Upload . with huggingface_hub

.summary/0/events.out.tfevents.1677113085.f54eb2240718

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc55a4c1084efc746a5ed4cf48120476b155daceb4bb870a4e15b9cf6c660c1d
+size 2683

README.md

@@ -15,7 +15,7 @@ model-index:
       type: doom_health_gathering_supreme
     metrics:
     - type: mean_reward
-      value: 4.00 +/- 0.68
+      value: 4.19 +/- 0.56
       name: mean_reward
       verified: false
 ---

checkpoint_p0/checkpoint_000002447_10022912.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0788c774d56aa22ced225fdc77679ab03a5f12285c48857899a0cb39e9935424
+size 34929220

config.json

@@ -23,7 +23,7 @@
   "rollout": 32,
   "recurrence": 32,
   "shuffle_minibatches": false,
-  "gamma": 0.98,
+  "gamma": 0.97,
   "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 sha256:70258433234a67ba882f938454e050765928c1bce517d1c7bb922800bacacd92
-size 4994059
+oid sha256:fa318558a81f3083a89ea6e50af8629d4ddd3569b8c1d9991c6bc53794fa001e
+size 5366056

sf_log.txt

@@ -3024,3 +3024,566 @@ main_loop: 38.4461
 [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!
+[2023-02-23 00:40:14,495][05631] The model has been pushed to https://huggingface.co/pittawat/rl_course_vizdoom_health_gathering_supreme
+[2023-02-23 00:44:45,344][05631] Environment doom_basic already registered, overwriting...
+[2023-02-23 00:44:45,347][05631] Environment doom_two_colors_easy already registered, overwriting...
+[2023-02-23 00:44:45,351][05631] Environment doom_two_colors_hard already registered, overwriting...
+[2023-02-23 00:44:45,352][05631] Environment doom_dm already registered, overwriting...
+[2023-02-23 00:44:45,353][05631] Environment doom_dwango5 already registered, overwriting...
+[2023-02-23 00:44:45,356][05631] Environment doom_my_way_home_flat_actions already registered, overwriting...
+[2023-02-23 00:44:45,357][05631] Environment doom_defend_the_center_flat_actions already registered, overwriting...
+[2023-02-23 00:44:45,360][05631] Environment doom_my_way_home already registered, overwriting...
+[2023-02-23 00:44:45,361][05631] Environment doom_deadly_corridor already registered, overwriting...
+[2023-02-23 00:44:45,362][05631] Environment doom_defend_the_center already registered, overwriting...
+[2023-02-23 00:44:45,366][05631] Environment doom_defend_the_line already registered, overwriting...
+[2023-02-23 00:44:45,367][05631] Environment doom_health_gathering already registered, overwriting...
+[2023-02-23 00:44:45,368][05631] Environment doom_health_gathering_supreme already registered, overwriting...
+[2023-02-23 00:44:45,369][05631] Environment doom_battle already registered, overwriting...
+[2023-02-23 00:44:45,371][05631] Environment doom_battle2 already registered, overwriting...
+[2023-02-23 00:44:45,373][05631] Environment doom_duel_bots already registered, overwriting...
+[2023-02-23 00:44:45,376][05631] Environment doom_deathmatch_bots already registered, overwriting...
+[2023-02-23 00:44:45,377][05631] Environment doom_duel already registered, overwriting...
+[2023-02-23 00:44:45,378][05631] Environment doom_deathmatch_full already registered, overwriting...
+[2023-02-23 00:44:45,381][05631] Environment doom_benchmark already registered, overwriting...
+[2023-02-23 00:44:45,382][05631] register_encoder_factory: <function make_vizdoom_encoder at 0x7f0a330101f0>
+[2023-02-23 00:44:45,417][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:44:45,419][05631] Overriding arg 'gamma' with value 0.97 passed from command line
+[2023-02-23 00:44:45,426][05631] Experiment dir /content/train_dir/default_experiment already exists!
+[2023-02-23 00:44:45,428][05631] Resuming existing experiment from /content/train_dir/default_experiment...
