[Feature] Add Stack transform
#2567
Open
+650
−5
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2567
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ❌ 6 New Failures, 18 Unrelated FailuresAs of commit f443812 with merge base 408cf7d ( NEW FAILURES - The following jobs have failed:
FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
Looks like there is a minor bug if I try to use this on |
vmoens
reviewed
Nov 15, 2024
kurtamohler
force-pushed
the
Stack-Transform-0
branch
2 times, most recently
from
dc5cceb
to
23f7e1b
Compare
kurtamohler
force-pushed
the
Stack-Transform-0
branch
from
23f7e1b
to
f443812
Compare
vmoens
reviewed
Nov 19, 2024
There was a problem hiding this comment.
Thanks this is superb
I'd like to discuss the inverse transform:
Would it make sense in the inverse to get an entry (from the input_spec) and unbind it?
Like: you have a single action with leading dim of 2, and map it to ("agent0", "action"), ("agent1", "action"). The spec seen from the outside is the stack of the 2 specs (as it is for stuff processed in forward).
Would that make sense?
| forward = _call | ||
|
|
||
| def _inv_call(self, tensordict: TensorDictBase) -> TensorDictBase: | ||
| values = torch.split(tensordict[self.in_keys_inv[0]], 1, dim=self.dim) |
There was a problem hiding this comment.
does this work too if the key isn't there?
There was a problem hiding this comment.
maybe use unbind instead, not to call squeeze afterwards?
| values = torch.split(tensordict[self.in_keys_inv[0]], 1, dim=self.dim) | ||
| for value, out_key_inv in _zip_strict(values, self.out_keys_inv): | ||
| tensordict.set(out_key_inv, value.squeeze(self.dim)) | ||
| tensordict.exclude(self.in_keys_inv[0], inplace=True) |
There was a problem hiding this comment.
maybe we don't want to do that inplace? Is there a specific reason to use that?
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.
Description
Adds a transform that stacks tensors and specs from different keys of a tensordict into a common key.
Motivation and Context
close #2566
Types of changes
What types of changes does your code introduce? Remove all that do not apply:
Checklist
Go over all the following points, and put an
xin all the boxes that apply.If you are unsure about any of these, don't hesitate to ask. We are here to help!