Spaces:
No application file
No application file
File size: 2,130 Bytes
2ae875f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 |
# Load model directly
from transformers import AutoTokenizer, AutoModelForTokenClassification
import json
#tokenizer = AutoTokenizer.from_pretrained("CAMeL-Lab/bert-base-arabic-camelbert-msa-ner")
#model = AutoModelForTokenClassification.from_pretrained("CAMeL-Lab/bert-base-arabic-camelbert-msa-ner")
#print("Model loaded successfully")
from nltk.tokenize import word_tokenize
# Function to extract names using MAREFA NER model
def extract_arabic_names(json_data, model, tokenizer):
arabic_names = set()
for entry in json_data:
if "Arabic Text" in entry:
text = entry["Arabic Text"]
tokenized_text = tokenizer.tokenize(text)
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
predictions = outputs.logits.argmax(dim=-1)
predicted_labels = [model.config.id2label[label_id] for label_id in predictions[0]]
current_name = ""
for token, label in zip(tokenized_text, predicted_labels):
if label == "B-person":
current_name = token
elif label == "I-person":
current_name += " " + token
elif label != "O" and current_name:
arabic_names.add(current_name)
current_name = ""
if current_name:
arabic_names.add(current_name)
return arabic_names
# Load the MAREFA NER model and tokenizer
model_name = "CAMeL-Lab/bert-base-arabic-camelbert-msa-ner"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForTokenClassification.from_pretrained(model_name)
basic='cache/output/basic_info_frame.json'
# Load JSON data from the file
with open(basic, "r", encoding="utf-8") as file:
json_data = json.load(file)
# Extract names from the JSON data using MAREFA model
arabic_names = extract_arabic_names(json_data, model, tokenizer)
# Print the extracted names
if arabic_names:
print("Arabic names extracted:")
for name in arabic_names:
print("Name:", name)
else:
print("No Arabic names found.") |