unsloth/GLM-4.6-REAP-268B-A32B-GGUF

Same weights, converted vs cerebras/GLM-4.6-REAP-268B-A32B. Same weights stored in another dtype (BF16/F32 → BF16/F32), plus 356 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/369, MLP 0/365, Attention 0/644, lm_head 0/1

On the Hub · Downloads on Model Pulse