Tinghong Ye

HFD研究持续交互中的人形机器人全双工行为：接收语音与可选的图像/视频，连续生成动作，并在视觉问答时生成语音回答。 任务包括 turn-taking、interruption（含stop）、backchannel、hesitation、pause。视觉问答叠加到这些交互任务中，不另设独立的SWA任务。代码中保留的swa名称用于模块与历史配置兼容。 直接执行bash inferrenderauto.sh会选择兼容的最新run并持续监听。 该入口接替原inferrenderautomultiturn.sh；任务集合、train/val样例、多卡队列、 checkpoint间隔和异步渲染逻辑不变。无需另起内部worker，也不要同时启动两份自动推理。 训练环境复用.venv-psi。依赖、模型和数据路径必须可用，启动器会检查；不要把配置打印成功当成完整训练验证。关键参数在训练脚本顶部，固定配置与环境变量映射在当前配方。完整续训与WandB续接参数见训练脚本末尾说明。 backup/preparefullduplexcontinuousaudio.sh用于连续背景音数据重建（旧流程还会导出hidden）， backup/prepareomnitrainingdata.sh及其调用的backup/cachevideoomniembeds.py 不要为清理目录而重新执行造数据。旧canonical-motion消融、旧OSS索引预构建、 早期hidden包装入口已退役，恢复方式见清理记录。…

Contributors: Songlin Wei, Hongyi Jing, Boqian Li, Zhenyu Zhao, Jiageng Mao, Zhenhao Ni, Sicheng He, Jie Liu, Xiawei Liu, Kaidi Kang, Sheng Zang,Weiduo Yuan, Marco Pavone, Di Huang, Yue Wang $\Psi0$ is an open vision-language-action (VLA) model for dexterous humanoid loco-manipulation. Our model first learns task semantics and visual representation from large-scale human egocentic videos, and then is post-trained on a smaller amount of real-world teleoperated robot data, to learn general dynamics of the embodiment. Our foundation model is capable of acquiring new long-horizontal dexterous loco-manipulation skill by fine-tuning using as few as 80 trajectories. Our key finding is that scaling…