Open stevenyslins opened 6 years ago
Dear all,
I want to convert inception-resnet-v2 model from caffe to caffe2, I use this project to download deploy_inception-resnet-v2.prototxt and inception-resnet-v2.caffemodel, then launch this command:
python -m caffe2.python.caffe_translator deploy_inception-resnet-v2.prototxt inception-resnet-v2.caffemodel
but it will get fail, it seems like not supported, so has the problem been solved ? Thank you for your help.
root@b8626bef4308:/opt/caffe2# python -m caffe2.python.caffe_translator deploy_inception-resnet-v2.prototxt inception-resnet-v2.caffemodel INFO:caffe_translator:Translate layer conv1_3x3_s2 INFO:caffe_translator:Translate layer conv1_3x3_s2_bn INFO:caffe_translator:Translate layer conv1_3x3_s2_scale INFO:caffe_translator:Translate layer conv1_3x3_relu INFO:caffe_translator:Translate layer conv2_3x3_s1 INFO:caffe_translator:Translate layer conv2_3x3_s1_bn INFO:caffe_translator:Translate layer conv2_3x3_s1_scale INFO:caffe_translator:Translate layer conv2_3x3_relu INFO:caffe_translator:Translate layer conv3_3x3_s1 INFO:caffe_translator:Translate layer conv3_3x3_s1_bn INFO:caffe_translator:Translate layer conv3_3x3_s1_scale INFO:caffe_translator:Translate layer conv3_3x3_relu INFO:caffe_translator:Translate layer pool1_3x3_s2 INFO:caffe_translator:Translate layer conv4_3x3_reduce INFO:caffe_translator:Translate layer conv4_3x3_reduce_bn INFO:caffe_translator:Translate layer conv4_3x3_reduce_scale INFO:caffe_translator:Translate layer conv4_3x3_reduce_relu INFO:caffe_translator:Translate layer conv4_3x3 INFO:caffe_translator:Translate layer conv4_3x3_bn INFO:caffe_translator:Translate layer conv4_3x3_scale INFO:caffe_translator:Translate layer conv4_relu_3x3 INFO:caffe_translator:Translate layer pool2_3x3_s2 INFO:caffe_translator:Translate layer conv5_1x1 INFO:caffe_translator:Translate layer conv5_1x1_bn INFO:caffe_translator:Translate layer conv5_1x1_scale INFO:caffe_translator:Translate layer conv5_1x1_relu INFO:caffe_translator:Translate layer conv5_5x5_reduce INFO:caffe_translator:Translate layer conv5_5x5_reduce_bn INFO:caffe_translator:Translate layer conv5_5x5_reduce_scale INFO:caffe_translator:Translate layer conv5_5x5_reduce_relu INFO:caffe_translator:Translate layer conv5_5x5 INFO:caffe_translator:Translate layer conv5_5x5_bn INFO:caffe_translator:Translate layer conv5_5x5_scale INFO:caffe_translator:Translate layer conv5_5x5_relu INFO:caffe_translator:Translate layer conv5_3x3_reduce INFO:caffe_translator:Translate layer conv5_3x3_reduce_bn INFO:caffe_translator:Translate layer conv5_3x3_reduce_scale INFO:caffe_translator:Translate layer conv5_3x3_reduce_relu INFO:caffe_translator:Translate layer conv5_3x3 INFO:caffe_translator:Translate layer conv5_3x3_bn INFO:caffe_translator:Translate layer conv5_3x3_scale INFO:caffe_translator:Translate layer conv5_3x3_relu INFO:caffe_translator:Translate layer conv5_3x3_2 INFO:caffe_translator:Translate layer conv5_3x3_2_bn INFO:caffe_translator:Translate layer conv5_3x3_2_scale INFO:caffe_translator:Translate layer conv5_3x3_2_relu INFO:caffe_translator:Translate layer ave_pool INFO:caffe_translator:Translate layer conv5_1x1_ave INFO:caffe_translator:Translate layer conv5_1x1_ave_bn INFO:caffe_translator:Translate layer conv5_1x1_ave_scale INFO:caffe_translator:Translate layer conv5_1x1_ave_relu INFO:caffe_translator:Translate layer stem_concat INFO:caffe_translator:Translate layer inception_resnet_v2_a1_1x1 INFO:caffe_translator:Translate layer inception_resnet_v2_a1_1x1_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_1x1_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_1x1_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_reduce INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_reduce_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_reduce_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_reduce_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3 INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_reduce INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_reduce_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_reduce_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_reduce_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2 INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_2_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_3 INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_3_bn INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_3_scale INFO:caffe_translator:Translate layer inception_resnet_v2_a1_3x3_3_relu INFO:caffe_translator:Translate layer inception_resnet_v2_a1_concat INFO:caffe_translator:Translate layer inception_resnet_v2_a1_up INFO:caffe_translator:Translate layer inception_resnet_v2_a1_residual_eltwise Traceback (most recent call last): File "/usr/lib/python2.7/runpy.py", line 174, in _run_module_as_main "__main__", fname, loader, pkg_name) File "/usr/lib/python2.7/runpy.py", line 72, in _run_code exec code in run_globals File "/opt/caffe2/caffe2/python/caffe_translator.py", line 928, in <module> input_dims=args.input_dims File "/opt/caffe2/caffe2/python/caffe_translator.py", line 299, in TranslateModel return TranslatorRegistry.TranslateModel(*args, **kwargs) File "/opt/caffe2/caffe2/python/caffe_translator.py", line 288, in TranslateModel net_params=net_params, input_dims=input_dims) File "/opt/caffe2/caffe2/python/caffe_translator.py", line 228, in TranslateLayer layer, pretrained_blobs, is_test, **kwargs) File "/opt/caffe2/caffe2/python/caffe_translator.py", line 753, in TranslateElementWise raise RuntimeError("This eltwise layer is not yet supported.") RuntimeError: This eltwise layer is not yet supported.
Meet the same problem when convert Flownet2.0. Maybe because the eltwise layer is not that popular, hope it come soon.
Dear all,
I want to convert inception-resnet-v2 model from caffe to caffe2, I use this project to download deploy_inception-resnet-v2.prototxt and inception-resnet-v2.caffemodel, then launch this command:
but it will get fail, it seems like not supported, so has the problem been solved ? Thank you for your help.