Return an example's size as a float or tuple. This value is used when filtering a dataset with ``--max-positions``.

Transform a human-readable string into a sequence of int ids. The ids should be in the range [num_reserved_ids, vocab_size). Ids [0,

:param sample: :param batch_idx: :return: output: dict

Sanity check a few things before starting actual training. :param model:

Initialize vocabulary with tokens from token_generator.

Run evaluation code. :param model: PT model :param dataloaders: list of PT dataloaders :param max_batches: Scalar

Build sinusoidal embeddings. This matches the implementation in tensor2tensor, but differs slightly from the description in Sectio

Logs the metric dict passed in. :param metrics: :param grad_norm_dic:

r"""Applies each `module` in :attr:`modules` in parallel on arguments contained in :attr:`inputs` (positional) and :attr:`kwargs_tup` (keyword)

Reduces output according to the training mode. Separates loss from logging and tqdm metrics :param output: :return:

Strip ids_to_strip from the end ids.

Saves a checkpoint atomically, avoiding the creation of incomplete checkpoints. This will create a temporary checkpoint with a suffix of ``

r""" Recursively find all tensors contained in the specified object.

Load vocab from a file. Args: filename: The file to load vocabulary from.

Initialize tokens from a list of tokens. It is ok if reserved tokens appear in the vocab list. They will be removed. The set of to

Repeat each frame according to duration. Examples: >>> x = torch.tensor([[1], [2], [3]]) tensor([[1],

:param sample: :param batch_idx: :return: total loss: torch.Tensor, loss_log: dict

:param outputs: :return: loss_output: dict

Transform a sequence of int ids into a human-readable string. EOS is not expected in ids. Args: ids: list of integers t

Transform a sequence of int ids into a their string versions. This method supports transforming individual input/output ids to their