Upload . with huggingface_hub

.summary/0/events.out.tfevents.1677112709.f54eb2240718

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41a402a1f4c0d94882b613aea62f44328064236c87352c850bf7f95b86eccbff
+size 2343

README.md

@@ -15,7 +15,7 @@ model-index:
       type: doom_health_gathering_supreme
     metrics:
     - type: mean_reward
-      value: 4.32 +/- 0.70
+      value: 4.00 +/- 0.68
       name: mean_reward
       verified: false
 ---

checkpoint_p0/checkpoint_000002445_10014720.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e39d4dd75fee0ec62008470cbd583761a51ab57a687a270bc53bb6034e566d0
+size 34929220

config.json

@@ -23,7 +23,7 @@
   "rollout": 32,
   "recurrence": 32,
   "shuffle_minibatches": false,
-  "gamma": 0.99,
+  "gamma": 0.98,
   "reward_scale": 1.0,
   "reward_clip": 1000.0,
   "value_bootstrap": false,

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:09537b156b7b8af3b288cbd6c49470d1689b5ff069d806ac5711235bf15ac062
-size 5619577
+oid sha256:70258433234a67ba882f938454e050765928c1bce517d1c7bb922800bacacd92
+size 4994059

sf_log.txt

@@ -2459,3 +2459,568 @@ main_loop: 1813.1331
 [2023-02-23 00:26:52,988][05631] Avg episode rewards: #0: 5.020, true rewards: #0: 4.320
 [2023-02-23 00:26:52,991][05631] Avg episode reward: 5.020, avg true_objective: 4.320
 [2023-02-23 00:27:13,979][05631] Replay video saved to /content/train_dir/default_experiment/replay.mp4!
+[2023-02-23 00:27:19,636][05631] The model has been pushed to https://huggingface.co/pittawat/rl_course_vizdoom_health_gathering_supreme
+[2023-02-23 00:38:29,240][05631] Environment doom_basic already registered, overwriting...
+[2023-02-23 00:38:29,244][05631] Environment doom_two_colors_easy already registered, overwriting...
+[2023-02-23 00:38:29,246][05631] Environment doom_two_colors_hard already registered, overwriting...
+[2023-02-23 00:38:29,248][05631] Environment doom_dm already registered, overwriting...
+[2023-02-23 00:38:29,250][05631] Environment doom_dwango5 already registered, overwriting...
+[2023-02-23 00:38:29,251][05631] Environment doom_my_way_home_flat_actions already registered, overwriting...
+[2023-02-23 00:38:29,253][05631] Environment doom_defend_the_center_flat_actions already registered, overwriting...
+[2023-02-23 00:38:29,254][05631] Environment doom_my_way_home already registered, overwriting...
+[2023-02-23 00:38:29,256][05631] Environment doom_deadly_corridor already registered, overwriting...
+[2023-02-23 00:38:29,258][05631] Environment doom_defend_the_center already registered, overwriting...
+[2023-02-23 00:38:29,259][05631] Environment doom_defend_the_line already registered, overwriting...
+[2023-02-23 00:38:29,261][05631] Environment doom_health_gathering already registered, overwriting...
+[2023-02-23 00:38:29,262][05631] Environment doom_health_gathering_supreme already registered, overwriting...
+[2023-02-23 00:38:29,263][05631] Environment doom_battle already registered, overwriting...
+[2023-02-23 00:38:29,265][05631] Environment doom_battle2 already registered, overwriting...
+[2023-02-23 00:38:29,266][05631] Environment doom_duel_bots already registered, overwriting...
+[2023-02-23 00:38:29,268][05631] Environment doom_deathmatch_bots already registered, overwriting...
+[2023-02-23 00:38:29,269][05631] Environment doom_duel already registered, overwriting...
+[2023-02-23 00:38:29,271][05631] Environment doom_deathmatch_full already registered, overwriting...
+[2023-02-23 00:38:29,272][05631] Environment doom_benchmark already registered, overwriting...
+[2023-02-23 00:38:29,274][05631] register_encoder_factory: <function make_vizdoom_encoder at 0x7f0a330101f0>
+[2023-02-23 00:38:29,308][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:38:29,311][05631] Overriding arg 'gamma' with value 0.98 passed from command line
+[2023-02-23 00:38:29,317][05631] Experiment dir /content/train_dir/default_experiment already exists!
+[2023-02-23 00:38:29,319][05631] Resuming existing experiment from /content/train_dir/default_experiment...
