).astype(np.float32)

    def data_gen() -> repr_dataset.RepresentativeDataset:
      for _ in range(50):
        yield {'input_tensor': random_tensor_gen_fn()}

    dataset_path = self.create_tempfile('tfrecord').full_path
    path_map = {'serving_default': dataset_path}
    repr_dataset.TfRecordRepresentativeDatasetSaver(path_map).save(
        {'serving_default': data_gen()}
    )

      dtype: The type of the generated input - usually dtypes.float32 for float
        and dtypes.int64 for int
      num_examples: Number of examples in the representative dataset.

    Yields:
      data_gen: A `quantize_model._RepresentativeSample` filled with random
        values.
    """
    for _ in range(num_examples):
      yield {input_key: random_ops.random_uniform(shape, minval, maxval, dtype)}

			listSym.WriteCURelativeAddr(ctxt, listSym.Size, startPC, int64(end))
		}

		i += 2 * ctxt.Arch.PtrSize
		datalen := 2 + int(ctxt.Arch.ByteOrder.Uint16(list[i:]))
		listSym.WriteBytes(ctxt, listSym.Size, list[i:i+datalen]) // copy datalen and location encoding
		i += datalen
	}

	// Location list contents, now with real PCs.
	// End entry.
	listSym.WriteInt(ctxt, listSym.Size, ctxt.Arch.PtrSize, 0)

}

// calculateSignedChunkLength - calculates the length of the overall stream (data + metadata)
func calculateStreamContentLength(dataLen, chunkSize int64) int64 {
	if dataLen <= 0 {
		return 0
	}
	chunksCount := dataLen / chunkSize
	remainingBytes := dataLen % chunkSize
	var streamLen int64
	streamLen += chunksCount * calculateSignedChunkLength(chunkSize)
	if remainingBytes > 0 {