metadata
datasets:
- guyyanko/hebrew-trc-special-markers
language:
- he
from transformers import TextClassificationPipeline
from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassification
class TemporalRelationClassificationPipeline(TextClassificationPipeline):
def check_model_type(self, supported_models):
pass
pretrained_checkpoint = "guyyanko/split-3-hebrew-trc-alephbert-base-EMP"
model = AutoModelForSequenceClassification.from_pretrained(pretrained_checkpoint, trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(pretrained_checkpoint, trust_remote_code=True)
classifier = pipeline(task='text-classification', model=model, tokenizer=tokenizer)
txt = "诪讞专 [讗1] 讗转讗诪谉 [/讗1] 讗诐 [讗2] 讗住讬讬诐 [/讗2] 讗转 讻诇 讛诪砖讬诪讜转 砖诇讬"
print(classifier(txt))
txt = "讗讞专讬 [讗1] 砖讗住讬讬诐 [/讗1] 讗转 讻诇 讛诪砖讬诪讜转 砖诇讬 [讗2] 讗诇讱 [/讗2] 诇讛转讗诪谉 讘讞讚专 讛讻讜砖专"
print(classifier(txt))