diff --git a/.gitattributes b/.gitattributes
index 7cc4348f11476b28fbd9d47f4adebb36851608e4..9b29fb88c16360011ee5fcc72b50b050fabbeffd 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -274,3 +274,4 @@ factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:
 factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs2:obs8p75:stride4:seed14/episode_metrics.jsonl filter=lfs diff=lfs merge=lfs -text
 factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed11/episode_metrics.jsonl filter=lfs diff=lfs merge=lfs -text
 factor_sweeps/flappy/context_window/train/factor_sweep:flappy:frame_stack:fixed_l2:fs1:obs30:stride1:seed14/episode_metrics.jsonl filter=lfs diff=lfs merge=lfs -text
+factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/episode_metrics.jsonl filter=lfs diff=lfs merge=lfs -text
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/best_000047292_24213504_reward_1951.365.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/best_000047292_24213504_reward_1951.365.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d7146e372abf42199c445b558efe8a5d6daac04f
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/best_000047292_24213504_reward_1951.365.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78d0129e64b110e977384f61847d42ac632a03f636cc898231daaaa996248f2d
+size 40643436
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000033944_17379328.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000033944_17379328.pth
new file mode 100644
index 0000000000000000000000000000000000000000..19095325369b3605efd4a284c9c93dd66282806a
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000033944_17379328.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bda462b61b291091122b2a2509a7edbe37bc1ce0ede5c982546a6f1ba4eb78c
+size 40643874
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000037824_19365888.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000037824_19365888.pth
new file mode 100644
index 0000000000000000000000000000000000000000..da8a2cc7b5c25dbc1e280ae56ef54851f8d1cc3d
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000037824_19365888.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cb7d03b064a32933a0d6f29831625560c5ac2ea5c5478a83a0c2a6648c4478d
+size 40643874
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000041464_21229568.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000041464_21229568.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4ef9f7d55bf1638dc998c19b0c173f6a091b9afa
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000041464_21229568.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1762a7d496e74aeecb2a65d3408e437a719ec9037ccdb43e336d90077327a1b8
+size 40643874
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000045736_23416832.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000045736_23416832.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fc4233e83fbe40f9ec14f5545f8d3ba201bc170a
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000045736_23416832.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e6eca4cb41105a4c65e91521e1b55603f3fb97883d36fd5c3e3217f5c253113
+size 40643874
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000048836_25004032.pth b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000048836_25004032.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1904517e3faf966e7f8dd129e4c385ab500b7eee
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/checkpoint_p0/checkpoint_000048836_25004032.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a20f911c8f67418ae04087792bc117b06bf9047248e583724ef4f8c66b8be0f7
+size 40643874
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/config.json b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c519b7ce15f2e7b698d02eb406df89347a6fc056
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/config.json
@@ -0,0 +1,263 @@
+{
+  "help": false,
+  "algo": "APPO",
+  "env": "doom_deadly_corridor",
+  "experiment": "deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10",
+  "train_dir": "results/checkpoints_factor_sweeps/deadly_corridor/context_window",
+  "restart_behavior": "resume",
+  "device": "gpu",
+  "seed": 10,
+  "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": 32,
+  "num_envs_per_worker": 4,
+  "batch_size": 1024,
+  "num_batches_per_epoch": 2,
+  "num_epochs": 2,
+  "rollout": 128,
+  "recurrence": 128,
+  "shuffle_minibatches": false,
+  "gamma": 0.99,
+  "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": null,
+  "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": 1,
+  "flush_summaries_interval": 30,
+  "stats_avg": 100,
+  "summaries_use_frameskip": true,
+  "heartbeat_interval": 20,
+  "heartbeat_reporting_interval": 600,
+  "train_for_env_steps": 25000000,
+  "train_for_seconds": 10000000000,
+  "save_every_sec": 600,
+  "keep_checkpoints": 5,
+  "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,
+  "with_wandb": true,
+  "wandb_user": null,
+  "wandb_project": "latency-sensitive-bench",
+  "wandb_group": "deadly-corridor-fs12-uniform_u2_6",
+  "wandb_job_type": "sample_factory",
+  "wandb_tags": [
+    "factor_sweep",
+    "deadly-corridor",
+    "frame_stack",
+    "uniform",
+    "uniform_u2_6",
+    "fs12",
+    "seed10"
+  ],
+  "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,
+  "fps": 35,
+  "eval_env_frameskip": 1,
+  "no_render": false,
+  "save_video": false,
+  "video_frames": 1000000000.0,
+  "video_name": null,
+  "max_num_frames": 1000000000.0,
+  "max_num_episodes": 1000000000.0,
+  "push_to_hub": false,
+  "hf_repository": null,
+  "policy_index": 0,
+  "eval_deterministic": true,
+  "train_script": null,
+  "enjoy_script": null,
+  "num_agents": -1,
+  "num_humans": 0,
+  "num_bots": -1,
+  "start_bot_difficulty": null,
+  "timelimit": null,
+  "res_w": 128,
+  "res_h": 72,
+  "wide_aspect_ratio": false,
+  "record_to": null,
+  "mode": "train",
+  "env_fps": 35.0,
+  "obs_fps": 8.75,
+  "latency_type": "uniform_distribution",
+  "fixed_latency_ms": null,
+  "mean_latency_ms": null,
+  "std_latency_ms": null,
+  "min_latency_ms": 57.142857142857146,
+  "max_latency_ms": 171.42857142857144,
+  "latency_seed": 10,
+  "add_latency_info": false,
+  "max_pending_actions": null,
+  "hold_policy": "hold",
+  "ordering_policy": "latest_ready",
+  "eval_episodes": 100,
+  "eval_parallel_envs": 32,
+  "eval_latency_raw_frame_values": "0,1,2,3,4,5,10,15,40",
+  "eval_max_steps": 30000,
+  "eval_raw_reward": true,
+  "disable_reward_scaling": true,
+  "frame_stack": 12,
+  "max_episode_steps": 0,
+  "simulator": "cpu",
+  "envpool_batch_size": 1,
+  "episode_metrics_path": "results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/episode_metrics.jsonl",
+  "vizdoom_init_max_parallel": 10,
+  "vizdoom_init_lock_timeout_sec": 5.0,
+  "vizdoom_init_lock_max_wait_sec": 60.0,
+  "vizdoom_init_retry_jitter_sec": 0.25,
+  "command_line": "--mode train --algo APPO --env doom_deadly_corridor --experiment deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10 --train_dir results/checkpoints_factor_sweeps/deadly_corridor/context_window --restart_behavior resume --device gpu --seed 10 --episode_metrics_path results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/episode_metrics.jsonl --train_for_env_steps 25000000 --num_workers 32 --num_envs_per_worker 4 --num_policies 1 --batch_size 1024 --rollout 128 --recurrence 128 --num_epochs 2 --num_batches_per_epoch 2 --worker_num_splits 2 --learning_rate 0.0001 --gamma 0.99 --gae_lambda 0.95 --ppo_clip_ratio 0.1 --ppo_clip_value 0.2 --exploration_loss symmetric_kl --exploration_loss_coeff 0.001 --value_loss_coeff 0.5 --max_grad_norm 4.0 --rnn_type gru --rnn_size 512 --save_every_sec 600 --keep_checkpoints 5 --stats_avg 100 --experiment_summaries_interval 1 --async_rl True --use_rnn True --normalize_returns True --normalize_input True --latency-type uniform_distribution --min-latency-ms 57.142857142857146 --max-latency-ms 171.42857142857144 --latency-seed 10 --add-latency-info False --eval-episodes 100 --eval-parallel-envs 32 --eval-max-steps 30000 --eval-deterministic True --eval-raw-reward --with_wandb True --wandb_project latency-sensitive-bench --wandb_group deadly-corridor-fs12-uniform_u2_6 --wandb_job_type sample_factory --wandb_tags factor_sweep deadly-corridor frame_stack uniform uniform_u2_6 fs12 seed10 --env-fps 35 --obs-fps 8.75 --res_w 128 --res_h 72 --wide_aspect_ratio False --frame_stack 12 --simulator cpu --disable-reward-scaling True --hold-policy hold --ordering-policy latest_ready",
+  "cli_args": {
+    "algo": "APPO",
+    "env": "doom_deadly_corridor",
+    "experiment": "deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10",
+    "train_dir": "results/checkpoints_factor_sweeps/deadly_corridor/context_window",
+    "restart_behavior": "resume",
+    "device": "gpu",
+    "seed": 10,
+    "num_policies": 1,
+    "async_rl": true,
+    "worker_num_splits": 2,
+    "num_workers": 32,
+    "num_envs_per_worker": 4,
+    "batch_size": 1024,
+    "num_batches_per_epoch": 2,
+    "num_epochs": 2,
+    "rollout": 128,
+    "recurrence": 128,
+    "gamma": 0.99,
+    "normalize_returns": true,
+    "exploration_loss_coeff": 0.001,
+    "value_loss_coeff": 0.5,
+    "exploration_loss": "symmetric_kl",
+    "gae_lambda": 0.95,
+    "ppo_clip_ratio": 0.1,
+    "ppo_clip_value": 0.2,
+    "max_grad_norm": 4.0,
+    "learning_rate": 0.0001,
+    "normalize_input": true,
+    "experiment_summaries_interval": 1,
+    "stats_avg": 100,
+    "train_for_env_steps": 25000000,
+    "save_every_sec": 600,
+    "keep_checkpoints": 5,
+    "use_rnn": true,
+    "rnn_size": 512,
+    "rnn_type": "gru",
+    "with_wandb": true,
+    "wandb_project": "latency-sensitive-bench",
+    "wandb_group": "deadly-corridor-fs12-uniform_u2_6",
+    "wandb_job_type": "sample_factory",
+    "wandb_tags": [
+      "factor_sweep",
+      "deadly-corridor",
+      "frame_stack",
+      "uniform",
+      "uniform_u2_6",
+      "fs12",
+      "seed10"
+    ],
+    "eval_deterministic": true,
+    "res_w": 128,
+    "res_h": 72,
+    "wide_aspect_ratio": false,
+    "mode": "train",
+    "env_fps": 35.0,
+    "obs_fps": 8.75,
+    "latency_type": "uniform_distribution",
+    "min_latency_ms": 57.142857142857146,
+    "max_latency_ms": 171.42857142857144,
+    "latency_seed": 10,
+    "add_latency_info": false,
+    "hold_policy": "hold",
+    "ordering_policy": "latest_ready",
+    "eval_episodes": 100,
+    "eval_parallel_envs": 32,
+    "eval_max_steps": 30000,
+    "eval_raw_reward": true,
+    "disable_reward_scaling": true,
+    "frame_stack": 12,
+    "simulator": "cpu",
+    "episode_metrics_path": "results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/episode_metrics.jsonl",
+    "env_frameskip": 1,
+    "eval_env_frameskip": 1
+  },
+  "git_hash": "eb3a2e1efbd2aa03a60d7f44f5e18d8fdd0f5a2d",
+  "git_repo_name": "git@github.com:ZihanWang314/latency-sensitive-bench.git",
+  "output_dir": "outputs/factor_sweeps/deadly_corridor/context_window/train/frame_stack/uniform_u2_6/fs12/seed_10",
+  "wandb_unique_id": "deadly-corridor-fs12-uniform_u2_6-s10"
+}
\ No newline at end of file
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/episode_metrics.jsonl b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/episode_metrics.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0042c4e02da1c59f86e8eefbcf01cd9ba76f07ad
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/episode_metrics.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3be5c0aefc972cffaba7a9e364997d50878bcc1a3bddf90b62e80605af5dea95
+size 164717983
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/git.diff b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/git.diff
new file mode 100644
index 0000000000000000000000000000000000000000..29108e5b29266817ecd22d4ca51bcbe8f6aaa23b
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/git.diff
@@ -0,0 +1,7 @@
+diff --git a/starVLA b/starVLA
+index ab3380d..9d8c567 160000
+--- a/starVLA
++++ b/starVLA
+@@ -1 +1 @@
+-Subproject commit ab3380dfbd1de9649c15d154cc41b97788674537
++Subproject commit 9d8c567188a3aa2a825296016cf17f3977101d8f
diff --git a/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/sf_log.txt b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/sf_log.txt
new file mode 100644
index 0000000000000000000000000000000000000000..abce440dac1063e6f5132714e00a1c05d711f7b7
--- /dev/null
+++ b/factor_sweeps/deadly_corridor/context_window/train/factor_sweep:deadly_corridor:frame_stack:uniform_u2_6:fs12:obs8p75:stride4:seed10/sf_log.txt
@@ -0,0 +1,7928 @@
+[2026-06-06 20:49:20,634][261293] Saving configuration to results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/config.json...
