cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Observer
Observer
685 Views
Registered: ‎07-02-2019

error when I use decent_q to quantizing faster rcnn

WechatIMG2.jpeg

I try to use decent_q to quantize faster rcnn.And the input node is Placeholder,the output nodes are resnet_v1_50/cls_score/weights and so on.

But when I finished the calibration,there was the mistake in quantize_utils.cc line 357.But I can't find the file,what's the problem? If the input node was wrong?

thank you!

0 Kudos
5 Replies
Highlighted
Xilinx Employee
Xilinx Employee
649 Views
Registered: ‎05-24-2019

Re: error when I use decent_q to quantizing faster rcnn

@tongtyr  I am surprised to see this message for op: "Placeholder" as typically "Placeholder" is the input node to the graph.  Are you able to attached your pb file or an excerpt of the first few layers of the graph from a pbtxt?

One thing that you can do if you haven't already is run decent_q in inspect mode.  This will give you an indication of what decent_q believes are the correct input/output nodes.  Example:

decent_q inspect --input_frozen_graph rcnngraph.pb

In general the message that you are seeing appears when one or more layers in the graph is not supported by the quantizer.  I generally modify the output node for the quantization script progressively in an effort to determine which layer(s) are causing problems.  In this case, the behaviour seems odd if indeed Placeholder is the first layer as you would expect it to be.

--Quenton

 

 

0 Kudos
Highlighted
Observer
Observer
630 Views
Registered: ‎07-02-2019

Re: error when I use decent_q to quantizing faster rcnn

Here is my pb.txt.

It shows that the placeholder'shape is "1,?,?,3",and I have a question that if the shape of input_node should be the same as the input_images' shape.

Are there same layers like "rpn" and nodes in faster-rcnn that DNNDK doesn't support?

I am very thankful that you answered me!

thanks!

0 Kudos
Highlighted
Observer
Observer
629 Views
Registered: ‎07-02-2019

Re: error when I use decent_q to quantizing faster rcnn

0 Kudos
Highlighted
Observer
Observer
572 Views
Registered: ‎07-04-2019

Re: error when I use decent_q to quantizing faster rcnn

Hello, I have encountered the same problem as you. Have you solved it ?
0 Kudos
Observer
Observer
346 Views
Registered: ‎07-02-2019

Re: error when I use decent_q to quantizing faster rcnn

have you solved it? I have on idea about it

0 Kudos