ZiyueHuang commented on issue #8338: master branch cannot build on centos 7
with cuda-8.0
URL:
https://github.com/apache/incubator-mxnet/issues/8338#issuecomment-338387761
I tried the latest master but it fails too.
I replaced smooth_l1_* with returning DType(0), the warnings
0x6a62 opened a new pull request #8376: Fix Typo (classification)
URL: https://github.com/apache/incubator-mxnet/pull/8376
## Description ##
Fix a typo in the example readme.
## Checklist ##
### Essentials ###
- [NA ] Passed code style checking (`make lint`)
- [X] Changes
piiswrong closed pull request #8345: Misc fixes for sparse distributed training
URL: https://github.com/apache/incubator-mxnet/pull/8345
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is
piiswrong commented on a change in pull request #8364: Fix typo
URL: https://github.com/apache/incubator-mxnet/pull/8364#discussion_r146099197
##
File path: example/README.md
##
@@ -62,7 +62,7 @@ If you want to contribute to this list and the examples,
please open a new
piiswrong commented on issue #8373: distribute training in fp16
URL: https://github.com/apache/incubator-mxnet/pull/8373#issuecomment-338371778
@eric-haibin-lin @rahul003 I think this should be merged with the 2bit PR to
make a more general n-bit gradient compression feature.
piiswrong commented on a change in pull request #8364: Fix typo
URL: https://github.com/apache/incubator-mxnet/pull/8364#discussion_r146099200
##
File path: example/README.md
##
@@ -86,7 +86,7 @@ If you want to contribute to this list and the examples,
please open a new
reminisce opened a new pull request #8371: Add note in the doc for using naive
engine in multithreading environment
URL: https://github.com/apache/incubator-mxnet/pull/8371
## Description ##
Add note for using naive engine in multithreading environment per the
request of @bhavinthaker.
jiarenyf commented on issue #8347: CTC Example Problem
URL:
https://github.com/apache/incubator-mxnet/issues/8347#issuecomment-338108475
??
This is an automated message from the Apache Git Service.
To respond to the message,
wisdomdeng opened a new issue #8372: PyPi wheel installation failed on Centos
7.5
URL: https://github.com/apache/incubator-mxnet/issues/8372
System information
```
(mxnet) [ruizhid@cedar5 mxnet]$ lsb_release -a
LSB Version: n/a
Distributor ID: CentOS
Description: CentOS
piiswrong commented on issue #8368: A hybrid model with sync and async in
kvstore
URL: https://github.com/apache/incubator-mxnet/pull/8368#issuecomment-338371317
This is too hacky
This is an automated message from the Apache
eric-haibin-lin commented on issue #5326: Train-accuracy=0.00
URL:
https://github.com/apache/incubator-mxnet/issues/5326#issuecomment-338371172
FYI there's an example for layer-wise pretrain for auto-encoder
jiarenyf commented on issue #8347: CTC Example Problem
URL:
https://github.com/apache/incubator-mxnet/issues/8347#issuecomment-337843827
@pluskid @thinxer
This is an automated message from the Apache Git Service.
To respond
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146099236
##
File path: src/operator/contrib/two_bit_quantize.cc
##
@@ -0,0 +1,122 @@
+/*
+ * Licensed to the Apache
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146099245
##
File path: python/mxnet/gluon/trainer.py
##
@@ -65,7 +74,10 @@ def __init__(self, params, optimizer,
ZiyueHuang opened a new pull request #8374: fix condition when CSRNDArray is
used in NDArrayIter with shuffle=True
URL: https://github.com/apache/incubator-mxnet/pull/8374
## Description ##
Currently if `NDArrayIter({'data':csr}, ..., shuffle=True)`, there is no
`AssertionError`.
eric-haibin-lin commented on issue #8189: Feed forward pass memory leaks (using
htop)
URL:
https://github.com/apache/incubator-mxnet/issues/8189#issuecomment-338371727
forward/backward/update are all asynchronous operation. It just pushes the
operations to the backend engine and returns
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112817
##
File path: python/mxnet/gluon/trainer.py
##
@@ -44,14 +44,23 @@ class Trainer(object):
kvstore :
piiswrong commented on issue #8182: Make gluon.Block cooperative in multiple
inheritance setting
URL: https://github.com/apache/incubator-mxnet/pull/8182#issuecomment-338425611
Let's put this on hold to see if there is wide demand for this feature.
cjolivier01 commented on issue #8361: Simplified unary/binary math operators
URL: https://github.com/apache/incubator-mxnet/pull/8361#issuecomment-338426767
We see it in the nightly benchmark runs, but I was unable to reproduce
locally after spending quite a bit of time on it. However
cjolivier01 commented on issue #8361: Simplified unary/binary math operators
URL: https://github.com/apache/incubator-mxnet/pull/8361#issuecomment-338426981
For example, a unit test that takes 3 seconds on my machine, I?m told it
takes 15-20 minutes on a c4.8xlarge.
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112744
##
File path: python/mxnet/kvstore.py
##
@@ -349,6 +349,101 @@ def row_sparse_pull(self, key, out=None,
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112750
##
File path: python/mxnet/kvstore.py
##
@@ -349,6 +349,101 @@ def row_sparse_pull(self, key, out=None,
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112886
##
File path: src/ndarray/ndarray.cc
##
@@ -558,6 +558,101 @@ void CopyFromTo(const NDArray& from, const
mseeger commented on issue #8361: Simplified unary/binary math operators
URL: https://github.com/apache/incubator-mxnet/pull/8361#issuecomment-338430529
Which functions in mshadow_op.h would be used in an LSTM? sigmoid, or relu?
