pdelobelle
commited on
Commit
•
e1eaeb2
1
Parent(s):
4c7ad12
Update README.md
Browse files
README.md
CHANGED
@@ -34,7 +34,15 @@ tags:
|
|
34 |
|
35 |
# RobBERTje finetuned for sentiment analysis on DBRD
|
36 |
|
37 |
-
This is a finetuned model based on [RobBERTje (
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
# Training data and setup
|
40 |
We used the [Dutch Book Reviews Dataset (DBRD)](https://huggingface.co/datasets/dbrd) from van der Burgh et al. (2019).
|
|
|
34 |
|
35 |
# RobBERTje finetuned for sentiment analysis on DBRD
|
36 |
|
37 |
+
This is a finetuned model based on [RobBERTje (merged)](https://huggingface.co/DTAI-KULeuven/robbertje-1-gb-non-shuffled). We used [DBRD](https://huggingface.co/datasets/dbrd), which consists of book reviews from [hebban.nl](hebban.nl). Hence our example sentences about books. We did some limited experiments to test if this also works for other domains, but this was not exactly amazing.
|
38 |
+
|
39 |
+
We released a distilled model and a `base`-sized model. Both models perform quite well, so there is only a slight performance tradeoff:
|
40 |
+
|
41 |
+
|
42 |
+
| Model | #Params. | Accuracy |
|
43 |
+
|-----------------------------------------------------------------------------------------|-----------|-----------|
|
44 |
+
| [RobBERT (v2)](https://huggingface.co/DTAI-KULeuven/robbert-v2-dutch-sentiment) | 110 M |94.4 |
|
45 |
+
| [Merged (p=0.5)](https://huggingface.co/DTAI-KULeuven/robbertje-merged-dutch-sentiment) | 74 M |92.9 |
|
46 |
|
47 |
# Training data and setup
|
48 |
We used the [Dutch Book Reviews Dataset (DBRD)](https://huggingface.co/datasets/dbrd) from van der Burgh et al. (2019).
|