** aux_states, + ExecutorHandle shared_exec_handle, + ExecutorHandle* out) { + return mxnet::SimpleBindExMaster(symbol_handle, +dev_type, dev_id, +num_g2c_keys, g
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338282712 ## File path: src/operator/numpy/np_matrix_op-inl.h ## @@ -71,6 +79,80 @@ void NumpyTranspose(const nnvm::NodeAttrs& attrs, } } +template +void NumpyColumnStackForward(const nnvm::NodeAttrs& attrs, +const OpContext& ctx, +const std::vector& inputs, +const std::vector& req, +const std::vector& outputs) { Review comment: Same applies to all other function signatures. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338276663 ## File path: src/operator/numpy/np_matrix_op-inl.h ## @@ -71,6 +79,80 @@ void NumpyTranspose(const nnvm::NodeAttrs& attrs, } } +template +void NumpyColumnStackForward(const nnvm::NodeAttrs& attrs, +const OpContext& ctx, +const std::vector& inputs, +const std::vector& req, +const std::vector& outputs) { Review comment: Alignment: ```c++ void NumpyColumnStackForward(const nnvm::NodeAttrs& attrs, const OpContext& ctx, const std::vector& inputs, const std::vector& req, const std::vector& outputs) { ``` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338276320 ## File path: python/mxnet/numpy_dispatch_protocol.py ## @@ -119,6 +119,7 @@ def _run_with_array_ufunc_proto(*args, **kwargs): 'var', 'vdot', 'vstack', +# 'column_stack', Review comment: Why is this commented out? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338276158 ## File path: python/mxnet/numpy/multiarray.py ## @@ -55,7 +55,7 @@ 'swapaxes', 'clip', 'argmax', 'std', 'var', 'indices', 'copysign', 'ravel', 'hanning', 'hamming', 'blackman', 'flip', 'around', 'arctan2', 'hypot', 'rad2deg', 'deg2rad', 'unique', 'lcm', 'tril', 'identity', 'take', 'ldexp', 'vdot', 'inner', 'outer', 'equal', 'not_equal', 'greater', 'less', - 'greater_equal', 'less_equal', 'hsplit', 'rot90', 'einsum', 'true_divide'] + 'greater_equal', 'less_equal', 'hsplit', 'rot90', 'einsum', 'true_divide', 'column_stack'] Review comment: Same comment on the positions of `column_stack`. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #16566: [CI test failure] test_fast_lars fails on windows gpu
ptrendx commented on issue #16566: [CI test failure] test_fast_lars fails on windows gpu URL: https://github.com/apache/incubator-mxnet/issues/16566#issuecomment-545631852 @Caenorst This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338272478 ## File path: python/mxnet/ndarray/numpy/_op.py ## @@ -4761,3 +4760,39 @@ def einsum(*operands, **kwargs): subscripts = operands[0] operands = operands[1:] return _npi.einsum(*operands, subscripts=subscripts, out=out, optimize=int(optimize_arg)) + Review comment: 2 blank lines between all Python functions. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack
haojin2 commented on a change in pull request #16594: [Numpy] implement np.column_stack URL: https://github.com/apache/incubator-mxnet/pull/16594#discussion_r338272300 ## File path: python/mxnet/ndarray/numpy/_op.py ## @@ -38,8 +38,7 @@ 'std', 'var', 'indices', 'copysign', 'ravel', 'hanning', 'hamming', 'blackman', 'flip', 'around', 'hypot', 'rad2deg', 'deg2rad', 'unique', 'lcm', 'tril', 'identity', 'take', 'ldexp', 'vdot', 'inner', 'outer', 'equal', 'not_equal', 'greater', 'less', 'greater_equal', 'less_equal', - 'hsplit', 'rot90', 'einsum', 'true_divide'] - + 'hsplit', 'rot90', 'einsum', 'true_divide', 'column_stack'] Review comment: Please move 'column_stack' to after `vstack` in this list. Also please move the function definition to after `vstack` as well in the file. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on issue #16587: How could I set the location of openblas/lapack when I compile mxnet from source?
haojin2 commented on issue #16587: How could I set the location of openblas/lapack when I compile mxnet from source? URL: https://github.com/apache/incubator-mxnet/issues/16587#issuecomment-545629395 Does the CMake run without explicitly specifying your openblas and lapack locations? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 edited a comment on issue #16596: How to initialize a CPU tensor in custom cu file?
haojin2 edited a comment on issue #16596: How to initialize a CPU tensor in custom cu file? URL: https://github.com/apache/incubator-mxnet/issues/16596#issuecomment-545628138 @vasusingla619 If you're asking how to create a temporary CPU memory space in a .cu file, you can simply do this: ```c++ void YourFunction(...) { // Your code vector temp_buffer(workspace_size * 5, 0); Tensor workspace(temp_buffer.data(), Shape1(workspace * 5)); // Your code } ``` Lemme know if this solves your question. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 commented on issue #16596: How to initialize a CPU tensor in custom cu file?
haojin2 commented on issue #16596: How to initialize a CPU tensor in custom cu file? URL: https://github.com/apache/incubator-mxnet/issues/16596#issuecomment-545628138 @vasusingla619 If you're asking how to create a temporary CPU memory space in a .cu file, you can simply do this: ```c++ void YourFunction(...) { vector temp_buffer(workspace_size * 5, 0); Tensor workspace(temp_buffer.data(), Shape1(workspace * 5)); } ``` Lemme know if this solves your question. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services