add a new DAG 'maxtext_multi_tier_sav01_save_local' - #81
Open
ernie-chang wants to merge 1 commit into
Open
Conversation
ernie-chang
force-pushed
the
sav01-save-local
branch
3 times, most recently
from
June 26, 2025 07:46
30513ab to
50c20e1
Compare
camiloCienet
force-pushed
the
sav01-save-local
branch
from
August 8, 2025 03:41
79e7507 to
9708252
Compare
camiloCienet
requested review from
andrewyct,
severus-ho and
xibinliu
as code owners
August 8, 2025 03:41
camiloCienet
force-pushed
the
sav01-save-local
branch
2 times, most recently
from
August 8, 2025 03:57
270038b to
ab6d305
Compare
CJ-Tsai-cienet
left a comment
CJ-Tsai-cienet
left a comment
Collaborator
There was a problem hiding this comment.
I think we need to update the commit message
| ABHINAV_S = "ABHINAV S." | ||
| XUEFENG_G = "XUEFENG G." | ||
| ERNIE_C = "ERNIE C." | ||
| CAMILO= "CAMILO Q." |
Collaborator
There was a problem hiding this comment.
I found that Severus has another PR, updating the owner's name to github ID., Maybe we need to update these once his PR is merged.
camiloCienet
force-pushed
the
sav01-save-local
branch
4 times, most recently
from
August 8, 2025 09:15
4e89d8f to
765ab6a
Compare
[UPDATE] Update with comment [NEW] Add orbax utils: apply cpc and delete cpc to the dag [UPDATE] add delete cpc and apply cpc in orbax [UPDATE] Update format and coding style [UPDATE] [UPDATE] Modify the dag structure and add cpc operation [UPDATE] Modify the dag structure and add cpc operation [UPDATE] update for the code comment [UPDATE] update for the code comment add a new DAG 'maxtext_multi_tier_sav01_save_local' GoogleCloudPlatform#804 Add comment validate() Change log explorer text filter string to be discarded to be discarded to be discarded
camiloCienet
force-pushed
the
sav01-save-local
branch
from
August 8, 2025 09:18
765ab6a to
9e00aa3
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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
This test verifies the Orbax Multitier Checkpointing local saving function, with phase 2 replicator enabled. It conducts the following tasks:
Run the MaxText training by enabling the checkpointing, until the local checkpoints have been saved.
Clean up the saved local ram checkpoints.
Using logging explorer API to query log entries during the tests, and verify they contain the logs that local checkpoints have been saved.
Note: The local checkpoints validation is through the logging explorer API, instead of going into the pod and checking the saved checkpoints files.
Tests
cluster env:
project: cloud-tpu-multipod-dev
name: b425674043-v5p64
zone: europe-west4-b
highScaleCheckpointingConfig: enabled
GcsFuseCsiDriver: enabled
Workload Identity Federation: enabled
gcs bucket:
name: mtc-automation-bucket
Hierarchical namespace: enabled
airflow/composer env:
project: cloud-ml-auto-solutions
name: erniechang-test
composer version: 2.13.4
airflow version: 2.10.5
test result:
Tests have been conducted in the "cloud-ml-auto-solutions" project with cloud composer.
Test link
Checklist
Before submitting this PR, please make sure (put X in square brackets):