xChar
undefined Avatar
@undefined
·2 months ago

其实要本地部署的目的有什么意义呢?没有足够的算力还不如直接用网页的

undefined Avatar
@undefined
·2 months ago

有些资料内容是不想通过联网的大模型去问答处理的

undefined Avatar
@undefined
·2 months ago

大模型的精度对应的显存大小我一直没搞明白之间的关系,比如千问这个30b和70b的模型一般都对应多少显存?

undefined Avatar
@undefined
·2 months ago

模型的显存占用推理和训练阶段的显存需求不一样,精度不一样的占用也不同

undefined Avatar
@undefined
·2 months ago
undefined Avatar
@undefined
·3 months ago
undefined Avatar
@undefined
·4 months ago
undefined Avatar
@undefined
·5 months ago
undefined Avatar
@undefined
·5 months ago
undefined Avatar
@undefined
·6 months ago
undefined Avatar
@undefined
·6 months ago
undefined Avatar
@undefined
·6 months ago