| { | |
| "help": false, | |
| "algo": "APPO", | |
| "env": "challenge", | |
| "experiment": "default_experiment", | |
| "train_dir": "/net/tscratch/people/plgbartekcupial/mrunner_scratch/sf2_nethack/20_03-18_43-laughing_jang/2024-03-20-monk-appo-ks-t-baseline_983r_1/train_dir", | |
| "restart_behavior": "resume", | |
| "device": "gpu", | |
| "seed": 1, | |
| "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": 1000, | |
| "num_workers": 16, | |
| "num_envs_per_worker": 32, | |
| "batch_size": 4096, | |
| "num_batches_per_epoch": 4, | |
| "num_epochs": 1, | |
| "rollout": 128, | |
| "recurrence": 128, | |
| "shuffle_minibatches": false, | |
| "warmup": 0, | |
| "gamma": 1.0, | |
| "reward_scale": 1.0, | |
| "reward_clip": 10.0, | |
| "value_bootstrap": false, | |
| "normalize_returns": true, | |
| "exploration_loss_coeff": 0.0, | |
| "value_loss_coeff": 1.0, | |
| "kl_loss_coeff": 0.0, | |
| "exploration_loss": "entropy", | |
| "gae_lambda": 1.0, | |
| "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-07, | |
| "adam_beta1": 0.9, | |
| "adam_beta2": 0.999, | |
| "momentum": 0.9, | |
| "max_grad_norm": 4, | |
| "learning_rate": 0.0001, | |
| "lr_schedule": "linear_decay", | |
| "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": false, | |
| "normalize_input_keys": null, | |
| "decorrelate_experience_max_seconds": 0, | |
| "decorrelate_envs_on_one_worker": false, | |
| "actor_worker_gpus": [], | |
| "set_workers_cpu_affinity": true, | |
| "force_envs_single_thread": false, | |
| "default_niceness": 0, | |
| "log_to_file": true, | |
| "experiment_summaries_interval": 50, | |
| "flush_summaries_interval": 30, | |
| "stats_avg": 100, | |
| "summaries_use_frameskip": true, | |
| "heartbeat_interval": 20, | |
| "heartbeat_reporting_interval": 180, | |
| "train_for_env_steps": 500000000, | |
| "train_for_seconds": 10000000000, | |
| "save_every_sec": 120, | |
| "keep_checkpoints": 2, | |
| "load_checkpoint_kind": "latest", | |
| "save_milestones_sec": -1, | |
| "save_milestones_ith": 25000000, | |
| "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": 1738, | |
| "rnn_type": "lstm", | |
| "rnn_num_layers": 1, | |
| "decoder_mlp_layers": [], | |
| "nonlinearity": "relu", | |
| "policy_initialization": "orthogonal", | |
| "policy_init_gain": 1.0, | |
| "actor_critic_share_weights": true, | |
| "adaptive_stddev": false, | |
| "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": "bartekcupial", | |
| "wandb_project": "sf2_nethack", | |
| "wandb_group": "gmum", | |
| "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, | |
| "character": "mon-hum-neu-mal", | |
| "max_episode_steps": 100000, | |
| "penalty_step": 0.0, | |
| "penalty_time": 0.0, | |
| "fn_penalty_step": "constant", | |
| "savedir": null, | |
| "save_ttyrec_every": 0, | |
| "gameloaddir": null, | |
| "state_counter": null, | |
| "add_image_observation": true, | |
| "crop_dim": 18, | |
| "pixel_size": 6, | |
| "reward_shaping": false, | |
| "use_tty_only": true, | |
| "use_prev_action": true, | |
| "h_dim": 1738, | |
| "msg_hdim": 64, | |
| "color_edim": 16, | |
| "char_edim": 16, | |
| "use_crop": true, | |
| "use_crop_norm": true, | |
| "screen_kernel_size": 3, | |
| "no_max_pool": false, | |
| "screen_conv_blocks": 2, | |
| "blstats_hdim": 512, | |
| "fc_after_cnn_hdim": 512, | |
| "use_resnet": true, | |
| "use_dataset": false, | |
| "behavioral_clone": false, | |
| "data_path": "/nle/nld-aa/nle_data", | |
| "db_path": "/ttyrecs/ttyrecs.db", | |
| "dataset_name": "autoascend", | |
| "dataset_num_splits": 2, | |
| "dataset_warmup": 0, | |
| "dataset_rollout": 32, | |
| "dataset_batch_size": 1024, | |
| "dataset_num_workers": 8, | |
| "dataset_demigod": false, | |
| "dataset_highscore": false, | |
| "dataset_midscore": false, | |
| "dataset_deep": false, | |
| "dataset_shuffle": true, | |
| "reset_on_rollout_boundary": false, | |
