-
Notifications
You must be signed in to change notification settings - Fork 5.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
support inference for quantized matmul_v2 #36594
Conversation
Thanks for your contribution! |
4968505
to
ce1e650
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
need to add unit testing
@@ -621,7 +642,8 @@ void QuantDequantFusePass::ApplyImpl(ir::Graph* graph) const { | |||
std::unordered_set<std::string> quant_types = { | |||
"fake_quantize_range_abs_max", "fake_quantize_moving_average_abs_max"}; | |||
std::unordered_set<std::string> quantized_op_types = { | |||
"conv2d", "mul", "matmul", "depthwise_conv2d", "fc", "conv2d_transpose"}; | |||
"conv2d", "mul", "matmul", "matmul_v2", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这里多余的空格处理一下。
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
done
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这里代码风格要求是第一行放满,然后上下对齐,所以空格无法去掉,也无法写成每个单独成行
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
LGTM |
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
* support inference for quantized matmul_v2 * undate code style * code style
PR types
New features
PR changes
Others
Describe
支持了量化后的matmul_v2在tensorrt上的推理