0.2.15

Fixed some bugs.

0.2.14

Adaptive memory to avoid out of memory error.

0.2.13

Fixed the bug in evaluation always output 100.0.

0.2.12

Fixed the bug in the evaluation.
limited the memory usage with Constant.MAX_MODEL_SIZE.

0.2.11

Changed dependency to tensorflow 1.10.0.

0.2.10

Fixed the bug in setup.py

0.2.9

Update setup.py (#147 )

0.2.8

Reformat the standard output.
Further limited the search space.

0.2.7

Fixed a bug of deeper dense layer.

0.2.6

Use torch.multiprocessing to avoid the error caused by multiprocessing in Windows.

Limit the model size to avoid memory overflow.