We could easily change them back to the old code and compare
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112917
##
File path: src/ndarray/ndarray.cc
##
@@ -558,6 +558,101 @@ void CopyFromTo(const NDArray& from, const
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112984
##
File path: src/operator/contrib/two_bit_quantize.cc
##
@@ -0,0 +1,122 @@
+/*
+ * Licensed to the Apache
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112556
##
File path: src/operator/contrib/two_bit_quantize.cc
##
@@ -0,0 +1,122 @@
+/*
+ * Licensed to the Apache
piiswrong closed pull request #8369: Fix the Readme
URL: https://github.com/apache/incubator-mxnet/pull/8369
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from
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 ad20d91 Update cudnn_algoreg-inl.h
piiswrong closed pull request #7988: solve problem in print "cudnn autotune"
URL: https://github.com/apache/incubator-mxnet/pull/7988
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a
szha closed pull request #8171: add profile option for frontend profiling to
image script
URL: https://github.com/apache/incubator-mxnet/pull/8171
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
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 590d7b0 add profile option for
reminisce commented on a change in pull request #8371: Add note in the doc for
using naive engine in multithreading environment
URL: https://github.com/apache/incubator-mxnet/pull/8371#discussion_r146114557
##
File path: docs/faq/env_var.md
##
@@ -56,6 +56,9 @@ export
mseeger commented on issue #8361: Simplified unary/binary math operators
URL: https://github.com/apache/incubator-mxnet/pull/8361#issuecomment-338430381
Due to the unary/binary math functions?
In principle, at least for the majority of all expressions, we should go
through the same
piiswrong commented on a change in pull request #8342: [WIP] 2bit gradient
compression
URL: https://github.com/apache/incubator-mxnet/pull/8342#discussion_r146112785
##
File path: python/mxnet/kvstore.py
##
@@ -349,6 +349,101 @@ def row_sparse_pull(self, key, out=None,
piiswrong commented on issue #8361: Simplified unary/binary math operators
URL: https://github.com/apache/incubator-mxnet/pull/8361#issuecomment-338425272
Chris observed 4x performance regression for LSTMs that might be related to
the previous PR.
@cjolivier01 have you found 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 72b0051 Allow test to converge (#8351)
piiswrong closed pull request #8351: Allow test to converge
URL: https://github.com/apache/incubator-mxnet/pull/8351
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull
piiswrong closed pull request #8192: [Perl] emulate Python zip() for Perl
URL: https://github.com/apache/incubator-mxnet/pull/8192
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a
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 d6062f7 [Perl] emulate Python zip() for
yajiedesign commented on issue #4746: ImportError for mxnet: cannot import name
libinfo
URL:
https://github.com/apache/incubator-mxnet/issues/4746#issuecomment-338445615
@Bumblebee1964 do you install cuda 8.0?
This is an
reminisce commented on a change in pull request #8371: Add note in the doc for
using naive engine in multithreading environment
URL: https://github.com/apache/incubator-mxnet/pull/8371#discussion_r146114557
##
File path: docs/faq/env_var.md
##
@@ -56,6 +56,9 @@ export
szha opened a new pull request #8377: re-enable tests
URL: https://github.com/apache/incubator-mxnet/pull/8377
## Description ##
Re-enable tests that were disabled in #7648, #8265, #7829, #8045. The tests
are marked as flaky, and are allowed maximum of three runs.
## Checklist ##
szha opened a new pull request #8378: upgrade MKL
URL: https://github.com/apache/incubator-mxnet/pull/8378
## Description ##
upgrade mklml dependency
## Checklist ##
### Essentials ###
- [x] Changes are complete (i.e. I finished coding on this PR)
- [x] All changes have
piiswrong closed pull request #8376: Fix Typo (classification)
URL: https://github.com/apache/incubator-mxnet/pull/8376
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull
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 725a542 Fix Typo (classification)
szha commented on issue #8340: Fill optimizations
URL: https://github.com/apache/incubator-mxnet/pull/8340#issuecomment-338442944
The symbol-side can be updated to use the new op as well. The frontend
definition that needs update is here:
cjolivier01 commented on a change in pull request #8340: Fill optimizations
URL: https://github.com/apache/incubator-mxnet/pull/8340#discussion_r146119887
##
File path: src/operator/tensor/init_op.h
##
@@ -164,19 +164,38 @@ inline bool InitStorageType(const
cjolivier01 commented on issue #8340: Fill optimizations
URL: https://github.com/apache/incubator-mxnet/pull/8340#issuecomment-338444522
done
This is an automated message from the Apache Git Service.
To respond to the
cjolivier01 opened a new pull request #8379: Use omp_get_max_threads() when
OMP_NUM_THREADS environment set
URL: https://github.com/apache/incubator-mxnet/pull/8379
Using wrong API call here. Only relevant if OMP_NUM_THREADS environment
variable is set.
## Description ##
(Brief
ZhichengHuang commented on issue #7593: why gluon is slower than PyTorch?
URL:
https://github.com/apache/incubator-mxnet/issues/7593#issuecomment-338448789
@piiswrong This is issue haven't solve ,can you give me some suggestions
about how to make the transform faster? or How to solve this
cjolivier01 opened a new pull request #8380: Memset/memcpy/omp profiling tester
URL: https://github.com/apache/incubator-mxnet/pull/8380
## Description ##
This doesn't assert, but tests the threshold where memset or memcpy becomes
slower than the equivalent OMP implementation.
##
szha commented on issue #8152: gluon improvement
URL: https://github.com/apache/incubator-mxnet/pull/8152#issuecomment-338452966
added shape completion after loading parameters or finishing deferred init.
This is an automated
53 matches
Mail list logo