Merge pull request #530 from doofenshirmtz/drmario-env

dr mario env

Author: jsuarez5341

config/ocean/drmario.ini

@@ -0,0 +1,45 @@
+[base]
+env_name = drmario
+
+[vec]
+total_agents = 1536          
+num_buffers = 4
+num_threads = 4
+
+[env]
+n_rows = 16
+n_cols = 8
+n_init_viruses = 4
+
+[policy]
+hidden_size = 128            
+num_layers = 1
+
+[legacy]
+torch_deterministic = 1
+cpu_offload = 0
+compile = 1
+compile_fullgraph = 0
+
+[train]
+total_timesteps = 500_000_000
+learning_rate = 0.001
+gamma = 0.99
+gae_lambda = 0.95
+clip_coef = 0.2
+vf_coef = 0.5
+ent_coef = 0.01
+minibatch_size = 16384       
+horizon = 128
+use_rnn = 0                  
+
+#copied from Breakout
+beta1 = 0.9
+beta2 = 0.999
+eps = 1e-8
+max_grad_norm = 1.0
+replay_ratio = 2
+vtrace_rho_clip = 2.0
+vtrace_c_clip = 2.0
+prio_alpha = 0.6
+prio_beta0 = 0.9

ocean/drmario/binding.c

@@ -0,0 +1,25 @@
+#include "drmario.h"
+
+#define OBS_SIZE 133     
+#define NUM_ATNS 1       
+#define ACT_SIZES {7}     
+#define OBS_TENSOR_T FloatTensor
+
+#define Env DrMario
+#include "vecenv.h"
+
+void my_init(Env* env, Dict* kwargs) {
+    env->num_agents=1;
+    env->n_rows = dict_get(kwargs, "n_rows")->value;
+    env->n_cols = dict_get(kwargs, "n_cols")->value;
+    env->n_init_viruses = dict_get(kwargs, "n_init_viruses")->value;
+        c_init(env);
+}
+
+void my_log(Log* log, Dict* out) {
+    dict_set(out, "perf", log->perf);
+    dict_set(out, "score", log->score);
+    dict_set(out, "episode_return", log->episode_return);
+    dict_set(out, "episode_length", log->episode_length);
+    dict_set(out, "viruses_cleared", log->viruses_cleared);
+}