---
license: cc-by-nc-4.0
language:
  - en
library_name: llama.cpp
pipeline_tag: text-generation
base_model: Auroraventures/cipher-sft25-merged
tags:
  - gemma4
  - gemma-4-31b
  - cipher
  - kin
  - gguf
  - q4_k_m
  - quantized
  - llama.cpp
  - ollama
  - retired
  - creative-coding
  - web-design
  - text-generation
inference: false
---

# Cipher SFT 2.5 (Synthetic, retired) — Q4_K_M GGUF 🦑⚠️

4-bit GGUF of [`Auroraventures/cipher-sft25-merged`](https://huggingface.co/Auroraventures/cipher-sft25-merged). **Retired** due to template collapse on synthetic data.

> 🔄 Use [`cipher-sft25-real-merged-Q4_K_M-GGUF`](https://huggingface.co/Auroraventures/cipher-sft25-real-merged-Q4_K_M-GGUF) when published. It's trained on real scraped Awwwards-grade source code and recovers the long-tail diversity synthetic SFT lost.

## License

CC-BY-NC-4.0. Gemma-4 base terms apply.

---

*Built with 🦑 by [Aurora Ventures](https://huggingface.co/Auroraventures).*