jerrywind commented on issue #7235: How to use a classification pretrain model
to train a triple-loss?
URL:
https://github.com/apache/incubator-mxnet/issues/7235#issuecomment-319010069
@thinxer @pluskid @darxriggs @ry @jamesliu
szha opened a new pull request #7265: add getitem to container layers
URL: https://github.com/apache/incubator-mxnet/pull/7265
This is to add `__getitem__` to the container layers so that it's easier to
locate the child layers along with their weights. For example:
```
In [1]: from
leezu opened a new pull request #7263: Fix gluon zero grad
URL: https://github.com/apache/incubator-mxnet/pull/7263
self._grad is a dictionary. `for i in self._grad:` iterates over keys.
This is an automated message from the
szha opened a new pull request #7264: gluon conv rnns
URL: https://github.com/apache/incubator-mxnet/pull/7264
@sxjscience
This is an automated message from the Apache Git Service.
To respond to the message, please log on
sxjscience commented on a change in pull request #7264: gluon conv rnns
URL: https://github.com/apache/incubator-mxnet/pull/7264#discussion_r130315604
##
File path: python/mxnet/gluon/rnn/rnn_cell.py
##
@@ -777,3 +785,329 @@ def unroll(self, length, inputs,
janejane11 opened a new issue #7262: use R3.3.0 and im2rec.py to run cnn and
always failed
URL: https://github.com/apache/incubator-mxnet/issues/7262
this is my code below,I have run it many times,but it always failed and my
computer stopped working, I check my memories,my memories are
al-rigazzi commented on issue #6548: [feature req / question ] slurm scheduler
support for launch.py? (distrubted processing / multi-machines support)
URL:
https://github.com/apache/incubator-mxnet/issues/6548#issuecomment-319025647
Hi, yes, indeed the module has to be used for launch.py
sxjscience commented on a change in pull request #7264: gluon conv rnns
URL: https://github.com/apache/incubator-mxnet/pull/7264#discussion_r130318613
##
File path: tests/python/unittest/test_gluon_rnn.py
##
@@ -192,6 +196,59 @@ def test_rnn_cells():
formath commented on issue #7184: Is there any monitor tool for cluster system
resources ?
URL:
https://github.com/apache/incubator-mxnet/issues/7184#issuecomment-319034280
I think `MXNet` should be just for computation. Other parts like resource
management and schedule should be done in
Harmonicahappy opened a new pull request #7266: Fix a spelling mistake
URL: https://github.com/apache/incubator-mxnet/pull/7266
This is an automated message from the Apache Git Service.
To respond to the message, please log
nikrao commented on issue #7251: update sparse ndarray api
URL: https://github.com/apache/incubator-mxnet/pull/7251#issuecomment-319118001
It's best to keep it consistent with MXNet nomenclature in general, so if
it's "X" and "BaseX" for other classes, I think it's nice to have
piiswrong closed pull request #7266: Fix a spelling mistake
URL: https://github.com/apache/incubator-mxnet/pull/7266
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub
piiswrong commented on issue #7265: add getitem to container layers
URL: https://github.com/apache/incubator-mxnet/pull/7265#issuecomment-319119975
also add len
This is an automated message from the Apache Git Service.
To
sergeykolychev commented on issue #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227#issuecomment-319125873
@eric-haibin-lin I'll do it tomorrow
Can't do it easily, the sparse branch is very old and the perl state is as
of
sergeykolychev commented on issue #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227#issuecomment-319125873
@eric-haibin-lin
Can't do it easily, the sparse branch is very old and the perl state is as
of May 14
Are you
skywayoo closed issue #6821: im2rec.py can not work
URL: https://github.com/apache/incubator-mxnet/issues/6821
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and
leezu opened a new issue #7268: Autograd retain_graph=True bugs
URL: https://github.com/apache/incubator-mxnet/issues/7268
Consider the following example
```
import mxnet as mx
from mxnet import autograd
from mxnet import gluon
encoder =
piiswrong closed pull request #7263: Fix gluon zero grad
URL: https://github.com/apache/incubator-mxnet/pull/7263
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and
Bumblebee1964 closed issue #6726: How to save a learned network in C++ API
URL: https://github.com/apache/incubator-mxnet/issues/6726
This is an automated message from the Apache Git Service.
To respond to the message,
Bumblebee1964 opened a new issue #7267: Looking for a C++ alternative for
symbol.get_internals() (in Python API)
URL: https://github.com/apache/incubator-mxnet/issues/7267
For bugs or installation issues, please provide the following information.
