| --- |
| tags: |
| - uzu |
| - mirai |
| - 8-bit |
| library_name: uzu |
| license: other |
| license_name: lfm-open-license |
| license_link: https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct/blob/main/LICENSE |
| pipeline_tag: text-generation |
| base_model: |
| - LiquidAI/LFM2.5-1.2B-Instruct |
| base_model_relation: quantized |
| --- |
| |
| <div style="display:flex;align-items:center;justify-content:space-between;gap:20px;flex-wrap:wrap"> |
| <div style="flex:1;min-width:260px"> |
| <h1 style="margin:0 0 12px">Mirai's LFM2.5-1.2B-Instruct Large Quantization</h1> |
| <p style="margin:0">A large, high-quality quantization of LFM2.5-1.2B-Instruct prepared for efficient local inference on Apple silicon.</p> |
| </div> |
| <img src="https://assets.trymirai.com/images/logo/ml_small_logo.svg" alt="Mirai Labs" width="64" height="64" style="width:80px;height:80px;object-fit:contain;flex:0 0 auto"> |
| </div> |
| |
|  |
|
|
| Mirai-L is on the size-KL Pareto frontier: we found no checkpoint that is smaller while also having lower KL divergence. |
|
|
| *Evaluation data mixture: 45% public agentic, 30% public SFT/long-context, 25% private chat data.* |
|
|
| ### Quickstart |
|
|
| If you are on macOS, the easiest way is to install the `mirai` Homebrew package and then run the CLI: |
|
|
| ```sh |
| brew install mirai |
| mirai --model trymirai/LFM2.5-1.2B-Instruct-L |
| ``` |
|
|
| Currently only Apple silicon inference is supported. If you want to build things from source, read this [overview](https://github.com/trymirai/uzu/blob/how-to/docs/how-to-run-uzu.md). |
|
|
| <div style="display:flex;flex-wrap:wrap;gap:12px;align-items:center;margin-top:32px"> |
| <a href="https://trymirai.com/local-models/liquidai-lfm2-5-1-2b-instruct-mirai-mirai-l-8" style="display:inline-flex;align-items:center;justify-content:space-between;width:220px;min-height:53px;padding:0 20px;box-sizing:border-box;border:1px solid #cccccc;border-radius:0;background:#ffffff;color:#3d3d3d;text-decoration:none;font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Inter,sans-serif;font-size:15px;font-weight:500;line-height:1.5"> |
| <span>Learn more</span> |
| <span aria-hidden="true" style="font-size:16px;line-height:1">→</span> |
| </a> |
| </div> |
| |
| --- |
|
|
| ### Method |
|
|
| Mirai Large uses 8-bit symmetric integer quantization with bfloat16 scales and group size 64. Block-diagonal Random Hadamard Transforms with block size 32 are used to reduce activation and weight outliers. The checkpoint was prepared with post-training quantization. |
|
|
|
|
|
|
| ### Citation |
|
|
| If you find our work helpful, feel free to give us a cite. |
|
|
| ```bibtex |
| @misc{mirai-quant, |
| title = {{Mirai Quantization}: Redefining the speed-quality frontier for local LLMs on Apple silicon}, |
| author = {Artur Chakhvadze and Ryan Mathieu and Roman Knyazhitskiy and Nikolai Voinilenko and Chen-Chen Yeh and Artur Mullakhmetov and Eugene Bokhan and others}, |
| note = {In collaboration with others at Mirai Labs}, |
| month = {June}, |
| year = {2026}, |
| url = {https://trymirai.com/blog/quantization} |
| } |
| ``` |
|
|
| ### Original model |
|
|
| This is a quantized version of [LiquidAI/LFM2.5-1.2B-Instruct](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct). For architecture details, intended use, evaluations, and limitations, see the [original model card](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct/blob/main/README.md). |
|
|
