2048

Author: Joseph Suarez

pufferlib/config/ocean/g2048.ini

@@ -1,24 +1,24 @@
 [base]
 package = ocean
 env_name = puffer_g2048
-policy_name = G2048
+policy_name = Policy
 rnn_name = Recurrent
 
 [policy]
-hidden_size = 256
+hidden_size = 128
 
 [rnn]
-input_size = 256
-hidden_size = 256
+input_size = 128
+hidden_size = 128
 
 [vec]
 num_envs = 4
 
 [env]
-num_envs = 4024
+num_envs = 4096
 
 [train]
-total_timesteps = 600_000_000
+total_timesteps = 5_000_000_000
 adam_beta1 = 0.9529488439604378
 adam_beta2 = 0.9993901829477296
 adam_eps = 2.745365927413118e-7
@@ -27,7 +27,8 @@ clip_coef = 0.596573170393339
 ent_coef = 0.02107417730003862
 gae_lambda = 0.9940613415815854
 gamma = 0.9889857974154952
-learning_rate = 0.0032402460796988127
+#learning_rate = 0.0032402460796988127
+learning_rate = 0.001
 max_grad_norm = 1.0752406726589745
 minibatch_size = 16384
 prio_alpha = 0.25297099593586336
@@ -36,4 +37,4 @@ vf_clip_coef = 0.1
 vf_coef = 1.6362878279900643
 vtrace_c_clip = 0
 vtrace_rho_clip = 1.2917509971869054
-anneal_lr = False
+anneal_lr = False

pufferlib/ocean/g2048/2048.h

@@ -44,19 +44,24 @@ typedef struct {
 } Game;
 
 // Precomputed color table for rendering optimization
+const Color PUFF_BACKGROUND = (Color){6, 24, 24, 255};
+const Color PUFF_WHITE = (Color){241, 241, 241, 241};
+const Color PUFF_RED = (Color){187, 0, 0, 255};
+const Color PUFF_CYAN = (Color){0, 187, 187, 255};
+
 static Color tile_colors[12] = {
-    {205, 193, 180, 255}, // Empty/background
-    {238, 228, 218, 255}, // 2
-    {237, 224, 200, 255}, // 4
-    {242, 177, 121, 255}, // 8
-    {245, 149, 99, 255},  // 16
-    {246, 124, 95, 255},  // 32
-    {246, 94, 59, 255},   // 64
-    {237, 207, 114, 255}, // 128
-    {237, 204, 97, 255},  // 256
-    {237, 200, 80, 255},  // 512
-    {237, 197, 63, 255},  // 1024
-    {237, 194, 46, 255}   // 2048+
+    {6, 24, 24, 255}, // Empty/background
+    {187, 187, 187, 255}, // 2
+    {170, 187, 187, 255}, // 4
+    {150, 187, 187, 255}, // 8
+    {130, 187, 187, 255},  // 16
+    {110, 187, 187, 255},  // 32
+    {90, 187, 187, 255},   // 64
+    {70, 187, 187, 255}, // 128
+    {50, 187, 187, 255},  // 256
+    {30, 187, 187, 255},  // 512
+    {10, 187, 187, 255},  // 1024
+    {0, 187, 187, 255}   // 2048+
 };
 
 // --- Logging ---
@@ -326,7 +331,7 @@ void c_render(Game* game) {
 
     if (!window_initialized) {
         InitWindow(px * SIZE, px * SIZE + 50, "2048");
-        SetTargetFPS(10); // Increased for smoother rendering
+        SetTargetFPS(30); // Increased for smoother rendering
         window_initialized = true;
     }
 
@@ -336,7 +341,7 @@ void c_render(Game* game) {
     }
 
     BeginDrawing();
-    ClearBackground(RAYWHITE);
+    ClearBackground(PUFF_BACKGROUND);
 
     // Draw grid
     for (int i = 0; i < SIZE; i++) {
@@ -354,14 +359,18 @@ void c_render(Game* game) {
                 int display_val = 1 << val; // Power of 2
                 // Pre-format text to avoid repeated formatting
                 snprintf(score_text, sizeof(score_text), "%d", display_val);
-                DrawText(score_text, j * px + 30, i * px + 40, 32, BLACK);
+                if (display_val < 1000) {
+                    DrawText(score_text, j * px + 30, i * px + 40, 32, PUFF_WHITE);
+                } else {
+                    DrawText(score_text, j * px + 20, i * px + 40, 32, PUFF_WHITE);
+                }
             }
         }
     }
 
     // Draw score (format once per frame)
     snprintf(score_text, sizeof(score_text), "Score: %d", 1 << game->score);
-    DrawText(score_text, 10, px * SIZE + 10, 24, DARKGRAY);
+    DrawText(score_text, 10, px * SIZE + 10, 24, PUFF_WHITE);
 
     EndDrawing();
 }
@@ -370,4 +379,4 @@ void c_close(Game* game) {
     if (IsWindowReady()) {
         CloseWindow();
     }
-}
+}

pufferlib/ocean/g2048/g2048.c

@@ -1,4 +1,5 @@
 #include "2048.h"
+#include "puffernet.h"
 #include <stdio.h>
 #include <stdlib.h>
 #include <termios.h>
@@ -18,25 +19,38 @@ int main() {
     env.actions = actions;
     env.rewards = rewards;
 
+    Weights* weights = load_weights("resources/g2048/g2048_weights.bin", 134917);
+    int logit_sizes[1] = {4};
+    LinearLSTM* net = make_linearlstm(weights, 1, 16, logit_sizes, 1);
     c_reset(&env);
 
     // Main game loop
+    int frame = 0;
     while (1) {
         c_render(&env);
+        frame++;
 
         int action = 0;
-        if (IsWindowReady()) {
+        if (IsKeyDown(KEY_LEFT_SHIFT)) {
             if (IsKeyPressed(KEY_W) || IsKeyPressed(KEY_UP)) action = UP;
             else if (IsKeyPressed(KEY_S) || IsKeyPressed(KEY_DOWN)) action = DOWN;
             else if (IsKeyPressed(KEY_A) || IsKeyPressed(KEY_LEFT)) action = LEFT;
             else if (IsKeyPressed(KEY_D) || IsKeyPressed(KEY_RIGHT)) action = RIGHT;
+            env.actions[0] = action - 1;
+        } else if (frame % 10 != 0) {
+            continue;
+        } else {
+            action = 1;
+            for (int i = 0; i < 16; i++) {
+                net->obs[i] = env.observations[i];
+            }
+            forward_linearlstm(net, net->obs, env.actions);
         }
 
         if (action != 0) {
-            env.actions[0] = action - 1;
             c_step(&env);
             if (!IsWindowReady()) {
-                print_grid(&env);
+                //print_grid(&env);
                 printf("Reward: %.0f\n", env.rewards[0]);
             }
         }
@@ -46,6 +60,7 @@ int main() {
         }
     }
 
+    free_linearlstm(net);
     c_close(&env);
     printf("Game Over! Final Max Tile: %d\n", env.score);
     return 0;