在我司服務器上部署，但是嘗試了其他文章的docker鏡像，出現算子錯誤等問題，后面又用回

參考文獻
https://blog.csdn.net/sccum/article/details/145995250
https://blog.csdn.net/qq_41994821/article/details/147223327
https://ascend.readthedocs.io/zh-cn/latest/sources/sentence_transformers/install.html

docker run -it -d --net=host --shm-size=20g  \
  --name Xinf \
  --device /dev/davinci4 \
  --device /dev/davinci5 \
  --device /dev/davinci_manager \
  --device /dev/devmm_svm \
  --device /dev/hisi_hdc \
  -v /usr/local/dcmi:/usr/local/dcmi \
  -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
  -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
  -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
  -v /etc/ascend_install.info:/etc/ascend_install.info \
  -v /root/.cache:/root/.cache \
  -v /share/models/embedding/bge-reranker-v2-m3:/bge-reranker-v2-m3:ro \
  -p 9997:9997 \
  -it $IMAGE bash

# 終端1
export XINFERENCE_ENDPOINT=http://0.0.0.0:9997
xinference-local --host 0.0.0.0 --port 9997
# 終端2
export XINFERENCE_ENDPOINT=http://127.0.0.1:9997
## 本地下載重排序模型之后
xinference launch --model-name bge-reranker-v2-m3 --model-type rerank --model-path /bge-reranker-v2-m3

后面我直接通過打開http://127.0.0.1:9997配置embedding和rerank模型。