moka-ai/m3e-base · Discussions

#37 opened 6 months ago by

zzk520

可以使用Transformer去启动模型吗？

#35 opened 6 months ago by

wss3390418

微调效果和在simcse上一样

#34 opened 7 months ago by

zhao26413

Ai editing

#31 opened 8 months ago by

mustyalaharsha

感觉相似度不是很准啊，难道要自己训练？

#30 opened 8 months ago by

xyqxyq

有最新的向量模型吗？

4

#29 opened 8 months ago by

LycheeX

🚩 Report: Ethical issue(s)

#28 opened 8 months ago by

LS001

转TorchScript

#27 opened 9 months ago by

JeanWhite

感谢作者的贡献，烦请问下代码数据的能力有时间点安排吗？

#26 opened 9 months ago by

mbear

在实例化 model = SentenceTransformer('moka-ai/m3e-base') 时，为什么显示 ConnectionError呢？

#25 opened 9 months ago by

justYeah

请问在CPU上推理有什么加速方案呢

#23 opened 9 months ago by

Ark0810

如何同时使用多个GPU进行预测推理

#22 opened 9 months ago by

yuqi-zhang

关于使用上的一些疑问

#21 opened 9 months ago by

Vaportang

请问用m3e-base作为embedding模型，对于文本的长度限制有没有比较推荐的设置

#20 opened 9 months ago by

demonai

embedding维度问题

8

#19 opened 9 months ago by

vicp

微调自己数据集泛化性较差

#18 opened 10 months ago by

chriske

训练数据集

#17 opened 10 months ago by

zzlife

请问加载模型后产出embedding很慢有什么办法么？

#15 opened 10 months ago by

EEik

m3e 如何来做中英文混合的文本分类

#14 opened 10 months ago by

anbangli

归一化

4

#13 opened 10 months ago by

nightlan

转onnx。

4

#12 opened 10 months ago by

angellee

感谢作者和问题请教

#11 opened 10 months ago by

rookieaca

请问计算相似性的时候，用向量的欧式距离还是内积比较好呢？

#10 opened 10 months ago by

firdota

缺少modules.json文件

#9 opened 10 months ago by

GodsDusk

最长支持多少token的embedding？

7

#8 opened 10 months ago by

cluo

m3e在大并发场景下（如30并发）CPU是否可以支撑，是否支持华为atlas昇腾的GPU卡

#7 opened 10 months ago by

wayne7356

m3e-base embedding api.ipynb - Colaboratory