+[2026-06-06 20:49:20,701][261293] Rollout worker 0 uses device cpu
+[2026-06-06 20:49:20,702][261293] Rollout worker 1 uses device cpu
+[2026-06-06 20:49:20,702][261293] Rollout worker 2 uses device cpu
+[2026-06-06 20:49:20,702][261293] Rollout worker 3 uses device cpu
+[2026-06-06 20:49:20,702][261293] Rollout worker 4 uses device cpu
+[2026-06-06 20:49:20,702][261293] Rollout worker 5 uses device cpu
+[2026-06-06 20:49:20,703][261293] Rollout worker 6 uses device cpu
+[2026-06-06 20:49:20,703][261293] Rollout worker 7 uses device cpu
+[2026-06-06 20:49:20,703][261293] Rollout worker 8 uses device cpu
+[2026-06-06 20:49:20,703][261293] Rollout worker 9 uses device cpu
+[2026-06-06 20:49:20,704][261293] Rollout worker 10 uses device cpu
+[2026-06-06 20:49:20,704][261293] Rollout worker 11 uses device cpu
+[2026-06-06 20:49:20,705][261293] Rollout worker 12 uses device cpu
+[2026-06-06 20:49:20,705][261293] Rollout worker 13 uses device cpu
+[2026-06-06 20:49:20,706][261293] Rollout worker 14 uses device cpu
+[2026-06-06 20:49:20,706][261293] Rollout worker 15 uses device cpu
+[2026-06-06 20:49:20,707][261293] Rollout worker 16 uses device cpu
+[2026-06-06 20:49:20,707][261293] Rollout worker 17 uses device cpu
+[2026-06-06 20:49:20,707][261293] Rollout worker 18 uses device cpu
+[2026-06-06 20:49:20,708][261293] Rollout worker 19 uses device cpu
+[2026-06-06 20:49:20,708][261293] Rollout worker 20 uses device cpu
+[2026-06-06 20:49:20,709][261293] Rollout worker 21 uses device cpu
+[2026-06-06 20:49:20,709][261293] Rollout worker 22 uses device cpu
+[2026-06-06 20:49:20,710][261293] Rollout worker 23 uses device cpu
+[2026-06-06 20:49:20,710][261293] Rollout worker 24 uses device cpu
+[2026-06-06 20:49:20,710][261293] Rollout worker 25 uses device cpu
+[2026-06-06 20:49:20,710][261293] Rollout worker 26 uses device cpu
+[2026-06-06 20:49:20,711][261293] Rollout worker 27 uses device cpu
+[2026-06-06 20:49:20,711][261293] Rollout worker 28 uses device cpu
+[2026-06-06 20:49:20,711][261293] Rollout worker 29 uses device cpu
+[2026-06-06 20:49:20,711][261293] Rollout worker 30 uses device cpu
+[2026-06-06 20:49:20,712][261293] Rollout worker 31 uses device cpu
+[2026-06-06 20:49:27,014][261293] Using GPUs [0] for process 0 (actually maps to GPUs [6])
+[2026-06-06 20:49:27,015][261293] InferenceWorker_p0-w0: min num requests: 10
+[2026-06-06 20:49:27,200][261293] Starting all processes...
+[2026-06-06 20:49:27,201][261293] Starting process learner_proc0
+[2026-06-06 20:49:29,502][261293] Starting all processes...
+[2026-06-06 20:49:29,511][261293] Starting process inference_proc0-0
+[2026-06-06 20:49:29,511][261293] Starting process rollout_proc0
+[2026-06-06 20:49:29,511][261293] Starting process rollout_proc1
+[2026-06-06 20:49:29,512][261293] Starting process rollout_proc2
+[2026-06-06 20:49:29,533][262115] Using GPUs [0] for process 0 (actually maps to GPUs [6])
+[2026-06-06 20:49:29,533][262115] Set environment var CUDA_VISIBLE_DEVICES to '6' (GPU indices [0]) for learning process 0
+[2026-06-06 20:49:29,512][261293] Starting process rollout_proc3
+[2026-06-06 20:49:29,512][261293] Starting process rollout_proc4
+[2026-06-06 20:49:29,512][261293] Starting process rollout_proc5
+[2026-06-06 20:49:29,521][261293] Starting process rollout_proc6
+[2026-06-06 20:49:29,523][261293] Starting process rollout_proc7
+[2026-06-06 20:49:29,524][261293] Starting process rollout_proc8
+[2026-06-06 20:49:29,531][261293] Starting process rollout_proc9
+[2026-06-06 20:49:29,533][261293] Starting process rollout_proc10
+[2026-06-06 20:49:29,533][261293] Starting process rollout_proc11
+[2026-06-06 20:49:29,534][261293] Starting process rollout_proc12
+[2026-06-06 20:49:29,535][261293] Starting process rollout_proc13
+[2026-06-06 20:49:29,535][261293] Starting process rollout_proc14
+[2026-06-06 20:49:29,724][262115] Num visible devices: 1
+[2026-06-06 20:49:29,727][262115] Setting fixed seed 10
+[2026-06-06 20:49:29,730][262115] Using GPUs [0] for process 0 (actually maps to GPUs [6])
+[2026-06-06 20:49:29,730][262115] Initializing actor-critic model on device cuda:0
+[2026-06-06 20:49:29,731][262115] RunningMeanStd input shape: (36, 72, 128)
+[2026-06-06 20:49:29,776][262115] RunningMeanStd input shape: (1,)
+[2026-06-06 20:49:29,794][262115] ConvEncoder: input_channels=36
+[2026-06-06 20:49:29,905][262115] Conv encoder output size: 512
+[2026-06-06 20:49:29,905][262115] Policy head output size: 512
+[2026-06-06 20:49:29,917][262115] Created Actor Critic model with architecture:
+[2026-06-06 20:49:29,917][262115] 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=11, bias=True)
+  )
+)
+[2026-06-06 20:49:30,289][262115] Using optimizer <class 'torch.optim.adam.Adam'>
+[2026-06-06 20:49:31,860][261293] Starting process rollout_proc15
+[2026-06-06 20:49:31,863][262199] Worker 0 uses CPU cores [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]
+[2026-06-06 20:49:31,930][261293] Starting process rollout_proc16
+[2026-06-06 20:49:31,932][262205] Worker 5 uses CPU cores [60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71]
+[2026-06-06 20:49:32,073][262115] No checkpoints found
+[2026-06-06 20:49:32,073][262115] Did not load from checkpoint, starting from scratch!
+[2026-06-06 20:49:32,074][262115] Initialized policy 0 weights for model version 0
+[2026-06-06 20:49:32,076][262115] LearnerWorker_p0 finished initialization!
+[2026-06-06 20:49:32,076][262115] Using GPUs [0] for process 0 (actually maps to GPUs [6])
+[2026-06-06 20:49:32,111][261293] Starting process rollout_proc17
+[2026-06-06 20:49:32,113][262203] Worker 3 uses CPU cores [36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47]
+[2026-06-06 20:49:32,286][261293] Starting process rollout_proc18
+[2026-06-06 20:49:32,289][262201] Worker 8 uses CPU cores [96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107]
+[2026-06-06 20:49:32,328][261293] Starting process rollout_proc19
+[2026-06-06 20:49:32,331][262198] Using GPUs [0] for process 0 (actually maps to GPUs [6])
+[2026-06-06 20:49:32,331][262198] Set environment var CUDA_VISIBLE_DEVICES to '6' (GPU indices [0]) for inference process 0
+[2026-06-06 20:49:32,348][261293] Starting process rollout_proc20
+[2026-06-06 20:49:32,350][262212] Worker 11 uses CPU cores [132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143]
+[2026-06-06 20:49:32,367][261293] Starting process rollout_proc21
+[2026-06-06 20:49:32,369][262208] Worker 10 uses CPU cores [120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131]
+[2026-06-06 20:49:32,376][261293] Starting process rollout_proc22
+[2026-06-06 20:49:32,379][262214] Worker 14 uses CPU cores [168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179]
+[2026-06-06 20:49:32,391][261293] Starting process rollout_proc23
+[2026-06-06 20:49:32,393][262213] Worker 13 uses CPU cores [156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167]
+[2026-06-06 20:49:32,409][261293] Starting process rollout_proc24
+[2026-06-06 20:49:32,412][262210] Worker 6 uses CPU cores [72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83]
+[2026-06-06 20:49:32,418][262211] Worker 12 uses CPU cores [144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155]
+[2026-06-06 20:49:32,418][261293] Starting process rollout_proc25
+[2026-06-06 20:49:32,419][262209] Worker 7 uses CPU cores [84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95]
+[2026-06-06 20:49:32,418][261293] Starting process rollout_proc26
+[2026-06-06 20:49:32,425][262204] Worker 9 uses CPU cores [108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119]
+[2026-06-06 20:49:32,430][261293] Starting process rollout_proc27
+[2026-06-06 20:49:32,470][261293] Starting process rollout_proc28
+[2026-06-06 20:49:32,472][262206] Worker 4 uses CPU cores [48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59]
+[2026-06-06 20:49:32,480][261293] Starting process rollout_proc29
+[2026-06-06 20:49:32,481][262200] Worker 1 uses CPU cores [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]
+[2026-06-06 20:49:32,510][261293] Starting process rollout_proc30
+[2026-06-06 20:49:32,512][262202] Worker 2 uses CPU cores [24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35]
+[2026-06-06 20:49:32,530][262198] Num visible devices: 1
+[2026-06-06 20:49:32,871][262198] RunningMeanStd input shape: (36, 72, 128)
+[2026-06-06 20:49:32,906][262198] RunningMeanStd input shape: (1,)
+[2026-06-06 20:49:32,919][262198] ConvEncoder: input_channels=36
+[2026-06-06 20:49:33,049][262198] Conv encoder output size: 512
+[2026-06-06 20:49:33,049][262198] Policy head output size: 512
+[2026-06-06 20:49:34,158][261293] Starting process rollout_proc31
+[2026-06-06 20:49:34,163][263226] Worker 15 uses CPU cores [180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191]
+[2026-06-06 20:49:34,331][263227] Worker 16 uses CPU cores [192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203]
+[2026-06-06 20:49:34,427][263228] Worker 17 uses CPU cores [204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215]
+[2026-06-06 20:49:34,604][263230] Worker 19 uses CPU cores [228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239]
+[2026-06-06 20:49:34,697][263229] Worker 18 uses CPU cores [216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227]
+[2026-06-06 20:49:34,917][263232] Worker 21 uses CPU cores [252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263]
+[2026-06-06 20:49:34,983][263233] Worker 22 uses CPU cores [264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275]
+[2026-06-06 20:49:34,989][263242] Worker 30 uses CPU cores [360, 361, 362, 363, 364, 365, 366, 367, 368, 369, 370, 371]
+[2026-06-06 20:49:34,990][263235] Worker 23 uses CPU cores [276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287]
+[2026-06-06 20:49:35,005][263236] Worker 24 uses CPU cores [288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299]
+[2026-06-06 20:49:35,010][263238] Worker 26 uses CPU cores [312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323]
+[2026-06-06 20:49:35,017][263239] Worker 27 uses CPU cores [324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335]
+[2026-06-06 20:49:35,033][263231] Worker 20 uses CPU cores [240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251]
+[2026-06-06 20:49:35,052][263237] Worker 25 uses CPU cores [300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311]
+[2026-06-06 20:49:35,199][263241] Worker 29 uses CPU cores [348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359]
+[2026-06-06 20:49:35,268][263240] Worker 28 uses CPU cores [336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347]
+[2026-06-06 20:49:36,139][264253] Worker 31 uses CPU cores [372, 373, 374, 375, 376, 377, 378, 379, 380, 381, 382, 383]
+[2026-06-06 20:49:36,143][261293] Inference worker 0-0 is ready!
+[2026-06-06 20:49:36,143][261293] All inference workers are ready! Signal rollout workers to start!
+[2026-06-06 20:49:36,145][261293] Fps is (10 sec: nan, 60 sec: nan, 300 sec: nan). Total num frames: 0. Throughput: 0: nan. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2026-06-06 20:49:36,174][264253] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,175][262212] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,176][263239] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,176][263238] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,177][263231] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,177][262206] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,177][263242] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,177][263229] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,177][263232] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,178][262208] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,179][262203] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,180][262214] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,180][263228] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,181][263230] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,182][262200] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,182][263227] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,183][262210] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,183][262202] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,184][262213] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,184][262204] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,184][263237] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,184][263241] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,185][263240] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,187][263233] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,188][263235] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,188][262199] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,189][262201] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,190][263236] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,192][262211] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,194][263226] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,197][262205] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:36,197][262209] Doom resolution: 160x120, resize resolution: (128, 72)
+[2026-06-06 20:49:37,217][264253] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,221][263239] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,278][263229] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,280][262203] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,295][263238] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,298][262213] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,302][263240] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,305][263230] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,305][263227] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,307][262200] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,912][262214] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,920][262208] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:37,920][261293] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 0. Throughput: 0: 0.0. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2026-06-06 20:49:38,157][263242] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:38,221][262203] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,259][263238] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,263][263230] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,265][263227] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,265][262205] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:38,266][263229] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,302][263236] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:38,354][262200] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,597][263242] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:38,641][263241] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:38,890][263237] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,048][263226] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,077][262213] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,080][262214] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,091][263239] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,109][262205] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,110][263232] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,111][262206] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,152][262199] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,309][262204] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,610][263226] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,713][262203] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:39,794][262211] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,843][263238] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:39,911][263232] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:39,916][263242] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:39,933][262202] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,950][263231] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:39,971][263237] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:40,132][262213] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:40,310][262211] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:40,328][262209] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:40,414][262201] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:40,598][263239] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:40,601][262214] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:40,612][264253] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:40,615][263226] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:40,628][263228] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:40,662][263235] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:40,754][262202] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:40,813][262201] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,250][263233] Another process currently holds the lock /tmp/sf2_root/doom_006.lockfile, attempt: 1, max_wait_sec=60.000
+[2026-06-06 20:49:41,253][262210] Another process currently holds the lock /tmp/sf2_root/doom_003.lockfile, attempt: 1, max_wait_sec=60.000
+[2026-06-06 20:49:41,333][263242] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:41,366][262200] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:41,548][263238] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:41,580][262212] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:41,591][263235] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,598][263231] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,610][262199] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,611][263230] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:41,614][263228] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,692][263237] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:41,801][262206] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:41,846][262202] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:42,170][263226] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:42,295][262203] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:42,320][263240] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:42,412][263241] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:42,418][262205] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:42,443][263233] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:42,462][262209] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:42,507][263229] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:42,529][262201] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:42,534][263235] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:42,846][262208] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:42,903][263239] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:42,920][261293] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 0. Throughput: 0: 0.0. Samples: 0. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2026-06-06 20:49:42,952][263237] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,210][263231] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:43,274][262204] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:43,276][262200] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,300][262212] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:43,302][263236] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:43,307][262202] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,390][263230] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,391][263233] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:43,480][262205] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,696][262201] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:43,740][264253] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,034][262213] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:44,145][263232] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,167][262206] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,183][262210] Decorrelating experience for 0 frames...
