Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
13 |
|
14 |
# RobBERTBestModelOct13
|
15 |
|
16 |
-
This model is a fine-tuned version of
|
17 |
It achieves the following results on the evaluation set:
|
18 |
- eval_loss: 0.3695
|
19 |
- eval_precisions: 0.8647
|
@@ -35,7 +35,9 @@ More information needed
|
|
35 |
|
36 |
## Training and evaluation data
|
37 |
|
38 |
-
|
|
|
|
|
39 |
|
40 |
## Training procedure
|
41 |
|
|
|
13 |
|
14 |
# RobBERTBestModelOct13
|
15 |
|
16 |
+
This model is a fine-tuned version of pdelobelle/robbert-v2-dutch-base (https://huggingface.co/ pdelobelle/robbert-v2-dutch-base) on the annotated part of the Moroccorp.
|
17 |
It achieves the following results on the evaluation set:
|
18 |
- eval_loss: 0.3695
|
19 |
- eval_precisions: 0.8647
|
|
|
35 |
|
36 |
## Training and evaluation data
|
37 |
|
38 |
+
Both training and evalutation data are sampled from the Moroccorp, a dataset that consists of chat conversation from an internet forum for Moroccan-Dutch people called maroc.nl
|
39 |
+
The dataset is labeled on word-level with labels for the three most common languages in the dataset: Dutch (NL), English (ENG), Moroccan Languages (MOR).
|
40 |
+
Additionally, labels for Named entities (NAME), language independent utterances (NON) and words from other languages (OTH) are used.
|
41 |
|
42 |
## Training procedure
|
43 |
|