mseth10 commented on issue #14016: fix test_depthwise_convoltuion for
occasional CI failures
URL: https://github.com/apache/incubator-mxnet/pull/14016#issuecomment-458673690
@apeforest We already do that. We actually compare between default contexts.
So when default context is cpu, we
lichen11 closed issue #14021: Load mxnet model trained in R to Python errors
imply inconsistency in R and Python or bug in mx.io.ImageRecordIter
URL: https://github.com/apache/incubator-mxnet/issues/14021
This is an
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 22ccda7 Bump the publish
piyushghai commented on issue #13853: Error while running the mxnet spark
examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458657852
Here's what I did to build the libmxnet.so from source :
```
git clone --recursive
anirudhacharya commented on issue #14021: Load mxnet model trained in R to
Python errors imply inconsistency in R and Python or bug in
mx.io.ImageRecordIter
URL:
https://github.com/apache/incubator-mxnet/issues/14021#issuecomment-458694326
@lichen11 can you please share a reproducible
vdantu commented on issue #14021: Load mxnet model trained in R to Python
errors imply inconsistency in R and Python or bug in mx.io.ImageRecordIter
URL:
https://github.com/apache/incubator-mxnet/issues/14021#issuecomment-458686266
@anirudhacharya : Would you be able to answer about
lichen11 opened a new issue #14021: Load mxnet model trained in R to Python
errors imply inconsistency in R and Python or bug in mx.io.ImageRecordIter
URL: https://github.com/apache/incubator-mxnet/issues/14021
I trained a model in R and want to load it in Python. I found that 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 5ef7ce6 Bump the publish
aashudwivedi edited a comment on issue #13853: Error while running the mxnet
spark examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458672045
@piyushghai
I am building it on macOS High Sierra 10.13.6.
Here's what I do to build the
hetong007 commented on issue #13859: Newly updated Makefile generate path errors
URL:
https://github.com/apache/incubator-mxnet/issues/13859#issuecomment-458723248
@lichen11 It sounds like you are having two different R installations. They
may partially share the `.libPaths()`, which may
aashudwivedi commented on issue #13853: Error while running the mxnet spark
examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458672045
@piyushghai
I am building it on macOS High Sierra 10.13.6.
Here's what I do to build the
aashudwivedi edited a comment on issue #13853: Error while running the mxnet
spark examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458672045
@piyushghai
I am building it on macOS High Sierra 10.13.6.
Here's what I do to build the
stu1130 commented on a change in pull request #14014: [WIP] Export resize and
support batch size
URL: https://github.com/apache/incubator-mxnet/pull/14014#discussion_r252028412
##
File path: src/operator/contrib/bilinear_resize.cu
##
@@ -181,9 +108,10 @@ void
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 b083a28 Bump the publish
lichen11 commented on issue #14021: Load mxnet model trained in R to Python
errors imply inconsistency in R and Python or bug in mx.io.ImageRecordIter
URL:
https://github.com/apache/incubator-mxnet/issues/14021#issuecomment-458711890
I figured out what went wrong. The problem was that
lichen11 commented on issue #13859: Newly updated Makefile generate path errors
URL:
https://github.com/apache/incubator-mxnet/issues/13859#issuecomment-458719621
Yes I installed roxygen2 on Rserver. I am still receiving the error
Loading required package: mxnet
[1] "Loading
vdantu commented on issue #14019: SequenceLast undefined behavior
URL:
https://github.com/apache/incubator-mxnet/issues/14019#issuecomment-458611718
@stephenrawls : I am not an expert here but index not being checked seems
more like an error that isn't handled. I would assume you could
PistonY opened a new issue #13709: Why FP16 training speed is too slow on Tesla
T4 in Gluon?
URL: https://github.com/apache/incubator-mxnet/issues/13709
Hi, I tried to train with FP16 on Tesla T4, but it's speed is slower than
GTX 1070 with FP32.
Could you please give me some suggests
stephenrawls opened a new issue #14019: SequenceLast undefined behavior
URL: https://github.com/apache/incubator-mxnet/issues/14019
SequenceLast currently causes undefined behavior if `use_sequence_length` is
true and any of the values in the `sequence_length` array are 0.
