Return tuple of loop state and transformed state #382
Closed
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
seanmor5
reviewed
Oct 12, 2022
lib/axon/loop.ex
Outdated
| @@ -536,7 +536,7 @@ defmodule Axon.Loop do | |||
| # Build loss now so we can use it as a metric | |||
| loss_fn = build_loss_fn(loss) | |||
| {init_fn, step_fn} = train_step(model, loss_fn, optimizer) | |||
| output_transform = fn state -> state.step_state[:model_state] end | |||
| output_transform = fn state -> {state, state.step_state[:model_state]} end | |||
There was a problem hiding this comment.
What I have in mind is keeping this the same fn state -> state.step_state[:model_state] end
And rather in run returning:
{state, output_transform.(state)}
Does that make sense?
There was a problem hiding this comment.
Yes, that does make sense. I've made the changes here bd92764 . Should specific tests be added? I've just updated the current ones.
|
@Ian-GL I am still considering the implications of this...it may be that we end up essentially copying the model state twice which is inefficient. I will revisit this after I finish a few other things for the Axon release |
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Outdated suggestions cannot be applied.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.
If I understood correctly #362 , the new output of the transform should be
{%Axon.Loop.State{...}, transformed_state}, so this PR implements said change and updates the tests. This should be the case for other default transforms too? So far I just changedtrainerbut I guessevaluatorshould follow suit. Is that right @seanmor5 ?Closes #362