---
license: apache-2.0
base_model: google/gemma-4-E2B-it
pipeline_tag: text-generation
library_name: rkllm
tags:
- rkllm
- rockchip
- rk3576
- rk3588
- gemma
- gemma-4
---

# gemma-4-E2B-it-RKLLM

RKLLM-converted Gemma 4 E2B IT language-model artifacts for Rockchip RK3576 and RK3588 NPUs.

Important: the upstream Gemma 4 E2B IT model is multimodal/VLM-capable, but this repository contains only the RKLLM language-model binaries. The converted model exposed here is LLM-only: it does not accept images and does not include an RKNN vision encoder.

These hardware-specific `.rkllm` files require a compatible Rockchip RKLLM runtime. They are not Transformers checkpoints and cannot be loaded directly with Transformers, llama.cpp, or Ollama.

## Base model

- Upstream model: [google/gemma-4-E2B-it](https://huggingface.co/google/gemma-4-E2B-it)
- License: Apache-2.0
- Upstream type: VLM/multimodal
- RKLLM type: LLM-only

## Conversion and variants

### Toolkit version

**RKLLM Toolkit: v1.3.0**

Use a file built for the exact target SoC.

| Target | Quantization | File | SHA256 |
| --- | --- | --- | --- |
| RK3576 | W4A16 (g128) | [gemma-4-E2B-it_RK3576_w4a16_g128.rkllm](RK3576/gemma-4-E2B-it_RK3576_w4a16_g128.rkllm) | `973008fd7ebed81e2a26dcf7dcc1c28fde631837e3e021c7f3007d6e82647165` |
| RK3576 | W8A8 | [gemma-4-E2B-it_RK3576_w8a8.rkllm](RK3576/gemma-4-E2B-it_RK3576_w8a8.rkllm) | `b7357c01fcd51c8f896f06ee3a3746bc233031abd04e2fa0cebb5908b1382a01` |
| RK3588 | W8A8 | [gemma-4-E2B-it_RK3588_w8a8.rkllm](RK3588/gemma-4-E2B-it_RK3588_w8a8.rkllm) | `ae81d1e37eb54b40d2d366ef622a0c05acfb15537272e982efc1dfdd60557979` |

The repository also includes `gemma-4-E2B-it_data_quant.json`, used as calibration data during conversion.

## Usage

```bash
hf download HanzoHuang/gemma-4-E2B-it-RKLLM \
  RK3576/gemma-4-E2B-it_RK3576_w4a16_g128.rkllm \
  --local-dir gemma-4-E2B-it-RKLLM
```

Run the file with the RKLLM runtime and the upstream Gemma instruction prompt format. For Docker deployment, see [Hanzo-Huang/rkllm-docker](https://github.com/Hanzo-Huang/rkllm-docker).

## Limitations

Vision input is not supported by these artifacts. Conversion can change output quality relative to the upstream model; validate the result on your target device.

## Acknowledgements

Thanks to Google, Rockchip, and the RKLLM community.