The more information you provide, the
mli opened a new pull request #7271: [Doc] Add the new doc link to to hybrid.md
URL: https://github.com/apache/incubator-mxnet/pull/7271
This is an automated message from the Apache Git Service.
To respond to the message,
b0noI opened a new issue #7272: jenkins/incubator-mxnet pipeline is broken on
the master
URL: https://github.com/apache/incubator-mxnet/issues/7272
https://builds.apache.org/blue/organizations/jenkins/incubator-mxnet/detail/master/83/pipeline/
output:
> [amalgamation] Running
stefanhenneking commented on a change in pull request #7223: GPU implementation
of cast_storage (dense to rsp)
URL: https://github.com/apache/incubator-mxnet/pull/7223#discussion_r130438387
##
File path: src/operator/tensor/cast_storage-inl.cuh
##
@@ -14,13 +14,288 @@
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130272497
##
File path: python/mxnet/symbol.py
##
@@ -17,8 +17,10 @@
from .base import NDArrayHandle,
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130258392
##
File path: python/mxnet/kvstore.py
##
@@ -214,11 +217,88 @@ def pull(self, key, out=None,
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130406399
##
File path: src/operator/elemwise_op_common.h
##
@@ -90,6 +126,33 @@ inline bool ElemwiseType(const
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130404224
##
File path: src/executor/attach_op_execs_pass.cc
##
@@ -22,28 +23,39 @@ const OperatorProperty*
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130258387
##
File path: python/mxnet/kvstore.py
##
@@ -214,11 +217,88 @@ def pull(self, key, out=None,
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130258581
##
File path: python/mxnet/ndarray/ndarray_utils.py
##
@@ -0,0 +1,127 @@
+# coding: utf-8
Review
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130403724
##
File path: src/c_api/c_api_ndarray.cc
##
@@ -301,10 +384,25 @@ void PushOperator(const OpStatePtr&
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130272291
##
File path: python/mxnet/ndarray/sparse_ndarray.py
##
@@ -0,0 +1,637 @@
+# coding: utf-8
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130272349
##
File path: python/mxnet/ndarray/sparse_ndarray.py
##
@@ -0,0 +1,637 @@
+# coding: utf-8
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130257348
##
File path: include/mxnet/kvstore.h
##
@@ -155,6 +156,29 @@ class KVStore {
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130272244
##
File path: python/mxnet/ndarray/sparse_ndarray.py
##
@@ -0,0 +1,637 @@
+# coding: utf-8
piiswrong commented on a change in pull request #7082: Sparse Tensor: request
for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130403722
##
File path: src/c_api/c_api_ndarray.cc
##
@@ -301,10 +384,25 @@ void PushOperator(const OpStatePtr&
leezu opened a new pull request #7261: Add grad_req parameter to Block that is
passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261
This allows specifying the grad_req for Blocks that is then used when
creating Parameters.
piiswrong commented on a change in pull request #7261: Add grad_req parameter
to Block that is passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#discussion_r130414628
##
File path: python/mxnet/gluon/parameter.py
##
@@ -374,6 +374,15 @@
piiswrong commented on a change in pull request #7265: add getitem to container
layers
URL: https://github.com/apache/incubator-mxnet/pull/7265#discussion_r130414949
##
File path: python/mxnet/gluon/block.py
##
@@ -162,6 +162,9 @@ def __setattr__(self, name, value):
szha commented on a change in pull request #7264: gluon conv rnns
URL: https://github.com/apache/incubator-mxnet/pull/7264#discussion_r130425435
##
File path: tests/python/unittest/test_gluon_rnn.py
##
@@ -192,6 +196,59 @@ def test_rnn_cells():
rahul003 commented on a change in pull request #7236: change unnecessarily
longer array
URL: https://github.com/apache/incubator-mxnet/pull/7236#discussion_r130406293
##
File path: src/io/iter_image_recordio_2.cc
##
@@ -360,10 +360,10 @@ inline void
thirdwing commented on issue #7262: use R3.3.0 and im2rec.py to run cnn and
always failed
URL:
https://github.com/apache/incubator-mxnet/issues/7262#issuecomment-319146812
Can you tell us how installed the mxnet and the version of mxnet?
Please also show us the error message and
sagaragrawal13 opened a new issue #7269: Bayesian Optimization for MXNET
URL: https://github.com/apache/incubator-mxnet/issues/7269
Could you please provide a good reference to get started with Bayesian
Optimization on models built through MXNET R package.