+[2023-02-23 00:44:45,429][05631] Weights and Biases integration disabled
+[2023-02-23 00:44:45,438][05631] Environment var CUDA_VISIBLE_DEVICES is 0
+
+[2023-02-23 00:44:47,591][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=0.97
+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 00:44:47,598][05631] Saving configuration to /content/train_dir/default_experiment/config.json...
+[2023-02-23 00:44:47,602][05631] Rollout worker 0 uses device cpu
+[2023-02-23 00:44:47,603][05631] Rollout worker 1 uses device cpu
+[2023-02-23 00:44:47,606][05631] Rollout worker 2 uses device cpu
+[2023-02-23 00:44:47,608][05631] Rollout worker 3 uses device cpu
+[2023-02-23 00:44:47,609][05631] Rollout worker 4 uses device cpu
+[2023-02-23 00:44:47,610][05631] Rollout worker 5 uses device cpu
+[2023-02-23 00:44:47,612][05631] Rollout worker 6 uses device cpu
+[2023-02-23 00:44:47,613][05631] Rollout worker 7 uses device cpu
+[2023-02-23 00:44:47,732][05631] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:44:47,734][05631] InferenceWorker_p0-w0: min num requests: 2
+[2023-02-23 00:44:47,890][05631] Starting all processes...
+[2023-02-23 00:44:47,893][05631] Starting process learner_proc0
+[2023-02-23 00:44:48,025][05631] Starting all processes...
+[2023-02-23 00:44:48,033][05631] Starting process inference_proc0-0
+[2023-02-23 00:44:48,033][05631] Starting process rollout_proc0
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc1
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc2
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc3
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc4
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc5
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc6
+[2023-02-23 00:44:48,035][05631] Starting process rollout_proc7
+[2023-02-23 00:44:56,168][39829] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:44:56,168][39829] Set environment var CUDA_VISIBLE_DEVICES to '0' (GPU indices [0]) for learning process 0
+[2023-02-23 00:44:56,202][39829] Num visible devices: 1
+[2023-02-23 00:44:56,239][39829] Starting seed is not provided
+[2023-02-23 00:44:56,240][39829] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:44:56,241][39829] Initializing actor-critic model on device cuda:0
+[2023-02-23 00:44:56,242][39829] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:44:56,245][39829] RunningMeanStd input shape: (1,)
+[2023-02-23 00:44:56,288][39829] ConvEncoder: input_channels=3
+[2023-02-23 00:44:57,277][39829] Conv encoder output size: 512
+[2023-02-23 00:44:57,280][39829] Policy head output size: 512
+[2023-02-23 00:44:57,450][39829] Created Actor Critic model with architecture:
+[2023-02-23 00:44:57,468][39829] 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 00:44:57,531][39843] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:44:57,537][39843] Set environment var CUDA_VISIBLE_DEVICES to '0' (GPU indices [0]) for inference process 0
+[2023-02-23 00:44:57,689][39843] Num visible devices: 1
+[2023-02-23 00:44:58,288][39844] Worker 1 uses CPU cores [1]
+[2023-02-23 00:44:58,437][39850] Worker 0 uses CPU cores [0]
+[2023-02-23 00:44:58,552][39852] Worker 3 uses CPU cores [1]
+[2023-02-23 00:44:58,941][39854] Worker 2 uses CPU cores [0]
+[2023-02-23 00:44:59,465][39864] Worker 5 uses CPU cores [1]
+[2023-02-23 00:44:59,470][39857] Worker 4 uses CPU cores [0]
+[2023-02-23 00:44:59,564][39860] Worker 6 uses CPU cores [0]
+[2023-02-23 00:44:59,766][39866] Worker 7 uses CPU cores [1]
+[2023-02-23 00:45:02,236][39829] Using optimizer <class 'torch.optim.adam.Adam'>
+[2023-02-23 00:45:02,236][39829] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002445_10014720.pth...
+[2023-02-23 00:45:02,270][39829] Loading model from checkpoint
+[2023-02-23 00:45:02,274][39829] Loaded experiment state at self.train_step=2445, self.env_steps=10014720
+[2023-02-23 00:45:02,275][39829] Initialized policy 0 weights for model version 2445
+[2023-02-23 00:45:02,278][39829] LearnerWorker_p0 finished initialization!
