GGUF quants of Qwen 9B, 27B, 35B-A3B and Flash-Next from Nace.AI, measured at 128K context.
AI & ML interests
Meta Learning
Recent Activity
View all activity
GGUF quants of Qwen 9B, 27B, 35B-A3B and Flash-Next from Nace.AI, measured at 128K context.
Nace's calibrated MLX packs for Mac: closer to the original model than standard 4-bit packs at about the same size. Lower KL is closer.
Nace AI Decision models
Datasets used for hypernetwork knowledge-injection and Wiki5M synthetic multihop QA experiments.