Proof:
PistonY commented on issue #13709: Why FP16 training speed is too slow on Tesla
T4 in Gluon?
URL:
https://github.com/apache/incubator-mxnet/issues/13709#issuecomment-458579142
Ok,I’ll test it later.
This is an automated
DickJC123 commented on issue #14006: Dual stream cudnn Convolution backward()
with MXNET_GPU_WORKER_NSTREAMS=2.
URL: https://github.com/apache/incubator-mxnet/pull/14006#issuecomment-458656852
I think we're not at the point of having the framework be smart enough to
make this trade-off,
piyushghai commented on issue #13853: Error while running the mxnet spark
examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458642620
@aashudwivedi Thanks for making the fix to the calsspath problem for the
Spark issue.
Can you point
aashudwivedi opened a new pull request #14020: Fix jar path and add missing
ones for spark jobs
URL: https://github.com/apache/incubator-mxnet/pull/14020
Fix path of jars / add missing jars in spark job
remove print, reduce clutter
## Description ##
(Brief description on
aashudwivedi commented on issue #13853: Error while running the mxnet spark
examples and test cases
URL:
https://github.com/apache/incubator-mxnet/issues/13853#issuecomment-458610016
@piyushghai The path of the jars are still incorrect / incomplete. I've
fixed this in this pull request,
mxnet-label-bot commented on issue #14019: SequenceLast undefined behavior
URL:
https://github.com/apache/incubator-mxnet/issues/14019#issuecomment-458579520
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
that the
vdantu commented on issue #14017: Loading parameters from pretrained gluon model
URL:
https://github.com/apache/incubator-mxnet/issues/14017#issuecomment-458591757
@mxnet-label-bot add [gluon, question]
This is an automated
DickJC123 commented on issue #13749: Add NHWC layout support to Pooling (cpu,
gpu cuda, gpu cuDNN)
URL: https://github.com/apache/incubator-mxnet/pull/13749#issuecomment-458645654
Thanks for your detailed thoughts. I view this PR as filling in a missing
piece in the MXNet support for
sandeep-krishnamurthy commented on a change in pull request #14014: [WIP]
Export resize and support batch size
URL: https://github.com/apache/incubator-mxnet/pull/14014#discussion_r251948372
##
File path: src/operator/image/resize-inl.h
##
@@ -0,0 +1,217 @@
+/*
+*
sandeep-krishnamurthy commented on a change in pull request #14014: [WIP]
Export resize and support batch size
URL: https://github.com/apache/incubator-mxnet/pull/14014#discussion_r251946245
##
File path: src/operator/contrib/bilinear_resize.cu
##
@@ -181,9 +108,10 @@
sandeep-krishnamurthy commented on a change in pull request #14014: [WIP]
Export resize and support batch size
URL: https://github.com/apache/incubator-mxnet/pull/14014#discussion_r251947622
##
File path: src/operator/image/image_utils.h
##
@@ -0,0 +1,164 @@
+/*
+*
hetong007 commented on issue #13859: Newly updated Makefile generate path errors
URL:
https://github.com/apache/incubator-mxnet/issues/13859#issuecomment-458751327
I'm saying that your R and R server are different R installation, but they
could have shared `.libPaths()`. Therefore some
sandeep-krishnamurthy commented on issue #13837: Image ToTensor operator - GPU
support, 3D/4D inputs
URL: https://github.com/apache/incubator-mxnet/pull/13837#issuecomment-458751847
@szha - I have addressed your comments. Can you please take a look at this
PR? Thanks.
anirudhacharya commented on issue #13589: Can't build R-package from master
since #13062
URL:
https://github.com/apache/incubator-mxnet/issues/13589#issuecomment-458770307
@jeremiedb indeed it should have. Also I notice the current recipe for
building R-package in the Makefile has a few
DickJC123 commented on issue #13142: [WIP] cudnnFind** API does not return
deterministic algorithms
URL: https://github.com/apache/incubator-mxnet/pull/13142#issuecomment-458775835
I have a possible explanation. When cudnnFind measures the runtime of algos
that have similar runtimes, the
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 e37ff53 Return value docs for
szha merged pull request #13994: Return value docs for nd.random.* and
sym.random.*
URL: https://github.com/apache/incubator-mxnet/pull/13994
This is an automated message from the Apache Git Service.
