Browse Source

modify quant export annotation

tags/v1.1.0
yuchaojie 5 years ago
parent
commit
d0fd18cff3
1 changed files with 2 additions and 2 deletions
  1. +2
    -2
      mindspore/train/quant/quant.py

+ 2
- 2
mindspore/train/quant/quant.py View File

@@ -487,9 +487,9 @@ def export(network, *inputs, file_name, mean=127.5, std_dev=127.5, file_format='
network (Cell): MindSpore network produced by `convert_quant_network`.
inputs (Tensor): Inputs of the `quantization aware training network`.
file_name (str): File name of model to export.
mean (int): Input data mean. Default: 127.5.
mean (int, float): Input data mean. Default: 127.5.
std_dev (int, float): Input data variance. Default: 127.5.
file_format (str): MindSpore currently supports 'AIR', 'ONNX' and 'MINDIR' format for exported
file_format (str): MindSpore currently supports 'AIR' and 'MINDIR' format for exported
quantization aware model. Default: 'AIR'.

- AIR: Graph Engine Intermidiate Representation. An intermidiate representation format of


Loading…
Cancel
Save