Jiarui Fang a590ed0ba3 [zero] improve the accuracy of get_memory_usage of sharded param (#538)

2022-03-28 16:19:19 +08:00

..

amp

[hotfix] fix initialize bug with zero (#442)

2022-03-17 13:16:22 +08:00

builder

add pytorch hooks (#179)

2022-01-25 22:20:54 +08:00

communication

fix format (#332)

2022-03-11 15:50:28 +08:00

context

[polish] polish singleton and global context (#500)

2022-03-23 18:03:39 +08:00

engine

[refactor] remove old zero code (#517)

2022-03-25 14:54:39 +08:00

kernel

[cuda] modify the fused adam, support hybrid of fp16 and fp32 (#497)

2022-03-25 14:15:53 +08:00

logging

[log] polish disable_existing_loggers (#519)

2022-03-25 12:30:55 +08:00

nn

[zero]added hybrid adam, removed loss scale in adam (#527)

2022-03-25 18:03:54 +08:00

registry

add pytorch hooks (#179)

2022-01-25 22:20:54 +08:00

testing

[test] fixed rerun_on_exception and adapted test cases (#487)

2022-03-25 17:25:12 +08:00

trainer

Added profiler communication operations

2022-03-11 15:50:28 +08:00

utils

[zero] fix grad offload (#528)

2022-03-25 18:23:25 +08:00

zero

[zero] improve the accuracy of get_memory_usage of sharded param (#538)

2022-03-28 16:19:19 +08:00

__init__.py

Develop/experiments (#59)

2021-12-09 15:08:29 +08:00

constants.py

fix format constants.py (#358)

2022-03-11 15:50:28 +08:00

core.py

[polish] polish singleton and global context (#500)

2022-03-23 18:03:39 +08:00

global_variables.py

[MOE] add unitest for MOE experts layout, gradient handler and kernel (#469)

2022-03-21 13:35:04 +08:00

initialize.py

[polish] polish singleton and global context (#500)

2022-03-23 18:03:39 +08:00