Named entity recognition On Persian dataset

traindataset=20484 persian sentense

valdataset=2561

AutoTokenizer=HooshvareLab/bert-fa-base-uncased

ner_tags= ['O', 'B-pro', 'I-pro', 'B-pers', 'I-pers', 'B-org', 'I-org', 'B-loc', 'I-loc', 'B-fac', 'I-fac', 'B-event', 'I-event']

training_args= learning_rate=2e-5,

per_device_train_batch_size=16,

per_device_eval_batch_size=16,

num_train_epochs=4,

weight_decay=0.01

Training Loss=0.001000

sample1: 'entity': 'B-loc', 'score': 0.9998902, 'index': 2, 'word': 'تهران',

sample2: 'entity': 'B-pers', 'score': 0.99988234, 'index': 2, 'word': 'عباس',

for use this model:

from transformers import pipeline

pipe = pipeline("token-classification", model="NLPclass/Named_entity_recognition_persian")

sentence = ""

predicted_ner = pipe(sentence)

for entity in predicted_ner:

    print(f"Entity: {entity['word']}, Label: {entity['entity']}")
Downloads last month
11
Safetensors
Model size
162M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.