double4tar

@luoyetx I got this strange problem when call `Net::CopyTrainedLayersFrom` under different mode(caffe::GPU vs caffe CPU): Error info is: ### C:\workspace\opensource\mini-caffe\src\net.cpp:277: Cannot copy param 0 weights from layer '221'; shape mismatch....

Recently I did some code reading and debug. The result shows that: 1、Net::Reshape() was called during Net::Forward, so the shape of my Convolution changed to mismatching status. 2、I put some...

I got the relative issue. I'd like to train with my own dataset with inceptionv3. Did you download the model here?  Is this model for trainning or for inferrence?

I also downloaded the model, and my scripts looks like this, and it worked. `python demo.py --gpu 0 --network inceptionv3 --data-shape 512 --epoch 215` maybe you miss `--data-shape 512` and...

Is the forward speed relative to the trainning convergence？ I used mxnet `PredictorHandle` (https://mxnet.incubator.apache.org/doxygen/c__predict__api_8h.html) to do the test. And calculated the forward time, And I got the experiments result below:...

Thank you. It helped. I checked the output num of multiboxdetection layer. And found the bounding box number range from 160(good model) to ~3500(bad model). I set the `threshold` paramter...

Test with GPU， I changed the parameter in `xx-symbol.json` by adding "threshold": "0.1", ``` { "op": "_contrib_MultiBoxDetection", "name": "detection", "attrs": { "force_suppress": "True", "nms_threshold": "0.5", "threshold": "0.1", "nms_topk": "400", "variances":...

Dear biggies, sorry to bother, I am solving a problem that might similar to yours. So would you please help? I currenty have `NormalizedLandmark_A`(which could be drawn on image once...

Excuse me, I found that both y and xc are subtracted by 120, but the style imge is not. Is there anything wrong when calculate L_style with feature_hat(subtracted by 120)...

@cryptexis Thank you. Yes，you're right. My bad, I didn't read code carefully.