How to search the run id of an experiment run created in another notebook?

[yorabhir]

Hello,

I have created an experiment using 

with mlflow.start_run(run_name='experment_1'):

in a notebook say 'notebook_1'.  In the 'Experiments' tab if I click on 'notebook_1', I am able to see 'experiment_1'.
 
Now I am trying to search the experiment in another notebook using
mlflow.search_runs(filter_string='tags.mlflow.runName = "experiment_1"') which is giving me None but if I run the same in 'notebook_1' I am able to see the details of the run (even after I Detach & re-attach or restart the cluster)

I have also tried 
mlflow.search_experiments(filter_string="name = 'experiment_1'") and MlflowClient.search_experiments(filter_string="name = 'experiment_1'") but they also giving output None.

[atmcqueen · Accepted solution]

Hi,

I believe the run name is an attribute, not a tag.

Try:
my_run = mlflow.search_runs(
search_all_experiments=True,
filter_string="attributes.run_name='experiment_1'"
)

To search for runs in specific notebooks, you can add "tags.environment='notebook_name'" to the filter string, so the filter string would then be:
"attributes.run_name='experiment_1' AND tags.enviroment='your_notebook'"

I think your issue with using search_experiments() was that you were providing the name of the run and not the name of the experiment, though I'm not sure if you were using the same name for both.

Here are the docs for mlflow.search_runs() and mlflow.search_experiments().

Hope this helps.

[yorabhir]

Thank you @atmcqueen , the solution is working.