+[2023-02-23 00:45:02,280][39829] Using GPUs [0] for process 0 (actually maps to GPUs [0])
+[2023-02-23 00:45:02,390][39843] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:45:02,392][39843] RunningMeanStd input shape: (1,)
+[2023-02-23 00:45:02,405][39843] ConvEncoder: input_channels=3
+[2023-02-23 00:45:02,507][39843] Conv encoder output size: 512
+[2023-02-23 00:45:02,507][39843] Policy head output size: 512
+[2023-02-23 00:45:04,694][05631] Inference worker 0-0 is ready!
+[2023-02-23 00:45:04,696][05631] All inference workers are ready! Signal rollout workers to start!
+[2023-02-23 00:45:04,798][39844] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,799][39852] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,790][39857] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,792][39860] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,797][39866] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,795][39854] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,799][39850] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:04,795][39864] Doom resolution: 160x120, resize resolution: (128, 72)
+[2023-02-23 00:45:05,438][05631] Fps is (10 sec: nan, 60 sec: nan, 300 sec: nan). Total num frames: 10014720. Throughput: 0: nan. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:45:05,630][39864] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:05,633][39866] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:05,940][39854] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:05,942][39860] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:05,945][39857] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:06,265][39844] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:06,275][39860] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:06,729][39854] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:07,336][39864] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:07,354][39866] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:07,407][39852] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:07,509][39860] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:07,686][39854] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:07,711][39844] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:07,725][05631] Heartbeat connected on Batcher_0
+[2023-02-23 00:45:07,728][05631] Heartbeat connected on LearnerWorker_p0
+[2023-02-23 00:45:07,759][05631] Heartbeat connected on InferenceWorker_p0-w0
+[2023-02-23 00:45:08,275][39850] Decorrelating experience for 0 frames...
+[2023-02-23 00:45:08,280][39857] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:08,857][39852] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:09,043][39866] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:09,122][39860] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:09,189][39864] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:09,271][39857] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:09,351][05631] Heartbeat connected on RolloutWorker_w6
+[2023-02-23 00:45:09,472][39844] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:09,955][39850] Decorrelating experience for 32 frames...
+[2023-02-23 00:45:10,106][39854] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:10,365][05631] Heartbeat connected on RolloutWorker_w2
+[2023-02-23 00:45:10,438][05631] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 10014720. Throughput: 0: 0.0. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:45:10,596][39852] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:10,741][39864] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:10,986][05631] Heartbeat connected on RolloutWorker_w5
+[2023-02-23 00:45:11,290][39844] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:11,683][05631] Heartbeat connected on RolloutWorker_w1
+[2023-02-23 00:45:12,330][39857] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:12,849][05631] Heartbeat connected on RolloutWorker_w4
+[2023-02-23 00:45:13,241][39852] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:13,738][05631] Heartbeat connected on RolloutWorker_w3
+[2023-02-23 00:45:15,355][39866] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:15,439][05631] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 10014720. Throughput: 0: 137.8. Samples: 1378. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2023-02-23 00:45:15,444][05631] Avg episode reward: [(0, '2.665')]
+[2023-02-23 00:45:16,721][05631] Heartbeat connected on RolloutWorker_w7
+[2023-02-23 00:45:16,893][39829] Signal inference workers to stop experience collection...
+[2023-02-23 00:45:16,943][39843] InferenceWorker_p0-w0: stopping experience collection
+[2023-02-23 00:45:17,440][39850] Decorrelating experience for 64 frames...
+[2023-02-23 00:45:18,212][39850] Decorrelating experience for 96 frames...
+[2023-02-23 00:45:18,373][05631] Heartbeat connected on RolloutWorker_w0
+[2023-02-23 00:45:18,423][39829] Signal inference workers to resume experience collection...
+[2023-02-23 00:45:18,424][39843] InferenceWorker_p0-w0: resuming experience collection
+[2023-02-23 00:45:18,435][39829] Stopping Batcher_0...
+[2023-02-23 00:45:18,436][39829] Loop batcher_evt_loop terminating...
+[2023-02-23 00:45:18,437][05631] Component Batcher_0 stopped!
+[2023-02-23 00:45:18,467][05631] Component RolloutWorker_w7 stopped!
+[2023-02-23 00:45:18,489][05631] Component RolloutWorker_w1 stopped!
