[GitHub] tianzening commented on issue #4221: construct symbol for converting ZF net

[gi...@git.apache.org]

tianzening commented on issue #4221: construct symbol for converting ZF net
URL: https://github.com/apache/incubator-mxnet/pull/4221#issuecomment-318811030
 
 
   @liangfu Thanks for your work.  when I use your convert_zf.py,https://github.com/apache/incubator-mxnet/pull/4221/commits/fcd1249e8d5aec0ac1c7dda704cd07c0b0b3eab7
    I face some problems. Can you guide me how to modify the test.prototxt. ?I want to use the py-faster-rcnn/models/pascall_voc/ZF/faster_rcnn_end2end/test.prototxt, and pretraind ZF.v2.caffemodel to generate ZF-0000.params ?.
   
   Error:  Message type "caffe.LayerParameter" has no field named "roi_pooling_param"
   
   #========= RCNN ============
   # layer {
   #   name: "roi_pool_conv5"
   #   type: "ROIPooling"
   #   bottom: "conv5"
   #   bottom: "rois"
   #   top: "roi_pool_conv5"
   #   # roi_pooling_param {
   #   #   pooled_w: 6
   #   #   pooled_h: 6
   #   #   spatial_scale: 0.0625 # 1/16
   #   # }
   # }
   Then I comment the above part?Now  there is another error?
   
   Cannot copy param 0 weights from layer 'fc6'; shape mismatch.  Source param shape is 4096 9216 (37748736); target param shape is 4096 57600 (235929600). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.
   
   Thanks
   
   
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services