tensorflow/models · error · ValueError

Unknown pos encoding %s

Error message

Unknown pos encoding %s

What it means

Error "Unknown pos encoding %s" thrown in tensorflow/models.

Source

Thrown at official/projects/pix2seq/modeling/pix2seq_model.py:61

def add_seq_pos_emb(
    self, pos_encoding, max_seq_len, dim, name_prefix=None, initializer=None
):
  """Add seq_pos_emb variable/tensor to model instance referenced by `self`."""
  if name_prefix is None:
    name_prefix = self.name
  if initializer is None:
    initializer = get_variable_initializer()
  if pos_encoding == "learned":
    self.seq_pos_emb = self.add_weight(
        shape=(max_seq_len + 1, dim),
        initializer=initializer,
        name="%s/seq_pos_embedding" % name_prefix,
    )
  # (gunho) currently only 'learned' positional encoding is supported
  elif pos_encoding == "sin_cos":
    self.seq_pos_emb = None
  else:
    raise ValueError("Unknown pos encoding %s" % pos_encoding)


def add_vocab_token_emb(
    self,
    vocab_size,
    dim,
    output_bias,
    name_prefix=None,
    initializer=None,
):
  """Add token_embedding variable to model instance referenced by `self`."""
  if name_prefix is None:
    name_prefix = self.name
  if initializer is None:
    initializer = get_variable_initializer()
  self.token_embedding = self.add_weight(
      shape=[vocab_size, dim],
      initializer=initializer,

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/pix2seq/modeling/pix2seq_model.py:61 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24). Data as JSON: /api/errors/b38b40d37e0ea5dd. Report an issue: GitHub.