+[2023-02-23 00:45:18,472][39866] Stopping RolloutWorker_w7...
+[2023-02-23 00:45:18,496][39866] Loop rollout_proc7_evt_loop terminating...
+[2023-02-23 00:45:18,494][39844] Stopping RolloutWorker_w1...
+[2023-02-23 00:45:18,505][05631] Component RolloutWorker_w5 stopped!
+[2023-02-23 00:45:18,511][39864] Stopping RolloutWorker_w5...
+[2023-02-23 00:45:18,511][39864] Loop rollout_proc5_evt_loop terminating...
+[2023-02-23 00:45:18,517][05631] Component RolloutWorker_w3 stopped!
+[2023-02-23 00:45:18,522][39852] Stopping RolloutWorker_w3...
+[2023-02-23 00:45:18,523][39852] Loop rollout_proc3_evt_loop terminating...
+[2023-02-23 00:45:18,526][39844] Loop rollout_proc1_evt_loop terminating...
+[2023-02-23 00:45:18,534][39857] Stopping RolloutWorker_w4...
+[2023-02-23 00:45:18,534][05631] Component RolloutWorker_w4 stopped!
+[2023-02-23 00:45:18,545][39857] Loop rollout_proc4_evt_loop terminating...
+[2023-02-23 00:45:18,557][39850] Stopping RolloutWorker_w0...
+[2023-02-23 00:45:18,557][39850] Loop rollout_proc0_evt_loop terminating...
+[2023-02-23 00:45:18,557][05631] Component RolloutWorker_w0 stopped!
+[2023-02-23 00:45:18,585][39854] Stopping RolloutWorker_w2...
+[2023-02-23 00:45:18,585][05631] Component RolloutWorker_w2 stopped!
+[2023-02-23 00:45:18,591][39860] Stopping RolloutWorker_w6...
+[2023-02-23 00:45:18,592][39860] Loop rollout_proc6_evt_loop terminating...
+[2023-02-23 00:45:18,586][39854] Loop rollout_proc2_evt_loop terminating...
+[2023-02-23 00:45:18,591][05631] Component RolloutWorker_w6 stopped!
+[2023-02-23 00:45:18,613][39843] Weights refcount: 2 0
+[2023-02-23 00:45:18,629][05631] Component InferenceWorker_p0-w0 stopped!
+[2023-02-23 00:45:18,633][39843] Stopping InferenceWorker_p0-w0...
+[2023-02-23 00:45:18,633][39843] Loop inference_proc0-0_evt_loop terminating...
+[2023-02-23 00:45:21,008][39829] Saving /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002447_10022912.pth...
+[2023-02-23 00:45:21,173][39829] Removing /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002443_10006528.pth
+[2023-02-23 00:45:21,183][39829] Saving /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002447_10022912.pth...
+[2023-02-23 00:45:21,420][39829] Stopping LearnerWorker_p0...
+[2023-02-23 00:45:21,421][05631] Component LearnerWorker_p0 stopped!
+[2023-02-23 00:45:21,421][39829] Loop learner_proc0_evt_loop terminating...
+[2023-02-23 00:45:21,424][05631] Waiting for process learner_proc0 to stop...
+[2023-02-23 00:45:22,957][05631] Waiting for process inference_proc0-0 to join...
+[2023-02-23 00:45:22,959][05631] Waiting for process rollout_proc0 to join...
+[2023-02-23 00:45:22,962][05631] Waiting for process rollout_proc1 to join...
+[2023-02-23 00:45:23,075][05631] Waiting for process rollout_proc2 to join...
+[2023-02-23 00:45:23,077][05631] Waiting for process rollout_proc3 to join...
+[2023-02-23 00:45:23,083][05631] Waiting for process rollout_proc4 to join...
+[2023-02-23 00:45:23,085][05631] Waiting for process rollout_proc5 to join...
+[2023-02-23 00:45:23,088][05631] Waiting for process rollout_proc6 to join...
+[2023-02-23 00:45:23,089][05631] Waiting for process rollout_proc7 to join...
