You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

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		1 year ago
..
grad_scaler	[zero] fix gradient clipping in hybrid parallelism (#2521)	2 years ago
mixed_precision_mixin	[bf16] add bf16 support (#3882)	2 years ago
__init__.py	[NFC] polish colossalai/amp/naive_amp/__init__.py code style (#1905)	2 years ago
_fp16_optimizer.py	[setup] support pre-build and jit-build of cuda kernels (#2374)	2 years ago
_utils.py	[NFC] polish colossalai/amp/naive_amp/_utils.py code style (#1816)	2 years ago
mixed_precision_optimizer.py	[pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354)	1 year ago
naive_amp.py	[amp] add gradient clipping for unit tests (#2283)	2 years ago