Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,30 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
---
|
4 |
+
This is a NER model meant to be used to detect/extract citations from American legal documents.
|
5 |
+
|
6 |
+
Ignore the widget on the model card page; see below for usage.
|
7 |
+
|
8 |
+
## How to Use the Model
|
9 |
+
|
10 |
+
This model outputs token-level predictions, which should be processed as follows to obtain meaningful labels for each token:
|
11 |
+
|
12 |
+
```python
|
13 |
+
from transformers import AutoTokenizer, AutoModelForTokenClassification
|
14 |
+
import torch
|
15 |
+
|
16 |
+
tokenizer = AutoTokenizer.from_pretrained("ss108/legal-citation-bert")
|
17 |
+
model = AutoModelForTokenClassification.from_pretrained("ss108/legal-citation-bert")
|
18 |
+
|
19 |
+
text = "Your example text here"
|
20 |
+
inputs = tokenizer(text, return_tensors="pt", padding=True)
|
21 |
+
outputs = model(**inputs)
|
22 |
+
|
23 |
+
logits = outputs.logits
|
24 |
+
predictions = torch.argmax(logits, dim=-1)
|
25 |
+
|
26 |
+
tokens = tokenizer.convert_ids_to_tokens(inputs['input_ids'][0])
|
27 |
+
predicted_labels = [model.config.id2label[p.item()] for p in predictions[0]]
|
28 |
+
|
29 |
+
for token, label in zip(tokens, predicted_labels):
|
30 |
+
print(f"{token}: {label}")
|