mradermacher/next2-fast-GGUF
Same weights, converted (fewer layers) vs thelamapi/next2-fast. mradermacher/next2-fast-GGUF matches part of thelamapi/next2-fast (444 of its 883 tensors; the rest of thelamapi/next2-fast has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/205, MLP 0/102, Attention 0/136