[RLlib] Fix SampleBatch to_device() #27572
Conversation
This reverts commit 74686a8.
kouroshHakha
requested review from
sven1977,
gjoliver,
avnishn,
ArturNiederfahrenhorst,
smorad,
maxpumperla and
krfricke
as code owners
| # Floatify all float64 tensors. | ||
| if tensor.dtype == torch.double: | ||
| if tensor.is_floating_point(): |
There was a problem hiding this comment.
any chance this is a new api? if yes, maybe make sure it works with the version we pin for core.
There was a problem hiding this comment.
good point. I checked quickly and it's even there on torch=1.8.0 Where do we specify the torch pinned version in ray?
https://pytorch.org/docs/1.8.0/search.html?q=IS_FLOATING_POINT&check_keywords=yes&area=default
| # Numpy arrays. | ||
| if isinstance(item, np.ndarray): |
There was a problem hiding this comment.
what were we doing before ... didn't we check these conditions before we get in here :)
There was a problem hiding this comment.
the else is added because I moved a bit of logic to clean the code. e.g torch.is_tensor() is brought down after dealing with RepeatedValues type.
| "f": RepeatedValues(np.array([[1, 2, 0, 0]]), lengths=[2], max_len=4), | ||
| SampleBatch.SEQ_LENS: np.array([2, 3, 1]), | ||
| "state_in_0": np.array([1.0, 3.0, 4.0]), | ||
| SampleBatch.INFOS: np.array([{"a": 1}, {"b": 2}, {"c": 3}]), |
There was a problem hiding this comment.
INFOS's dtype is object right? and we basically don't do anything to it if I understand correctly.
Signed-off-by: Huaiwei Sun <[email protected]>
Signed-off-by: Stefan van der Kleij <[email protected]>
Why are these changes needed?
Related issue number
closes #26593
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.