|
From this old issue it seems no - #1164 ; can you confirm there is still no multi-device nor multi-node data parallel support in GluonTS? |
Answered by
lostella
Oct 7, 2022
Replies: 1 comment
|
This is not supported in MXNet-based models (the training loop we use does not support that currently). For PyTorch models, we rely on Lightning for training, so there it may be possible (but I haven't done it, so this needs to be validated). |
0 replies
Answer selected by
la-cruche
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This is not supported in MXNet-based models (the training loop we use does not support that currently). For PyTorch models, we rely on Lightning for training, so there it may be possible (but I haven't done it, so this needs to be validated).