| --- |
| license: apache-2.0 |
| language: |
| - en |
| - zh |
| base_model: |
| - openbmb/MiniCPM5-2B |
| pipeline_tag: text-generation |
| library_name: litert |
| tags: |
| - minicpm |
| - minicpm5 |
| - litert |
| - tflite |
| - on-device |
| - edge-ai |
| --- |
| |
| # MiniCPM5-2B (LiteRT-LM) |
|
|
| This repository hosts a [**LiteRT-LM**](https://ai.google.dev/edge/litert-lm) (LiteRT, formerly TensorFlow Lite) conversion of **[openbmb/MiniCPM5-2B](https://huggingface.co/openbmb/MiniCPM5-2B)**, optimized for fully on-device inference on mobile and edge hardware. |
|
|
| The model was re-authored with `ai-edge-torch` (Llama-architecture decoder, since MiniCPM5-2B ships as a standard `LlamaForCausalLM`), verified for numerical parity against the original HF checkpoint (cosine similarity 0.999995 on last-token logits), converted to a multi-signature (`prefill_8/64/128/256/512/1024` + `decode`) `.tflite` with INT8 dynamic weight-only quantization, and packaged into `.litertlm` with `litert-lm-builder`. |
|
|
| ## Available Files |
|
|
| * **`MiniCPM5-2B_wi8_afp32.litertlm`**: dynamic weight-only INT8 quantization (`wi8`) with FP32 activations (`afp32`). KV cache buffer length 1280 tokens. |
|
|
| ## Model Information |
|
|
| | Item | Value | |
| | --- | --- | |
| | Type | Causal Language Model | |
| | Architecture | Standard `LlamaForCausalLM` | |
| | Layers | 42 | |
| | Hidden size | 2048 | |
| | Attention Heads (GQA) | 16 (Q) / 2 (KV), head_dim 128 | |
| | Vocab size | 130,560 | |
| | Context Length (native) | 131,072 | |
| | Exported KV cache length | 1,280 | |
| |
| ## Verification |
| |
| - **Numerical parity**: re-authored `ai-edge-torch` model vs. original HF `transformers` forward pass — cosine similarity **0.999995** on last-token logits (identical top-5 token predictions). |
| - **End-to-end generation**: verified with `litert-lm run`, including correct `<think>...</think>` reasoning-channel parsing and coherent completions (factual QA + code generation tested). |
| |
| ## Use the model |
| |
| ### Desktop/CLI |
| |
| Install `uv` and run the model directly from the LiteRT-LM command line: |
| |
| ```bash |
| uv tool install litert-lm |
| uvx litert-lm run --from-huggingface-repo=luispoveda93/MiniCPM5-2B-litert-lm MiniCPM5-2B_wi8_afp32.litertlm --prompt="What is the capital of France?" |
| ``` |
| |
| ### Edge Gallery App (Android) |
| |
| 1. Install the [app](https://play.google.com/store/apps/details?id=com.google.ai.edge.gallery) from Google Play or the [GitHub releases page](https://github.com/google-ai-edge/gallery/releases). |
| 2. In **Model manager**, tap **+ → Import from HF**, paste this repo's `.litertlm` URL, and download. |
| |
| ## Links |
| |
| - 🤗 Original model (BF16): [openbmb/MiniCPM5-2B](https://huggingface.co/openbmb/MiniCPM5-2B) |
| - 📦 GitHub: [OpenBMB/MiniCPM](https://github.com/OpenBMB/MiniCPM) |
| - 🛠️ LiteRT docs: [ai.google.dev/edge/litert](https://ai.google.dev/edge/litert) |
| - Reference conversion for the smaller checkpoint: [litert-community/MiniCPM5-1B](https://huggingface.co/litert-community/MiniCPM5-1B) |
| |
| ## License |
| |
| Released under the **Apache-2.0 License**, consistent with the upstream [openbmb/MiniCPM5-2B](https://huggingface.co/openbmb/MiniCPM5-2B). |
| |
| ## Citation |
| |
| ```bibtex |
| @article{minicpm4, |
| title={MiniCPM4: Ultra-efficient LLMs on end devices}, |
| author={MiniCPM, Team}, |
| journal={arXiv preprint arXiv:2506.07900}, |
| year={2025} |
| } |
| ``` |
| |
