Skip to content

T
tensorflow

xxadev / tensorflow
与 Fork 源项目一致

从无法访问的项目Fork

通知 3
Star 1
Fork 0
T
tensorflow

前往新版Gitcode，体验更适合开发者的 AI 搜索 >>

提交 6ab2c521 编写于 作者: E Elena Zhelezina
浏览文件
操作

Added 16x8 op tests for conv_activations. 

Change-Id: Ie5e53c48284eac0ad80b75b1514cd5e3e4494232
上级 3b8ef731
隐藏空白更改
内联 并排
Showing with 3 addition and 0 deletion
tensorflow/lite/testing/op_tests/conv_activation.py
浏览文件 @ 6ab2c521
...@@ -40,6 +40,7 @@ def make_conv_activation_tests(activation_op): ...@@ -40,6 +40,7 @@ def make_conv_activation_tests(activation_op):
"constant_filter": [True, False], "constant_filter": [True, False],
"channel_multiplier": [1, 2], "channel_multiplier": [1, 2],
"fully_quantize": [False], "fully_quantize": [False],
"quant_16x8": [False],
"dynamic_range_quantize": [False], "dynamic_range_quantize": [False],
}, },
# TODO(b/134702301): The fully_quantize param is just ignored by the # TODO(b/134702301): The fully_quantize param is just ignored by the
...@@ -55,6 +56,7 @@ def make_conv_activation_tests(activation_op): ...@@ -55,6 +56,7 @@ def make_conv_activation_tests(activation_op):
"constant_filter": [True], "constant_filter": [True],
"channel_multiplier": [1, 2], "channel_multiplier": [1, 2],
"fully_quantize": [True], "fully_quantize": [True],
"quant_16x8": [False, True],
"dynamic_range_quantize": [False], "dynamic_range_quantize": [False],
}, },
{ {
...@@ -67,6 +69,7 @@ def make_conv_activation_tests(activation_op): ...@@ -67,6 +69,7 @@ def make_conv_activation_tests(activation_op):
"constant_filter": [True], "constant_filter": [True],
"channel_multiplier": [1, 2], "channel_multiplier": [1, 2],
"fully_quantize": [False], "fully_quantize": [False],
"quant_16x8": [False],
"dynamic_range_quantize": [True], "dynamic_range_quantize": [True],
}, },
] ]
......
Markdown is supported
0% 或 .
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 或