unsloth/MAI-DS-R1-GGUF

Same weights, converted vs microsoft/MAI-DS-R1. Same weights stored in another dtype (BF16 → BF16/F32), plus 354 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/245, MLP 0/241, Attention 0/244, lm_head 0/1

On the Hub · Downloads on Model Pulse