birgermoell
commited on
Commit
•
c0449c4
1
Parent(s):
8119e66
Updated model
Browse files- README.md +14 -3
- eval_script.py +13 -0
README.md
CHANGED
@@ -70,9 +70,9 @@ model-index:
|
|
70 |
|
71 |
should probably proofread and complete it, then remove this comment. -->
|
72 |
|
73 |
-
#
|
74 |
|
75 |
-
This model is a fine-tuned version of [
|
76 |
|
77 |
eval F1-Score: **83,78**
|
78 |
|
@@ -127,8 +127,19 @@ It achieves the following results on the evaluation set:
|
|
127 |
- Loss: 0.3156
|
128 |
|
129 |
- Precision: 0.8332
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
130 |
|
131 |
-
|
|
|
|
|
132 |
|
133 |
- F1: 0.8378
|
134 |
|
|
|
70 |
|
71 |
should probably proofread and complete it, then remove this comment. -->
|
72 |
|
73 |
+
# ner-swedish-wikiann
|
74 |
|
75 |
+
This model is a fine-tuned version of [nordic-roberta-wiki](hhttps://huggingface.co/flax-community/nordic-roberta-wiki) trained for NER on the wikiann dataset.
|
76 |
|
77 |
eval F1-Score: **83,78**
|
78 |
|
|
|
127 |
- Loss: 0.3156
|
128 |
|
129 |
- Precision: 0.8332
|
130 |
+
from transformers import AutoTokenizer, AutoModelForTokenClassification
|
131 |
+
|
132 |
+
from transformers import pipeline
|
133 |
+
|
134 |
+
tokenizer = AutoTokenizer.from_pretrained("birgermoell/ner-swedish-wikiann")
|
135 |
+
|
136 |
+
model = AutoModelForTokenClassification.from_pretrained("birgermoell/ner-swedish-wikiann")
|
137 |
+
|
138 |
+
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
|
139 |
|
140 |
+
example = "Jag heter Per och jag jobbar på KTH"
|
141 |
+
|
142 |
+
nlp(example)
|
143 |
|
144 |
- F1: 0.8378
|
145 |
|
eval_script.py
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import AutoTokenizer, AutoModelForTokenClassification
|
2 |
+
|
3 |
+
from transformers import pipeline
|
4 |
+
|
5 |
+
tokenizer = AutoTokenizer.from_pretrained("birgermoell/ner-swedish-wikiann")
|
6 |
+
|
7 |
+
model = AutoModelForTokenClassification.from_pretrained("birgermoell/ner-swedish-wikiann")
|
8 |
+
|
9 |
+
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
|
10 |
+
|
11 |
+
example = "Jag heter Per och jag jobbar på KTH"
|
12 |
+
|
13 |
+
print(nlp(example))
|