SAVRN
Search Contact SAVRN

Open-weight model · Any to any

RK182X-MLLM-Gemma-4-E2B-it

by RKNNAI RKNNAI/RK182X-MLLM-Gemma-4-E2B-it

RK182X-MLLM-Gemma-4-E2B-it is an open-weight model for any to any from RKNNAI, released under Apache License 2.0. Its published files total 15.9 GB.

本仓库提供由 google/gemma-4-E2B-it 转换得到的 RKNN MLLM 模型。 - Model ID:RKNNAI/RK182X-MLLM-Gemma-4-E2B-it - 模型显示名称:RK182X-MLLM-Gemma-4-E2B-it - 源模型:google/gemma-4-E2B-it - 模型类型:MLLM ModelScope 完整下载: Hugging Face 完整下载: ModelScope 指定配置下载: Hugging Face 指定配置下载: - 使用配套 RKNN…

Parameters—
Context—
Weights11.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—

Model Card

By RKNNAI, published under apache-2.0, revision c55977065f06.

本仓库提供由 google/gemma-4-E2B-it 转换得到的 RKNN MLLM 模型。 - Model ID:RKNNAI/RK182X-MLLM-Gemma-4-E2B-it - 模型显示名称:RK182X-MLLM-Gemma-4-E2B-it - 源模型:google/gemma-4-E2B-it - 模型类型:MLLM ModelScope 完整下载: Hugging Face 完整下载: ModelScope 指定配置下载: Hugging Face 指定配置下载: - 使用配套 RKNN Runtime 和驱动;运行前用 rknn-smi -v 检查设备端版本。 源模型许可证见根目录 LICENSE;同时遵守 RKNN Toolkit 和 RKNN Runtime 许可条款。

Read RKNNAI's full model card

1. 模型介绍

本仓库提供由 google/gemma-4-E2B-it 转换得到的 RKNN MLLM 模型。

  • Model ID:RKNNAI/RK182X-MLLM-Gemma-4-E2B-it
  • 模型显示名称:RK182X-MLLM-Gemma-4-E2B-it
  • 发布版本:v1.1.0
  • 源模型:google/gemma-4-E2B-it
  • 模型类型:MLLM
  • 芯片字段:RK182X
  • 具体支持芯片:RK1828

可用模型

发布版本 配置目录 支持芯片 分辨率 量化方式 NPU 核数 上下文长度(tokens) KVCache 音频 chunk 大小
v1.1.0 Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4 RK1828 Vision: 384x384 Vision: w4a16;Audio: w4a16;LLM: w4a16 Vision: 8;Audio: 8;LLM: 8 LLM: 102912(100.5k) LLM: int4 Audio: 756 mel_frames
v1.1.0 Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568 RK1828 Vision: 384x384 Vision: w4a16;Audio: w4a16;LLM: w4a16 Vision: 8;Audio: 8;LLM: 8 LLM: 45568(44.5k) LLM: fp16 Audio: 756 mel_frames

2. 文件说明

文件或目录 说明
LICENSE 源模型许可证
<配置目录>/ 配套模型文件
子目录 README.md 当前配置说明
子目录 config.json 模型配置与文件清单
子目录 SHA256SUMS 当前目录交付文件的 SHA-256 校验值(不包含自身)

3. 模型下载

ModelScope 完整下载:

modelscope download --model RKNNAI/RK182X-MLLM-Gemma-4-E2B-it --revision v1.1.0 --local_dir ./RK182X-MLLM-Gemma-4-E2B-it

Hugging Face 完整下载:

hf download RKNNAI/RK182X-MLLM-Gemma-4-E2B-it --revision v1.1.0 --local-dir ./RK182X-MLLM-Gemma-4-E2B-it

ModelScope 指定配置下载:

from modelscope import snapshot_download

snapshot_download(
    "RKNNAI/RK182X-MLLM-Gemma-4-E2B-it",
    revision="v1.1.0",
    allow_patterns=["README.md", "LICENSE", "Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/**"],
    local_dir="./RK182X-MLLM-Gemma-4-E2B-it",
)

Hugging Face 指定配置下载:

hf download RKNNAI/RK182X-MLLM-Gemma-4-E2B-it --revision v1.1.0 --include "README.md" "LICENSE" "Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/**" --local-dir ./RK182X-MLLM-Gemma-4-E2B-it

4. SHA-256 校验

在配置目录执行:

cd ./RK182X-MLLM-Gemma-4-E2B-it/Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4
sha256sum -c SHA256SUMS

所有条目显示 OK 后再部署。

5. 兼容性与限制

  • 支持芯片:RK1828。
  • 使用配套 RKNN Runtime 和驱动;运行前用 rknn-smi -v 检查设备端版本。
  • 所选配置目录内的文件须配套使用。

6. 版权与许可证

源模型许可证见根目录 LICENSE;同时遵守 RKNN Toolkit 和 RKNN Runtime 许可条款。

Identity and Version

Repository
RKNNAI/RK182X-MLLM-Gemma-4-E2B-it
Publisher
RKNNAI
Task
Any to any
Modality
Multimodal
Library
Not stated by the source
Parameters
Not stated by the source
Languages
Not stated by the source
Revision
c55977065f06b30f34d0a018c513c63a8bad4187
First published
2026-09-24
Last updated
2026-09-28

Files and Weights

35 files, 15.9 GB in total. The weights are 8 files totalling 11.5 GB in bin, gguf, safetensors.

