diff --git a/fbgemm_gpu/codegen/genscript/jinja_environment.py b/fbgemm_gpu/codegen/genscript/jinja_environment.py
index d682208989..51cf32cbd6 100644
--- a/fbgemm_gpu/codegen/genscript/jinja_environment.py
+++ b/fbgemm_gpu/codegen/genscript/jinja_environment.py
@@ -295,13 +295,7 @@ def has_experimental_support(
     Check if the given combination of configs has TBE v2 support
     - TBE v2 does not support dense, nobag, vbe, is_index_select, is_rocm, and ssd
     """
-    return (
-        not dense
-        and not nobag
-        and not vbe
-        and not is_index_select
-        and not ssd
-    )
+    return not dense and not nobag and not vbe and not is_index_select and not ssd
 
 
 def is_valid_gwd_config(
diff --git a/fbgemm_gpu/include/fbgemm_gpu/utils/cuda_prelude.cuh b/fbgemm_gpu/include/fbgemm_gpu/utils/cuda_prelude.cuh
index 9612e13198..ec21076b3e 100644
--- a/fbgemm_gpu/include/fbgemm_gpu/utils/cuda_prelude.cuh
+++ b/fbgemm_gpu/include/fbgemm_gpu/utils/cuda_prelude.cuh
@@ -127,8 +127,9 @@ DEVICE_INLINE uint32_t ballot_sync(
 
 /// Sums a register value across all warp threads
 template <typename T, int ReduceWidth = kWarpSize>
-DEVICE_INLINE T
-warpReduceAllSum(T val, unsigned shfl_sync_mask = static_cast<unsigned>(kFullWarpMask)) {
+DEVICE_INLINE T warpReduceAllSum(
+    T val,
+    unsigned shfl_sync_mask = static_cast<unsigned>(kFullWarpMask)) {
 #pragma unroll
   for (int mask = ReduceWidth / 2; mask > 0; mask >>= 1) {
     val += shfl_xor(val, mask, ReduceWidth, shfl_sync_mask);