diff --git a/cpp/src/wholememory_ops/functions/embedding_cache_func.cuh b/cpp/src/wholememory_ops/functions/embedding_cache_func.cuh
index c46bf6e65..e3589285f 100644
--- a/cpp/src/wholememory_ops/functions/embedding_cache_func.cuh
+++ b/cpp/src/wholememory_ops/functions/embedding_cache_func.cuh
@@ -151,7 +151,6 @@ class CacheLineInfo {
   uint32_t lfu_count_;
 };
 
-
 template <typename NodeIDT>
 class CacheSetUpdater {
  public:
@@ -160,13 +159,12 @@ class CacheSetUpdater {
   static constexpr int kScaledCounterBits = 14;
 
  private:
-
   using warp_bq_t =
     raft::matrix::detail::select::warpsort::warp_sort_immediate<kCacheSetSize, false, int64_t, int>;
 
   static constexpr int WARP_SIZE  = 32;
   static constexpr int BLOCK_SIZE = kCacheSetSize;
-  static_assert(kCacheSetSize == WARP_SIZE,"only support CacheSetSize==32,and BLOCK_SIZE==32\n");
+  static_assert(kCacheSetSize == WARP_SIZE, "only support CacheSetSize==32,and BLOCK_SIZE==32\n");
 
  public:
   struct TempStorage {
@@ -242,7 +240,7 @@ class CacheSetUpdater {
     //        candidate_local_id_,
     //        has_local_id_count);
     int64_t candidate_lfu_count0 = -1;
-    int candidate_local_id0  = -1;
+    int candidate_local_id0      = -1;
     unsigned int match_flag;
     // match_flag = WarpMatchLocalIDPairSync(candidate_local_id_[0], cached_local_id);
     int64_t estimated_lfu_count = cache_line_info.LfuCountSync();
@@ -348,15 +346,14 @@ class CacheSetUpdater {
                                                TempStorage& temp_storage,
                                                int cached_local_id)
   {
-
     warp_bq_t warp_queue(kCacheSetSize);
     const int per_thread_lim = id_count + raft::laneId();
 
     int has_local_id_count = 0;
     for (int idx = threadIdx.x; idx < per_thread_lim; idx += BLOCK_SIZE) {
-      int local_id            = -1;
+      int local_id                = -1;
       int64_t candidate_lfu_count = -1;
-      int candidate_local_id  = -1;
+      int candidate_local_id      = -1;
       if (idx < id_count) {
         local_id            = gids != nullptr ? gids[idx] - cache_set_start_id : idx;
         candidate_lfu_count = cache_set_coverage_counter[local_id];
@@ -381,7 +378,6 @@ class CacheSetUpdater {
     }
     __syncthreads();
 
-
     return has_local_id_count;
   }
 };
diff --git a/cpp/tests/wholegraph_ops/wholegraph_csr_weighted_sample_without_replacement_tests.cu b/cpp/tests/wholegraph_ops/wholegraph_csr_weighted_sample_without_replacement_tests.cu
index fa8cd4f10..eac1723af 100644
--- a/cpp/tests/wholegraph_ops/wholegraph_csr_weighted_sample_without_replacement_tests.cu
+++ b/cpp/tests/wholegraph_ops/wholegraph_csr_weighted_sample_without_replacement_tests.cu
@@ -446,7 +446,7 @@ INSTANTIATE_TEST_SUITE_P(WholeGraphCSRWeightedSampleWithoutReplacementOpTests,
                                              .set_center_node_count(35)
                                              .set_graph_node_count(23289)
                                              .set_graph_edge_couont(689403),
-                                              WholeGraphCSRWeightedSampleWithoutReplacementTestParam()
+                                           WholeGraphCSRWeightedSampleWithoutReplacementTestParam()
                                              .set_memory_type(WHOLEMEMORY_MT_CONTINUOUS)
                                              .set_max_sample_count(300)
                                              .set_center_node_count(256)
diff --git a/python/pylibwholegraph/pylibwholegraph/tests/wholegraph_torch/ops/test_wholegraph_weighted_sample_without_replacement.py b/python/pylibwholegraph/pylibwholegraph/tests/wholegraph_torch/ops/test_wholegraph_weighted_sample_without_replacement.py
index 0f53044bd..138163a87 100644
--- a/python/pylibwholegraph/pylibwholegraph/tests/wholegraph_torch/ops/test_wholegraph_weighted_sample_without_replacement.py
+++ b/python/pylibwholegraph/pylibwholegraph/tests/wholegraph_torch/ops/test_wholegraph_weighted_sample_without_replacement.py
@@ -46,8 +46,7 @@ def host_weighted_sample_without_replacement_func(
     output_center_localid_tensor = torch.empty((total_sample_count,), dtype=torch.int32)
     output_edge_gid_tensor = torch.empty((total_sample_count,), dtype=torch.int64)
     center_nodes_count = center_nodes.size(0)
-    block_size = 128 if max_sample_count <=256 else 256
-
+    block_size = 128 if max_sample_count <= 256 else 256
 
     for i in range(center_nodes_count):
         node_id = center_nodes[i]
@@ -66,23 +65,25 @@ def host_weighted_sample_without_replacement_func(
             edge_weight_corresponding_ids = torch.tensor([], dtype=col_id_dtype)
             for j in range(block_size):
                 local_gidx = gidx + j
-                local_edge_weights = torch.tensor( [],dtype=csr_weight_dtype
-                )
+                local_edge_weights = torch.tensor([], dtype=csr_weight_dtype)
                 generated_edge_weight_count = 0
-                for id in range(j,neighbor_count,block_size):
+                for id in range(j, neighbor_count, block_size):
                     local_edge_weights = torch.cat(
-                    (
-                        local_edge_weights,
-                         torch.tensor([host_csr_weight_ptr[start + id]], dtype=csr_weight_dtype),
-                     )
+                        (
+                            local_edge_weights,
+                            torch.tensor(
+                                [host_csr_weight_ptr[start + id]],
+                                dtype=csr_weight_dtype,
+                            ),
+                        )
                     )
                     generated_edge_weight_count += 1
                     edge_weight_corresponding_ids = torch.cat(
-                            (
-                                edge_weight_corresponding_ids,
-                                torch.tensor([id], dtype=col_id_dtype),
-                            )
+                        (
+                            edge_weight_corresponding_ids,
+                            torch.tensor([id], dtype=col_id_dtype),
                         )
+                    )
                 random_values = (
                     wg_ops.generate_exponential_distribution_negative_float_cpu(
                         random_seed, local_gidx, generated_edge_weight_count