+[2026-06-06 20:49:44,184][263229] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:44,209][262209] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,229][263235] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:44,235][263227] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,262][263233] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,737][262214] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:44,819][262212] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,941][262204] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,948][263231] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:44,974][262208] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,979][263241] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:44,980][262210] Decorrelating experience for 128 frames...
+[2026-06-06 20:49:45,197][262211] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:45,369][263233] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:45,437][263232] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:45,486][264253] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:45,502][263240] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:45,606][262206] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:45,773][262210] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:45,844][262209] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:45,925][262212] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,049][262204] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,135][263227] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,232][263241] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,264][262211] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,347][262208] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,536][263240] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:46,615][262199] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:46,781][262210] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:47,001][261293] Heartbeat connected on Batcher_0
+[2026-06-06 20:49:47,007][261293] Heartbeat connected on LearnerWorker_p0
+[2026-06-06 20:49:47,028][263236] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:47,029][261293] Heartbeat connected on RolloutWorker_w1
+[2026-06-06 20:49:47,035][261293] Heartbeat connected on RolloutWorker_w2
+[2026-06-06 20:49:47,042][261293] Heartbeat connected on RolloutWorker_w3
+[2026-06-06 20:49:47,049][261293] Heartbeat connected on RolloutWorker_w4
+[2026-06-06 20:49:47,054][261293] Heartbeat connected on RolloutWorker_w5
+[2026-06-06 20:49:47,067][261293] Heartbeat connected on RolloutWorker_w7
+[2026-06-06 20:49:47,072][261293] Heartbeat connected on RolloutWorker_w8
+[2026-06-06 20:49:47,078][261293] Heartbeat connected on RolloutWorker_w9
+[2026-06-06 20:49:47,088][261293] Heartbeat connected on RolloutWorker_w11
+[2026-06-06 20:49:47,098][261293] Heartbeat connected on RolloutWorker_w13
+[2026-06-06 20:49:47,103][261293] Heartbeat connected on RolloutWorker_w14
+[2026-06-06 20:49:47,108][261293] Heartbeat connected on RolloutWorker_w15
+[2026-06-06 20:49:47,126][261293] Heartbeat connected on RolloutWorker_w18
+[2026-06-06 20:49:47,130][261293] Heartbeat connected on RolloutWorker_w19
+[2026-06-06 20:49:47,135][261293] Heartbeat connected on RolloutWorker_w20
+[2026-06-06 20:49:47,141][261293] Heartbeat connected on RolloutWorker_w21
+[2026-06-06 20:49:47,152][261293] Heartbeat connected on RolloutWorker_w22
+[2026-06-06 20:49:47,156][261293] Heartbeat connected on InferenceWorker_p0-w0
+[2026-06-06 20:49:47,158][261293] Heartbeat connected on RolloutWorker_w23
+[2026-06-06 20:49:47,163][261293] Heartbeat connected on RolloutWorker_w25
+[2026-06-06 20:49:47,171][261293] Heartbeat connected on RolloutWorker_w12
+[2026-06-06 20:49:47,177][261293] Heartbeat connected on RolloutWorker_w27
+[2026-06-06 20:49:47,181][261293] Heartbeat connected on RolloutWorker_w26
+[2026-06-06 20:49:47,187][261293] Heartbeat connected on RolloutWorker_w29
+[2026-06-06 20:49:47,192][261293] Heartbeat connected on RolloutWorker_w30
+[2026-06-06 20:49:47,197][261293] Heartbeat connected on RolloutWorker_w31
+[2026-06-06 20:49:47,304][261293] Heartbeat connected on RolloutWorker_w16
+[2026-06-06 20:49:47,352][261293] Heartbeat connected on RolloutWorker_w10
+[2026-06-06 20:49:47,441][263228] Decorrelating experience for 256 frames...
+[2026-06-06 20:49:47,475][261293] Heartbeat connected on RolloutWorker_w28
+[2026-06-06 20:49:47,661][261293] Heartbeat connected on RolloutWorker_w6
+[2026-06-06 20:49:47,753][262199] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:47,921][261293] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 0. Throughput: 0: 130.9. Samples: 1542. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2026-06-06 20:49:47,923][261293] Avg episode reward: [(0, '-61.446')]
+[2026-06-06 20:49:48,015][263236] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:48,411][263228] Decorrelating experience for 384 frames...
+[2026-06-06 20:49:48,739][261293] Heartbeat connected on RolloutWorker_w0
+[2026-06-06 20:49:48,886][261293] Heartbeat connected on RolloutWorker_w24
+[2026-06-06 20:49:49,373][261293] Heartbeat connected on RolloutWorker_w17
+[2026-06-06 20:49:51,900][262115] Signal inference workers to stop experience collection...
+[2026-06-06 20:49:51,932][262198] InferenceWorker_p0-w0: stopping experience collection
+[2026-06-06 20:49:52,920][261293] Fps is (10 sec: 0.0, 60 sec: 0.0, 300 sec: 0.0). Total num frames: 0. Throughput: 0: 704.8. Samples: 11824. Policy #0 lag: (min: -1.0, avg: -1.0, max: -1.0)
+[2026-06-06 20:49:52,922][261293] Avg episode reward: [(0, '-58.956')]
+[2026-06-06 20:49:53,593][262115] Signal inference workers to resume experience collection...
+[2026-06-06 20:49:53,595][262198] InferenceWorker_p0-w0: resuming experience collection
+[2026-06-06 20:49:56,497][262198] Updated weights for policy 0, policy_version 24 (0.0056)
+[2026-06-06 20:49:57,920][261293] Fps is (10 sec: 1638.5, 60 sec: 752.4, 300 sec: 752.4). Total num frames: 16384. Throughput: 0: 968.0. Samples: 21078. Policy #0 lag: (min: 5.0, avg: 15.1, max: 17.0)
+[2026-06-06 20:49:57,921][261293] Avg episode reward: [(0, '-51.771')]
+[2026-06-06 20:49:59,810][262198] Updated weights for policy 0, policy_version 37 (0.0011)
+[2026-06-06 20:50:02,039][262198] Updated weights for policy 0, policy_version 56 (0.0011)
+[2026-06-06 20:50:02,920][261293] Fps is (10 sec: 3276.8, 60 sec: 1223.8, 300 sec: 1223.8). Total num frames: 32768. Throughput: 0: 1199.3. Samples: 32112. Policy #0 lag: (min: 10.0, avg: 18.2, max: 42.0)
+[2026-06-06 20:50:02,921][261293] Avg episode reward: [(0, '-16.436')]
+[2026-06-06 20:50:03,748][262198] Updated weights for policy 0, policy_version 66 (0.0010)
+[2026-06-06 20:50:05,667][262198] Updated weights for policy 0, policy_version 83 (0.0010)
+[2026-06-06 20:50:07,920][261293] Fps is (10 sec: 3276.7, 60 sec: 1546.9, 300 sec: 1546.9). Total num frames: 49152. Throughput: 0: 1649.0. Samples: 52398. Policy #0 lag: (min: 3.0, avg: 23.9, max: 35.0)
+[2026-06-06 20:50:07,923][261293] Avg episode reward: [(0, '-11.219')]
+[2026-06-06 20:50:08,225][262198] Updated weights for policy 0, policy_version 98 (0.0009)
+[2026-06-06 20:50:10,148][262198] Updated weights for policy 0, policy_version 113 (0.0010)
+[2026-06-06 20:50:11,327][262198] Updated weights for policy 0, policy_version 128 (0.0009)
+[2026-06-06 20:50:12,921][261293] Fps is (10 sec: 3686.3, 60 sec: 1893.4, 300 sec: 1893.4). Total num frames: 69632. Throughput: 0: 2103.0. Samples: 77340. Policy #0 lag: (min: 0.0, avg: 17.2, max: 32.0)
+[2026-06-06 20:50:12,922][261293] Avg episode reward: [(0, '28.661')]
+[2026-06-06 20:50:13,954][262198] Updated weights for policy 0, policy_version 144 (0.0011)
+[2026-06-06 20:50:15,578][262198] Updated weights for policy 0, policy_version 160 (0.0011)
+[2026-06-06 20:50:17,920][261293] Fps is (10 sec: 3891.2, 60 sec: 2108.0, 300 sec: 2108.0). Total num frames: 88064. Throughput: 0: 2097.8. Samples: 87636. Policy #0 lag: (min: 3.0, avg: 17.0, max: 35.0)
+[2026-06-06 20:50:17,922][261293] Avg episode reward: [(0, '47.568')]
+[2026-06-06 20:50:17,925][262198] Updated weights for policy 0, policy_version 173 (0.0010)
+[2026-06-06 20:50:19,054][262198] Updated weights for policy 0, policy_version 187 (0.0015)
+[2026-06-06 20:50:21,139][262198] Updated weights for policy 0, policy_version 197 (0.0010)
+[2026-06-06 20:50:22,921][261293] Fps is (10 sec: 3891.2, 60 sec: 2320.5, 300 sec: 2320.5). Total num frames: 108544. Throughput: 0: 2513.6. Samples: 113112. Policy #0 lag: (min: 1.0, avg: 15.5, max: 35.0)
+[2026-06-06 20:50:22,923][261293] Avg episode reward: [(0, '42.848')]
+[2026-06-06 20:50:22,980][262198] Updated weights for policy 0, policy_version 213 (0.0011)
+[2026-06-06 20:50:23,092][262115] Saving new best policy, reward=42.848!
+[2026-06-06 20:50:24,821][262198] Updated weights for policy 0, policy_version 225 (0.0009)
+[2026-06-06 20:50:26,224][262198] Updated weights for policy 0, policy_version 240 (0.0011)
+[2026-06-06 20:50:27,746][262198] Updated weights for policy 0, policy_version 256 (0.0010)
+[2026-06-06 20:50:27,921][261293] Fps is (10 sec: 4300.5, 60 sec: 2531.5, 300 sec: 2531.5). Total num frames: 131072. Throughput: 0: 3025.9. Samples: 136168. Policy #0 lag: (min: 3.0, avg: 24.8, max: 37.0)
+[2026-06-06 20:50:27,923][261293] Avg episode reward: [(0, '90.282')]
+[2026-06-06 20:50:27,929][262115] Saving new best policy, reward=90.282!
+[2026-06-06 20:50:29,683][262198] Updated weights for policy 0, policy_version 266 (0.0010)
+[2026-06-06 20:50:30,925][262198] Updated weights for policy 0, policy_version 282 (0.0011)
+[2026-06-06 20:50:32,539][262198] Updated weights for policy 0, policy_version 292 (0.0009)
+[2026-06-06 20:50:32,920][261293] Fps is (10 sec: 4096.2, 60 sec: 2633.3, 300 sec: 2633.3). Total num frames: 149504. Throughput: 0: 3316.2. Samples: 150768. Policy #0 lag: (min: 7.0, avg: 20.1, max: 39.0)
+[2026-06-06 20:50:32,922][261293] Avg episode reward: [(0, '117.139')]
+[2026-06-06 20:50:32,988][262115] Saving new best policy, reward=117.139!
+[2026-06-06 20:50:33,885][262198] Updated weights for policy 0, policy_version 303 (0.0011)
+[2026-06-06 20:50:35,400][262198] Updated weights for policy 0, policy_version 317 (0.0011)
+[2026-06-06 20:50:36,941][262198] Updated weights for policy 0, policy_version 327 (0.0010)
+[2026-06-06 20:50:37,920][261293] Fps is (10 sec: 3891.4, 60 sec: 2833.1, 300 sec: 2751.7). Total num frames: 169984. Throughput: 0: 3625.8. Samples: 174984. Policy #0 lag: (min: 0.0, avg: 23.4, max: 40.0)
+[2026-06-06 20:50:37,923][261293] Avg episode reward: [(0, '147.547')]
+[2026-06-06 20:50:38,062][262115] Saving new best policy, reward=147.547!
+[2026-06-06 20:50:38,884][262198] Updated weights for policy 0, policy_version 345 (0.0010)
+[2026-06-06 20:50:40,317][262198] Updated weights for policy 0, policy_version 355 (0.0008)
+[2026-06-06 20:50:41,843][262198] Updated weights for policy 0, policy_version 368 (0.0010)
+[2026-06-06 20:50:42,920][261293] Fps is (10 sec: 4505.6, 60 sec: 3242.7, 300 sec: 2913.7). Total num frames: 194560. Throughput: 0: 3999.5. Samples: 201054. Policy #0 lag: (min: 13.0, avg: 27.1, max: 45.0)
+[2026-06-06 20:50:42,922][261293] Avg episode reward: [(0, '176.866')]
+[2026-06-06 20:50:43,014][262198] Updated weights for policy 0, policy_version 383 (0.0009)
+[2026-06-06 20:50:43,074][262115] Saving new best policy, reward=176.866!
+[2026-06-06 20:50:45,207][262198] Updated weights for policy 0, policy_version 396 (0.0009)
+[2026-06-06 20:50:45,810][262115] Signal inference workers to stop experience collection... (50 times)
+[2026-06-06 20:50:45,904][262115] Signal inference workers to resume experience collection... (50 times)
+[2026-06-06 20:50:45,961][262198] InferenceWorker_p0-w0: stopping experience collection (50 times)
+[2026-06-06 20:50:45,961][262198] InferenceWorker_p0-w0: resuming experience collection (50 times)
+[2026-06-06 20:50:46,519][262198] Updated weights for policy 0, policy_version 411 (0.0011)
+[2026-06-06 20:50:47,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3549.9, 300 sec: 2967.5). Total num frames: 212992. Throughput: 0: 4055.7. Samples: 214620. Policy #0 lag: (min: 7.0, avg: 20.0, max: 39.0)
+[2026-06-06 20:50:47,922][261293] Avg episode reward: [(0, '181.702')]
+[2026-06-06 20:50:47,926][262115] Saving new best policy, reward=181.702!
+[2026-06-06 20:50:48,512][262198] Updated weights for policy 0, policy_version 421 (0.0010)
+[2026-06-06 20:50:50,408][262198] Updated weights for policy 0, policy_version 440 (0.0011)
+[2026-06-06 20:50:51,932][262198] Updated weights for policy 0, policy_version 450 (0.0009)
+[2026-06-06 20:50:52,920][261293] Fps is (10 sec: 3891.1, 60 sec: 3891.2, 300 sec: 3041.0). Total num frames: 233472. Throughput: 0: 4120.5. Samples: 237822. Policy #0 lag: (min: 3.0, avg: 23.5, max: 35.0)
+[2026-06-06 20:50:52,922][261293] Avg episode reward: [(0, '221.929')]
+[2026-06-06 20:50:53,202][262115] Saving new best policy, reward=221.929!
+[2026-06-06 20:50:53,882][262198] Updated weights for policy 0, policy_version 465 (0.0009)
+[2026-06-06 20:50:55,063][262198] Updated weights for policy 0, policy_version 479 (0.0009)
+[2026-06-06 20:50:57,119][262198] Updated weights for policy 0, policy_version 489 (0.0010)
+[2026-06-06 20:50:57,920][261293] Fps is (10 sec: 4300.8, 60 sec: 3993.6, 300 sec: 3130.5). Total num frames: 256000. Throughput: 0: 4143.1. Samples: 263778. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 20:50:57,922][261293] Avg episode reward: [(0, '255.039')]
+[2026-06-06 20:50:58,187][262115] Saving new best policy, reward=255.039!
+[2026-06-06 20:50:58,437][262198] Updated weights for policy 0, policy_version 507 (0.0010)
+[2026-06-06 20:51:00,172][262198] Updated weights for policy 0, policy_version 517 (0.0010)
+[2026-06-06 20:51:01,717][262198] Updated weights for policy 0, policy_version 532 (0.0011)
+[2026-06-06 20:51:02,831][262198] Updated weights for policy 0, policy_version 544 (0.0009)
+[2026-06-06 20:51:02,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4096.0, 300 sec: 3209.8). Total num frames: 278528. Throughput: 0: 4188.3. Samples: 276110. Policy #0 lag: (min: 7.0, avg: 20.4, max: 43.0)
+[2026-06-06 20:51:02,922][261293] Avg episode reward: [(0, '246.174')]
+[2026-06-06 20:51:04,987][262198] Updated weights for policy 0, policy_version 556 (0.0010)
+[2026-06-06 20:51:06,643][262198] Updated weights for policy 0, policy_version 574 (0.0013)
+[2026-06-06 20:51:07,920][261293] Fps is (10 sec: 4095.9, 60 sec: 4130.1, 300 sec: 3235.7). Total num frames: 296960. Throughput: 0: 4160.5. Samples: 300332. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 20:51:07,923][261293] Avg episode reward: [(0, '227.735')]
+[2026-06-06 20:51:07,990][262198] Updated weights for policy 0, policy_version 584 (0.0010)
+[2026-06-06 20:51:10,006][262198] Updated weights for policy 0, policy_version 603 (0.0010)
+[2026-06-06 20:51:11,582][262198] Updated weights for policy 0, policy_version 613 (0.0010)
+[2026-06-06 20:51:12,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4164.3, 300 sec: 3301.3). Total num frames: 319488. Throughput: 0: 4266.0. Samples: 328136. Policy #0 lag: (min: 6.0, avg: 26.7, max: 46.0)
+[2026-06-06 20:51:12,923][261293] Avg episode reward: [(0, '290.873')]
+[2026-06-06 20:51:13,196][262115] Saving new best policy, reward=290.873!
+[2026-06-06 20:51:13,620][262198] Updated weights for policy 0, policy_version 632 (0.0013)
+[2026-06-06 20:51:15,444][262198] Updated weights for policy 0, policy_version 642 (0.0011)
+[2026-06-06 20:51:17,084][262198] Updated weights for policy 0, policy_version 654 (0.0011)
+[2026-06-06 20:51:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4198.4, 300 sec: 3340.4). Total num frames: 339968. Throughput: 0: 4175.7. Samples: 338676. Policy #0 lag: (min: 1.0, avg: 17.1, max: 34.0)
+[2026-06-06 20:51:17,922][261293] Avg episode reward: [(0, '263.421')]
+[2026-06-06 20:51:18,046][262198] Updated weights for policy 0, policy_version 668 (0.0010)
+[2026-06-06 20:51:19,734][262198] Updated weights for policy 0, policy_version 678 (0.0011)
+[2026-06-06 20:51:21,631][262198] Updated weights for policy 0, policy_version 693 (0.0009)
+[2026-06-06 20:51:22,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4198.4, 300 sec: 3375.8). Total num frames: 360448. Throughput: 0: 4200.8. Samples: 364022. Policy #0 lag: (min: 10.0, avg: 21.0, max: 42.0)
+[2026-06-06 20:51:22,922][261293] Avg episode reward: [(0, '256.893')]
+[2026-06-06 20:51:23,017][262198] Updated weights for policy 0, policy_version 705 (0.0010)
+[2026-06-06 20:51:24,759][262198] Updated weights for policy 0, policy_version 720 (0.0010)
+[2026-06-06 20:51:26,139][262198] Updated weights for policy 0, policy_version 734 (0.0010)
+[2026-06-06 20:51:27,280][262198] Updated weights for policy 0, policy_version 744 (0.0010)
+[2026-06-06 20:51:27,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4164.3, 300 sec: 3408.0). Total num frames: 380928. Throughput: 0: 4212.3. Samples: 390608. Policy #0 lag: (min: 3.0, avg: 19.4, max: 35.0)
+[2026-06-06 20:51:27,924][261293] Avg episode reward: [(0, '301.988')]
+[2026-06-06 20:51:28,088][262115] Saving new best policy, reward=301.988!
+[2026-06-06 20:51:28,983][262115] Signal inference workers to stop experience collection... (100 times)
+[2026-06-06 20:51:29,009][262198] InferenceWorker_p0-w0: stopping experience collection (100 times)
+[2026-06-06 20:51:29,127][262115] Signal inference workers to resume experience collection... (100 times)
+[2026-06-06 20:51:29,128][262198] InferenceWorker_p0-w0: resuming experience collection (100 times)
+[2026-06-06 20:51:29,131][262198] Updated weights for policy 0, policy_version 760 (0.0011)
+[2026-06-06 20:51:30,651][262198] Updated weights for policy 0, policy_version 770 (0.0008)
+[2026-06-06 20:51:32,256][262198] Updated weights for policy 0, policy_version 784 (0.0009)
+[2026-06-06 20:51:32,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4232.5, 300 sec: 3455.0). Total num frames: 403456. Throughput: 0: 4187.6. Samples: 403064. Policy #0 lag: (min: 10.0, avg: 21.8, max: 42.0)
+[2026-06-06 20:51:32,922][261293] Avg episode reward: [(0, '319.439')]
+[2026-06-06 20:51:33,192][262115] Saving new best policy, reward=319.439!
+[2026-06-06 20:51:33,668][262198] Updated weights for policy 0, policy_version 797 (0.0011)
+[2026-06-06 20:51:34,903][262198] Updated weights for policy 0, policy_version 807 (0.0010)
+[2026-06-06 20:51:36,786][262198] Updated weights for policy 0, policy_version 824 (0.0011)
+[2026-06-06 20:51:37,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4266.6, 300 sec: 3498.1). Total num frames: 425984. Throughput: 0: 4244.4. Samples: 428820. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 20:51:37,924][261293] Avg episode reward: [(0, '321.493')]
+[2026-06-06 20:51:37,930][262115] Saving new best policy, reward=321.493!
+[2026-06-06 20:51:38,294][262198] Updated weights for policy 0, policy_version 834 (0.0009)
+[2026-06-06 20:51:39,871][262198] Updated weights for policy 0, policy_version 849 (0.0011)
+[2026-06-06 20:51:40,567][262198] Updated weights for policy 0, policy_version 859 (0.0010)
+[2026-06-06 20:51:42,401][262198] Updated weights for policy 0, policy_version 869 (0.0010)
+[2026-06-06 20:51:42,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4198.4, 300 sec: 3521.7). Total num frames: 446464. Throughput: 0: 4262.5. Samples: 455590. Policy #0 lag: (min: 3.0, avg: 22.1, max: 35.0)
+[2026-06-06 20:51:42,922][261293] Avg episode reward: [(0, '351.306')]
+[2026-06-06 20:51:43,229][262115] Saving new best policy, reward=351.306!
+[2026-06-06 20:51:44,051][262198] Updated weights for policy 0, policy_version 886 (0.0013)
+[2026-06-06 20:51:45,202][262198] Updated weights for policy 0, policy_version 896 (0.0013)
+[2026-06-06 20:51:47,149][262198] Updated weights for policy 0, policy_version 908 (0.0011)
+[2026-06-06 20:51:47,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4266.6, 300 sec: 3559.0). Total num frames: 468992. Throughput: 0: 4255.4. Samples: 467604. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 20:51:47,922][261293] Avg episode reward: [(0, '303.212')]
+[2026-06-06 20:51:48,993][262198] Updated weights for policy 0, policy_version 925 (0.0011)
+[2026-06-06 20:51:50,408][262198] Updated weights for policy 0, policy_version 935 (0.0010)
+[2026-06-06 20:51:52,315][262198] Updated weights for policy 0, policy_version 954 (0.0009)
+[2026-06-06 20:51:52,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4266.7, 300 sec: 3578.6). Total num frames: 489472. Throughput: 0: 4284.3. Samples: 493128. Policy #0 lag: (min: 3.0, avg: 18.4, max: 35.0)
+[2026-06-06 20:51:52,924][261293] Avg episode reward: [(0, '345.263')]
+[2026-06-06 20:51:53,797][262198] Updated weights for policy 0, policy_version 964 (0.0009)
+[2026-06-06 20:51:56,032][262198] Updated weights for policy 0, policy_version 984 (0.0010)
+[2026-06-06 20:51:57,430][262198] Updated weights for policy 0, policy_version 994 (0.0010)
+[2026-06-06 20:51:57,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4232.5, 300 sec: 3596.9). Total num frames: 509952. Throughput: 0: 4230.9. Samples: 518526. Policy #0 lag: (min: 3.0, avg: 22.5, max: 35.0)
+[2026-06-06 20:51:57,924][261293] Avg episode reward: [(0, '289.605')]
+[2026-06-06 20:51:58,934][262198] Updated weights for policy 0, policy_version 1005 (0.0009)
+[2026-06-06 20:52:00,570][262198] Updated weights for policy 0, policy_version 1022 (0.0011)
+[2026-06-06 20:52:02,079][262198] Updated weights for policy 0, policy_version 1032 (0.0012)
+[2026-06-06 20:52:02,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4198.4, 300 sec: 3613.9). Total num frames: 530432. Throughput: 0: 4271.3. Samples: 530888. Policy #0 lag: (min: 10.0, avg: 23.5, max: 43.0)
+[2026-06-06 20:52:02,923][261293] Avg episode reward: [(0, '334.375')]
+[2026-06-06 20:52:03,759][262198] Updated weights for policy 0, policy_version 1050 (0.0010)
+[2026-06-06 20:52:05,073][262198] Updated weights for policy 0, policy_version 1060 (0.0009)
+[2026-06-06 20:52:06,868][262198] Updated weights for policy 0, policy_version 1073 (0.0010)
+[2026-06-06 20:52:07,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4300.8, 300 sec: 3656.8). Total num frames: 555008. Throughput: 0: 4314.3. Samples: 558164. Policy #0 lag: (min: 2.0, avg: 19.7, max: 38.0)
+[2026-06-06 20:52:07,923][261293] Avg episode reward: [(0, '348.829')]
+[2026-06-06 20:52:08,004][262198] Updated weights for policy 0, policy_version 1086 (0.0009)
+[2026-06-06 20:52:09,431][262198] Updated weights for policy 0, policy_version 1096 (0.0009)
+[2026-06-06 20:52:11,156][262198] Updated weights for policy 0, policy_version 1115 (0.0010)
+[2026-06-06 20:52:12,920][261293] Fps is (10 sec: 4505.9, 60 sec: 4266.7, 300 sec: 3670.8). Total num frames: 575488. Throughput: 0: 4296.6. Samples: 583952. Policy #0 lag: (min: 3.0, avg: 22.3, max: 35.0)
+[2026-06-06 20:52:12,923][261293] Avg episode reward: [(0, '410.955')]
+[2026-06-06 20:52:12,929][262115] Saving new best policy, reward=410.955!
