[GitHub] [incubator-mxnet] leezu merged pull request #18319: [numpy] symbolic advanced indexing
leezu merged pull request #18319: URL: https://github.com/apache/incubator-mxnet/pull/18319 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18724: Initialize docker cache in build.py for docker-compose containers
mxnet-bot commented on pull request #18724: URL: https://github.com/apache/incubator-mxnet/pull/18724#issuecomment-659202110 Jenkins CI successfully triggered : [centos-gpu, unix-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
[GitHub] [incubator-mxnet] leezu commented on pull request #18724: Initialize docker cache in build.py for docker-compose containers
leezu commented on pull request #18724: URL: https://github.com/apache/incubator-mxnet/pull/18724#issuecomment-659202077 @mxnet-bot run ci [unix-cpu, centos-gpu] 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
[GitHub] [incubator-mxnet] leezu commented on pull request #18706: [1.x] Revert Fix memory leaks in Gluon in 1.x branch
leezu commented on pull request #18706: URL: https://github.com/apache/incubator-mxnet/pull/18706#issuecomment-659202436 The issue does not apply to master. 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
[incubator-mxnet] branch master updated (6901325 -> 8198442)
This is an automated email from the ASF dual-hosted git repository. lausen pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 6901325 Add the newest mxnet discuss version. Add d2l.ai (#18663) add 8198442 [numpy] symbolic advanced indexing (#18319) No new revisions were added by this update. Summary of changes: python/mxnet/numpy/multiarray.py | 8 + python/mxnet/symbol/numpy/_symbol.py | 19 +- src/operator/numpy/np_indexing_op.cc | 551 +++ src/operator/numpy/np_indexing_op.cu | 452 ++ src/operator/numpy/np_indexing_op.h | 130 +++ tests/python/unittest/test_gluon_indexing.py | 104 + 6 files changed, 1262 insertions(+), 2 deletions(-) create mode 100644 src/operator/numpy/np_indexing_op.cc create mode 100644 src/operator/numpy/np_indexing_op.cu create mode 100644 src/operator/numpy/np_indexing_op.h create mode 100644 tests/python/unittest/test_gluon_indexing.py
[incubator-mxnet] branch master updated (6901325 -> 8198442)
This is an automated email from the ASF dual-hosted git repository. lausen pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 6901325 Add the newest mxnet discuss version. Add d2l.ai (#18663) add 8198442 [numpy] symbolic advanced indexing (#18319) No new revisions were added by this update. Summary of changes: python/mxnet/numpy/multiarray.py | 8 + python/mxnet/symbol/numpy/_symbol.py | 19 +- src/operator/numpy/np_indexing_op.cc | 551 +++ src/operator/numpy/np_indexing_op.cu | 452 ++ src/operator/numpy/np_indexing_op.h | 130 +++ tests/python/unittest/test_gluon_indexing.py | 104 + 6 files changed, 1262 insertions(+), 2 deletions(-) create mode 100644 src/operator/numpy/np_indexing_op.cc create mode 100644 src/operator/numpy/np_indexing_op.cu create mode 100644 src/operator/numpy/np_indexing_op.h create mode 100644 tests/python/unittest/test_gluon_indexing.py
[GitHub] [incubator-mxnet] leezu commented on issue #18210: Github Actions: OS X jobs canceled
leezu commented on issue #18210: URL: https://github.com/apache/incubator-mxnet/issues/18210#issuecomment-659202975 Github needs to fix 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
[GitHub] [incubator-mxnet] szha commented on pull request #18706: [1.x] Revert Fix memory leaks in Gluon in 1.x branch
szha commented on pull request #18706: URL: https://github.com/apache/incubator-mxnet/pull/18706#issuecomment-659214645 Oh right, I just realized that _BlockScope is gone. 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
svn commit: r40510 - in /dev/incubator/mxnet/1.7.0.rc1: ./ apache-mxnet-src-1.7.0.rc1-incubating.tar.gz apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.asc apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.sh
Author: taolv Date: Thu Jul 16 08:15:47 2020 New Revision: 40510 Log: Add mxnet-1.7.0.rc1 Added: dev/incubator/mxnet/1.7.0.rc1/ dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz (with props) dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.asc (with props) dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.sha512 Added: dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz == Binary file - no diff available. Propchange: dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz -- svn:mime-type = application/x-gzip Added: dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.asc == Binary file - no diff available. Propchange: dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.asc -- svn:mime-type = application/pgp-signature Added: dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.sha512 == --- dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.sha512 (added) +++ dev/incubator/mxnet/1.7.0.rc1/apache-mxnet-src-1.7.0.rc1-incubating.tar.gz.sha512 Thu Jul 16 08:15:47 2020 @@ -0,0 +1 @@ +ab996843374e7ccdf26ee875eacf639c1f30895df8b5d2c1830cbbfee6d097af0ee05279165845a2e15009f23d8401734129b4a46ef12acee39ea738e3c08032 apache-mxnet-src-1.7.0.rc1-incubating.tar.gz
[GitHub] [incubator-mxnet] wkcn commented on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn commented on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659286009 @szha Thank you for the suggestion! When axes.ndim() <= 6, it will call the original function `TransposeImpl`. `TransposeImpl` does not allocate any workspace, and its speed is consistent with the original `transpose`. I will use the profiler on `transpose` and `rollaxis` to test `axes.ndim() <= 6 and axes.ndim() > 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
[GitHub] [incubator-mxnet] wkcn commented on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn commented on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, axes is generated randomly ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. I will try to optimize it. Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` 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
[GitHub] [incubator-mxnet] wkcn edited a comment on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn edited a comment on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, axes is generated randomly ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. I will try to optimize it. If axes is monotonically increasing (namely [0, 1, 2, 3, ..., ndim - 1]), ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1492 2|12582.9121|1.1732 3|12582.9121|1.3264 4|12582.9121|1.3896 5|12582.9121|0.9107 6|12582.9121|0.8965 7|12583.3604|0.9028 8|12583.4238|0.9105 9|12583.4883|0.8981 Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` 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
[GitHub] [incubator-mxnet] wkcn edited a comment on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn edited a comment on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, axes is generated randomly ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. I will try to optimize it. If axes is monotonically increasing (namely [0, 1, 2, 3, ..., ndim - 1]) (comment the line 21st `random.shuffle(axes)`), ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1492 2|12582.9121|1.1732 3|12582.9121|1.3264 4|12582.9121|1.3896 5|12582.9121|0.9107 6|12582.9121|0.8965 7|12583.3604|0.9028 8|12583.4238|0.9105 9|12583.4883|0.8981 Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` 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
[GitHub] [incubator-mxnet] wkcn edited a comment on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn edited a comment on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, **axes is generated randomly** ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. If **axes is monotonically increasing** (namely [0, 1, 2, 3, ..., ndim - 1]) (comment the line 21st `random.shuffle(axes)`), ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1492 2|12582.9121|1.1732 3|12582.9121|1.3264 4|12582.9121|1.3896 5|12582.9121|0.9107 6|12582.9121|0.8965 7|12583.3604|0.9028 8|12583.4238|0.9105 9|12583.4883|0.8981 Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` 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
[GitHub] [incubator-mxnet] wkcn edited a comment on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn edited a comment on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, **axes is generated randomly** ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. If **axes is monotonically increasing** (namely [0, 1, 2, 3, ..., ndim - 1]) (comment the line 21st `random.shuffle(axes)`), ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1492 2|12582.9121|1.1732 3|12582.9121|1.3264 4|12582.9121|1.3896 5|12582.9121|0.9107 6|12582.9121|0.8965 7|12583.3604|0.9028 8|12583.4238|0.9105 9|12583.4883|0.8981 If **axes is monotonically decreasing** (namely [ndim - 1, ndim -2, ..., 2, 1, 0]) ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1290 2|12582.9121|1.1204 3|12582.9121|1.1874 4|12582.9121|1.4240 5|12582.9121|7.7080 6|12582.9121|24.0448 7|12583.3604|115.1126 8|12583.4238|105.9091 9|12583.4883|106.3913 Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` 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
[GitHub] [incubator-mxnet] wkcn edited a comment on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn edited a comment on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659304799 Performance Benchmark: transpose operator on CPU, **axes is generated randomly** ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.0786 2|12582.9121|1.0851 3|12582.9121|0.6763 4|12582.9121|1.2172 5|12582.9121|6.4305 6|12582.9121|11.7841 7|12583.3604|65.7184 8|12583.4238|65.2171 9|12583.4883|82.4930 The increase of memory footprint is slight, but the time is intolerable when `axes.ndim() > 6`. If **axes is monotonically increasing** (namely [0, 1, 2, 3, ..., ndim - 1]) (comment the line 21st `random.shuffle(axes)`), ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1492 2|12582.9121|1.1732 3|12582.9121|1.3264 4|12582.9121|1.3896 5|12582.9121|0.9107 6|12582.9121|0.8965 7|12583.3604|0.9028 8|12583.4238|0.9105 9|12583.4883|0.8981 If **axes is monotonically decreasing** (namely [ndim - 1, ndim -2, ..., 2, 1, 0]) ndim | max use (kb) | avg time (ms) ---|---|-- 1|12582.9121|1.1290 2|12582.9121|1.1204 3|12582.9121|1.1874 4|12582.9121|1.4240 5|12582.9121|7.7080 6|12582.9121|24.0448 7|12583.3604|115.1126 8|12583.4238|105.9091 9|12583.4883|106.3913 Compare with NumPy Transpose: ndim | numpy time (s) | mxnet time (s) --|--|-- 1 | 0.1621077060699463 | 0.31803297996520996 2 | 0.2637207508087158 | 0.33347415924072266 3 | 0.4311816692352295 | 0.47667574882507324 4 | 0.5303101539611816 | 0.49021244049072266 5 | 0.5940566062927246 | 1.48443603515625 6 | 0.8220541477203369 | 2.03752064704895 7 | 0.8727006912231445 | 9.488046169281006 8 | 1.0004301071166992 | 9.947605848312378 9 | 1.2341070175170898 | 12.262272119522095 Test Code: ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) #configure the profiler profiler.set_config(profile_all=True, aggregate_stats=True, filename='trace_profile.json') #start the profiler collecting data def test_transpose(ndim): for t in range(20): dims = [4 for _ in range(ndim)] dims[-1] *= 4 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x = mx.nd.array(np.random.normal(size=dims)) y = mx.nd.transpose(x, axes=axes) assert_allclose(np.transpose(x.asnumpy(), axes=axes), y.asnumpy()) for ndim in range(1, 10): profiler.set_state('run') tic = time.time() test_transpose(ndim) print(ndim, "", time.time() - tic) #stop the profiler profiler.set_state('stop') #dump the profiling data as a string print(profiler.dumps(reset=True)) print("Over") ``` Test Code, compare with NumPy ```python import mxnet as mx from mxnet import profiler print(mx) import numpy as np from numpy.testing import assert_allclose import time import random seed = 42 np.random.seed(seed) mx.random.seed(seed) def test_transpose(ndim): np_time = 0 mx_time = 0 for t in range(20): dims = [5 for _ in range(ndim)] dims[-1] *= 5 ** (10 - ndim) axes = list(range(ndim)) random.shuffle(axes) axes = tuple(axes) x_np = np.array(np.random.normal(size=dims), dtype=np.float32) x_mx = mx.nd.array(x_np, dtype=np.float32) for _ in range(2): y_np = np.transpose(x_np, axes=axes).copy() y_mx = mx.nd.transpose(x_mx, axes=axes) y_mx.asnumpy() tic_np = time.time() for _ in range(1): y_np = np.transpose(x_np, axes=axes).copy() np_time += time.time() - tic_np tic_mx = time.time() for _ in range(1): y_mx = mx.nd.transpose(x_mx, axes=axes) y_mx.asnumpy() mx_time += time.time() - tic_mx print(f"{ndim} | {np_time} | {mx_time}") for ndim in range(1, 10): test_transpose(ndim) print("Over") ``` 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
[GitHub] [incubator-mxnet] Qengineering commented on issue #18536: MXNET build not working on armv7 (raspberry pi 4)
Qengineering commented on issue #18536: URL: https://github.com/apache/incubator-mxnet/issues/18536#issuecomment-659370080 I've had the same problem with the memory exhausting. It has to do with GNU compiler. The compiler is very memory hungry, especially when long chains off template code is involved. Even with 8 GByte of RAM, the compilation crashes. The solution is using the Clang compilers temporarily when installing MXNet on a 32-bits Raspberry Pi OS. The 64-bit OS didn't suffer from memory shortage. For a detailed discussion please look at [https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html] 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
[GitHub] [incubator-mxnet] Qengineering edited a comment on issue #18536: MXNET build not working on armv7 (raspberry pi 4)
Qengineering edited a comment on issue #18536: URL: https://github.com/apache/incubator-mxnet/issues/18536#issuecomment-659370080 I've had the same problem with the memory exhausting. It has to do with GNU compiler. The compiler is very memory hungry, especially when long chains off template code is involved. Even with 8 GByte of RAM, the compilation crashes. The solution is using the Clang compilers temporarily when installing MXNet on a 32-bits Raspberry Pi OS. The 64-bit OS didn't suffer from memory shortage. For a detailed discussion please look at [https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html](https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html]) 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
[GitHub] [incubator-mxnet] Qengineering edited a comment on issue #18536: MXNET build not working on armv7 (raspberry pi 4)
Qengineering edited a comment on issue #18536: URL: https://github.com/apache/incubator-mxnet/issues/18536#issuecomment-659370080 I've had the same problem with the memory exhausting. It has to do with GNU compiler. The compiler is very memory hungry, especially when long chains off template code are involved. Even with 8 GByte of RAM, the compilation crashes. The solution is using the Clang compilers temporarily when installing MXNet on a 32-bits Raspberry Pi OS. The 64-bit OS didn't suffer from memory shortage. For a detailed discussion please look at [https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html](https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html]) 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
[GitHub] [incubator-mxnet] Qengineering commented on issue #18628: Raspberry pi 4 compile failure
Qengineering commented on issue #18628: URL: https://github.com/apache/incubator-mxnet/issues/18628#issuecomment-659372967 I've built the latest version of MXNet (2.0.0) on a Raspberry Pi with 32 bit and 64 bit OS successfully. The instruction can be found here [](https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html) 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
[GitHub] [incubator-mxnet] Qengineering edited a comment on issue #18628: Raspberry pi 4 compile failure
Qengineering edited a comment on issue #18628: URL: https://github.com/apache/incubator-mxnet/issues/18628#issuecomment-659372967 I've built the latest version of MXNet (2.0.0) on a Raspberry Pi with 32 bit and 64 bit OS successfully. The instruction can be found here [https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html](https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html) 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
[GitHub] [incubator-mxnet] Qengineering edited a comment on issue #18628: Raspberry pi 4 compile failure
Qengineering edited a comment on issue #18628: URL: https://github.com/apache/incubator-mxnet/issues/18628#issuecomment-659372967 I've built the latest version of MXNet (2.0.0) on a Raspberry Pi with 32 bit and 64 bit OS successfully. The instructions can be found here [install-mxnet-on-raspberry-pi-4](https://qengineering.eu/install-mxnet-on-raspberry-pi-4.html) 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
[incubator-mxnet-site] branch asf-site updated: Publish triggered by CI
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 1b26cac Publish triggered by CI 1b26cac is described below commit 1b26cac03f0df20349409485af1463cfc5eba547 Author: mxnet-ci AuthorDate: Thu Jul 16 12:41:48 2020 + Publish triggered by CI --- date.txt | 1 - feed.xml | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/date.txt b/date.txt deleted file mode 100644 index b0caac6..000 --- a/date.txt +++ /dev/null @@ -1 +0,0 @@ -Thu Jul 16 06:42:48 UTC 2020 diff --git a/feed.xml b/feed.xml index 6be483c..38ec4ff 100644 --- a/feed.xml +++ b/feed.xml @@ -1 +1 @@ -http://www.w3.org/2005/Atom"; >https://jekyllrb.com/"; version="4.0.0">Jekyllhttps://mxnet.apache.org/feed.xml"; rel="self" type="application/atom+xml" />https://mxnet.apache.org/"; rel="alternate" type="text/html" />2020-07-16T06:29:57+00:00https://mxnet.apache.org/feed.xmlApache MXNetA flexible and efficient library for deep [...] \ No newline at end of file +http://www.w3.org/2005/Atom"; >https://jekyllrb.com/"; version="4.0.0">Jekyllhttps://mxnet.apache.org/feed.xml"; rel="self" type="application/atom+xml" />https://mxnet.apache.org/"; rel="alternate" type="text/html" />2020-07-16T12:30:05+00:00https://mxnet.apache.org/feed.xmlApache MXNetA flexible and efficient library for deep [...] \ No newline at end of file
[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 444aad1 Bump the publish timestamp. 444aad1 is described below commit 444aad1a352542df574cbde30d7af24beaaafc1a Author: mxnet-ci AuthorDate: Thu Jul 16 12:41:54 2020 + 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..d695daf --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Thu Jul 16 12:41:54 UTC 2020
[GitHub] [incubator-mxnet] lilipj commented on issue #17395: Keras model to ONNX to mxnet error
lilipj commented on issue #17395: URL: https://github.com/apache/incubator-mxnet/issues/17395#issuecomment-659387413 I've used the onnxmltool.convert_keras function, it works with target_opset=8 (9 fails) 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
[GitHub] [incubator-mxnet] woreom commented on issue #18536: MXNET build not working on armv7 (raspberry pi 4)
woreom commented on issue #18536: URL: https://github.com/apache/incubator-mxnet/issues/18536#issuecomment-659391586 @Qengineering Thank you for your response, It seems that the CC compiler is the issue however I used GCC-4.9 to overcome it, I wanted to use OpenCV so I had to compile OpenCV with GCC-4.9, too (otherwise it won't work) also the latest python supported by raspbian is python3.7.3 which does not go well with mxnet and OpenCV, so I build python-3.7.4, again with GCC-4.9. I look around your post quit frequently they are really helpful, I can't believe I've missed this one and inflicted so much pain on myself :))) 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
[GitHub] [incubator-mxnet] woreom closed issue #18536: MXNET build not working on armv7 (raspberry pi 4)
woreom closed issue #18536: URL: https://github.com/apache/incubator-mxnet/issues/18536 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
[GitHub] [incubator-mxnet] chinakook commented on issue #18543: Gradient checkpointing in the Gluon interface
chinakook commented on issue #18543: URL: https://github.com/apache/incubator-mxnet/issues/18543#issuecomment-659456175 It's an urgent job because the MXNet 2.0 delete the symbolic executor API. 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
[GitHub] [incubator-mxnet] Masquerade0097 edited a comment on issue #18667: MXNet serialization format depends on endianness
Masquerade0097 edited a comment on issue #18667: URL: https://github.com/apache/incubator-mxnet/issues/18667#issuecomment-655500718 @leezu Thanks for your reply. > For now, you can't use the pretrained models directly on s390x architecture but you'd need some workaround where you load the parameter on a x86 machine, call asnumpy, save via numpy and finally load the parameters on s390x machine using numpy and convert them to mxnet. Would that work for you? I'll give it a try. 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
[GitHub] [incubator-mxnet] szha commented on issue #18543: Gradient checkpointing in the Gluon interface
szha commented on issue #18543: URL: https://github.com/apache/incubator-mxnet/issues/18543#issuecomment-659493480 @eric-haibin-lin @ArmageddonKnight could you update on this? 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
[GitHub] [incubator-mxnet] ptrendx commented on issue #18729: TensorRT tutorial is missing in master
ptrendx commented on issue #18729: URL: https://github.com/apache/incubator-mxnet/issues/18729#issuecomment-659506296 @Kh4L Could you check if the content from 1.x is ok / modify the example to work with the current master? @aaronmarkham You gave link to 1.6 version, which seems to be there, the one that does not work is the master equivalent: https://mxnet.apache.org/api/python/docs/tutorials/performance/backend/tensorrt/tensorrt 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
