diff --git a/fbgemm_gpu/fbgemm_gpu/quantize_comm.py b/fbgemm_gpu/fbgemm_gpu/quantize_comm.py
index 2875494a2e..8535374b7b 100644
--- a/fbgemm_gpu/fbgemm_gpu/quantize_comm.py
+++ b/fbgemm_gpu/fbgemm_gpu/quantize_comm.py
@@ -13,13 +13,12 @@
 
 
 import logging
+from dataclasses import dataclass
 from typing import TypeVar
 
-_T = TypeVar("_T")
-
 import torch
 
-# fmt:skip
+import fbgemm_gpu.quantize.quantize_ops  # noqa: F401
 from fbgemm_gpu.quantize_utils import (
     bf16_to_fp32,
     fp16_to_fp32,
@@ -33,9 +32,8 @@
 )
 from fbgemm_gpu.split_embedding_configs import SparseType
 from torch.autograd.profiler import record_function  # usort:skip
-from dataclasses import dataclass
 
-import fbgemm_gpu.quantize.quantize_ops  # noqa F401
+_T = TypeVar("_T")
 
 logger: logging.Logger = logging.getLogger()