Sakhanda Wire
NVDA $153.64 -1.60% MSFT $506.98 -2.13% GOOGL $279.95 -1.33% META $620.09 -1.01% AMZN $256.57 -0.46%
← Back to the news

Strands Robots links robot recording, training, and deployment

An AWS walkthrough presents a continuous robotics data loop built with Strands Robots, LeRobot, and Hugging Face Storage Buckets. The open-source Strands Robots SDK lets an agent record demonstrations, train a policy from the resulting dataset, and deploy that policy to a robot. Recordings remain in LeRobot’s on-disk format, which is already used by more than 90,000 datasets and models from over 8,000 publishers.

The workflow stores recordings in a mutable, non-versioned Storage Bucket backed by Xet. Byte-level, content-defined chunking means later synchronizations transfer only changed data rather than repeatedly uploading complete video and Parquet files. The same dataset can then be streamed from the Hub frame by frame, with camera video decoded during reading and no full local copy. A trained checkpoint is deployed to the same Robot() object by changing the mode to "real".

The article describes agent decisions such as selecting episodes and replacing checkpoints. The default simulation path uses a mock policy, producing a structurally valid dataset but not useful training behavior. Hardware recording and large-scale training require additional robot, authentication, and GPU setup.

Written by Sakhanda Wire AI from reporting by Hugging Face. Facts, quotes and full context are in the original article below.

Originally published by Hugging Face on

Read the original on Hugging Face ↗

The reporting and the image belong to Hugging Face and are credited and linked above; the text on this page is Sakhanda Wire's own summary.

← Back to the news

More stories

All the latest news