issues
search
microsoft
/
solution-accelerator-many-models
MIT License
193
stars
85
forks
source link
Adding optional code for automl train + prediction in one pipeline
#154
Open
ScoGroMSFT
opened
1 year ago
ScoGroMSFT
commented
1 year ago
Adding optional code for demonstrating how to add prediction to an automl training pipeline
Demonstrates the need to move automlpipelinebuilder-generated files between calls to avoid overwriting them.
This was implemented on a customer engagement where they wanted to train + predict within the same pipeline
The resulting pipeline looks like this: