-
Notifications
You must be signed in to change notification settings - Fork 125
New issue
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
caffe 训练速度仍然很慢,请问作者后来有发现这个问题吗, #26
Comments
你用的什么网络训练的,使用的什么数据,数据规模多大? |
除了 lost adaface 不一样 ,其他都一样的情况下,收敛速度loss好慢啊, |
说实话,这个损失函数的效果并没有论文说的那么好,感觉还是原始的arcface的比较work。 |
谢大神 |
|
我自己训练的时候,acc=0.98左右, loss = 5.8左右,感觉是有点不是很正常,loss太高了。 |
loss 开始高 没有关系,跟数据有关系,主要是降的太慢,又回到loss=15.1,晚上跑一晚上看看,能不能降到1以内 |
|
可以试一下调整动量momentum。 |
大神 昨晚把动量从0.92调到0.9 m 0.35->0.4 acc=0.99 不知道收敛到0.996+ |
@wavelet2008 你们说的acc是fc6层作为输入的acc么?请问训练过程中AdaCosAddmScale的acc大概能到多少 |
抱歉,caffe 训练慢的问题没有解决,还是那么慢,也没有花时间研究,但是很多时间都消耗在网络的运行,我试着减少打印损失值的输出,节省时间
…------------------ 原始邮件 ------------------
发件人: "xialuxi"<notifications@github.com>;
发送时间: 2019年6月17日(星期一) 下午5:48
收件人: "xialuxi/arcface-caffe"<arcface-caffe@noreply.github.com>;
抄送: "史诗"<892781037@qq.com>;"Author"<author@noreply.github.com>;
主题: Re: [xialuxi/arcface-caffe] caffe 训练速度仍然很慢,请问作者后来有发现这个问题吗, (#26)
说实话,这个损失函数的效果并没有论文说的那么好,感觉还是原始的arcface的比较work。
建议使用arcface
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
|
同问~ |
训练速度多少? 每秒86个samples? |
@xialuxi ,谢谢您的回复,我上午看了adaface的论文的实现
用caffe 训练很慢的原因请问作者后来有发现吗,前向传播我觉得挺快的
我设置的参数是batch size=56 l两块1080 iter_size:6
/57597324-8428ad00-7581-11e9-9fa1-e2d72d0446f7.png)
12:49:24.228211 23509 solver.cpp:243] Iteration 0, loss = 24.1503
I0513 12:49:24.228235 23509 solver.cpp:259] Train net output #0: accuracy-t = 0.839286
I0513 12:49:24.228257 23509 solver.cpp:259] Train net output #1: softmax_loss = 22.5511 (* 1 = 22.5511 loss)
I0513 12:49:24.228299 23509 sgd_solver.cpp:138] Iteration 0, lr = 0.01
I0513 12:54:50.852994 23509 solver.cpp:243] Iteration 100, loss = 20.3324
I0513 12:54:50.853057 23509 solver.cpp:259] Train net output #0: accuracy-t = 0.928571
I0513 12:54:50.853081 23509 solver.cpp:259] Train net output #1: softmax_loss = 17.7896 (* 1 = 17.7896 loss)
I0513 12:54:50.923504 23509 sgd_solver.cpp:138] Iteration 100, lr = 0.01
I0513 13:00:39.458894 23509 solver.cpp:243] Iteration 200, loss = 18.8438
I0513 13:00:39.459019 23509 solver.cpp:259] Train net output #0: accuracy-t = 0.964286
I0513 13:00:39.459044 23509 solver.cpp:259] Train net output #1: softmax_loss = 16.4004 (* 1 = 16.4004 loss)
I0513 13:00:39.500185 23509 sgd_solver.cpp:138] Iteration 200, lr = 0.01
I0513 13:06:26.364652 23509 solver.cpp:243] Iteration 300, loss = 18.461
I0513 13:06:26.364759 23509 solver.cpp:259] Train net output #0: accuracy-t = 0.946429
I0513 13:06:26.364783 23509 solver.cpp:259] Train net output #1: softmax_loss = 16.8125 (* 1 = 16.8125 loss)
The text was updated successfully, but these errors were encountered: