Add Job ID Appending to Output Directories and Maximum Folders Threshold #331
+236
−7
Conversation
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
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
|
@AnuradhaKaruppiah Hi Ms. Karuppiah, could you review my first PR when you get a chance? Thank you! |
ZhongxuanWang
force-pushed
the
develop
branch
2 times, most recently
from
3a280fe to
7324ff9
Compare
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
AnuradhaKaruppiah
added
improvement
There was a problem hiding this comment.
Pull Request Overview
Adds functionality to append unique job IDs to output directories to avoid overwriting results across runs.
- Introduces
append_job_id_to_output_dirin the evaluation output config. - Implements UUID-based job ID generation in
run_and_evaluate. - Updates tests and documentation to cover and describe the new feature.
Reviewed Changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
| tests/aiq/eval/test_evaluate.py | Adds unit tests that simulate job ID generation logic |
| src/aiq/eval/evaluate.py | Inserts logic to generate and log a job ID when enabled |
| src/aiq/data_models/evaluate.py | Defines the new append_job_id_to_output_dir field |
| docs/source/reference/evaluate.md | Documents usage of the new output-dir option |
Comments suppressed due to low confidence (1)
docs/source/reference/evaluate.md:455
- The documentation uses
JOB_{UUID}(uppercase) but the implementation prefixes withjob_(lowercase). Update the example to match the actual lowercase prefix for consistency.
When `append_job_id_to_output_dir` is set to `true`, a unique job ID (`JOB_{UUID}`) is automatically generated for each evaluation run and appended to the output directory path.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Signed-off-by: Zhongxuan (Daniel) Wang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
|
In the new commits, I just made the two new parameters for EvalOutputConfig, namely # Maximum number of jobs to keep in the output directory. Oldest jobs will be evicted.
# A value of 0 or None means no limit.
max_jobs: int | None = None
# Policy for evicting old jobs. Defaults to using time_created.
eviction_policy: JobEvictionPolicy = JobEvictionPolicy.TIME_CREATED # or JobEvictionPolicy.TIME_MODIFIEDI also added corresponding logics in Now, the uses can choose whether the output folder removal policy should be either based on the time the folders are created ( I also made corresponding unit tests. I also made the parameters to have a default value so they will be backward compatible. |
ZhongxuanWang
changed the title
Signed-off-by: zhongxuanwang <cndanielwang@gmail.com>
|
I accidentally missed changing the documentation during adding my 2nd feature. Now the documentation is updated with the two new parameters |
|
/ok to test d78530c |
|
/ok to test 5deaf25 |
This commit: 1. Adds a job_management subsecion to group job config 2. Fixes a problem with the parent directory searched for eviction 3. Updates docs to reflect these changes Signed-off-by: Anuradha Karuppiah <anuradhak@nvidia.com>
AnuradhaKaruppiah
approved these changes
Jun 17, 2025
|
/ok to test 4ce53c6 |
|
/merge |
|
Merged, thx for the contribution @ZhongxuanWang |
AnuradhaKaruppiah
pushed a commit
to AnuradhaKaruppiah/oss-agentiq
that referenced
this pull request
Aug 4, 2025
…old (NVIDIA#331) Closes NVIDIA#325 This PR adds a new feature to append unique job IDs to output directories in the evaluation configuration. This helps prevent overwriting results across multiple runs by ensuring each run has a unique output directory. 1. Added new configuration option `append_job_id_to_output_dir` to `EvalOutputConfig` 2. Implemented job ID generation logic: - Generates UUID when feature is enabled and no job ID provided - Uses custom job ID when provided - No job ID generation when feature is disabled 3. Added comprehensive test coverage for all scenarios 4. Added documentation for the new feature This PR is also backward compatible by making the `append_job_id_to_output_dir` optional. ## By Submitting this PR I confirm: - I am familiar with the [Contributing Guidelines](https://github.com/NVIDIA/AIQToolkit/blob/develop/docs/source/resources/contributing.md). - We require that all contributors "sign-off" on their commits. This certifies that the contribution is your original work, or you have rights to submit it under the same license, or a compatible license. - Any contribution which contains commits that are not Signed-Off will not be accepted. - When the PR is ready for review, new or existing tests cover these changes. - When the PR is ready for review, the documentation is up to date with these changes. Authors: - Zhongxuan (Daniel) Wang (https://github.com/ZhongxuanWang) - Anuradha Karuppiah (https://github.com/AnuradhaKaruppiah) Approvers: - Anuradha Karuppiah (https://github.com/AnuradhaKaruppiah) URL: NVIDIA#331
scheckerNV
pushed a commit
to scheckerNV/aiq-factory-reset
that referenced
this pull request
Aug 22, 2025
…old (NVIDIA#331) Closes NVIDIA#325 This PR adds a new feature to append unique job IDs to output directories in the evaluation configuration. This helps prevent overwriting results across multiple runs by ensuring each run has a unique output directory. 1. Added new configuration option `append_job_id_to_output_dir` to `EvalOutputConfig` 2. Implemented job ID generation logic: - Generates UUID when feature is enabled and no job ID provided - Uses custom job ID when provided - No job ID generation when feature is disabled 3. Added comprehensive test coverage for all scenarios 4. Added documentation for the new feature This PR is also backward compatible by making the `append_job_id_to_output_dir` optional. ## By Submitting this PR I confirm: - I am familiar with the [Contributing Guidelines](https://github.com/NVIDIA/AIQToolkit/blob/develop/docs/source/resources/contributing.md). - We require that all contributors "sign-off" on their commits. This certifies that the contribution is your original work, or you have rights to submit it under the same license, or a compatible license. - Any contribution which contains commits that are not Signed-Off will not be accepted. - When the PR is ready for review, new or existing tests cover these changes. - When the PR is ready for review, the documentation is up to date with these changes. Authors: - Zhongxuan (Daniel) Wang (https://github.com/ZhongxuanWang) - Anuradha Karuppiah (https://github.com/AnuradhaKaruppiah) Approvers: - Anuradha Karuppiah (https://github.com/AnuradhaKaruppiah) URL: NVIDIA#331
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
Closes #325
This PR adds a new feature to append unique job IDs to output directories in the evaluation configuration. This helps prevent overwriting results across multiple runs by ensuring each run has a unique output directory.
append_job_id_to_output_dirtoEvalOutputConfigThis PR is also backward compatible by making the
append_job_id_to_output_diroptional.By Submitting this PR I confirm: