metadata
license: mit
tags:
- gguf
- llama.cpp
- q4_k_m
- cypherium
- cph
- local-ai
datasets:
- cypherium_raw
language:
- en
- ja
pipeline_tag: text-generation
CPH-Community-7B (Q4_K_M)
A compact 7B-model fine-tuned for Cypherium blockchain operations, validator support, node configuration, RPC troubleshooting, and general-purpose lightweight reasoning.
This model is optimized for CPU-only inference using llama.cpp and provides fast responses on low-resource servers such as VPS instances (2–6 vCPUs, 8–16GB RAM).
Model Description
- Base model: Qwen2-7B
- Fine-tuning: QLoRA
- Domain: Cypherium blockchain RPC, node operations, validator troubleshooting
- Format: GGUF (
Q4_K_M) - Intended use: lightweight on-device assistant for Cypherium node operators
Example Inference Command (llama.cpp)
./llama-cli \
-m cph-community-7b-q4_k_m.gguf \
-c 4096 \
-n 256 \
--system-prompt "You are a helpful Cypherium assistant." \
--prompt "Explain how to resync a Cypherium validator node."