nlpaueb / multi-eurlex Goto Github PK
View Code? Open in Web Editor NEWMultiEURLEX - A multi-lingual and multi-label legal document classification dataset for zero-shot cross-lingual transfer
MultiEURLEX - A multi-lingual and multi-label legal document classification dataset for zero-shot cross-lingual transfer
Hi, when I get the dataset by load_dataset('multi_eurlex', language='en', label_level='level_1')
The label in these data samples are all small numbers from 0 to around 20, which I assume that they are all re-labelled by sequence.
But in the eurovoc_concepts file, the label is still 100XXX for level_1. Could you please specify the mapping relationship between them? (Same question for level 2 and level 3 as well)
Or is there any other way to interpret the labels?
Thank you very much.
Hi,
I got a key error when I try to have the original label of the data.
The code I used is load_dataset('multi_eurlex', language='en', label_level='all')
The error message is KeyError: 'all'
Could you please check about that?
Thank you very much.
Hey,
Thanks for releasing this code. I believe that there is a small inaccuracy in your code.
In the file experiments/trainer.py
, line number: 111, the following line of code
dataset=eval_dataset['validation'][:eval_samples if eval_samples else len(eval_dataset)]
must be replaced with
dataset=eval_dataset['validation'][:eval_samples if eval_samples else len(eval_dataset['validation'])]
I believe that len(eval_dataset)
would be 3 as it indicates the number of splits associated with the dataset. Please let me know whether this rectification is valid.
Hello, Thanks for sharing the code. I wanted to know the argument values for reproducing the results in the paper "LEGAL-BERT: The Muppets straight out of Law School". I tried executing the code in its default setting (eval_langs
was set to ['en']
). This resulted in a very poor performance. I am attaching the log file with this message (
MULTI-EURLEX_2022_01_19_05_01_19.txt). Thanks.
Hello,
Thank you for your research and your code which are very interesting. I'm trying to reproduce your results on your paper. I called python3 trainer.py --bert_path 'camembert-base' --native_bert True --batch_size 1 --train_lang 'fr' --eval_langs 'fr' --label_level 'level_3'
, but got NotImplementedError: Model type camembert is not supported for adaptation
.
Thank you in advance for any reply.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.