fix: broadcast ref/old model state_dict for FSDP

- Add broadcast_state_dict to sync state_dict from rank-0 to all ranks
- Fix create_ref_model returning None on non-rank-0 under FSDP
- Fix sync_old_model only updating old_model on rank-0 under FSDP
- Split skip_no_cuda/skip_no_kernel markers and hoist to top-level conftest
- Add distributed tests for broadcast_state_dict and create_ref_model

tests/helpers.py

@@ -165,6 +165,8 @@ class FakeTokenizer:
 class FakeExecutor:
     """Executor stub tracking ``sync_gradients`` and providing ``unwrap_model``."""
 
+    use_distributed = False
+
     def __init__(self, sync_gradients=True):
         self._sync_gradients = sync_gradients