SmolVLA: Efficient Vision-Language-Action Model trained on Lerobot Community Data

SmolVLA-450M is an open-source, compact Vision-Language-Action model for robotics that runs on consumer hardware. It is pretrained on publicly licensed community data and trainable on a single consumer GPU, with asynchronous inference boosting throughput and strong performance on simulation and real-world tasks.

Source

HuggingFace Blog

Ingéré

03 JUIN 2025 · 19:10

Score édito

4.0 / 5