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 sliced PRB when only traj is provided #2228

Merged
merged 4 commits into from
Jun 14, 2024
Merged

[BugFix] Fix sliced PRB when only traj is provided #2228

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
b0ea50d
init
vmoens Jun 14, 2024
15f1f0d
amend
vmoens Jun 14, 2024
7c0eaa4
amend
vmoens Jun 14, 2024
3b32f20
amend
vmoens Jun 14, 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
17 changes: 9 additions & 8 deletions test/test_rb.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2230,12 +2230,12 @@ def test_slice_sampler(
def test_slice_sampler_at_capacity(self, sampler):
torch.manual_seed(0)

trajectory0 = torch.tensor([3, 3, 0, 1, 1, 1, 2, 2, 2, 3])
trajectory1 = torch.arange(2).repeat_interleave(5)
trajectory0 = torch.tensor([0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3])
trajectory1 = torch.arange(2).repeat_interleave(6)
trajectory = torch.stack([trajectory0, trajectory1], 0)

td = TensorDict(
{"trajectory": trajectory, "steps": torch.arange(10).expand(2, 10)}, [2, 10]
{"trajectory": trajectory, "steps": torch.arange(12).expand(2, 12)}, [2, 12]
)

rb = ReplayBuffer(
Expand Down Expand Up @@ -2469,7 +2469,8 @@ def test_slice_sampler_strictlength(self):
@pytest.mark.parametrize("ndim", [1, 2])
@pytest.mark.parametrize("strict_length", [True, False])
@pytest.mark.parametrize("circ", [False, True])
def test_slice_sampler_prioritized(self, ndim, strict_length, circ):
@pytest.mark.parametrize("at_capacity", [False, True])
def test_slice_sampler_prioritized(self, ndim, strict_length, circ, at_capacity):
torch.manual_seed(0)
out = []
for t in range(5):
Expand All @@ -2491,9 +2492,9 @@ def test_slice_sampler_prioritized(self, ndim, strict_length, circ):
if ndim == 2:
data = torch.stack([data, data])
rb = TensorDictReplayBuffer(
storage=LazyTensorStorage(data.numel(), ndim=ndim),
storage=LazyTensorStorage(data.numel() - at_capacity, ndim=ndim),
sampler=PrioritizedSliceSampler(
max_capacity=data.numel(),
max_capacity=data.numel() - at_capacity,
alpha=1.0,
beta=1.0,
end_key="done",
Expand Down Expand Up @@ -2530,8 +2531,8 @@ def test_slice_sampler_prioritized(self, ndim, strict_length, circ):
assert (samples["traj"] == 0).any()
# Check that all samples of the first traj contain all elements (since it's too short to fullfill 10 elts)
sc = samples[samples["traj"] == 0]["step_count"]
assert (sc == 0).sum() == (sc == 1).sum()
assert (sc == 0).sum() == (sc == 4).sum()
assert (sc == 1).sum() == (sc == 2).sum()
assert (sc == 1).sum() == (sc == 4).sum()
assert rb._sampler._cache
rb.extend(data)
assert not rb._sampler._cache
Expand Down
29 changes: 24 additions & 5 deletions torchrl/data/replay_buffers/samplers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1068,7 +1068,9 @@ def _get_stop_and_length(self, storage, fallback=True):
"Could not get a tensordict out of the storage, which is required for SliceSampler to compute the trajectories."
)
vals = self._find_start_stop_traj(
trajectory=trajectory, at_capacity=storage._is_full
trajectory=trajectory,
at_capacity=storage._is_full,
cursor=getattr(storage, "_last_cursor", None),
)
if self.cache_values:
self._cache["stop-and-length"] = vals
Expand Down Expand Up @@ -1803,7 +1805,7 @@ def _padded_indices(self, shapes, arange) -> torch.Tensor: # noqa: F811
.flip(0)
)

def _preceding_stop_idx(self, storage, lengths, seq_length):
def _preceding_stop_idx(self, storage, lengths, seq_length, start_idx):
preceding_stop_idx = self._cache.get("preceding_stop_idx")
if preceding_stop_idx is not None:
return preceding_stop_idx
Expand All @@ -1828,6 +1830,13 @@ def _preceding_stop_idx(self, storage, lengths, seq_length):
# Mask the rightmost values of that padded tensor
preceding_stop_idx = pad[:, -seq_length + 1 + span_right :]
preceding_stop_idx = preceding_stop_idx[preceding_stop_idx >= 0]
if storage._is_full:
preceding_stop_idx = (
preceding_stop_idx
+ np.ravel_multi_index(
tuple(start_idx[0].tolist()), storage._total_shape
)
) % storage._total_shape.numel()
if self.cache_values:
self._cache["preceding_stop_idx"] = preceding_stop_idx
return preceding_stop_idx
Expand All @@ -1838,7 +1847,9 @@ def sample(self, storage: Storage, batch_size: int) -> Tuple[torch.Tensor, dict]
start_idx, stop_idx, lengths = self._get_stop_and_length(storage)
seq_length, num_slices = self._adjusted_batch_size(batch_size)

preceding_stop_idx = self._preceding_stop_idx(storage, lengths, seq_length)
preceding_stop_idx = self._preceding_stop_idx(
storage, lengths, seq_length, start_idx
)
if storage.ndim > 1:
# we need to convert indices of the permuted, flatten storage to indices in a flatten storage (not permuted)
# This is because the lengths come as they would for a permuted storage
Expand All @@ -1851,12 +1862,14 @@ def sample(self, storage: Storage, batch_size: int) -> Tuple[torch.Tensor, dict]
)

# force to not sample index at the end of a trajectory
vals = torch.tensor(self._sum_tree[preceding_stop_idx.cpu().numpy()])
self._sum_tree[preceding_stop_idx.cpu().numpy()] = 0.0
# and no need to update self._min_tree

starts, info = PrioritizedSampler.sample(
self, storage=storage, batch_size=batch_size // seq_length
)
self._sum_tree[preceding_stop_idx.cpu().numpy()] = vals
# We must truncate the seq_length if (1) not strict length or (2) span[1]
if self.span[1] or not self.strict_length:
if not isinstance(starts, torch.Tensor):
Expand All @@ -1866,7 +1879,13 @@ def sample(self, storage: Storage, batch_size: int) -> Tuple[torch.Tensor, dict]
# Find the stop that comes after the start index
# say start_tensor has shape [N, X] and stop_idx has shape [M, X]
# diff will have shape [M, N, X]
diff = stop_idx.unsqueeze(1) - starts_tensor.unsqueeze(0)
stop_idx_corr = stop_idx.clone()
stop_idx_corr[:, 0] = torch.where(
stop_idx[:, 0] < start_idx[:, 0],
stop_idx[:, 0] + storage._len_along_dim0,
stop_idx[:, 0],
)
diff = stop_idx_corr.unsqueeze(1) - starts_tensor.unsqueeze(0)
# filter out all items that don't belong to the same dim in the storage
mask = (diff[:, :, 1:] != 0).any(-1)
diff = diff[:, :, 0]
Expand All @@ -1876,7 +1895,7 @@ def sample(self, storage: Storage, batch_size: int) -> Tuple[torch.Tensor, dict]
diff[diff < 0] = diff.max() + 1
# Take the arg min along dim 0 (thereby reducing dim M)
idx = diff.argmin(dim=0)
stops = stop_idx[idx, 0]
stops = stop_idx_corr[idx, 0]
# TODO: here things may not work bc we could have spanning trajs,
# though I cannot show that it breaks in the tests
if starts_tensor.ndim > 1:
Expand Down
Loading