Is device mapping supported?

[dancixx]

Hey Guys,

Is device mapping supported? I mean to share the model between the host and gpu automatically, especially in the case of Qwen models?

[sonam-pankaj95]

Hi @dancixx,

Embedanything has two pip, when you use embed-anything-gpu it runs on the GPU as the default device, else it runs on the CPU.
There is no mapping, but two different pips.

I know this isn't the best practice, but we wanted to release the GPU version quickly.

Let me know if you need any help.

Thanks,