+[2023-02-23 00:45:23,093][05631] Batcher 0 profile tree view:
+batching: 0.0454, releasing_batches: 0.0004
+[2023-02-23 00:45:23,095][05631] InferenceWorker_p0-w0 profile tree view:
+wait_policy: 0.0051
+  wait_policy_total: 8.5334
+update_model: 0.0442
+  weight_update: 0.0259
+one_step: 0.1178
+  handle_policy_step: 3.6239
+    deserialize: 0.0496, stack: 0.0089, obs_to_device_normalize: 0.3427, forward: 2.7631, send_messages: 0.1124
+    prepare_outputs: 0.2459
+      to_cpu: 0.1354
+[2023-02-23 00:45:23,097][05631] Learner 0 profile tree view:
+misc: 0.0000, prepare_batch: 6.3831
+train: 0.7580
+  epoch_init: 0.0000, minibatch_init: 0.0000, losses_postprocess: 0.0004, kl_divergence: 0.0005, after_optimizer: 0.0041
+  calculate_losses: 0.1495
+    losses_init: 0.0000, forward_head: 0.1162, bptt_initial: 0.0202, tail: 0.0017, advantages_returns: 0.0010, losses: 0.0060
+    bptt: 0.0038
+      bptt_forward_core: 0.0037
+  update: 0.6026
+    clip: 0.0081
+[2023-02-23 00:45:23,101][05631] RolloutWorker_w0 profile tree view:
+wait_for_trajectories: 0.0003, enqueue_policy_requests: 0.0006
+[2023-02-23 00:45:23,104][05631] RolloutWorker_w7 profile tree view:
+wait_for_trajectories: 0.0003, enqueue_policy_requests: 0.0175, env_step: 0.2402, overhead: 0.0012, complete_rollouts: 0.0000
+save_policy_outputs: 0.0009
+  split_output_tensors: 0.0004
+[2023-02-23 00:45:23,105][05631] Loop Runner_EvtLoop terminating...
+[2023-02-23 00:45:23,107][05631] Runner profile tree view:
+main_loop: 35.2166
+[2023-02-23 00:45:23,108][05631] Collected {0: 10022912}, FPS: 232.6
+[2023-02-23 00:51:07,960][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:51:07,965][05631] Overriding arg 'num_workers' with value 1 passed from command line
+[2023-02-23 00:51:07,967][05631] Adding new argument 'no_render'=True that is not in the saved config file!
+[2023-02-23 00:51:07,970][05631] Adding new argument 'save_video'=True that is not in the saved config file!
+[2023-02-23 00:51:07,972][05631] Adding new argument 'video_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:51:07,974][05631] Adding new argument 'video_name'=None that is not in the saved config file!
+[2023-02-23 00:51:07,976][05631] Adding new argument 'max_num_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:51:07,977][05631] Adding new argument 'max_num_episodes'=10 that is not in the saved config file!
+[2023-02-23 00:51:07,978][05631] Adding new argument 'push_to_hub'=False that is not in the saved config file!
+[2023-02-23 00:51:07,979][05631] Adding new argument 'hf_repository'=None that is not in the saved config file!
+[2023-02-23 00:51:07,980][05631] Adding new argument 'policy_index'=0 that is not in the saved config file!
+[2023-02-23 00:51:07,981][05631] Adding new argument 'eval_deterministic'=False that is not in the saved config file!
+[2023-02-23 00:51:07,983][05631] Adding new argument 'train_script'=None that is not in the saved config file!
+[2023-02-23 00:51:07,984][05631] Adding new argument 'enjoy_script'=None that is not in the saved config file!
+[2023-02-23 00:51:07,985][05631] Using frameskip 1 and render_action_repeat=4 for evaluation
+[2023-02-23 00:51:08,013][05631] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:51:08,016][05631] RunningMeanStd input shape: (1,)
+[2023-02-23 00:51:08,038][05631] ConvEncoder: input_channels=3
+[2023-02-23 00:51:08,096][05631] Conv encoder output size: 512
+[2023-02-23 00:51:08,100][05631] Policy head output size: 512
+[2023-02-23 00:51:08,130][05631] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002447_10022912.pth...
+[2023-02-23 00:51:08,807][05631] Num frames 100...
+[2023-02-23 00:51:08,941][05631] Num frames 200...
+[2023-02-23 00:51:09,054][05631] Num frames 300...