| "exp_tags": "['2024_03_20_monk-APPO-KS-T-baseline']", | |
| "exp_point": "monk-APPO-KS-T", | |
| "group": "monk-APPO-KS-T", | |
| "use_pretrained_checkpoint": true, | |
| "model": "ScaledNet", | |
| "model_path": "/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained", | |
| "supervised_loss_coeff": 0.0, | |
| "kickstarting_loss_coeff": 0.5, | |
| "distillation_loss_coeff": 0.0, | |
| "supervised_loss_decay": 1.0, | |
| "kickstarting_loss_decay": 0.99998, | |
| "distillation_loss_decay": 1.0, | |
| "min_supervised_loss_coeff": 0.0, | |
| "min_kickstarting_loss_coeff": 0.0, | |
| "min_distillation_loss_coeff": 0.0, | |
| "substitute_regularization_with_exploration": false, | |
| "exploration_coeff_on_supervised_loss_coeff": 0.0, | |
| "exploration_coeff_on_kickstarting_loss_coeff": 0.0, | |
| "exploration_coeff_on_distillation_loss_coeff": 0.0, | |
| "teacher_path": "/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained", | |
| "run_teacher_hs": false, | |
| "add_stats_to_info": true, | |
| "capture_video": false, | |
| "capture_video_ith": 100, | |
| "freeze": { | |
| "encoder": 0 | |
| }, | |
| "unfreeze": {}, | |
| "freeze_batch_norm": false, | |
| "skip_train": 25000000, | |
| "target_batch_size": 128, | |
| "optim_step_every_ith": 1, | |
| "command_line": "--env=challenge --exp_tags=['2024_03_20_monk-APPO-KS-T-baseline'] --exp_point=monk-APPO-KS-T --train_for_env_steps=500000000 --group=monk-APPO-KS-T --character=mon-hum-neu-mal --num_workers=16 --num_envs_per_worker=32 --worker_num_splits=2 --rollout=128 --batch_size=4096 --async_rl=True --serial_mode=False --wandb_user=bartekcupial --wandb_project=sf2_nethack --wandb_group=gmum --with_wandb=True --use_pretrained_checkpoint=True --model_path=/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained --kickstarting_loss_coeff=0.5 --teacher_path=/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained --run_teacher_hs=False --use_prev_action=True --model=ScaledNet --use_resnet=True --learning_rate=0.0001 --rnn_size=1738 --h_dim=1738 --exploration_loss_coeff=0.0 --gamma=1.0 --skip_train=25000000 --lr_schedule=linear_decay --save_milestones_ith=25000000 --kickstarting_loss_decay=0.99998 --seed=1 --freeze={'encoder': 0} --num_batches_per_epoch=4 --optim_step_every_ith=1 --target_batch_size=128", | |
| "cli_args": { | |
| "env": "challenge", | |
| "seed": 1, | |
| "async_rl": true, | |
| "serial_mode": false, | |
| "worker_num_splits": 2, | |
| "num_workers": 16, | |
| "num_envs_per_worker": 32, | |
| "batch_size": 4096, | |
| "num_batches_per_epoch": 4, | |
| "rollout": 128, | |
| "gamma": 1.0, | |
| "exploration_loss_coeff": 0.0, | |
| "learning_rate": 0.0001, | |
| "lr_schedule": "linear_decay", | |
| "train_for_env_steps": 500000000, | |
| "save_milestones_ith": 25000000, | |
| "rnn_size": 1738, | |
| "with_wandb": true, | |
| "wandb_user": "bartekcupial", | |
| "wandb_project": "sf2_nethack", | |
| "wandb_group": "gmum", | |
| "character": "mon-hum-neu-mal", | |
| "use_prev_action": true, | |
| "h_dim": 1738, | |
| "use_resnet": true, | |
| "exp_tags": "['2024_03_20_monk-APPO-KS-T-baseline']", | |
| "exp_point": "monk-APPO-KS-T", | |
| "group": "monk-APPO-KS-T", | |
| "use_pretrained_checkpoint": true, | |
| "model": "ScaledNet", | |
| "model_path": "/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained", | |
| "kickstarting_loss_coeff": 0.5, | |
| "kickstarting_loss_decay": 0.99998, | |
| "teacher_path": "/net/pr2/projects/plgrid/plgggmum_crl/bcupial/sf_checkpoints/amzn-AA-BC_pretrained", | |
| "run_teacher_hs": false, | |
| "freeze": { | |
| "encoder": 0 | |
| }, | |
| "skip_train": 25000000, | |
| "target_batch_size": 128, | |
| "optim_step_every_ith": 1 | |
| }, | |
| "git_hash": "unknown", | |
| "git_repo_name": "not a git repository", | |
| "wandb_unique_id": "default_experiment_20240321_101612_080200" | |
| } |