- Code quality has been promoted by introducing linting check in CI and auto code format as helper script. For linting, the tools,
cpplint
andpylint
, are used and configured to comply google coding styles details intool/linting/
. Similarily, formating tools,clang-format
andyapf
configured with google coding styles, are the recommended one for developers to clean code before submitting changes, details intool/code-format/
. - Tensor APIs adds some new ones for naming consitency, and feature enhancement:
- size(), mem_size(), get_value(), to_proto(), l1(), l2(): added for the sake of naming consitency
- AsType(): convert data type between
float
andint
- ceil(): perform element-wise ceiling of the input
- concat(): concatenate two tensor
- index selector: e.g. tensor1[:,:,1:,1:]
- softmax(in, axis): allow to perform softmax on a axis on a multi-dimensionals tensor
- DNNL(Deep Neural Network Library), powered by Intel, is integrated into
model/opertaions/[batchnorm|pooling|convolution]
, the changes is opaque to the end users. The current version is dnnl v1.1 which replaced previous integration of mkl-dnn v0.18. The framework could boost the performance of dl operations when executing on CPU. The dnnl dependency is installed through conda. - marks some Tensor APIs as deprecated which could be replaced by broadcast, and it can support better on multi-dimensional operations
- add_column(), add_row(), div_column(), div_row(), mult_column(), mult_row()