13c724a1b3dad1d5eddb51b851c39671982dbb69,sonnet/python/modules/conv.py,CausalConv1D,_build,#CausalConv1D#Any#,1679

Before Change

      raise base.UnderspecifiedError(
          "Number of input channels must be known at module build time")

    if not tf.float32.is_compatible_with(inputs.dtype):
      raise TypeError("Input must have dtype tf.float32, but dtype was {}".
                      format(inputs.dtype))

    weight_shape = (self._kernel_shape[0], input_channels, self.output_channels)

    bias_shape = (self.output_channels,)

    if "w" not in self._initializers:
      self._initializers["w"] = create_weight_initializer(weight_shape[:2])

    if "b" not in self._initializers and self._use_bias:
      self._initializers["b"] = create_bias_initializer(bias_shape)

    self._w = tf.get_variable(
        "w",
        shape=weight_shape,

After Change

      raise base.UnderspecifiedError(
          "Number of input channels must be known at module build time")

    _verify_inputs_dtype(inputs)

    weight_shape = (self._kernel_shape[0], input_channels, self.output_channels)

    bias_shape = (self.output_channels,)

    if "w" not in self._initializers:
      self._initializers["w"] = create_weight_initializer(weight_shape[:2],
                                                          dtype=inputs.dtype)

    if "b" not in self._initializers and self._use_bias:
      self._initializers["b"] = create_bias_initializer(bias_shape,
                                                        dtype=inputs.dtype)

    self._w = tf.get_variable(
        "w",
        shape=weight_shape,
        dtype=inputs.dtype,
        initializer=self._initializers["w"],
        partitioner=self._partitioners.get("w", None),
        regularizer=self._regularizers.get("w", None))

    pad_amount = int((self._kernel_shape[0] - 1) * self._rate[0])
    if self._data_format == DATA_FORMAT_NCW:
      padded_inputs = tf.pad(inputs, paddings=[[0, 0], [0, 0], [pad_amount, 0]])
    else:
      padded_inputs = tf.pad(inputs, paddings=[[0, 0], [pad_amount, 0], [0, 0]])

    outputs = tf.nn.convolution(
        padded_inputs,
        self._w,
        strides=self._stride,
        padding=VALID,
        dilation_rate=self._rate,
        data_format=self._data_format)

    if self._use_bias:
      self._b = tf.get_variable(
          "b",
          shape=bias_shape,
          dtype=inputs.dtype,
          initializer=self._initializers["b"],
          partitioner=self._partitioners.get("b", None),
          regularizer=self._regularizers.get("b", None))

Instances