Distributed layers #1270
Draft
Distributed layers #1270
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
awni
reviewed
Jul 17, 2024
| sl = cls(input_dims, output_dims, False, group) | ||
| # The multiplication with 1.0 forces a copy, perhaps change to | ||
| # something better when available. | ||
| sl.weight = linear_layer.weight[r * step : (r + 1) * step] * 1 |
There was a problem hiding this comment.
Is it possible the input buffer could be donated so we'd still hold on to the memory?
If so, maybe another option is to do sl.weight[ ... ] = ... that will force the copy since it's a slice update?
There was a problem hiding this comment.
Nice! That does sound better actually!
|
I kind of like this design. I like that it's all quite simple and easy to follow and we have a lot of control over how to shard the model (as in ml-explore/mlx-examples#890). We could possibly find a way to reduce the code needed for adding a new custom linear-like layer.. but the simplicity is nice, I wouldn't want to give that up. |
angeloskath
force-pushed
the
distributed-layers
branch
from
6090542
to
fea9644
Compare
angeloskath
force-pushed
the
distributed-layers
branch
2 times, most recently
from
061d214
to
b32ce2c
Compare
angeloskath
force-pushed
the
distributed-layers
branch
from
b32ce2c
to
ab26116
Compare
angeloskath
force-pushed
the
distributed-layers
branch
from
ab26116
to
3d431c0
Compare
Merged
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Outdated suggestions cannot be applied.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.
Adds linear layers that allow training and inference of a model sharded across several devices. The main things added are
simply changing linear layers to the above results in a model that works out of the box with distributed inference and training.
I am starting it as a draft so that we can iterate a bit on the design. The negative aspects of the above design are that we have yet another linear layer to think about when implementing LoRA and friends or weird new quantizations for instance. Perhaps it would be better to make the above layers with an internal linear layer so model surgery that swaps linear layers would still work out of the box.