-
-
Notifications
You must be signed in to change notification settings - Fork 1.6k
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
InvalidArgumentError: LossTensor is inf or nan : Tensor had NaN values #1017
Comments
HI @DuanWei-fudan - what version of tensorflow are you using? I would ask you to fill out the template completely so I can better help! |
hello,thanks for helping firstly,i use the tensorflow-gpu 1.13.1 with cudatoolkit 10.0.130,cudnn 7.6.5 right now.I used the tensorflow-gpu 2.0 ,but the code had some problem,so i change it. I just know a little about the deeplabcut, so i don't know how to solve it. Can I connect with you with wechat or others?Maybe it will be more problems in the future. maybe the tensorflow with 1.13.1 isn't best for my computer? |
Hi @DuanWei-fudan -- easy fix! The 3000 series does not work with tensorflow 1.x; so in short you need to use tensorflow 2, but this requires a new DLC update. Here is the blog about this: http://www.mousemotorlab.org/deeplabcutblog/2020/11/23/rolling-up-to-tensorflow-2 What you can do is install this package: #Install the branch with tf2.x support:
pip install git+https://github.com/DeepLabCut/DeepLabCut-core.git@tf2.2alpha
pip install tf_slim and use the project you already made, just make a new training_dataset! Here is a COLAB notebook to show you what to do: https://colab.research.google.com/github/DeepLabCut/DeepLabCut-core/blob/tf2.2alpha/Colab_TrainNetwork_VideoAnalysis_TF2.ipynb also see this issue: #944 |
my code
The text was updated successfully, but these errors were encountered: