Hugging Face's logo

daveey
/
metta

Reinforcement Learning
sample-factory
TensorBoard
deep-reinforcement-learning
Model card Files Metrics Community
metta / config.json
daveey's picture
daveey
Upload folder using huggingface_hub
535f508 verified
raw history blame contribute delete
No virus
7.98 kB
{
"help": false,
"algo": "APPO",
"env": "GDY-PowerGrid",
"experiment": "p2.objt_atn.4",
"train_dir": "/workspace/metta/train_dir",
"restart_behavior": "resume",
"device": "gpu",
"seed": null,
"num_policies": 1,
"async_rl": true,
"serial_mode": false,
"batched_sampling": false,
"num_batches_to_accumulate": 2,
"worker_num_splits": 2,
"policy_workers_per_policy": 1,
"max_policy_lag": 20000,
"num_workers": 50,
"num_envs_per_worker": 2,
"batch_size": 32768,
"num_batches_per_epoch": 1,
"num_epochs": 1,
"rollout": 256,
"recurrence": 256,
"shuffle_minibatches": false,
"gamma": 0.99,
"reward_scale": 1.0,
"reward_clip": 1000.0,
"value_bootstrap": false,
"normalize_returns": true,
"exploration_loss_coeff": 0.002,
"value_loss_coeff": 0.976,
"kl_loss_coeff": 0.0,
"aux_loss_coeff": 0.0,
"exploration_loss": "symmetric_kl",
"gae_lambda": 0.95,
"ppo_clip_ratio": 0.1,
"ppo_clip_value": 1.0,
"with_vtrace": false,
"vtrace_rho": 1.0,
"vtrace_c": 1.0,
"optimizer": "adam",
"adam_eps": 1e-06,
"adam_beta1": 0.9,
"adam_beta2": 0.999,
"max_grad_norm": 4.0,
"learning_rate": 0.0001,
"lr_schedule": "constant",
"lr_schedule_kl_threshold": 0.008,
"lr_adaptive_min": 1e-06,
"lr_adaptive_max": 0.01,
"obs_subtract_mean": 0.0,
"obs_scale": 1.0,
"normalize_input": true,
"normalize_input_keys": null,
"decorrelate_experience_max_seconds": 150,
"decorrelate_envs_on_one_worker": true,
"actor_worker_gpus": [],
"set_workers_cpu_affinity": true,
"force_envs_single_thread": false,
"default_niceness": 0,
"log_to_file": true,
"experiment_summaries_interval": 10,
"flush_summaries_interval": 30,
"stats_avg": 100,
"summaries_use_frameskip": true,
"heartbeat_interval": 20,
"heartbeat_reporting_interval": 180,
"train_for_env_steps": 10000000000,
"train_for_seconds": 10000000000,
"save_every_sec": 120,
"keep_checkpoints": 2,
"load_checkpoint_path": null,
"init_checkpoint_path": "train_dir/p2.objt_atn.3/checkpoint_p0/checkpoint_000046241_1515225088.pth",
"load_checkpoint_kind": "latest",
"save_milestones_sec": -1,
"save_best_every_sec": 5,
"save_best_metric": "reward",
"save_best_after": 100000,
"benchmark": false,
"encoder_mlp_layers": [
512,
512
],
"encoder_conv_architecture": "convnet_simple",
"encoder_conv_mlp_layers": [
512
],
"use_rnn": true,
"rnn_size": 512,
"rnn_type": "gru",
"rnn_num_layers": 1,
"decoder_mlp_layers": [],
"nonlinearity": "elu",
"policy_initialization": "orthogonal",
"policy_init_gain": 1.0,
"actor_critic_share_weights": true,
"adaptive_stddev": true,
"continuous_tanh_scale": 0.0,
"initial_stddev": 1.0,
"use_env_info_cache": false,
"env_gpu_actions": false,
"env_gpu_observations": true,
"env_frameskip": 1,
"env_framestack": 1,
"pixel_format": "CHW",
"use_record_episode_statistics": false,
"episode_counter": false,
"with_wandb": true,
"wandb_user": null,
"wandb_project": "sample_factory",
"wandb_group": null,
"wandb_job_type": "SF",
"wandb_tags": [],
"with_pbt": false,
"pbt_mix_policies_in_one_env": true,
"pbt_period_env_steps": 5000000,
"pbt_start_mutation": 20000000,
"pbt_replace_fraction": 0.3,
"pbt_mutation_rate": 0.15,
"pbt_replace_reward_gap": 0.1,
"pbt_replace_reward_gap_absolute": 1e-06,
"pbt_optimize_gamma": false,
"pbt_target_objective": "true_objective",
"pbt_perturb_min": 1.1,
"pbt_perturb_max": 1.5,
"env_num_agents": 100,
"env_max_steps": 1000,
"scenario": null,
"agent": null,
"env_altar:cooldown": [
2,
5
],
"env_altar:cost": [
100.0
],
"env_charger:cooldown": [
