unsloth/Qwen3.8-2.4T-A95B-GGUF

Same weights, converted vs Qwen/Qwen3.8-2.4T-A95B. Same weights stored in another dtype (BF16 → BF16/F32), plus 365 new tensors.

Tensors changed: Norms 0/231, MLP 0/460, Attention 0/92, Embeddings 0/1, lm_head 0/1

3 derivatives checked