mirror of https://github.com/hpcaitech/ColossalAI
![]() * support existing sharded and unsharded parameters in zero * add unitest for moe-zero model init * polish moe gradient handler |
||
---|---|---|
.. | ||
__init__.py | ||
_base_gradient_handler.py | ||
_data_parallel_gradient_handler.py | ||
_moe_gradient_handler.py | ||
_pipeline_parallel_gradient_handler.py | ||
_sequence_parallel_gradient_handler.py | ||
_zero_gradient_handler.py | ||
utils.py |