I plan to use Bayesian
sergeykolychev commented on issue #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227#issuecomment-318976336
@eric-haibin-lin I'll do it tomorrow
This is an
eric-haibin-lin closed pull request #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227
This is an automated message from the Apache Git Service.
To respond to
eric-haibin-lin commented on issue #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227#issuecomment-318976432
awesome thank you very much!
This is an automated
eric-haibin-lin commented on issue #7227: MXExecutorSimpleBindEx for backward
compatability
URL: https://github.com/apache/incubator-mxnet/pull/7227#issuecomment-318975582
@sergeykolychev Oh I misunderstood your message. I thought you're going to
make the change to master branch. If you
piiswrong closed pull request #7261: Add grad_req parameter to Block that is
passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261
This is an automated message from the Apache Git Service.
To
piiswrong closed pull request #7271: [Doc] Add the new doc link to to hybrid.md
URL: https://github.com/apache/incubator-mxnet/pull/7271
This is an automated message from the Apache Git Service.
To respond to the message,
thirdwing commented on issue #6925: Layer Specific Learning Rate For R
URL:
https://github.com/apache/incubator-mxnet/issues/6925#issuecomment-319213255
Currently, this hasn't been support in R.
This is an automated message
thirdwing commented on issue #7019: R Character RNN program question
URL:
https://github.com/apache/incubator-mxnet/issues/7019#issuecomment-319212189
I think the best method to get the dimension is to write a small helper
function. Just like we I did in the unit test
thirdwing closed issue #6918: mxnet R 3.4 installation for CentOS release 6.9
(Final)
URL: https://github.com/apache/incubator-mxnet/issues/6918
This is an automated message from the Apache Git Service.
To respond to the
thirdwing closed issue #6945: predict(model, val) produces fewer number of
samples compared to original sample size
URL: https://github.com/apache/incubator-mxnet/issues/6945
This is an automated message from the Apache
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318986185
Sure, but thats in my opinion just a workaround. In that case one needs to
manually create all parameter
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318986185
Sure, but thats in my opinion just a workaround. In that case one needs to
manually create all parameter
piiswrong commented on issue #7261: Add grad_req parameter to Block that is
passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318986721
There are too many attributes for each parameter and too many parameters.
Adding one argument for each
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318988444
Arguably though the library should take care of this and one shouldn't rely
on the user setting it for each
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318988444
Arguably though the library should take care of this and one shouldn't rely
on the user setting it for each
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318988444
Arguably though the library should take care of this and one shouldn't rely
on the user setting it for each
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318988444
Arguably though the library should take care of this and one shouldn't rely
on the user setting it for each
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318988444
Arguably though the library should take care of this and one shouldn't rely
on the user setting it for each
piiswrong commented on issue #7261: Add grad_req parameter to Block that is
passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318989119
you can do it with
params = ParameterDict('conv0')
params.get('weight', grad_req = 'null')
conv0
piiswrong commented on issue #7261: Add grad_req parameter to Block that is
passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318989918
we can also add a setattr method to parameterdict that sets the attributes
for each parameter
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318991048
The idea behind this PR is that as Blocks can be arbitrarily composed, and
as each block abstracts away the
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318991048
The idea behind this PR is that as Blocks can be arbitrarily composed, and
as each block abstracts away the
leezu commented on issue #7261: Add grad_req parameter to Block that is passed
to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#issuecomment-318992462
Regarding setattr for parameterdict, it should work if we always want to
change the attributes of all
leezu commented on issue #7268: Autograd retain_graph=True bugs
URL:
https://github.com/apache/incubator-mxnet/issues/7268#issuecomment-319238610
Thanks. Using `autograd.backward` works for me. Your first proposed solution
doesn't allow for making manual changes to the grad arrays before
This is an automated email from the ASF dual-hosted git repository.
qkou 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 92a93b8 [R][DOC] make sure all code in
bhavinthaker opened a new pull request #7276: Add support for port package
manager on macOS
URL: https://github.com/apache/incubator-mxnet/pull/7276
add support for port package manager along with brew;
show compile-time configs;
print run-time env vars;
print link to MXNet
bhavinthaker commented on issue #7166: add support for port package manager
URL: https://github.com/apache/incubator-mxnet/pull/7166#issuecomment-319243998
I have modified the script to include user's PATH to have precedence over
the script's PATH so that the user's python (for example,
bhavinthaker commented on issue #7166: add support for port package manager
URL: https://github.com/apache/incubator-mxnet/pull/7166#issuecomment-319244661
See: https://github.com/apache/incubator-mxnet/pull/7276
This is an
Harmonicahappy opened a new pull request #7277: Fix a spelling mistake
URL: https://github.com/apache/incubator-mxnet/pull/7277
This is an automated message from the Apache Git Service.
