Commit 807c8bd5 authored by Jean Ibarz's avatar Jean Ibarz
Browse files

Removed non-linear activation function 'sigmoid' in models, as my brother told...

Removed non-linear activation function 'sigmoid' in models, as my brother told me that it is not appropriate for regression. It effectively improves the training of the 'default' model, while it doesn't change the behavior of 'one_sided' model.
parent 7b483138
......@@ -36,8 +36,8 @@ def default_model_creator(model_config):
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=1, activation='sigmoid'),
tkl.experimental.preprocessing.Rescaling(scale=360, offset=0.0)
tkl.Dense(units=1),
tkl.experimental.preprocessing.Rescaling(scale=360)
])
return model
......@@ -92,8 +92,8 @@ def one_sided_model_creator(model_config):
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=10, activation='relu'),
tkl.Dense(units=1, activation='sigmoid'),
tkl.experimental.preprocessing.Rescaling(scale=180, offset=0.0)
tkl.Dense(units=1),
tkl.experimental.preprocessing.Rescaling(scale=180)
])
return model
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment