Hi Sergio,
we are only collecting ICLAs from people if they become a committer.
Otherwise, we don't request any.
-Marco
On Thu, Jul 12, 2018 at 8:07 PM Sergio Fernández wrote:
> Hi,
>
> one detail I've notices: is the Podling collecting all ICLA when people
> contribute to the project?
>
> I c
Quick update: We are now at 27 disabled tests
https://github.com/apache/incubator-mxnet/issues?q=is%3Aopen+is%3Aissue+label%3A%22Disabled+test%22
-Marco
On Thu, Jun 28, 2018 at 2:31 AM Marco de Abreu
wrote:
> Hello,
>
> we currently have a false-failure rate of about 59% in our CI. T
ween these two labels.
> >
> > It would make sense to merge the "Feature" label into "Feature Request".
> >
> >
> > Thanks
> > Anirudh
> >
> >
> > On Wed, Jun 27, 2018 at 3:50 PM Hagay Lupesko wrote:
> >
> &
Hi Naveen,
I'm in favour of the squashing, considering the number of commits in some
PRs and especially because of some people making commit messages a la "fix"
"fix" "fix" all the time. Additionally, it gets hard (not impossible, just
more inconvenient) to determine the atomic states of master -
We could add that doc to our CI nightly pipeline or auto generate it by
extracting certain parts from our website, thus reducing the maintenance
overhead.
In general, I'd think that it's not feasible to add a redirect. But I think
it would be desirable to get the readthedocs page under our control
nvironment configuration (e.g. for the package
> `rJava` one needs to take care of $PATH and the permission of the
> installation).
>
>
> Best regards,
>
> Tong He
>
> 2018-07-04 1:20 GMT-07:00 Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid>:
>
> >
t; Thanks
> Anirudh
>
>
> On Tue, Jul 3, 2018 at 2:48 AM Marco de Abreu
> wrote:
>
> > Hello,
> >
> > do we have a release process for our R frontend? I noticed the issue at
> [1]
> > and it seems like we're only publishing to an S3 bucket which i
Thanks for raising this issue Sheng.
My proposal would be to always print a warning message when this function
is called with the ssl check disabled. This functionality would be tested
by a unit test which mocks the network access.
Additionally, I'd like to propose that we set a policy for oursel
Hello,
do we have a release process for our R frontend? I noticed the issue at [1]
and it seems like we're only publishing to an S3 bucket which is not under
Apache. Is there another channel for our users to retrieve that package or
is this our only supported official way?
Best regards,
Marco
[1
Okay, the issue has been resolved and master is stable again. Please just
add a new commit to your pull requests so they take effect.
Thanks to Pedro, Anton and Carin!
-Marco
Marco de Abreu schrieb am So., 1. Juli 2018,
11:57:
> I'm on it.
>
> Pedro Larroy schrieb am So., 1. J
018 at 6:55 AM, Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid> wrote:
>
> > PR is available at https://github.com/apache/incubator-mxnet/pull/11512.
> > So
> > far, it's looking good. None the less, please review :)
> >
> > -Marco
> >
>
gt;
> On Sun, Jul 1, 2018 at 5:56 AM, Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid> wrote:
>
> > Congratulations! Unfortunately was the last CI run of that PR stale and
> did
> > not include the latest updates to our pipeline. We just added the rat
> > lic
I'm on it.
Pedro Larroy schrieb am So., 1. Juli 2018,
09:03:
> Hi
>
> Master is not building due to RAT license check on the clojure package. Is
> anyone having a look at this?
>
> Pedro.
>
Congratulations! Unfortunately was the last CI run of that PR stale and did
not include the latest updates to our pipeline. We just added the rat
license check which apparently did not run on your PR send is now failing
on master as well as blocking all pull request builds. I will try to add
the li
Very interesting! I especially like the different levels of engagement and
beam committers actively engaging potential committers to mentor them
towards reaching the next level. This fact together with the feedback loop
sounds like a very promising and encouraging system.
