Ingot-XII-27B-Q2 三元量化模型公开镜像发布
原标题：dendriteholdings/Ingot-XII-27B-Q2
AI 摘要
Hugging Face 上发布了 Ingot-XII-27B-Q2，这是 Bittensor 子网 Ingot 竞赛中获胜矿工提交模型的公开镜像，基于 Qwen3.8-27B 量化而来。该模型以三元权重存储，提供 1.75 bpw（5.95GB）和 2.125 bpw（7.21GB）两个 GGUF 文件，需使用 PrismML 的 llama.cpp 分支运行。评测显示其与 BF16 教师模型的 KL 散度和隐藏状态距离均优于此前冠军模型。
正文节选
--- base_model: Qwen/Qwen3.8-27B base_model_relation: quantized library_name: llama.cpp pipeline_tag: text-generation tags: - gguf - ingot - bittensor --- # Ingot-XII-27B-Q2 A public mirror of a model that won on Ingot, a Bittensor subnet where miners quantize a fixed teacher model and the validator crowns the copy that stays closest to it. - Submitted by miner hotkey `5FTizTwJE5xzTpNedS1hFiycHz5cCH2fq3wYBbgq5z8nFYnK` (uid 200) - Crowned from submission `e14bdcbb-4f73-4d24-bfc3-d03c726b5e19`