+[2023-02-23 00:38:29,322][05631] Weights and Biases integration disabled
+[2023-02-23 00:38:29,326][05631] Environment var CUDA_VISIBLE_DEVICES is 0
+
+[2023-02-23 00:38:32,154][05631] Starting experiment with the following configuration:
+help=False
+algo=APPO
+env=doom_health_gathering_supreme
+experiment=default_experiment
+train_dir=/content/train_dir
+restart_behavior=resume
+device=gpu
+seed=None
+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=8
+num_envs_per_worker=4
+batch_size=1024
+num_batches_per_epoch=1
+num_epochs=1
+rollout=32
+recurrence=32
+shuffle_minibatches=False
+gamma=0.98
+reward_scale=1.0
+reward_clip=1000.0
+value_bootstrap=False
+normalize_returns=True
+exploration_loss_coeff=0.001
+value_loss_coeff=0.5
+kl_loss_coeff=0.0
+exploration_loss=symmetric_kl
+gae_lambda=0.95
+ppo_clip_ratio=0.1
+ppo_clip_value=0.2
+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=255.0
+normalize_input=True
+normalize_input_keys=None
+decorrelate_experience_max_seconds=0
+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=600
+train_for_env_steps=10000000
+train_for_seconds=10000000000
+save_every_sec=120
+keep_checkpoints=2
+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=4
+env_framestack=1
+pixel_format=CHW
+use_record_episode_statistics=False
+with_wandb=False
+wandb_user=None
+wandb_project=sample_factory
+wandb_group=None
+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
+num_agents=-1
+num_humans=0
+num_bots=-1
+start_bot_difficulty=None
+timelimit=None
+res_w=128
+res_h=72
+wide_aspect_ratio=False
+eval_env_frameskip=1
+fps=35
+command_line=--env=doom_health_gathering_supreme --num_workers=8 --num_envs_per_worker=4 --train_for_env_steps=4000000
+cli_args={'env': 'doom_health_gathering_supreme', 'num_workers': 8, 'num_envs_per_worker': 4, 'train_for_env_steps': 4000000}
+git_hash=unknown
+git_repo_name=not a git repository
+[2023-02-23 00:38:32,160][05631] Saving configuration to /content/train_dir/default_experiment/config.json...
+[2023-02-23 00:38:32,167][05631] Rollout worker 0 uses device cpu
+[2023-02-23 00:38:32,172][05631] Rollout worker 1 uses device cpu
+[2023-02-23 00:38:32,174][05631] Rollout worker 2 uses device cpu
+[2023-02-23 00:38:32,177][05631] Rollout worker 3 uses device cpu
+[2023-02-23 00:38:32,180][05631] Rollout worker 4 uses device cpu
+[2023-02-23 00:38:32,182][05631] Rollout worker 5 uses device cpu
+[2023-02-23 00:38:32,184][05631] Rollout worker 6 uses device cpu
+[2023-02-23 00:38:32,193][05631] Rollout worker 7 uses device cpu
+[2023-02-23 00:38:32,318][05631] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:38:32,323][05631] InferenceWorker_p0-w0: min num requests: 2
+[2023-02-23 00:38:32,354][05631] Starting all processes...
+[2023-02-23 00:38:32,359][05631] Starting process learner_proc0
+[2023-02-23 00:38:32,505][05631] Starting all processes...
+[2023-02-23 00:38:32,516][05631] Starting process inference_proc0-0
+[2023-02-23 00:38:32,517][05631] Starting process rollout_proc0
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc1
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc2
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc3
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc4
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc5
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc6
+[2023-02-23 00:38:32,521][05631] Starting process rollout_proc7
+[2023-02-23 00:38:44,016][34379] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:38:44,017][34379] Set environment var CUDA_VISIBLE_DEVICES to '0' (GPU indices [0]) for learning process 0
+[2023-02-23 00:38:44,094][34379] Num visible devices: 1
+[2023-02-23 00:38:44,127][34379] Starting seed is not provided
+[2023-02-23 00:38:44,127][34379] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:38:44,128][34379] Initializing actor-critic model on device cuda:0
+[2023-02-23 00:38:44,129][34379] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:38:44,136][34379] RunningMeanStd input shape: (1,)
+[2023-02-23 00:38:44,260][34379] ConvEncoder: input_channels=3
+[2023-02-23 00:38:44,635][34394] Worker 0 uses CPU cores [0]
+[2023-02-23 00:38:44,821][34393] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:38:44,822][34393] Set environment var CUDA_VISIBLE_DEVICES to '0' (GPU indices [0]) for inference process 0
+[2023-02-23 00:38:44,894][34393] Num visible devices: 1
+[2023-02-23 00:38:45,181][34379] Conv encoder output size: 512
+[2023-02-23 00:38:45,182][34379] Policy head output size: 512
+[2023-02-23 00:38:45,290][34379] Created Actor Critic model with architecture:
+[2023-02-23 00:38:45,292][34395] Worker 2 uses CPU cores [0]