+[2026-06-06 20:52:13,158][262198] Updated weights for policy 0, policy_version 1125 (0.0009)
+[2026-06-06 20:52:14,503][262198] Updated weights for policy 0, policy_version 1142 (0.0011)
+[2026-06-06 20:52:15,530][262198] Updated weights for policy 0, policy_version 1152 (0.0009)
+[2026-06-06 20:52:17,379][262115] Signal inference workers to stop experience collection... (150 times)
+[2026-06-06 20:52:17,407][262198] InferenceWorker_p0-w0: stopping experience collection (150 times)
+[2026-06-06 20:52:17,488][262115] Signal inference workers to resume experience collection... (150 times)
+[2026-06-06 20:52:17,489][262198] InferenceWorker_p0-w0: resuming experience collection (150 times)
+[2026-06-06 20:52:17,491][262198] Updated weights for policy 0, policy_version 1164 (0.0010)
+[2026-06-06 20:52:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4300.8, 300 sec: 3696.6). Total num frames: 598016. Throughput: 0: 4304.0. Samples: 596744. Policy #0 lag: (min: 3.0, avg: 18.2, max: 39.0)
+[2026-06-06 20:52:17,922][261293] Avg episode reward: [(0, '394.533')]
+[2026-06-06 20:52:18,848][262198] Updated weights for policy 0, policy_version 1179 (0.0009)
+[2026-06-06 20:52:20,494][262198] Updated weights for policy 0, policy_version 1189 (0.0010)
+[2026-06-06 20:52:21,997][262198] Updated weights for policy 0, policy_version 1206 (0.0009)
+[2026-06-06 20:52:22,886][262198] Updated weights for policy 0, policy_version 1216 (0.0010)
+[2026-06-06 20:52:22,920][261293] Fps is (10 sec: 4710.3, 60 sec: 4369.1, 300 sec: 3733.1). Total num frames: 622592. Throughput: 0: 4334.6. Samples: 623876. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 20:52:22,922][261293] Avg episode reward: [(0, '400.719')]
+[2026-06-06 20:52:24,811][262198] Updated weights for policy 0, policy_version 1228 (0.0010)
+[2026-06-06 20:52:26,050][262198] Updated weights for policy 0, policy_version 1242 (0.0010)
+[2026-06-06 20:52:27,222][262198] Updated weights for policy 0, policy_version 1252 (0.0009)
+[2026-06-06 20:52:27,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4334.9, 300 sec: 3731.8). Total num frames: 641024. Throughput: 0: 4332.9. Samples: 650572. Policy #0 lag: (min: 1.0, avg: 20.5, max: 34.0)
+[2026-06-06 20:52:27,922][261293] Avg episode reward: [(0, '359.691')]
+[2026-06-06 20:52:29,310][262198] Updated weights for policy 0, policy_version 1270 (0.0009)
+[2026-06-06 20:52:30,476][262198] Updated weights for policy 0, policy_version 1280 (0.0010)
+[2026-06-06 20:52:32,264][262198] Updated weights for policy 0, policy_version 1292 (0.0009)
+[2026-06-06 20:52:32,921][261293] Fps is (10 sec: 4095.7, 60 sec: 4334.9, 300 sec: 3753.6). Total num frames: 663552. Throughput: 0: 4344.0. Samples: 663084. Policy #0 lag: (min: 6.0, avg: 19.5, max: 38.0)
+[2026-06-06 20:52:32,923][261293] Avg episode reward: [(0, '417.097')]
+[2026-06-06 20:52:33,198][262115] Saving new best policy, reward=417.097!
+[2026-06-06 20:52:33,764][262198] Updated weights for policy 0, policy_version 1307 (0.0010)
+[2026-06-06 20:52:35,677][262198] Updated weights for policy 0, policy_version 1317 (0.0012)
+[2026-06-06 20:52:37,141][262198] Updated weights for policy 0, policy_version 1336 (0.0010)
+[2026-06-06 20:52:37,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4334.9, 300 sec: 3774.3). Total num frames: 686080. Throughput: 0: 4361.7. Samples: 689402. Policy #0 lag: (min: 3.0, avg: 17.8, max: 35.0)
+[2026-06-06 20:52:37,923][261293] Avg episode reward: [(0, '390.198')]
+[2026-06-06 20:52:38,657][262198] Updated weights for policy 0, policy_version 1346 (0.0009)
+[2026-06-06 20:52:40,435][262198] Updated weights for policy 0, policy_version 1360 (0.0010)
+[2026-06-06 20:52:41,824][262198] Updated weights for policy 0, policy_version 1373 (0.0010)
+[2026-06-06 20:52:42,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4334.9, 300 sec: 3782.9). Total num frames: 706560. Throughput: 0: 4357.8. Samples: 714626. Policy #0 lag: (min: 3.0, avg: 22.0, max: 37.0)
+[2026-06-06 20:52:42,922][261293] Avg episode reward: [(0, '426.892')]
+[2026-06-06 20:52:42,932][262115] Saving new best policy, reward=426.892!
+[2026-06-06 20:52:43,256][262198] Updated weights for policy 0, policy_version 1383 (0.0009)
+[2026-06-06 20:52:44,772][262198] Updated weights for policy 0, policy_version 1400 (0.0010)
+[2026-06-06 20:52:46,223][262198] Updated weights for policy 0, policy_version 1410 (0.0009)
+[2026-06-06 20:52:47,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4300.8, 300 sec: 3791.1). Total num frames: 727040. Throughput: 0: 4361.6. Samples: 727156. Policy #0 lag: (min: 3.0, avg: 23.6, max: 39.0)
+[2026-06-06 20:52:47,923][261293] Avg episode reward: [(0, '390.135')]
+[2026-06-06 20:52:48,066][262198] Updated weights for policy 0, policy_version 1424 (0.0010)
+[2026-06-06 20:52:49,235][262198] Updated weights for policy 0, policy_version 1435 (0.0010)
+[2026-06-06 20:52:50,865][262198] Updated weights for policy 0, policy_version 1445 (0.0009)
+[2026-06-06 20:52:52,522][262198] Updated weights for policy 0, policy_version 1464 (0.0010)
+[2026-06-06 20:52:52,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4335.0, 300 sec: 3809.3). Total num frames: 749568. Throughput: 0: 4337.1. Samples: 753332. Policy #0 lag: (min: 7.0, avg: 20.0, max: 39.0)
+[2026-06-06 20:52:52,922][261293] Avg episode reward: [(0, '404.801')]
+[2026-06-06 20:52:54,010][262198] Updated weights for policy 0, policy_version 1474 (0.0008)
+[2026-06-06 20:52:55,661][262198] Updated weights for policy 0, policy_version 1488 (0.0009)
+[2026-06-06 20:52:57,022][262198] Updated weights for policy 0, policy_version 1501 (0.0009)
+[2026-06-06 20:52:57,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4369.1, 300 sec: 3826.5). Total num frames: 772096. Throughput: 0: 4333.0. Samples: 778940. Policy #0 lag: (min: 3.0, avg: 17.9, max: 36.0)
+[2026-06-06 20:52:57,922][261293] Avg episode reward: [(0, '465.764')]
+[2026-06-06 20:52:57,929][262115] Saving new best policy, reward=465.764!
+[2026-06-06 20:52:58,574][262198] Updated weights for policy 0, policy_version 1511 (0.0010)
+[2026-06-06 20:52:59,882][262198] Updated weights for policy 0, policy_version 1526 (0.0010)
+[2026-06-06 20:53:01,003][262198] Updated weights for policy 0, policy_version 1536 (0.0009)
+[2026-06-06 20:53:02,779][262198] Updated weights for policy 0, policy_version 1548 (0.0010)
+[2026-06-06 20:53:02,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4369.1, 300 sec: 3833.0). Total num frames: 792576. Throughput: 0: 4327.2. Samples: 791470. Policy #0 lag: (min: 3.0, avg: 17.1, max: 35.0)
+[2026-06-06 20:53:02,924][261293] Avg episode reward: [(0, '415.947')]
+[2026-06-06 20:53:04,343][262198] Updated weights for policy 0, policy_version 1563 (0.0010)
+[2026-06-06 20:53:05,846][262198] Updated weights for policy 0, policy_version 1573 (0.0009)
+[2026-06-06 20:53:07,075][262198] Updated weights for policy 0, policy_version 1588 (0.0010)
+[2026-06-06 20:53:07,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4334.9, 300 sec: 3848.9). Total num frames: 815104. Throughput: 0: 4339.3. Samples: 819146. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 20:53:07,922][261293] Avg episode reward: [(0, '440.259')]
+[2026-06-06 20:53:08,339][262198] Updated weights for policy 0, policy_version 1598 (0.0010)
+[2026-06-06 20:53:09,719][262198] Updated weights for policy 0, policy_version 1608 (0.0010)
+[2026-06-06 20:53:11,224][262198] Updated weights for policy 0, policy_version 1624 (0.0010)
+[2026-06-06 20:53:11,960][262115] Signal inference workers to stop experience collection... (200 times)
+[2026-06-06 20:53:11,974][262115] Signal inference workers to resume experience collection... (200 times)
+[2026-06-06 20:53:11,975][262198] InferenceWorker_p0-w0: stopping experience collection (200 times)
+[2026-06-06 20:53:11,993][262198] InferenceWorker_p0-w0: resuming experience collection (200 times)
+[2026-06-06 20:53:12,686][262198] Updated weights for policy 0, policy_version 1634 (0.0009)
+[2026-06-06 20:53:12,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4369.0, 300 sec: 3864.0). Total num frames: 837632. Throughput: 0: 4309.8. Samples: 844512. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 20:53:12,925][261293] Avg episode reward: [(0, '426.513')]
+[2026-06-06 20:53:14,276][262198] Updated weights for policy 0, policy_version 1647 (0.0011)
+[2026-06-06 20:53:15,426][262198] Updated weights for policy 0, policy_version 1659 (0.0012)
+[2026-06-06 20:53:16,983][262198] Updated weights for policy 0, policy_version 1669 (0.0010)
+[2026-06-06 20:53:17,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4334.9, 300 sec: 3869.3). Total num frames: 858112. Throughput: 0: 4322.8. Samples: 857610. Policy #0 lag: (min: 3.0, avg: 22.3, max: 39.0)
+[2026-06-06 20:53:17,923][261293] Avg episode reward: [(0, '441.742')]
+[2026-06-06 20:53:18,473][262198] Updated weights for policy 0, policy_version 1686 (0.0009)
+[2026-06-06 20:53:19,639][262198] Updated weights for policy 0, policy_version 1696 (0.0008)
+[2026-06-06 20:53:21,334][262198] Updated weights for policy 0, policy_version 1708 (0.0009)
+[2026-06-06 20:53:22,745][262198] Updated weights for policy 0, policy_version 1723 (0.0010)
+[2026-06-06 20:53:22,920][261293] Fps is (10 sec: 4505.9, 60 sec: 4335.0, 300 sec: 3892.4). Total num frames: 882688. Throughput: 0: 4366.8. Samples: 885904. Policy #0 lag: (min: 3.0, avg: 23.5, max: 39.0)
+[2026-06-06 20:53:22,921][261293] Avg episode reward: [(0, '419.136')]
+[2026-06-06 20:53:24,177][262198] Updated weights for policy 0, policy_version 1733 (0.0009)
+[2026-06-06 20:53:25,738][262198] Updated weights for policy 0, policy_version 1750 (0.0011)
+[2026-06-06 20:53:26,954][262198] Updated weights for policy 0, policy_version 1760 (0.0010)
+[2026-06-06 20:53:27,920][261293] Fps is (10 sec: 4505.8, 60 sec: 4369.1, 300 sec: 3896.7). Total num frames: 903168. Throughput: 0: 4392.2. Samples: 912274. Policy #0 lag: (min: 3.0, avg: 23.5, max: 43.0)
+[2026-06-06 20:53:27,922][261293] Avg episode reward: [(0, '435.248')]
+[2026-06-06 20:53:28,670][262198] Updated weights for policy 0, policy_version 1772 (0.0012)
+[2026-06-06 20:53:29,664][262198] Updated weights for policy 0, policy_version 1784 (0.0010)
+[2026-06-06 20:53:31,286][262198] Updated weights for policy 0, policy_version 1794 (0.0010)
+[2026-06-06 20:53:32,735][262198] Updated weights for policy 0, policy_version 1807 (0.0011)
+[2026-06-06 20:53:32,920][261293] Fps is (10 sec: 4300.6, 60 sec: 4369.1, 300 sec: 3909.6). Total num frames: 925696. Throughput: 0: 4377.0. Samples: 924122. Policy #0 lag: (min: 3.0, avg: 23.3, max: 39.0)
+[2026-06-06 20:53:32,922][261293] Avg episode reward: [(0, '458.385')]
+[2026-06-06 20:53:34,178][262198] Updated weights for policy 0, policy_version 1820 (0.0009)
+[2026-06-06 20:53:35,820][262198] Updated weights for policy 0, policy_version 1830 (0.0009)
+[2026-06-06 20:53:37,146][262198] Updated weights for policy 0, policy_version 1845 (0.0010)
+[2026-06-06 20:53:37,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4369.1, 300 sec: 3921.9). Total num frames: 948224. Throughput: 0: 4386.1. Samples: 950708. Policy #0 lag: (min: 5.0, avg: 21.1, max: 36.0)
+[2026-06-06 20:53:37,924][261293] Avg episode reward: [(0, '436.556')]
+[2026-06-06 20:53:38,339][262198] Updated weights for policy 0, policy_version 1855 (0.0009)
+[2026-06-06 20:53:39,920][262198] Updated weights for policy 0, policy_version 1868 (0.0009)
+[2026-06-06 20:53:40,767][262198] Updated weights for policy 0, policy_version 1879 (0.0009)
+[2026-06-06 20:53:42,487][262198] Updated weights for policy 0, policy_version 1889 (0.0011)
+[2026-06-06 20:53:42,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.1, 300 sec: 3925.4). Total num frames: 968704. Throughput: 0: 4399.7. Samples: 976928. Policy #0 lag: (min: 10.0, avg: 23.1, max: 42.0)
+[2026-06-06 20:53:42,923][261293] Avg episode reward: [(0, '431.989')]
+[2026-06-06 20:53:43,742][262198] Updated weights for policy 0, policy_version 1901 (0.0010)
+[2026-06-06 20:53:45,128][262198] Updated weights for policy 0, policy_version 1913 (0.0009)
+[2026-06-06 20:53:46,206][262198] Updated weights for policy 0, policy_version 1923 (0.0009)
+[2026-06-06 20:53:47,883][262198] Updated weights for policy 0, policy_version 1937 (0.0011)
+[2026-06-06 20:53:47,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4403.2, 300 sec: 3937.0). Total num frames: 991232. Throughput: 0: 4420.1. Samples: 990372. Policy #0 lag: (min: 7.0, avg: 21.7, max: 40.0)
+[2026-06-06 20:53:47,923][261293] Avg episode reward: [(0, '521.860')]
+[2026-06-06 20:53:48,044][262115] Saving new best policy, reward=521.860!
