You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Implement operations commonly found in standard neural network frameworks/libraries. This would allow for benchmarking and comparison testing against other frameworks.
Basic arithmetic
2D Convolution
2D Max pooling
Element-wise nonlinearities (tanh, sin, cos, softplus, log, exp, etc.)
Basic linear algebra (matrix multiplication, transpose, determinant, trace, inverse)
Implement operations commonly found in standard neural network frameworks/libraries. This would allow for benchmarking and comparison testing against other frameworks.
Basic arithmetic
2D Convolution
2D Max pooling
Element-wise nonlinearities (
tanh,sin,cos,softplus,log,exp, etc.)Basic linear algebra (matrix multiplication, transpose, determinant, trace, inverse)
Basic reductions (sum, product, etc.)
Basic logical operators (
and,or,not,<=,>=, etc.)Logical reductions (all true, any true...)
Basic shape operators (reorder, reshape, etc.)
Broadcasting as an explicit operation