---
tags:
- gguf
- llama.cpp
- quantized
- jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF
license: apache-2.0
---

# jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF

This model was converted to GGUF format from [`jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF`](https://huggingface.co/jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF) using llama.cpp via
[Convert Model to GGUF](https://github.com/ruslanmv/convert-model-to-GGUF).

**Key Features:**

* Quantized for reduced file size (GGUF format)
* Optimized for use with llama.cpp
* Compatible with llama-server for efficient serving

Refer to the [original model card](https://huggingface.co/jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF) for more details on the base model.

## Usage with llama.cpp

**1. Install llama.cpp:**

```bash
brew install llama.cpp  # For macOS/Linux
```

**2. Run Inference:**

**CLI:**

```bash
llama-cli --hf-repo jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF --hf-file SummLlama3.2-3B-Q4_K_M-GGUF-4bit.gguf -p "Your prompt here"
```

**Server:**

```bash
llama-server --hf-repo jsjeon/SummLlama3.2-3B-Q4_K_M-GGUF --hf-file SummLlama3.2-3B-Q4_K_M-GGUF-4bit.gguf -c 2048
```

For more advanced usage, refer to the [llama.cpp repository](https://github.com/ggerganov/llama.cpp).