Jim Fan 团队发布并开源 Flora,一套解析物理世界、把原始视频转为训练信号的开放模型,性能对标 Astra 而成本低 125 倍。Flora 已在其内部数据引擎中使用,每天用 500 张 GPU 标注 20 万小时视频,提供带时间戳的描述、用于 RL 后训练的密集进度奖励和真实世界评测的 rubric 打分。
We built a robotics data engine with Astra-level performance at 125x lower cost. And we are open-sourcing the weights today! Introducing Flora, a suite of open models to parse the physical world and turn raw footage into rich training signals.
Flora already powers our in-house data engine, labeling 200K hours of videos a day on 500 GPUs. We created it to scale the frontier for robot learning. Now you own the weights too.
Flora gives you:
- High-quality, timestamped captions to curate and filter human egocentric and robot data;
- Dense progress rewards for quality control and RL post-training;
- Rubrics scoring for real-world benchmarking
Try it today, links in thread:
来源:Jim Fan · x.com