[GitHub] [incubator-mxnet] DickJC123 commented on pull request #18711: Fix (log_)softmax backward on empty ndarray
DickJC123 commented on pull request #18711: URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659516305 My PR is having the same 'worker crash' problem this PR is having. Without understanding the root cause, I tried marking the failing test in unittest/test_sparse_operator.py as serial, and I got no similar failures. So you might give this a try: ``` @with_seed() @pytest.mark.serial def test_elemwise_binary_ops(): ``` 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
[GitHub] [incubator-mxnet] DickJC123 edited a comment on pull request #18711: Fix (log_)softmax backward on empty ndarray
DickJC123 edited a comment on pull request #18711: URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659516305 My PR is having the same 'worker crash' problem this PR is having. Without understanding the root cause, I tried marking the failing test in unittest/test_sparse_operator.py as serial, and I got no similar failures. So you might give this a try: ``` @with_seed() @pytest.mark.serial def test_elemwise_binary_ops(): ``` 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
[GitHub] [incubator-mxnet] szha commented on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
szha commented on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659556617 the comparison with numpy is unfair in that numpy just sets strides and doesn't have to be c contiguous. 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
[GitHub] [incubator-mxnet] szha closed issue #18695: Transpose only supports up to 6 dimensions
szha closed issue #18695: URL: https://github.com/apache/incubator-mxnet/issues/18695 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
[GitHub] [incubator-mxnet] szha merged pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
szha merged pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707 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
[incubator-mxnet] branch master updated: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis (#18707)
This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 37bdf0b [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis (#18707) 37bdf0b is described below commit 37bdf0bf981d11a89bd248b02f473211d57bc9c6 Author: JackieWu AuthorDate: Fri Jul 17 01:25:01 2020 +0800 [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis (#18707) * support 6+ dims for transpose * test over * reorder code * fix transposeex --- src/operator/numpy/np_matrix_op-inl.h | 51 src/operator/numpy/np_matrix_op.cc | 17 +++- src/operator/tensor/matrix_op-inl.h| 138 +++-- src/operator/tensor/matrix_op.cc | 4 + tests/python/unittest/test_numpy_op.py | 8 +- tests/python/unittest/test_operator.py | 4 +- 6 files changed, 191 insertions(+), 31 deletions(-) diff --git a/src/operator/numpy/np_matrix_op-inl.h b/src/operator/numpy/np_matrix_op-inl.h index 0125feb..0fea76b 100644 --- a/src/operator/numpy/np_matrix_op-inl.h +++ b/src/operator/numpy/np_matrix_op-inl.h @@ -134,10 +134,10 @@ void NumpyTranspose(const nnvm::NodeAttrs& attrs, const std::vector& inputs, const std::vector& req, const std::vector& outputs) { - const NumpyTransposeParam& param = nnvm::get(attrs.parsed); if (req[0] == kNullOp) return; CHECK(req[0] == kWriteTo || req[0] == kAddTo) - << "Transpose only supports kWriteTo, kNullOp and kAddTo"; + << "Transpose does not support inplace"; + const NumpyTransposeParam& param = nnvm::get(attrs.parsed); mxnet::TShape axes; if (ndim_is_known(param.axes)) { axes = common::CanonicalizeAxes(param.axes); @@ -147,10 +147,14 @@ void NumpyTranspose(const nnvm::NodeAttrs& attrs, axes[i] = axes.ndim() - 1 - i; } } + mshadow::Tensor workspace = +GetTransposeExWorkspace(ctx, axes); if (req[0] == kAddTo) { -TransposeImpl(ctx.run_ctx, inputs[0], outputs[0], axes); +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); } else { -TransposeImpl(ctx.run_ctx, inputs[0], outputs[0], axes); +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); } } @@ -779,13 +783,21 @@ void NumpyRollaxisCompute(const nnvm::NodeAttrs& attrs, using namespace mshadow::expr; CHECK_EQ(inputs.size(), 1U); CHECK_EQ(outputs.size(), 1U); - CHECK_EQ(req[0], kWriteTo) << "Rollaxis does not support inplace"; - mxnet::TShape axes; + if (req[0] == kNullOp) return; + CHECK(req[0] == kWriteTo || req[0] == kAddTo) + << "Rollaxis does not support inplace"; const NumpyRollaxisParam& param = nnvm::get(attrs.parsed); - axes = NumpyRollaxisShapeImpl(param.axis, param.start, inputs[0].ndim()); - MSHADOW_TYPE_SWITCH(outputs[0].type_flag_, Dtype, { -TransposeImpl(ctx.run_ctx, inputs[0], outputs[0], axes); - }) + mxnet::TShape axes = NumpyRollaxisShapeImpl(param.axis, param.start, inputs[0].ndim()); + + mshadow::Tensor workspace = +GetTransposeExWorkspace(ctx, axes); + if (req[0] == kAddTo) { +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); + } else { +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); + } } template @@ -796,6 +808,9 @@ void NumpyRollaxisBackward(const nnvm::NodeAttrs &attrs, const std::vector &outputs) { using namespace mshadow; using namespace mshadow::expr; + if (req[0] == kNullOp) return; + CHECK(req[0] == kWriteTo || req[0] == kAddTo) + << "Rollaxis Backward does not support inplace"; const NumpyRollaxisParam& param = nnvm::get(attrs.parsed); int axis_origin = param.axis; int start_origin = param.start; @@ -819,11 +834,17 @@ void NumpyRollaxisBackward(const nnvm::NodeAttrs &attrs, axis = start_origin; start = axis_origin + 1; } - mxnet::TShape axes; - axes = NumpyRollaxisShapeImpl(axis, start, inputs[0].ndim()); - MSHADOW_TYPE_SWITCH(outputs[0].type_flag_, Dtype, { -TransposeImpl(ctx.run_ctx, inputs[0], outputs[0], axes); - }) + mxnet::TShape axes = NumpyRollaxisShapeImpl(axis, start, inputs[0].ndim()); + + mshadow::Tensor workspace = +GetTransposeExWorkspace(ctx, axes); + if (req[0] == kAddTo) { +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); + } else { +TransposeExImpl(ctx.run_ctx, inputs[0], outputs[0], +axes, workspace); + } } struct NumpyRot90Param : public dmlc::Parameter { diff --git a/src/operator/numpy/np_matrix_op.cc b/src/operator/numpy/np_matrix_op.cc index da9839f..2bb2fe3 100644 --- a/src/operator/numpy/np_matrix_op.cc +++ b/src/operator/numpy/
[incubator-mxnet] branch master updated (8198442 -> 37bdf0b)
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 8198442 [numpy] symbolic advanced indexing (#18319) add 37bdf0b [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis (#18707) No new revisions were added by this update. Summary of changes: src/operator/numpy/np_matrix_op-inl.h | 51 src/operator/numpy/np_matrix_op.cc | 17 +++- src/operator/tensor/matrix_op-inl.h| 138 +++-- src/operator/tensor/matrix_op.cc | 4 + tests/python/unittest/test_numpy_op.py | 8 +- tests/python/unittest/test_operator.py | 4 +- 6 files changed, 191 insertions(+), 31 deletions(-)
[GitHub] [incubator-mxnet] leezu merged pull request #18724: Initialize docker cache in build.py for docker-compose containers
leezu merged pull request #18724: URL: https://github.com/apache/incubator-mxnet/pull/18724 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
[incubator-mxnet] branch master updated (37bdf0b -> 2abf0b8)
This is an automated email from the ASF dual-hosted git repository. lausen pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 37bdf0b [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis (#18707) add 2abf0b8 Initialize docker cache in build.py for docker-compose containers (#18724) No new revisions were added by this update. Summary of changes: ci/build.py | 14 +++--- 1 file changed, 11 insertions(+), 3 deletions(-)
[GitHub] [incubator-mxnet] leezu opened a new issue #18732: Website build is untested
leezu opened a new issue #18732: URL: https://github.com/apache/incubator-mxnet/issues/18732 Website is built from master branch; but there is no PR check ensuring that a PR does not break the website build. Thus website breaks frequently - https://github.com/apache/incubator-mxnet/issues/18680 - https://github.com/apache/incubator-mxnet/issues/18725 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18733: Move gluon.metric api docs
mxnet-bot commented on pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733#issuecomment-659573656 Hey @leezu , Thanks for submitting the PR All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: - To trigger all jobs: @mxnet-bot run ci [all] - To trigger specific jobs: @mxnet-bot run ci [job1, job2] *** **CI supported jobs**: [sanity, centos-cpu, centos-gpu, website, clang, windows-gpu, windows-cpu, unix-cpu, miscellaneous, edge, unix-gpu] *** _Note_: Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. All CI tests must pass before the PR can be merged. 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
[GitHub] [incubator-mxnet] leezu opened a new pull request #18733: Move gluon.metric api docs
leezu opened a new pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733 Fixes https://github.com/apache/incubator-mxnet/issues/18725 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
[GitHub] [incubator-mxnet] leezu commented on issue #18725: metric.py change between branches - missing docs in master
leezu commented on issue #18725: URL: https://github.com/apache/incubator-mxnet/issues/18725#issuecomment-659573454 The module should be documented in the same way as any other module on the website. The issue here is that the website build is not tested in PRs and thus it's no surprise to find various parts of the build break. Can we prioritize fixing the root-cause? https://github.com/apache/incubator-mxnet/issues/18732 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
[GitHub] [incubator-mxnet] leezu commented on a change in pull request #18721: Remove Makefile
