embedding输出是否经过归一化处理

#10
by lehuangg - opened

请问bge-large-zh-v1.5输出的结果经过归一化处理了吗?

Beijing Academy of Artificial Intelligence org

使用FlagModel和sentence transformers的话,默认归一化了。使用huggingface, 需要sentence_embeddings = torch.nn.functional.normalize(sentence_embeddings, p=2, dim=1)

好的,感谢您的回答。

lehuangg changed discussion status to closed

Sign up or log in to comment