stsukrov commented on issue #13846: Unit tests are broken on OSX
URL:
https://github.com/apache/incubator-mxnet/issues/13846#issuecomment-468576825
CLion uses CMake, so we cannot ignore it completely.
I'm not even sure, that the default make even makes tests. Last time, I
found
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a change to branch numpy
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from f9c436b Fix req=null in SliceLikeBackward (#14209)
add 862cbc6 softmax for fp16 with fp32
apeforest commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261502101
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
apeforest commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261502101
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
reminisce closed issue #14266: Move TShape definition and necessary passes out
of NNVM
URL: https://github.com/apache/incubator-mxnet/issues/14266
This is an automated message from the Apache Git Service.
To respond to the
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 c8fd793 Bump the publish
wkcn commented on issue #14284: Allow gluon blocks to register lists of blocks
as attributes
URL: https://github.com/apache/incubator-mxnet/pull/14284#issuecomment-468563353
The PR is good.
I do not know whether there is a conflict between this PR and
HybridConcurrent.
For
wkcn commented on issue #11635: Aborted python session when using nd.concat
URL:
https://github.com/apache/incubator-mxnet/issues/11635#issuecomment-468559877
Thanks for your report.
Current MXNet does not support 0-size tensor, like mx.nd.empty((0, 1)).
We will fix it.
anirudhacharya commented on issue #14190: [Flaky Test] Python3: MKLDNN-GPU
test_kvstore_gpu.test_rsp_push_pull
URL: https://github.com/apache/incubator-mxnet/pull/14190#issuecomment-468553422
@Chancebair is there an issue to track the re-enabling of this test?
access2rohit edited a comment on issue #11635: Aborted python session when
using nd.concat
URL:
https://github.com/apache/incubator-mxnet/issues/11635#issuecomment-468552555
I am able to repro the issue. Will raise a PR for the fix by tomorrow
access2rohit commented on issue #11635: Aborted python session when using
nd.concat
URL:
https://github.com/apache/incubator-mxnet/issues/11635#issuecomment-468552555
I am able to repro the issue. Will raise a PR by tomorrow
edfall closed issue #14293: SOFTMAX is slow
URL: https://github.com/apache/incubator-mxnet/issues/14293
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
edfall commented on issue #14293: SOFTMAX is slow
URL:
https://github.com/apache/incubator-mxnet/issues/14293#issuecomment-468549821
I tested, it is much faster. thank you!
This is an automated message from the Apache Git
wkcn commented on issue #14242: Large array support for randint
URL: https://github.com/apache/incubator-mxnet/pull/14242#issuecomment-468548554
Could you please retrigger the CI? Thanks!
This is an automated message from the
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261482565
##
File path: src/operator/quantization/quantized_fully_connected.cc
##
@@
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261482517
##
File path: src/operator/quantization/mkldnn/mkldnn_quantized_fully_connected.cc
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261482177
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
wkcn commented on issue #14293: SOFTMAX is slow
URL:
https://github.com/apache/incubator-mxnet/issues/14293#issuecomment-468546320
There are two softmax implementations in MXNet: softmax and
SoftmaxActivation, the
former is a naive implementation, but the latter is a cudnn one.
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261482044
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261481852
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261481799
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected-inl.h
##
@@
ciyongch commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261481627
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected-inl.h
##
@@
TaoLv commented on issue #14202: [Doc] Start the tutorials for MKL-DNN backend
URL: https://github.com/apache/incubator-mxnet/pull/14202#issuecomment-468543817
@aaronmarkham If yes, do you think we can do that in another PR? In this PR,
I just try to move the document and make it visible
TaoLv commented on issue #14202: [Doc] Start the tutorials for MKL-DNN backend
URL: https://github.com/apache/incubator-mxnet/pull/14202#issuecomment-468543355
@aaronmarkham Thank you for your kind suggestion. Currently, `pip install
mxnet` doesn't have MKL-DNN enabled but `mxnet-mkl`
ZhennanQin commented on a change in pull request #14150: Fix entropy for uint8
URL: https://github.com/apache/incubator-mxnet/pull/14150#discussion_r261477970
##
File path: tests/python/quantization/test_quantization.py
##
@@ -673,8 +673,9 @@ def
ZhennanQin commented on issue #14123: Add int8 data loader
URL: https://github.com/apache/incubator-mxnet/pull/14123#issuecomment-468539956
PR is updated to address all comments received. Please review again. Thanks.
