search

NVIDIA / fsi-samples

A collection of open-source GPU accelerated Python tools and examples for quantitative analyst tasks and leverages RAPIDS AI project, Numba, cuDF, and Dask.
271 stars 115 forks source link

updated docker build and version number #83

Closed jbaron closed 4 years ago

jbaron commented 4 years ago

Few changes/updates:

1) Added support for CUDA 10.2 (as now is also supported by rapids) in the docker build file.

2) The docker build was missing a copy for download_data.sh that made most of the included notebooks not usable. Now it works again.

3) The docker build provides a lot of warnings due to incompatible versions. Fixed them and removed packages already available in the runtime rapids docker image. Docker build is much cleaner now.

4) The version in the setup was not correctly updated. So I set it so it is 0.4 now (at least that is my understanding).

5) Added two lines to .gitignore so that docker build artifacts don't get checked in by mistake.

Hope this also helps others.

GPUtester commented 4 years ago

Can one of the admins verify this patch?

yidong72 commented 4 years ago

@jbaron , Thanks for your PR. Could you change it to base on gQuant:develop instead of gQuant:master ? We would like to keep master branch stable. develop branch is used for new developments. We can merge it to master in a separate PR.

jbaron commented 4 years ago

Send another pull request, this time against the dev branch as requested. So closing this one.