qwen3.8-flash-next-lil-rascal-uncensored
abliterated with homebrew workflow.
- Quant: local-inference-lab/Qwen3.8-Flash-Next-NVFP4 — the NVFP4/MXFP8 quantized QAD checkpoint this uncensor is built directly on (
template_modelinexport-manifest.json). - Upstream: Qwen/Qwen3.8-Flash-Next
- License: Qwen Community License 1.0 (inherited)
Serve with the single-Spark vLLM recipe (TP=1, MTP 3, fp8 KV, MAX_MODEL_LEN 262144).