question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

MXNet shared weights

See original GitHub issue

Hello, I’ve tried this tool recently, and feel it’s really wonderful. But i find that there is some room for improvement of visualization when weights are sharing between convs in mxnet. In the below, i share the weights of three convs. While the weight id of both the middle and the right conv are id: [563,0] , the left conv’s weight id is id: stage4_unit2_conv1_weight. 图片 图片 图片 By guessing, I can figure out the left conv weight id is id: [563,0]. But it will be much better if the weight id of the original conv is given in digital form explicitly as well, along with its string name.

Thank you~

Issue Analytics

  • State:closed
  • Created 4 years ago
  • Comments:5 (3 by maintainers)

github_iconTop GitHub Comments

1reaction
lutzroedercommented, Mar 28, 2019

@huangh12 can you verify the fix? Steps here.

0reactions
huangh12commented, Mar 28, 2019

Problem is fixed. Thx!

Read more comments on GitHub >

github_iconTop Results From Across the Web

Best practice for shared weights between different layers?
Dear all, I want to use the same weights (only) across different layers. For example, assuming I have two layers (say a convolution...
Read more >
softmax - Avoiding weight sharing among certain layers in ...
Is there any way to train such a model in mxnet, where I want to share the weights for all the layers but...
Read more >
Gluon Package — mxnet documentation
Based on the Gluon API specification, the Gluon API in Apache MXNet provides a clear, concise, ... ParameterDict for sharing weights with the...
Read more >
What is Apache MxNet and How it works? An Overview and Its ...
Apache MXNet is an open-source deep learning software framework, ... In this design note, we share the rationale for the specific choices ...
Read more >
Transfer Learning using MXNet (Part 7) - Heartbeat
Transfer Learning using MXNet (Part 7) — Transferring Pre-Trained Weights to a ... find out how to use the other two approaches from...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found