- Newest
- Most votes
- Most comments
Thanks for your question
There is some guidance and information about selecting a proper baseline from here that may help: https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-feature-attribute-shap-baselines.html
if I am passing a single row as input to shap_baseline parameter, the schedule is failing by concatenating 2 rows
For this, you may want to open a ticket via AWS Support in the AWS console so we can take a better look at this internally, and include your processing job details if possible (such as your analysis configurations, Job ARN, etc.)
If i ignore the shap_baseline (as it is optional), the schedule is taking forever to run
There is a num_clusters
parameter in the SHAPConfig which you can explicitly set to reduce the size of the baseline dataset that is generated. A lower number (low single digits) will generally provide faster runtime
Relevant content
- asked 3 years ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated a year ago