Apache 2.0
Model weights: (c) Alibaba Group / Qwen team. Abliterated (uncensored)
weights: (c) huihui-ai / Huihui-Qwen3.8-27B-abliterated. Quantized with the
IQ1_M 1-bit technique from unsloth / llama.cpp. YaRN rope scaling baked into
GGUF metadata provides 1M context (factor 4.0 over the 262144 base).