We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
在ptranse.cpp 中,有一些地方实在看不明白,希望麻烦解答下:
dScope=rand(0.0,normal(miu,miu,sigma));
relationVec[r1* dimension + ii]+=belta*ptranseAlpha*x
relationVec[r1 * dimension + ii]-=belta*ptranseAlpha*x;
relationVec[j * dimension + i]-=belta*ptranseAlpha*x
relationVec[j * dimension + ii]+=belta*ptranseAlpha*x;
if (randd(id) % 1000 < pr)
pr_path = 0.99*pr_path + 0.01;
train_path(trainList[i]->r, j, relPath, 2*margin, pr*pr_path);
2*margin
The text was updated successfully, but these errors were encountered:
No branches or pull requests
在ptranse.cpp 中,有一些地方实在看不明白,希望麻烦解答下:
dScope=rand(0.0,normal(miu,miu,sigma));
的作用是什么?整个randn()函数实现的是什么功能?relationVec[r1* dimension + ii]+=belta*ptranseAlpha*x
;我在算梯度时感觉应该是relationVec[r1 * dimension + ii]-=belta*ptranseAlpha*x;
同样,relationVec[j * dimension + i]-=belta*ptranseAlpha*x
;我觉得应该是relationVec[j * dimension + ii]+=belta*ptranseAlpha*x;
。不知道这个是不是我计算有问题,麻烦查看下,谢谢。if (randd(id) % 1000 < pr)
表示什么意思?pr_path = 0.99*pr_path + 0.01;
为什么要做这个操作?train_path(trainList[i]->r, j, relPath, 2*margin, pr*pr_path);
中为什么要用2*margin
而不是margin?The text was updated successfully, but these errors were encountered: