ajitrajasekharan
commited on
Commit
•
a23c1dc
1
Parent(s):
c22c86d
Update README.md
Browse files
README.md
CHANGED
@@ -25,7 +25,7 @@ This model was pretrained from scratch using a custom vocabulary on the followin
|
|
25 |
- Clinical trials corpus
|
26 |
- and a small subset of Bookcorpus
|
27 |
|
28 |
-
|
29 |
|
30 |
[Github link](https://github.com/ajitrajasekharan/unsupervised_NER) to perform NER using this model in an ensemble with bert-base cased.
|
31 |
|
@@ -33,8 +33,12 @@ The ensemble detects 69 entity subtypes (17 broad entity groups)
|
|
33 |
|
34 |
<img src="https://ajitrajasekharan.github.io/images/1.png" width="600">
|
35 |
|
36 |
-
|
37 |
|
38 |
<img src="https://ajitrajasekharan.github.io/images/6.png" width="600">
|
39 |
|
|
|
|
|
|
|
|
|
40 |
|
|
|
25 |
- Clinical trials corpus
|
26 |
- and a small subset of Bookcorpus
|
27 |
|
28 |
+
The pretrained model was used to do NER **as is, with no fine-tuning**. The approach is described [in this post](https://ajitrajasekharan.github.io/2021/01/02/my-first-post.html). [Towards Data Science review](https://twitter.com/TDataScience/status/1486300137366466560?s=20)
|
29 |
|
30 |
[Github link](https://github.com/ajitrajasekharan/unsupervised_NER) to perform NER using this model in an ensemble with bert-base cased.
|
31 |
|
|
|
33 |
|
34 |
<img src="https://ajitrajasekharan.github.io/images/1.png" width="600">
|
35 |
|
36 |
+
### Ensemble model performance
|
37 |
|
38 |
<img src="https://ajitrajasekharan.github.io/images/6.png" width="600">
|
39 |
|
40 |
+
### License
|
41 |
+
|
42 |
+
MIT license
|
43 |
+
|
44 |
|