Open zhl95 opened 6 years ago
@zhl95 hello. I came across the same problem. I think need a little change to the codes. Did u solve this problem?
use code as adioshun suggested. Then delete the redundant arguments for Batchloading and trainer.
thanks a lot. It works now.
@Anida-qin
@zhl95
hi,Have you ever encountered mistakes like
after execute the very beginning steps which before python data.py&python train.py: source activate tensorflow sudo chmod 755 ./make.sh ./make.sh or i shouldnt have these steps?
No. Did you try pip install the missing module?
have you ever use the pretrained weights in checkpoint mudule(which can be download in baiduyun)? i have problem with restoring the model.... always shows the error "Converting sparse IndexedSlices to a dense Tensor of unknown shape. " @zhl95 @Anida-qin do you know how to solve it ? thanks a lot...
can you tell me which version of tensorflow are you using ? @zhl95 @Anida-qin
When I do "python train.py". It gives "NameError: name 'data_splitter' is not defined". But data_splitter is only defined when data is didi. How to run train.py with kitti data?