Mia-AiLab / Qwen3.8-27B-EXL3-3.5bpw

EXL3 quantization of Qwen/Qwen3.8-27B, the hybrid linear-attention / full-attention (gated-deltanet + full-attn) dense 27B model. 14.2 GB of weights — sized to leave room for large KV caches on 24 GB GPUs: with the built-in MTP draft head the full native 262k context fits full…