mradermacher/Horiza-RAG-base-8b-i1-GGUF
Quantization (fewer layers) vs Lazysniper/Horiza-RAG-base-8b. mradermacher/Horiza-RAG-base-8b-i1-GGUF matches part of Lazysniper/Horiza-RAG-base-8b (809 of its 1676 tensors; the rest of Lazysniper/Horiza-RAG-base-8b has no counterpart). On the shared part: mradermacher/Horiza-RAG-base-8b-i1-GGUF is a GGUF quantization (Q6_K) of the same weights: 493 tensors match exactly and 316 within quantization error.
Tensors changed: Embeddings 0/1, other 0/246, Norms 0/317, MLP 0/105, Attention 0/140