医疗领域中文命名实体识别
项目地址:https://github.com/iioSnail/chinese_medical_ner
使用方法:
from transformers import AutoModelForTokenClassification, BertTokenizerFast
tokenizer = BertTokenizerFast.from_pretrained('iioSnail/bert-base-chinese-medical-ner')
model = AutoModelForTokenClassification.from_pretrained("iioSnail/bert-base-chinese-medical-ner")
sentences = ["瘦脸针、水光针和玻尿酸详解!", "半月板钙化的病因有哪些?"]
inputs = tokenizer(sentences, return_tensors="pt", padding=True, add_special_tokens=False)
outputs = model(**inputs)
outputs = outputs.logits.argmax(-1) * inputs['attention_mask']
print(outputs)
输出结果:
tensor([[1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 4, 4],
[1, 2, 2, 2, 3, 4, 4, 4, 4, 4, 4, 4, 0, 0]])
其中 1=B, 2=I, 3=E, 4=O
。1, 3
表示一个二字医疗实体,1,2,3
表示一个3字医疗实体, 1,2,2,3
表示一个4字医疗实体,依次类推。
可以使用项目中的MedicalNerModel.format_outputs(sentences, outputs)
来将输出进行转换。
效果如下:
[
[
{'start': 0, 'end': 3, 'word': '瘦脸针'},
{'start': 4, 'end': 7, 'word': '水光针'},
{'start': 8, 'end': 11, 'word': '玻尿酸'}、
],
[
{'start': 0, 'end': 5, 'word': '半月板钙化'}
]
]
- Downloads last month
- 83
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.