lgrobol commited on
Commit
59c8b37
1 Parent(s): 22f63f7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +31 -0
README.md CHANGED
@@ -1,3 +1,34 @@
1
  ---
2
  license: mit
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
+ language:
4
+ - fro
5
+ pipeline_tag: fill-mask
6
+ library_name: transformers
7
  ---
8
+
9
+ BERTrade-base
10
+ =============
11
+
12
+ This contextual embeddings model for Old French was developped as part of the work of Grobol, Regnault, Ortiz-Suarez et al. ([2022](https://aclanthology.org/2022.lrec-1.119/)), we refer to it for a list of the resources and the settings used to develop this model as well as assessment of their suitability for syntactic dependency parsing of Old French.
13
+
14
+ If you use it, please cite the original paper:
15
+
16
+ ```bibtex
17
+ @inproceedings{grobol-etal-2022-bertrade,
18
+ title = "{BERT}rade: Using Contextual Embeddings to Parse {O}ld {F}rench",
19
+ author = {Grobol, Lo{\"\i}c and
20
+ Regnault, Mathilde and
21
+ Ortiz Suarez, Pedro and
22
+ Sagot, Beno{\^\i}t and
23
+ Romary, Laurent and
24
+ Crabb{\'e}, Benoit},
25
+ booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
26
+ month = jun,
27
+ year = "2022",
28
+ address = "Marseille, France",
29
+ publisher = "European Language Resources Association",
30
+ url = "https://aclanthology.org/2022.lrec-1.119",
31
+ pages = "1104--1113",
32
+ abstract = "The successes of contextual word embeddings learned by training large-scale language models, while remarkable, have mostly occurred for languages where significant amounts of raw texts are available and where annotated data in downstream tasks have a relatively regular spelling. Conversely, it is not yet completely clear if these models are also well suited for lesser-resourced and more irregular languages. We study the case of Old French, which is in the interesting position of having relatively limited amount of available raw text, but enough annotated resources to assess the relevance of contextual word embedding models for downstream NLP tasks. In particular, we use POS-tagging and dependency parsing to evaluate the quality of such models in a large array of configurations, including models trained from scratch from small amounts of raw text and models pre-trained on other languages but fine-tuned on Medieval French data.",
33
+ }
34
+ ```