To respond to the message, please log
leezu commented on issue #7268: Autograd retain_graph=True bugs
URL:
https://github.com/apache/incubator-mxnet/issues/7268#issuecomment-319238610
Thanks. Using `autograd.backward` works for me. Your first proposed solution
doesn't allow for making manual changes to the grad arrays before
leezu commented on a change in pull request #7261: Add grad_req parameter to
Block that is passed to ParameterDict.get
URL: https://github.com/apache/incubator-mxnet/pull/7261#discussion_r130500156
##
File path: python/mxnet/gluon/parameter.py
##
@@ -374,6 +374,15 @@ def
This is an automated email from the ASF dual-hosted git repository.
lxn2 pushed a change to branch asf-site
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-test.git.
No new revisions were added by this update.
Summary of changes:
--
To stop receiving notification emails
b0noI commented on issue #7272: jenkins/incubator-mxnet pipeline is broken on
the master
URL:
https://github.com/apache/incubator-mxnet/issues/7272#issuecomment-319229778
I was not able to reproduce it by running manually.
piiswrong closed pull request #7265: add getitem to container layers
URL: https://github.com/apache/incubator-mxnet/pull/7265
This is an automated message from the Apache Git Service.
To respond to the message, please log
piiswrong closed pull request #7277: Fix a spelling mistake
URL: https://github.com/apache/incubator-mxnet/pull/7277
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub
piiswrong closed pull request #7275: Jenkins correctly cleans build when
incremental build failed
URL: https://github.com/apache/incubator-mxnet/pull/7275
This is an automated message from the Apache Git Service.
To
This is an automated email from the ASF dual-hosted git repository.
jxie pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from f965542 Fix a spelling mistake (#7277)
add f187d5b Now Jenkins correctly executes clean command when
piiswrong closed pull request #7260: Gluon RNN fixes for seqlen 1
URL: https://github.com/apache/incubator-mxnet/pull/7260
This is an automated message from the Apache Git Service.
To respond to the message, please log on
piiswrong closed pull request #7270: Update cub for CUDA 9
URL: https://github.com/apache/incubator-mxnet/pull/7270
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub
This is an automated email from the ASF dual-hosted git repository.
jxie 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 43ea355 Update cub for CUDA 9 (#7270)
dsqx71 commented on issue #7203: Add document for BilinearSampler Op
URL: https://github.com/apache/incubator-mxnet/pull/7203#issuecomment-319265849
@piiswrong Looks like this PR is nothing to do with the error happened in
CI. Please check and merge.
piiswrong closed pull request #7230: Update NNVM
URL: https://github.com/apache/incubator-mxnet/pull/7230
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
This is an automated email from the ASF dual-hosted git repository.
jxie 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 1784f62 fix random sized crop (#7173)
This is an automated email from the ASF dual-hosted git repository.
jxie pushed a commit to branch sparse
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/sparse by this push:
new f77f31a Operator add_n for row sparse
eric-haibin-lin commented on issue #7082: Sparse Tensor: request for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#issuecomment-319272276
@jermainewang Thanks for the great reference! I can see how the proposed
storage format in the paper could be used to express many
eric-haibin-lin commented on a change in pull request #7082: Sparse Tensor:
request for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130522825
##
File path: include/mxnet/kvstore.h
##
@@ -155,6 +156,29 @@ class KVStore {
eric-haibin-lin commented on a change in pull request #7082: Sparse Tensor:
request for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130522936
##
File path: python/mxnet/_ctypes/ndarray.py
##
@@ -61,22 +70,29 @@ def
YvesHarrison opened a new issue #7279: Batch processing for rcnn prediction
URL: https://github.com/apache/incubator-mxnet/issues/7279
I want to send pictures to rcnn network for prediction by batch. Does the
current rcnn function like im_detect() and Predictor() support this. If they do
eric-haibin-lin commented on a change in pull request #7082: Sparse Tensor:
request for reviews
URL: https://github.com/apache/incubator-mxnet/pull/7082#discussion_r130523297
##
File path: python/mxnet/ndarray/sparse_ndarray.py
##
@@ -0,0 +1,637 @@
+# coding: utf-8
91 matches
Mail list logo