Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add Adapt.adapt_structure method for Optimisers.Leaf #180

Merged
merged 9 commits into from
Nov 12, 2024
Merged

Add Adapt.adapt_structure method for Optimisers.Leaf #180

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
8077900
Adapt.adapt_structure method for Optimisers.Leaf
vpuri3 Oct 3, 2024
8906e50
import Adapt.jl
vpuri3 Oct 3, 2024
0c66a52
add Adapt.jl to Project.toml
vpuri3 Oct 3, 2024
f48c128
adapt compat
vpuri3 Oct 3, 2024
e8ba26e
based on discussion: adapt_structure method does not maintain IdDict …
vpuri3 Oct 4, 2024
89c6fd0
Update ext/OptimisersAdaptExt.jl
vpuri3 Nov 11, 2024
3269bd9
Merge branch 'master' into patch-1
vpuri3 Nov 11, 2024
8c1c073
edit warning to indicate that this is a correctness issue
vpuri3 Nov 12, 2024
978ed6f
Update ext/OptimisersAdaptExt.jl
vpuri3 Nov 12, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions Project.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,12 +11,15 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"

[weakdeps]
Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
EnzymeCore = "f151be2c-9106-41f4-ab19-57ee4f262869"

[extensions]
OptimisersAdaptExt = ["Adapt"]
OptimisersEnzymeCoreExt = "EnzymeCore"

[compat]
Adapt = "4"
ChainRulesCore = "1"
EnzymeCore = "0.8.5"
Functors = "0.4.9, 0.5"
Expand Down
20 changes: 20 additions & 0 deletions ext/OptimisersAdaptExt.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
module OptimisersAdaptExt

import Adapt
import Optimisers: Leaf

function Adapt.adapt_structure(to, leaf::Leaf)
@warn """`Optimisers.Leaf` object does not support device transfer via
`Adapt.jl`. This is because `Adapt.jl` does not handle shared parameters (i.e. the same parameter array
appearing more than once in the model), and in such cases this will lead to incorrect gradient updates.
Avoid this by calling `Flux.gpu/cpu` or `MLDataDevices.cpu_device()/gpu_device()` on the
optimiser state object.
""" maxlog=1
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we want this only once? It's a potential correctness bug, not a performance issue.

Suggested change
""" maxlog=1
"""

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is your worry that the message will get lost in the shuffle somehow? My thought was that people may have a valid use for this, and as long as they know what they're getting into the library doesn't have to remind them on every call.

Another practical concern would be what happens when someone tries to call cu(large state tree). Not setting maxlog would mean other logging is drowned out because this warning would trigger for every Leaf.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I agree that printing 100 times on a big model is too much. Ideal would be once, on every invocation, IMO... but that's hard to make happen.

It's not the world's biggest correctness bug to ignore shared parameters, so maybe we should live with it. Maybe the message should say that's what the problem is?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@mcabbott i edited the warning to say that this is a correctness issue.


rule = Adapt.adapt(to, leaf.rule)
state = Adapt.adapt(to, leaf.state)

Leaf(rule, state, leaf.frozen)
end

end
Loading