+[2023-02-23 00:38:45,291][34379] ActorCriticSharedWeights(
+  (obs_normalizer): ObservationNormalizer(
+    (running_mean_std): RunningMeanStdDictInPlace(
+      (running_mean_std): ModuleDict(
+        (obs): RunningMeanStdInPlace()
+      )
+    )
+  )
+  (returns_normalizer): RecursiveScriptModule(original_name=RunningMeanStdInPlace)
+  (encoder): VizdoomEncoder(
+    (basic_encoder): ConvEncoder(
+      (enc): RecursiveScriptModule(
+        original_name=ConvEncoderImpl
+        (conv_head): RecursiveScriptModule(
+          original_name=Sequential
+          (0): RecursiveScriptModule(original_name=Conv2d)
+          (1): RecursiveScriptModule(original_name=ELU)
+          (2): RecursiveScriptModule(original_name=Conv2d)
+          (3): RecursiveScriptModule(original_name=ELU)
+          (4): RecursiveScriptModule(original_name=Conv2d)
+          (5): RecursiveScriptModule(original_name=ELU)
+        )
+        (mlp_layers): RecursiveScriptModule(
+          original_name=Sequential
+          (0): RecursiveScriptModule(original_name=Linear)
+          (1): RecursiveScriptModule(original_name=ELU)
+        )
+      )
+    )
+  )
+  (core): ModelCoreRNN(
+    (core): GRU(512, 512)
+  )
+  (decoder): MlpDecoder(
+    (mlp): Identity()
+  )
+  (critic_linear): Linear(in_features=512, out_features=1, bias=True)
+  (action_parameterization): ActionParameterizationDefault(
+    (distribution_linear): Linear(in_features=512, out_features=5, bias=True)
+  )
+)
+[2023-02-23 00:38:45,460][34404] Worker 4 uses CPU cores [0]
+[2023-02-23 00:38:45,507][34400] Worker 1 uses CPU cores [1]
+[2023-02-23 00:38:45,837][34406] Worker 3 uses CPU cores [1]
+[2023-02-23 00:38:45,918][34415] Worker 5 uses CPU cores [1]
+[2023-02-23 00:38:46,026][34408] Worker 6 uses CPU cores [0]
+[2023-02-23 00:38:46,052][34416] Worker 7 uses CPU cores [1]
+[2023-02-23 00:38:49,594][34379] Using optimizer <class 'torch.optim.adam.Adam'>
+[2023-02-23 00:38:49,595][34379] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002443_10006528.pth...
+[2023-02-23 00:38:49,629][34379] Loading model from checkpoint
+[2023-02-23 00:38:49,634][34379] Loaded experiment state at self.train_step=2443, self.env_steps=10006528
+[2023-02-23 00:38:49,635][34379] Initialized policy 0 weights for model version 2443
+[2023-02-23 00:38:49,637][34379] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:38:49,644][34379] LearnerWorker_p0 finished initialization!
+[2023-02-23 00:38:49,753][34393] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:38:49,754][34393] RunningMeanStd input shape: (1,)
+[2023-02-23 00:38:49,767][34393] ConvEncoder: input_channels=3
+[2023-02-23 00:38:49,868][34393] Conv encoder output size: 512
+[2023-02-23 00:38:49,869][34393] Policy head output size: 512
+[2023-02-23 00:38:52,129][05631] Inference worker 0-0 is ready!
+[2023-02-23 00:38:52,131][05631] All inference workers are ready! Signal rollout workers to start!
+[2023-02-23 00:38:52,234][34406] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,236][34408] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,235][34400] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,237][34404] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,232][34416] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,237][34394] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,232][34395] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,239][34415] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:38:52,310][05631] Heartbeat connected on Batcher_0
+[2023-02-23 00:38:52,315][05631] Heartbeat connected on LearnerWorker_p0
+[2023-02-23 00:38:52,366][05631] Heartbeat connected on InferenceWorker_p0-w0
+[2023-02-23 00:38:52,742][34406] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:53,295][34415] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:53,847][34395] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:53,851][34408] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:53,857][34404] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:53,859][34394] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:54,328][05631] Fps is (10 sec: nan, 60 sec: nan, 300 sec: nan). Total num frames: 10006528. Throughput: 0: nan. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:38:55,102][34404] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:55,113][34394] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:55,146][34395] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:55,194][34400] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:55,186][34416] Decorrelating experience for 0 frames...
