alpindale/Mistral-7B-Instruct-v0.2-EETQ

Model quantized using a modified EETQ repo. Currently working on decoupling its kernels from CUTLASS to make this a bit easier to use.

8bits.