Weights8 files · 11.5 GB
Configuration2 files · 3.7 KB
Documentation4 files · 20.1 KB
Other20 files · 4.4 GB
Repository1 file · 3.1 KB
Every file
FileTypeSizeSHA-256
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it.embed.binWeights805.3 MB c7a23c49fa7c
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it.safetensorsWeights316.1 MB bc43f37846bb
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it.tokenizer.ggufWeights15.8 MB 16ac32feeb42
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it_per_layer_inputs.embed.binWeights4.7 GB f2a3b25ebdf9
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it.embed.binWeights805.3 MB c7a23c49fa7c
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it.safetensorsWeights140.0 MB fd5ca76ed672
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it.tokenizer.ggufWeights15.8 MB 16ac32feeb42
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it_per_layer_inputs.embed.binWeights4.7 GB f2a3b25ebdf9
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/config.jsonConfiguration1.8 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/config.jsonConfiguration1.8 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/README.mdDocumentation2.8 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/README.mdDocumentation2.8 KB —
LICENSEDocumentation11.3 KB —
README.mdDocumentation3.2 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it-audio-8core.rknnOther10.7 MB 87d675534088
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it-audio-8core.weightOther626.8 MB cf37d0057d9b
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it.rknnOther25.1 MB cfda0389e5ac
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-it.weightOther1.4 GB 2fb157e30ba8
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-vision.rknnOther5.1 MB 53d11138fffd
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/Gemma-4-E2B-vision.weightOther91.4 MB 6066167823e1
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/SHA256SUMSOther1.4 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/audio_model_report_8core.htmlOther86.1 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/model_report.htmlOther528.7 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-102912-kv-int4/vision_model_report.htmlOther85.0 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it-audio-8core.rknnOther10.7 MB 87d675534088
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it-audio-8core.weightOther626.8 MB cf37d0057d9b
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it.rknnOther23.5 MB a97fc4e63a9f
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-it.weightOther1.4 GB 2fb157e30ba8
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-vision.rknnOther5.1 MB 53d11138fffd
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/Gemma-4-E2B-vision.weightOther91.4 MB 6066167823e1
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/SHA256SUMSOther1.4 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/audio_model_report_8core.htmlOther86.1 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/model_report.htmlOther507.9 KB —
Gemma-4-E2B-it-384x384-756frames-w4a16-8-45568/vision_model_report.htmlOther85.0 KB —
.gitattributesRepository3.1 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
11.5 GB
Download from RKNNAI

Released by RKNNAI through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published11.5 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About RK182X-MLLM-Gemma-4-E2B-it

Can I use RK182X-MLLM-Gemma-4-E2B-it commercially?

Yes. RK182X-MLLM-Gemma-4-E2B-it is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

Similar Models

Model · Any to any

gemma-4-12B-it-qat-GGUF

Unsloth AI

This model ships a Multi-Token Prediction drafter at the repo root (mtp-gemma-4-12B-it.gguf, a near-lossless smart Q40). A recent llama.cpp auto-discovers it from -hf, so you do not pass --model-draft: The drafter shares the target's KV cache and does not change the output (the target verifies every drafted token). See the MTP/ folder for the other precisions and explicit usage. Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context…

Open weights apache-2.0 transformers

Model · Any to any

gemma-4-E4B-it-GGUF

GGML Org

Run with https://llama.app - https://huggingface.co/google/gemma-4-E4B-it - https://huggingface.co/google/gemma-4-E4B-it-assistant - https://huggingface.co/google/gemma-4-E4B-it-qat-q40-unquantized-assistant - https://huggingface.co/google/gemma-4-E4B-it-qat-q40-unquantized - add info - add dflash

Open weights apache-2.0

Model · Any to any

gemma-4-12B-it-qat-q4_0-gguf

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 transformers

Model · Any to any

gemma-4-E4B-it-qat-q4_0-gguf

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0

Model · Any to any

gemma-4-E4B-it-qat-GGUF

Unsloth AI

This model ships a Multi-Token Prediction drafter at the repo root (mtp-gemma-4-E4B-it.gguf, a near-lossless smart Q40). A recent llama.cpp auto-discovers it from -hf, so you do not pass --model-draft: The drafter shares the target's KV cache and does not change the output (the target verifies every drafted token). See the MTP/ folder for the other precisions and explicit usage. Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context…

Open weights apache-2.0 131,072 tokens transformers

Model · Any to any

gemma-4-E2B-it-qat-GGUF

Unsloth AI

This model ships a Multi-Token Prediction drafter at the repo root (mtp-gemma-4-E2B-it.gguf, a near-lossless smart Q40). A recent llama.cpp auto-discovers it from -hf, so you do not pass --model-draft: The drafter shares the target's KV cache and does not change the output (the target verifies every drafted token). See the MTP/ folder for the other precisions and explicit usage. Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context…

Open weights apache-2.0 131,072 tokens transformers