feat(pt): add HybridMuonOptimizer (#5149)

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added HybridMuon optimizer option with configurable hyperparameters
(momentum alias, lr_adjust, lr_adjust_coeff, min_2d_dim, muon_2d_only,
weight decay) and integrated training support (initialization, LR
tracking, scheduler compatibility, checkpoint resume).
* **Tests**
* Added comprehensive tests covering orthogonalization, optimizer step
behavior, weight-decay effects, fallback cases, LR-adjust modes, and
state save/load.

<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

Author: OutisLi

deepmd/pt/optimizer/__init__.py

@@ -2,11 +2,19 @@
 from .adamuon import (
     AdaMuonOptimizer,
 )
+from .hybrid_muon import (
+    HybridMuonOptimizer,
+)
 from .KFWrapper import (
     KFOptimizerWrapper,
 )
 from .LKF import (
     LKFOptimizer,
 )
 
-__all__ = ["AdaMuonOptimizer", "KFOptimizerWrapper", "LKFOptimizer"]
+__all__ = [
+    "AdaMuonOptimizer",
+    "HybridMuonOptimizer",
+    "KFOptimizerWrapper",
+    "LKFOptimizer",
+]