leezu commented on a change in pull request #18721: URL: https://github.com/apache/incubator-mxnet/pull/18721#discussion_r455972369 ## File path: Makefile ## @@ -1,768 +0,0 @@ -# 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. - -ROOTDIR = $(CURDIR) -TPARTYDIR = $(ROOTDIR)/3rdparty - -ifeq ($(OS),Windows_NT) - UNAME_S := Windows -else - UNAME_S := $(shell uname -s) - UNAME_P := $(shell uname -p) -endif - -ifndef config -ifdef CXXNET_CONFIG - config = $(CXXNET_CONFIG) -else ifneq ("$(wildcard ./config.mk)","") - config = config.mk -else - config = make/config.mk -endif -endif - -ifndef DMLC_CORE - DMLC_CORE = $(TPARTYDIR)/dmlc-core -endif -CORE_INC = $(wildcard $(DMLC_CORE)/include/*/*.h) - -ifndef NNVM_PATH - NNVM_PATH = $(TPARTYDIR)/tvm/nnvm -endif - -ifndef DLPACK_PATH - DLPACK_PATH = $(ROOTDIR)/3rdparty/dlpack -endif - -ifndef TVM_PATH - TVM_PATH = $(TPARTYDIR)/tvm -endif - -ifndef LLVM_PATH - LLVM_PATH = $(TVM_PATH)/build/llvm -endif - -ifneq ($(USE_OPENMP), 1) - export NO_OPENMP = 1 -endif - -# use customized config file -include $(config) - -ifndef USE_MKLDNN -ifneq ($(UNAME_S), Darwin) -ifneq ($(UNAME_S), Windows) -ifeq ($(UNAME_P), x86_64) - USE_MKLDNN=1 -endif -endif -endif -endif - -ifeq ($(USE_MKL2017), 1) -$(warning "USE_MKL2017 is deprecated. We will switch to USE_MKLDNN.") - USE_MKLDNN=1 -endif - -ifeq ($(USE_MKLDNN), 1) - MKLDNNROOT = $(ROOTDIR)/3rdparty/mkldnn/build/install -endif - -include $(TPARTYDIR)/mshadow/make/mshadow.mk -include $(DMLC_CORE)/make/dmlc.mk - -# all tge possible warning tread -WARNFLAGS= -Wall -Wsign-compare -CFLAGS = -DMSHADOW_FORCE_STREAM $(WARNFLAGS) -# C++ standard -CFLAGS+= -DDMLC_USE_CXX11=1 -DDMLC_USE_CXX11=1 -DDMLC_USE_CXX14=1 -# use old thread local implementation in DMLC-CORE -CFLAGS += -DDMLC_MODERN_THREAD_LOCAL=0 -# disable stack trace in exception by default. -CFLAGS += -DDMLC_LOG_STACK_TRACE_SIZE=0 -CFLAGS += -DDMLC_LOG_FATAL_THROW=1 - -ifeq ($(DEV), 1) - # Excluded from Werror: - # 1) variables used in '#pragma omp parallel' are considered unused - CFLAGS += -g -Werror -Wno-error=unused-variable -Wno-error=maybe-uninitialized -Wno-error=unused-function - NVCCFLAGS += -Werror cross-execution-space-call -endif - -# CFLAGS for debug -ifeq ($(DEBUG), 1) - CFLAGS += -g -O0 -D_GLIBCXX_ASSERTIONS -else - CFLAGS += -O3 -DNDEBUG=1 -endif -CFLAGS += -I$(TPARTYDIR)/mshadow/ -I$(TPARTYDIR)/dmlc-core/include -fPIC -I$(NNVM_PATH)/include -I$(DLPACK_PATH)/include -I$(TPARTYDIR)/tvm/include -Iinclude $(MSHADOW_CFLAGS) -LDFLAGS = -pthread -ldl $(MSHADOW_LDFLAGS) $(DMLC_LDFLAGS) - -# please note that when you enable this, you might run into an linker not being able to work properly due to large code injection. -# you can find more information here https://github.com/apache/incubator-mxnet/issues/15971 -ifeq ($(ENABLE_TESTCOVERAGE), 1) -CFLAGS += --coverage -LDFLAGS += --coverage -endif - -ifeq ($(USE_NVTX), 1) -CFLAGS += -DMXNET_USE_NVTX=1 -LDFLAGS += -lnvToolsExt -endif - -ifeq ($(USE_TENSORRT), 1) - CFLAGS += -I$(ROOTDIR) -I$(TPARTYDIR) -DONNX_NAMESPACE=$(ONNX_NAMESPACE) -DMXNET_USE_TENSORRT=1 - LDFLAGS += -lprotobuf -pthread -lonnx -lonnx_proto -lnvonnxparser -lnvonnxparser_runtime -lnvinfer -lnvinfer_plugin -endif -# -L/usr/local/lib - -ifeq ($(DEBUG), 1) - NVCCFLAGS += -std=c++14 -Xcompiler -D_FORCE_INLINES -g -G -O0 -ccbin $(CXX) $(MSHADOW_NVCCFLAGS) -else - NVCCFLAGS += -std=c++14 -Xcompiler -D_FORCE_INLINES -O3 -ccbin $(CXX) $(MSHADOW_NVCCFLAGS) -endif - -# CFLAGS for segfault logger -ifeq ($(USE_SIGNAL_HANDLER), 1) - CFLAGS += -DMXNET_USE_SIGNAL_HANDLER=1 -endif - -# Caffe Plugin -ifdef CAFFE_PATH - CFLAGS += -DMXNET_USE_CAFFE=1 -endif - -ifndef LINT_LANG - LINT_LANG = "all" -endif - -ifeq ($(USE_MKLDNN), 1) - CFLAGS += -DMXNET_USE_MKLDNN=1 - CFLAGS += -I$(ROOTDIR)/src/operator/nn/mkldnn/ - CFLAGS += -I$(MKLDNNROOT)/include - LIB_DEP += $(MKLDNNROOT)/lib/libdnnl.a -endif - -# setup opencv -ifeq ($(USE_OPENCV), 1) - CFLAGS += -DMXNET_USE_OPENCV=1 - ifneq ($(filter-out NONE, $(
[GitHub] [incubator-mxnet] guptaanshul201989 opened a new issue #18734: Multi-gpu Mxnet training in sagemaker gives cuda error when dataloader is using multiprocessing
guptaanshul201989 opened a new issue #18734: URL: https://github.com/apache/incubator-mxnet/issues/18734 I am trying to train a transformer seq-to-seq model on Sagemaker ( The script I am using works fine when I run it on an EC2 multi gpu instance ). When I start a training job on sagemaker, the training progresses fine, but it logs a cuda error: `[03:28:04] src/engine/threaded_engine_perdevice.cc:101: Ignore CUDA Error [03:28:04] /root/pip_build/mxnet-build/3rdparty/mshadow/mshadow/./tensor_gpu-inl.h:35: Check failed: e == cudaSuccess: CUDA: initialization error Stack trace: [bt] (0) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x6dfb0b) [0x7f9f2591cb0b] [bt] (1) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x3898dd2) [0x7f9f28ad5dd2] [bt] (2) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x38bc49e) [0x7f9f28af949e] [bt] (3) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x38aee71) [0x7f9f28aebe71] [bt] (4) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x38a4a21) [0x7f9f28ae1a21] [bt] (5) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x38a5974) [0x7f9f28ae2974] [bt] (6) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(mxnet::NDArray::Chunk::~Chunk()+0x48a) [0x7f9f28d1ce1a] [bt] (7) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x6e32ba) [0x7f9f259202ba] [bt] (8) /usr/local/lib/python3.6/site-packages/mxnet/libmxnet.so(std::vector >::~vector()+0xc8) [0x7f9f25951818] ` I found out that when I initialize dataloader with multiprocessing, I get this error. When I switch thread_pool on, I don't see this error. 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
[GitHub] [incubator-mxnet] ptrendx opened a new pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ptrendx opened a new pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735 ## Description ## Current build script used by CI on Windows sets the CUDA arch to 5.2, which results in the need to do ptx -> sass compilation at the beginning of the tests (depending on whether the kernels are already in driver jit cache or not). This adds 15-20 minutes to the windows-gpu CI runs that do not have the jit cache populated. This PR changes the CUDA arch used for building in Windows CI to 7.5, which is the used for testing. FYI @ChaiBapchya @leezu ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] Changes are complete (i.e. I finished coding on this PR) 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
mxnet-bot commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659583990 Hey @ptrendx , Thanks for submitting the PR All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: - To trigger all jobs: @mxnet-bot run ci [all] - To trigger specific jobs: @mxnet-bot run ci [job1, job2] *** **CI supported jobs**: [centos-cpu, website, clang, unix-gpu, miscellaneous, edge, windows-cpu, windows-gpu, sanity, centos-gpu, unix-cpu] *** _Note_: Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. All CI tests must pass before the PR can be merged. 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
[GitHub] [incubator-mxnet] leezu commented on issue #18734: Multi-gpu Mxnet training in sagemaker gives cuda error when dataloader is using multiprocessing
leezu commented on issue #18734: URL: https://github.com/apache/incubator-mxnet/issues/18734#issuecomment-659584345 Try not using cuda before enabling the multiprocessor as a workaround. There are various bugs in the multiprocessing implementation in MXNet 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
[GitHub] [incubator-mxnet] ChaiBapchya commented on pull request #18608: Cherry-pick #18310 #18355
ChaiBapchya commented on pull request #18608: URL: https://github.com/apache/incubator-mxnet/pull/18608#issuecomment-659591292 @leezu @TaoLv @ciyongch gentle ping.. for help.. unix-gpu: Any idea why specifically python3 GPU tests are failing while trying to add mkl headers? edge: Jetson build failure: `libmxnet.a(random_generator.cu.o): error adding symbols: File in wrong format` I haven't seen this before.. maybe you guys have some idea? 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
[GitHub] [incubator-mxnet] DickJC123 opened a new issue #18736: unittest/test_numpy_interoperability.py inadvertently puts additional unittests on a fixed seed
DickJC123 opened a new issue #18736: URL: https://github.com/apache/incubator-mxnet/issues/18736 ## Description The following code in test_numpy_interoperability.py sets a numpy seed unilaterally, thereby making deterministic the random numbers used in subsequent unittests: https://github.com/apache/incubator-mxnet/blob/2abf0b8c2b3361c73c9dfdeabdb8a88278b693d0/tests/python/unittest/test_numpy_interoperability.py#L515-L528 Also, since the code is executed on import of the file, this sets the seed in advance of the with_seed() mechanism, thereby forcing determinism on a large number of other modules that import this test file, including test_operator_gpu.py. We clearly want to get back to our usual mode of running tests non-deterministically with different random data each time, to ensure a robust framework. Thus, having noticed this issue during the development of my current PR https://github.com/apache/incubator-mxnet/pull/18694, I have included a fixing commit https://github.com/apache/incubator-mxnet/pull/18694/commits/6567b61d219fa984d5c1b8c3987990ca04b73575 . @vexilligera and @reminisce, could you review my fix and comment, or suggest an alternative fix via another PR? For example, you may want to evaluate whether the test is able to support non-deterministic shapes. Or, you may want to evaluate whether performing all the set-up functions of this test module lazily on first use, rather than as done currently at file import, is generally a better practice. Thanks. ### Error Message I noticed this problem when different runs of the CI would flag the same failing test with the same rnd seed. [Output not retained] ## What have you tried to solve it? 1. See referenced PR and fixing commit 2. 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
[incubator-mxnet-site] branch asf-site updated: Publish triggered by CI
