piotr-rybak
commited on
Commit
•
6953ff8
1
Parent(s):
4816334
Update README.md
Browse files
README.md
CHANGED
@@ -34,7 +34,7 @@ Tokenizer utilizes `XLMTokenizer` implementation for that reason, one should loa
|
|
34 |
HerBERT was evaluated on the [KLEJ](https://klejbenchmark.com/) benchmark, publicly available set of nine evaluation tasks for the Polish language understanding.
|
35 |
It had the best average performance and obtained the best results for three of them.
|
36 |
|
37 |
-
| Model | Average | NKJP-NER | CDSC-E | CDSC-R | CBD | PolEmo2.0-IN
|
38 |
| :------ | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: |
|
39 |
| herbert-klej-cased-v1 | **80.5** | 92.7 | 92.5 | 91.9 | **50.3** | **89.2** |**76.3** |52.1 |95.3 | 84.5 |
|
40 |
|
@@ -68,18 +68,24 @@ CC BY-SA 4.0
|
|
68 |
## Citation
|
69 |
If you use this model, please cite the following paper:
|
70 |
```
|
71 |
-
@
|
72 |
-
title={KLEJ: Comprehensive Benchmark for
|
73 |
-
author=
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
78 |
}
|
79 |
```
|
80 |
-
Paper is accepted at ACL 2020, as soon as proceedings appear, we will update the BibTeX.
|
81 |
|
82 |
## Authors
|
83 |
-
|
84 |
|
85 |
You can contact us at: <a href="mailto:klejbenchmark@allegro.pl">klejbenchmark@allegro.pl</a>
|
|
|
34 |
HerBERT was evaluated on the [KLEJ](https://klejbenchmark.com/) benchmark, publicly available set of nine evaluation tasks for the Polish language understanding.
|
35 |
It had the best average performance and obtained the best results for three of them.
|
36 |
|
37 |
+
| Model | Average | NKJP-NER | CDSC-E | CDSC-R | CBD | PolEmo2.0-IN\t|PolEmo2.0-OUT | DYK | PSC | AR\t|
|
38 |
| :------ | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: | ------: |
|
39 |
| herbert-klej-cased-v1 | **80.5** | 92.7 | 92.5 | 91.9 | **50.3** | **89.2** |**76.3** |52.1 |95.3 | 84.5 |
|
40 |
|
|
|
68 |
## Citation
|
69 |
If you use this model, please cite the following paper:
|
70 |
```
|
71 |
+
@inproceedings{rybak-etal-2020-klej,
|
72 |
+
title = "{KLEJ}: Comprehensive Benchmark for {P}olish Language Understanding",
|
73 |
+
author = "Rybak, Piotr and
|
74 |
+
Mroczkowski, Robert and
|
75 |
+
Tracz, Janusz and
|
76 |
+
Gawlik, Ireneusz",
|
77 |
+
booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
|
78 |
+
month = jul,
|
79 |
+
year = "2020",
|
80 |
+
address = "Online",
|
81 |
+
publisher = "Association for Computational Linguistics",
|
82 |
+
url = "https://www.aclweb.org/anthology/2020.acl-main.111",
|
83 |
+
doi = "10.18653/v1/2020.acl-main.111",
|
84 |
+
pages = "1191--1201",
|
85 |
}
|
86 |
```
|
|
|
87 |
|
88 |
## Authors
|
89 |
+
The model was trained by **Allegro Machine Learning Research** team.
|
90 |
|
91 |
You can contact us at: <a href="mailto:klejbenchmark@allegro.pl">klejbenchmark@allegro.pl</a>
|