To respond to the
szha commented on issue #13994: Return value docs for nd.random.* and
sym.random.*
URL: https://github.com/apache/incubator-mxnet/pull/13994#issuecomment-458780779
Thanks for the contribution, @zboldyga. The preview is at
juliusshufan commented on a change in pull request #13895: Add a inference
script providing both accuracy and benchmark result for original wide_n_deep
example
URL: https://github.com/apache/incubator-mxnet/pull/13895#discussion_r252095585
##
File path:
matthewberryman closed issue #13965: mxnet_coreml_converter unsupported type
error
URL: https://github.com/apache/incubator-mxnet/issues/13965
This is an automated message from the Apache Git Service.
To respond to the
lanking520 commented on a change in pull request #13807: [MXNET-1180] Java
Image API
URL: https://github.com/apache/incubator-mxnet/pull/13807#discussion_r252053042
##
File path:
scala-package/core/src/main/scala/org/apache/mxnet/javaapi/Image.scala
##
@@ -0,0 +1,95 @@
mseth10 commented on issue #12203: flaky test:
test_operator_gpu.test_depthwise_convolution
URL:
https://github.com/apache/incubator-mxnet/issues/12203#issuecomment-45874
Fix in #14016
This is an automated message from
lichen11 commented on issue #13859: Newly updated Makefile generate path errors
URL:
https://github.com/apache/incubator-mxnet/issues/13859#issuecomment-458749938
I checked the R versions. I am using R 3.5.1 (2018-07-02) Feather Spray on
both R and R server.
szha commented on issue #14006: Dual stream cudnn Convolution backward() with
MXNET_GPU_WORKER_NSTREAMS=2.
URL: https://github.com/apache/incubator-mxnet/pull/14006#issuecomment-458767213
Will take a look soon. RNN can benefit from the additional stream too.
ZhennanQin commented on issue #14022: Segmentation fault during inference with
quantization
URL:
https://github.com/apache/incubator-mxnet/issues/14022#issuecomment-458780253
@nuslq log indicates that there's no implementation for quantized
convolution. MXNet quantization flow only works
juliusshufan commented on issue #13895: Add a inference script providing both
accuracy and benchmark result for original wide_n_deep example
URL: https://github.com/apache/incubator-mxnet/pull/13895#issuecomment-458786487
@szha and @TaoLv @pengzhao-intel @larroy may I have your approval
wkcn commented on issue #9880: TVM bridge support to JIT NDArray Function by TVM
URL: https://github.com/apache/incubator-mxnet/pull/9880#issuecomment-458790311
@tqchen Hi!
I would like to use `MXTVMBridge` to accelerate my project, but there is a
flaky problem.
I wrote a
PistonY edited a comment on issue #13709: Why FP16 training speed is too slow
on Tesla T4 in Gluon?
URL:
https://github.com/apache/incubator-mxnet/issues/13709#issuecomment-458792488
I tried to use fixed input,FP32 work well but FP16 out of memory.
This is my script.
```python
PistonY commented on issue #13709: Why FP16 training speed is too slow on Tesla
T4 in Gluon?
URL:
https://github.com/apache/incubator-mxnet/issues/13709#issuecomment-458792488
I tried to fixed input,FP32 work well but FP16 out of memory.
This is my script.
```python
from mxnet
mseth10 commented on issue #14016: fix test_depthwise_convoltuion for
occasional CI failures
URL: https://github.com/apache/incubator-mxnet/pull/14016#issuecomment-458749553
@Vikas89 I checked. We do have GPU no cuDNN environment running on our CI. I
will make sure this test is running on
matthewberryman commented on issue #13965: mxnet_coreml_converter unsupported
type error
URL:
https://github.com/apache/incubator-mxnet/issues/13965#issuecomment-458783532
Hi @ChaiBapchya - many thanks for opening the feature request. Closing this
issue in favour of that one.
juliusshufan commented on a change in pull request #13895: Add a inference
script providing both accuracy and benchmark result for original wide_n_deep
example
URL: https://github.com/apache/incubator-mxnet/pull/13895#discussion_r252095585
##
File path:
pengzhao-intel commented on issue #14022: Segmentation fault during inference
with quantization
URL:
https://github.com/apache/incubator-mxnet/issues/14022#issuecomment-458792386
Thanks @ZhennanQin :)
@nuslq you can refer
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 4b69a37 Bump the publish
lanking520 commented on a change in pull request #13816: Add default
parameters for Scala NDArray.arange
URL: https://github.com/apache/incubator-mxnet/pull/13816#discussion_r252084337
##
File path: scala-package/core/src/main/scala/org/apache/mxnet/NDArray.scala
##
@@
juliusshufan commented on a change in pull request #13895: Add a inference
script providing both accuracy and benchmark result for original wide_n_deep
example
URL: https://github.com/apache/incubator-mxnet/pull/13895#discussion_r250185628
##
File path:
nuslq opened a new issue #14022: Segmentation fault during inference with
quantization
URL: https://github.com/apache/incubator-mxnet/issues/14022
Note: Providing complete information in the most concise form is the best
way to get help. This issue template serves as the checklist for
eric-haibin-lin closed issue #11796: Batch_dot does not support FP16 well
URL: https://github.com/apache/incubator-mxnet/issues/11796
This is an automated message from the Apache Git Service.
To respond to the message,
eric-haibin-lin commented on issue #11796: Batch_dot does not support FP16 well
URL:
https://github.com/apache/incubator-mxnet/issues/11796#issuecomment-458821472
Added in https://github.com/apache/incubator-mxnet/pull/13716
This is an automated email from the ASF dual-hosted git repository.
iblis 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 36a3cb8 Julia: split ndarray.jl into
This is an automated email from the ASF dual-hosted git repository.
iblis pushed a change to branch ib/jl-ndarray-split
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
was 548636d Julia: split ndarray.jl into several snippets
The revisions that were on this branch
iblis17 merged pull request #14001: Julia: split ndarray.jl into several
snippets
URL: https://github.com/apache/incubator-mxnet/pull/14001
This is an automated message from the Apache Git Service.
To respond to the
szha edited a comment on issue #13964: Addresses comments in runtime feature
discovery API
URL: https://github.com/apache/incubator-mxnet/pull/13964#issuecomment-457739621
@larroy I will need more information to offer advice on how to make it work.
Alternatively, I can fork the branch and
marcoabreu commented on issue #13589: Can't build R-package from master since
#13062
URL:
https://github.com/apache/incubator-mxnet/issues/13589#issuecomment-458820161
Yeah
This is an automated message from the Apache Git
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 cb0c9f2 Bump the publish
yzhliu commented on a change in pull request #13807: [MXNET-1180] Java Image API
URL: https://github.com/apache/incubator-mxnet/pull/13807#discussion_r252114510
##
File path:
scala-package/core/src/main/scala/org/apache/mxnet/javaapi/Image.scala
##
@@ -0,0 +1,95 @@
+/*
+
anirudhacharya opened a new pull request #14023: [R] Add NAG optimizer
URL: https://github.com/apache/incubator-mxnet/pull/14023
## Description ##
Nesterov accelerated SGD for R API.
## Checklist ##
### Essentials ###
Please feel free to remove inapplicable items for your
rajeshii commented on issue #13952: Fix MXNet R package build
URL: https://github.com/apache/incubator-mxnet/pull/13952#issuecomment-458838497
Rebase three times with different kinds of CI failures:(
This is an automated
marcoabreu commented on issue #14006: Dual stream cudnn Convolution backward()
with MXNET_GPU_WORKER_NSTREAMS=2.
URL: https://github.com/apache/incubator-mxnet/pull/14006#issuecomment-458820344
Good point, thanks
This is an
iblis17 opened a new pull request #14024: Julia: split symbolic-node.jl into
several snippets
URL: https://github.com/apache/incubator-mxnet/pull/14024
- `symbolic-node/type.jl`
- `symbolic-node/show.jl`
- `symbolic-node/arithmetic.jl`
- `symbolic-node/io.jl`
-
anirudhacharya commented on issue #13609: [MXNET-1258]fix unittest for ROIAlign
Operator
URL: https://github.com/apache/incubator-mxnet/pull/13609#issuecomment-458454950
@sandeep-krishnamurthy have already approved this PR, above.
szha commented on issue #13896: Cudnn dropout
URL: https://github.com/apache/incubator-mxnet/pull/13896#issuecomment-458445531
As of 7f76582 the speed-up still depends on the size of the input because of
set descriptor overhead. I'm currently looking into adding a runtime resource
manager
szha edited a comment on issue #13896: Cudnn dropout
URL: https://github.com/apache/incubator-mxnet/pull/13896#issuecomment-458445531
As of 7f76582 the speed-up still depends on the size of the input because of
set descriptor overhead. I'm currently looking into adding a runtime resource
szha edited a comment on issue #13896: Cudnn dropout
URL: https://github.com/apache/incubator-mxnet/pull/13896#issuecomment-458445531
As of 7f76582 the speed-up still depends on the size of the input because of
set descriptor overhead. Since it’s a stateful op, it’s possible to utilize
anirudhacharya commented on issue #13401: [MXNET-1227] Adding CornerPooling
operator
URL: https://github.com/apache/incubator-mxnet/pull/13401#issuecomment-458454670
@sandeep-krishnamurthy As mentioned above my comments have already been
addressed. If @anirudh2290's comments are
rajeshii commented on issue #13952: Fix MXNet R package build
URL: https://github.com/apache/incubator-mxnet/pull/13952#issuecomment-458501799
@marcoabreu addressed your comments. It seems that all tests have passed but
the CI is pending...
kiendang commented on a change in pull request #13816: Add default parameters
for Scala NDArray.arange
URL: https://github.com/apache/incubator-mxnet/pull/13816#discussion_r251755999
##
File path: scala-package/core/src/main/scala/org/apache/mxnet/NDArray.scala
##
@@
This is an automated email from the ASF dual-hosted git repository.
marcoabreu 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 b9e02ab remove debug; add support
marcoabreu merged pull request #14015: website publish updates
URL: https://github.com/apache/incubator-mxnet/pull/14015
This is an automated message from the Apache Git Service.
To respond to the message, please log on
pengzhao-intel commented on issue #13749: Add NHWC layout support to Pooling
(cpu, gpu cuda, gpu cuDNN)
URL: https://github.com/apache/incubator-mxnet/pull/13749#issuecomment-458566007
@DickJC123 It's great to see the improvements and thank you to cover the
MKLDNN backend as well :)
Do
pengzhao-intel commented on issue #14017: Loading parameters from pretrained
gluon model
URL:
https://github.com/apache/incubator-mxnet/issues/14017#issuecomment-458513852
@zhanghang1989 @xinyu-intel to help :)
This is an
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251813404
##
File path: python/mxnet/contrib/quantization.py
##
@@ -80,8 +80,7 @@ def
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251821222
##
File path: include/mxnet/c_api.h
##
@@ -1566,7 +1566,7 @@ MXNET_DLL int
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251822891
##
File path: src/operator/quantization/quantize_v2-inl.h
##
@@ -0,0 +1,220 @@
+/*
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251822762
##
File path: src/operator/quantization/mkldnn/mkldnn_dequantize-inl.h
##
@@ -74,6
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251822255
##
File path: python/mxnet/contrib/quantization.py
##
@@ -80,8 +80,7 @@ def
ZhennanQin commented on a change in pull request #13697: [MKLDNN] Enable signed
int8 support for convolution.
URL: https://github.com/apache/incubator-mxnet/pull/13697#discussion_r251822984
##
File path: src/operator/quantization/quantize_v2.cc
##
@@ -0,0 +1,106 @@
+/*
+
86 matches
Mail list logo