+[2023-02-23 00:38:56,646][34408] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:56,994][34416] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:57,000][34400] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:57,003][34406] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:57,016][34394] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:57,018][34395] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:58,783][34404] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:58,802][34415] Decorrelating experience for 32 frames...
+[2023-02-23 00:38:59,076][34408] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:59,141][34406] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:59,160][34400] Decorrelating experience for 64 frames...
+[2023-02-23 00:38:59,261][34395] Decorrelating experience for 96 frames...
+[2023-02-23 00:38:59,275][34394] Decorrelating experience for 96 frames...
+[2023-02-23 00:38:59,327][05631] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 10006528. Throughput: 0: 0.0. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:38:59,623][05631] Heartbeat connected on RolloutWorker_w2
+[2023-02-23 00:38:59,671][05631] Heartbeat connected on RolloutWorker_w0
+[2023-02-23 00:39:00,510][34404] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:00,708][05631] Heartbeat connected on RolloutWorker_w4
+[2023-02-23 00:39:00,726][34408] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:00,911][05631] Heartbeat connected on RolloutWorker_w6
+[2023-02-23 00:39:01,609][34415] Decorrelating experience for 64 frames...
+[2023-02-23 00:39:01,635][34416] Decorrelating experience for 64 frames...
+[2023-02-23 00:39:01,747][34406] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:01,758][34400] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:02,043][05631] Heartbeat connected on RolloutWorker_w3
+[2023-02-23 00:39:02,062][05631] Heartbeat connected on RolloutWorker_w1
+[2023-02-23 00:39:04,220][34415] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:04,249][34416] Decorrelating experience for 96 frames...
+[2023-02-23 00:39:04,327][05631] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 10006528. Throughput: 0: 166.6. Samples: 1666. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:39:04,329][05631] Avg episode reward: [(0, '1.925')]
+[2023-02-23 00:39:04,908][05631] Heartbeat connected on RolloutWorker_w5
+[2023-02-23 00:39:04,955][05631] Heartbeat connected on RolloutWorker_w7
+[2023-02-23 00:39:05,985][34379] Signal inference workers to stop experience collection...
+[2023-02-23 00:39:06,010][34393] InferenceWorker_p0-w0: stopping experience collection
+[2023-02-23 00:39:07,046][34379] Signal inference workers to resume experience collection...
+[2023-02-23 00:39:07,049][34393] InferenceWorker_p0-w0: resuming experience collection
+[2023-02-23 00:39:07,049][34379] Stopping Batcher_0...
+[2023-02-23 00:39:07,052][34379] Loop batcher_evt_loop terminating...
+[2023-02-23 00:39:07,050][05631] Component Batcher_0 stopped!
+[2023-02-23 00:39:07,084][34406] Stopping RolloutWorker_w3...
+[2023-02-23 00:39:07,085][34406] Loop rollout_proc3_evt_loop terminating...
+[2023-02-23 00:39:07,084][05631] Component RolloutWorker_w3 stopped!
+[2023-02-23 00:39:07,091][34415] Stopping RolloutWorker_w5...
+[2023-02-23 00:39:07,092][34400] Stopping RolloutWorker_w1...
+[2023-02-23 00:39:07,093][34415] Loop rollout_proc5_evt_loop terminating...
+[2023-02-23 00:39:07,094][34400] Loop rollout_proc1_evt_loop terminating...
+[2023-02-23 00:39:07,091][05631] Component RolloutWorker_w5 stopped!
+[2023-02-23 00:39:07,097][34416] Stopping RolloutWorker_w7...
+[2023-02-23 00:39:07,098][34416] Loop rollout_proc7_evt_loop terminating...
+[2023-02-23 00:39:07,098][05631] Component RolloutWorker_w1 stopped!
+[2023-02-23 00:39:07,105][05631] Component RolloutWorker_w7 stopped!
+[2023-02-23 00:39:07,143][05631] Component RolloutWorker_w0 stopped!
+[2023-02-23 00:39:07,150][34394] Stopping RolloutWorker_w0...
+[2023-02-23 00:39:07,150][34394] Loop rollout_proc0_evt_loop terminating...
+[2023-02-23 00:39:07,160][05631] Component RolloutWorker_w4 stopped!
+[2023-02-23 00:39:07,166][34404] Stopping RolloutWorker_w4...
+[2023-02-23 00:39:07,166][34404] Loop rollout_proc4_evt_loop terminating...
+[2023-02-23 00:39:07,159][34393] Weights refcount: 2 0
+[2023-02-23 00:39:07,182][05631] Component InferenceWorker_p0-w0 stopped!
+[2023-02-23 00:39:07,191][34408] Stopping RolloutWorker_w6...
+[2023-02-23 00:39:07,191][05631] Component RolloutWorker_w6 stopped!
+[2023-02-23 00:39:07,193][34393] Stopping InferenceWorker_p0-w0...
+[2023-02-23 00:39:07,194][34393] Loop inference_proc0-0_evt_loop terminating...
+[2023-02-23 00:39:07,200][34408] Loop rollout_proc6_evt_loop terminating...
+[2023-02-23 00:39:07,204][05631] Component RolloutWorker_w2 stopped!
+[2023-02-23 00:39:07,206][34395] Stopping RolloutWorker_w2...
+[2023-02-23 00:39:07,208][34395] Loop rollout_proc2_evt_loop terminating...
+[2023-02-23 00:39:09,166][34379] Saving /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002445_10014720.pth...
+[2023-02-23 00:39:09,278][34379] Removing /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002434_9969664.pth
+[2023-02-23 00:39:09,286][34379] Saving /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002445_10014720.pth...
+[2023-02-23 00:39:09,425][05631] Component LearnerWorker_p0 stopped!
+[2023-02-23 00:39:09,431][05631] Waiting for process learner_proc0 to stop...
+[2023-02-23 00:39:09,426][34379] Stopping LearnerWorker_p0...
+[2023-02-23 00:39:09,436][34379] Loop learner_proc0_evt_loop terminating...
+[2023-02-23 00:39:10,754][05631] Waiting for process inference_proc0-0 to join...
+[2023-02-23 00:39:10,758][05631] Waiting for process rollout_proc0 to join...
+[2023-02-23 00:39:10,763][05631] Waiting for process rollout_proc1 to join...
+[2023-02-23 00:39:10,765][05631] Waiting for process rollout_proc2 to join...
+[2023-02-23 00:39:10,770][05631] Waiting for process rollout_proc3 to join...
+[2023-02-23 00:39:10,774][05631] Waiting for process rollout_proc4 to join...
+[2023-02-23 00:39:10,778][05631] Waiting for process rollout_proc5 to join...
+[2023-02-23 00:39:10,781][05631] Waiting for process rollout_proc6 to join...
+[2023-02-23 00:39:10,783][05631] Waiting for process rollout_proc7 to join...
+[2023-02-23 00:39:10,787][05631] Batcher 0 profile tree view:
+batching: 0.0630, releasing_batches: 0.0022
+[2023-02-23 00:39:10,793][05631] InferenceWorker_p0-w0 profile tree view:
+update_model: 0.0221
+wait_policy: 0.0000
+  wait_policy_total: 9.8845
+one_step: 0.0029
+  handle_policy_step: 3.6858
+    deserialize: 0.0487, stack: 0.0120, obs_to_device_normalize: 0.3783, forward: 2.8836, send_messages: 0.0690
+    prepare_outputs: 0.2014
+      to_cpu: 0.1193
+[2023-02-23 00:39:10,795][05631] Learner 0 profile tree view:
+misc: 0.0000, prepare_batch: 5.0141
+train: 0.6550
+  epoch_init: 0.0000, minibatch_init: 0.0000, losses_postprocess: 0.0004, kl_divergence: 0.0004, after_optimizer: 0.0048
+  calculate_losses: 0.1413
+    losses_init: 0.0000, forward_head: 0.1155, bptt_initial: 0.0169, tail: 0.0014, advantages_returns: 0.0011, losses: 0.0034
+    bptt: 0.0026
+      bptt_forward_core: 0.0025
+  update: 0.5072
+    clip: 0.0054
+[2023-02-23 00:39:10,796][05631] RolloutWorker_w0 profile tree view:
+wait_for_trajectories: 0.0233, enqueue_policy_requests: 0.9685, env_step: 3.1668, overhead: 0.0812, complete_rollouts: 0.0138
+save_policy_outputs: 0.0673
+  split_output_tensors: 0.0201
+[2023-02-23 00:39:10,797][05631] RolloutWorker_w7 profile tree view:
+wait_for_trajectories: 0.0003, enqueue_policy_requests: 0.1870, env_step: 0.7674, overhead: 0.0097, complete_rollouts: 0.0001
+save_policy_outputs: 0.0074
+  split_output_tensors: 0.0037
+[2023-02-23 00:39:10,799][05631] Loop Runner_EvtLoop terminating...
+[2023-02-23 00:39:10,800][05631] Runner profile tree view:
+main_loop: 38.4461
+[2023-02-23 00:39:10,801][05631] Collected {0: 10014720}, FPS: 213.1
+[2023-02-23 00:39:10,846][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:39:10,850][05631] Overriding arg 'num_workers' with value 1 passed from command line
+[2023-02-23 00:39:10,854][05631] Adding new argument 'no_render'=True that is not in the saved config file!