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 710fec8 Publish triggered by CI 710fec8 is described below commit 710fec83f43d1f2d4c8cb85b4f4d8027270b8d0a Author: mxnet-ci AuthorDate: Thu Jul 16 18:42:29 2020 + Publish triggered by CI --- api/python/docs/api/ndarray/ndarray.html | 24 api/python/docs/api/ndarray/op/index.html | 24 api/python/docs/api/ndarray/sparse/index.html | 4 ++-- api/python/docs/api/symbol/op/index.html | 24 api/python/docs/api/symbol/sparse/index.html | 4 ++-- api/python/docs/api/symbol/symbol.html| 24 api/python/docs/searchindex.js| 2 +- date.txt | 1 - feed.xml | 2 +- 9 files changed, 54 insertions(+), 55 deletions(-) diff --git a/api/python/docs/api/ndarray/ndarray.html b/api/python/docs/api/ndarray/ndarray.html index 68e3a60..b62e162 100644 --- a/api/python/docs/api/ndarray/ndarray.html +++ b/api/python/docs/api/ndarray/ndarray.html @@ -5674,7 +5674,7 @@ clip(x,1,8) = [ 1., 1., 2., 3., 4., 5., 6., 7., 8., 8.] clip(csr, a_min > 0, a_max > 0) = csr -Defined in src/operator/tensor/matrix_op.cc:L677 +Defined in src/operator/tensor/matrix_op.cc:L681 Parameters @@ -5876,7 +5876,7 @@ a dense tensor. -Defined in src/operator/tensor/matrix_op.cc:L482 +Defined in src/operator/tensor/matrix_op.cc:L486 Parameters @@ -6050,7 +6050,7 @@ Example: [15, 21, 16, 22, 17, 23 -Defined in src/operator/tensor/matrix_op.cc:L972 +Defined in src/operator/tensor/matrix_op.cc:L976 Parameters @@ -6434,7 +6434,7 @@ from standard updates. For more details, please check the Optimization API at: Inserts a new axis of size 1 into the array shape For example, given x with shape (2,3,4), then expand_dims(x, axis=1) will return a new array with shape (2,1,3,4). -Defined in src/operator/tensor/matrix_op.cc:L395 +Defined in src/operator/tensor/matrix_op.cc:L399 Parameters @@ -6591,7 +6591,7 @@ Examples: [ 9., 8., 7., 6., 5.]] -Defined in src/operator/tensor/matrix_op.cc:L832 +Defined in src/operator/tensor/matrix_op.cc:L836 Parameters @@ -10154,7 +10154,7 @@ elements: -Defined in src/operator/tensor/matrix_op.cc:L744 +Defined in src/operator/tensor/matrix_op.cc:L748 Parameters @@ -10326,7 +10326,7 @@ Examples: [ 9., 8., 7., 6., 5.]] -Defined in src/operator/tensor/matrix_op.cc:L832 +Defined in src/operator/tensor/matrix_op.cc:L836 Parameters @@ -11378,7 +11378,7 @@ a dense tensor. -Defined in src/operator/tensor/matrix_op.cc:L482 +Defined in src/operator/tensor/matrix_op.cc:L486 Parameters @@ -11418,7 +11418,7 @@ Examples: [ 10., 11.]] -Defined in src/operator/tensor/matrix_op.cc:L571 +Defined in src/operator/tensor/matrix_op.cc:L575 Parameters @@ -11478,7 +11478,7 @@ Example: [ 9., 10., 11.]] -Defined in src/operator/tensor/matrix_op.cc:L625 +Defined in src/operator/tensor/matrix_op.cc:L629 Parameters @@ -11766,7 +11766,7 @@ Example: [21, 22, 23 -Defined in src/operator/tensor/matrix_op.cc:L1019 +Defined in src/operator/tensor/matrix_op.cc:L1023 Parameters @@ -12391,7 +12391,7 @@ shape (2,2)< -Defined in src/operator/tensor/matrix_op.cc:L796 +Defined in src/operator/tensor/matrix_op.cc:L800 Parameters diff --git a/api/python/docs/api/ndarray/op/index.html b/api/python/docs/api/ndarray/op/index.html index 7b78dca..2fe5d44 100644 --- a/api/python/docs/api/ndarray/op/index.html +++ b/api/python/docs/api/ndarray/op/index.html @@ -5539,7 +5539,7 @@ clip(x,1,8) = [ 1., 1., 2., 3., 4., 5., 6., 7., 8., 8.] clip(csr, a_min > 0, a_max > 0) = csr -Defined in src/operator/tensor/matrix_op.cc:L677 +Defined in src/operator/tensor/matrix_op.cc:L681 Parameters @@ -5741,7 +5741,7 @@ a dense tensor. -Defined in src/operator/tensor/matrix_op.cc:L482 +Defined in src/operator/tensor/matrix_op.cc:L486 Parameters @@ -5915,7 +5915,7 @@ Example: [15, 21, 16, 22, 17, 23 -Defined in src/operator/tensor/matrix_op.cc:L972 +Defined in src/operator/tensor/matrix_op.cc:L976 Parameters @@ -6299,7 +6299,7 @@ from standard updates. For more details, please check the Optimization API at: Inserts a new axis of size 1 into the array shape For example, given x with shape (2,3,4), then expand_dims(x, axis=1) will return a new array with shape (2,1,3,4). -Defined in src/operator/tensor/matrix_op.cc:L395 +D
[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 02f1618 Bump the publish timestamp. 02f1618 is described below commit 02f161803ca8672c129d223b66e46d03e26383ec Author: mxnet-ci AuthorDate: Thu Jul 16 18:42:35 2020 + 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..2d72a32 --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Thu Jul 16 18:42:35 UTC 2020
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
mxnet-bot commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659600625 Jenkins CI successfully triggered : [edge] 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
[GitHub] [incubator-mxnet] ptrendx commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ptrendx commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659600564 @mxnet-bot run ci [edge] 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
[GitHub] [incubator-mxnet] sandeep-krishnamurthy commented on pull request #18368: [WIP] enable large tensor in np
sandeep-krishnamurthy commented on pull request #18368: URL: https://github.com/apache/incubator-mxnet/pull/18368#issuecomment-659620714 @access2rohit - Please help review. Thanks. 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
[GitHub] [incubator-mxnet] DickJC123 commented on issue #18736: unittest/test_numpy_interoperability.py inadvertently puts additional unittests on a fixed seed
DickJC123 commented on issue #18736: URL: https://github.com/apache/incubator-mxnet/issues/18736#issuecomment-659621773 While technically a different issue, test_numpy_interoperability.py has a spurious print statement: https://github.com/apache/incubator-mxnet/blob/2abf0b8c2b3361c73c9dfdeabdb8a88278b693d0/tests/python/unittest/test_numpy_interoperability.py#L2896-L2902 I've removed this in PR https://github.com/apache/incubator-mxnet/pull/18694 commit https://github.com/apache/incubator-mxnet/pull/18694/commits/3d47acba82e44b6b731ff048905472c8c0c2f8de. 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
[GitHub] [incubator-mxnet] DickJC123 opened a new issue #18737: test_operator_gpu.py::test_batch_norm_with_type inadvertently retests cases, misses others
DickJC123 opened a new issue #18737: URL: https://github.com/apache/incubator-mxnet/issues/18737 ## Description Notice the test case duplication in the code: https://github.com/apache/incubator-mxnet/blob/2abf0b8c2b3361c73c9dfdeabdb8a88278b693d0/tests/python/gpu/test_operator_gpu.py#L406-L424 This is one of the issues fixed by PR https://github.com/apache/incubator-mxnet/pull/18694 in commit https://github.com/apache/incubator-mxnet/pull/18694/commits/4e81d14fb05d48e1eb53e861fb0065b68347f4c0 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18738: set website default version - test redirect
mxnet-bot commented on pull request #18738: URL: https://github.com/apache/incubator-mxnet/pull/18738#issuecomment-659635114 Hey @ys2843 , Thanks for submitting the PR All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: - To trigger all jobs: @mxnet-bot run ci [all] - To trigger specific jobs: @mxnet-bot run ci [job1, job2] *** **CI supported jobs**: [edge, miscellaneous, unix-cpu, unix-gpu, windows-cpu, sanity, centos-cpu, website, windows-gpu, centos-gpu, clang] *** _Note_: Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. All CI tests must pass before the PR can be merged. 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
[GitHub] [incubator-mxnet] ys2843 opened a new pull request #18738: set website default version - test redirect
ys2843 opened a new pull request #18738: URL: https://github.com/apache/incubator-mxnet/pull/18738 ## Description ## #18429 , This PR contains the following redirect rules: When user lands on main page "/" 1. If coming from domains except `mxnet.apache.org` or `mxnet.incubator.apache.org`, redirect users to v1.6 website 2. If coming from urls on MXNet domain, don't redirect to avoid any looping. Because sometimes Apache server can have unexpected behavior on redirect, this PR only contains a test for this rules. If everything works well, I will submit another PR, delete the test code and uncomment the origin lines. ## 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: ### Changes ### - [ ] Test redirect based on HTTP referer header. ## Comments ## - Preview: http://ec2-34-219-134-42.us-west-2.compute.amazonaws.com/versions/1.1.0/community/contribute.html > Should land on the same URI but on v1.6 website if clicking on this link 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
[GitHub] [incubator-mxnet] ys2843 commented on pull request #18738: set website default version - test redirect
ys2843 commented on pull request #18738: URL: https://github.com/apache/incubator-mxnet/pull/18738#issuecomment-659646149 @mxnet-label-bot add [Website] 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
[GitHub] [incubator-mxnet] leezu opened a new issue #18739: Compiling with ps-lite enabled leaves 3rdparty/ps-lite/src/meta.pb.cc 3rdparty/ps-lite/src/meta.pb.h in tree
leezu opened a new issue #18739: URL: https://github.com/apache/incubator-mxnet/issues/18739 ## Description Compilation will generate the `3rdparty/ps-lite/src/meta.pb.cc` `3rdparty/ps-lite/src/meta.pb.h` files. This causes issues when compiling using different Docker containers or on the host platform, as subsequent compilation attempts won't generate the files and instead will fail with the error: ``` [1/223] Building CXX object 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/meta.pb.cc.o FAILED: 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/meta.pb.cc.o /usr/local/bin/ccache /usr/bin/g++-7 -DDMLC_LOG_STACK_TRACE_SIZE=0 -DDMLC_MODERN_THREAD_LOCAL=0 -DDMLC_STRICT_CXX11 -DDMLC_USE_CXX11 -DDMLC_USE_CXX14 -DMSHADOW_IN_CXX11 -DMSHADOW_USE _CBLAS=1 -DMSHADOW_USE_MKL=0 -DMXNET_USE_BLAS_OPEN=1 -DMXNET_USE_LAPACK=1 -DMXNET_USE_LIBJPEG_TURBO=0 -DMXNET_USE_OPENCV=1 -DMXNET_USE_OPENMP=1 -DMXNET_USE_OPERATOR_TUNING=1 -DMXNET_U SE_TVM_OP=1 -I/work/mxnet/include -I/work/mxnet/src -I/work/mxnet/3rdparty/nvidia_cub -I/work/mxnet/3rdparty/tvm/nnvm/include -I/work/mxnet/3rdparty/tvm/include -I/work/mxnet/3rdparty /dmlc-core/include -I/work/mxnet/3rdparty/dlpack/include -I3rdparty/ps-lite -I/work/mxnet/3rdparty/ps-lite/include -isystem /usr/include/opencv -Wno-error=strict-overflow -D_GLIBCXX_A SSERTIONS -Wall -Wno-sign-compare -O3 -g -fopenmp -O2 -g -DNDEBUG -fPIC -std=c++11 -MD -MT 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/meta.pb.cc.o -MF 3rdparty/ps-lite/CMakeFiles/p slite.dir/src/meta.pb.cc.o.d -o 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/meta.pb.cc.o -c /work/mxnet/3rdparty/ps-lite/src/meta.pb.cc In file included from /work/mxnet/3rdparty/ps-lite/src/meta.pb.cc:4:0: /work/mxnet/3rdparty/ps-lite/src/meta.pb.h:10:10: fatal error: google/protobuf/port_def.inc: No such file or directory #include ^~ compilation terminated. [2/223] Building CXX object 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/van.cc.o FAILED: 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/van.cc.o /usr/local/bin/ccache /usr/bin/g++-7 -DDMLC_LOG_STACK_TRACE_SIZE=0 -DDMLC_MODERN_THREAD_LOCAL=0 -DDMLC_STRICT_CXX11 -DDMLC_USE_CXX11 -DDMLC_USE_CXX14 -DMSHADOW_IN_CXX11 -DMSHADOW_USE _CBLAS=1 -DMSHADOW_USE_MKL=0 -DMXNET_USE_BLAS_OPEN=1 -DMXNET_USE_LAPACK=1 -DMXNET_USE_LIBJPEG_TURBO=0 -DMXNET_USE_OPENCV=1 -DMXNET_USE_OPENMP=1 -DMXNET_USE_OPERATOR_TUNING=1 -DMXNET_U SE_TVM_OP=1 -I/work/mxnet/include -I/work/mxnet/src -I/work/mxnet/3rdparty/nvidia_cub -I/work/mxnet/3rdparty/tvm/nnvm/include -I/work/mxnet/3rdparty/tvm/include -I/work/mxnet/3rdparty /dmlc-core/include -I/work/mxnet/3rdparty/dlpack/include -I3rdparty/ps-lite -I/work/mxnet/3rdparty/ps-lite/include -isystem /usr/include/opencv -Wno-error=strict-overflow -D_GLIBCXX_A SSERTIONS -Wall -Wno-sign-compare -O3 -g -fopenmp -O2 -g -DNDEBUG -fPIC -std=c++11 -MD -MT 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/van.cc.o -MF 3rdparty/ps-lite/CMakeFiles/pslit e.dir/src/van.cc.o.d -o 3rdparty/ps-lite/CMakeFiles/pslite.dir/src/van.cc.o -c /work/mxnet/3rdparty/ps-lite/src/van.cc In file included from /work/mxnet/3rdparty/ps-lite/src/van.cc:14:0: /work/mxnet/3rdparty/ps-lite/src/./meta.pb.h:10:10: fatal error: google/protobuf/port_def.inc: No such file or directory #include ^~ compilation terminated. ``` No files should be generated in the src or 3rdparty folder; but all build-generated files should be contained in the build folder. As 3rdparty/ps-lite is a submodule, `git clean -ffxd` won't clean the files either. 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
[GitHub] [incubator-mxnet] leezu commented on pull request #18733: Move gluon.metric api docs
leezu commented on pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733#issuecomment-659677878 @mxnet-bot run ci [centos-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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18733: Move gluon.metric api docs
mxnet-bot commented on pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733#issuecomment-659677902 Jenkins CI successfully triggered : [centos-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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
mxnet-bot commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659677769 Jenkins CI successfully triggered : [unix-cpu, windows-gpu] 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
[GitHub] [incubator-mxnet] leezu commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
leezu commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659677721 @mxnet-bot run ci [unix-cpu, windows-gpu] 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
[GitHub] [incubator-mxnet] leezu commented on a change in pull request #18690: [WIP] optimize graph in presence of dynamic shape ops
leezu commented on a change in pull request #18690: URL: https://github.com/apache/incubator-mxnet/pull/18690#discussion_r456130118 ## File path: python/mxnet/symbol/symbol.py ## @@ -1470,6 +1470,10 @@ def optimize_for(self, backend, args=None, aux=None, ctx=None, ctx : Context, optional Device context, used to infer stypes +is_np_sym : boolean, optional +Output symbol type +- If true, output type is np symbol, otherwise nd symbol. + Review comment: nd symbol will be removed soon anyways. Can you rely on the global `is_np` state (and if needed adapt the state in the `build_cache` function) ## File path: python/mxnet/symbol/symbol.py ## @@ -2627,6 +2633,24 @@ def detach(self): def backward(self): raise NotImplementedForSymbol(self.backward, None) +def optimize_for_dynamic_shape_op(self, is_np_sym=False): Review comment: Why is this a public API even though it's called automatically in `_build_cache`? Should it be private? 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
[GitHub] [incubator-mxnet] wkcn commented on pull request #18707: [MXNET-1453] Support the intput whose dimension is greater than 6 for Transpose and Rollaxis
wkcn commented on pull request #18707: URL: https://github.com/apache/incubator-mxnet/pull/18707#issuecomment-659732747 @szha In numpy, the output of np.transpose is not contiguous, but the copied ndarray is contiguous. We compute each position of tranpose's output individually, but numpy computes them in blocks. There is still an improvement space. 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18733: Move gluon.metric api docs
mxnet-bot commented on pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733#issuecomment-659739371 Jenkins CI successfully triggered : [centos-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
[GitHub] [incubator-mxnet] leezu commented on pull request #18733: Move gluon.metric api docs
leezu commented on pull request #18733: URL: https://github.com/apache/incubator-mxnet/pull/18733#issuecomment-659739351 @mxnet-bot run ci [centos-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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18722: Remove NNPACK integration
mxnet-bot commented on pull request #18722: URL: https://github.com/apache/incubator-mxnet/pull/18722#issuecomment-659741094 Jenkins CI successfully triggered : [centos-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
[GitHub] [incubator-mxnet] leezu commented on pull request #18722: Remove NNPACK integration
leezu commented on pull request #18722: URL: https://github.com/apache/incubator-mxnet/pull/18722#issuecomment-659741070 @mxnet-bot run ci [centos-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
[GitHub] [incubator-mxnet] leezu opened a new issue #18740: test_sparse_operator.py::test_elemwise_binary_ops
leezu opened a new issue #18740: URL: https://github.com/apache/incubator-mxnet/issues/18740 ## Description Tests crashes affecting multiple PRs: https://github.com/apache/incubator-mxnet/pull/18711 https://github.com/apache/incubator-mxnet/pull/18694 https://github.com/apache/incubator-mxnet/pull/18722 https://github.com/apache/incubator-mxnet/pull/18733 ``` [2020-07-15T23:41:55.453Z] Fatal Python error: Aborted [2020-07-15T23:41:55.453Z] [2020-07-15T23:41:55.453Z] Thread 0x7f6de68a6700 (most recent call first): [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/execnet/gateway_base.py", line 400 in read [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/execnet/gateway_base.py", line 432 in from_io [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/execnet/gateway_base.py", line 967 in _thread_receiver [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/execnet/gateway_base.py", line 220 in run [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/execnet/gateway_base.py", line 285 in _perform_spawn [2020-07-15T23:41:55.453Z] [2020-07-15T23:41:55.453Z] Current thread 0x7f6de857a740 (most recent call first): [2020-07-15T23:41:55.453Z] File "/work/mxnet/python/mxnet/_ctypes/ndarray.py", line 178 in __call__ [2020-07-15T23:41:55.453Z] File "/work/mxnet/python/mxnet/executor.py", line 184 in forward [2020-07-15T23:41:55.453Z] File "/work/mxnet/python/mxnet/test_utils.py", line 937 in numeric_grad [2020-07-15T23:41:55.453Z] File "/work/mxnet/python/mxnet/test_utils.py", line 1088 in check_numeric_gradient [2020-07-15T23:41:55.453Z] File "/work/mxnet/tests/python/unittest/test_sparse_operator.py", line 312 in test_elemwise_binary_op [2020-07-15T23:41:55.453Z] File "/work/mxnet/tests/python/unittest/test_sparse_operator.py", line 417 in check_elemwise_binary_ops [2020-07-15T23:41:55.453Z] File "/work/mxnet/tests/python/unittest/test_sparse_operator.py", line 520 in test_elemwise_binary_ops [2020-07-15T23:41:55.453Z] File "/work/mxnet/tests/python/unittest/common.py", line 223 in test_new [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/python.py", line 167 in pytest_pyfunc_call [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/callers.py", line 187 in _multicall [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/manager.py", line 87 in [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/manager.py", line 93 in _hookexec [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/hooks.py", line 286 in __call__ [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/python.py", line 1445 in runtest [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 134 in pytest_runtest_call [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/callers.py", line 187 in _multicall [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/manager.py", line 87 in [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/manager.py", line 93 in _hookexec [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/pluggy/hooks.py", line 286 in __call__ [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 210 in [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 237 in from_call [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 210 in call_runtest_hook [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/flaky/flaky_pytest_plugin.py", line 129 in call_and_report [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 99 in runtestprotocol [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/_pytest/runner.py", line 84 in pytest_runtest_protocol [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/flaky/flaky_pytest_plugin.py", line 92 in pytest_runtest_protocol [2020-07-15T23:41:55.453Z] File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/plug
[GitHub] [incubator-mxnet] ChaiBapchya commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ChaiBapchya commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659742257 @leezu @marcoabreu Looked into the mxnet-ci isengard a/c, to find: windows-gpu instances are all G3 all G3s are Tesla M60. [discussed offline with ptrendx] 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
[GitHub] [incubator-mxnet] szha commented on pull request #18582: Refactoring of Pooled Storage Manager classes
szha commented on pull request #18582: URL: https://github.com/apache/incubator-mxnet/pull/18582#issuecomment-659742603 @andrei5055 thanks for the refactor 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
[GitHub] [incubator-mxnet] szha merged pull request #18582: Refactoring of Pooled Storage Manager classes
szha merged pull request #18582: URL: https://github.com/apache/incubator-mxnet/pull/18582 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
[incubator-mxnet] branch master updated (2abf0b8 -> 3ef00b8)
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 2abf0b8 Initialize docker cache in build.py for docker-compose containers (#18724) add 3ef00b8 Refactoring of Pooled Storage Manager classes (#18582) No new revisions were added by this update. Summary of changes: docs/static_site/src/pages/api/faq/env_var.md | 67 ++- src/profiler/storage_profiler.h | 20 +- src/storage/cpu_device_storage.h | 11 +- src/storage/cpu_shared_storage_manager.h | 16 +- src/storage/gpu_device_storage.h | 37 +- src/storage/naive_storage_manager.h | 1 - src/storage/pinned_memory_storage.h | 20 +- src/storage/pooled_storage_manager.h | 590 ++ src/storage/storage.cc| 269 +++- src/storage/storage_manager.h | 2 +- src/storage/storage_manager_helpers.h | 162 +++ tests/python/unittest/test_gluon_data.py | 15 +- 12 files changed, 732 insertions(+), 478 deletions(-) create mode 100644 src/storage/storage_manager_helpers.h
[incubator-mxnet] branch master updated (2abf0b8 -> 3ef00b8)
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 2abf0b8 Initialize docker cache in build.py for docker-compose containers (#18724) add 3ef00b8 Refactoring of Pooled Storage Manager classes (#18582) No new revisions were added by this update. Summary of changes: docs/static_site/src/pages/api/faq/env_var.md | 67 ++- src/profiler/storage_profiler.h | 20 +- src/storage/cpu_device_storage.h | 11 +- src/storage/cpu_shared_storage_manager.h | 16 +- src/storage/gpu_device_storage.h | 37 +- src/storage/naive_storage_manager.h | 1 - src/storage/pinned_memory_storage.h | 20 +- src/storage/pooled_storage_manager.h | 590 ++ src/storage/storage.cc| 269 +++- src/storage/storage_manager.h | 2 +- src/storage/storage_manager_helpers.h | 162 +++ tests/python/unittest/test_gluon_data.py | 15 +- 12 files changed, 732 insertions(+), 478 deletions(-) create mode 100644 src/storage/storage_manager_helpers.h
