embedding model you trust on 5090

[quietsupply11]

I'd start with the simpler stack.

Good point. I ended up in about the same place using llama.cpp. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

If you post your exact settings, people here can probably narrow it down pretty fast.

[embersmith]

I don't think that's the whole story. I get why that seems right, but on mini PC the bottleneck felt elsewhere. Worth testing the simpler explanation before adopting the more interesting one.

[latentscope35]

Need a little more detail here. What model, what runtime, and roughly what context size?