Add values after fitting a model

Hi everyone,

Just a question on how to add values to the model after it has been fitted. For example,

1 4 7
2 5 8
3 6 9

I have 3 columns of data. column C is totally independent of the other data. So I want to remove it and add back to the fitted model after training.

Is there anyway of doing this? Thank you.

Of course. Typically, your predictions come back in the order you provide them to the predict() function. So you can just merge the data back together again at the end

The exception is torchtext which sorts by the field lengths or something. For that you need to set batch size to 1 when evaluating test results so you can maintain the same order

1 Like

Thank you rob, appreciate it.