language:

• 希伯来语 pipeline_tag: 零样本分类 datasets:
• HeTree/MevakerConcTree license: apache-2.0 widget:
• text: "上周我升级了手机系统版本。" candidate_labels: "通话设备, 账单支付, 足球狂热" multi_class: false example_title: "设备升级"

希伯来语交叉编码器模型

使用方法

from sentence_transformers import CrossEncoder
model = CrossEncoder('HeTree/HeCross')

# 分数（已通过sigmoid处理）
scores = model.predict([('柏林有多少人口？', '柏林在891.82平方公里区域内登记有3,520,031名居民。'), 
                        ('柏林有多少人口？', '纽约市以大都会艺术博物馆闻名。')])
print(scores)

使用Transformers的AutoModel

您也可以直接使用Transformers库（无需SentenceTransformers库）：

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
import numpy as np

# 对分数应用sigmoid的函数
def sigmoid(x):
    return 1 / (1 + np.exp(-x))

model = AutoModelForSequenceClassification.from_pretrained('HeTree/HeCross')
tokenizer = AutoTokenizer.from_pretrained('HeTree/HeCross')
features = tokenizer(['柏林有多少人口？', '柏林有多少人口？'],
                     ['柏林在891.82平方公里区域内登记有3,520,031名居民。', '纽约市以大都会艺术博物馆闻名。'],
                     padding=True, truncation=True, return_tensors="pt")
model.eval()
with torch.no_grad():
    scores = sigmoid(model(**features).logits)
    print(scores)

零样本分类

此模型也可用于零样本分类：

from transformers import pipeline
classifier = pipeline("zero-shot-classification", model='HeTree/HeCross')
sent = "上周我升级了手机系统版本。"
candidate_labels = ["通话设备", "网站", "账单支付", "银行账户访问"]
res = classifier(sent, candidate_labels)
print(res)

引用

如果您在研究中使用了HeCross，请引用Mevaker: Conclusion Extraction and Allocation Resources for the Hebrew Language。

@article{shalumov2024mevaker,
      title={Mevaker: Conclusion Extraction and Allocation Resources for the Hebrew Language}, 
      author={Vitaly Shalumov and Harel Haskey and Yuval Solaz},
      year={2024},
      eprint={2403.09719},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}