mradermacher/TinyDolphin-3x-MoE-GGUF

Same weights, converted vs jtatman/TinyDolphin-3x-MoE. Same weights stored in another dtype (F16 → F16/F32), plus 66 new tensors.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/45, other 0/22, Attention 0/88

On the Hub · Downloads on Model Pulse