gitUser_DL

@yangxue0827 Ok, Thanks! By the way, could you add vgg16 support? Because the Res50 and Res101 have out-of-memory on my GPU device.

@yangxue0827 I tried the python3 run train1.py, I got error like this: Traceback (most recent call last): File "train1.py", line 21, in from libs.fast_rcnn import build_fast_rcnn1 File "/home/hardDisk/research/workspace/R2CNN_FPN_Tensorflow/tools/../libs/fast_rcnn/build_fast_rcnn1.py", line 14,...

I using the given data and coverted the airplane dataset into LMDB, then starting the training. However the training loss change from 7.0+ up to 300.0+! after about 70000 iterations....

Thanks, I will try your suggestions. 2018-06-21 21:49 GMT+08:00 chensonglu : > @gitUserGoodLeaner I encounter the > same problem and changing the "base_lr" from 0.001 to 0.0001 works for me....