Published by TokForge for on-device inference.
TokForge
- Website: https://tokforge.ai
- Discord: https://discord.gg/Acv3CBtfVm
- Google Play: https://play.google.com/store/apps/details?id=dev.tokforge
- iOS TestFlight: https://testflight.apple.com/join/jnufjzRr
Runs on-device in the TokForge app.
gemma-4-E4B-it-MNN
Introduction
This model is a 4-bit quantized version of the MNN model exported from gemma-4-E4B-it using llmexport.
Usage
# clone MNN source
git clone https://github.com/alibaba/MNN.git
# compile
cd MNN
mkdir build && cd build
cmake .. -DMNN_LOW_MEMORY=true -DMNN_CPU_WEIGHT_DEQUANT_GEMM=true -DMNN_BUILD_LLM=true -DMNN_SUPPORT_TRANSFORMER_FUSE=true
make -j
# run
./llm_demo /path/to/gemma-4-E4B-it-MNN/config.json prompt.txt
Document
Credits
- Gemma 4 E4B IT by Google (Hugging Face). Used under the Apache-2.0 license.
- The vision and audio encoders and the tokenizer are byte-identical to the MNN conversion by the taobao-mnn team (Hugging Face). Used under the Apache-2.0 license.
- Converted with llmexport from MNN by Alibaba (Apache-2.0).