mradermacher/nekomata-14b-instruction-GGUF
Quantization vs rinna/nekomata-14b-instruction. mradermacher/nekomata-14b-instruction-GGUF is a GGUF quantization (Q8_0) of the same weights: 121 tensors match exactly and 202 within quantization error.
Tensors changed: lm_head 0/1, Attention 0/120, Norms 0/81, MLP 0/120, Embeddings 0/1