GitHub · HTML
demiralpdev/Caphlon-Model
A 6.87 GiB GGUF build of Qwen 3.8 27B — full 27B quality on 8 GB RAM. 96.7% HumanEval, 92.9% MMLU retention, no distillation.
- Stars
- 19
- Forks
- 2
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: 27b, edge-llm, gguf, iq2s, layer-pruning, lightweight-llm, llama-cpp, llm, local-llm, model-card
Latest data point: Nov 22, 2026 · 19 stars · 2 forks