+[2023-02-23 00:39:10,856][05631] Adding new argument 'save_video'=True that is not in the saved config file!
+[2023-02-23 00:39:10,858][05631] Adding new argument 'video_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:39:10,859][05631] Adding new argument 'video_name'=None that is not in the saved config file!
+[2023-02-23 00:39:10,865][05631] Adding new argument 'max_num_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:39:10,867][05631] Adding new argument 'max_num_episodes'=10 that is not in the saved config file!
+[2023-02-23 00:39:10,872][05631] Adding new argument 'push_to_hub'=False that is not in the saved config file!
+[2023-02-23 00:39:10,874][05631] Adding new argument 'hf_repository'=None that is not in the saved config file!
+[2023-02-23 00:39:10,875][05631] Adding new argument 'policy_index'=0 that is not in the saved config file!
+[2023-02-23 00:39:10,877][05631] Adding new argument 'eval_deterministic'=False that is not in the saved config file!
+[2023-02-23 00:39:10,878][05631] Adding new argument 'train_script'=None that is not in the saved config file!
+[2023-02-23 00:39:10,880][05631] Adding new argument 'enjoy_script'=None that is not in the saved config file!
+[2023-02-23 00:39:10,883][05631] Using frameskip 1 and render_action_repeat=4 for evaluation
+[2023-02-23 00:39:10,923][05631] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:39:10,926][05631] RunningMeanStd input shape: (1,)
+[2023-02-23 00:39:10,947][05631] ConvEncoder: input_channels=3
+[2023-02-23 00:39:11,007][05631] Conv encoder output size: 512
+[2023-02-23 00:39:11,009][05631] Policy head output size: 512
+[2023-02-23 00:39:11,044][05631] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002445_10014720.pth...
+[2023-02-23 00:39:11,890][05631] Num frames 100...
+[2023-02-23 00:39:12,066][05631] Num frames 200...
+[2023-02-23 00:39:12,256][05631] Num frames 300...
+[2023-02-23 00:39:12,465][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:39:12,468][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:39:12,504][05631] Num frames 400...
+[2023-02-23 00:39:12,689][05631] Num frames 500...
+[2023-02-23 00:39:12,859][05631] Num frames 600...
+[2023-02-23 00:39:13,033][05631] Num frames 700...
+[2023-02-23 00:39:13,213][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:39:13,216][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:39:13,278][05631] Num frames 800...
+[2023-02-23 00:39:13,449][05631] Num frames 900...
+[2023-02-23 00:39:13,626][05631] Num frames 1000...
+[2023-02-23 00:39:13,797][05631] Num frames 1100...
+[2023-02-23 00:39:13,944][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:39:13,947][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:39:14,040][05631] Num frames 1200...
+[2023-02-23 00:39:14,212][05631] Num frames 1300...
+[2023-02-23 00:39:14,396][05631] Num frames 1400...
+[2023-02-23 00:39:14,571][05631] Num frames 1500...
+[2023-02-23 00:39:14,703][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:39:14,706][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:39:14,832][05631] Num frames 1600...
+[2023-02-23 00:39:14,955][05631] Num frames 1700...
+[2023-02-23 00:39:15,076][05631] Num frames 1800...
+[2023-02-23 00:39:15,196][05631] Num frames 1900...
+[2023-02-23 00:39:15,312][05631] Num frames 2000...
+[2023-02-23 00:39:15,443][05631] Num frames 2100...
+[2023-02-23 00:39:15,570][05631] Num frames 2200...
+[2023-02-23 00:39:15,676][05631] Avg episode rewards: #0: 5.280, true rewards: #0: 4.480
+[2023-02-23 00:39:15,678][05631] Avg episode reward: 5.280, avg true_objective: 4.480
+[2023-02-23 00:39:15,760][05631] Num frames 2300...
+[2023-02-23 00:39:15,888][05631] Num frames 2400...
+[2023-02-23 00:39:16,008][05631] Num frames 2500...
+[2023-02-23 00:39:16,129][05631] Num frames 2600...
+[2023-02-23 00:39:16,262][05631] Avg episode rewards: #0: 5.260, true rewards: #0: 4.427
+[2023-02-23 00:39:16,264][05631] Avg episode reward: 5.260, avg true_objective: 4.427
+[2023-02-23 00:39:16,324][05631] Num frames 2700...
+[2023-02-23 00:39:16,453][05631] Num frames 2800...
+[2023-02-23 00:39:16,577][05631] Num frames 2900...
+[2023-02-23 00:39:16,695][05631] Num frames 3000...
+[2023-02-23 00:39:16,797][05631] Avg episode rewards: #0: 5.057, true rewards: #0: 4.343
+[2023-02-23 00:39:16,799][05631] Avg episode reward: 5.057, avg true_objective: 4.343
+[2023-02-23 00:39:16,880][05631] Num frames 3100...
+[2023-02-23 00:39:17,017][05631] Num frames 3200...
+[2023-02-23 00:39:17,146][05631] Num frames 3300...
+[2023-02-23 00:39:17,277][05631] Num frames 3400...
+[2023-02-23 00:39:17,451][05631] Avg episode rewards: #0: 5.110, true rewards: #0: 4.360
+[2023-02-23 00:39:17,454][05631] Avg episode reward: 5.110, avg true_objective: 4.360
+[2023-02-23 00:39:17,476][05631] Num frames 3500...
+[2023-02-23 00:39:17,595][05631] Num frames 3600...
+[2023-02-23 00:39:17,733][05631] Num frames 3700...
+[2023-02-23 00:39:17,839][05631] Avg episode rewards: #0: 4.827, true rewards: #0: 4.160
+[2023-02-23 00:39:17,842][05631] Avg episode reward: 4.827, avg true_objective: 4.160
+[2023-02-23 00:39:17,915][05631] Num frames 3800...
+[2023-02-23 00:39:18,037][05631] Num frames 3900...
+[2023-02-23 00:39:18,159][05631] Num frames 4000...
+[2023-02-23 00:39:18,282][05631] Num frames 4100...
+[2023-02-23 00:39:18,412][05631] Avg episode rewards: #0: 4.860, true rewards: #0: 4.160
+[2023-02-23 00:39:18,415][05631] Avg episode reward: 4.860, avg true_objective: 4.160
+[2023-02-23 00:39:41,642][05631] Replay video saved to /content/train_dir/default_experiment/replay.mp4!
+[2023-02-23 00:39:41,852][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:39:41,854][05631] Overriding arg 'num_workers' with value 1 passed from command line
+[2023-02-23 00:39:41,857][05631] Adding new argument 'no_render'=True that is not in the saved config file!
+[2023-02-23 00:39:41,859][05631] Adding new argument 'save_video'=True that is not in the saved config file!
+[2023-02-23 00:39:41,865][05631] Adding new argument 'video_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:39:41,866][05631] Adding new argument 'video_name'=None that is not in the saved config file!
+[2023-02-23 00:39:41,868][05631] Adding new argument 'max_num_frames'=100000 that is not in the saved config file!
+[2023-02-23 00:39:41,869][05631] Adding new argument 'max_num_episodes'=10 that is not in the saved config file!
+[2023-02-23 00:39:41,870][05631] Adding new argument 'push_to_hub'=True that is not in the saved config file!
+[2023-02-23 00:39:41,871][05631] Adding new argument 'hf_repository'='pittawat/rl_course_vizdoom_health_gathering_supreme' that is not in the saved config file!
+[2023-02-23 00:39:41,873][05631] Adding new argument 'policy_index'=0 that is not in the saved config file!
+[2023-02-23 00:39:41,877][05631] Adding new argument 'eval_deterministic'=False that is not in the saved config file!
+[2023-02-23 00:39:41,878][05631] Adding new argument 'train_script'=None that is not in the saved config file!
+[2023-02-23 00:39:41,880][05631] Adding new argument 'enjoy_script'=None that is not in the saved config file!
+[2023-02-23 00:39:41,881][05631] Using frameskip 1 and render_action_repeat=4 for evaluation
+[2023-02-23 00:39:41,906][05631] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:39:41,909][05631] RunningMeanStd input shape: (1,)
+[2023-02-23 00:39:41,928][05631] ConvEncoder: input_channels=3
+[2023-02-23 00:39:41,987][05631] Conv encoder output size: 512
+[2023-02-23 00:39:41,990][05631] Policy head output size: 512
+[2023-02-23 00:39:42,019][05631] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002445_10014720.pth...
+[2023-02-23 00:39:42,823][05631] Num frames 100...
+[2023-02-23 00:39:43,007][05631] Num frames 200...
+[2023-02-23 00:39:43,184][05631] Num frames 300...
+[2023-02-23 00:39:43,391][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:39:43,394][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:39:43,429][05631] Num frames 400...
+[2023-02-23 00:39:43,622][05631] Num frames 500...
+[2023-02-23 00:39:43,820][05631] Num frames 600...
+[2023-02-23 00:39:44,016][05631] Num frames 700...
+[2023-02-23 00:39:44,226][05631] Num frames 800...
+[2023-02-23 00:39:44,433][05631] Num frames 900...
+[2023-02-23 00:39:44,553][05631] Avg episode rewards: #0: 5.640, true rewards: #0: 4.640
+[2023-02-23 00:39:44,555][05631] Avg episode reward: 5.640, avg true_objective: 4.640
+[2023-02-23 00:39:44,709][05631] Num frames 1000...
+[2023-02-23 00:39:44,901][05631] Num frames 1100...
+[2023-02-23 00:39:45,072][05631] Num frames 1200...
+[2023-02-23 00:39:45,252][05631] Num frames 1300...
+[2023-02-23 00:39:45,337][05631] Avg episode rewards: #0: 5.040, true rewards: #0: 4.373
+[2023-02-23 00:39:45,340][05631] Avg episode reward: 5.040, avg true_objective: 4.373
+[2023-02-23 00:39:45,500][05631] Num frames 1400...
+[2023-02-23 00:39:45,630][05631] Num frames 1500...
+[2023-02-23 00:39:45,751][05631] Num frames 1600...
+[2023-02-23 00:39:45,867][05631] Num frames 1700...
+[2023-02-23 00:39:45,955][05631] Avg episode rewards: #0: 5.070, true rewards: #0: 4.320
+[2023-02-23 00:39:45,963][05631] Avg episode reward: 5.070, avg true_objective: 4.320
+[2023-02-23 00:39:46,059][05631] Num frames 1800...
+[2023-02-23 00:39:46,186][05631] Num frames 1900...
+[2023-02-23 00:39:46,324][05631] Num frames 2000...
+[2023-02-23 00:39:46,443][05631] Num frames 2100...
+[2023-02-23 00:39:46,551][05631] Avg episode rewards: #0: 5.088, true rewards: #0: 4.288
+[2023-02-23 00:39:46,553][05631] Avg episode reward: 5.088, avg true_objective: 4.288
+[2023-02-23 00:39:46,628][05631] Num frames 2200...
+[2023-02-23 00:39:46,751][05631] Num frames 2300...
+[2023-02-23 00:39:46,870][05631] Num frames 2400...
+[2023-02-23 00:39:46,990][05631] Num frames 2500...
+[2023-02-23 00:39:47,080][05631] Avg episode rewards: #0: 4.880, true rewards: #0: 4.213
+[2023-02-23 00:39:47,082][05631] Avg episode reward: 4.880, avg true_objective: 4.213
+[2023-02-23 00:39:47,182][05631] Num frames 2600...
+[2023-02-23 00:39:47,318][05631] Num frames 2700...
+[2023-02-23 00:39:47,455][05631] Num frames 2800...
+[2023-02-23 00:39:47,581][05631] Num frames 2900...
+[2023-02-23 00:39:47,727][05631] Avg episode rewards: #0: 4.966, true rewards: #0: 4.251
+[2023-02-23 00:39:47,729][05631] Avg episode reward: 4.966, avg true_objective: 4.251
+[2023-02-23 00:39:47,764][05631] Num frames 3000...
+[2023-02-23 00:39:47,890][05631] Num frames 3100...
+[2023-02-23 00:39:48,018][05631] Num frames 3200...
+[2023-02-23 00:39:48,138][05631] Num frames 3300...
+[2023-02-23 00:39:48,265][05631] Avg episode rewards: #0: 4.825, true rewards: #0: 4.200
+[2023-02-23 00:39:48,267][05631] Avg episode reward: 4.825, avg true_objective: 4.200
+[2023-02-23 00:39:48,322][05631] Num frames 3400...
+[2023-02-23 00:39:48,453][05631] Num frames 3500...
+[2023-02-23 00:39:48,573][05631] Num frames 3600...
+[2023-02-23 00:39:48,693][05631] Num frames 3700...
+[2023-02-23 00:39:48,808][05631] Avg episode rewards: #0: 4.716, true rewards: #0: 4.160
+[2023-02-23 00:39:48,811][05631] Avg episode reward: 4.716, avg true_objective: 4.160
+[2023-02-23 00:39:48,888][05631] Num frames 3800...
+[2023-02-23 00:39:49,016][05631] Num frames 3900...
+[2023-02-23 00:39:49,139][05631] Num frames 4000...
+[2023-02-23 00:39:49,195][05631] Avg episode rewards: #0: 4.500, true rewards: #0: 4.000
+[2023-02-23 00:39:49,198][05631] Avg episode reward: 4.500, avg true_objective: 4.000
+[2023-02-23 00:40:09,146][05631] Replay video saved to /content/train_dir/default_experiment/replay.mp4!