You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

floordiv.py 1.7 kB

5 years ago
12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849
  1. #!/usr/bin/env python3
  2. # coding: utf-8
  3. # Copyright 2019 Huawei Technologies Co., Ltd
  4. #
  5. # Licensed under the Apache License, Version 2.0 (the "License");
  6. # you may not use this file except in compliance with the License.
  7. # You may obtain a copy of the License at
  8. #
  9. # http://www.apache.org/licenses/LICENSE-2.0
  10. #
  11. # Unless required by applicable law or agreed to in writing, software
  12. # distributed under the License is distributed on an "AS IS" BASIS,
  13. # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14. # See the License for the specific language governing permissions and
  15. # limitations under the License.
  16. """operator dsl function:floordiv"""
  17. import akg.topi
  18. import akg
  19. import akg.lang.cce
  20. from akg.ops.math.reciprocal import reciprocal
  21. from akg.utils import validation_check as vc_util, kernel_exec as utils
  22. @vc_util.check_input_type(akg.tvm.tensor.Tensor, akg.tvm.tensor.Tensor)
  23. def floordiv(data1, data2):
  24. """
  25. Calculate x/y, and always returns an integer which is floored.
  26. Args:
  27. data1 (tvm.tensor.Tensor): Tensor of type float16, float32.
  28. data2 (tvm.tensor.Tensor): Tensor of type float16, float32.
  29. Returns:
  30. tvm.tensor.Tensor, has type of int32.
  31. """
  32. vc_util.ops_dtype_check([data1.dtype, data2.dtype], vc_util.DtypeForDavinci.ALL_FLOAT)
  33. shape1 = [x.value for x in data1.shape]
  34. vc_util.check_shape(shape1)
  35. shape2 = [x.value for x in data2.shape]
  36. vc_util.check_shape(shape2)
  37. if utils.product_is_mini():
  38. rec = reciprocal(data2, high_precision=True)
  39. res = data1 * rec
  40. else:
  41. res = akg.topi.divide(data1, data2)
  42. res = akg.lang.cce.floor(res)
  43. return res