+[2026-06-06 20:53:49,128][262198] Updated weights for policy 0, policy_version 1949 (0.0010)
+[2026-06-06 20:53:50,481][262198] Updated weights for policy 0, policy_version 1959 (0.0009)
+[2026-06-06 20:53:51,671][262198] Updated weights for policy 0, policy_version 1974 (0.0010)
+[2026-06-06 20:53:52,878][262198] Updated weights for policy 0, policy_version 1984 (0.0009)
+[2026-06-06 20:53:52,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4437.3, 300 sec: 3956.0). Total num frames: 1015808. Throughput: 0: 4418.0. Samples: 1017956. Policy #0 lag: (min: 9.0, avg: 22.2, max: 40.0)
+[2026-06-06 20:53:52,924][261293] Avg episode reward: [(0, '468.373')]
+[2026-06-06 20:53:54,444][262198] Updated weights for policy 0, policy_version 1996 (0.0009)
+[2026-06-06 20:53:55,991][262198] Updated weights for policy 0, policy_version 2010 (0.0009)
+[2026-06-06 20:53:57,036][262198] Updated weights for policy 0, policy_version 2020 (0.0010)
+[2026-06-06 20:53:57,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4403.2, 300 sec: 3958.7). Total num frames: 1036288. Throughput: 0: 4466.4. Samples: 1045500. Policy #0 lag: (min: 8.0, avg: 20.9, max: 40.0)
+[2026-06-06 20:53:57,923][261293] Avg episode reward: [(0, '503.386')]
+[2026-06-06 20:53:59,016][262198] Updated weights for policy 0, policy_version 2037 (0.0011)
+[2026-06-06 20:54:00,301][262198] Updated weights for policy 0, policy_version 2047 (0.0010)
+[2026-06-06 20:54:02,056][262198] Updated weights for policy 0, policy_version 2061 (0.0010)
+[2026-06-06 20:54:02,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4471.5, 300 sec: 3976.6). Total num frames: 1060864. Throughput: 0: 4445.4. Samples: 1057652. Policy #0 lag: (min: 3.0, avg: 18.8, max: 39.0)
+[2026-06-06 20:54:02,924][261293] Avg episode reward: [(0, '599.382')]
+[2026-06-06 20:54:02,928][262115] Saving new best policy, reward=599.382!
+[2026-06-06 20:54:03,414][262198] Updated weights for policy 0, policy_version 2073 (0.0010)
+[2026-06-06 20:54:04,453][262198] Updated weights for policy 0, policy_version 2083 (0.0009)
+[2026-06-06 20:54:06,240][262115] Signal inference workers to stop experience collection... (250 times)
+[2026-06-06 20:54:06,264][262198] InferenceWorker_p0-w0: stopping experience collection (250 times)
+[2026-06-06 20:54:06,266][262198] Updated weights for policy 0, policy_version 2097 (0.0010)
+[2026-06-06 20:54:06,370][262115] Signal inference workers to resume experience collection... (250 times)
+[2026-06-06 20:54:06,370][262198] InferenceWorker_p0-w0: resuming experience collection (250 times)
+[2026-06-06 20:54:07,891][262198] Updated weights for policy 0, policy_version 2109 (0.0011)
+[2026-06-06 20:54:07,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4403.2, 300 sec: 3971.3). Total num frames: 1079296. Throughput: 0: 4403.4. Samples: 1084056. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 20:54:07,923][261293] Avg episode reward: [(0, '530.053')]
+[2026-06-06 20:54:09,369][262198] Updated weights for policy 0, policy_version 2120 (0.0010)
+[2026-06-06 20:54:10,528][262198] Updated weights for policy 0, policy_version 2134 (0.0011)
+[2026-06-06 20:54:11,879][262198] Updated weights for policy 0, policy_version 2144 (0.0009)
+[2026-06-06 20:54:12,920][261293] Fps is (10 sec: 3891.3, 60 sec: 4369.1, 300 sec: 3973.5). Total num frames: 1099776. Throughput: 0: 4372.7. Samples: 1109044. Policy #0 lag: (min: 3.0, avg: 19.8, max: 38.0)
+[2026-06-06 20:54:12,923][261293] Avg episode reward: [(0, '548.553')]
+[2026-06-06 20:54:13,579][262198] Updated weights for policy 0, policy_version 2158 (0.0009)
+[2026-06-06 20:54:14,951][262198] Updated weights for policy 0, policy_version 2170 (0.0009)
+[2026-06-06 20:54:16,190][262198] Updated weights for policy 0, policy_version 2180 (0.0009)
+[2026-06-06 20:54:17,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4437.4, 300 sec: 3990.2). Total num frames: 1124352. Throughput: 0: 4404.2. Samples: 1122312. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 20:54:17,923][261293] Avg episode reward: [(0, '534.461')]
+[2026-06-06 20:54:17,935][262198] Updated weights for policy 0, policy_version 2197 (0.0010)
+[2026-06-06 20:54:19,256][262198] Updated weights for policy 0, policy_version 2207 (0.0010)
+[2026-06-06 20:54:20,783][262198] Updated weights for policy 0, policy_version 2218 (0.0010)
+[2026-06-06 20:54:21,667][262198] Updated weights for policy 0, policy_version 2231 (0.0010)
+[2026-06-06 20:54:22,920][261293] Fps is (10 sec: 4505.5, 60 sec: 4369.0, 300 sec: 3992.1). Total num frames: 1144832. Throughput: 0: 4397.5. Samples: 1148592. Policy #0 lag: (min: 3.0, avg: 20.5, max: 39.0)
+[2026-06-06 20:54:22,923][261293] Avg episode reward: [(0, '526.046')]
+[2026-06-06 20:54:23,523][262198] Updated weights for policy 0, policy_version 2241 (0.0009)
+[2026-06-06 20:54:24,708][262198] Updated weights for policy 0, policy_version 2254 (0.0010)
+[2026-06-06 20:54:26,216][262198] Updated weights for policy 0, policy_version 2266 (0.0010)
+[2026-06-06 20:54:27,368][262198] Updated weights for policy 0, policy_version 2276 (0.0009)
+[2026-06-06 20:54:27,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4403.2, 300 sec: 4000.9). Total num frames: 1167360. Throughput: 0: 4415.6. Samples: 1175628. Policy #0 lag: (min: 1.0, avg: 17.1, max: 33.0)
+[2026-06-06 20:54:27,926][261293] Avg episode reward: [(0, '605.540')]
+[2026-06-06 20:54:28,184][262115] Saving new best policy, reward=605.540!
+[2026-06-06 20:54:29,071][262198] Updated weights for policy 0, policy_version 2293 (0.0010)
+[2026-06-06 20:54:30,368][262198] Updated weights for policy 0, policy_version 2303 (0.0009)
+[2026-06-06 20:54:31,975][262198] Updated weights for policy 0, policy_version 2313 (0.0013)
+[2026-06-06 20:54:32,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4403.2, 300 sec: 4033.5). Total num frames: 1189888. Throughput: 0: 4402.2. Samples: 1188470. Policy #0 lag: (min: 1.0, avg: 17.4, max: 34.0)
+[2026-06-06 20:54:32,923][261293] Avg episode reward: [(0, '519.523')]
+[2026-06-06 20:54:33,621][262198] Updated weights for policy 0, policy_version 2330 (0.0011)
+[2026-06-06 20:54:34,930][262198] Updated weights for policy 0, policy_version 2340 (0.0010)
+[2026-06-06 20:54:36,257][262198] Updated weights for policy 0, policy_version 2353 (0.0008)
+[2026-06-06 20:54:37,480][262198] Updated weights for policy 0, policy_version 2364 (0.0009)
+[2026-06-06 20:54:37,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4369.1, 300 sec: 4102.9). Total num frames: 1210368. Throughput: 0: 4365.3. Samples: 1214392. Policy #0 lag: (min: 0.0, avg: 19.1, max: 36.0)
+[2026-06-06 20:54:37,923][261293] Avg episode reward: [(0, '577.298')]
+[2026-06-06 20:54:39,256][262198] Updated weights for policy 0, policy_version 2374 (0.0010)
+[2026-06-06 20:54:40,474][262198] Updated weights for policy 0, policy_version 2388 (0.0011)
+[2026-06-06 20:54:41,821][262198] Updated weights for policy 0, policy_version 2398 (0.0010)
+[2026-06-06 20:54:42,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4369.1, 300 sec: 4172.4). Total num frames: 1230848. Throughput: 0: 4320.4. Samples: 1239914. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 20:54:42,923][261293] Avg episode reward: [(0, '519.123')]
+[2026-06-06 20:54:43,327][262198] Updated weights for policy 0, policy_version 2409 (0.0009)
+[2026-06-06 20:54:44,133][262198] Updated weights for policy 0, policy_version 2420 (0.0009)
+[2026-06-06 20:54:45,413][262198] Updated weights for policy 0, policy_version 2430 (0.0008)
+[2026-06-06 20:54:46,843][262198] Updated weights for policy 0, policy_version 2440 (0.0009)
+[2026-06-06 20:54:47,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4369.1, 300 sec: 4248.7). Total num frames: 1253376. Throughput: 0: 4344.1. Samples: 1253138. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 20:54:47,923][261293] Avg episode reward: [(0, '507.275')]
+[2026-06-06 20:54:48,168][262198] Updated weights for policy 0, policy_version 2456 (0.0010)
+[2026-06-06 20:54:50,024][262198] Updated weights for policy 0, policy_version 2466 (0.0009)
+[2026-06-06 20:54:50,580][262115] Signal inference workers to stop experience collection... (300 times)
+[2026-06-06 20:54:50,594][262115] Signal inference workers to resume experience collection... (300 times)
+[2026-06-06 20:54:50,605][262198] InferenceWorker_p0-w0: stopping experience collection (300 times)
+[2026-06-06 20:54:50,622][262198] InferenceWorker_p0-w0: resuming experience collection (300 times)
+[2026-06-06 20:54:51,121][262198] Updated weights for policy 0, policy_version 2477 (0.0009)
+[2026-06-06 20:54:52,441][262198] Updated weights for policy 0, policy_version 2490 (0.0010)
+[2026-06-06 20:54:52,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4334.9, 300 sec: 4269.6). Total num frames: 1275904. Throughput: 0: 4345.5. Samples: 1279606. Policy #0 lag: (min: 3.0, avg: 18.2, max: 35.0)
+[2026-06-06 20:54:52,922][261293] Avg episode reward: [(0, '480.894')]
+[2026-06-06 20:54:54,016][262198] Updated weights for policy 0, policy_version 2500 (0.0010)
+[2026-06-06 20:54:55,572][262198] Updated weights for policy 0, policy_version 2518 (0.0011)
+[2026-06-06 20:54:56,897][262198] Updated weights for policy 0, policy_version 2528 (0.0009)
+[2026-06-06 20:54:57,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4334.9, 300 sec: 4283.4). Total num frames: 1296384. Throughput: 0: 4366.9. Samples: 1305556. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 20:54:57,923][261293] Avg episode reward: [(0, '500.193')]
+[2026-06-06 20:54:58,510][262198] Updated weights for policy 0, policy_version 2540 (0.0011)
+[2026-06-06 20:55:00,076][262198] Updated weights for policy 0, policy_version 2554 (0.0009)
+[2026-06-06 20:55:01,619][262198] Updated weights for policy 0, policy_version 2564 (0.0009)
+[2026-06-06 20:55:02,658][262198] Updated weights for policy 0, policy_version 2576 (0.0010)
+[2026-06-06 20:55:02,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4300.8, 300 sec: 4304.3). Total num frames: 1318912. Throughput: 0: 4343.3. Samples: 1317762. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 20:55:02,922][261293] Avg episode reward: [(0, '524.930')]
+[2026-06-06 20:55:03,756][262198] Updated weights for policy 0, policy_version 2587 (0.0009)
+[2026-06-06 20:55:05,329][262198] Updated weights for policy 0, policy_version 2597 (0.0010)
+[2026-06-06 20:55:06,253][262198] Updated weights for policy 0, policy_version 2608 (0.0008)
+[2026-06-06 20:55:07,408][262198] Updated weights for policy 0, policy_version 2619 (0.0009)
+[2026-06-06 20:55:07,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4369.0, 300 sec: 4311.2). Total num frames: 1341440. Throughput: 0: 4362.3. Samples: 1344898. Policy #0 lag: (min: 5.0, avg: 18.5, max: 39.0)
+[2026-06-06 20:55:07,922][261293] Avg episode reward: [(0, '521.398')]
+[2026-06-06 20:55:09,178][262198] Updated weights for policy 0, policy_version 2629 (0.0011)
+[2026-06-06 20:55:10,258][262198] Updated weights for policy 0, policy_version 2641 (0.0011)
+[2026-06-06 20:55:11,256][262198] Updated weights for policy 0, policy_version 2651 (0.0009)
+[2026-06-06 20:55:12,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4369.1, 300 sec: 4318.2). Total num frames: 1361920. Throughput: 0: 4334.5. Samples: 1370678. Policy #0 lag: (min: 4.0, avg: 17.2, max: 37.0)
+[2026-06-06 20:55:12,924][261293] Avg episode reward: [(0, '532.956')]
+[2026-06-06 20:55:12,934][262198] Updated weights for policy 0, policy_version 2662 (0.0012)
+[2026-06-06 20:55:13,942][262198] Updated weights for policy 0, policy_version 2673 (0.0010)
+[2026-06-06 20:55:14,851][262198] Updated weights for policy 0, policy_version 2683 (0.0009)
+[2026-06-06 20:55:17,071][262198] Updated weights for policy 0, policy_version 2697 (0.0011)
+[2026-06-06 20:55:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4334.9, 300 sec: 4325.1). Total num frames: 1384448. Throughput: 0: 4347.3. Samples: 1384098. Policy #0 lag: (min: 4.0, avg: 16.4, max: 36.0)
+[2026-06-06 20:55:17,922][261293] Avg episode reward: [(0, '618.515')]
+[2026-06-06 20:55:18,134][262198] Updated weights for policy 0, policy_version 2711 (0.0009)
+[2026-06-06 20:55:18,164][262115] Saving new best policy, reward=618.515!
