_summary_ Args: in_channels (int): feature size of input (3 or 6) input_transform (bool, optional): whether to use tr

_summary_ Args: in_channels (int): feature size of input (3 or 6) input_transform (bool, optional): whether to use tr

Dummy constructor. Use copy_from* and create_from* class methods instead.

key_first_k: dict str: int Only take first k data from these keys (to improve perf)

Initialize the EfficientScalarRecorder. Parameters: - K: Number of largest scalars to consider when computing the av

Ensure that arrays and tensors in nested dictionary or list or tuple have size @size in dim @dim. Args: x (dict or list or tupl

Clones all torch tensors and numpy arrays in nested dictionary or list or tuple and returns a new nested structure. Args: x (dic

Makes all torch tensors and numpy arrays contiguous in nested dictionary or list or tuple and returns a new nested structure. Args:

Open a on-disk zarr directly (for dataset larger than memory). Slower.

Detaches all torch tensors in nested dictionary or list or tuple and returns a new nested structure. Args: x (dict or list or tu

x : [ batch_size x in_channels x horizon ] cond : [ batch_size x cond_dim] returns: out : [ batch_si

x: (B,T,input_dim) timestep: (B,) or int, diffusion step local_cond: (B,T,local_cond_dim) global_cond: (B,global_cond

x : [ batch_size x in_channels x horizon ] cond : [ batch_size x cond_dim] returns: out : [ batch_si

x: (B,T,input_dim) timestep: (B,) or int, diffusion step local_cond: (B,T,local_cond_dim) global_cond: (B,global_cond

Given a nested dictionary or list or tuple, gathers an element from each sequence of the batch for tensors with leading dimensions [B, T, ...

Given a tensor with leading dimensions [B, T, ...], gather an element from each sequence in the batch given an index for each sequence.