Thanks a lot Yasser and K
Hello,
so far, we've been running RAT license checks as part of our nightly
pipeline. To avoid back and forth before a release, we'd like to make it
part of the sanity stage of our PR validation.
In future, you will receive a notification if you added a file which
requires an Apache license. The
Hello,
we encountered a test failure on Windows which caused memory corruption.
This resulted in spilled over test failures, causing further CUDA kernels
to fail to launch [1]. Thanks to Dick Carter who debugged the issue and
tracked it down to a bad implementation of SequenceLastKernel, leading t
ng links, I
> > see MXNet already has configured project management boards in Jira and it
> > seems people have assigned related tasks there, I mean I see it's already
> > working and we don't need an extra To Do, right?
> >
> > Regards.
>
Hello,
we currently have a false-failure rate of about 59% in our CI. This leads
to a lot of PRs failing due to flaky tests. We're currently in the process
to fix problematic tests, but we're still not at a point which we can
consider as stable.
A few colleagues proposed to disable tests temporar
qskuze8K2d+ck2p/t9fsuXLMYE7Se6f7T/wCjkP2e/wDhH1+q5cnXYnhBXHfdZ/ashx39n932XLksuwro8bsPP7plS3HmuXIBOtf6j/JVD+o3yXq5AB1Hc+ZUbb+p6H6rxcszDNvvfnReVP5XLljDC12CKXLkxj//2Q==
>
>
> 2018-06-27 14:45 GMT+01:00 Chen HY :
>
> > Is that the cat you need?
> >
> > 2018-06-27 13:03 GMT+01:0
ke to make
sure that we re-enable this test before the next release.
-Marco
[1]: https://github.com/apache/incubator-mxnet/pull/11419
On Wed, Jun 27, 2018 at 1:53 PM Marco de Abreu
wrote:
> Hello,
>
> I have noticed an increased number of CI failures. The reason seems to be
> htt
;
> On 6/11/2018 8:53 PM, Marco de Abreu wrote:
> > How about you set up a personal instance of JIRA (let us know if the
> public
> > ones differ too much from the one we're running under Apache) and play
> > around a bit? We can then review the setup and migrate th
Hello,
I have noticed an increased number of CI failures. The reason seems to be
https://github.com/apache/incubator-mxnet/issues/11407. A third party
dependency (a cat image) has been deleted from the webserver and thus
causes all our CI runs to fail. Is anybody familiar with that test and has
a
+1 to renaming to Backend
On Mon, Jun 25, 2018 at 10:13 AM Hagay Lupesko wrote:
> Thanks Lin for your feedback.
> Bumping again to get more feedback before concluding.
>
> On Fri, Jun 22, 2018 at 8:53 AM Lin Yuan wrote:
>
> > I agree with Hagay. Using "Backend" as label makes it much easier to
value explained here (
>
> https://stackoverflow.com/questions/14599670/what-error-code-does-a-process-that-segfaults-return
> )
> set. Then you can mark the full suite as crashed.
>
> Pedro.
>
> On Thu, Jun 21, 2018 at 1:35 PM Marco de Abreu
> wrote:
>
> > Hello,
> >
We had 868 successful runs and no failures for that test so far, Pedro.
On Thu, Jun 21, 2018 at 11:17 PM Pedro Larroy
wrote:
> Observed just one failure in OSX in test_imdecode:
>
>
> ==
> FAIL: test_imdecode (test_image.TestIma
Hello,
is anybody aware of a way to catch segmentation faults as part of the
nosetests execution and log them as ERROR? Right now, the nosetests process
gets terminated without a stack trace or further test execution. This has
additional significance because our result-recording that has been
intr
ersonally like it a
> lot even with some limitations with it. It doesn't allow us to merge a
> PR with a not tested code. It reports us where are those not covered new
> codes.
>
> Regards.