+[2023-02-23 00:51:09,207][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:51:09,209][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:51:09,233][05631] Num frames 400...
+[2023-02-23 00:51:09,366][05631] Num frames 500...
+[2023-02-23 00:51:09,487][05631] Num frames 600...
+[2023-02-23 00:51:09,611][05631] Num frames 700...
+[2023-02-23 00:51:09,743][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:51:09,746][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:51:09,788][05631] Num frames 800...
+[2023-02-23 00:51:09,912][05631] Num frames 900...
+[2023-02-23 00:51:10,031][05631] Num frames 1000...
+[2023-02-23 00:51:10,159][05631] Num frames 1100...
+[2023-02-23 00:51:10,318][05631] Avg episode rewards: #0: 4.280, true rewards: #0: 3.947
+[2023-02-23 00:51:10,322][05631] Avg episode reward: 4.280, avg true_objective: 3.947
+[2023-02-23 00:51:10,352][05631] Num frames 1200...
+[2023-02-23 00:51:10,477][05631] Num frames 1300...
+[2023-02-23 00:51:10,596][05631] Num frames 1400...
+[2023-02-23 00:51:10,710][05631] Num frames 1500...
+[2023-02-23 00:51:10,824][05631] Num frames 1600...
+[2023-02-23 00:51:10,877][05631] Avg episode rewards: #0: 4.500, true rewards: #0: 4.000
+[2023-02-23 00:51:10,879][05631] Avg episode reward: 4.500, avg true_objective: 4.000
+[2023-02-23 00:51:11,019][05631] Num frames 1700...
+[2023-02-23 00:51:11,141][05631] Num frames 1800...
+[2023-02-23 00:51:11,273][05631] Avg episode rewards: #0: 4.112, true rewards: #0: 3.712
+[2023-02-23 00:51:11,275][05631] Avg episode reward: 4.112, avg true_objective: 3.712
+[2023-02-23 00:51:11,332][05631] Num frames 1900...
+[2023-02-23 00:51:11,459][05631] Num frames 2000...
+[2023-02-23 00:51:11,582][05631] Num frames 2100...
+[2023-02-23 00:51:11,713][05631] Num frames 2200...
+[2023-02-23 00:51:11,817][05631] Avg episode rewards: #0: 4.067, true rewards: #0: 3.733
+[2023-02-23 00:51:11,820][05631] Avg episode reward: 4.067, avg true_objective: 3.733
+[2023-02-23 00:51:11,896][05631] Num frames 2300...
+[2023-02-23 00:51:12,021][05631] Num frames 2400...
+[2023-02-23 00:51:12,141][05631] Num frames 2500...
+[2023-02-23 00:51:12,259][05631] Num frames 2600...
+[2023-02-23 00:51:12,342][05631] Avg episode rewards: #0: 4.034, true rewards: #0: 3.749
+[2023-02-23 00:51:12,344][05631] Avg episode reward: 4.034, avg true_objective: 3.749
+[2023-02-23 00:51:12,452][05631] Num frames 2700...
+[2023-02-23 00:51:12,567][05631] Num frames 2800...
+[2023-02-23 00:51:12,689][05631] Num frames 2900...
+[2023-02-23 00:51:12,803][05631] Num frames 3000...
+[2023-02-23 00:51:12,932][05631] Num frames 3100...
+[2023-02-23 00:51:12,993][05631] Avg episode rewards: #0: 4.255, true rewards: #0: 3.880
+[2023-02-23 00:51:12,995][05631] Avg episode reward: 4.255, avg true_objective: 3.880
+[2023-02-23 00:51:13,113][05631] Num frames 3200...
+[2023-02-23 00:51:13,236][05631] Num frames 3300...
+[2023-02-23 00:51:13,369][05631] Num frames 3400...
+[2023-02-23 00:51:13,526][05631] Avg episode rewards: #0: 4.209, true rewards: #0: 3.876
+[2023-02-23 00:51:13,528][05631] Avg episode reward: 4.209, avg true_objective: 3.876
+[2023-02-23 00:51:13,554][05631] Num frames 3500...
+[2023-02-23 00:51:13,678][05631] Num frames 3600...
+[2023-02-23 00:51:13,801][05631] Num frames 3700...
