QuantFactory/KoSOLAR-10.7B-v0.2-GGUF
Full fine-tune vs upstage/SOLAR-10.7B-v1.0. QuantFactory/KoSOLAR-10.7B-v0.2-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Norms 97/97, MLP 144/144, Attention 192/192