I am trying to load a network configuration at the TEST stage, which first has a memory data layer and then a convolution level. Creating a MemoryData layer succeeds, but creating a convolution level fails at the following location:
LOG(INFO) << "Creating layer " << param.name(); const string& type = param.type(); CreatorRegistry& registry = Registry(); CHECK_EQ(registry.count(type), 1) << "Unknown layer type: " << type << " (known types: " << LayerTypeList() << ")";
Print error:
F0519 14: 54: 12.494139 14504 layer_factory.hpp: 77] Verification failed: registry.count (t ype) == 1 (0 vs 1) Unknown layer type: Convolution (known types: MemoryData)
In the registry
there is only one entry, even with MemoryData. When you enter the registry creation functions, it looks first (and finally, since it is a singleton) called from
REGISTER_LAYER_CLASS(MemoryData);
in memory_data_later.cpp.
I see similar REGISTER_LAYER_CLASS calls for other supported levels, but it looks like they are never called. How can i solve this?
Thanks!
c ++ deep-learning caffe layer
rkellerm
source share