SDK_CLI_test - #3623
Open
raghu-microsoft wants to merge 74 commits into
Open
Conversation
* Add grpo job example * Replace client info with generics * Add missing copyright * newline at end of aml_setup.py * Fix black formatting issues and reduce dataset * Update sdk/python/jobs/grpo/src/grpo_trainer_rewards.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update sdk/python/jobs/grpo/src/BldDemo_Reasoning_Train.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update sdk/python/jobs/grpo/src/BldDemo_Reasoning_Train.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Fix duplicated cell * Add to CODEOWNERS * Fix agenda image and dataset string * Replace demo with example * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Address deployment related comments * Add README * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Remove duplicated model info section * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update sdk/python/jobs/grpo/launch_grpo_command_job-med-mcqa-commented.ipynb Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Change dataset description --------- Co-authored-by: Sharvin Jondhale <shjondhale@microsoft.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com>
Co-authored-by: Mohd Javed Ansari <moansa@microsoft.com>
Add yeshwanth and harsha to code owners for GRPO
Changed Agenda to 'Plan of Action'
* Followup improvements to GRPO job example * Add more docs to the readme * Update sdk/python/jobs/grpo/README.md Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update sdk/python/jobs/grpo/README.md Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com> * Update README.md --------- Co-authored-by: Sharvin Jondhale <shjondhale@microsoft.com> Co-authored-by: Gayatri Penumetsa <181455625+gpenumetsa-msft@users.noreply.github.com>
Co-authored-by: Mohd Javed Ansari <moansa@microsoft.com>
Co-authored-by: Mohd Javed Ansari <moansa@microsoft.com>
* fail sdk installation workflow temporarily
* Replace azureml-defaults with azureml-inference-server-http==1.4.0 in environment configuration files * update azureml-inference-server-http * update inference-schema and joblib in conda.yaml * Replace azureml-defaults with azureml-inference-server-http==1.4.0 in environment configuration files * update azureml-inference-server-http * update inference-schema and joblib in conda.yaml --------- Co-authored-by: Salman Arshad <v-saarshad@microsoft.com>
* Adding mcp server content * updating content * add more content to it * adding more details * adding pr details * formating * updating readme file * updating readme * formating the content * formatting * applying formatng * applying formating * applying formatting * adding more ocntent
* Add config for single node runs * Fix notebook metadata * Run black formatter --------- Co-authored-by: Sharvin Jondhale <shjondhale@microsoft.com>
* [Fix]: Add accuracy also and reword informative assertions
* remove responsibleai-vision notebooks and related workflows
raghu-microsoft
requested review from
achauhan-scc,
imatiach-msft,
jayesh-tanna and
kingernupur
as code owners
June 20, 2025 06:45
|
This pull request sets up GitHub code scanning for this repository. Once the scans have completed and the checks have passed, the analysis results for this pull request branch will appear on this overview. Once you merge this pull request, the 'Security' tab will show more code scanning analysis results (for example, for the default branch). Depending on your configuration and choice of analysis tool, future pull requests will be annotated with code scanning analysis results. For more information about GitHub code scanning, check out the documentation. |
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| from azure.identity import InteractiveBrowserCredential, DefaultAzureCredential | ||
|
|
||
| try: | ||
| credential = DefaultAzureCredential() |
Check failure
Code scanning / CodeQL
Detect unsafe use of DefaultAzureCredential in python application Error
kshitij-microsoft
requested review from
a team,
ahughes-msft,
nemanjarajic and
yunjie-hub
as code owners
June 27, 2025 09:15
sdgilley
approved these changes
Jun 30, 2025
sdgilley
left a comment
sdgilley
left a comment
Contributor
There was a problem hiding this comment.
these changes will not break the docs builds.
* update rai-tabular environment version in rai tabular notebooks
…efore recreating model (#3743) * Update model cleanup to use DELETE API Replaced model archiving with DELETE API calls for cleanup.
