---
license: mit
library_name: llama-cpp
tags:
  - gguf
  - q8_0
  - llama-cpp
  - local-llm
  - build-small-hackathon
---

# Phase-3 GGUF

This repository hosts the Phase-3 GGUF artifact used by the public Gradio
Space `build-small-hackathon/First-Principle-AI`.

## File

- `model-Q8_0.gguf`
- Local source before upload:
  `/Users/user/.lmstudio/models/owenisas/Phase-3-GGUF/model-Q8_0.gguf`
- Quantization label: Q8_0
- Approximate local size: 31 GB

## Intended Use

The paired Space loads this artifact with `llama-cpp-python` and exposes a
dense model-lab chatbot with sampling controls and runtime diagnostics.

## Runtime Caveat

This is a large Q8 GGUF. It needs substantially more memory than a typical
free CPU Space can provide. The paired Space is configured to target ZeroGPU,
but ZeroGPU compatibility with llama.cpp/GGUF should be verified at runtime.