作者为 Reachy Mini 打造开源实时动作生成框架
作者为 Reachy Mini 机器人打造了开源的实时动作生成框架，用 Astra 合成超过 10000 条动作数据集，微调 Qwen 3.5 4B，再训练一个流匹配 Transformer 头，把低于 1Hz 的稀疏动作规划转成 25Hz 的密集动作。
I just made the most expressive real-time motion harness for the Reachy Mini.
It was created by:
> synthesizing a massive motion dataset using Astra (over 10000 samples)
> fine-tuning Qwen 3.5 4B using such dataset
> then training a flow matching transformer head to transform sparse motion plan (<1hz) into dense expressive motion (25Hz)
The average inference time for a full motion is less than 200ms, enabling real-time interaction.
@pollenrobotics and @huggingface only recorded 85 emotions for the reachy mini. With my work, you can now prompt and animate any motion you want.
Everything is open-source; write-up and links attached below. Here is a video of Reachy explaining itself through my harness.
来源：pham_blnh · x.com