Add XGlue by patrickvonplaten · Pull Request #802 · huggingface/datasets

patrickvonplaten · 2020-11-04T17:29:54Z

Dataset is ready to merge. An important feature of this dataset is that for each config the train data is in English, while dev and test data are in multiple languages. Therefore, @lhoestq and I decided offline that we will give the dataset the following API, e.g. for

load_dataset("xglue", "ner")  # would give the splits 'train', 'validation.en', 'test.en', 'validation.es', 'test.es', ...

=> therefore one can load a single language test via

load_dataset("xglue", "ner", split="test.es")

Close #749.

lhoestq · 2020-11-18T10:35:37Z

Really cool to add XGlue, this will be a nice addition !

Splits shouldn't depend on the language. There must be configurations for each language, as we're doing for xnli, xtreme, etc.
For example for XGlue we'll have these configurations: NER.de, NER.en etc.

patrickvonplaten · 2020-12-01T11:20:30Z

@@ -0,0 +1,328 @@
+---


This README.md is not fully complete. But in a good state now. @yjernite maybe the authors feel like filling out the rest.

lhoestq

Awesome good job !

I left a few comments about feature naming and types.
Also It looks like the dummy data files are quite heavy (450KB each). It should be possible to remove some data files inside the zip files to make them lighter (for example remove the mlqa files from the dummy_data.zip for for ner)

lhoestq · 2020-12-01T13:15:16Z

+            features = {
+                "news_title": datasets.Value("string"),
+                "news_body": datasets.Value("string"),
+                "news_category": datasets.Value("string"),


is is possible to use ClassLabel here ?

Awesome good job !

I left a few comments about feature naming and types.
Also It looks like the dummy data files are quite heavy (450KB each). It should be possible to remove some data files inside the zip files to make them lighter (for example remove the mlqa files from the dummy_data.zip for for ner)

Tried to reduce as much as possible. It totals now around ~2MB. Think that's fine given that we have 11 configs a 19 splits

Thank you so much ! This makes cloning the library faster

lhoestq

Awesome thanks ! Impressive job

Feel free to merge if you have no other changes to make and CI is green

lhoestq · 2020-12-01T15:46:04Z

+      "\u0909\u0924\u094d\u0924\u0930 \u092a\u0942\u0930\u094d\u0935"
+    ]
+  },
+  "context": "\u0909\u0938\u0940 \"\u090f\u0930\u093f\u092f\u093e XX \" \u0928\u093e\u092e\u0915\u0930\u0923 \u092a\u094d\u0930\u0923\u093e\u0932\u0940 \u0915\u093e \u092a\u094d\u0930\u092f\u094b\u0917 \u0928\u0947\u0935\u093e\u0926\u093e \u092a\u0930\u0940\u0915\u094d\u0937\u0923 \u0938\u094d\u0925\u0932 \u0915\u0947 \u0905\u0928\u094d\u092f \u092d\u093e\u0917\u094b\u0902 \u0915\u0947 \u0932\u093f\u090f \u0915\u093f\u092f\u093e \u0917\u092f\u093e \u0939\u0948\u0964\u092e\u0942\u0932 \u0930\u0942\u092a \u092e\u0947\u0902 6 \u092c\u091f\u0947 10 \u092e\u0940\u0932 \u0915\u093e \u092f\u0939 \u0906\u092f\u0924\u093e\u0915\u093e\u0930 \u0905\u0921\u094d\u0921\u093e \u0905\u092c \u0924\u0925\u093e\u0915\u0925\u093f\u0924 '\u0917\u094d\u0930\u0942\u092e \u092c\u0949\u0915\u094d\u0938 \" \u0915\u093e \u090f\u0915 \u092d\u093e\u0917 \u0939\u0948, \u091c\u094b \u0915\u093f 23 \u092c\u091f\u0947 25.3 \u092e\u0940\u0932 \u0915\u093e \u090f\u0915 \u092a\u094d\u0930\u0924\u093f\u092c\u0902\u0927\u093f\u0924 \u0939\u0935\u093e\u0908 \u0915\u094d\u0937\u0947\u0924\u094d\u0930 \u0939\u0948\u0964 \u092f\u0939 \u0915\u094d\u0937\u0947\u0924\u094d\u0930 NTS \u0915\u0947 \u0906\u0902\u0924\u0930\u093f\u0915 \u0938\u0921\u093c\u0915 \u092a\u094d\u0930\u092c\u0902\u0927\u0928 \u0938\u0947 \u091c\u0941\u0921\u093c\u093e \u0939\u0948, \u091c\u093f\u0938\u0915\u0940 \u092a\u0915\u094d\u0915\u0940 \u0938\u0921\u093c\u0915\u0947\u0902 \u0926\u0915\u094d\u0937\u093f\u0923 \u092e\u0947\u0902 \u092e\u0930\u0915\u0930\u0940 \u0915\u0940 \u0913\u0930 \u0914\u0930 \u092a\u0936\u094d\u091a\u093f\u092e \u092e\u0947\u0902 \u092f\u0941\u0915\u094d\u0915\u093e \u092b\u094d\u0932\u0948\u091f \u0915\u0940 \u0913\u0930 \u091c\u093e\u0924\u0940 \u0939\u0948\u0902\u0964 \u091d\u0940\u0932 \u0938\u0947 \u0909\u0924\u094d\u0924\u0930 \u092a\u0942\u0930\u094d\u0935 \u0915\u0940 \u0913\u0930 \u092c\u0922\u093c\u0924\u0947 \u0939\u0941\u090f \u0935\u094d\u092f\u093e\u092a\u0915 \u0914\u0930 \u0914\u0930 \u0938\u0941\u0935\u094d\u092f\u0935\u0938\u094d\u0925\u093f\u0924 \u0917\u094d\u0930\u0942\u092e \u091d\u0940\u0932 \u0915\u0940 \u0938\u0921\u093c\u0915\u0947\u0902 \u090f\u0915 \u0926\u0930\u094d\u0930\u0947 \u0915\u0947 \u091c\u0930\u093f\u092f\u0947 \u092a\u0947\u091a\u0940\u0926\u093e \u092a\u0939\u093e\u0921\u093c\u093f\u092f\u094b\u0902 \u0938\u0947 \u0939\u094b\u0915\u0930 \u0917\u0941\u091c\u0930\u0924\u0940 \u0939\u0948\u0902\u0964 \u092a\u0939\u0932\u0947 \u0938\u0921\u093c\u0915\u0947\u0902 \u0917\u094d\u0930\u0942\u092e \u0918\u093e\u091f\u0940",


