提交 02a8c938 编写于 作者: L liyin

Remove Identity at before and after fake quantize

上级 6d12272f
......@@ -527,6 +527,7 @@ class ConverterOption(object):
else:
self._transformer_option = [
# Model structure related transformation
TransformerRule.REMOVE_IDENTITY_OP,
TransformerRule.TRANSFORM_FAKE_QUANTIZE,
TransformerRule.REMOVE_IDENTITY_OP,
TransformerRule.TRANSFORM_GLOBAL_POOLING,
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册