Papuguinho/gemmuguinho-636m
MoE built from dense models vs google/gemma-3-270m. Papuguinho/gemmuguinho-636m is a mixture of 8 experts with the same shape as google/gemma-3-270m's MLP, but none of them is a copy of it: they come from other models or were trained further. Shared tensors (attention, norms, embeddings): 0 of 181 identical.
1 derivatives checked
- Papuguinho/gemmuguinho-636m-GGUF: Same weights, converted