mlx-community/Kimi-K2-Thinking

Same weights, repacked vs moonshotai/Kimi-K2-Thinking. Same weights in different files (re-saved, re-sharded or new metadata), plus 540 new tensors.

Tensors changed: Norms 0/245, MLP 0/303, Attention 0/305, lm_head 0/1, Embeddings 0/1

On the Hub · Downloads on Model Pulse