更新了langchain框架后 嵌入报错 v3 和 v3.5都出现同样的错误

#3
by hellokid200 - opened

NFO:main:Now you have 423 documents
INFO:sentence_transformers.SentenceTransformer:Load pretrained SentenceTransformer: /media/nas/emb_models/stella-large-zh-v35-1792d
INFO:sentence_transformers.SentenceTransformer:Use pytorch device_name: cuda
INFO:chromadb.telemetry.product.posthog:Anonymized telemetry enabled. See https://docs.trychroma.com/telemetry for more information.
^Z
[2]+ 已停止 /home/aivet/anaconda3/envs/chroma/bin/python /media/nas/langchain/emb_server/stella-large-zh-v35-1792d/chroma_embedding_chmed.py

我在虚拟环境下运行的

hellokid200 changed discussion title from 更新了langchain框架后 其纳入模型报错 v3 和 v3.5都出现同样的错误 to 更新了langchain框架后 嵌入报错 v3 和 v3.5都出现同样的错误
StellaEncoder org

@hellokid200 你好,先测试用sentence-transformer加载并编码是否报错,如果没报错,那就是language-chain的问题,是否有具体的报错信息呢?

6
sentence-transformer加载测试是正常的,我去检查langchain
没有返回任何报错信息,这次更新langchain后 出了好多问题

langchain 0.0.179
chromadb 0.3.27
unstructured 0.10.7
python 3.12
修正版本后,问题解决

hellokid200 changed discussion status to closed

Sign up or log in to comment