Unified release repository for Qwen3.8-27B quantizations, fine-tunes, and speculative decoding configurations optimized for llamAmpere on SM86 (RTX 3090 / 3090 Ti). Run with native MTP speculative drafting
Independent publisher
Jake
jakeatx
distillation, MoE tuning + optimizing, offloading
Models in Library1
Datasets in Library0
Models on Hugging Face23
Followers12