mxnet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lausen, Leonard" <lau...@amazon.com.INVALID>
Subject Re: Stopping nightly releases to Pypi
Date Mon, 02 Dec 2019 06:15:07 GMT
Quoting Dustin from Pypi: "Hi folks, this is a really big ask. The mxnet-*
projects already represent a huge portion of PyPI's total size on disk and in
terms of bandwidth. Per https://pypi.org/stats/, the mxnet-* projects total more
than 1.5TB of PyPI's 6.5TB total size."

Given these numbers, reducing size by 20% or 30% percent may not be the right
way to address the concerns. It would still be helpful for the releases and
release candidates to improve user experience (download time, disk space). So
cutting down on the SMs we release may be helpful.

On Mon, 2019-12-02 at 05:53 +0000, Sunderland, Kellen wrote:
> Makes sense to me to release nightlies to s3 only.  Can we reduce size by
> cutting down on the SMs we release?  Was the main complaint around cuda
> release sizes?
> 
> On Dec 1, 2019 9:43 PM, "Lausen, Leonard" <lausen@amazon.com.INVALID> wrote:
> Hi MXNet Community,
> 
> since more than 2 months our binary Python nightly releases published on Pypi
> are broken. The problem is that our binaries exceed Pypi's size limit.
> Decreasing the binary size by adding compression breaks third-party libraries
> loading libmxnet.so https://github.com/apache/incubator-mxnet/issues/16193
> 
> Sheng requested Pypi to increase their size limit:
> https://github.com/pypa/pypi-support/issues/50
> 
> Currently "the biggest cost for PyPI from [the many MXNet binaries with
> nightly
> release to Pypi] is the bandwidth consumed when several hundred mirrors
> attempt
> to mirror each release immediately after it's published". So Pypi is not
> inclined to allow us to upload even larger binaries on a nightly schedule.
> Their compromise is to allow it on a weekly cadence.
> 
> However, I would like the community to revisit the necessity of releasing pre-
> release binaries to Pypi on a nightly (or weekly) cadence. Instead, we can
> release nightly binaries ONLY to a public S3 bucket and instruct users to
> install from there. On our side, we only need to prepare a html document that
> contains links to all released nightly binaries.
> Finally users will install the nightly releases via
> 
>   pip install --pre mxnet-cu101 -f http://mxnet.s3.amazonaws.com/mxnet-cu101/
> nightly.html
> 
> Instead of
> 
>   pip install --pre mxnet-cu101
> 
> Of course proper releases and release candidates should still be made
> available
> via Pypi. Thus releases would be installed via
> 
>   pip install mxnet-cu101
> 
> And release candidates via
> 
>   pip install --pre mxnet-cu101
> 
> This will substantially reduce the costs of the Pypi project and in fact
> matches
> the installation experience provided by PyTorch. I don't think the benefit of
> not including "-f http://mxnet.s3.amazonaws.com/mxnet-cu101/nightly.html"
> matches the costs we currently externalize to the Pypi team.
> 
> This suggestion seems uncontroversial to me. Thus I would like to start lazy
> consensus. If there are no objections, I will assume lazy consensus on
> stopping
> nightly releases to Pypi in 72hrs.
> 
> Best regards
> Leonard
Mime
View raw message