QomSSLab/Anonymizer-v3

This repository hosts an XLM-RoBERTa token-classification head trained.

Usage

from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline

model_id = "QomSSLab/Anonymizer-v3"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForTokenClassification.from_pretrained(model_id)
tagger = pipeline("token-classification", model=model, tokenizer=tokenizer, aggregation_strategy="simple")

text = "مثال از یک ورودی فارسی"
for entity in tagger(text):
    print(entity)

Labels

• ACOUNT
• ADDRESS
• AMOUNT
• DATE
• DOCUMENT_ID
• ID
• JOB
• O
• ORG
• ORG_BRANCH
• PERSON
• PHONEـNUMBER
• PLATEـNUMBER

Metrics

Validation Metrics

• Precision: 0.9337
• Recall: 0.9488
• F1: 0.9412
• Accuracy: 0.9919

Per-label Breakdown

Label	Precision	Recall	F1	Support
ACOUNT	1.0000	1.0000	1.0000	0
ADDRESS	0.9683	0.9934	0.9807	1970
AMOUNT	1.0000	0.9947	0.9973	188
DATE	1.0000	0.9734	0.9865	338
DOCUMENT_ID	0.9888	0.9902	0.9895	711
ID	0.9718	1.0000	0.9857	69
JOB	1.0000	0.4364	0.6076	55
O	0.9953	0.9952	0.9952	18610
ORG	0.8037	0.8866	0.8431	97
ORG_BRANCH	0.9833	0.9584	0.9707	553
PERSON	0.9978	0.9928	0.9953	1385
PHONEـNUMBER	1.0000	0.9944	0.9972	178
PLATEـNUMBER	0.0000	1.0000	0.0000	0