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

[BugFix] Fix _reset data passing in parallel env #1880

Merged
merged 5 commits into from
Feb 6, 2024
Merged

[BugFix] Fix _reset data passing in parallel env #1880

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
f30d544
init
vmoens Feb 6, 2024
e02feef
tests
vmoens Feb 6, 2024
5810253
tests
vmoens Feb 6, 2024
fbc101b
Merge remote-tracking branch 'origin/main' into fix-update-reset-penv
vmoens Feb 6, 2024
7e95ffc
amend
vmoens Feb 6, 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
Next Next commit
init
  • Loading branch information
@vmoens
vmoens committed Feb 6, 2024
commit f30d544cbd67646b8a9ef2963e63a1b268dad016
11 changes: 9 additions & 2 deletions torchrl/envs/batched_envs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1295,7 +1295,12 @@ def _reset(self, tensordict: TensorDictBase, **kwargs) -> TensorDictBase:
tensordict_, keys_to_update=list(self._selected_reset_keys)
)
continue
out = ("reset", tensordict_)
if tensordict_ is not None:
self.shared_tensordicts[i].update_(tensordict_)
out = ("reset", True)
else:
out = ("reset", False)

channel.send(out)
workers.append(i)

Expand Down Expand Up @@ -1522,7 +1527,9 @@ def look_for_cuda(tensor, has_cuda=has_cuda):
torchrl_logger.info(f"resetting worker {pid}")
if not initialized:
raise RuntimeError("call 'init' before resetting")
cur_td = env.reset(tensordict=data)
# we use 'data' to tell the env whether there is anything
# to read in the input tensordict
cur_td = env.reset(tensordict=root_shared_tensordict if data else None)
shared_tensordict.update_(
cur_td,
keys_to_update=list(_selected_reset_keys),
Expand Down
Loading