https://gitcode.net/paddlepaddle/Paddle/-/commit/8d42540f595de44146715687e23dd1b82e7b3e3eadd paddle.async_save to reduce time cost by checkpoint saving (#55115)2023-07-21T11:50:52+08:00Tian121000916+SylarTiaNII@users.noreply.github.com
* add paddle.async_save to reduce time cost by checkpoint saving
* adapt save_for_auto_inference to paddle.async_save
* modify UT
* modify UT
* fix on cpu only version
* revert commit on save_auto_inference
* fix threadinghttps://gitcode.net/paddlepaddle/Paddle/-/commit/9daba606d9e6c3c0b8f6acdb5fba505240e5bc2bmake sharding reduce mode by default (#55529)2023-07-22T10:20:15+08:00sneaxiy32832641+sneaxiy@users.noreply.github.com
* make sharding reduce mode by default
* Update dygraph_sharding_optimizer.py
* Update hybrid_parallel_optimizer.py
* Update pipeline_parallel.pyhttps://gitcode.net/paddlepaddle/Paddle/-/commit/8520a5b352d4e44d8c3cf796df5b7d4b107ef640add check for cembedding (#55621)2023-07-22T14:48:54+08:00ShenLiang1422485404@qq.comhttps://gitcode.net/paddlepaddle/Paddle/-/commit/f275ad2becd4c56475166ae1ee0bb14e0127e82a[Distributed] Support dp/sharding overlap in virtual pp (#55651)2023-07-26T11:39:29+08:00ShenLiang1422485404@qq.com
* Add virtual pp and dp overlap
* add sharding/dp overlap
* add dp/vpp overlap
* fix code
* fix log