Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Rollup changes with MOE #1545

Merged
merged 33 commits into from
Dec 7, 2017
Merged

Rollup changes with MOE #1545

merged 33 commits into from
Dec 7, 2017

Conversation

s-deitz
Copy link
Member

@s-deitz s-deitz commented Dec 7, 2017

No description provided.

asaksena and others added 30 commits December 7, 2017 08:47
…nning on the delete status, while the service is asynchronously decommissioned. This avoids running the delete command multiple times and correctly tracks the delete time and status of the resource.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=175892551
(903eb82)

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=175900388
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=175988443
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176033925
…es (https://cloud.google.com/sql/docs/postgres/high-availability). High availability instances can be created by using the --managed_db_high_availability=true flag.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176038709
…in the current directory. Increasing the default timeout for resource creation to handle multi TB clusters. Adding the snapshot details to the metadata.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176141018
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176377490
…sort the

metadata dict by key before converting it to the 'labels' string. This will
result in the 'labels' string being sorted by key in alphabetical order.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176385451
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176449000
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=176580390
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177072996
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177073589
…The VMs are run

independently, i.e, are not configured to use distributed training.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177091389
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177210785
…n. The environment vars ensure that the

required CUDA libraries can be found.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177383869
… config file provided by #156)

- Add a flag allowing change build tool version
- Make lint happy.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177388988
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177479960
…user (aka

a pod with a restricted security context) can set the GPU clock speed and
autoboost policy.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177481407
…is equal

to the current value. This allows users to run GPU benchmarks in the case that
they do not have permission to modify the GPU's clock speed or autoboost,
with the condition that they specify the default values for both settings.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177501235
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177535383
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177609813
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177635647
resources, but still works as intended. This allows all GPUs on the node to be useable by pods. Furthermore, pods can set the clock speed of the GPUs, as this daemonset grants them permissions to do so.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177654682
…-smi is

available on PATH. This prevents the daemonset from exiting in error if it runs
before nvidia-smi becomes available.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=177662581
accelerators.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=178023686
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=178167242
tohaowu and others added 2 commits December 7, 2017 08:59
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=178173609
The unit test was expecting FLAGS.run_uri to be a string, but it was
None.
@s-deitz s-deitz merged commit b87e6cb into master Dec 7, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

9 participants