tensorflow/models · error · ValueError

%s does not match any files.

Error message

%s does not match any files.

What it means

Error "%s does not match any files." thrown in tensorflow/models.

Source

Thrown at official/legacy/bert/input_pipeline.py:89

          tf.io.FixedLenFeature([seq_length], tf.int64),
      'segment_ids':
          tf.io.FixedLenFeature([seq_length], tf.int64),
      'masked_lm_positions':
          tf.io.FixedLenFeature([max_predictions_per_seq], tf.int64),
      'masked_lm_ids':
          tf.io.FixedLenFeature([max_predictions_per_seq], tf.int64),
      'masked_lm_weights':
          tf.io.FixedLenFeature([max_predictions_per_seq], tf.float32),
  }
  if use_next_sentence_label:
    name_to_features['next_sentence_labels'] = tf.io.FixedLenFeature([1],
                                                                     tf.int64)
  if use_position_id:
    name_to_features['position_ids'] = tf.io.FixedLenFeature([seq_length],
                                                             tf.int64)
  for input_pattern in input_patterns:
    if not tf.io.gfile.glob(input_pattern):
      raise ValueError('%s does not match any files.' % input_pattern)

  dataset = tf.data.Dataset.list_files(input_patterns, shuffle=is_training)

  if input_pipeline_context and input_pipeline_context.num_input_pipelines > 1:
    dataset = dataset.shard(input_pipeline_context.num_input_pipelines,
                            input_pipeline_context.input_pipeline_id)
  if is_training:
    dataset = dataset.repeat()

    # We set shuffle buffer to exactly match total number of
    # training files to ensure that training data is well shuffled.
    input_files = []
    for input_pattern in input_patterns:
      input_files.extend(tf.io.gfile.glob(input_pattern))
    dataset = dataset.shuffle(len(input_files))

  # In parallel, create tf record dataset for each train files.
  # cycle_length = 8 means that up to 8 files will be read and deserialized in

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/legacy/bert/input_pipeline.py:89 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/050ce42bc4203af9. Report an issue: GitHub.