{"record":{"id":"b70f28ce64789599","repo":"tensorflow/models","slug":"self-class-does-not-support-block-sparse-att","errorCode":null,"errorMessage":"{self.__class__} does not support block sparse attention.","messagePattern":"(.+?) does not support block sparse attention\\.","errorType":"exception","errorClass":"ValueError","httpStatus":null,"severity":"error","filePath":"official/nlp/modeling/layers/pack_optimization.py","lineNumber":146,"sourceCode":"    if self._norm_first:\n      return source_attention_output + layer_output  # pyrefly: ignore[unbound-name]\n\n    layer_output = tf.cast(layer_output, tf.float32)\n    return self._output_layer_norm(layer_output + attention_output)\n\n\n@tf_keras.utils.register_keras_serializable(package='Text')\nclass StridedReZeroTransformer(rezero_transformer.ReZeroTransformer):\n  \"\"\"ReZeroTransformer for packing optimization to stride over inputs.\"\"\"\n\n  def __init__(self, *args, **kwargs):\n    super().__init__(*args, **kwargs)\n    if self._output_range is not None:\n      raise ValueError(f'{self.__class__} does not '\n                       'support `output_range` argument.')\n    # TODO(b/337888023): Support block sparse attention with strided inputs.\n    if self._src_block_size is not None:\n      raise ValueError(f'{self.__class__} does not '\n                       'support block sparse attention.')\n\n  def call(self, inputs, stride: tf.Tensor):\n    if isinstance(inputs, (list, tuple)):\n      if len(inputs) == 2:\n        input_tensor, attention_mask = inputs\n        key_value = None\n      elif len(inputs) == 3:\n        input_tensor, key_value, attention_mask = inputs\n      else:\n        raise ValueError(f'Unexpected inputs to {self.__class__} with '\n                         f'length at {len(inputs)}.')\n    else:\n      input_tensor, key_value, attention_mask = (inputs, None, None)\n\n    target_tensor = input_tensor[:, ::stride, :]\n    if attention_mask is not None:\n      attention_mask = attention_mask[:, ::stride, :]","sourceCodeStart":128,"sourceCodeEnd":164,"githubUrl":"https://github.com/tensorflow/models/blob/e006f5f0d534913e49c1f1dae87364039fa607e2/official/nlp/modeling/layers/pack_optimization.py#L128-L164","documentation":"Error \"{self.__class__} does not support block sparse attention.\" thrown in tensorflow/models.","triggerScenarios":"Thrown at official/nlp/modeling/layers/pack_optimization.py:146 when the library encounters an invalid state.","commonSituations":"See trigger scenarios.","solutions":[],"exampleFix":null,"handlingStrategy":null,"validationCode":null,"typeGuard":null,"tryCatchPattern":null,"preventionTips":[],"tags":[],"backgroundTag":null,"analyzedSha":"e006f5f0d534913e49c1f1dae87364039fa607e2","analyzedAt":"2026-08-24T14:09:15.576Z","schemaVersion":2},"datasetVersion":"2026-08-24T17:17:21.512Z"}