| --- |
| base_model: litert-community/gemma-4-E2B-it-litert-lm |
| language: |
| - en |
| - hi |
| - ta |
| - te |
| - bn |
| - mr |
| - gu |
| - kn |
| - ml |
| - or |
| - pa |
| - ur |
| license: gemma |
| tags: |
| - gemma |
| - litert |
| - android |
| - on-device |
| - offline |
| - education |
| --- |
| |
| # Gemma 4 E2B — LiteRT-LM (EdSparkAI) |
|
|
| INT4-quantized `.litertlm` model file used by **EdSparkAI**, a fully offline Android AI tutor. |
|
|
| Based on [litert-community/gemma-4-E2B-it-litert-lm](https://huggingface.co/litert-community/gemma-4-E2B-it-litert-lm). |
|
|
| ## File |
|
|
| | | | |
| |---|---| |
| | **Filename** | `model.litertlm` | |
| | **Format** | LiteRT-LM (MediaPipe) | |
| | **Quantization** | INT4 | |
| | **Size** | ~2.4 GB | |
|
|
| ## Usage |
|
|
| This model runs on Android via **MediaPipe LiteRT 0.10.14** on the CPU backend. Used in [EdSparkAI](https://github.com/Destroyer1543/EdSparkAI) — an offline bilingual classroom assistant that explains textbook pages in 12 Indian languages. |
|
|
| ```kotlin |
| val cfg = EngineConfig( |
| modelPath = "/path/to/model.litertlm", |
| backend = Backend.CPU(), |
| maxNumTokens = 2048, |
| ) |
| val engine = Engine(cfg) |
| engine.initialize() |
| ``` |
|
|
| ### Device requirements |
| - Android 8.0+ (API 26+) |
| - 4 GB RAM minimum |
| - 3 GB free storage |
|
|
| ## License |
|
|
| Gemma is a trademark of Google LLC. Use is subject to the [Gemma Terms of Use](https://ai.google.dev/gemma/terms). |
|
|
