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.

Training Error : bg_num_rois = 0 and fg_num_rois = 0

See original GitHub issue

selection_003

ValueError: bg_num_rois = 0 and fg_num_rois = 0, this should not happen!

The error comes after 2500 iterations into training. Here is the command line I used for training.

CUDA_VISIBLE_DEVICES=0 python trainval_net.py --dataset coco --net vgg16 --bs 4 --lr 0.001 --cuda

I tried changing FG_THRESH and BG_THRESH but with no luck. Anything I am missing something ?

Issue Analytics

  • State:closed
  • Created 6 years ago
  • Comments:10 (3 by maintainers)

github_iconTop GitHub Comments

2reactions
jwyangcommented, Feb 18, 2018

I see, I guess there is a minor bug in data loader, already found it, will fix that.

0reactions
xiexijuncommented, Aug 26, 2021

Hi, all I have fixed that bug for a while, please pull the repo again.

Could you tell me where you fixed? I faced the same problem in other codes that based on your code

Read more comments on GitHub >

github_iconTop Results From Across the Web

Does zero training error mean zero bias? - Cross Validated
I was implicitly thinking of classification and 0/1 error – if a model is unbiased, does it classify all training points with perfect...
Read more >
Lecture 6: Bias and variance - MS&E 226: “Small” Data
We saw how we can estimate prediction error using validation or test sets. ... In particular, if Cov( ˆf(Xi),Yi|X) > 0, then training...
Read more >
powereedge 2850 pcie training error - Dell Community
i have a poweredge 2850 that was working but i messed around in the bios and ended up changing something that i dont...
Read more >
Why You Should Ignore the Training Error - RapidMiner
Figure 2: Calculating the training error of a 1-Nearest Neighbor classifier always leads to a 0% training error – even on completely random...
Read more >
CPSC 340: Data Mining Machine Learning
0. 1. 1. X = y = Egg Milk Fish Wheat Shellfish Peanuts … 0.5. 0. 1. 0.6. 2. 1. 0. 0.7. 0....
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