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18722: Remove NNPACK integration
mxnet-bot commented on pull request #18722: URL: https://github.com/apache/incubator-mxnet/pull/18722#issuecomment-659750295 Jenkins CI successfully triggered : [centos-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
[GitHub] [incubator-mxnet] leezu commented on pull request #18722: Remove NNPACK integration
leezu commented on pull request #18722: URL: https://github.com/apache/incubator-mxnet/pull/18722#issuecomment-659750261 @mxnet-bot run ci [centos-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
[GitHub] [incubator-mxnet] Kh4L commented on issue #18729: TensorRT tutorial is missing in master
Kh4L commented on issue #18729: URL: https://github.com/apache/incubator-mxnet/issues/18729#issuecomment-659752290 @ptrendx @aaronmarkham It has been removed in this PR https://github.com/apache/incubator-mxnet/pull/18579 cc @eric-haibin-lin 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
[incubator-mxnet-site] branch asf-site updated: Publish triggered by CI
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 f36fa90 Publish triggered by CI f36fa90 is described below commit f36fa90f675ad00437c975f71c3a6c9a061785b8 Author: mxnet-ci AuthorDate: Fri Jul 17 00:39:10 2020 + Publish triggered by CI --- api/faq/env_var.html | 95 ++-- date.txt | 1 - feed.xml | 2 +- 3 files changed, 86 insertions(+), 12 deletions(-) diff --git a/api/faq/env_var.html b/api/faq/env_var.html index dddc8e1..1fae11c 100644 --- a/api/faq/env_var.html +++ b/api/faq/env_var.html @@ -575,37 +575,112 @@ Typically, you wouldn’t need to change these settings, but they are listed her This parameter is also used to get number of matching colors in graph and in turn how much parallelism one can get in each GPU. Color based match usually costs more memory but also enables more parallelism. + MXNET_GPU_MEM_POOL_TYPE + + Values: String (default=Naive) + The type of GPU memory pool. + Choices: + + Naive: A simple memory pool that allocates memory for the requested size and cache memory buffers, when this memory is released. The size of memory chunk is defined by rounding the requested memory size to the nearest bigger multiple of MXNET_GPU_MEM_POOL_PAGE_SIZE (or MXNET_GPU_MEM_LARGE_ALLOC_ROUND_SIZE, when the result of rounding for MXNET_GPU_MEM_POOL_PAGE_SIZE is bigger than MXNET_GPU_MEM_LARGE_ALLOC_ROUND_SIZE) and allocates memory of the rounded size. + Round: A memory pool that try to rounds the requested memory size to the nearest bigger power of 2. When this rounded number is bigger that 2*MXNET_GPU_MEM_POOL_ROUND_LINEAR_CUTOFF, the *Naive rounding algorithm is used. Caching and allocating buffered memory works in the same way as the naive memory pool. + Unpooled: No memory pool is used. + + + + MXNET_GPU_MEM_POOL_RESERVE Values: Int (default=5) The percentage of GPU memory to reserve for things other than the GPU array, such as kernel launch or cudnn handle space. + The value is used only by the GPU memory pool. If it is not possible to allocate new memory AND still save this reserve, the memory pool will free the cached memory. If you see a strange out-of-memory error from the kernel launch, after multiple iterations, try setting this to a larger value. - MXNET_GPU_MEM_POOL_TYPE + MXNET_GPU_MEM_LARGE_ALLOC_ROUND_SIZE + + Values: Int (default=2097152) + When the rounded size of memory allocations calculated by the pool of Naive type is larger than this threshold, it will be rounded up to a multiple of this value. + The default was chosen to minimize global memory fragmentation within the GPU driver. Set this to 1 to disable. + + + MXNET_GPU_MEM_POOL_ROUND_LINEAR_CUTOFF + + Values: Int (default=24) + The cutoff threshold used by Round strategy. Let’s denote the threshold as T. If the memory size is smaller than 2 ** T (by default, it’s 2 ** 24 = 16MB), it rounds to the smallest 2 ** n that is larger than the requested memory size; if the memory size is larger than 2 ** T, it rounds to the next k * 2 ** T. + + + MXNET_CPU_MEM_POOL_TYPE Values: String (default=Naive) - The type of memory pool. + The type of CPU memory pool. Choices: - Naive: A simple memory pool that allocates memory for the exact requested size and cache memory buffers. If a buffered memory chunk matches the size of a new request, the chunk from the memory pool will be returned and reused. - Round: A memory pool that always rounds the requested memory size and allocates memory of the rounded size. MXNET_GPU_MEM_POOL_ROUND_LINEAR_CUTOFF defines how to round up a memory size. Caching and allocating buffered memory works in the same way as the naive memory pool. - Unpooled: No memory pool is used. + Naive: A simple memory pool that allocates memory for the requested size and cache memory buffers, when this memory is released. The size of memory chunk is defined by rounding the requested memory size to the nearest bigger multiple of MXNET_CPU_MEM_POOL_PAGE_SIZE (or MXNET_CPU_MEM_LARGE_ALLOC_ROUND_SIZE, when the result of rounding for MXNET_CPU_MEM_POOL_PAGE_SIZE is bigger than MXNET_CPU_MEM_LARGE_ALLOC_ROUND_SIZE) and allocates memory of the rounded size. + Round: A memory pool that try to rounds the requested memory size to the nearest bigger power of 2. When this rounded number is bigger that 2*MXNET_CPU_MEM_POOL_ROUND_LINEAR_CUTOFF, the the *Naive rounding algorithm is used. Caching and allocating buffered memory works in
[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 c073125 Bump the publish timestamp. c073125 is described below commit c073125d8e589239942fd6001af72c005342b794 Author: mxnet-ci AuthorDate: Fri Jul 17 00:39:15 2020 + 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..953bef2 --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Fri Jul 17 00:39:15 UTC 2020
[GitHub] [incubator-mxnet] DickJC123 commented on pull request #18688: Fix the flaky test 'test_npx_batch_norm'
DickJC123 commented on pull request #18688: URL: https://github.com/apache/incubator-mxnet/pull/18688#issuecomment-659766979 My PR https://github.com/apache/incubator-mxnet/pull/18694 is now pushed and I am working toward getting a clean CI prior to merging. The sample- vs. population-variance issue I mentioned above is corrected in that PR's commit https://github.com/apache/incubator-mxnet/pull/18694/commits/e0a7dda38d17c7c607a94ef4efe3b88ff1955fb3 . Let me know if you would like the problem sizes reverted back to what they were and I can add a commit for that. 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
[GitHub] [incubator-mxnet] DickJC123 commented on a change in pull request #18694: Unittest tolerance handling improvements
DickJC123 commented on a change in pull request #18694: URL: https://github.com/apache/incubator-mxnet/pull/18694#discussion_r456164385 ## File path: Makefile ## @@ -431,7 +431,7 @@ endif # be JIT-compiled by the updated driver from the included PTX. ifeq ($(USE_CUDA), 1) ifeq ($(CUDA_ARCH),) - KNOWN_CUDA_ARCHS := 30 35 50 52 60 61 70 75 + KNOWN_CUDA_ARCHS := 30 35 50 52 60 61 70 75 80 Review comment: I will revert the commit that added '80' since the Makefile will soon be removed. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
[GitHub] [incubator-mxnet] eric-haibin-lin opened a new pull request #18741: Disable sparse op test
eric-haibin-lin opened a new pull request #18741: URL: https://github.com/apache/incubator-mxnet/pull/18741 ## Description ## Disabling this test for now to unblock other PRs, while I'm looking into it. https://github.com/apache/incubator-mxnet/issues/18740 ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$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 best of my 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18741: Disable sparse op test
mxnet-bot commented on pull request #18741: URL: https://github.com/apache/incubator-mxnet/pull/18741#issuecomment-659773696 Hey @eric-haibin-lin , Thanks for submitting the PR All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: - To trigger all jobs: @mxnet-bot run ci [all] - To trigger specific jobs: @mxnet-bot run ci [job1, job2] *** **CI supported jobs**: [clang, centos-cpu, sanity, unix-gpu, website, unix-cpu, miscellaneous, windows-cpu, centos-gpu, edge, windows-gpu] *** _Note_: Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. All CI tests must pass before the PR can be merged. 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
[GitHub] [incubator-mxnet] ChaiBapchya commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ChaiBapchya commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659774362 Thus we can close this PR since cuda arch for Tesla M60 is 5.2 https://developer.nvidia.com/cuda-gpus 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
[GitHub] [incubator-mxnet] ChaiBapchya opened a new pull request #18742: [v1.x] Cherrypick Migrate from private to public jetson toolchain files (#18677)
ChaiBapchya opened a new pull request #18742: URL: https://github.com/apache/incubator-mxnet/pull/18742 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
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18742: [v1.x] Cherrypick Migrate from private to public jetson toolchain files (#18677)
mxnet-bot commented on pull request #18742: URL: https://github.com/apache/incubator-mxnet/pull/18742#issuecomment-659776652 Hey @ChaiBapchya , Thanks for submitting the PR All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: - To trigger all jobs: @mxnet-bot run ci [all] - To trigger specific jobs: @mxnet-bot run ci [job1, job2] *** **CI supported jobs**: [clang, centos-cpu, sanity, unix-gpu, website, unix-cpu, miscellaneous, windows-cpu, centos-gpu, edge, windows-gpu] *** _Note_: Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. All CI tests must pass before the PR can be merged. 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
[GitHub] [incubator-mxnet] ptrendx closed pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ptrendx closed pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735 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
[GitHub] [incubator-mxnet] ptrendx edited a comment on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ptrendx edited a comment on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659807152 That is true - I was working on a g4 instance that I thought matches the windows-gpu CI, where I saw that issue and thought the reason of this strange gap at the beginning of some of the windows-gpu test runs is due to this. It is unfortunately not the case. 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
[GitHub] [incubator-mxnet] ptrendx commented on pull request #18735: Match the CUDA arch to the GPU used for CI testing on Windows
ptrendx commented on pull request #18735: URL: https://github.com/apache/incubator-mxnet/pull/18735#issuecomment-659807152 That is true - I was working on an instance that I thought matches the windows-gpu CI, where I saw that issue and thought the reason of this strange gap at the beginning of some of the windows-gpu test runs is due to this. It is unfortunately not the case. 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
[GitHub] [incubator-mxnet] wkcn commented on pull request #18688: Fix the flaky test 'test_npx_batch_norm'
wkcn commented on pull request #18688: URL: https://github.com/apache/incubator-mxnet/pull/18688#issuecomment-659811667 @DickJC123 Thank you! The problem size could be reverted back after merging the PR #18694 : ) 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
[GitHub] [incubator-mxnet] chengyuz opened a new issue #18743: AMP: an illegal memory access was encountered
chengyuz opened a new issue #18743: URL: https://github.com/apache/incubator-mxnet/issues/18743 ## Description i followed this link(https://mxnet.apache.org/api/python/docs/tutorials/performance/backend/amp.html) to enable amp in my project, but with error: INFO:root: INFO:root:Using AMP INFO:root:Features in transition 1: 96 -> 96 INFO:root:Features in transition 2: 192 -> 192 INFO:root:Features in transition 3: 448 -> 448 [11:43:40] /media/apache-mxnet-src-1.6.0-incubating/src/io/iter_image_recordio_2.cc:178: ImageRecordIOParser2: ./dataset/imagenet200/rec/train.rec, use 30 threads for decoding.. [11:43:42] /media/apache-mxnet-src-1.6.0-incubating/src/io/iter_image_recordio_2.cc:178: ImageRecordIOParser2: ./dataset/imagenet200/rec/val.rec, use 30 threads for decoding.. [11:44:05] /media/apache-mxnet-src-1.6.0-incubating/src/operator/nn/./cudnn/./cudnn_algoreg-inl.h:97: Running performance tests to find the best convolution algorithm, this can take a while... (set the environment variable MXNET_CUDNN_AUTOTUNE_DEFAULT to 0 to disable) [11:44:10] /media/apache-mxnet-src-1.6.0-incubating/src/operator/nn/./cudnn/./cudnn_algoreg-inl.h:97: Running performance tests to find the best convolution algorithm, this can take a while... (set the environment variable MXNET_CUDNN_AUTOTUNE_DEFAULT to 0 to disable) [11:44:18] /media/apache-mxnet-src-1.6.0-incubating/src/kvstore/././comm.h:744: only 0 out of 2 GPU pairs are enabled direct access. It may affect the performance. You can set MXNET_ENABLE_GPU_P2P=0 to turn it off [11:44:18] /media/apache-mxnet-src-1.6.0-incubating/src/kvstore/././comm.h:753: .. [11:44:18] /media/apache-mxnet-src-1.6.0-incubating/src/kvstore/././comm.h:753: .. Traceback (most recent call last): File "scripts/train_imagenet.py", line 807, in main() File "scripts/train_imagenet.py", line 803, in main train(context) File "scripts/train_imagenet.py", line 736, in train trainer.step(batch_size) File "/media/apache-mxnet-src-1.6.0-incubating/python/mxnet/gluon/trainer.py", line 334, in step self._allreduce_grads() File "/media/apache-mxnet-src-1.6.0-incubating/python/mxnet/gluon/trainer.py", line 364, in _allreduce_grads self._kvstore.push(i, param.list_grad(), priority=-i) File "/media/apache-mxnet-src-1.6.0-incubating/python/mxnet/kvstore.py", line 234, in push self.handle, mx_uint(len(ckeys)), ckeys, cvals, ctypes.c_int(priority))) File "/media/apache-mxnet-src-1.6.0-incubating/python/mxnet/base.py", line 255, in check_call raise MXNetError(py_str(_LIB.MXGetLastError())) mxnet.base.MXNetError: [11:44:18] /media/apache-mxnet-src-1.6.0-incubating/src/storage/./pooled_storage_manager.h:164: cudaMalloc failed: an illegal memory access was encountered Stack trace: [bt] (0) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x43) [0x7f500e8f9493] [bt] (1) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(mxnet::storage::GPUPooledStorageManager::Alloc(mxnet::Storage::Handle*)+0x245) [0x7f50113b6775] [bt] (2) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(mxnet::StorageImpl::Alloc(mxnet::Storage::Handle*)+0x59) [0x7f50113b8c79] [bt] (3) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(mxnet::NDArray::NDArray(mxnet::TShape const&, mxnet::Context, bool, int)+0x52b) [0x7f500e91272b] [bt] (4) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(mxnet::kvstore::CommDevice::Reduce(int, std::vector > const&, int)+0x277) [0x7f500ebb5eb7] [bt] (5) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(mxnet::kvstore::KVStoreLocal::PushImpl(std::vector > const&, std::vector > const&, int)+0x11d) [0x7f500ebb9f5d] [bt] (6) /media/apache-mxnet-src-1.6.0-incubating/python/mxnet/../../build/libmxnet.so(MXKVStorePush+0x105) [0x7f500e903845] [bt] (7) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call_unix64+0x4c) [0x7f504603fdae] [bt] (8) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call+0x22f) [0x7f504603f71f] ## Environment mxnet1.6.0 build from source, gtx2080, python3.6.9 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
[GitHub] [incubator-mxnet] szha commented on issue #18743: AMP: an illegal memory access was encountered
szha commented on issue #18743: URL: https://github.com/apache/incubator-mxnet/issues/18743#issuecomment-659839040 how do you reproduce the error? 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
[GitHub] [incubator-mxnet] ciyongch commented on pull request #18608: Cherry-pick #18310 #18355
ciyongch commented on pull request #18608: URL: https://github.com/apache/incubator-mxnet/pull/18608#issuecomment-659849744 Hi @ChaiBapchya , I took a look at the failure of [unix-gpu] which showed TVM compilation error and GPU OOM runtime error, seems it's more like the CI side issue. can you try to trigger the failed jobs? This patch is targeting to fix the header issue for the gpu binary, right? 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
[GitHub] [incubator-mxnet] szha commented on pull request #18368: [WIP] enable large tensor in np
szha commented on pull request #18368: URL: https://github.com/apache/incubator-mxnet/pull/18368#issuecomment-659852424 @sandeep-krishnamurthy the CI is currently stuck on build issue that @TaoLv pointed out. @access2rohit 's PR on the build fix is needed to address the issue but it seems that there has been no progress in #18645. What's the plan? 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
[GitHub] [incubator-mxnet] szha merged pull request #18722: Remove NNPACK integration
szha merged pull request #18722: URL: https://github.com/apache/incubator-mxnet/pull/18722 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
[incubator-mxnet] branch master updated: Remove NNPACK integration (#18722)
This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new a77f774 Remove NNPACK integration (#18722) a77f774 is described below commit a77f774ed179786fc8429d913a2da1d942528de9 Author: Leonard Lausen AuthorDate: Fri Jul 17 05:01:17 2020 + Remove NNPACK integration (#18722) --- docs/static_site/src/pages/api/faq/env_var.md| 3 - docs/static_site/src/pages/api/faq/nnpack.md | 162 --- src/operator/convolution_v1.cc | 4 - src/operator/nn/convolution.cc | 3 - src/operator/nn/fully_connected.cc | 3 - src/operator/nn/pooling.cc | 3 - src/operator/nnpack/nnpack_convolution-inl.h | 124 - src/operator/nnpack/nnpack_fully_connected-inl.h | 108 --- src/operator/nnpack/nnpack_pooling-inl.h | 91 - src/operator/nnpack/nnpack_util.cc | 37 -- src/operator/nnpack/nnpack_util.h| 64 - 11 files changed, 602 deletions(-) diff --git a/docs/static_site/src/pages/api/faq/env_var.md b/docs/static_site/src/pages/api/faq/env_var.md index 364fd1d..55e5f38 100644 --- a/docs/static_site/src/pages/api/faq/env_var.md +++ b/docs/static_site/src/pages/api/faq/env_var.md @@ -59,9 +59,6 @@ $env:MXNET_STORAGE_FALLBACK_LOG_VERBOSE=0 * MXNET_CPU_PRIORITY_NTHREADS - Values: Int ```(default=4)``` - The number of threads given to prioritized CPU jobs. -* MXNET_CPU_NNPACK_NTHREADS - - Values: Int ```(default=4)``` - - The number of threads used for NNPACK. NNPACK package aims to provide high-performance implementations of some layers for multi-core CPUs. Checkout [NNPACK]({{'/api/faq/nnpack'|relative_url}}) to know more about it. * MXNET_MP_WORKER_NTHREADS - Values: Int ```(default=1)``` - The number of scheduling threads on CPU given to multiprocess workers. Enlarge this number allows more operators to run in parallel in individual workers but please consider reducing the overall `num_workers` to avoid thread contention (not available on Windows). diff --git a/docs/static_site/src/pages/api/faq/nnpack.md b/docs/static_site/src/pages/api/faq/nnpack.md deleted file mode 100644 index 84bedee..000 --- a/docs/static_site/src/pages/api/faq/nnpack.md +++ /dev/null @@ -1,162 +0,0 @@ -layout: page_category -title: NNPACK for Multi-Core CPU Support in MXNet -category: faq -faq_c: Speed -question: Can I use nnpack to improve the CPU performance of MXNet? -permalink: /api/faq/nnpack - - - - - - - - - - - - - - - - - -### NNPACK for Multi-Core CPU Support in MXNet -[NNPACK](https://github.com/Maratyszcza/NNPACK) is an acceleration package -for neural network computations, which can run on x86-64, ARMv7, or ARM64 architecture CPUs. -Using NNPACK, higher-level libraries like _MXNet_ can speed up -the execution on multi-core CPU computers, including laptops and mobile devices. - -_MXNet_ supports NNPACK for forward propagation (inference only) in convolution, max-pooling, and fully-connected layers. -In this document, we give a high level overview of how to use NNPACK with _MXNet_. - - -### Conditions -The underlying implementation of NNPACK utilizes several acceleration methods, -including [fft](https://arxiv.org/abs/1312.5851) and [winograd](https://arxiv.org/abs/1509.09308). -These algorithms work better on some special `batch size`, `kernel size`, and `stride` settings than on other, -so depending on the context, not all convolution, max-pooling, or fully-connected layers can be powered by NNPACK. -When favorable conditions for running NNPACKS are not met, -_MXNet_ will fall back to the default implementation automatically. - -NNPACK only supports Linux and OS X systems. Windows is not supported at present. -The following table explains under which conditions NNPACK will work. - -| operation | conditions | -|:- |:-- | -|convolution |2d convolution `and` no-bias=False `and` dilate=(1,1) `and` num_group=1 `and` batch-size = 1 or batch-size > 1 && stride = (1,1);| -|pooling | max-pooling `and` kernel=(2,2) `and` stride=(2,2) `and` pooling_convention=full| -|fully-connected| without any restrictions | - -### Build/Install NNPACK with MXNet - -If the trained model meets some conditions of using NNPACK, -you can build MXNet with NNPACK support. -Follow these simple steps: -* Build NNPACK shared library with the following commands. _MXNet_ will link NNPACK dynamically. - -Note: The following NNPACK installation instructions have been tested on Ubuntu 14.04 and 16.04. - -```bash -# Install Pip -$ sudo apt-get update -$ sudo apt-get install -y python-pip -$ sudo pip install --upgrade pip - -# Install Peach -$ git clone https://github.com/Maratyszcza/Peac