Giter VIP home page Giter VIP logo

Comments (14)

VixeruntR avatar VixeruntR commented on May 20, 2024 1

@VixeruntR 您这边往后训练多个epoch后是正常的吗，我训练也是，第三轮开始recall就是0了，看着有点慌

我刚开始数据集大概2000张图像出现了这种情况，后来加了大概4000的icdar里面的数据，学习率设置的5e-4，batch_size=8，主干是resnet18，就没出现这个情况了，训练了20个epoch，recall 大概0.6，实际效果就已经非常好了。希望对你有帮助

from dbnet.pytorch.

WenmuZhou avatar WenmuZhou commented on May 20, 2024

100个epoch之后再看

from dbnet.pytorch.

VixeruntR avatar VixeruntR commented on May 20, 2024

100个epoch之后再看

好的~

from dbnet.pytorch.

VixeruntR avatar VixeruntR commented on May 20, 2024

100个epoch之后再看

方便的话，能否提供下您的预训练模型？

from dbnet.pytorch.

WenmuZhou avatar WenmuZhou commented on May 20, 2024

目前没有预训练

from dbnet.pytorch.

VixeruntR avatar VixeruntR commented on May 20, 2024

目前没有预训练

好的，仍然感谢您！

from dbnet.pytorch.

WenmuZhou avatar WenmuZhou commented on May 20, 2024

那我先关了这个了

from dbnet.pytorch.

kunyao2015 avatar kunyao2015 commented on May 20, 2024

@VixeruntR 您这边往后训练多个epoch后是正常的吗，我训练也是，第三轮开始recall就是0了，看着有点慌

from dbnet.pytorch.

kunyao2015 avatar kunyao2015 commented on May 20, 2024

@VixeruntR 您这边往后训练多个epoch后是正常的吗，我训练也是，第三轮开始recall就是0了，看着有点慌

我刚开始数据集大概2000张图像出现了这种情况，后来加了大概4000的icdar里面的数据，学习率设置的5e-4，batch_size=8，主干是resnet18，就没出现这个情况了，训练了20个epoch，recall 大概0.6，实际效果就已经非常好了。希望对你有帮助

感谢！我在resnet18下面训收敛比较快，换成resnet50就很慢了，100个epoch后也不见多好，效果还比上训练十几轮的resnet18

from dbnet.pytorch.

VixeruntR avatar VixeruntR commented on May 20, 2024

@VixeruntR 您这边往后训练多个epoch后是正常的吗，我训练也是，第三轮开始recall就是0了，看着有点慌

我刚开始数据集大概2000张图像出现了这种情况，后来加了大概4000的icdar里面的数据，学习率设置的5e-4，batch_size=8，主干是resnet18，就没出现这个情况了，训练了20个epoch，recall 大概0.6，实际效果就已经非常好了。希望对你有帮助

感谢！我在resnet18下面训收敛比较快，换成resnet50就很慢了，100个epoch后也不见多好，效果还比上训练十几轮的resnet18

我也是。。训练20轮的resnet18是最香的。。

from dbnet.pytorch.

xxllp avatar xxllp commented on May 20, 2024

训练几百张数据是不是容易过拟合啊

from dbnet.pytorch.

VixeruntR avatar VixeruntR commented on May 20, 2024

训练几百张数据是不是容易过拟合啊

几百张有点太少了。。

from dbnet.pytorch.

619268640 avatar 619268640 commented on May 20, 2024

@VixeruntR 你好，请问，我用了几百张图片训练，1200轮都跑完了，P，R，F仍然是0，是什么原因呢？几百张可能效果不好，不不至于全都是0啊

from dbnet.pytorch.

remeberWei avatar remeberWei commented on May 20, 2024

@VixeruntR 你好，请问，我用了几百张图片训练，1200轮都跑完了，P，R，F仍然是0，是什么原因呢？几百张可能效果不好，不不至于全都是0啊

有可能就是你的pyrorch版本跟cuda版本不匹配

from dbnet.pytorch.

Related Issues (20)

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.