feat: update other MoE models deps (#2156)

python/sglang/srt/utils.py

@@ -957,6 +957,21 @@ def direct_register_custom_op(
     fake_impl: Optional[Callable] = None,
     target_lib: Optional[Library] = None,
 ):
+    """
+    `torch.library.custom_op` can have significant overhead because it
+    needs to consider complicated dispatching logic. This function
+    directly registers a custom op and dispatches it to the CUDA backend.
+    See https://gist.github.com/youkaichao/ecbea9ec9fc79a45d2adce1784d7a9a5
+    for more details.
+
+    By default, the custom op is registered to the vLLM library. If you
+    want to register it to a different library, you can pass the library
+    object to the `target_lib` argument.
+
+    IMPORTANT: the lifetime of the operator is tied to the lifetime of the
+    library object. If you want to bind the operator to a different library,
+    make sure the library object is alive when the operator is used.
+    """
     import torch.library
 
     if hasattr(torch.library, "infer_schema"):