feat(train): parallel training framework — FSDP2, HSDP, gradient accumulation, and DCP checkpoints (#4010)

* feat(train): parallel training engine with FSDP2, HSDP, and DCP checkpoints

Replace the FSDP1 training path with a config-owned parallel-training
engine:

- Topology and runtime configs (--parallelism.*, --accelerator.*):
  dp_replicate x dp_shard degrees select single-process, DDP (unchanged
  default), FSDP2, or HSDP; mixed precision, first-class gradient
  accumulation, and FSDP/DDP tuning knobs are mirrored as plain
  dataclasses that build the accelerate objects at runtime, so every
  run is reproducible from its train_config.json alone. Accelerate env
  vars are guarded against configuring the engine behind the config
  system's back.
- Declarative policy surface: policies declare FSDP2 wrap units
  (_fsdp_wrap_modules) and non-forward entry points
  (_fsdp_forward_methods); a shared engine resolves them around
  accelerator.prepare(). Context-parallel fields are reserved and
  validated to 1.
- Checkpoints: selectable --checkpoint_format (safetensors | dcp |
  safetensors_dcp); the sharded optimizer channel is always DCP;
  two-phase resume (step+RNG before prepare, DCP model/optimizer after)
  reshards across GPU-topology changes; lerobot-convert-dcp merges DCP
  shards into a distributable model.safetensors offline.
- Publishing: PreTrainedPolicy.push_model_to_hub is replaced by the
  free publish_trained_model (model + processors + card + train config,
  all-ranks gather with main-rank writes);
  PreTrainedPolicy._save_pretrained gathers state dicts internally,
  removing the state_dict= threading from save_pretrained.
- lerobot_train is restructured around the engine: optimizer built
  before the single prepare() call, deferred weight load on DCP
  resumes, collective save_checkpoint with no call-site rank branches,
  dp-world-size-based sample accounting.

Breaking changes: FSDP checkpoints from lerobot <= 0.6.x are not
resumable (weights stay loadable via from_pretrained; pin
lerobot==0.6.x to finish old runs); the `accelerate launch
--config_file` yaml flow is superseded by the config flags; training
autocast is owned exclusively by --accelerator.mixed_precision
(policy.dtype only casts parameters).

Also fixes: reward-model hub publishing crash (TypeError on extra
kwargs).

Verified by ~200 new CPU tests (config round-trips, checkpoint
round-trips per format, two-phase resume, publisher contracts,
converter equivalence, accelerate canaries), a 5-test 4-GPU suite
(FSDP2 save/resume bit-exactness, HSDP/DDP loss parity,
changed-topology resume, all-ranks save_pretrained, grad-accum
equivalence), and end-to-end ACT (1/4/8 GPUs) + FastWAM 6B
(FSDP2 + HSDP) training runs.

tests/policies/test_policies.py

@@ -301,8 +301,12 @@ def test_save_and_load_pretrained(dummy_dataset_metadata, tmp_path, policy_name:
     torch.testing.assert_close(list(policy.parameters()), list(loaded_policy.parameters()), rtol=0, atol=0)
 
 
-def test_save_pretrained_with_state_dict(dummy_dataset_metadata, tmp_path):
-    """Exercise the FSDP checkpoint path: save_pretrained with a pre-gathered state_dict."""
+def test_save_pretrained_single_file_artifact(dummy_dataset_metadata, tmp_path):
+    """The distributable checkpoint is one unsharded safetensors file.
+
+    The former `state_dict=` variant of this test died with the #3810 save override: the
+    kwarg would now be silently swallowed by HubMixin's **push_to_hub_kwargs.
+    """
     policy_cls = get_policy_class("act")
     policy_cfg = make_policy_config("act")
     features = dataset_to_policy_features(dummy_dataset_metadata.features)
@@ -313,8 +317,8 @@ def test_save_pretrained_with_state_dict(dummy_dataset_metadata, tmp_path):
     policy = policy_cls(policy_cfg)
     policy.to(policy_cfg.device)
 
-    save_dir = tmp_path / "fsdp_state_dict"
-    policy.save_pretrained(save_dir, state_dict=policy.state_dict())
+    save_dir = tmp_path / "single_file_artifact"
+    policy.save_pretrained(save_dir)
 
     # A single, unsharded safetensors file (no sharded set + index).
     assert (save_dir / SAFETENSORS_SINGLE_FILE).is_file()