JThomas-CoE/coe-gemma4-biology-mmlu_pro-14b-a4b-q4
Base model: google/gemma-4-26b-it Architecture: MoE — 26B total / ≈4B active parameters (1 shared expert + 8 routed from a pool of 128 per MoE layer, 30 MoE layers) Method: Activation-directed expert surgery — 128 → 64 experts per layer (50% reduction) Quantization: Q4KM (≈9.7 GB on disk) Tags:…
README
license: gemma language: en basemodel: google/gemma-4-26b-it tags: text-generation gguf moe mixture-of-experts domain-specialist expert-surgery mmlu-pro college-of-experts ollama math physics chemistry engineering computer-science biology economics business psychology law pipelinetag: text-generation libraryname: gguf Gemma4 College of Experts — MMLU-Pro Domain Specialists (Automated Pipeline) Base model: google/gemma-4-26b-it Architecture: MoE — 26B total / ≈4B active parameters (1 shared…
Source attribution
- HuggingFace — JThomas-CoE/coe-gemma4-biology-mmlu_pro-14b-a4b-q4
Related resources
Junhauwong/Surge-Cognition-4x8B
by JunhauwongVerdugie/STEM-Oracle-27B
by Verdugie# or·a·cle /ˈôrəkəl/ — a source of wise counsel; one who provides authoritative knowledge. From Latin ōrāculum, meaning divine announcement. In computer science, an oracle is a black box that always returns the correct answer — you don't ask it how it knows, you ask and it answers.
Welcome to the repository for Nidum-Limitless-Gemma-2B-GGUF, an advanced language model that provides unrestricted and versatile responses across a wide range of topics. This version is designed for maximum flexibility, allowing you to run it on both CPU and GPU.
jackxinning/Leanly_AI
by jackxinningUsing llama.cpp release b5466 for quantization.