Edgeai-gio/SmolCpp-MoE-360M

Same weights, converted vs HuggingFaceTB/SmolLM2-360M. Same weights stored in another dtype (BF16 → F16/F32), plus 128 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/65, Attention 0/128

On the Hub · Downloads on Model Pulse