mradermacher/Shivik-1B-MEGA-GGUF
Same weights, converted vs navflyer/Shivik-1B-MEGA. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/33, MLP 0/48, Attention 0/64
Same weights, converted vs navflyer/Shivik-1B-MEGA. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/33, MLP 0/48, Attention 0/64