You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am getting a strange error during prediction:
Failed to execute query. Error: Message(s) from 'PREDICT' engine: Input column 1 (1-based) has shape [1,2] which is not supported.
Total execution time: 00:00:03.780
The text was updated successfully, but these errors were encountered:
When the lab train the model, it define the input as shape[1,2]
initial_types = [
("features", FloatTensorType([None, 2]))
I am not sure whether Predict can do this kind of transform before, in order to pass the lab you need to modify and re-train the model to accept two separate column 'customerkey', 'stockitemkey' as input.
Today, I got this issue.
I was able to find the root cause. The table [wwi_ml].[SampleData] (in my case), nor table [wwi].[SampleData] (in the original case), does not exists in Synapses.
I am getting a strange error during prediction:
Failed to execute query. Error: Message(s) from 'PREDICT' engine: Input column 1 (1-based) has shape [1,2] which is not supported.
Total execution time: 00:00:03.780
The text was updated successfully, but these errors were encountered: