feat: NGC+ Image Template #235
Merged
+286
−198
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
MikhailKardash
force-pushed
the
ngc_images
branch
from
bc415d0
to
a9df79b
Compare
ioga
reviewed
Nov 30, 2023
cdrickett
reviewed
Dec 11, 2023
cdrickett
reviewed
Dec 11, 2023
cdrickett
reviewed
Dec 11, 2023
| @@ -19,6 +19,6 @@ cd /tmp && \ | |||
| ./packages/build-deb-packages.sh -t -k -d && \ | |||
| cd / && \ | |||
| dpkg -i \ | |||
| /tmp/gdrcopy-$GDR_VER/libgdrapi_$GDR_VER-1_amd64.Ubuntu20_04.deb && \ | |||
There was a problem hiding this comment.
Is this necessary anymore? The base NGC images I used for the GH200 nodes already had libgdr* in /usr.
There was a problem hiding this comment.
I think we need this for our other image offerings. (i.e. Pytorch 2.0.1 and Pytorch 1.12) But this can be removed once we make these NGC images the default.
MikhailKardash
force-pushed
the
ngc_images
branch
from
f0e961a
to
f3a4b2d
Compare
MikhailKardash
force-pushed
the
ngc_images
branch
from
ea7b10d
to
34d6e1a
Compare
rb-determined-ai
approved these changes
Feb 8, 2024
4 tasks
4 tasks
soohoonchoi
added a commit
that referenced
this pull request
Mar 29, 2024
* fix: Tensorboard Profilers (#240) * Add tensorboard profilers back into images * we don't need 3.9 yet * wrong tag * build hpc/ngc together and update makefile * version matrix and update comment * profiler arg relocation * address some duplicates * formatting and libnss * yaml formatting * use actual yaml linter * relocate again * backport additional-requirements-torch and bump VERSION * additional-requirements for tf * bash syntax * cleanup dockerfiles, remove duplicate publishing steps, correct a dockerfile * try different syntax * semicolons * version pin and revert * pip * try python 3.10 * maybe it's a concurrency thing * no more version pin * ngc dockerfile cleanup * bump version file, minor formatting, publish artifacts * debian frontend google * google_cloud_cli... * cloud cli? * minor cleanup * version-matrix update and lots of formatting * unparametrize deepspeed * oops * update nvidia drivers to 535.161.07 (#246) minor version upgrade * version bump * feat: NGC+ Image Template (#235) * add templates for NGC+ images * add image matrix * backport lots of improvements to scripts * remove tf2.8 images * Removed a duplicate line * Added WITH_NCCL option to the Dockerfile-ngc-tf --------- Co-authored-by: Michael Kardash <[email protected]> Co-authored-by: Hamid Zare <[email protected]>
soohoonchoi
added a commit
that referenced
this pull request
Apr 30, 2024
* fix: Tensorboard Profilers (#240) * Add tensorboard profilers back into images * update nvidia drivers to 535.161.07 (#246) minor version upgrade * feat: NGC+ Image Template (#235) * add templates for NGC+ images * add image matrix * backport lots of improvements to scripts * remove tf2.8 images * fix: dependabot alert for `jupyterlab-3.6.7`. (#241) * Add support to build the tf2-gpu image for Libfabric(OFI) (#251) * Add support to build the tf2-gpu image for Libfabric(OFI), which incorporates the AWS libfabric plug-in for NCCL to use on Slingshot 11(SS11) networks. * Increment the VERSION to 0.31.1 * feat: update ngc version (#253) * feat: update ngc version * feat: Update naming (#252) * renaming a bunch of stuff, removing py3.8 and old cuda, changes to CI job names --------- Co-authored-by: Michael Kardash <[email protected]> Co-authored-by: Hamid Zare <[email protected]> Co-authored-by: Ilia Glazkov <[email protected]> Co-authored-by: Jerry G <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
version-matrix.yamltf28andCUDA111imagesinstall_google_cloud.sdkThis PR also moves some Dockerfile steps into scripts and moves some framework-based pip installs into requirements files. These updates are backported to our existing images.
Checklist
bumpenvsprocedure in the determined repo. See README.