mirror of
https://github.com/modelscope/modelscope.git
synced 2025-12-24 20:19:22 +01:00
16 lines
601 B
Python
16 lines
601 B
Python
from modelscope.models import SbertForNLI
|
|
from modelscope.pipelines import pipeline
|
|
from modelscope.preprocessors import NLIPreprocessor
|
|
|
|
model = SbertForNLI('../nlp_structbert_nli_chinese-base')
|
|
print(model)
|
|
tokenizer = NLIPreprocessor(model.model_dir)
|
|
|
|
semantic_cls = pipeline('nli', model=model, preprocessor=tokenizer)
|
|
print(type(semantic_cls))
|
|
|
|
print(
|
|
semantic_cls(
|
|
input=('我想还有一件事也伤害到了老师的招聘，那就是他们在课堂上失去了很多的权威',
|
|
'教师在课堂上失去权威，导致想要进入这一职业的人减少了。')))
|