ColossalAI/colossalai
Jianghai 376533a564
[shardformer] zero1+pp and the corresponding tests (#4517)
* pause

* finish pp+zero1

* Update test_shard_vit.py
2023-08-28 10:51:16 +08:00
..
_C
_analyzer
amp [pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354) 2023-08-15 23:25:14 +08:00
auto_parallel
autochunk
booster [shardformer] support sharded checkpoint IO for models of HybridParallelPlugin (#4506) 2023-08-25 22:04:57 +08:00
builder
checkpoint_io [shardformer] support sharded checkpoint IO for models of HybridParallelPlugin (#4506) 2023-08-25 22:04:57 +08:00
cli fix localhost measurement (#4320) 2023-08-01 10:14:00 +08:00
cluster [shardformer] support interleaved pipeline (#4448) 2023-08-16 19:29:03 +08:00
communication [NFC] fix: format (#4270) 2023-07-26 14:12:57 +08:00
context
device
engine
fx
interface [pipeline] refactor 1f1b schedule (#4115) 2023-08-15 23:25:14 +08:00
kernel [shardformer] update shardformer to use flash attention 2 (#4392) 2023-08-15 23:25:14 +08:00
lazy [shardformer] support lazy init (#4202) 2023-08-15 23:25:14 +08:00
logging
nn [doc] add Series A Funding and NeurIPS news (#4377) 2023-08-04 17:42:07 +08:00
pipeline [shardformer] zero1+pp and the corresponding tests (#4517) 2023-08-28 10:51:16 +08:00
registry
shardformer [shardformer] support sharded checkpoint IO for models of HybridParallelPlugin (#4506) 2023-08-25 22:04:57 +08:00
tensor [pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354) 2023-08-15 23:25:14 +08:00
testing
trainer
utils [test] remove useless tests (#4359) 2023-08-01 18:52:14 +08:00
zero [shardformer] zero1+pp and the corresponding tests (#4517) 2023-08-28 10:51:16 +08:00
__init__.py
constants.py
core.py
global_variables.py
initialize.py