@anirudh2290 @zhreshold @reminisce
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261475674
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
+
+
+
+
mxnet-label-bot commented on issue #14293: SOFTMAX is slow
URL:
https://github.com/apache/incubator-mxnet/issues/14293#issuecomment-468537039
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
that the appropriate MXNet
edfall opened a new issue #14293: SOFTMAX is slow
URL: https://github.com/apache/incubator-mxnet/issues/14293
## Description
Trying to implement an attention model with mxnet, a lot of softmax
operation is used.
I find softmax operation is `too slow`, slower than a convolution
pengzhao-intel commented on a change in pull request #14286: Add examples of
running MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261473796
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
yuxihu commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261471920
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
yuxihu commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261471920
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261471550
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
+
+
+
+
ZhennanQin commented on issue #14123: Add int8 data loader
URL: https://github.com/apache/incubator-mxnet/pull/14123#issuecomment-468524784
@anirudh2290 Thanks for having a look on this. I agree with your point that
user experience is important. Then I will add operator support without
pengzhao-intel commented on a change in pull request #14286: Add examples of
running MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261463572
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261463076
##
File path: src/operator/quantization/quantized_fully_connected.cc
##
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261462614
##
File path: src/operator/quantization/mkldnn/mkldnn_quantized_fully_connected.cc
anirudhacharya commented on issue #14274: added mkldnn dependency for plugin
compile target
URL: https://github.com/apache/incubator-mxnet/pull/14274#issuecomment-468519551
@mxnet-label-bot add [pr-awaiting-review]
This is
anirudhacharya edited a comment on issue #14277: Enhance PartitionGraph
URL: https://github.com/apache/incubator-mxnet/pull/14277#issuecomment-468519240
@mxnet-label-bot add [pr-awaiting-review]
@ZhennanQin could you please rebase your PR and resolve the conflicts?
anirudhacharya commented on issue #14275: Register fake grad to subgraph and
quantized operators
URL: https://github.com/apache/incubator-mxnet/pull/14275#issuecomment-468519500
@mxnet-label-bot add [pr-awaiting-review]
This
anirudhacharya commented on issue #14276: Skip inference only subgraph pass
when gradient is needed.
URL: https://github.com/apache/incubator-mxnet/pull/14276#issuecomment-468519412
@mxnet-label-bot add [pr-awaiting-review]
anirudhacharya commented on issue #14278: Use cudnn for dropout by default
URL: https://github.com/apache/incubator-mxnet/pull/14278#issuecomment-468519222
@mxnet-label-bot add [pr-awaiting-review]
This is an automated
anirudhacharya commented on issue #14277: Enhance PartitionGraph
URL: https://github.com/apache/incubator-mxnet/pull/14277#issuecomment-468519240
@mxnet-label-bot add [pr-awaiting-review]
This is an automated message from the
anirudhacharya commented on issue #14284: Allow gluon blocks to register lists
of blocks as attributes
URL: https://github.com/apache/incubator-mxnet/pull/14284#issuecomment-468519204
@mxnet-label-bot add [pr-awaiting-review]
anirudhacharya commented on issue #14286: Add examples of running MXNet with
Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#issuecomment-468519140
@mxnet-label-bot add [pr-awaiting-review]
This is an
anirudhacharya commented on issue #14287: FIX: flaky test exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#issuecomment-468519110
@mxnet-label-bot add [pr-awaiting-review]
This is an automated
anirudh2290 commented on issue #14123: Add int8 data loader
URL: https://github.com/apache/incubator-mxnet/pull/14123#issuecomment-468518573
I guess there are good reasons for not allowing inplace operations for
inputs. One reason could be that the data allocation for inputs is controlled
anirudh2290 commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261461869
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261461699
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261461488
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261461488
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
pengzhao-intel commented on issue #14128: MKLDNN based Quantized FullyConnected
Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#issuecomment-468516453
@reminisce @zheng-da @anirudh2290 could you help to take a review?
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261460230
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected.cc
##
@@
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261459758
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected-inl.h
##
@@
ZhennanQin commented on a change in pull request #14128: MKLDNN based Quantized
FullyConnected Operator and its fusion
URL: https://github.com/apache/incubator-mxnet/pull/14128#discussion_r261459375
##
File path: src/operator/nn/mkldnn/mkldnn_fully_connected-inl.h
##
@@
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from c319ae5 MXNet Java bug fixes and experience improvement (#14213)
add 992c3c0 [MXNET-1330] Bring
reminisce merged pull request #14270: [MXNET-1330] Bring nnvm::Tuple to
mxnet::Tuple
URL: https://github.com/apache/incubator-mxnet/pull/14270
This is an automated message from the Apache Git Service.
To respond to the
wkcn commented on a change in pull request #14278: Use cudnn for dropout by
default
URL: https://github.com/apache/incubator-mxnet/pull/14278#discussion_r261444796
##
File path: tests/cpp/include/test_op.h
##
@@ -72,7 +72,7 @@ struct GPUStreamScope {
:
wkcn commented on a change in pull request #14278: Use cudnn for dropout by
default
URL: https://github.com/apache/incubator-mxnet/pull/14278#discussion_r261444796
##
File path: tests/cpp/include/test_op.h
##
@@ -72,7 +72,7 @@ struct GPUStreamScope {
:
junrushao1994 opened a new issue #14292: [Flaky] test_operator_gpu.test_adamax
URL: https://github.com/apache/incubator-mxnet/issues/14292
mxnet-label-bot commented on issue #14292: [Flaky] test_operator_gpu.test_adamax
URL:
https://github.com/apache/incubator-mxnet/issues/14292#issuecomment-468497734
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
that the
wkcn commented on issue #14278: Use cudnn for dropout by default
URL: https://github.com/apache/incubator-mxnet/pull/14278#issuecomment-468494491
Thanks for your contribution!
LGTM.
Wait for the finish of CI.
This is an
access2rohit commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261441217
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def
yuxihu commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261440007
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
junrushao1994 commented on a change in pull request #14192: [MXNET-1324] Add
NaiveRunGraph to imperative utils
URL: https://github.com/apache/incubator-mxnet/pull/14192#discussion_r261430888
##
File path: src/imperative/cached_op.cc
##
@@ -262,6 +262,32 @@ std::vector
access2rohit commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261430218
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def
access2rohit commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261430218
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def
mxnet-label-bot commented on issue #14291: [Clojure] mxnet-scala cannot be
loaded without legacy openssl
URL:
https://github.com/apache/incubator-mxnet/issues/14291#issuecomment-468482804
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest
cbowdon opened a new issue #14291: [Clojure] mxnet-scala cannot be loaded
without legacy openssl
URL: https://github.com/apache/incubator-mxnet/issues/14291
## Description
I have a very similar (probably related) issue to #12822 when using the
pre-built jar. Same error message as
access2rohit commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261428445
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def
junrushao1994 opened a new issue #14289: [Flaky]
test_operator_gpu.test_convolution_multiple_streams
URL: https://github.com/apache/incubator-mxnet/issues/14289
mxnet-label-bot commented on issue #14289: [Flaky]
test_operator_gpu.test_convolution_multiple_streams
URL:
https://github.com/apache/incubator-mxnet/issues/14289#issuecomment-468480929
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some
apeforest edited a comment on issue #14253: [RFC] Introducing NumPy-compatible
coding experience into MXNet
URL:
https://github.com/apache/incubator-mxnet/issues/14253#issuecomment-468479524
@reminisce I am fine with keeping those two namespaces till 2.0 for backward
compatibility.
apeforest commented on issue #14253: [RFC] Introducing NumPy-compatible coding
experience into MXNet
URL:
https://github.com/apache/incubator-mxnet/issues/14253#issuecomment-468479524
@reminisce I am fine with keeping those two namespaces till 2.0 for backward
compatibility. Starting
junrushao1994 commented on issue #14288: Flaky test:
test_operator_gpu.test_dropout
URL:
https://github.com/apache/incubator-mxnet/issues/14288#issuecomment-468475371
See also
mxnet-label-bot commented on issue #14288: Flaky test:
test_operator_gpu.test_dropout
URL:
https://github.com/apache/incubator-mxnet/issues/14288#issuecomment-468474933
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
that
junrushao1994 opened a new issue #14288: Flaky test:
test_operator_gpu.test_dropout
URL: https://github.com/apache/incubator-mxnet/issues/14288
ctcyang commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261421989
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
+
+
+
+
apeforest commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261420941
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
zheng-da commented on a change in pull request #14192: [MXNET-1324] Add
NaiveRunGraph to imperative utils
URL: https://github.com/apache/incubator-mxnet/pull/14192#discussion_r261420719
##
File path: src/imperative/cached_op.cc
##
@@ -262,6 +262,32 @@ std::vector
mseth10 commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261419808
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
stu1130 commented on a change in pull request #14259: Add Gluon Transformer Crop
URL: https://github.com/apache/incubator-mxnet/pull/14259#discussion_r261419411
##
File path: python/mxnet/gluon/data/vision/transforms.py
##
@@ -228,6 +228,54 @@ def forward(self, x):
yuxihu commented on a change in pull request #14287: FIX: flaky test
exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#discussion_r261418060
##
File path: python/mxnet/test_utils.py
##
@@ -1933,7 +1933,7 @@ def chi_square_check(generator,
access2rohit commented on a change in pull request #14259: Add Gluon
Transformer Crop
URL: https://github.com/apache/incubator-mxnet/pull/14259#discussion_r261415307
##
File path: python/mxnet/gluon/data/vision/transforms.py
##
@@ -228,6 +228,54 @@ def forward(self, x):
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261411926
##
File path: example/distributed_training-horovod/mxnet_imagenet_resnet50.py
##
@@ -0,0 +1,456
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261412200
##
File path: example/distributed_training-horovod/mxnet_mnist.py
##
@@ -0,0 +1,142 @@
+# Step
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261413090
##
File path: example/distributed_training-horovod/README.md
##
@@ -0,0 +1,207 @@
+
+
+
+
+
+
+
yuxihu commented on a change in pull request #14286: Add examples of running
MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#discussion_r261412084
##
File path: example/distributed_training-horovod/mxnet_imagenet_resnet50.py
##
@@ -0,0 +1,456
anirudhacharya commented on issue #14262: Fix NaN value comparisons in relu,
max and min ops
URL: https://github.com/apache/incubator-mxnet/pull/14262#issuecomment-468464486
@szha yes I think the relu grad should also be handled in a special way,
thanks for pointing it out.
access2rohit commented on issue #14287: FIX: falky test exponential generator
URL: https://github.com/apache/incubator-mxnet/pull/14287#issuecomment-468463591
@anirudh2290 @eric-haibin-lin @apeforest Can you please review ?
access2rohit opened a new pull request #14287: FIX: falky test exponential
generator
URL: https://github.com/apache/incubator-mxnet/pull/14287
## Description ##
revert success_rate to 0.15(earlier value) since new value is much stricter
and fails random number generator tests.
apeforest commented on issue #14286: Add examples of running MXNet with Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286#issuecomment-468462707
@yuxihu @rahul003 @ctcyang Please help to review.
This is an
apeforest opened a new pull request #14286: Add examples of running MXNet with
Horovod
URL: https://github.com/apache/incubator-mxnet/pull/14286
## Description ##
Added a mnist and an imagenet example to show how to run MXNet with Horovod.
README page is also added.
### Changes
stephenrawls commented on issue #14264: nd.reshape truncate values
URL:
https://github.com/apache/incubator-mxnet/issues/14264#issuecomment-468462262
I'm okay if we add a boolean flag to turn this feature on. I'm even okay if
we move this feature to a different operator if you want 100%
access2rohit commented on issue #13846: Unit tests are broken on OSX
URL:
https://github.com/apache/incubator-mxnet/issues/13846#issuecomment-468459759
@zachgk Thanks for the link.
@stsukrov If you want to use CMAKE only you can try instructions here:
MaxBareiss commented on issue #14263: MXNet Master build for CUDA with DEBUG=1
failing
URL:
https://github.com/apache/incubator-mxnet/issues/14263#issuecomment-468458677
I have the same issue on PPC64LE using master (c319ae57), built with CMake
on Linux. I don't have this issue with
danithaca edited a comment on issue #13598: More fine-grained operator
implementation dispatch & memory planning flow
URL:
https://github.com/apache/incubator-mxnet/issues/13598#issuecomment-468456105
We observed poor inference latency with MKL-DNN (see
danithaca edited a comment on issue #13598: More fine-grained operator
implementation dispatch & memory planning flow
URL:
https://github.com/apache/incubator-mxnet/issues/13598#issuecomment-468456105
We observed poor inference latency with MKL-DNN (see
danithaca commented on issue #13598: More fine-grained operator implementation
dispatch & memory planning flow
URL:
https://github.com/apache/incubator-mxnet/issues/13598#issuecomment-468456105
We observed poor inference latency with MKL-DNN, and maybe this proposal
will help address
junrushao1994 commented on issue #14192: [MXNET-1324] Add NaiveRunGraph to
imperative utils
URL: https://github.com/apache/incubator-mxnet/pull/14192#issuecomment-468455458
@szha Hey I added a unit test for this PR :-)
This
1 - 100 of 145 matches
Mail list logo