Spaces:
Sleeping
Sleeping
Update main.py
Browse files
main.py
CHANGED
|
@@ -1,13 +1,22 @@
|
|
| 1 |
import chardet
|
| 2 |
import spacy
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
|
| 5 |
# Load SpaCy and add GLiNER to the pipeline
|
| 6 |
nlp = spacy.load("en_core_web_lg")
|
| 7 |
-
nlp.add_pipe("gliner_spacy", config=
|
| 8 |
-
"labels": ["PERSON", "ORGANIZATION", "LOCATION", "DISEASE"],
|
| 9 |
-
"model": "urchade/gliner_multi_pii-v1"
|
| 10 |
-
}, last=True)
|
| 11 |
|
| 12 |
def detect_encoding(file_bytes):
|
| 13 |
result = chardet.detect(file_bytes)
|
|
|
|
| 1 |
import chardet
|
| 2 |
import spacy
|
| 3 |
+
|
| 4 |
+
# ------------------------
|
| 5 |
+
# CONFIGURATION
|
| 6 |
+
# ------------------------
|
| 7 |
+
custom_spacy_config = {
|
| 8 |
+
"gliner_model": "urchade/gliner_multi_pii-v1",
|
| 9 |
+
"labels": [
|
| 10 |
+
"person", "organization", "company", "country",
|
| 11 |
+
"medical condition", "credit card brand",
|
| 12 |
+
],
|
| 13 |
+
"threshold": 0.39,
|
| 14 |
+
"style": "ent",
|
| 15 |
+
}
|
| 16 |
|
| 17 |
# Load SpaCy and add GLiNER to the pipeline
|
| 18 |
nlp = spacy.load("en_core_web_lg")
|
| 19 |
+
nlp.add_pipe("gliner_spacy", config=custom_spacy_config)
|
|
|
|
|
|
|
|
|
|
| 20 |
|
| 21 |
def detect_encoding(file_bytes):
|
| 22 |
result = chardet.detect(file_bytes)
|