[BUG] 运行本地知识库对话报错：Internal Server Error

[RyanBin0126]

如题，
对话模型：qwen1.5-chat
嵌入模型：bce-embedding-base_v1
系统：win11

2024-06-26 12:56:07,384 model_providers.bootstrap_web.openai_bootstrap_web 52588 INFO Received create_embeddings request: {'encoding_format': 'base64',
'input': ['this is a test'],
'model': 'bce-embedding-base_v1'}
2024-06-26 12:56:07,388 model_providers.bootstrap_web.openai_bootstrap_web 52588 WARNING Warning: model not found. Using cl100k_base encoding.
2024-06-26 12:56:07,711 uvicorn.access 52588 INFO 127.0.0.1:6713 - "POST /xinference/v1/embeddings HTTP/1.1" 200
2024-06-26 12:56:07,712 httpx 69388 INFO HTTP Request: POST http://127.0.0.1:20000/xinference/v1/embeddings "HTTP/1.1 200 OK"
2024-06-26 12:56:07,719 root 69388 INFO loading vector store in 'grammar_zh/vector_store/bce-embedding-base_v1' from disk.
2024-06-26 12:56:07,959 faiss.loader 69388 INFO Loading faiss with AVX2 support.
2024-06-26 12:56:08,236 faiss.loader 69388 INFO Successfully loaded faiss with AVX2 support.
INFO: 127.0.0.1:6712 - "POST /chat/chat/completions HTTP/1.1" 500 Internal Server Error
2024-06-26 12:56:08,388 httpx 69620 INFO HTTP Request: POST http://127.0.0.1:7861/chat/chat/completions "HTTP/1.1 500 Internal Server Error"
2024-06-26 12:56:08,389 openai._base_client 69620 INFO Retrying request to /chat/completions in 0.992779 seconds

[perfece]

同样的问题，定位问题在向量化那里。我试了bge-large-zh-v1.5 和m3e-base都会报错，chatchat服务端会超时，xinference和transformers那里有问题

[huangyuanzhuo-coder]

同样的问题，定位问题在向量化那里。我试了bge-large-zh-v1.5 和m3e-base都会报错，chatchat服务端会超时，xinference和transformers那里有问题

有解决吗

[jackchen556]

碰到了相同的问题

[binxuan98]

same error

[Modas-Li]

same issue

[anderson-gznu]

I encount the same issue also.