---
license: gemma
base_model: Catter58/CASELLM-26b-a4b-evaluation-full
tags:
  - gguf
  - gemma4
  - quantized
  - q4_k_m
---

# CASELLM-26b-a4b-evaluation (GGUF, Q4_K_M)

Q4_K_M quantization of `Catter58/CASELLM-26b-a4b-evaluation-full`.

- Architecture: Gemma4 (MoE, 26B total / 4B active)
- Quantization: Q4_K_M (~16 GB)
- Converted with `llama.cpp` `convert_hf_to_gguf.py`

## Usage (llama.cpp)

```bash
llama-cli -m casellm-26b-a4b-Q4_K_M.gguf -p "Hello"
```

## Usage (Ollama)

```bash
ollama run reinhardbit/casellm-26b-a4b-evaluation
```