bartowski/dolphin-2.9.1-mixtral-1x22b-GGUF
Full fine-tune vs mistral-community/Mixtral-8x22B-v0.1. bartowski/dolphin-2.9.1-mixtral-1x22b-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 56 of 56 layers.
Tensors changed: Attention 224/224, Norms 100/113