Quanted using lmdeploy - auto_awq. Going to run benchmarks to see performence.

Ran it up on 4x A4000 GPUs with 32k context. Got 30 t/s per request.

Downloads last month
16
Safetensors
Model size
2B params
Tensor type
I32
·
F16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support