mradermacher/Nyanko-64-GGUF
Same weights, converted vs beyoru/Nyanko-64. Same weights stored in another dtype (F32 → F16/F32), plus 84 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/113, MLP 0/28, Attention 0/112
Same weights, converted vs beyoru/Nyanko-64. Same weights stored in another dtype (F32 → F16/F32), plus 84 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/113, MLP 0/28, Attention 0/112