+[2023-02-23 00:51:13,924][05631] Num frames 3800...
+[2023-02-23 00:51:14,038][05631] Num frames 3900...
+[2023-02-23 00:51:14,135][05631] Avg episode rewards: #0: 4.336, true rewards: #0: 3.936
+[2023-02-23 00:51:14,137][05631] Avg episode reward: 4.336, avg true_objective: 3.936
+[2023-02-23 00:51:35,049][05631] Replay video saved to /content/train_dir/default_experiment/replay.mp4!
+[2023-02-23 00:51:35,237][05631] Loading existing experiment configuration from /content/train_dir/default_experiment/config.json
+[2023-02-23 00:51:35,239][05631] Overriding arg 'num_workers' with value 1 passed from command line
+[2023-02-23 00:51:35,246][05631] Adding new argument 'no_render'=True that is not in the saved config file!
+[2023-02-23 00:51:35,248][05631] Adding new argument 'save_video'=True that is not in the saved config file!
+[2023-02-23 00:51:35,250][05631] Adding new argument 'video_frames'=1000000000.0 that is not in the saved config file!
+[2023-02-23 00:51:35,252][05631] Adding new argument 'video_name'=None that is not in the saved config file!
+[2023-02-23 00:51:35,257][05631] Adding new argument 'max_num_frames'=100000 that is not in the saved config file!
+[2023-02-23 00:51:35,258][05631] Adding new argument 'max_num_episodes'=10 that is not in the saved config file!
+[2023-02-23 00:51:35,259][05631] Adding new argument 'push_to_hub'=True that is not in the saved config file!
+[2023-02-23 00:51:35,261][05631] Adding new argument 'hf_repository'='pittawat/rl_course_vizdoom_health_gathering_supreme' that is not in the saved config file!
+[2023-02-23 00:51:35,262][05631] Adding new argument 'policy_index'=0 that is not in the saved config file!
+[2023-02-23 00:51:35,263][05631] Adding new argument 'eval_deterministic'=False that is not in the saved config file!
+[2023-02-23 00:51:35,269][05631] Adding new argument 'train_script'=None that is not in the saved config file!
+[2023-02-23 00:51:35,270][05631] Adding new argument 'enjoy_script'=None that is not in the saved config file!
+[2023-02-23 00:51:35,271][05631] Using frameskip 1 and render_action_repeat=4 for evaluation
+[2023-02-23 00:51:35,306][05631] RunningMeanStd input shape: (3, 72, 128)
+[2023-02-23 00:51:35,308][05631] RunningMeanStd input shape: (1,)
+[2023-02-23 00:51:35,336][05631] ConvEncoder: input_channels=3
+[2023-02-23 00:51:35,400][05631] Conv encoder output size: 512
+[2023-02-23 00:51:35,402][05631] Policy head output size: 512
+[2023-02-23 00:51:35,434][05631] Loading state from checkpoint /content/train_dir/default_experiment/checkpoint_p0/checkpoint_000002447_10022912.pth...
+[2023-02-23 00:51:36,290][05631] Num frames 100...
+[2023-02-23 00:51:36,467][05631] Num frames 200...
+[2023-02-23 00:51:36,651][05631] Num frames 300...
+[2023-02-23 00:51:36,865][05631] Avg episode rewards: #0: 3.840, true rewards: #0: 3.840
+[2023-02-23 00:51:36,868][05631] Avg episode reward: 3.840, avg true_objective: 3.840
+[2023-02-23 00:51:36,908][05631] Num frames 400...
+[2023-02-23 00:51:37,124][05631] Num frames 500...
+[2023-02-23 00:51:37,321][05631] Num frames 600...
+[2023-02-23 00:51:37,519][05631] Num frames 700...
+[2023-02-23 00:51:37,712][05631] Num frames 800...
+[2023-02-23 00:51:37,896][05631] Num frames 900...
+[2023-02-23 00:51:38,009][05631] Avg episode rewards: #0: 5.640, true rewards: #0: 4.640
+[2023-02-23 00:51:38,012][05631] Avg episode reward: 5.640, avg true_objective: 4.640
+[2023-02-23 00:51:38,141][05631] Num frames 1000...
+[2023-02-23 00:51:38,301][05631] Num frames 1100...
