Issues

• [BUG] Writes to an unpickled consolidated tensordict are lost when it is pickled again

  bug
  pytorch/tensordict#2503 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] TensorDict.from_consolidated(filename) loads the batch size and names from consolidate() time, not the current ones

  bug
  pytorch/tensordict#2502 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] NonTensorData.empty(names=...) ignores the names it is given

  bugtensorclass
  pytorch/tensordict#2501 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] LazyStackedTensorDict.empty() drops all dim names

  bug
  pytorch/tensordict#2500 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] apply(fn, batch_size=...) and to(dtype, batch_size=...) skip UnbatchedTensor entries

  bug
  pytorch/tensordict#2499 · Filter by author peterdsharpe opened on Oct 13, 2026
  1

• [BUG] to(attrs, inplace=True, num_threads=...) on a nested tensordict raises and leaves the nested leaves at the root

  bug
  pytorch/tensordict#2498 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] Reductions with reduce=True raise a torch.cat error on a tensordict without tensor leaves

  bugtriaged
  pytorch/tensordict#2497 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] TensorDict.max(dim=..., reduce=True) and min(dim=..., reduce=True) raise AttributeError

  bug
  pytorch/tensordict#2496 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] td.add(1, alpha=2) and td.sub(1, alpha=2) raise TypeError when other is a number

  bug
  pytorch/tensordict#2495 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] TypedTensorDict reshape, masked_select and repeat_interleave with inplace=True return a new object instead of self

  bug
  pytorch/tensordict#2494 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] reshape, flatten and repeat with inplace=True on an h5 PersistentTensorDict rewrite the file and then raise

  bugpersistent
  pytorch/tensordict#2493 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] reshape, flatten and unflatten with inplace=True keep dim names that no longer match the batch dims

  bug
  pytorch/tensordict#2492 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] reshape and view give wrong shapes on a lazy stack with a zero-size dim

  bug
  pytorch/tensordict#2491 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] pad_sequence raises TypeError on tensordicts that hold only non-tensor data

  bugfunctional
  pytorch/tensordict#2490 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] update_ on a shared tensordict drops its strings when the source also holds a tensor

  bug
  pytorch/tensordict#2487 · Filter by author peterdsharpe opened on Oct 13, 2026

• [CI] test_chunksize_num_chunks hangs at random on macOS

  CI
  pytorch/tensordict#2483 · Filter by author peterdsharpe opened on Oct 13, 2026

• [CI] macOS jobs fail at random when a child process aborts with "OMP: Error #15"

  CI
  pytorch/tensordict#2482 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] set_at_ of a string into a non-tensor entry fails on a lazy stack whose stack dim is not 0

  bug
  pytorch/tensordict#2459 · Filter by author peterdsharpe opened on Oct 13, 2026

• [BUG] torch.cat(..., out=<lazy stack>) raises when an input has size 0 along the cat dim

  bugtriaged
  pytorch/tensordict#2416 · Filter by author peterdsharpe opened on Oct 12, 2026

• [BUG] NonTensorStack.reshape gives the wrong batch size for a stack with a zero-size dim

  bugtriaged
  pytorch/tensordict#2415 · Filter by author peterdsharpe opened on Oct 12, 2026

• [BUG] to_pytree does not rebuild a list or dict value whose dict key is not a NestedKey

  bugtriaged
  pytorch/tensordict#2414 · Filter by author peterdsharpe opened on Oct 12, 2026

• [BUG] unflatten does not check that the sizes multiply up to the dim size when a tensordict has no tensor leaves

  bugtriaged
  pytorch/tensordict#2413 · Filter by author peterdsharpe opened on Oct 12, 2026

• [BUG] memmap(num_threads>0) ignores errors from its worker threads and returns a tensordict without the failed entries

  bugmemmaptriaged
  pytorch/tensordict#2406 · Filter by author peterdsharpe opened on Oct 12, 2026

• [BUG] store[i] of a locked LazyStackedTensorDictStore is not locked, so it can add keys

  bugstoretriaged
  pytorch/tensordict#2404 · Filter by author peterdsharpe opened on Oct 12, 2026
  1

• [BUG] update(..., inplace=True) on a locked LazyStackedTensorDictStore raises a bare NotImplementedError

  bugstoretriaged
  pytorch/tensordict#2403 · Filter by author peterdsharpe opened on Oct 12, 2026