feat(processors): Introduce processors for various policy types

- Added `make_processor` function to create processor instances for different policy types, including `tdmpc`, `diffusion`, `act`, `vqbet`, `pi0`, `pi0fast`, `sac`, and `reward_classifier`.
- Implemented corresponding processor files for each policy type, encapsulating normalization and unnormalization steps.
- Updated existing policies to remove direct normalization dependencies, enhancing modularity and clarity.
- Enhanced test coverage to validate the integration of new processors with existing policy configurations.

tests/policies/test_policies.py

@@ -39,6 +39,7 @@ from lerobot.policies.factory import (
     get_policy_class,
     make_policy,
     make_policy_config,
+    make_processor,
 )
 from lerobot.policies.normalize import Normalize, Unnormalize
 from lerobot.policies.pretrained import PreTrainedPolicy
@@ -151,6 +152,7 @@ def test_policy(ds_repo_id, env_name, env_kwargs, policy_name, policy_kwargs):
 
     # Check that we can make the policy object.
     dataset = make_dataset(train_cfg)
+    preprocessor, _ = make_processor(train_cfg.policy, None)
     policy = make_policy(train_cfg.policy, ds_meta=dataset.meta)
     assert isinstance(policy, PreTrainedPolicy)
 
@@ -224,6 +226,7 @@ def test_act_backbone_lr():
     assert cfg.policy.optimizer_lr_backbone == 0.001
 
     dataset = make_dataset(cfg)
+    preprocessor, _ = make_processor(cfg.policy, None)
     policy = make_policy(cfg.policy, ds_meta=dataset.meta)
     optimizer, _ = make_optimizer_and_scheduler(cfg, policy)
     assert len(optimizer.param_groups) == 2