MCPcopy Create free account
hub / github.com/SpatialVLA/SpatialVLA / apply_obs_transform

Function apply_obs_transform

data/rlds.py:396–399  ·  view source on GitHub ↗
(fn: Callable[[Dict], Dict], frame: Dict)

Source from the content-addressed store, hash-verified

394 # Convenience wrapper that takes a function that operates on a non-chunked "observation" dict and applies
395 # it to the chunked "observation" dict as well as the non-chunked "task" dict
396 def apply_obs_transform(fn: Callable[[Dict], Dict], frame: Dict) -> Dict:
397 frame["task"] = fn(frame["task"])
398 frame["observation"] = dl.vmap(fn)(frame["observation"])
399 return frame
400
401 # Decode + resize images (and depth images)
402 dataset = dataset.frame_map(

Callers 1

augFunction · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected