StarVLA open-sources a Lego-like codebase for VLA models, nearing 4,000 GitHub stars
wzenus · x · 2026-11-18
Jinhui Ye's team introduced StarVLA, an open-source (MIT) codebase for developing and experimenting with Vision-Language-Action models, already at nearly 4,000 GitHub stars with 60+ community projects built on it.
- Lego-like four-layer architecture: backbones (Qwen3-VL, QwenVL-MoE, InternVL, Florence-2, plus world models Cosmos-Predict2 and Wan2.2), action heads (FAST, OFT, π Flow, GR00T), training recipes (SFT, multimodal co-training, cross-embodiment, RL), and benchmarks (LIBERO, SimplerEnv, RoboTwin 2.0, BEHAVIOR-1K, CALVIN and 13 sim benchmarks total) all plug-and-play, sharing one trainer and policy server.
- Scale: 27 policy implementations, 10 robot embodiments, 41+ checkpoint repos, covering simulation and real robots.
- Users praise it as the clearest, easiest-to-use infra for VLA training.
More from Embodied
- VITURE launches $599 Phantom Beast XR glasses co-branded with Phantom Blade Zero — Scobleizer · 2026-11-18
- Museo, a whisperable AI ring for phone-free walks, ships next month with open dev platform — Scobleizer · 2026-11-18
- Claude Fable 5.2 tops LIBERO-10 LLM robot control benchmark, lifting success to 54% — YuXiang_IRVL · 2026-11-18
- RL-trained 5-tonne skid-steer loader balances on two wheels after sim training — Scobleizer · 2026-11-18
- Scale AI's FineART: 1,700-Hour Dual-Arm Robot Dataset Where Human Subtask Labels Enable Cross-Robot Transfer — yacinelearning · 2026-11-18
- Valkryn raises $17M led by a16z to build a Battle Lab for robotic warfare — espricewright · 2026-11-18