Comments (4)
写错了,是验证集F@3只有0.1
from bert-kpe.
Hi, 谢谢你的问题。请问你是不是把batch size或者accumulate steps改了,导致model没有很好的收敛。我们当时在training dataset上的Epoch Mean Loss=0.0471 (epoch=4), 然后在dev上测试F1@3=0.368, 你可以提供一下目前你的model在training dataset上的Epoch Mean Loss数值吗? 这样我们可以进一步确认问题,并帮助解决。
from bert-kpe.
我们在2块Tesla T4(2*16G)上跑的实验,因此直接将batch size调整为其最大容量。如果你使用不同容量的计算卡,我们建议可以稍微调整一下accumulate steps,让model可以充分收敛。如果还有问题欢迎提问,这个issue我们先close了。也可以邮件沟通: [email protected]
from bert-kpe.
我们在2块Tesla T4(2*16G)上跑的实验,因此直接将batch size调整为其最大容量。如果你使用不同容量的计算卡,我们建议可以稍微调整一下accumulate steps,让model可以充分收敛。如果还有问题欢迎提问,这个issue我们先close了。也可以邮件沟通: [email protected]
非常感谢你的回答,我看了一下我的mean loss确实没有降到你所说的那个值, 可能跟我使用的卡有关,我只用了一张24G的M40。我会再尝试一下,按照你的意见。
from bert-kpe.
Related Issues (17)
- 请问是否可以用于中文关键词提取? HOT 14
- What is the parameter `active_mask` mean?
- Unable to use checkpoints for inference HOT 4
- Installing the requirements.txt file does not work HOT 1
- Model loading failed HOT 2
- API for key phrase extraction inference HOT 1
- 变量 max_diff_gram_num 的含义 HOT 1
- When I ran the code, some errors occurred. HOT 3
- Script to generate keyphrases from arbitrary text? HOT 1
- Increasing max_phrase_words runs into memory issues HOT 1
- Does this repo support non-en language, such as Chinese, japan and korea?
- loss function is wrong
- Error while running test.sh HOT 4
- micro-f1 or macro-f1 ?
- Why include transformers? HOT 2
- Using kp20k dataset HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
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.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from bert-kpe.