load map changes to remove duplition on lots of inits and latest ini values

Author: Tim-phant

pufferlib/config/ocean/boxoban.ini

@@ -10,7 +10,7 @@ num_buffers = 8
 num_threads = 8 
 
 [policy]
-num_layers = 2 
+num_layers = 1 
 hidden_size = 256
 
 [env]
@@ -25,31 +25,33 @@ max_steps = 300
 
 
 [train]
-#EASY
 anneal_lr = 1
-beta1 = 0.9524080120922038
-beta2 = 0.9938151282804136
-clip_coef = 0.4770373231152628
-ent_coef = 0.00001
+beta1 = 0.9774372816193448
+beta2 = 0.9659403664380584
+clip_coef = 0.6046560670053024
+ent_coef = 0.00002079831529141607
 eps = 0.00000000000001
-gae_lambda = 0.8032589838049962
-gamma = 0.965698161834
+gae_lambda = 0.9258914518467392
+gamma = 0.9772998708784648
 gpus = 1
 horizon = 64
-learning_rate = 0.002557505571664044
-max_grad_norm = 0.1
+learning_rate = 0.004480255741933225
+max_grad_norm = 1.221684008665154
 min_lr_ratio = 0.37872027027338984
 minibatch_size = 8192
-prio_alpha = 0.9556021765385548
-prio_beta0 = 1
-replay_ratio = 4
+prio_alpha = 1
+prio_beta0 = 0.8789921736378042
+replay_ratio = 3.210300031048168
 seed = 42
-total_timesteps = 53251895
+total_timesteps = 55504884
 use_rnn = true
-vf_clip_coef = 4.856116827195164
-vf_coef = 5
-vtrace_c_clip = 2.5648427343944147
-vtrace_rho_clip = 3.258932255953648
+vf_clip_coef = 4.339748010438874
+vf_coef = 4.240274862679744
+vtrace_c_clip = 1.3625779006162615
+vtrace_rho_clip = 3.17260199042977
+
+
+#EASY
 
 [sweep]
 metric = perf

pufferlib/ocean/boxoban/boxoban.h

@@ -113,11 +113,15 @@ static inline const uint32_t get_random_puzzle_idx(const Boxoban *env) {
 
 
 void init (Boxoban* env) {
-    if (boxoban_configure_maps_from_env(env) != 0) {
-        fprintf(stderr, "Failed to configure Boxoban maps\n");
-        abort();
+    static int boxoban_maps_ready = 0;
+    if (!boxoban_maps_ready) {
+        if (boxoban_configure_maps_from_env(env) != 0) {
+            fprintf(stderr, "Failed to configure Boxoban maps\n");
+            abort();
+        }
+        ensure_map_loaded();
+        boxoban_maps_ready = 1;
     }
-    ensure_map_loaded();
     env->intermediate_rewards = calloc(env->size*env->size, sizeof(unsigned char));
     env->win = 0;
   }