EdgeRock-4B-LiteRT-LM-INT4

Public development preview for Google AI Edge Gallery device validation. The container includes the text decoder, tokenizer, ChatML prompt template and fixed-state executor metadata. The graph retains recurrent state, local attention and RMA reads and block-boundary commits. Full-model long-memory checks and Android device validation are in progress.

The model file is 2.635 GB (decimal file size, not runtime memory). Use the CPU backend in Gallery 1.0.19 / LiteRT-LM 0.18.0. Image and audio inputs are not included.

python -m pip install -r requirements.txt
python run_litert.py --model EdgeRock-4B-int4.litertlm --prompt 'Calculate 17 * 23. Give only the answer.'

Weights retain the EdgeRock Model License 1.0. Code retains Apache-2.0.