v1.2.3

• cli update

v1.2.2

• Support Custom C Structure input kernel arg
• add tests

v1.2.1

• more cli command features support
• tune can be start from cli

v1.2.0

support tuning:

• class Tuner
• docs

v1.1.4

• fix benchmark config
• update docs

v1.1.3

• add benchmark

v1.1.2

• support timer info return by run()
• add docs readthedocs

v1.1.1

• add release function for reload kernel, useful when using in a jupyter notebook
• update docs
• update pyi files

v1.1.0

• support different type for constant args
• support worksize -1

v1.0.2

• release gpu memory after run kernel
• update python api, support more way to load kernel
• update readme and example