LFM2.5-230M — Apple Neural Engine artifacts
Prebuilt Apple Neural Engine artifacts for LiquidAI/LFM2.5-230M.
The artifacts are portable Core ML packages; Core ML compiles them for the specific device on first load.
Contents
Each directory is a complete, independently runnable bundle. Pick one.
| directory | weights | notes |
|---|---|---|
int8/ |
8-bit, linear per-channel | half the size, and faster |
fp16/ |
16-bit float | the uncompressed reference |
Both are validated. They differ very slightly in wording on some prompts; pick int8/ for size and
speed, fp16/ to stay closest to the upstream weights.
Compute is fp16 in both cases; the directories differ only in how the weights are stored. Each
carries a PRECISION.json describing exactly how it was produced.
Runs on both iPhone and Mac.
For model behaviour, capabilities, intended use, limitations, and license, see the upstream LiquidAI/LFM2.5-230M model card. This repository adds no model capabilities and changes no model behaviour.
- Downloads last month
- -