Craftax prep for 5.0 release

config/craftax.ini

@@ -2,28 +2,37 @@
 env_name = craftax
 
 [vec]
-total_agents = 16384
-num_buffers = 16
-num_threads = 16
+total_agents = 8192
+num_buffers = 4
+num_threads = 8
 
 [env]
-seed_offset = 0
-# Pre-generated world pool. Each reset memcpys from a pool entry
-# instead of re-running generate_world (~ms -> ~us per reset).
-# Bounds world diversity: at most reset_pool_size unique maps are
-# ever seen per process. Set to 0 to disable (required for the
-# parity harness to maintain exact per-seed determinism).
 reset_pool_size = 1024
+action_mask = 1
+
+[policy]
+hidden_size = 1024
+num_layers = 3.95689535
 
 [train]
-total_timesteps = 20_000_000_000
-learning_rate = 0.008
-ent_coef = 0.02
-gamma = 0.997
-gae_lambda = 0.95
-horizon = 128
-minibatch_size = 32768
+gpus = 1
+total_timesteps = 200_000_000
+learning_rate = 0.00212870375
+anneal_lr = 1
+min_lr_ratio = 0.0
+gamma = 0.980109215
+gae_lambda = 0.741049647
+replay_ratio = 1.59645867
+clip_coef = 0.361835271
+vf_coef = 3.47198796
+vf_clip_coef = 3.29685092
+max_grad_norm = 1.08277392
+ent_coef = 8.97720674e-05
+anneal_ent_coef = 0
+min_ent_coef_ratio = 0.1
+momentum = 0.965376794
+minibatch_size = 4096
+horizon = 32
 
-[policy]
-hidden_size = 32
-num_layers = 3
+[sweep]
+max_suggestion_cost = 7200

ocean/craftax_clean/constants.h → ocean/craftax/constants.h

@@ -29,6 +29,7 @@
 // Environment parameters
 #define DEFAULT_MAX_TIMESTEPS 100000
 #define DAY_LENGTH 300
+#define VISIBLE_LIGHT_THRESHOLD 12
 #define MAX_ATTRIBUTE 5
 #define MOB_DESPAWN_DISTANCE 14
 #define MONSTERS_KILLED_TO_CLEAR_LEVEL 8
@@ -447,11 +448,13 @@ static const DungeonConfig DUNGEON_LEVEL_CONFIGS[3] = {
 };
 
 
-// Rendering parameters. tiles.png is 16x16 RGBA tiles, row-major, 16 columns:
+// Rendering parameters. textures.png is 16x16 RGBA tiles, row-major, 16 columns:
 //   [0..36]  blocks  [37..41] player  [42..46] items  [47..49] generic mobs
 //   [50..53] arrows  [54..61] armour  [62..70] tools  [71..76] potions
 //   [77..79] HUD     [80..87] melee   [88..90] passive [91..98] ranged
 //   [99..102] projectiles
+//   [103..104] sword enchant  [105..106] arrow enchant
+//   [107..110] armour fire overlay  [111..114] armour ice overlay
 #define TEX_TILE_PX 16
 #define TEX_SHEET_COLS 16
 #define TEX_SCALE 3

ocean/craftax/craftax.c

@@ -1,40 +1,30 @@
-// Standalone viewer for Craftax (random-action policy).
-//
-// Build:
-//   ./build.sh craftax --cpu          # optimized
-//   ./build.sh craftax --debug        # debug with sanitizers
-// Run:
-//   ./craftax
-
-#define CRAFTAX_ENABLE_ENV_IMPL
 #include "craftax.h"
-#include "step_crafting.h"
-#include "step_update_mobs.h"
-#include "step_spawn_mobs.h"
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <time.h>
-
-int main(int argc, char** argv) {
-    uint64_t seed = (argc > 1) ? strtoull(argv[1], NULL, 10) : (uint64_t)time(NULL);
 
+int main(void) {
     Craftax env;
     memset(&env, 0, sizeof(env));
     env.num_agents = 1;
-    env.seed = seed;
-    env.rng = (uint32_t)seed;
-
-    // Minimal buffers for a single agent
-    env.agents[0].observations = calloc(CRAFTAX_OBS_SIZE, sizeof(float));
-    env.agents[0].actions = calloc(1, sizeof(float));
-    env.agents[0].rewards = calloc(1, sizeof(float));
-    env.agents[0].terminals = calloc(1, sizeof(float));
-
-    c_init(&env);
+    env.rng = 1;
+    env.seed = 1;
+    env.use_action_mask = 1;
+
+    env.agents[0].observations = (obs_t*)calloc(OBS_SIZE, sizeof(obs_t));
+    env.agents[0].actions = (float*)calloc(1, sizeof(float));
+    env.agents[0].rewards = (float*)calloc(1, sizeof(float));
+    env.agents[0].terminals = (float*)calloc(1, sizeof(float));
+    env.agents[0].action_mask = (unsigned char*)calloc(ATN_DIM, 1);
     puf_reset(&env);
+    env.agents[0].actions[0] = -1.0f;
 
+    puf_render(&env);
     while (!WindowShouldClose()) {
+        int action = key_to_action();
+        if (action < 0) {
+            env.agents[0].actions[0] = -1.0f;
+            puf_render(&env);
+            continue;
+        }
+        env.agents[0].actions[0] = (float)action;
         puf_step(&env);
         puf_render(&env);
     }
@@ -44,5 +34,6 @@ int main(int argc, char** argv) {
     free(env.agents[0].actions);
     free(env.agents[0].rewards);
     free(env.agents[0].terminals);
+    free(env.agents[0].action_mask);
     return 0;
 }