>
> On 6/21/2018 7:13 PM, Marco de Abreu wrote:
> > To avoid any confu
Hello Hagay,
as far as I understand, the CPP Package label is for the C++ Frontend which
is generated at [1] while the C++ label is intended for the backend. But I
agree that we don't really have to label backend bugs as C++, so we could
just deprecate the CPP Package label and move everything ove
On Thu, Jun 21, 2018 at 6:43 AM Marco de Abreu
wrote:
> Hi Kellen,
>
> Thanks for your feedback.
>
> The same argument about a hosted service could be applied to GitHub - we
> could just use the git repository hosted under Apache Infra then and
> completely stick to JIRA. This
t; (i.e. Scala) might also be nice eventually.
>
> -Kellen
>
> On Thu, Jun 21, 2018 at 4:59 AM Chris Olivier
> wrote:
>
> > we should have a betting pool on what the current coverage is.
> >
> > On Wed, Jun 20, 2018 at 7:54 PM Naveen Swamy
> > > Firstly, this would only be triggered if there is an operator
> > changes
> > > > > (Only general operators).
> > > > > Secondly, it is simple to go through. Just need to change 1
> line
> > of
> > > > code
> > >
on the pull request merging.
>
> Tianqi
>
> On Wed, Jun 20, 2018 at 7:14 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid> wrote:
>
> > Hello,
> >
> > I'd like to introduce test coverage metrics of PRs using
> > https://codecov.io/.
Hello,
I'd like to introduce test coverage metrics of PRs using https://codecov.io/.
This tool will aggregate coverage reports across multiple runs, platforms
and technologies and gives contributors as well as reviewers a new tool
that allows to improve the quality of a pull request.
Since we nee
working together on this. The idea is that we fail the build
> if there is a operator change on the backend and have not synced to the
> Scala API. We want to catch this before breaking the user's code which will
> be a pretty bad experience.
>
>
>
> On Tue, Jun 1
Hi Qing,
thank you for working on improving the compatibility of our APIs!
Your linked proposal does not describe the mentioned FILEHASH. Could you
elaborate a bit? Would this be a hash of the entire file, some hash created
based on the signature of the underlying C++ methods or maybe a different
GPUs. After reading your email, I'm
> thinking I'll run most of the notebooks on P3.2xl and only the ones that
> require multiple GPUs on P3.8xl. Thanks for the reply.
>
>
> On Mon, Jun 18, 2018 at 2:24 PM Marco de Abreu
> wrote:
>
> > In future (next few days
In future (next few days) the nightlies will be running on our regular CI
at http://jenkins.mxnet-ci.amazon-ml.com/
On there, we support C5.18xlarge (Ubuntu, CentOS within Docker and
Windows), G3.8xlarge (Ubuntu, CentOS within Docker and Windows), P3.2xlarge
(Ubuntu, CentOS within docker) and (upc
> we
> > > > > > stick
> > > > > > > with that given the user community prefers that
> > > > > > >
> > > > > > > Tianqi
> > > > > > >
> > > > > > > On Fri, Jun 15, 2018
I think nobody was opposed to it in the past, right?
I'd propose that all emails automatically get copied to dev@ to ensure high
visibility initially. What do you think?
Sebastian schrieb am Fr., 15. Juni 2018, 20:51:
> I have already proposed this many times in the past and would strongly
> en
c Xie wrote:
> Hi Marco de Abreu,
>
> CI has been totally broken recently. It randomly fails for no good reason
> more often than it passes. For example the ccache/efs failure has been
> really annoying.
>
> Looks like there has been many changes to Jenkins and Docker lately. Do
&g
tly. It breaks a not
> recommended usage of the API from an unmaintained tutorial, I don't think
> adding more reviewers will help it.
>
> Besides, I'm less sure if we can find enough reviewers to provide useful
> feedbacks for major changes.
>
> On Fri, Jun 15, 2018
that brings improvements in the long term
>
> Tianqi
>
> On Fri, Jun 15, 2018 at 2:15 PM, Mu Li wrote:
>
> > Why reverting instead of fixing the bugs? Static memory aims to
> reduce
> > memory allocation, it's a key feature to bridge the perf
Mentors, do you know if it is possible to get a bot account with these
restricted permissions? I know that previous requests have been declined by
Apache Infra. Shall we try it again or how shall we continue moving forward?
-Marco
On Fri, Jun 15, 2018 at 2:09 PM Qing Lan wrote:
> Hi All,
> I wo
it's a key feature to bridge the perf gap between gluon
> and symbol.
>
> On Fri, Jun 15, 2018 at 2:06 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid> wrote:
>
> > Hello,
> >
> > I'm reverting https://github.com/apache/incubator-mxnet/pul
Hello,
I'm reverting https://github.com/apache/incubator-mxnet/pull/10817 as of
https://github.com/apache/incubator-mxnet/pull/11311 due to regressions
described in https://github.com/apache/incubator-mxnet/issues/11171 and
https://github.com/apache/incubator-mxnet/pull/10817.
The pull request ha
; also seen it happen on Ubuntu, there are probably some links in the
> issue
> >>> (on my phone right now).
> >>>
> >>> Anirudh schrieb am Mi., 13. Juni 2018, 22:16:
> >>>
> >>> > By segfaulting test do you mean : test_gru_bidirect
itters not having to modify the action (PR,
> proposed-answer, proposed-wiki-change etc).
>
> Hen
>
> On Fri, Jun 15, 2018 at 12:26 AM, Marco de Abreu <
> marco.g.ab...@googlemail.com.invalid> wrote:
>
> > Hen,
> >
> > As you stated, it's of signifi
Hen,
As you stated, it's of significance of how much a PR has to be changed as a
result of a review. I think this is what this project defines as quality.
If people submit a bunch of PRs and we reviewers spend a lot of time on
every single one of them to give the contributors advice about how to
in main
default=default_ccache_dir(),
File "ci/build.py", line 127, in default_ccache_dir
return ccache_dirpython
NameError: name 'ccache_dirpython' is not defined
script returned exit code 1
Best regards,
Marco
[1]: https://github.com/apache/incubator-mxnet/pull/11269
On Thu, Jun 1
Hello,
I'd like to explain the recent errors we had in the past 12 hours as part
of CI. This was caused by my preparation to enable shared ccache volumes.
Due to a mistake from my side, the option got applied to all runs instead
of just my PR, causing all runs to fail. I have reverted all changes
don't see the
> segfault in the logs. Can you point me to the test.
> Also, this seems to be specific to the master and not in 1.2:
> https://github.com/apache/incubator-mxnet/pull/10311
>
> Anirudh
>
> On Wed, Jun 13, 2018 at 10:00 PM, Marco de Abreu <
> marco.g.ab.
I can confirm that this segfaulting test has a big impact.
On Wed, Jun 13, 2018 at 9:39 PM Aaron Markham
wrote:
> I'd keep an eye on this one... Flaky test: test_gru_bidirectional #11219
>
> https://github.com/apache/incubator-mxnet/issues/11219
>
> Just reran several PR's CI runs that all had
I think Sheng manages the publishing to PyPi and we should be fine with
just adding it as another flavour.
-Marco
On Wed, Jun 13, 2018 at 6:36 AM Pedro Larroy
wrote:
> Hi
>
> We would like to start publishing pre-built beta pip wheels of MXNet for
> Raspberry pi and ARM.
>
> What's the procedur
Hello Cathy,
that's a great proposal. Thank you!
A few comments from my side:
- Good idea with the alias. We should have a special email-list for
automated reports to prevent spamming dev@.
- "Create weekly email to internal team members:" -> email-list
- "Part II - Label Bot - Amazon cloudwatch e
it
> up.
>
>
> On Tue, Jun 12, 2018 at 12:07 AM, Marco de Abreu <
> marco.g.ab...@googlemail.com> wrote:
>
> > The problem I see here is that the migration will have different type of
> > challenges which should be handled isolated. Trying to solve them all at
res iteratively well
> tested and well documented. MXNet already has many different language
> bindings which has quite a bit of tech-debt, I don't want just to add more
> tech-debt to the code base with new language bindings as well.
