Comments (6)
Hi @caoxu915683474 , which task do you want to train? NER? Generally set the iteration as 100 is enough, 15000 is way too much! If you trained your model on CoNLL 2003 English NER data, the F1=0.7 is too low, you need to check you input data.
The calculation of F1-score is based on precision and recall. Different implementations may have different name, but the equation should be same. You can refer https://en.wikipedia.org/wiki/F1_score for more information about F1-score.
from ncrfpp.
Thanks for your quick answer , I just use the data that in your work. And thanks for the help with F1. So what I only do is clone your project and set the iteration as 15000 and run it without any change.
from ncrfpp.
The provided data is a sample data. It does not contain the whole dataset in CoNLL03. You need to get the original full data and the embeddings to reproduce my result.
from ncrfpp.
I think get_ner_BIO() in metric.py is wrong.
consider the example where label_list = [I-MISC, I-MISC, O, I-PER, I-PER, O, O, O, O, O I-ORG, O] according to current function the following will happen :
Since there is no tag involving B-, whole_tag and tag_index will always be [] and hence the output of the function is [] which is wrong?
I am unable to find CoNLL-2003 in BIOMES format, is it available from official CoNLL website? If not can you provide reference to convert BIOMES to BIO?
from ncrfpp.
@udion The metric.py is not wrong, but you are using wrong data format. BE RESPECT TO OTHERS' WORK!
You should learn the difference of BIO/BIOES/IOB and then you can write a script to convert the data format between these tag scheme.
from ncrfpp.
@jiesutd yes, sorry foolish of me
the above mentioned example is not even valid BIOES or BIO format so obviously get_ner_BIO()
won't function the way it should be. forgive me for my noobness.
Thanks, I wrote a script to convert (what seems like IOB) to BIOES. I am a noob though and got to know about CoNLL few days back, I was wondering how you guys check if different versions (BIOES/BIO/IOB) data are correct, do you guys have an exhaustive set of tagged sequences on which you check your conversion scripts?
( P.S. please don't mind dumb questions )
from ncrfpp.
Related Issues (20)
- Conll 使用内存大 HOT 2
- nbest score HOT 1
- Provision for Custom Features? HOT 1
- Bug in forward method when calling _viterbi_decode, mask is not provided
- about requirement HOT 1
- Can I just use the CRF layer? HOT 1
- 大神你好,请教一些关于报错的问题 HOT 1
- About "tcmalloc: large alloc" message and training aborted HOT 3
- Difference between main_parse.py and main.py HOT 1
- Different results on CPU versus GPU HOT 1
- Test Data Format for Chunking
- Experimenting with Transformer models? HOT 1
- Decode config parameter when not using CRF HOT 1
- Reading fasttext word embeddings HOT 2
- About MAX_SENTENCE_LENGTH parameter HOT 2
- Decode error with nbest=0 when not using CRF
- Resume model training from a checkpoint
- 训练一段时间后,f1为-1 HOT 2
- Unknown labels in decoding
- Support trained model serialization to ONNX format 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 ncrfpp.