refactor(policies): remove reward model branches from policy factory and __init__

src/lerobot/policies/__init__.py

@@ -12,11 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from lerobot.rewards.classifier.configuration_classifier import (
-    RewardClassifierConfig as RewardClassifierConfig,
-)
-from lerobot.rewards.sarm.configuration_sarm import SARMConfig as SARMConfig
-
 from .act.configuration_act import ACTConfig as ACTConfig
 from .diffusion.configuration_diffusion import DiffusionConfig as DiffusionConfig
 from .factory import get_policy_class, make_policy, make_policy_config, make_pre_post_processors
@@ -48,9 +43,7 @@ __all__ = [
     "PI0Config",
     "PI0FastConfig",
     "PI05Config",
-    "RewardClassifierConfig",
     "SACConfig",
-    "SARMConfig",
     "SmolVLAConfig",
     "TDMPCConfig",
     "VQBeTConfig",

src/lerobot/policies/factory.py

@@ -36,8 +36,6 @@ from lerobot.processor import (
     transition_to_batch,
     transition_to_policy_action,
 )
-from lerobot.rewards.classifier.configuration_classifier import RewardClassifierConfig
-from lerobot.rewards.sarm.configuration_sarm import SARMConfig
 from lerobot.types import PolicyAction
 from lerobot.utils.constants import (
     ACTION,
@@ -89,7 +87,7 @@ def get_policy_class(name: str) -> type[PreTrainedPolicy]:
 
     Args:
         name: The name of the policy. Supported names are "tdmpc", "diffusion", "act",
-            "multi_task_dit", "vqbet", "pi0", "pi05", "sac", "reward_classifier", "smolvla", "wall_x".
+            "multi_task_dit", "vqbet", "pi0", "pi05", "sac", "smolvla", "wall_x".
     Returns:
         The policy class corresponding to the given name.
 
@@ -132,18 +130,10 @@ def get_policy_class(name: str) -> type[PreTrainedPolicy]:
         from .sac.modeling_sac import SACPolicy
 
         return SACPolicy
-    elif name == "reward_classifier":
-        from lerobot.rewards.classifier.modeling_classifier import Classifier
-
-        return Classifier
     elif name == "smolvla":
         from .smolvla.modeling_smolvla import SmolVLAPolicy
 
         return SmolVLAPolicy
-    elif name == "sarm":
-        from lerobot.rewards.sarm.modeling_sarm import SARMRewardModel
-
-        return SARMRewardModel
     elif name == "groot":
         from .groot.modeling_groot import GrootPolicy
 
@@ -173,7 +163,7 @@ def make_policy_config(policy_type: str, **kwargs) -> PreTrainedConfig:
     Args:
         policy_type: The type of the policy. Supported types include "tdmpc",
                      "multi_task_dit", "diffusion", "act", "vqbet", "pi0", "pi05", "sac",
-                     "smolvla", "reward_classifier", "wall_x".
+                     "smolvla", "wall_x".
         **kwargs: Keyword arguments to be passed to the configuration class constructor.
 
     Returns:
@@ -376,14 +366,6 @@ def make_pre_post_processors(
             dataset_stats=kwargs.get("dataset_stats"),
         )
 
-    elif isinstance(policy_cfg, RewardClassifierConfig):
-        from lerobot.rewards.classifier.processor_classifier import make_classifier_processor
-
-        processors = make_classifier_processor(
-            config=policy_cfg,
-            dataset_stats=kwargs.get("dataset_stats"),
-        )
-
     elif isinstance(policy_cfg, SmolVLAConfig):
         from .smolvla.processor_smolvla import make_smolvla_pre_post_processors
 
@@ -392,15 +374,6 @@ def make_pre_post_processors(
             dataset_stats=kwargs.get("dataset_stats"),
         )
 
-    elif isinstance(policy_cfg, SARMConfig):
-        from lerobot.rewards.sarm.processor_sarm import make_sarm_pre_post_processors
-
-        processors = make_sarm_pre_post_processors(
-            config=policy_cfg,
-            dataset_stats=kwargs.get("dataset_stats"),
-            dataset_meta=kwargs.get("dataset_meta"),
-        )
-
     elif isinstance(policy_cfg, GrootConfig):
         from .groot.processor_groot import make_groot_pre_post_processors