| .. |
|
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
|
[MOE] remove old MoE legacy (#493)
|
2022-03-22 17:37:16 +08:00 |
|
engine
|
[zero] polish sharded param name (#484)
|
2022-03-22 14:36:16 +08:00 |
|
kernel
|
[formart] format fixed for kernel\cuda_native codes (#335)
|
2022-03-11 15:50:28 +08:00 |
|
logging
|
[log] better logging display with rich (#426)
|
2022-03-16 09:51:15 +08:00 |
|
nn
|
[zero] sharded model support the reuse of fp16 shard (#495)
|
2022-03-23 14:59:59 +08:00 |
|
registry
|
add pytorch hooks (#179)
|
2022-01-25 22:20:54 +08:00 |
|
testing
|
[test] added rerun on exception for testing (#475)
|
2022-03-21 15:51:57 +08:00 |
|
trainer
|
Added profiler communication operations
|
2022-03-11 15:50:28 +08:00 |
|
utils
|
[MOE] remove old MoE legacy (#493)
|
2022-03-22 17:37:16 +08:00 |
|
zero
|
[zero] sharded model support the reuse of fp16 shard (#495)
|
2022-03-23 14:59:59 +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
|
[MOE] polish moe_env (#467)
|
2022-03-19 15:36:25 +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
|
[MOE] add unitest for MOE experts layout, gradient handler and kernel (#469)
|
2022-03-21 13:35:04 +08:00 |
9ec1ce6ab1