Ziyue Jiang 59e343328d [Pipeline Middleware ] Fix deadlock when num_microbatch=num_stage (#2156) ...

* add splitter

* polish code

* remove comment

* fix async nan by moving to cpu first

Co-authored-by: Ziyue Jiang <ziyue.jiang@gmail.com>

2022-12-23 11:38:43 +08:00

..

algorithms

[fx] refactor memory utils and extend shard utils. (#1754)

2022-10-26 14:24:41 +08:00

experimental

[autoparallel] refactor the runtime apply pass and add docstring to passes (#1757)

2022-10-25 14:32:22 +08:00

__init__.py

[fx] metainfo_trace as an API. (#1873)

2022-11-10 20:58:37 +08:00

adding_split_node_pass.py

[Pipeline Middleware ] Fix deadlock when num_microbatch=num_stage (#2156)

2022-12-23 11:38:43 +08:00

concrete_info_prop.py

[fx] refactor memory utils and extend shard utils. (#1754)

2022-10-26 14:24:41 +08:00

meta_info_prop.py

[SC] add GPT example for auto checkpoint (#1889)

2022-11-11 23:17:25 +08:00

passes_for_gpt2_test.py

[hotfix] fix some bugs during gpt2 testing (#1379)

2022-07-28 17:21:07 +08:00

shard_1d_pass.py

[Doc] add more doc for ColoTensor. (#1458)

2022-08-16 10:38:41 +08:00

split_module.py

[Pipeline] Add Topo Class (#2059)

2022-12-02 18:13:20 +08:00

utils.py

[Pipeline] Add Topo Class (#2059)

2022-12-02 18:13:20 +08:00