Handle special case for tensorflow cat input

README.md

@@ -14,6 +14,7 @@ Nowadays, the cost of computing many different models and hyperparameters is muc
 * Correct serialization of data pipeline encoders on disk (i.e. no pickled Python objects!)
 * Retrain the generated model on new data without making any code/pipeline changes.
 * Quit the hyperparameter search at any time, as the results are saved after each trial.
+* Training progress bars with ETAs for both the overall experiment and per-epoch during the experiment.
 
 The models generated by automl-gs are intended to give a very strong *baseline* for solving a given problem; they're not the end-all-be-all that often accompanies the AutoML hype, but the resulting code is easily tweakable to improve from the baseline.
 

automl-gs/templates/models/tensorflow/categorical.py

@@ -1,3 +1,4 @@
     # {{ field_raw }}
-    input_{{ field }} = Input(shape=(len(encoders['{{ field }}_encoder'].classes_),), name="input_{{ field }}")
+    input_{{ field }}_size = len(encoders['{{ field }}_encoder'].classes_)
+    input_{{ field }} = Input(shape=(input_{{ field }}_size if input_{{ field }}_size != 2 else 1,), name="input_{{ field }}")