>
> On Mon, Jun 11, 2018 at 11:17 PM, M
I think we should try to separate this migration into multiple stages:
1. Move into contrib
2. Migrate release pipeline
3. Migrate tests
4. Improvements
5. Mark as stable and announce Julia officially
I know how much effort Carin and the other maintainers are putting into
that project and thus I'd
I don't know how to grant permission on Confluence. If somebody else knows
how to do so, please grant Marek the edit permissions.
-Marco
On Mon, Jun 11, 2018 at 11:05 AM Marek Kolodziej wrote:
> Hi Rajan,
>
> I wanted to share on Confluence, but it didn't allow me to create a new
> document. If
Hello Marek,
this sounds great! Definitely looking forward to it.
It seems like our mailing list destroyed your formatting. You might want to
consider putting it into a Google Docs document or uploading it to
confluence.
Best regards,
Marco
On Mon, Jun 11, 2018 at 10:50 AM Marek Kolodziej wrot
Thanks for this detailed description!
I think GitHub and JIRA are a good thing to co-exist. GitHub for issues and
JIRA for action items (as a result of the issues). I like GitHub for it's
good community integration - just mention somebody or link it in another
issue/PR and everything will automati
t work it off of a fork?
>
>
> On Thu, Jun 7, 2018 at 5:32 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com
> > wrote:
>
> > Hello,
> >
> > we are currently revisiting our setup for ARM and Android. Since we're
> not
> > in a stable state but need
hear that. I contacted several customers, most of them are
> waiting for 1.2 to be ready on the internal system. Did they fork by their
> own? If they changed the codes, will them submit their changes?
>
> On Sat, Jun 9, 2018 at 2:51 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com
committers.
-Marco
On Sat, Jun 9, 2018 at 5:32 PM Yasser Zamani
wrote:
>
>
> On 6/9/2018 5:01 PM, Marco de Abreu wrote:
> > Would you mind creating a proposal document at
> > [1], describing what you would have in mind and how project planning,
> > -management, developme
> 1.2 some day between May 21 and June 7 and want to PR their changes later.
> But it less likely happens.
>
> Best,
> Mu
>
> > On Jun 9, 2018, at 2:05 PM, Marco de Abreu
> wrote:
> >
> > Would there be any benefit besides cosmetics? I'd propose to just
change is pushed, and rebase
> against the code after squash, it will be fine. The problem will only
> happen if someone checked out 1.2.0 after these commits get in, push their
> own changes, then try to rebase
>
> Tianqi
>
> On Sat, Jun 9, 2018 at 1:25 PM, Marco de Abreu <
Jun 9, 2018 at 1:18 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com
> > wrote:
>
> > I think we should never ever force push to a publish repository since we
> > don't know what depends on it. I'd say we take this as a lesson learned
> and
> &g
It's good to have a stable
> > release only patched with meaningful commits.
> >
> > Best,
> > Mu
> >
> > > On Jun 7, 2018, at 2:55 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com>
> > wrote:
> > >
> > > Ah yeah that's e
s you can think of would look like then? Please let us know if you
need permissions to edit the wiki.
Best regards,
Marco
[1]: https://cwiki.apache.org/confluence/display/MXNET/Design+Proposals
On Sat, Jun 9, 2018 at 12:56 PM Yasser Zamani
wrote:
>
>
> On 6/9/2018 12:43 PM, Marco de Abre
e github issue, another for only
> contributer to access, which of course will have some more harder to answer
> or codding questions and will be published at JIRA.
> For example, the simple one can be documents, bug fix or new ideas, the the
> harder one can be new idea implementation.
Hi,
thanks for your feedback!
I think the big problem is that we have nobody who maintains JIRA properly.
If we had a dedicated person who owns that system, I'm sure we could excell
at it's usage and it would provide better value. About the plugins, I have
to agree that we can just request assist
+1
While the intention of JIRA was good, we have proven as a community that we
are not able to use it properly and it's only being a burden. The initial
idea was that it allows us to manage projects publicly, make plans and
allow everybody to engage into these projects by just picking small tasks.
Thanks a lot Aaron for addressing this in such a timely manner!
It would definitely be great to see the raspberry build fixes in that
branch since it's currently broken.
-Marco
Aaron Markham schrieb am Fr., 8. Juni 2018,
18:10:
> Hi Anirudh,
> Marco brought my attention to CI failures on the 1
ot build all components to reduce cost and the time it
> > > takes to run CI. A Scala build need not build everything and run tests
> > > related to Python, etc.,
> > >
> > > Thanks, Naveen
> > >
> > > On Thu, Jun 7, 2018 at 9:57 AM, Marco de Abreu <
> &g
7, 2018 at 2:41 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com
> > wrote:
>
> > Hello Naveen,
> >
> > thank you for addressing this. Is it possible to let the plugin point to
> a
> > private fork so you can review it's actions manually before they a
Hello Naveen,
thank you for addressing this. Is it possible to let the plugin point to a
private fork so you can review it's actions manually before they are
written to the main repository? That way, we could just open a PR and
double check everything is as expected.
Thanks a lot for taking care
Yeah, why not :)
-Marco
On Thu, Jun 7, 2018 at 7:03 PM singh.raja...@gmail.com <
singh.raja...@gmail.com> wrote:
> Hi All,
>
> On MXNet home page, the link to github is currently inside
> Community->GitHub. I think we should have link to GitHub readily available
> on homepage ( one click), simil
> that is killing the productivity, we can redesign the test pipeline to run
> integration and unit test in parallel and that would give us straight away
> a 30 minutes reduced time in the CI run. Then we'd be always at <2h for a
> build, which seems reasonable if it never fails
Hello,
we are currently revisiting our setup for ARM and Android. Since we're not
in a stable state but need some way to collaborate while having support
from CI, we proposed the usage of feature branches. Thus, I have create the
two branches devel-arm and devel-android into which Anton and Pedro
Yeah, I think we are at the point at which we have to disable tests..
If a test fails in nightly, the commit would not be reverted since it's
hard to pin a failure to a specific PR. We will have reporting for failures
on nightly (they have proven to be stable, so we can enable it right from
the be
Hello secretary,
could you please confirm that all committers have signed their ICLAs and
that the wiki (see history of this thread) is obsolete.
Best regards,
Marco de abreu
On behalf of Apache MXNet (incubating)
-- Forwarded message -
From: Steffen Rochel
Date: Mi., 6. Juni
Thanks a lot Isabel!
Considering ICLAs are automatically managed by the secretary of Apache
before anybody can become a committer and no CLA is required for
non-committers, I'd say that we can delete that wiki page.
Does anybody disagree?
-Marco
Isabel Drost-Fromm schrieb am Mi., 6. Juni 2018,
Henry, was this document only related to the migration towards Apache or do
we have to keep it up to date? I've heard that sometimes every contributor
has to sign it before a PR can be accepted.
I have seen other communities that use a bot to keep track of signed ICLA
documents, but I don't know
I definitely like this idea. We have also been discussing about the exact
same idea within our team and think that it allows MXNet to scale better
with the increasing number of backends. Imagine we get AMD-CPU, AMD-GPU,
ARM, Snapdragon and other vendors/chips. We are definitely running into
problem
Hello Sergio,
you are right. We are following semantic versioning and thus, every model
produced within the same major version (e.g.1.x) has to be backwards
compatible.
Could you please provide a small example so we can reproduce this? We
definitely do not want our users to retrain their model if
Hello MXNet community,
I'd like to announce the launch of restricted slaves and jobs for our CI
system.
The purpose of this feature is to allow separating slaves that execute
arbitrary code from verified code like on our version-branches and the
master. This step is necessary in order to increase
Hello,
I'd to inform you that our Jenkins Master and the underlying plugins have
all just been updated to the latest versions to include the latest security
updates.
We don't expect any problems caused by this, but if you experience any
issues due to the update, please let me know so we can inves
pache.org/confluence/display/MXNET/Release+Process#
> ReleaseProcess-Step1.12.CreateScalaMavenPackages(WIP)
>
> 2) https://cwiki.apache.org/confluence/display/MXNET/MXNet-Scala
>
> On Mon, May 21, 2018 at 5:37 PM, Marco de Abreu <
> marco.g.ab...@googlemail.com> wrote:
>
&g
Haha, Aaron could work in marketing - AI solves everything! :P
But yeah, a bit more details would be good. As an alternative approach, I'd
like to propose a bot that does actions on your behalf. In this case, we
could have a bit that listens for comments and (based on a whitelist), it
would apply
Agree, we should make a proper assessment of what all these packages are
before we try managing them. Do you know who published them before?
Moving forward, it'd be great if you could document the entire process (and
all issues you encounter) in confluence. This would allow us to re-visit
decision
Hi Naveen,
thank you for driving the releases for the Scala packages!
For CPU, I'd propose add MKL/MKLDNN builds.
For GPU, do we have to specify the minor version of CUDA up front or are
they interchangeable? CUDA 9.2, for example, received some quite
significant performance updates and it would
By the way. In general, I'd prefer to have a solution in C++ as part of our
Backend, as this would allow us to share this information with all frontend
APIs. Tobias is currently working on a PR which does exactly the same, but
for GPUs.
Marco de Abreu schrieb am Sa., 19. Mai 2018,
02:45:
I would have to check the license, but I assume it's under Apache. In that
case, it should be enough to copy the file and keep the original license
header. Additionally, you have to document any changes you made and where
you got it from. You can do this in the header, below the license.
In any ca
found during
> 1.2 rc voting
> - https://github.com/apache/incubator-mxnet/issues/10981 were reported by
> an user with CUDA 7
>
> Having these covered in CI will help catch the issues early. I don't recall
> if we decided to drop CUDA 7 support for MXNet.
>
>
at 3:47 AM, Thomas DELTEIL
wrote:
> Great news :) thanks Marco!
>
> On Tue, May 15, 2018, 18:29 Steffen Rochel
> wrote:
>
> > Thanks Marco, good step forward.
> > What is the expected, typical and worst case TAT time for PR checks?
> >
> > Steffen
> &
Hello,
I'd like to announce the deployment of auto scaling for our CI system (see
[1] for reference, setup documentation at [2]) for today at 11:00PM PST
05/15/18. I expect no downtime since these changes are happening outside of
Jenkins.
This system will increase the flexibility of our system to
on mxnet master. One thing on my wish
>> list is a database which stores all the occurrences of test failures and
>> their commit ids, which would be very helpful for initial diagnosing what
>> code changes potentially introduced bugs. Otherwise clicking all past
>> tests
l diagnosing what
> code changes potentially introduced bugs. Otherwise clicking all past tests
> and reading those logs requires a lot of manual work.
>
> Best,
> Haibin
>
> On Wed, May 9, 2018 at 5:32 AM, Marco de Abreu <
> marco.g.ab...@googlemail.com
> > wrote:
>
>
Hello,
I'm currently working on auto scaling and encountering a consistent test
failure on CPU. At the moment, I'm not really sure what's causing this,
considering the setup should be identical.
http://jenkins.mxnet-ci-dev.amazon-ml.com/blue/organizations/jenkins/incubator-mxnet/detail/ci-master/
@ accordingly?
> > Currently you are on record as -1. Just trying to help Anirud to get
> proper
> > vote count.
> >
> > Thanks
> > Steffen (MXNet contributor hat on)
> >
> > On Tue, May 8, 2018 at 6:37 AM Marco de Abreu <
> > marco.g.ab...
301 - 400 of 583 matches
Mail list logo