* fix: Update rerank notebook for latest models * Update sdk/python/foundation-models/cohere/rerank-cohere-client.ipynb Co-authored-by: Michael <michael.kozakov@gmail.com> * fix typo * remove empty cell * refine base_url for clarity * fix formatting --------- Co-authored-by: Michael <michael.kozakov@gmail.com>
* Add enable_rbac_authorization parameter to Key Vault * Update Docker image version for deployments
* Update Docker image version in entry.spec.yaml * Update Docker image version in entry.spec.yaml * Update Docker image to use CUDA 13.1 * Update Docker image to version 5.0 with CUDA 13.1 * Update conda.yaml * Update entry.spec.yaml * Update Docker image version in entry.spec.yaml * Update entry.spec.yaml * Update entry.spec.yaml * Update entry.spec.yaml * Update conda.yaml * Update Docker image to openmpi5.0-ubuntu24.04 * Update Docker image version in entry.spec.yaml * Change Python version from 3.10 to 3.8.12 * Update Docker image version in entry.spec.yaml * Update Docker image version in entry.spec.yaml * Update Docker image version in entry.spec.yaml * Update Docker image version in entry.spec.yaml
Co-authored-by: Arun <arunsu@microsoft.com>
…score_data failure (#3888) The sklearn-1.5/labels/latest curated environment updated to MLflow 2.19+ around Dec 16 2025. MLflow 2.19 uses the logged-models API which is not supported by AzureML tracking server, breaking mlflow.autolog() in the train step and mlflow.sklearn.load_model() in the predict/score_data step. Pin mlflow<2.19 at runtime in both train.yml and predict.yml commands.
… predict_step failure (#3889) The sklearn-1.5/labels/latest curated environment updated to MLflow 2.19+ around Dec 16 2025. MLflow 2.19 uses the logged-models API which is not supported by AzureML tracking server, breaking mlflow.autolog() in the train step and mlflow.sklearn.load_model() in the predict step. Pin mlflow<2.19 at runtime in both train.yml and predict.yml commands.
* Update command to install dependencies and run training * Fix command syntax for pip installation * update protobuf installation to avoid dependency conflicts * Update AzureML environment reference in notebook * Update AzureML environment reference in notebook
…nt dependencies (#3902) * Update e2e-ml-workflow.ipynb * Update AzureML environment reference in notebook * Update AzureML environment reference in notebook * Remove environment variables from online deployment Removed environment variables from deployment configuration. * Add environment variables to deployment configuration * Update azureml-mlflow version in notebook * Update environment version to 0.1.5 in ML workflow * Update Docker image version in ML workflow * Update environment version to 0.1.6 * update logging and saving of model artifact via MLflow * Fix deployment: custom score.py compatible with mlflow 3.x Three issues fixed: 1. Auto-generated MLflow scoring script imports azureml.ai.monitoring which is missing from auto-built env. Fixed with explicit conda env. 2. The mlflow.pyfunc.scoring_server.infer_and_parse_data and predictions_to_json functions were removed in mlflow 3.x. Rewrote score.py to parse input_data (split-oriented DataFrame) with pandas and serialize predictions with numpy/json directly. 3. Fixed model path: AZUREML_MODEL_DIR points to registration root but MLflow artifacts (MLmodel, model.pkl) are in a subdirectory. The score.py walks the directory to find the MLmodel file. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * Update e2e-ml-workflow.ipynb * Fix formatting in e2e-ml-workflow.ipynb * Format extra_pip_requirements for better readability * Update environment version to 0.1.7 * Update e2e-ml-workflow.ipynb --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
#3928) * fix * Fix AzureML deployment env deps for inference server * Update pipeline.ipynb * Update conda.yaml * Update requirements.txt
…e change (#3948) * Initial plan * Fix storage account key extraction in managed identity notebooks * Handle list_keys API shape differences in managed identity notebooks * Fix managed identity notebook key extraction --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
* Refactor code structure for improved readability and maintainability * Remove sample data download step and add input image for binary payloads deployment
* Fix uai notebook: handle azure-mgmt-msi Identity schema change * updated * Pin azure-mgmt-msi<8 and azure-mgmt-authorization<5 in UAI notebook * UAI notebook: extract principal/client id via as_dict() to support new azure-mgmt-msi schema
#3984) * version upgarde * json correct
#3980) * upgrade * image upgrade * upgrade * upgrade * try method added * fix * try method * fix
…te (#4035) The --set-default flag on batch-deployment create fails with a RequestInvalid error because the API can no longer find 'deployment_name' on BatchEndpointDefaults when sent through that code path. Replace --set-default with a two-step approach: create the deployment, then explicitly update the endpoint with --set defaults.deployment_name to set the default deployment. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* Fix amlsecscan root cron entrypoint ownership * Support cgroup v2 in amlsecscan wrapper * Harden amlsecscan install path * Format amlsecscan scanner changes * Move amlsecscan scheduled entrypoint under etc * Make amlsecscan reinstalls failure-safe Preserve trusted installations during package setup and stage root-owned scanner files with rollback before switching the cron schedule. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
|
You are seeing this message because GitHub Code Scanning has recently been set up for this repository, or this pull request contains the workflow file for the Code Scanning tool. What Enabling Code Scanning Means:
For more information about GitHub Code Scanning, check out the documentation. |
…4077) * Add SDK sample for batch deployment from registry pipeline component * Update batch endpoints readme with registry pipeline deployment sample * Format registry batch deployment sample notebook for black * Add CI workflow for from-registry batch pipeline notebook * Add discoverability README for registry batch deployment sample * Fix typo in autogenerated comment * Improve discoverability README for registry batch deployment sample * Add required discoverability keywords to registry sample README * Refine documentation and component loading logic Updated markdown and code comments for clarity. Added checks for component name and version. * Format registry batch deployment notebook * Normalize notebook JSON for black check * fix black format * Update Python version to use variable for flexibility * add owners for notebook * Remove azure-ai-ml installation command Removed the pip install command for azure-ai-ml from the notebook. * Fix JSON formatting in sdk-deploy-and-test.ipynb Fixed JSON formatting issue by removing an extraneous comma. * Remove azure-ai-ml==1.32.0 pin cell from from-registry batch pipeline notebook The notebook should validate against the CI-provided azure-ai-ml (dev-requirements, or the freshly-built release wheel via setup.sh), not downgrade to a pinned old SDK. Matches the sibling pipeline_with_components_from_yaml notebook, which has no install cell. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 48bc14df-6a6c-4fb3-bb66-4b187cc8163e * Fix from-registry batch sample: unique endpoint name, drop unsupported ComponentDeployment.Enabled property The endpoint used a fixed name and set properties={'ComponentDeployment.Enabled': True}. On re-run, begin_create_or_update tried to update the existing endpoint's immutable properties (also a True vs true case mismatch), failing with EndpointPropertiesUpdateNotSupported. Match sibling notebooks (hello-batch, training-with-components): append a random suffix so a fresh endpoint is created each run, and drop the unnecessary property (PipelineComponentBatchDeployment works without it). Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 48bc14df-6a6c-4fb3-bb66-4b187cc8163e --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 48bc14df-6a6c-4fb3-bb66-4b187cc8163e
* Add batch endpoint job list workarounds to all batch endpoint notebooks Added alternative methods to list batch endpoint jobs when the standard approach encounters issues: Option 1: Using MLflow to search for runs by experiment name (same as the batch endpoint name) Option 2: Using ml_client.jobs.list() and filtering by experiment_name References: - ICM 569867940 - ICM 632272288 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * Convert workaround code cells to markdown to prevent CI execution The workaround code cells were being executed by papermill during CI, causing ModuleNotFoundError for mlflow. Convert to a single markdown cell with fenced code blocks so the samples are visible to customers but not executed during CI runs. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> * Clarify snippets assume default experiment name in batch endpoint notebooks Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: eb4ba443-58fa-4e89-89e0-d77baa4b8c18 --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: eb4ba443-58fa-4e89-89e0-d77baa4b8c18
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
20 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
Checklist