Commit
•
6518761
1
Parent(s):
795f245
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
datasets:
|
3 |
+
- tner/bc5cdr
|
4 |
+
- tner/bionlp2004
|
5 |
+
- tner/btc
|
6 |
+
- tner/conll2003
|
7 |
+
- tner/fin
|
8 |
+
- tner/mit_movie_trivia
|
9 |
+
- tner/mit_restaurant
|
10 |
+
- tner/multinerd
|
11 |
+
- tner/ontonotes5
|
12 |
+
- tner/tweebank_ner
|
13 |
+
- tner/tweetner7
|
14 |
+
- tner/wikineural
|
15 |
+
- tner/wnut2017
|
16 |
+
language:
|
17 |
+
- en
|
18 |
+
metrics:
|
19 |
+
- accuracy
|
20 |
+
- f1
|
21 |
+
---
|
22 |
+
# RoBERTa Span Detection
|
23 |
+
This model is a fine-tuned model of [roberta-large](https://huggingface.co/roberta-large) after being trained on a **mixture of NER datasets**.
|
24 |
+
|
25 |
+
Basically, this model has learn can detect NER spans (with <u>no differenciation on classes</u>). Labels use the IBO format and are:
|
26 |
+
- 'B-TAG': beginning token of span
|
27 |
+
- 'I-TAG': inside token of span
|
28 |
+
- 'O': token not a span
|
29 |
+
|
30 |
+
# Evaluation results
|
31 |
+
|
32 |
+
| Data | Accuracy |
|
33 |
+
|:---:|:---------:|
|
34 |
+
| validation | 0.964 |
|