Impressive
not sure if it's possible to show the actual characters

I'm sure there is a way -> maybe can use this PR to introduce "First Good Issue" in datasets

* first version * delete unnecessary files * add hyphen to regex * update * add pos and ner * add mlqa * finish script * add dataset json info * typo * del log * add dummy data - very nice tool quentin!!! * use utf-8 encoding * make style * finish dataset * improve README.md * improve readme * Update datasets/xglue/README.md * correct tab * apply quentin suggestinos * make quentin happy Co-authored-by: Patrick von Platen <patrick@huggingface.co>

patrickvonplaten added 2 commits November 4, 2020 17:29

first version

ae31844

delete unnecessary files

5383d72

patrickvonplaten changed the title ~~first version~~ [WIP] Add XGlue Nov 4, 2020

patrickvonplaten mentioned this pull request Nov 4, 2020

[XGLUE] Adding new dataset #749

Closed

add hyphen to regex

94e678b

patrickvonplaten commented Nov 4, 2020

View reviewed changes

Comment thread src/datasets/arrow_reader.py Outdated

patrickvonplaten and others added 3 commits November 30, 2020 11:38

update

89287c2

add pos and ner

49ca00e

add mlqa

c14aed8

yjernite approved these changes Nov 30, 2020

View reviewed changes

Patrick von Platen and others added 8 commits November 30, 2020 23:20

finish script

6661ebd

add dataset json info

067afea

merge conflicts

0006d14

typo

497fd2e

del log

e4ec410

add dummy data - very nice tool quentin!!!

95ad43b

use utf-8 encoding

32608c4

make style

4b2ea02

patrickvonplaten commented Dec 1, 2020

View reviewed changes

Comment thread src/datasets/naming.py

finish dataset

8218372

patrickvonplaten changed the title ~~[WIP] Add XGlue~~ Add XGlue Dec 1, 2020

patrickvonplaten commented Dec 1, 2020

View reviewed changes

improve README.md

9f6feac

patrickvonplaten requested a review from lhoestq December 1, 2020 11:33

lhoestq reviewed Dec 1, 2020

View reviewed changes

improve readme

70191ef

patrickvonplaten commented Dec 1, 2020

View reviewed changes

Comment thread datasets/xglue/README.md Outdated

patrickvonplaten added 2 commits December 1, 2020 14:44

Update datasets/xglue/README.md

c52ca29

correct tab

eb719f3

patrickvonplaten added 2 commits December 1, 2020 13:50

fix tab

18192f6

apply quentin suggestinos

b3fb068

lhoestq approved these changes Dec 1, 2020

View reviewed changes

make quentin happy

5b8473a

patrickvonplaten merged commit 3248f79 into master Dec 1, 2020

patrickvonplaten deleted the add_xglue branch December 1, 2020 15:58

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Add XGlue#802

Add XGlue#802
patrickvonplaten merged 22 commits into
masterfrom
add_xglue

patrickvonplaten commented Nov 4, 2020 •

edited by albertvillanova

Loading

Uh oh!

Uh oh!

lhoestq commented Nov 18, 2020

Uh oh!

Uh oh!

patrickvonplaten Dec 1, 2020 •

edited

Loading

Uh oh!

lhoestq left a comment

Uh oh!

Uh oh!

Uh oh!

lhoestq Dec 1, 2020

Uh oh!

patrickvonplaten Dec 1, 2020 •

edited

Loading

Uh oh!

lhoestq Dec 1, 2020

Uh oh!

Uh oh!

lhoestq left a comment

Uh oh!

lhoestq Dec 1, 2020

Uh oh!

patrickvonplaten Dec 1, 2020 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

patrickvonplaten commented Nov 4, 2020 • edited by albertvillanova Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

lhoestq commented Nov 18, 2020

Uh oh!

Uh oh!

patrickvonplaten Dec 1, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lhoestq left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

lhoestq Dec 1, 2020

Choose a reason for hiding this comment

Uh oh!

patrickvonplaten Dec 1, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lhoestq Dec 1, 2020

Choose a reason for hiding this comment

Uh oh!

Uh oh!

lhoestq left a comment

Choose a reason for hiding this comment

Uh oh!

lhoestq Dec 1, 2020

Choose a reason for hiding this comment

Uh oh!

patrickvonplaten Dec 1, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

patrickvonplaten commented Nov 4, 2020 •

edited by albertvillanova

Loading

patrickvonplaten Dec 1, 2020 •

edited

Loading

patrickvonplaten Dec 1, 2020 •

edited

Loading

patrickvonplaten Dec 1, 2020 •

edited

Loading