diff --git a/onnxruntime/core/framework/prepacked_weights_container.h b/onnxruntime/core/framework/prepacked_weights_container.h
index a0880239bba5f..394dfdc5a93eb 100644
--- a/onnxruntime/core/framework/prepacked_weights_container.h
+++ b/onnxruntime/core/framework/prepacked_weights_container.h
@@ -82,7 +82,8 @@ class PrepackedWeightsContainer final {
 ///
 /// If saving is OFF, it is used to contain the weights memory mapped from disk.
 /// Those weights are then moved to the shared container if weight sharing is enabled.
-/// And also the interested kernels.
+/// If x-session weight sharing is not enabled, the weights are stored in this container,
+/// and shared with the interested kernels.
 /// </summary>
 class PrepackedForSerialization final {
  public:
@@ -91,11 +92,11 @@ class PrepackedForSerialization final {
 
   ORT_DISALLOW_COPY_ASSIGNMENT_AND_MOVE(PrepackedForSerialization);
 
+  // Maps a pre-packed weight blob key to PrepackedWeights instance
   using KeyToBlobMap = std::unordered_map<std::string, PrePackedWeights>;
 
-  // Maps weight name to iterators in key_to_blobs_. It associates a weight name with its pre-packs.
-  // Normally, a single weight produces a single PrePackedWeights. But it is possible that a weight
-  // is pre-packed by different kernels.
+  // WeightToPrePacksMap maps weight name to a set of pre-packed
+  // keys contained in the KeyToBlobMap
   using KeysPerWeight = std::unordered_set<std::string>;  // blob keys
   using WeightToPrePacksMap = std::unordered_map<std::string, KeysPerWeight>;
 
@@ -138,7 +139,7 @@ class PrepackedForSerialization final {
     // The function would add or replace existing entry with references to it.
     // If the entry is present, it would replace it with references to the existing entry.
     // If the entry is not present, it would add reference to refer_if_absent
-    // If present it would return the existing entry otherwise std::nullopt
+    // If the entry is present it would return the existing entry otherwise std::nullopt
     std::optional<PrePackedWeights> ReplaceWithReferenceIfSaving(const std::string& weight_name,
                                                                  const std::string& key,
                                                                  const PrePackedWeights& refer_if_absent);
diff --git a/onnxruntime/core/framework/session_state.cc b/onnxruntime/core/framework/session_state.cc
index 5edea481ba7f0..16d2e53732dde 100644
--- a/onnxruntime/core/framework/session_state.cc
+++ b/onnxruntime/core/framework/session_state.cc
@@ -495,7 +495,7 @@ Status SessionState::PrepackConstantInitializedTensors(
                       // everybody can share the same memory mapped entry
                       // the shared container takes ownership of the memory mapped entries
 
-                      // The next like replaces the existing entry with references to it
+                      // The next line replaces the existing entry with references to it
                       // and returns the container that holds the memory mapped entries
                       // so we can transfer it to shared container.
                       // if there is not an entry, we replace it with references to weights_to_be_filled_in
diff --git a/onnxruntime/core/framework/tensor_external_data_info.h b/onnxruntime/core/framework/tensor_external_data_info.h
index badd2137f3472..110c1122868da 100644
--- a/onnxruntime/core/framework/tensor_external_data_info.h
+++ b/onnxruntime/core/framework/tensor_external_data_info.h
@@ -2,14 +2,15 @@
 // Licensed under the MIT License.
 #pragma once
 
+#include <cmath>
 #include <filesystem>
 #include <ostream>
 #include <string>
 #include <tuple>
 
 #include <core/common/inlined_containers_fwd.h>
-#include "core/common/status.h"
 #include "core/common/path_string.h"
+#include "core/common/status.h"
 #include "core/framework/prepacked_weights_container.h"
 #include "core/graph/onnx_protobuf.h"
 
diff --git a/onnxruntime/core/graph/graph.cc b/onnxruntime/core/graph/graph.cc
index 758c717045537..e2e9987fc8940 100644
--- a/onnxruntime/core/graph/graph.cc
+++ b/onnxruntime/core/graph/graph.cc
@@ -4209,7 +4209,7 @@ Status Graph::ToGraphProtoWithExternalInitiallizersImpl(
         }
 
         if (!blob_keys_to_external_data.empty()) {
-          ORT_RETURN_IF_NOT(ExternalDataInfo::WritePrepackedToFileAndAddToProto(
+          ORT_RETURN_IF_NOT(!!ExternalDataInfo::WritePrepackedToFileAndAddToProto(
               *model_saving_options.prepacked_for_save, blob_keys_to_external_data,
               model_saving_options.align_offset,
               model_saving_options.allocation_granularity,
diff --git a/onnxruntime/test/framework/session_state_test.cc b/onnxruntime/test/framework/session_state_test.cc
index a7dcddfe61c00..aa01772d6fdd3 100644
--- a/onnxruntime/test/framework/session_state_test.cc
+++ b/onnxruntime/test/framework/session_state_test.cc
@@ -75,10 +75,10 @@ void PrepackedForSerialization::Subgraph::TestHarness<InspectLoadedSharedPrepack
 
     // We are expecting to load only one shared pre-packed weight
     // Saving is off, so we should not have any pre-packed weights for writing
-    const size_t expected_prepacks_for_writing = 0U;
+    constexpr const size_t expected_prepacks_for_writing = 0U;
     ASSERT_EQ(expected_prepacks_for_writing, sub.GetNumberOfWeightsForWriting());
 
-    const size_t expected_blobs_for_writing = 0U;
+    constexpr const size_t expected_blobs_for_writing = 0U;
     ASSERT_EQ(expected_blobs_for_writing, sub.GetNumberOfKeyedBlobsForWriting());
 
     ASSERT_EQ(0U, key_to_blobs_.size());
diff --git a/onnxruntime/test/framework/tensorutils_test.cc b/onnxruntime/test/framework/tensorutils_test.cc
index 6372756f5b1f8..5251b66edfe33 100644
--- a/onnxruntime/test/framework/tensorutils_test.cc
+++ b/onnxruntime/test/framework/tensorutils_test.cc
@@ -26,8 +26,8 @@ namespace test {
 TEST(TensorProtoUtilsTest, SetExternalDataInformation) {
   ONNX_NAMESPACE::TensorProto tensor_proto;
   const std::filesystem::path kExternalDataPath("test.bin");
-  const int64_t init_offset = 100;
-  const size_t init_length = 200;
+  constexpr const int64_t init_offset = 100;
+  constexpr const size_t init_length = 200;
 
   ExternalDataInfo::SetExternalLocationToProto(kExternalDataPath, init_offset, init_length, tensor_proto);
 
@@ -43,11 +43,11 @@ TEST(TensorProtoUtilsTest, SetExternalDataInformation) {
   PrepackedForSerialization prepacked_for_serialization;
   prepacked_for_serialization.SetSaveMode(true);
   PrePackedWeights prepacked_weights;
-  constexpr size_t buffer_size = 100;
   const std::string init_name = "test_initializer";
   const std::string blob_key = "test_key";
 
   std::array<float, 2> kData = {1.2345f, 2.4690f};
+  const size_t buffer_size = kData.size() * sizeof(float);
 
   prepacked_weights.buffers_.push_back(BufferUniquePtr(kData.data(), BufferDeleter(nullptr)));
   prepacked_weights.buffer_sizes_.push_back(buffer_size);
@@ -57,7 +57,7 @@ TEST(TensorProtoUtilsTest, SetExternalDataInformation) {
 
   prepacked_for_serialization.MainGraph().WritePacked(init_name, blob_key, std::move(prepacked_weights));
 
-  const int64_t starting_offset = 300;
+  constexpr const int64_t starting_offset = 300;
   int64_t external_offset = starting_offset;
   std::stringstream ss;
   const auto* blobs_for_weight = prepacked_for_serialization.MainGraph().GetBlobsForWeight(init_name);
diff --git a/orttraining/orttraining/training_api/module.cc b/orttraining/orttraining/training_api/module.cc
index 8f2d0f6531500..60708b05626c5 100644
--- a/orttraining/orttraining/training_api/module.cc
+++ b/orttraining/orttraining/training_api/module.cc
@@ -11,6 +11,7 @@
 #include "core/session/inference_session.h"
 #include "core/session/environment.h"
 #include "core/session/onnxruntime_session_options_config_keys.h"
+#include "core/graph/model_saving_options.h"
 #include "core/graph/graph_utils.h"
 
 #include "orttraining/training_api/checkpoint.h"