+[2023-02-23 00:51:38,469][05631] Num frames 1200...
+[2023-02-23 00:51:38,603][05631] Num frames 1300...
+[2023-02-23 00:51:38,674][05631] Avg episode rewards: #0: 5.040, true rewards: #0: 4.373
+[2023-02-23 00:51:38,675][05631] Avg episode reward: 5.040, avg true_objective: 4.373
+[2023-02-23 00:51:38,785][05631] Num frames 1400...
+[2023-02-23 00:51:38,906][05631] Num frames 1500...
+[2023-02-23 00:51:39,030][05631] Num frames 1600...
+[2023-02-23 00:51:39,208][05631] Avg episode rewards: #0: 4.740, true rewards: #0: 4.240
+[2023-02-23 00:51:39,210][05631] Avg episode reward: 4.740, avg true_objective: 4.240
+[2023-02-23 00:51:39,221][05631] Num frames 1700...
+[2023-02-23 00:51:39,334][05631] Num frames 1800...
+[2023-02-23 00:51:39,452][05631] Num frames 1900...
+[2023-02-23 00:51:39,574][05631] Num frames 2000...
+[2023-02-23 00:51:39,721][05631] Avg episode rewards: #0: 4.560, true rewards: #0: 4.160
+[2023-02-23 00:51:39,723][05631] Avg episode reward: 4.560, avg true_objective: 4.160
+[2023-02-23 00:51:39,752][05631] Num frames 2100...
+[2023-02-23 00:51:39,874][05631] Num frames 2200...
+[2023-02-23 00:51:39,995][05631] Num frames 2300...
+[2023-02-23 00:51:40,113][05631] Num frames 2400...
+[2023-02-23 00:51:40,283][05631] Avg episode rewards: #0: 4.493, true rewards: #0: 4.160
+[2023-02-23 00:51:40,286][05631] Avg episode reward: 4.493, avg true_objective: 4.160
+[2023-02-23 00:51:40,295][05631] Num frames 2500...
+[2023-02-23 00:51:40,408][05631] Num frames 2600...
+[2023-02-23 00:51:40,526][05631] Num frames 2700...
+[2023-02-23 00:51:40,641][05631] Num frames 2800...
+[2023-02-23 00:51:40,787][05631] Avg episode rewards: #0: 4.400, true rewards: #0: 4.114
+[2023-02-23 00:51:40,789][05631] Avg episode reward: 4.400, avg true_objective: 4.114
+[2023-02-23 00:51:40,818][05631] Num frames 2900...
+[2023-02-23 00:51:40,943][05631] Num frames 3000...
+[2023-02-23 00:51:41,058][05631] Num frames 3100...
+[2023-02-23 00:51:41,172][05631] Num frames 3200...
+[2023-02-23 00:51:41,286][05631] Num frames 3300...
+[2023-02-23 00:51:41,458][05631] Avg episode rewards: #0: 4.740, true rewards: #0: 4.240
+[2023-02-23 00:51:41,459][05631] Avg episode reward: 4.740, avg true_objective: 4.240
+[2023-02-23 00:51:41,473][05631] Num frames 3400...
+[2023-02-23 00:51:41,594][05631] Num frames 3500...
+[2023-02-23 00:51:41,710][05631] Num frames 3600...
+[2023-02-23 00:51:41,824][05631] Num frames 3700...
+[2023-02-23 00:51:41,944][05631] Num frames 3800...
+[2023-02-23 00:51:42,010][05631] Avg episode rewards: #0: 4.787, true rewards: #0: 4.231
+[2023-02-23 00:51:42,013][05631] Avg episode reward: 4.787, avg true_objective: 4.231
+[2023-02-23 00:51:42,125][05631] Num frames 3900...
+[2023-02-23 00:51:42,253][05631] Num frames 4000...
+[2023-02-23 00:51:42,370][05631] Num frames 4100...
+[2023-02-23 00:51:42,528][05631] Avg episode rewards: #0: 4.692, true rewards: #0: 4.192
+[2023-02-23 00:51:42,532][05631] Avg episode reward: 4.692, avg true_objective: 4.192
+[2023-02-23 00:52:02,869][05631] Replay video saved to /content/train_dir/default_experiment/replay.mp4!