In [ ]:
from fastai.gen_doc.nbdoc import *
from fastai.tabular import *
In [ ]:
show_doc(tabular_learner)
You can customize the automatic embeddings sizes picked by the library by passing a dictionary emb_szs
to match categorical variable names with an embedding size. emb_drop
, ps
. y_range
and use_bn
are passed to TabularModel
, the kwargs
are passed to Learner
. See tabular
for an example of use.