[PoC][MoE & EP] model code and various parallelisms #725
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
[ghstack-poisoned]
tianyu-l
added a commit
that referenced
this pull request
Dec 9, 2024
ghstack-source-id: ebaafd51eeab3a33f45012c5c204d7979a9c67f0 Pull Request resolved: #725
facebook-github-bot
added
the
CLA Signed
[ghstack-poisoned]
tianyu-l
added a commit
that referenced
this pull request
Dec 9, 2024
ghstack-source-id: 96435a0a0e70b5a2f6931608b3692ae264c89b58 Pull Request resolved: #725
tianyu-l
changed the title
The expert-choice MoE implementation is mostly from torchtune: pytorch/torchtune#1902 Temporary changes to unblock exploration - [pytorch] comment out the check at https://github.com/pytorch/pytorch/blob/main/torch/distributed/tensor/parallel/api.py#L66 - [torchtitan] for dp2ep, turn optimizers `foreach` and `clip_grad_norm_` off, as not all parameters are DTensors on the same meshes (e.g. in dp2ep `moe.router.gate` is a replicate torch.Tensor) - [torchtitan] for dp2ep, comment out `apply_fsdp` which would leave the non-expert parameters replicate Todo - FSDP / CP integration Haven't worked on - softmax scoring when Router Parallel is used (currently only sigmoid) - token-choice MoE - shared expert overlapping [ghstack-poisoned]
tianyu-l
added a commit
that referenced
this pull request
Dec 10, 2024
ghstack-source-id: f2ffdba21a4408392b1083ab8328909fb41fb16b Pull Request resolved: #725
|
closing as this is reorganized in #732 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):