nethack sweep

logs/nethack/mine50e_1b_s311.ini

@@ -0,0 +1,103 @@
+# PufferLib log v1
+
+[base]
+env_name = nethack
+rank = 0
+world_size = 1
+gpu_id = 0
+gpu_offset = 0
+nccl_id = None
+profile = False
+checkpoint_dir = checkpoints
+log_dir = logs
+checkpoint_interval = 500
+eval_episodes = 10000
+burnin_games = 0
+eval_agents = -1
+load_model_path = None
+load_enemy_model_path = None
+run_id = mine50e_1b_s311
+result_fd = 0
+wandb = True
+slowly = False
+cudagraphs = 1
+seed = 311
+reset_every_horizon = 0
+async = 1
+
+[vec]
+total_agents = 256
+num_buffers = 4
+num_threads = 32
+num_policies = 1
+hist_policy_hidden_size = 0
+hist_policy_num_layers = 0
+hist_policy_percent = 0
+action_mask_size = 734
+gpu_env = 0
+
+[selfplay]
+enabled = 0
+max_size = 16
+seed = 42
+opp_timeout_steps = 500_000_000
+eval_pool_size = 8
+eval_games = 0
+eval_agents = 8192
+
+[env]
+dr = 0
+num_agents = 1
+num_bots = 0
+gold_coef = 0.0953414813
+exp_coef = 9.99999975e-05
+descent_coef = 0.807671428
+floor_coef = 0.463313729
+xp_coef = 0.0555660129
+scout_coef = 0.0192291923
+ac_coef = 0.0
+death_penalty = -0.580139101
+scout_ready = 0.0
+ac_nospell = 1.0
+mask_search20 = 1
+mask_run = 1
+multi_role = 1
+
+[policy]
+hidden_size = 512
+num_layers = 3
+expansion_factor = 1
+
+[torch]
+network = MinGRU
+encoder = DefaultEncoder
+decoder = DefaultDecoder
+
+[train]
+gpus = 1
+seed = 42
+total_timesteps = 1000000000
+learning_rate = 0.000608655275
+anneal_lr = 1
+min_lr_ratio = 0.0
+gamma = 0.998621047
+gae_lambda = 0.536597371
+replay_ratio = 1.91373289
+clip_coef = 0.175347924
+vf_coef = 1.45746601
+vf_clip_coef = 0.00999999978
+max_grad_norm = 1.23201728
+ent_coef = 0.00844300818
+anneal_ent_coef = 0
+min_ent_coef_ratio = 0.1
+momentum = 0.984448254
+minibatch_size = 4096
+horizon = 256
+vtrace = 0
+vtrace_rho_clip = 1.0
+vtrace_c_clip = 1.0
+verb_eps = 0.005
+verb_eps_anneal_start = 0.4
+verb_eps_anneal_end = 1.0
+prio_alpha = 0
+prio_beta0 = 0.610757589

logs/nethack/sw0386_1b_s311.ini

@@ -0,0 +1,101 @@
+# PufferLib log v1
+
+[base]
+env_name = nethack
+rank = 0
+world_size = 1
+gpu_id = 0
+gpu_offset = 0
+nccl_id = None
+profile = False
+checkpoint_dir = checkpoints
+log_dir = logs
+checkpoint_interval = 500
+eval_episodes = 10000
+burnin_games = 0
+eval_agents = -1
+load_model_path = None
+load_enemy_model_path = None
+run_id = sw0386_1b_s311
+result_fd = 0
+wandb = True
+slowly = False
+cudagraphs = 1
+seed = 311
+reset_every_horizon = 0
+async = 1
+
+[vec]
+total_agents = 512
+num_buffers = 4
+num_threads = 32
+num_policies = 1
+hist_policy_hidden_size = 0
+hist_policy_num_layers = 0
+hist_policy_percent = 0
+action_mask_size = 734
+gpu_env = 0
+
+[selfplay]
+enabled = 0
+max_size = 16
+seed = 42
+opp_timeout_steps = 500_000_000
+eval_pool_size = 8
+eval_games = 0
+eval_agents = 8192
+
+[env]
+dr = 0
+num_agents = 1
+num_bots = 0
+gold_coef = 0.096887961
+exp_coef = 0.041234985
+descent_coef = 1
+floor_coef = 0.38962388
+xp_coef = 0.741106272
+scout_coef = 0.000813971157
+ac_coef = 0.0
+death_penalty = -0.286484599
+scout_ready = 0.0
+ac_nospell = 1.0
+mask_search20 = 0
+mask_run = 0
+multi_role = 1
+
+[policy]
+hidden_size = 512
+num_layers = 3
+expansion_factor = 1
+
+[torch]
+network = MinGRU
+encoder = DefaultEncoder
+decoder = DefaultDecoder
+
+[train]
+gpus = 1
+seed = 42
+total_timesteps = 1000000000
+learning_rate = 0.00154131022
+anneal_lr = 1
+min_lr_ratio = 0.0
+gamma = 0.999329984
+gae_lambda = 0.199999988
+replay_ratio = 1.20105004
+clip_coef = 0.36599344
+vf_coef = 0.882273436
+vf_clip_coef = 1.30713189
+max_grad_norm = 0.335048616
+ent_coef = 0.00555205066
+anneal_ent_coef = 0
+min_ent_coef_ratio = 0.1
+momentum = 0.956484258
+minibatch_size = 4096
+horizon = 256
+vtrace = 0
+vtrace_rho_clip = 1.0
+vtrace_c_clip = 1.0
+verb_eps = 0.005
+verb_eps_anneal_start = 0.4
+verb_eps_anneal_end = 1.0