Merge pull request #623 from FinlaySanders/nethack

Nethack hypers

Author: jsuarez5341

config/nethack.ini

@@ -8,16 +8,16 @@ num_buffers = 4
 num_threads = 32
 
 [env]
-gold_coef = 0.03540707780083026
-exp_coef = 0.25886063598802184
-descent_coef = 1.8884272008758112
-xp_coef = 0.28073502137812123
-scout_coef = 0.01689042846824247
-hp_coef = 0.018973887012487405
-hunger_coef = 0.023586377138546992
-illegal_penalty = -0.006951422236819335
-death_penalty = -0.6335992931154887
-ac_coef = 0.01
+gold_coef = 0.021659979757862095
+exp_coef = 0.08341832601491733
+descent_coef = 1.9586073647172169
+xp_coef = 0.5926087261188399
+scout_coef = 0.07425725927605491
+hp_coef = 0.016768633120843694
+hunger_coef = 0.04373640516193193
+illegal_penalty = -0.009138756699997636
+death_penalty = -0.8209825710553373
+ac_coef = 0.0508544124054636
 heal_coef = 0.001
 status_coef = 0.01
 
@@ -27,25 +27,25 @@ num_layers = 3
 expansion_factor = 1
 
 [train]
-total_timesteps = 179_453_112
+total_timesteps = 247765094
 minibatch_size = 32768
 horizon = 64
-learning_rate = 0.013102893141695623
-gamma = 0.995330546979216
-gae_lambda = 0.7296684290473575
-replay_ratio = 1.393745327111498
+learning_rate = 0.011854702184553156
+gamma = 0.9973714278257158
+gae_lambda = 0.7339946788354172
+replay_ratio = 1.036717410123433
 clip_coef = 1
-vf_coef = 2.967784385898158
-vf_clip_coef = 4.084322349420867
-max_grad_norm = 1.0968230967915535
-ent_coef = 0.005217403062762164
-beta1 = 0.8250073190394559
-beta2 = 0.9935894811507382
-eps = 3.82673948478513e-13
-vtrace_rho_clip = 3.8767312776569813
-vtrace_c_clip = 4.60481789092722
-prio_alpha = 0.7459213056962047
-prio_beta0 = 0.7387749352151263
+vf_coef = 1.8486010633248169
+vf_clip_coef = 4.624607731543118
+max_grad_norm = 0.902165688849037
+ent_coef = 0.008389210353641681
+beta1 = 0.805399553172738
+beta2 = 0.999139492311758
+eps = 1e-14
+vtrace_rho_clip = 2.770648671516198
+vtrace_c_clip = 4.5026502564349595
+prio_alpha = 0.6754855924309153
+prio_beta0 = 0.8672463883417074
 
 [sweep]
 metric = score