mradermacher/ultiima-108B-i1-GGUF

Quantization vs Sakalti/ultiima-108B. mradermacher/ultiima-108B-i1-GGUF is a GGUF quantization (IQ3_XXS) of the same weights: 601 tensors match exactly and 842 within quantization error.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/241, MLP 0/360, Attention 0/840

On the Hub · Downloads on Model Pulse