Giter VIP home page Giter VIP logo

Comments (6)

lzl2040 avatar lzl2040 commented on July 26, 2024

我用2个GPU训练的时候发现score loss的收敛似乎很困难,比较慢,这个正常吗?用的是源代码,没有修改

from qdmn.

yongliu20 avatar yongliu20 commented on July 26, 2024

不需要调整,我在预训练没有使用衰减,不过你也可以试一下加上衰减。score loss收敛应该是比较快的,有log吗

from qdmn.

lzl2040 avatar lzl2040 commented on July 26, 2024

不需要调整,我在预训练没有使用衰减,不过你也可以试一下加上衰减。score loss收敛应该是比较快的,有log吗

1
我迭代了2000多次,前面一排是每个batch中score的值,都很小,所以我感觉它几乎没收敛

from qdmn.

lzl2040 avatar lzl2040 commented on July 26, 2024

我没有用tps,这应该不会影响收敛的吧

from qdmn.

yongliu20 avatar yongliu20 commented on July 26, 2024

我刚才跑了一下代码,两张卡的话score loss在500--600ite之间会迅速收敛,500之前收敛的会比较慢。并且score的预测会从1e-7级别慢慢上升到1e-4然后迅速到达正常输出如0.7、0.8。这是由于sigmoid函数的梯度在输出值非常接近于0时很小,逐渐上涨之后梯度变大了收敛就迅速了。2000ite还未收敛不是正常现象,或许与您模型的初始值有关系,或许您改变了部分代码。我刚才跑的是原始代码,您也可以实验一下不加修改的原始代码。目前的代码是没有问题的。

from qdmn.

lzl2040 avatar lzl2040 commented on July 26, 2024

我刚才跑了一下代码,两张卡的话score loss在500--600ite之间会迅速收敛,500之前收敛的会比较慢。并且score的预测会从1e-7级别慢慢上升到1e-4然后迅速到达正常输出如0.7、0.8。这是由于sigmoid函数的梯度在输出值非常接近于0时很小,逐渐上涨之后梯度变大了收敛就迅速了。2000ite还未收敛不是正常现象,或许与您模型的初始值有关系,或许您改变了部分代码。我刚才跑的是原始代码,您也可以实验一下不加修改的原始代码。目前的代码是没有问题的。

ok,谢谢您的指导

from qdmn.

Related Issues (19)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.