Conversation
# Conflicts: # ray_lightning/ray_ddp.py
DDP Sharded and Horovod also need to be P0. We have to make sure the entire library works with 1.6. |
1 similar comment
DDP Sharded and Horovod also need to be P0. We have to make sure the entire library works with 1.6. |
Could we also merge in this PR to rename everything to strategy following the 1.6 convention. #129 |
from this test: ray_lightning/ray_lightning/tests/utils.py Lines 241 to 245 in 6aed848
|
for the Q1:
Yes, we need this. On the remote task, the weights is not going to dump / load from the ckpt from the hard-disk. And |
passed all the test except multi-gpu on one node for this version (https://github.com/JiahaoYao/ray_lightning/tree/3df599a8bb1ac917bf6352b8baef63ad64f21595) |
# https://github.com/PyTorchLightning/pytorch-lightning/discussions/8561 | ||
# is fixed. | ||
ddp_kwargs.pop("parallel_devices", None) | ||
ddp_kwargs.pop("cluster_environment", None) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Let's keep this?
|
Todos:
to_state_stream
/load_state_stream
P(0)