Baizhou Zhang 0ceec8f9a9 [pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354) ... * add naive optimizer for 3DPlugin/refactor gpt2 shardformer test * merge tests of PP/DP/TP combinations into one test file * fix bug when sync grad for dp in HybridPlugin * update supported precisions for 3DPlugin/fix bug when shifting tp_degree * improve the passing of lazy_init * modify lazy_init/use sync_shared_params		2023-08-15 23:25:14 +08:00
..
__init__.py	[plugin] add 3d parallel plugin (#4295)	2023-08-15 23:25:14 +08:00
dp_plugin_base.py	[booster] update prepare dataloader method for plugin (#3706)	2023-05-08 15:44:03 +08:00
gemini_plugin.py	[zero]support no_sync method for zero1 plugin (#4138)	2023-07-31 22:13:29 +08:00
hybrid_parallel_plugin.py	[pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354)	2023-08-15 23:25:14 +08:00
low_level_zero_plugin.py	[zero] support shard optimizer state dict of zero (#4194)	2023-07-31 22:13:29 +08:00
plugin_base.py	[zero]support no_sync method for zero1 plugin (#4138)	2023-07-31 22:13:29 +08:00
pp_plugin_base.py	[plugin] add 3d parallel plugin (#4295)	2023-08-15 23:25:14 +08:00
torch_ddp_plugin.py	[zero]support no_sync method for zero1 plugin (#4138)	2023-07-31 22:13:29 +08:00
torch_fsdp_plugin.py	[zero]support no_sync method for zero1 plugin (#4138)	2023-07-31 22:13:29 +08:00