GuminiResearch/Gumini-1B-Base-i1-GGUF
Same weights, converted vs GuminiResearch/Gumini-1B-Base. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/21, MLP 0/30, Attention 0/70
Same weights, converted vs GuminiResearch/Gumini-1B-Base. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/21, MLP 0/30, Attention 0/70