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

Is this a problem with the data set #32

Open
vaeXu opened this issue Apr 28, 2018 · 3 comments
Open

Is this a problem with the data set #32

vaeXu opened this issue Apr 28, 2018 · 3 comments

Comments

@vaeXu
Copy link

vaeXu commented Apr 28, 2018

I used two days of training ./rmpe_dataset_server ./train_dataset.h5 5555,but it keeps:
Total samples: 2645
Epoch 1
curr_sample/total_samples/curr_epoch = 1/2645/1
What is the reason for this?Is the amount of data too large? What I use is the provided COCO2014 data set.If someone knows the reason, please answer your questions. Be grateful.

@anatolix
Copy link
Owner

anatolix commented May 4, 2018

That is not correct dataset size. Something is wrong. Looks more like validation dataset

@abinjoabraham
Copy link

@anatolix @GAVANXU Is it mandatory to go with the coco2014 data?? I was using the latest version of the dataset coco2017

@anatolix
Copy link
Owner

anatolix commented May 7, 2018

No it is not, I also train with 2017. But original work uses coco2014 for some reason. Imho 2014 and 2017 datasets are same for 90% pictures.

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

No branches or pull requests

3 participants