Closed daholste closed 5 years ago
You can simply remove col=Ignore:TX:16,25
from your schema. This will not load those columns in the first place.
I did try that, but got an exception when trying to load col 17 (after first ignored col, ie w/o first loading/mentioning col 16)
Closing: Microsoft.ML.PipelineInference
has been removed from the repository.
System information
Issue
Contents of automl_.graph.json:
What happened? Even though Ignore:TX:16,25 is specified in the schema, the learners start expensive / slow text transforms / n-gram extractions on the columns
What did you expect? The columns to be ignored