Hugging Face's logo

anwesham
/
indic_with_mbert_tokens

Model card Files Community
anwesham commited on
Commit
2038bb8
•
1 Parent(s): f6b4225

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -6
README.md CHANGED
@@ -1,18 +1,17 @@
1
  AI4Bharat's IndicBERT finetuned for few-shot transfer learning by fine-tuning on Hindi training data with Urdu validation and test sets. Expected low accuracy. Leverages mbert's tokenizer in this implementation.
2
 
3
  ---
4
-
5
  language:
6
- - ru
7
- - en
8
  tags:
9
- - translation
 
10
  license: apache-2.0
11
  datasets:
12
- - wmt19
13
  metrics:
14
  - f1
15
  - recall
 
16
  - accuracy
17
-
18
  ---
1
  AI4Bharat's IndicBERT finetuned for few-shot transfer learning by fine-tuning on Hindi training data with Urdu validation and test sets. Expected low accuracy. Leverages mbert's tokenizer in this implementation.
2
 
3
  ---
 
4
  language:
5
+ - ur
 
6
  tags:
7
+ - named entity recognition
8
+ - ner
9
  license: apache-2.0
10
  datasets:
11
+ - wikiann
12
  metrics:
13
  - f1
14
  - recall
15
+ - precision
16
  - accuracy
 
17
  ---