noctrex/Kimi-Linear-48B-A3B-Instruct-MXFP4_MOE-GGUF

Same weights, converted vs moonshotai/Kimi-Linear-48B-A3B-Instruct. Same weights stored in another dtype (BF16/F32 → MXFP4/F16), plus 138 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/82, MLP 0/81, Attention 0/201, other 0/26, lm_head 0/1

On the Hub · Downloads on Model Pulse