You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Jul 3, 2025. It is now read-only.
在我的应用里,发现带BN+scale结构推理速度更快;但是,我发现scale.cpp不支持bias的参数预加载,即在网络初始化时,无法提前规划权重。我觉得可以采用caffe源码中的实现,将57-59行
bias_param_id_ = this->blobs_.size();
this->blobs_.resize(bias_param_id_ + 1);
this->blobs_[bias_param_id_] = bias_layer_->blobs()[0];
替换为:
if (this->blobs_.size() + bottom.size() < 3) {
// case: blobs.size == 1 && bottom.size == 1
// or blobs.size == 0 && bottom.size == 2
bias_param_id_ = this->blobs_.size();
this->blobs_.resize(bias_param_id_ + 1);
this->blobs_[bias_param_id_] = bias_layer_->blobs()[0];
}
else {
// bias param already initialized
bias_param_id_ = this->blobs_.size() - 1;
bias_layer_->blobs()[0] = this->blobs_[bias_param_id_];
}