edbeeching HF staff commited on
Commit
3833731
1 Parent(s): 01e86d9

Upload cfg.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. cfg.json +139 -0
cfg.json ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "help": false,
3
+ "algo": "APPO",
4
+ "env": "Cartpole",
5
+ "experiment": "Cartpole_1111",
6
+ "train_dir": "/gpfsssd/scratch/rech/ajs/utv52ia/isaacgym/sample-factory/train_dir",
7
+ "restart_behavior": "resume",
8
+ "device": "gpu",
9
+ "seed": 1111,
10
+ "num_policies": 1,
11
+ "async_rl": false,
12
+ "serial_mode": true,
13
+ "batched_sampling": true,
14
+ "num_batches_to_accumulate": 2,
15
+ "worker_num_splits": 1,
16
+ "policy_workers_per_policy": 1,
17
+ "max_policy_lag": 1000,
18
+ "num_workers": 1,
19
+ "num_envs_per_worker": 1,
20
+ "batch_size": 32768,
21
+ "num_batches_per_epoch": 8,
22
+ "num_epochs": 5,
23
+ "rollout": 16,
24
+ "recurrence": 1,
25
+ "shuffle_minibatches": false,
26
+ "gamma": 0.99,
27
+ "reward_scale": 0.01,
28
+ "reward_clip": 1000.0,
29
+ "value_bootstrap": true,
30
+ "normalize_returns": true,
31
+ "exploration_loss_coeff": 0.0,
32
+ "value_loss_coeff": 2.0,
33
+ "kl_loss_coeff": 0.0,
34
+ "exploration_loss": "entropy",
35
+ "gae_lambda": 0.95,
36
+ "ppo_clip_ratio": 0.2,
37
+ "ppo_clip_value": 1.0,
38
+ "with_vtrace": false,
39
+ "vtrace_rho": 1.0,
40
+ "vtrace_c": 1.0,
41
+ "optimizer": "adam",
42
+ "adam_eps": 1e-06,
43
+ "adam_beta1": 0.9,
44
+ "adam_beta2": 0.999,
45
+ "max_grad_norm": 1.0,
46
+ "learning_rate": 0.0003,
47
+ "lr_schedule": "kl_adaptive_epoch",
48
+ "lr_schedule_kl_threshold": 0.008,
49
+ "obs_subtract_mean": 0.0,
50
+ "obs_scale": 1.0,
51
+ "normalize_input": true,
52
+ "normalize_input_keys": null,
53
+ "decorrelate_experience_max_seconds": 0,
54
+ "decorrelate_envs_on_one_worker": true,
55
+ "actor_worker_gpus": [
56
+ 0
57
+ ],
58
+ "set_workers_cpu_affinity": true,
59
+ "force_envs_single_thread": false,
60
+ "default_niceness": 0,
61
+ "log_to_file": true,
62
+ "experiment_summaries_interval": 10,
63
+ "flush_summaries_interval": 30,
64
+ "stats_avg": 100,
65
+ "summaries_use_frameskip": true,
66
+ "heartbeat_interval": 20,
67
+ "heartbeat_reporting_interval": 180,
68
+ "train_for_env_steps": 1000000000,
69
+ "train_for_seconds": 10000000000,
70
+ "save_every_sec": 120,
71
+ "keep_checkpoints": 2,
72
+ "load_checkpoint_kind": "latest",
73
+ "save_milestones_sec": -1,
74
+ "save_best_every_sec": 5,
75
+ "save_best_metric": "reward",
76
+ "save_best_after": 5000000,
77
+ "benchmark": false,
78
+ "encoder_mlp_layers": [
79
+ 128,
80
+ 64,
81
+ 32
82
+ ],
83
+ "encoder_conv_architecture": "convnet_simple",
84
+ "encoder_conv_mlp_layers": [
85
+ 512
86
+ ],
87
+ "use_rnn": false,
88
+ "rnn_size": 512,
89
+ "rnn_type": "gru",
90
+ "rnn_num_layers": 1,
91
+ "decoder_mlp_layers": [],
92
+ "nonlinearity": "elu",
93
+ "policy_initialization": "torch_default",
94
+ "policy_init_gain": 1.0,
95
+ "actor_critic_share_weights": true,
96
+ "adaptive_stddev": false,
97
+ "continuous_tanh_scale": 0.0,
98
+ "initial_stddev": 1.0,
99
+ "use_env_info_cache": true,
100
+ "env_gpu_actions": true,
101
+ "env_gpu_observations": true,
102
+ "env_frameskip": 1,
103
+ "env_framestack": 1,
104
+ "pixel_format": "CHW",
105
+ "use_record_episode_statistics": false,
106
+ "with_wandb": false,
107
+ "wandb_user": null,
108
+ "wandb_project": "sample_factory",
109
+ "wandb_group": null,
110
+ "wandb_job_type": "SF",
111
+ "wandb_tags": [],
112
+ "with_pbt": false,
113
+ "pbt_mix_policies_in_one_env": true,
114
+ "pbt_period_env_steps": 5000000,
115
+ "pbt_start_mutation": 20000000,
116
+ "pbt_replace_fraction": 0.3,
117
+ "pbt_mutation_rate": 0.15,
118
+ "pbt_replace_reward_gap": 0.1,
119
+ "pbt_replace_reward_gap_absolute": 1e-06,
120
+ "pbt_optimize_gamma": false,
121
+ "pbt_target_objective": "true_objective",
122
+ "pbt_perturb_min": 1.1,
123
+ "pbt_perturb_max": 1.5,
124
+ "env_agents": -1,
125
+ "obs_key": "obs",
126
+ "subtask": null,
127
+ "ige_api_version": "preview4",
128
+ "eval_stats": false,
129
+ "command_line": "--seed=1111 --experiment=Cartpole_1111 --env=Cartpole --train_for_env_steps=1000000000",
130
+ "cli_args": {
131
+ "env": "Cartpole",
132
+ "experiment": "Cartpole_1111",
133
+ "seed": 1111,
134
+ "train_for_env_steps": 1000000000
135
+ },
136
+ "git_hash": "a79821cb7657bf600e5eb05db9b797b132737468",
137
+ "git_repo_name": "https://github.com/alex-petrenko/sample-factory.git",
138
+ "train_script": "sf_examples.isaacgym_examples.train_isaacgym"
139
+ }