mradermacher/Jupiter-G-8B-GGUF
Same weights, converted (fewer layers) vs locailabs/Jupiter-G-8B. mradermacher/Jupiter-G-8B-GGUF matches part of locailabs/Jupiter-G-8B (677 of its 2130 tensors; the rest of locailabs/Jupiter-G-8B has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/BF16), plus 43 new tensors.
Tensors changed: Embeddings 0/2, Norms 0/296, MLP 0/126, other 0/85, Attention 0/168