calcuis/dia-gguf
Same weights, converted vs nari-labs/Dia-1.6B. Same weights stored in another dtype (F32 → F16/F32).
Tensors changed: Embeddings 0/10, Attention 0/192, MLP 0/60, Norms 0/80, other 0/1
Same weights, converted vs nari-labs/Dia-1.6B. Same weights stored in another dtype (F32 → F16/F32).
Tensors changed: Embeddings 0/10, Attention 0/192, MLP 0/60, Norms 0/80, other 0/1