grimulkan/aurelian-alpha0.1-70b-rope8-32K-4.65bpw_h6_exl2

This is a 4-bit quantization of Aurelian v0.1alpha 70B 32K for testing & feedback. See that page for more details.

This quantization fits in 2x24GB (19/24) using Exllamav2 @ 16K context.