yifanmao 7537b19c73 [FSDP2] Update ignored_params docstring and add unit test (#149074) ... Fixes https://github.com/pytorch/pytorch/issues/148242 ignored_params won't be moved to devices in full_shard(), update docstring. Add unit test `test_move_states_to_device_ignored_param_device` to show that ignored_params won't be moved during full_shard(), but would be after `model.cuda()` Pull Request resolved: https://github.com/pytorch/pytorch/pull/149074 Approved by: https://github.com/awgu		2025-03-15 00:23:09 +00:00
..
fsdp	[FSDP2] Update ignored_params docstring and add unit test (#149074)	2025-03-15 00:23:09 +00:00
test_composability	PEP585 update - test (#145176)	2025-01-22 04:48:28 +00:00
test_checkpoint.py	PEP585 update - test (#145176)	2025-01-22 04:48:28 +00:00
test_contract.py	PEP585 update - test (#145176)	2025-01-22 04:48:28 +00:00
test_replicate_with_compiler.py	[ddp] decouple python reducer from compilation mode (#147123)	2025-02-19 15:51:40 +00:00
test_replicate.py	[FSDP2] Move to public torch.distributed.fsdp (#141868)	2024-12-07 01:24:28 +00:00