refactor(processors): unify import statements by consolidating pipeline imports into the main processor module (#1845)

src/lerobot/processor/tokenizer_processor.py

@@ -11,14 +11,11 @@ import torch
 
 from lerobot.configs.types import FeatureType, PolicyFeature
 from lerobot.constants import OBS_LANGUAGE_ATTENTION_MASK, OBS_LANGUAGE_TOKENS
-from lerobot.processor.pipeline import (
-    EnvTransition,
-    ObservationProcessor,
-    ProcessorStepRegistry,
-    TransitionKey,
-)
 from lerobot.utils.import_utils import _transformers_available
 
+from .core import EnvTransition, TransitionKey
+from .pipeline import ObservationProcessor, ProcessorStepRegistry
+
 if TYPE_CHECKING or _transformers_available:
     from transformers import AutoTokenizer
 else: