mirror of
https://github.com/huggingface/lerobot.git
synced 2026-07-17 23:11:45 +00:00
add reward output
This commit is contained in:
@@ -91,8 +91,8 @@ class SARMConfig(PreTrainedConfig):
|
|||||||
"state_features": PolicyFeature(shape=(9, 14), type=FeatureType.STATE) # Example: 7 DOF × 2 arms
|
"state_features": PolicyFeature(shape=(9, 14), type=FeatureType.STATE) # Example: 7 DOF × 2 arms
|
||||||
})
|
})
|
||||||
output_features: dict = field(default_factory=lambda: {
|
output_features: dict = field(default_factory=lambda: {
|
||||||
"stage": PolicyFeature(shape=(1,), type=FeatureType.OTHER),
|
"stage": PolicyFeature(shape=(1,), type=FeatureType.REWARD),
|
||||||
"progress": PolicyFeature(shape=(1,), type=FeatureType.OTHER)
|
"progress": PolicyFeature(shape=(1,), type=FeatureType.REWARD)
|
||||||
})
|
})
|
||||||
|
|
||||||
def __post_init__(self):
|
def __post_init__(self):
|
||||||
|
|||||||
Reference in New Issue
Block a user