2,
2
],
"env_charger:energy": [
50,
50
],
"env_generator:cooldown": [
20,
100
],
"env_agent:energy:regen": [
0,
0
],
"env_agent:energy:initial": [
10,
200
],
"env_agent:energy:max": [
500,
500
],
"env_gift:energy": [
20,
20
],
"env_cost:move:predator": [
0,
0
],
"env_cost:move:prey": [
0,
0
],
"env_cost:move": [
0,
0
],
"env_cost:jump": [
3,
3
],
"env_cost:rotate": [
0,
0
],
"env_cost:shield": [
0.0
],
"env_cost:shield:upkeep": [
1,
2
],
"env_cost:frozen": [
0,
0
],
"env_cost:attack": [
5,
40
],
"env_attack:damage": [
5,
40
],
"env_attack:freeze_duration": [
5,
100
],
"env_tile_size": [
32,
32
],
"env_width": [
100.0
],
"env_height": [
100.0
],
"env_wall_density": [
0.01
],
"env_milestone_density": [
0,
0.1
],
"env_num_altars": [
20.0
],
"env_num_chargers": [
30.0
],
"env_num_generators": [
200.0
],
"env_rsm_num_families": [
0,
0
],
"env_rsm_family_reward": [
0,
0
],
"env_reward_rank_steps": [
1000.0
],
"env_reward_prestige_weight": [
0.0
],
"agent_max_features": 50,
"agent_fc_layers": 4,
"agent_fc_size": 512,
"agent_embedding_size": 64,
"agent_embedding_layers": 3,
"agent_attention_size": 512,
"agent_attention_layers": 2,
"command_line": "--algo=APPO --env=GDY-PowerGrid --with_wandb=True --num_workers=25 --env_num_agents=5 --env_width=20:40 --env_height=20:40 --env_num_altars=1:5 --env_num_chargers=1:5 --env_num_generators=5:20 --env_wall_density=0:0.15 --env_reward_rank_steps=10 --agent_fc_layers=3 --agent_fc_size=512 --agent_embedding_size=64 --agent_embedding_layers=3 --agent_attention_size=512 --agent_attention_layers=2 --rnn_num_layers=2 --rnn_size=512 --rnn_type=gru --rollout=256 --batch_size=65536 --decorrelate_experience_max_seconds=150 --value_loss_coeff=0.976 --exploration_loss=symmetric_kl --exploration_loss_coeff=0.002 --policy_initialization=orthogonal --learning_rate=0.0001 --max_policy_lag=2000 --nonlinearity=elu --normalize_input=True --env_width=100 --env_height=100 --env_num_agents=100 --env_num_altars=20 --env_num_chargers=30 --env_num_generators=200 --env_wall_density=0.01 --env_reward_rank_steps=1000 --env_altar:cost=100 --env_reward_prestige_weight=0 --env_cost:shield=0 --agent_fc_layers=4 --agent_fc_size=512 --rnn_num_layers=1 --rnn_size=512 --rnn_type=gru --max_policy_lag=20000 --experiment=p2.objt_atn.4 --num_workers=50 --batch_size=32768 --init_checkpoint_path=workspace/metta/train_dir/p2.objt_atn.3/checkpoint_p0/checkpoint_000046241_1515225088.pth",
"cli_args": {
"algo": "APPO",
"env": "GDY-PowerGrid",
"experiment": "p2.objt_atn.4",
"max_policy_lag": 20000,
"num_workers": 50,
"batch_size": 32768,
"rollout": 256,
"exploration_loss_coeff": 0.002,
"value_loss_coeff": 0.976,
"exploration_loss": "symmetric_kl",
"learning_rate": 0.0001,
"normalize_input": true,
"decorrelate_experience_max_seconds": 150,
"init_checkpoint_path": "workspace/metta/train_dir/p2.objt_atn.3/checkpoint_p0/checkpoint_000046241_1515225088.pth",
"rnn_size": 512,
"rnn_type": "gru",
"rnn_num_layers": 1,
"nonlinearity": "elu",
"policy_initialization": "orthogonal",
"with_wandb": true,
"env_num_agents": 100,
"env_altar:cost": [
100.0
],
"env_cost:shield": [
0.0
],
"env_width": [
100.0
],
"env_height": [
100.0
],
"env_wall_density": [
0.01
],
"env_num_altars": [
20.0
],
"env_num_chargers": [
30.0
],
"env_num_generators": [
200.0
],
"env_reward_rank_steps": [
1000.0
],
"env_reward_prestige_weight": [
0.0
],
"agent_fc_layers": 4,
"agent_fc_size": 512,
"agent_embedding_size": 64,
"agent_embedding_layers": 3,
"agent_attention_size": 512,
"agent_attention_layers": 2
},
"git_hash": "960abca49c3c941d43bcf65df909840ad43c6156",
"git_repo_name": "https://github.com/daveey/metta.git",
"wandb_unique_id": "p2.objt_atn.4_20240320_224453_910734"
}