[BUG] Deffault hyper opts produce wrong models

[holinov]

Bug Description

Found thad deafult hyper-opt tuning mechanism producec invalid options combinations

Hyperparameter                                    |Value                                             |Best Value So Far
rnn_block_1/layer_type                            |gru                                               |gru
rnn_block_1/num_layers                            |1                                                 |1
rnn_block_2/layer_type                            |gru                                               |gru
rnn_block_2/num_layers                            |3                                                 |3
regression_head_1/dropout                         |0.5                                               |0
dense_block_1/use_batchnorm                       |False                                             |False
dense_block_1/num_layers                          |2                                                 |2
dense_block_1/units_0                             |32                                                |32
dense_block_1/dropout                             |0.5                                               |0.5
dense_block_1/units_1                             |32                                                |32
classification_head_1/dropout                     |0                                                 |0
optimizer                                         |adam                                              |adam
learning_rate                                     |0.001                                             |0.001
dense_block_1/units_2                             |1024                                              |None

As you could see it tries to build model with

dense_block_1/num_layers                          |2  
dense_block_1/units_0                             |32 
dense_block_1/units_1                             |32 
dense_block_1/units_2                             |1024 

Bug Reproduction

Code for reproducing the bug:

layer = ak.Input()
layer = ak.DenseBlock()(layer)
class = ak.ClassificationHead()(layer)

Data used by the code: Any data

Setup Details

Include the details about the versions of:

• OS type and version: Win10
• Python: 3.5, 3.8
• autokeras: 1.0.12
• keras-tuner: 1.0.2
• scikit-learn: 0.24.0
• numpy: 1.19.5
• pandas: 1.2.0
• tensorflow: 2.4.0

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.