test.py

# from gliner2 import GLiNER2

# extractor = GLiNER2.from_pretrained("fastino/gliner2-multi-v1")

# text = "En 1887, la Société suisse du Grutlis'est accrue de 40 sections; l'association compte actuellement 12,000 membres."

# labels = ["person", "location", "organization", "time", "production"]

# entities = extractor.extract_entities(text, labels, include_confidence=True, threshold=0.3)

# print(entities)

from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline

model_id = "emanuelaboros/historical-ner-baseline"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForTokenClassification.from_pretrained(model_id)

ner = pipeline(
    "token-classification",
    model=model,
    tokenizer=tokenizer,
    aggregation_strategy="simple",
)

text = "En 1887, la Société suisse du Grutlis'est accrue de 40 sections; en 1977, l'association compte actuellement 12,000 membres."

predictions = ner(text)

for entity in predictions:
    print(entity)