Skip to content
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 some ops for full quantization #40083

Merged
merged 5 commits into from
Mar 15, 2022

Conversation

yghstill
Copy link
Contributor

@yghstill yghstill commented Mar 2, 2022

PR types

Function optimization

PR changes

APIs

Describe

Support some op for full quantization

@paddle-bot-old
Copy link

paddle-bot-old bot commented Mar 2, 2022

Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

"hard_sigmoid", "conv2d_transpose", "gru", "bilinear_interp",
"nearest_interp", "trilinear_interp", "flatten", "flatten2", "transpose",
"pad2d", "reshape", "layer_norm", "split", "flatten_contiguous_range",
"squeeze2", "nearest_interp_v2", "fill_constant_batch_size_like"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

list后边加个逗号,修改文件后的diff会更直观。

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

done

@@ -1804,7 +1780,8 @@ class AddQuantDequantPass(object):
"squeeze", "elementwise_sub", "mul", "matmul", "relu", "relu6",
"leaky_relu", "tanh", "swish", "scale", "transpose", "transpose2",
"sigmoid", "pad2d", "flatten", "flatten2", "batch_norm", "layer_norm",
"matmul_v2"
"matmul_v2", "split", "flatten_contiguous_range", "squeeze2",
"nearest_interp_v2", "fill_constant_batch_size_like"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

加逗号

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

done

@@ -137,6 +109,10 @@
"flatten2": [["X"], ["Out"]],
"unsqueeze2": [["X"], ["Out"]],
"flatten_contiguous_range": [['X'], ["Out"]],
"split": [['X'], ["Out"]],
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

split需要特殊处理下么?类似于concat。

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

out_threshold只输出一个即可,其余暂不需要特殊处理。

@paddle-bot-old
Copy link

Sorry to inform you that 423912c's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

@yghstill yghstill changed the title Support some op for full quantization Support some ops for full quantization Mar 14, 2022
Copy link
Contributor

@wanghaoshuang wanghaoshuang left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@yghstill yghstill merged commit 7ced301 into PaddlePaddle:develop Mar 15, 2022
@yghstill yghstill deleted the add_full_quantization_op branch March 15, 2022 12:38
liqitong-a pushed a commit to liqitong-a/Paddle that referenced this pull request Mar 17, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants