Skip to content

issues Search Results · repo:lufficc/SSD language:Python

Filter by

209 results
 (56 ms)

209 results

inlufficc/SSD (press backspace or delete to remove)

lufficc/SSD
The problem of conversion from pth format model to onnx model

Hello, I hope there is a problem in the process of converting your pre training model to onnx. Can you help me solve it? I used: torch.onnx.export For weights / vgg_ ssd300_ Voc0712.pth,But,The program ...
contributions welcome
enhancement
  • NOBUGW
  • 3
  • Opened 
    on Oct 12, 2020
  • #176

lufficc/SSD
train time

How long does it take to train the coco dataset? @lufficc Thank you.
  • Keeping7
  • 1
  • Opened 
    on Oct 4, 2020
  • #175

lufficc/SSD
own dataset test

Hi,I have trained the model on my own datasets,but how can i test my trained result? if isinstance(dataset, MyDataset): return my_own_evaluation(**args) how to code the my_own_evaluation function ?could ...
  • ShiqiSong
  • 1
  • Opened 
    on Sep 23, 2020
  • #174

lufficc/SSD
Test mAP

Hi, @lufficc. I trained a DIOR dataset, 90k iterations passed. On the validation mAP is 79%, but the test mAP is 0%. What could have happened? Why overfit? Val mAP - 79.74810936215799 Classes: array([ ...
  • Vlad15lav
  • Opened 
    on Sep 21, 2020
  • #173

lufficc/SSD
mAP is 0. Is it normal?

I trained my dataset with ssd300( I changed the input size to 1280* 384 ), and do not use the pre-trained model, after 5000 iters, the mAP was still 0. Is this normal?
  • GuoxiongZhang
  • 1
  • Opened 
    on Sep 21, 2020
  • #172

lufficc/SSD
adding CoordConv on ssd

Hi @lufficc I ve trained the model using my own custom dataset and it worked perfect. Thank you for your work! I wanted to enhance the result so I m trying to add CoordConv on ssd. I added CoordConv ...
  • 91dana
  • Opened 
    on Sep 21, 2020
  • #171

lufficc/SSD
Learning rate is always 0.001

I m trying to change learning rate in defaults.py (particularly _C.SOLVER.LR = 1e-3). However, it remains the same even though I change this value to 1e-6 or 1e-1. Is this bug or something wrong with yacs? ...
  • priteshgohil
  • 1
  • Opened 
    on Sep 19, 2020
  • #170

lufficc/SSD
Question: Why logs are per iteration and not per epoch?

I know that it is easy to log per epoch by calculating the mean of all the iteration. But is there any specific reason for working with iteration and not per epoch?
  • priteshgohil
  • 2
  • Opened 
    on Sep 16, 2020
  • #169

lufficc/SSD
assign_priors return value when zero ground truths

Hi @lufficc I was trying to train on the custom dataset and many of the images have zero objects. In that case, how should I deal with assign_priors() function? Normally it throws the following error ...
  • priteshgohil
  • 8
  • Opened 
    on Sep 15, 2020
  • #168

lufficc/SSD
Training problem on custom dataset

Hi. I m training on a DIOR dataset, but I have a problem with that. Loss: { reg_loss : tensor(413458.5938, device= cuda:0 , grad_fn= DivBackward0 ), cls_loss : tensor(17.1876, device= cuda:0 , grad_fn= ...
  • Vlad15lav
  • 9
  • Opened 
    on Sep 13, 2020
  • #167
Issue origami icon

Learn how you can use GitHub Issues to plan and track your work.

Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub Issues
ProTip! 
Press the
/
key to activate the search input again and adjust your query.
Issue origami icon

Learn how you can use GitHub Issues to plan and track your work.

Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub Issues
ProTip! 
Press the
/
key to activate the search input again and adjust your query.