WeichenXu comments

Results 273 comments of


                                            WeichenXu

[BUG] Computing SHAP explanations fails during mlflow.evaluate()

You can also try seting `explainability_algorithm` to be `kernel`. The default shap tree explainer might be buggy.

[BUG] Computing SHAP explanations fails during mlflow.evaluate()

@khojarohan4 For the shap tree explainer, you can file ticket in shap repo, I think it is shap bug. :)

[FR]Delete child runs together with partent runs

I think this is a good proposal. We can add a param `delete_child_runs`, to keep backwards compatibilty, set the param to be false by default. @PhilipMay Are you interested in...

[FR] [Roadmap] Add utility for signature validation of logged model to dataset

@serena-ruan volunteers to do this task, assign it to her. :)

MlFlow load model form azure blob error: blob doesn't exist

@rjtshrm Thanks for reporting it ! Do you know the root cause of the issue ?

[BUG] Require validation for `MlflowClient.log_batch` when updating params

Make sense! @Mathanraj-Sharma Would you create a PR to fix it ? I will help review.

[FR] Make log_param() return the passed parameter value (expression for functional use)

@serena-ruan volunteers to do this task, assign it to her. :)

[FR] Make log_param() return the passed parameter value (expression for functional use)

@dbczumar Shall we make `MlflowClient.log_param` return param value as well ?

How to obtain the azureml mlflow tracking server URL

Could you try run "mlflow.get_tracking_uri()" to get the tracking server uri ?

How to obtain the azureml mlflow tracking server URL

Do you mean to access the server outside of the "AzureML" ? I am not sure whether "AzureML" provides you the permission to do that, i.e., you need the permission...