Model Terminal

act_so101_keyboard_gray_cube_clean11

A task-specific ACT (Action Chunking with Transformers) imitation-learning policy checkpoint for the SO-101 robotic arm, trained via the Hugging Face LeRobot stack to perform a tabletop keyboard/gray-cube manipulation task. It outputs chunked action sequences from sensory input to control the arm through the learned cleaning behavior. Source: supabase.

Value score
Not scored
Context
tokens
Max output
tokens
Price

Capability radar

Identity

Developer
Timmy
Openness
Modalities
Release
Knowledge cutoff
Deprecation
API docs

Benchmarks

No benchmark scores yet.

Compare nearby

Pricing

Input / 1M
Output / 1M
Speed