Peng Chen

I did'nt meet the zero result, for example my test on mobilnet_v2_yolo_v2_voc gives a mAP of 0.721956. Check my result at https://github.com/ShuangXieIrene/ssds.pytorch/issues/41 I use pytorch 0.4.1

@jinfagang Have you solved the problem? I have the same issue. @1453042287 I trained the yolov2-mobilenet-v2 from stratch. U mentioned 'pre-trained model', do y mean the pre-trained bone network model...

@1453042287 Hi, thanks for the advise. My current training seems working. In my previous training, I set 'base' and 'loc' so on all in the trainable_scope, and it does not...

Yes, set all parameter to re-trainable seems hard to converge. This year, Mr He did publish a paper named 'Rethinking ImageNet Pre-training' which claimed the pre-train on imagenet is not...

Hi, @1453042287 @cvtower I have another issue about the train precision and loss curve. The following is the result from tensorboardX.  It can be see that the precision slowly...

@funnyuday I found a similar problem. The NMS is very very slow on my machine. Could you share what you have revised to improve the inference detection speed?

I made the following revise: 1. epoch_size = len(data_loader) instead of len(dataset) // args.batch_size 2. add the try clause when load train data: try: images, targets = next(batch_iterator) except StopIteration:...

would be loss_x.data[0] >> loss_x.item() better?

@sh1r0 Hi, I met a similar problem when build the glag glog libs when the ANDROID_ABI is set to 'arm64-v8a'. If I set it as 'armeabi-v7a with NEON', everything is...

cmake --version: 3.6.0-rc4. Should this be exactly 3.3.2 or 3.5.2 as you mentioned before? ndk version: r10d