文档嵌入模型bett-base-chinese

Question

你好，我把bert-base-chinese模型下载到D:/Jupyter/bert-base-chinese文件夹下，但是在运行embedding_model = SentenceTransformer('D:/Jupyter/bert-base-chinese')时，为什么出现这个错误呢 ： No sentence-transformers model found with name D:/Jupyter/bert-base-chinese. Creating a new one with MEAN pooling.
Some weights of the model checkpoint at D:/Jupyter/bert-base-chinese were not used when initializing BertModel: ['cls.seq_relationship.bias', 'cls.predictions.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.decoder.weight', 'cls.predictions.transform.LayerNorm.weight', 'cls.seq_relationship.weight', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.dense.bias']

• This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
• This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).

Answer 1

引自讯飞星火
这个错误是因为你尝试从预训练模型的检查点初始化一个与预期完全相同的模型，但SentenceTransformer库无法找到与预期完全相同的模型。在这种情况下，它会创建一个新的模型并使用MEAN pooling。

如果你想要使用与预期完全相同的模型，可以尝试以下方法：

1. 确保你已经正确安装了sentence-transformers库。你可以使用以下命令安装：

pip install sentence-transformers

1. 使用与预期完全相同的预训练模型进行初始化。例如，如果你想使用BERT-Base-Chinese模型进行序列分类任务，可以这样做：

from sentence_transformers import SentenceTransformer, models

model_name = 'bert-base-chinese'
word_embedding_model = models.Transformer(model_name)
pooling_model = models.Pooling(word_embedding_model.get_word_embedding_dimension())
sentence_transformer_model = SentenceTransformer(modules=[word_embedding_model, pooling_model])

这样，你就可以使用与预期完全相同的BERT-Base-Chinese模型进行序列分类任务了。