CMakeLists.txt

@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.25)
-project(COSC-4P80-Final-Project VERSION 0.0.23)
+project(COSC-4P80-Final-Project VERSION 0.0.15)
 
 option(ENABLE_ADDRSAN "Enable the address sanitizer" OFF)
 option(ENABLE_UBSAN "Enable the ub sanitizer" OFF)

graph.py

@@ -21,6 +21,9 @@ def plot_stacked_graph(title, output, csv_file1, csv_file2, position, position2)
     ax.plot(x1, y1, label=f"{csv_file1}")
     ax.plot(x2, y2, label=f"{csv_file2}")
 
+    ax.fill_between(x1, y1, alpha=0.5)
+    ax.fill_between(x2, y2, alpha=0.5)
+    
     if position < 2 ** 32:
         ax.axvline(x=position, color='red', linestyle='--')
         ax.text(position, ax.get_ylim()[1] * 0.95, f"Feed-forward average # of epochs", color='red', fontsize=10, ha='right', va='top', backgroundcolor='white')
@@ -31,13 +34,13 @@ def plot_stacked_graph(title, output, csv_file1, csv_file2, position, position2)
 
     ax.set_xlabel(x1_label)
     ax.set_ylabel(y1_label)
-    # ax.legend()
+    ax.legend()
     ax.set_title(title)
 
     plt.savefig(output)
 
 if __name__ == "__main__":
-    if len(sys.argv) != 7:
+    if len(sys.argv) != 5:
         print("Usage: python script.py <title> <output_file> <csv_file1> <csv_file2> <position_feed_forward> <position_deep>")
         sys.exit(1)
 
@@ -48,4 +51,4 @@ if __name__ == "__main__":
     position = sys.argv[5]
     position2 = sys.argv[6]
 
-    plot_stacked_graph(title, output, csv_file1, csv_file2, int(position), int(position2))
+    plot_stacked_graph(title, output, csv_file1, csv_file2, position, position2)

src/MNIST.cpp

@@ -312,7 +312,6 @@ namespace fp
         friend std::ofstream& operator<<(std::ofstream& file, const network_stats_t& stats)
         {
             file << stats.epoch_stats.size();
-            file << '\n';
             for (const auto& v : stats.epoch_stats)
                 file << v << "\n";
             return file;
@@ -322,7 +321,6 @@ namespace fp
         {
             blt::size_t size;
             file >> size;
-            file.ignore();
             for (blt::size_t i = 0; i < size; i++)
             {
                 stats.epoch_stats.emplace_back();
@@ -371,7 +369,6 @@ namespace fp
         friend std::ofstream& operator<<(std::ofstream& file, const network_average_stats_t& stats)
         {
             file << stats.run_stats.size();
-            file << '\n';
             for (const auto& v : stats.run_stats)
                 file << v << "---\n";
             return file;
@@ -381,7 +378,6 @@ namespace fp
         {
             blt::size_t size;
             file >> size;
-            file.ignore();
             for (blt::size_t i = 0; i < size; i++)
             {
                 stats.run_stats.emplace_back();
@@ -505,7 +501,7 @@ namespace fp
             if (break_flag)
             {
                 break_flag = false;
-                last_epoch = epochs + 1;
+                last_epoch = epochs;
                 break;
             }
             // dlib::serialize("mnist_network_" + ident + ".dat") << network;
@@ -560,29 +556,20 @@ namespace fp
             }
 
             state >> i;
-            state.ignore();
             blt::i64 load_epoch = 0;
             state >> load_epoch;
-            state.ignore();
             last_epoch = load_epoch;
             state >> stats;
-            state.ignore();
             blt::size_t test_stats_size = 0;
             state >> test_stats_size;
-            state.ignore();
             for (blt::size_t _ = 0; _ < test_stats_size; _++)
             {
                 test_stats.emplace_back();
                 state >> test_stats.back();
-                state.ignore();
+            }
         }
 
-            BLT_TRACE("Restoring at run %d with epoch %ld", i, load_epoch);
-            BLT_TRACE("\tRestored state size %lu", stats.run_stats.size());
-            BLT_TRACE("\tRestored test size %lu", test_stats_size);
-        }
-
-        blt::i64 last_epoch_save = last_epoch;
+        blt::i64 last_epoch_save = -1;
         for (; i < runs; i++)
         {
             if (stop_flag)
@@ -600,20 +587,10 @@ namespace fp
                 }
                 catch (dlib::serialization_error&)
                 {
-                    goto train_label;
+                    stats += train_network(local_ident, network);
                 }
             else
-            {
-                train_label:
-                auto stat = train_network(local_ident, network);
-                if (last_epoch_save > 0)
-                {
-                    // add in all the new epochs
-                    auto& vec = stats.run_stats.back();
-                    vec.epoch_stats.insert(vec.epoch_stats.end(), stat.epoch_stats.begin(), stat.epoch_stats.end());
-                } else
-                    stats += stat;
-            }
+                stats += train_network(local_ident, network);
             last_epoch_save = last_epoch;
             last_epoch = -1;
             test_stats.push_back(test_network(network));
@@ -631,27 +608,12 @@ namespace fp
             std::exit(-1);
         }
 
-        // user can skip this if required.
-        state << std::max(i - 1, 0);
-        state << '\n';
+        state << i;
         state << last_epoch_save;
-        state << '\n';
         state << stats;
-        state << '\n';
-        blt::i32 remove = 0;
-        if (stop_flag)
-            remove = 1;
-        state << static_cast<blt::size_t>(std::max(static_cast<blt::i64>(test_stats.size()) - remove, 0l));
-        state << '\n';
-        if (!test_stats.empty())
-        {
-            // the last test stat will be recalculated on restore. keeping it is an error.
-            for (const auto& v : blt::iterate(test_stats).take(test_stats.size() - remove))
-            {
+        state << test_stats.size();
+        for (const auto& v : test_stats)
             state << v;
-                state << '\n';
-            }
-        }
 
         return {stats, average};
     }
@@ -794,7 +756,6 @@ namespace fp
             average_epochs << average_forward_size << "," << average_deep_size << std::endl;
         }
 
-        BLT_INFO("Running python!");
         run_python_line_graph("Feed-Forward vs Deep Learning, Average Loss over Epochs", "epochs.png", path + "/forward_train_results.csv",
                               path + "/deep_train_results.csv", average_forward_size, average_deep_size);