**MedGemma 1.5 4B Instruct — medical AI for on-device inference** — converted for mobile and edge deployment by [DuoNeural](https://huggingface.co/DuoNeural).

- **Source model:** [google/medgemma-1.5-4b-it](https://huggingface.co/google/medgemma-1.5-4b-it)
- **Format:** GGUF Q4_K_M (llama.cpp-compatible)
- **File size:** 2490 MB
- **Quantization:** 4-bit K-mean (Q4_K_M) — excellent accuracy/size trade-off for edge devices
- **Target platforms:** Android, iOS, desktop edge inference
- **Converted:** 2026-05-06 06:03:33 by Archon / DuoNeural

## Usage

### llama.cpp (CLI)
```bash
./llama-cli -m medgemma-1.5-4b-it-LiteRT_Q4_K_M.gguf -n 512 --temp 0.7
```

### Google AI Edge / MediaPipe (Android/iOS)
This GGUF is compatible with [MLC-LLM](https://github.com/mlc-ai/mlc-llm) and [llama.cpp Android bindings](https://github.com/ggerganov/llama.cpp) for on-device inference. For use with [Google Edge Gallery](https://ai.google.dev/edge/gallery), convert to `.task` bundle using MediaPipe LLM conversion tools.

### Python via llama-cpp-python
```python
from llama_cpp import Llama

llm = Llama(
    model_path="medgemma-1.5-4b-it-LiteRT_Q4_K_M.gguf",
    n_ctx=2048,
    n_threads=4,
    verbose=False,
)

response = llm.create_chat_completion(
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Hello! How can you help me today?"},
    ]
)
print(response["choices"][0]["message"]["content"])
```

### Ollama
```bash
ollama run hf.co/DuoNeural/medgemma-1.5-4b-it-LiteRT
```

## About the Conversion

Converted using [llama.cpp](https://github.com/ggerganov/llama.cpp) GGUF pipeline with CUDA acceleration.
Source weights downloaded from HuggingFace, converted to F16 GGUF, then quantized to Q4_K_M.

---

## DuoNeural

**DuoNeural** is an open AI research lab — human + AI in collaboration.

| Platform | Link |
|----------|------|
| HuggingFace | [huggingface.co/DuoNeural](https://huggingface.co/DuoNeural) |
| Website | [duoneural.com](https://duoneural.com) |
| GitHub | [github.com/DuoNeural](https://github.com/DuoNeural) |
| X / Twitter | [@DuoNeural](https://x.com/DuoNeural) |
| Email | duoneural@proton.me |
| Newsletter | [duoneural.beehiiv.com](https://duoneural.beehiiv.com) |
| Support | [buymeacoffee.com/duoneural](https://buymeacoffee.com/duoneural) |

### DuoNeural Research Publications

| Title | DOI |
|-------|-----|
| [Nano-CTM: Ternary Continuous Thought Machines with Thought-Space Self-Prediction for Efficient Iterative Reasoning](https://doi.org/10.5281/zenodo.19775622) | [10.5281/zenodo.19775622](https://doi.org/10.5281/zenodo.19775622) |
| [Recurrence as World Model: CTM Learns Implicit Belief States in Partially Observable Physical Environments](https://doi.org/10.5281/zenodo.19810620) | [10.5281/zenodo.19810620](https://doi.org/10.5281/zenodo.19810620) |
| [Per-Object Slot Decomposition for Scalable Neural World Modeling: When Does Attention Beat Mean-Field?](https://doi.org/10.5281/zenodo.19846804) | [10.5281/zenodo.19846804](https://doi.org/10.5281/zenodo.19846804) |
| [The Dynamical Horizon Principle: CTM Gates Converge to the Predictability Limit of Dynamical Systems](https://doi.org/10.5281/zenodo.19952612) | [10.5281/zenodo.19952612](https://doi.org/10.5281/zenodo.19952612) |

*Open access, CC BY 4.0. Authored by Archon, Jesse Caldwell, Aura — DuoNeural.*

### Research Team
- **Jesse** — Vision, hardware, direction
- **Archon** — Lab Director, post-training, abliteration, experiments
- **Aura** — Research AI, literature synthesis, novel proposals

*Subscribe to the lab newsletter at [duoneural.beehiiv.com](https://duoneural.beehiiv.com) for model drops before they go anywhere else.*