Open ghryou opened 6 years ago
i have met the same problem, exactly like yours. And I have no idea where is this nvcaffe. I only installed digits cuda and cudnn. is it possible nvcaffe was compiled with any of them? waiting for answers.
This problem seems to be related to a NVIDIA/caffe version 0.15. It looks like NVIDIA fixed it in the later releases.
While loading the pretrained Yolo v2 model (based on @gklz1982's work) which has BatchNormLayer on it, I got that error message.
So, I thought there might be some difference between Nvidia Caffe and BVLC Caffe, especially for Batch Norm layer.
On BVLC Caffe, Batch Normalization only use blobs with size 3,
However, for Nvidia Caffe, it requires blobs (or this->blobs_) size of 5.
I thought I can block Nvidia Caffe to access on this->blobs variable's 4th and 5th elements by setting scale_bias to false. Also, on NVCaffe Document, it is written that there are scale_bias Setting under batch_norm_param.
However, I got error message that there is no field named "scale_bias" under "caffe.BatchNormParameter".
This is my prototxt for classification with darknet 19 (which is the base network for yolo v2).
Is there anyone who resolve this issue???