+[2026-06-06 20:55:20,209][262198] Updated weights for policy 0, policy_version 2721 (0.0010)
+[2026-06-06 20:55:21,287][262198] Updated weights for policy 0, policy_version 2733 (0.0009)
+[2026-06-06 20:55:22,022][262198] Updated weights for policy 0, policy_version 2744 (0.0009)
+[2026-06-06 20:55:22,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4369.1, 300 sec: 4325.1). Total num frames: 1406976. Throughput: 0: 4343.1. Samples: 1409830. Policy #0 lag: (min: 7.0, avg: 19.1, max: 39.0)
+[2026-06-06 20:55:22,922][261293] Avg episode reward: [(0, '624.811')]
+[2026-06-06 20:55:22,928][262115] Saving new best policy, reward=624.811!
+[2026-06-06 20:55:24,188][262198] Updated weights for policy 0, policy_version 2756 (0.0011)
+[2026-06-06 20:55:25,225][262198] Updated weights for policy 0, policy_version 2768 (0.0011)
+[2026-06-06 20:55:26,512][262198] Updated weights for policy 0, policy_version 2779 (0.0013)
+[2026-06-06 20:55:27,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4300.8, 300 sec: 4325.1). Total num frames: 1425408. Throughput: 0: 4318.6. Samples: 1434252. Policy #0 lag: (min: 3.0, avg: 18.3, max: 40.0)
+[2026-06-06 20:55:27,923][261293] Avg episode reward: [(0, '600.261')]
+[2026-06-06 20:55:28,426][262198] Updated weights for policy 0, policy_version 2790 (0.0011)
+[2026-06-06 20:55:29,194][262115] Signal inference workers to stop experience collection... (350 times)
+[2026-06-06 20:55:29,223][262198] InferenceWorker_p0-w0: stopping experience collection (350 times)
+[2026-06-06 20:55:29,258][262115] Signal inference workers to resume experience collection... (350 times)
+[2026-06-06 20:55:29,258][262198] InferenceWorker_p0-w0: resuming experience collection (350 times)
+[2026-06-06 20:55:29,366][262198] Updated weights for policy 0, policy_version 2801 (0.0010)
+[2026-06-06 20:55:30,500][262198] Updated weights for policy 0, policy_version 2811 (0.0008)
+[2026-06-06 20:55:32,471][262198] Updated weights for policy 0, policy_version 2822 (0.0010)
+[2026-06-06 20:55:32,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4300.8, 300 sec: 4332.0). Total num frames: 1447936. Throughput: 0: 4306.0. Samples: 1446908. Policy #0 lag: (min: 6.0, avg: 18.3, max: 38.0)
+[2026-06-06 20:55:32,922][261293] Avg episode reward: [(0, '577.550')]
+[2026-06-06 20:55:33,382][262198] Updated weights for policy 0, policy_version 2834 (0.0010)
+[2026-06-06 20:55:34,286][262198] Updated weights for policy 0, policy_version 2844 (0.0010)
+[2026-06-06 20:55:36,221][262198] Updated weights for policy 0, policy_version 2856 (0.0010)
+[2026-06-06 20:55:37,361][262198] Updated weights for policy 0, policy_version 2868 (0.0009)
+[2026-06-06 20:55:37,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4334.9, 300 sec: 4325.1). Total num frames: 1470464. Throughput: 0: 4299.5. Samples: 1473084. Policy #0 lag: (min: 5.0, avg: 19.7, max: 38.0)
+[2026-06-06 20:55:37,924][261293] Avg episode reward: [(0, '610.098')]
+[2026-06-06 20:55:38,511][262198] Updated weights for policy 0, policy_version 2878 (0.0009)
+[2026-06-06 20:55:39,699][262198] Updated weights for policy 0, policy_version 2888 (0.0010)
+[2026-06-06 20:55:40,917][262198] Updated weights for policy 0, policy_version 2900 (0.0011)
+[2026-06-06 20:55:42,133][262198] Updated weights for policy 0, policy_version 2910 (0.0009)
+[2026-06-06 20:55:42,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4334.9, 300 sec: 4332.0). Total num frames: 1490944. Throughput: 0: 4334.6. Samples: 1500612. Policy #0 lag: (min: 4.0, avg: 20.3, max: 36.0)
+[2026-06-06 20:55:42,922][261293] Avg episode reward: [(0, '550.568')]
+[2026-06-06 20:55:43,618][262198] Updated weights for policy 0, policy_version 2923 (0.0010)
+[2026-06-06 20:55:44,705][262198] Updated weights for policy 0, policy_version 2935 (0.0009)
+[2026-06-06 20:55:46,476][262198] Updated weights for policy 0, policy_version 2945 (0.0009)
+[2026-06-06 20:55:47,638][262198] Updated weights for policy 0, policy_version 2957 (0.0010)
+[2026-06-06 20:55:47,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4369.1, 300 sec: 4345.9). Total num frames: 1515520. Throughput: 0: 4358.0. Samples: 1513874. Policy #0 lag: (min: 7.0, avg: 22.8, max: 40.0)
+[2026-06-06 20:55:47,923][261293] Avg episode reward: [(0, '526.853')]
+[2026-06-06 20:55:48,782][262198] Updated weights for policy 0, policy_version 2969 (0.0010)
+[2026-06-06 20:55:50,640][262198] Updated weights for policy 0, policy_version 2981 (0.0010)
+[2026-06-06 20:55:51,665][262198] Updated weights for policy 0, policy_version 2993 (0.0009)
+[2026-06-06 20:55:52,577][262198] Updated weights for policy 0, policy_version 3003 (0.0009)
+[2026-06-06 20:55:52,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4369.0, 300 sec: 4345.9). Total num frames: 1538048. Throughput: 0: 4341.0. Samples: 1540244. Policy #0 lag: (min: 3.0, avg: 22.4, max: 35.0)
+[2026-06-06 20:55:52,924][261293] Avg episode reward: [(0, '587.617')]
+[2026-06-06 20:55:54,406][262198] Updated weights for policy 0, policy_version 3015 (0.0009)
+[2026-06-06 20:55:55,397][262198] Updated weights for policy 0, policy_version 3026 (0.0009)
+[2026-06-06 20:55:56,343][262198] Updated weights for policy 0, policy_version 3036 (0.0010)
+[2026-06-06 20:55:57,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4369.1, 300 sec: 4339.0). Total num frames: 1558528. Throughput: 0: 4370.1. Samples: 1567332. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 20:55:57,924][261293] Avg episode reward: [(0, '685.362')]
+[2026-06-06 20:55:58,138][262115] Saving new best policy, reward=685.362!
+[2026-06-06 20:55:58,141][262198] Updated weights for policy 0, policy_version 3048 (0.0011)
+[2026-06-06 20:55:59,416][262198] Updated weights for policy 0, policy_version 3061 (0.0010)
+[2026-06-06 20:56:00,608][262198] Updated weights for policy 0, policy_version 3071 (0.0010)
+[2026-06-06 20:56:02,372][262198] Updated weights for policy 0, policy_version 3084 (0.0011)
+[2026-06-06 20:56:02,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.0, 300 sec: 4352.9). Total num frames: 1581056. Throughput: 0: 4351.3. Samples: 1579908. Policy #0 lag: (min: 7.0, avg: 19.1, max: 39.0)
+[2026-06-06 20:56:02,923][261293] Avg episode reward: [(0, '552.763')]
+[2026-06-06 20:56:03,623][262198] Updated weights for policy 0, policy_version 3097 (0.0010)
+[2026-06-06 20:56:05,463][262198] Updated weights for policy 0, policy_version 3109 (0.0010)
+[2026-06-06 20:56:06,419][262198] Updated weights for policy 0, policy_version 3121 (0.0010)
+[2026-06-06 20:56:07,413][262198] Updated weights for policy 0, policy_version 3131 (0.0010)
+[2026-06-06 20:56:07,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4369.1, 300 sec: 4352.9). Total num frames: 1603584. Throughput: 0: 4368.5. Samples: 1606416. Policy #0 lag: (min: 6.0, avg: 21.4, max: 40.0)
+[2026-06-06 20:56:07,923][261293] Avg episode reward: [(0, '612.149')]
+[2026-06-06 20:56:09,322][262198] Updated weights for policy 0, policy_version 3143 (0.0009)
+[2026-06-06 20:56:10,747][262198] Updated weights for policy 0, policy_version 3158 (0.0009)
+[2026-06-06 20:56:12,050][262198] Updated weights for policy 0, policy_version 3168 (0.0008)
+[2026-06-06 20:56:12,920][261293] Fps is (10 sec: 4301.1, 60 sec: 4369.1, 300 sec: 4352.9). Total num frames: 1624064. Throughput: 0: 4410.9. Samples: 1632740. Policy #0 lag: (min: 10.0, avg: 23.0, max: 41.0)
+[2026-06-06 20:56:12,921][261293] Avg episode reward: [(0, '632.719')]
+[2026-06-06 20:56:13,505][262115] Signal inference workers to stop experience collection... (400 times)
+[2026-06-06 20:56:13,526][262198] InferenceWorker_p0-w0: stopping experience collection (400 times)
+[2026-06-06 20:56:13,541][262115] Signal inference workers to resume experience collection... (400 times)
+[2026-06-06 20:56:13,542][262198] InferenceWorker_p0-w0: resuming experience collection (400 times)
+[2026-06-06 20:56:13,543][262198] Updated weights for policy 0, policy_version 3180 (0.0009)
+[2026-06-06 20:56:14,896][262198] Updated weights for policy 0, policy_version 3195 (0.0008)
+[2026-06-06 20:56:16,673][262198] Updated weights for policy 0, policy_version 3207 (0.0009)
+[2026-06-06 20:56:17,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4369.1, 300 sec: 4359.8). Total num frames: 1646592. Throughput: 0: 4425.3. Samples: 1646048. Policy #0 lag: (min: 1.0, avg: 22.1, max: 33.0)
+[2026-06-06 20:56:17,922][261293] Avg episode reward: [(0, '600.325')]
+[2026-06-06 20:56:18,065][262198] Updated weights for policy 0, policy_version 3221 (0.0009)
+[2026-06-06 20:56:19,121][262198] Updated weights for policy 0, policy_version 3231 (0.0008)
+[2026-06-06 20:56:20,928][262198] Updated weights for policy 0, policy_version 3245 (0.0010)
+[2026-06-06 20:56:21,942][262198] Updated weights for policy 0, policy_version 3257 (0.0010)
+[2026-06-06 20:56:22,921][261293] Fps is (10 sec: 4710.2, 60 sec: 4403.2, 300 sec: 4373.7). Total num frames: 1671168. Throughput: 0: 4454.3. Samples: 1673526. Policy #0 lag: (min: 3.0, avg: 21.7, max: 35.0)
+[2026-06-06 20:56:22,925][261293] Avg episode reward: [(0, '653.211')]
+[2026-06-06 20:56:23,840][262198] Updated weights for policy 0, policy_version 3269 (0.0009)
+[2026-06-06 20:56:25,215][262198] Updated weights for policy 0, policy_version 3282 (0.0009)
+[2026-06-06 20:56:26,337][262198] Updated weights for policy 0, policy_version 3293 (0.0009)
+[2026-06-06 20:56:27,844][262198] Updated weights for policy 0, policy_version 3304 (0.0009)
+[2026-06-06 20:56:27,923][261293] Fps is (10 sec: 4504.3, 60 sec: 4437.1, 300 sec: 4366.7). Total num frames: 1691648. Throughput: 0: 4431.0. Samples: 1700018. Policy #0 lag: (min: 7.0, avg: 19.6, max: 39.0)
+[2026-06-06 20:56:27,925][261293] Avg episode reward: [(0, '743.710')]
+[2026-06-06 20:56:27,930][262115] Saving new best policy, reward=743.710!
+[2026-06-06 20:56:29,487][262198] Updated weights for policy 0, policy_version 3319 (0.0011)
+[2026-06-06 20:56:31,278][262198] Updated weights for policy 0, policy_version 3329 (0.0010)
+[2026-06-06 20:56:32,601][262198] Updated weights for policy 0, policy_version 3344 (0.0011)
+[2026-06-06 20:56:32,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4437.3, 300 sec: 4366.8). Total num frames: 1714176. Throughput: 0: 4400.7. Samples: 1711902. Policy #0 lag: (min: 4.0, avg: 18.8, max: 38.0)
+[2026-06-06 20:56:32,923][261293] Avg episode reward: [(0, '679.306')]
+[2026-06-06 20:56:33,581][262198] Updated weights for policy 0, policy_version 3356 (0.0009)
+[2026-06-06 20:56:35,327][262198] Updated weights for policy 0, policy_version 3367 (0.0009)
+[2026-06-06 20:56:36,648][262198] Updated weights for policy 0, policy_version 3381 (0.0010)
+[2026-06-06 20:56:37,729][262198] Updated weights for policy 0, policy_version 3391 (0.0009)
+[2026-06-06 20:56:37,921][261293] Fps is (10 sec: 4506.8, 60 sec: 4437.3, 300 sec: 4373.7). Total num frames: 1736704. Throughput: 0: 4402.0. Samples: 1738332. Policy #0 lag: (min: 11.0, avg: 24.3, max: 43.0)
+[2026-06-06 20:56:37,923][261293] Avg episode reward: [(0, '562.494')]
+[2026-06-06 20:56:39,291][262198] Updated weights for policy 0, policy_version 3402 (0.0010)
+[2026-06-06 20:56:40,371][262198] Updated weights for policy 0, policy_version 3415 (0.0010)
+[2026-06-06 20:56:42,294][262198] Updated weights for policy 0, policy_version 3425 (0.0010)
+[2026-06-06 20:56:42,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4437.4, 300 sec: 4366.8). Total num frames: 1757184. Throughput: 0: 4419.4. Samples: 1766206. Policy #0 lag: (min: 7.0, avg: 23.4, max: 36.0)
+[2026-06-06 20:56:42,922][261293] Avg episode reward: [(0, '745.488')]
+[2026-06-06 20:56:43,111][262115] Saving new best policy, reward=745.488!
+[2026-06-06 20:56:43,599][262198] Updated weights for policy 0, policy_version 3441 (0.0008)
+[2026-06-06 20:56:44,509][262198] Updated weights for policy 0, policy_version 3452 (0.0009)
+[2026-06-06 20:56:46,272][262198] Updated weights for policy 0, policy_version 3462 (0.0009)
+[2026-06-06 20:56:47,356][262198] Updated weights for policy 0, policy_version 3476 (0.0010)
+[2026-06-06 20:56:47,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4437.4, 300 sec: 4380.6). Total num frames: 1781760. Throughput: 0: 4423.5. Samples: 1778964. Policy #0 lag: (min: 3.0, avg: 21.8, max: 35.0)
+[2026-06-06 20:56:47,923][261293] Avg episode reward: [(0, '783.254')]
+[2026-06-06 20:56:48,169][262115] Saving new best policy, reward=783.254!
+[2026-06-06 20:56:48,759][262198] Updated weights for policy 0, policy_version 3487 (0.0010)
+[2026-06-06 20:56:50,781][262198] Updated weights for policy 0, policy_version 3501 (0.0009)
+[2026-06-06 20:56:51,661][262115] Signal inference workers to stop experience collection... (450 times)
+[2026-06-06 20:56:51,678][262115] Signal inference workers to resume experience collection... (450 times)
+[2026-06-06 20:56:51,686][262198] InferenceWorker_p0-w0: stopping experience collection (450 times)
+[2026-06-06 20:56:51,705][262198] InferenceWorker_p0-w0: resuming experience collection (450 times)
+[2026-06-06 20:56:51,788][262198] Updated weights for policy 0, policy_version 3513 (0.0010)
+[2026-06-06 20:56:52,920][261293] Fps is (10 sec: 4505.5, 60 sec: 4403.2, 300 sec: 4380.6). Total num frames: 1802240. Throughput: 0: 4411.8. Samples: 1804946. Policy #0 lag: (min: 1.0, avg: 14.8, max: 33.0)
+[2026-06-06 20:56:52,923][261293] Avg episode reward: [(0, '676.916')]
+[2026-06-06 20:56:53,673][262198] Updated weights for policy 0, policy_version 3524 (0.0011)
+[2026-06-06 20:56:54,934][262198] Updated weights for policy 0, policy_version 3537 (0.0009)
+[2026-06-06 20:56:55,695][262198] Updated weights for policy 0, policy_version 3547 (0.0009)
+[2026-06-06 20:56:57,651][262198] Updated weights for policy 0, policy_version 3558 (0.0011)
+[2026-06-06 20:56:57,920][261293] Fps is (10 sec: 4095.9, 60 sec: 4403.2, 300 sec: 4380.6). Total num frames: 1822720. Throughput: 0: 4415.1. Samples: 1831420. Policy #0 lag: (min: 1.0, avg: 16.1, max: 33.0)
+[2026-06-06 20:56:57,922][261293] Avg episode reward: [(0, '760.363')]
+[2026-06-06 20:56:58,642][262198] Updated weights for policy 0, policy_version 3571 (0.0009)
+[2026-06-06 20:57:00,051][262198] Updated weights for policy 0, policy_version 3581 (0.0009)
+[2026-06-06 20:57:01,219][262198] Updated weights for policy 0, policy_version 3591 (0.0009)
+[2026-06-06 20:57:02,545][262198] Updated weights for policy 0, policy_version 3605 (0.0009)
+[2026-06-06 20:57:02,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4437.4, 300 sec: 4380.6). Total num frames: 1847296. Throughput: 0: 4414.3. Samples: 1844692. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 20:57:02,923][261293] Avg episode reward: [(0, '620.960')]
+[2026-06-06 20:57:03,591][262198] Updated weights for policy 0, policy_version 3615 (0.0009)
+[2026-06-06 20:57:05,496][262198] Updated weights for policy 0, policy_version 3629 (0.0010)
+[2026-06-06 20:57:06,506][262198] Updated weights for policy 0, policy_version 3641 (0.0011)
+[2026-06-06 20:57:07,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4403.2, 300 sec: 4380.6). Total num frames: 1867776. Throughput: 0: 4399.8. Samples: 1871516. Policy #0 lag: (min: 2.0, avg: 23.4, max: 40.0)
+[2026-06-06 20:57:07,923][261293] Avg episode reward: [(0, '710.750')]
+[2026-06-06 20:57:08,194][262198] Updated weights for policy 0, policy_version 3651 (0.0009)
+[2026-06-06 20:57:09,484][262198] Updated weights for policy 0, policy_version 3665 (0.0009)
+[2026-06-06 20:57:10,208][262198] Updated weights for policy 0, policy_version 3675 (0.0009)
+[2026-06-06 20:57:12,327][262198] Updated weights for policy 0, policy_version 3687 (0.0009)
+[2026-06-06 20:57:12,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4437.3, 300 sec: 4380.6). Total num frames: 1890304. Throughput: 0: 4414.4. Samples: 1898656. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 20:57:12,923][261293] Avg episode reward: [(0, '557.444')]
+[2026-06-06 20:57:13,342][262198] Updated weights for policy 0, policy_version 3700 (0.0011)
+[2026-06-06 20:57:14,762][262198] Updated weights for policy 0, policy_version 3710 (0.0009)
+[2026-06-06 20:57:15,753][262198] Updated weights for policy 0, policy_version 3720 (0.0009)
+[2026-06-06 20:57:16,960][262198] Updated weights for policy 0, policy_version 3734 (0.0010)
+[2026-06-06 20:57:17,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4471.5, 300 sec: 4380.6). Total num frames: 1914880. Throughput: 0: 4476.2. Samples: 1913330. Policy #0 lag: (min: 7.0, avg: 22.7, max: 41.0)
+[2026-06-06 20:57:17,922][261293] Avg episode reward: [(0, '702.632')]
+[2026-06-06 20:57:18,524][262198] Updated weights for policy 0, policy_version 3744 (0.0009)
+[2026-06-06 20:57:19,605][262198] Updated weights for policy 0, policy_version 3756 (0.0012)
+[2026-06-06 20:57:20,638][262198] Updated weights for policy 0, policy_version 3768 (0.0009)
+[2026-06-06 20:57:22,561][262198] Updated weights for policy 0, policy_version 3779 (0.0010)
+[2026-06-06 20:57:22,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4403.2, 300 sec: 4387.6). Total num frames: 1935360. Throughput: 0: 4481.2. Samples: 1939986. Policy #0 lag: (min: 3.0, avg: 24.2, max: 42.0)
+[2026-06-06 20:57:22,925][261293] Avg episode reward: [(0, '684.023')]
+[2026-06-06 20:57:23,560][262198] Updated weights for policy 0, policy_version 3790 (0.0009)
+[2026-06-06 20:57:24,593][262198] Updated weights for policy 0, policy_version 3801 (0.0011)
+[2026-06-06 20:57:26,561][262198] Updated weights for policy 0, policy_version 3815 (0.0011)
+[2026-06-06 20:57:27,007][262115] Signal inference workers to stop experience collection... (500 times)
+[2026-06-06 20:57:27,011][262115] Signal inference workers to resume experience collection... (500 times)
+[2026-06-06 20:57:27,026][262198] InferenceWorker_p0-w0: stopping experience collection (500 times)
+[2026-06-06 20:57:27,026][262198] InferenceWorker_p0-w0: resuming experience collection (500 times)
+[2026-06-06 20:57:27,538][262198] Updated weights for policy 0, policy_version 3827 (0.0010)
+[2026-06-06 20:57:27,922][261293] Fps is (10 sec: 4505.0, 60 sec: 4471.6, 300 sec: 4394.5). Total num frames: 1959936. Throughput: 0: 4460.0. Samples: 1966916. Policy #0 lag: (min: 4.0, avg: 21.9, max: 36.0)
+[2026-06-06 20:57:27,924][261293] Avg episode reward: [(0, '774.871')]
+[2026-06-06 20:57:29,591][262198] Updated weights for policy 0, policy_version 3837 (0.0010)
+[2026-06-06 20:57:30,740][262198] Updated weights for policy 0, policy_version 3850 (0.0009)
+[2026-06-06 20:57:31,726][262198] Updated weights for policy 0, policy_version 3863 (0.0010)
+[2026-06-06 20:57:32,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4437.3, 300 sec: 4387.6). Total num frames: 1980416. Throughput: 0: 4471.6. Samples: 1980188. Policy #0 lag: (min: 7.0, avg: 19.6, max: 43.0)
+[2026-06-06 20:57:32,924][261293] Avg episode reward: [(0, '802.898')]
+[2026-06-06 20:57:32,931][262115] Saving new best policy, reward=802.898!
+[2026-06-06 20:57:33,787][262198] Updated weights for policy 0, policy_version 3876 (0.0010)
+[2026-06-06 20:57:34,727][262198] Updated weights for policy 0, policy_version 3888 (0.0009)
+[2026-06-06 20:57:35,670][262198] Updated weights for policy 0, policy_version 3898 (0.0009)
+[2026-06-06 20:57:37,732][262198] Updated weights for policy 0, policy_version 3912 (0.0010)
+[2026-06-06 20:57:37,921][261293] Fps is (10 sec: 4301.1, 60 sec: 4437.3, 300 sec: 4394.5). Total num frames: 2002944. Throughput: 0: 4504.7. Samples: 2007660. Policy #0 lag: (min: 10.0, avg: 25.6, max: 43.0)
+[2026-06-06 20:57:37,925][261293] Avg episode reward: [(0, '693.594')]
+[2026-06-06 20:57:38,815][262198] Updated weights for policy 0, policy_version 3925 (0.0010)
+[2026-06-06 20:57:40,670][262198] Updated weights for policy 0, policy_version 3936 (0.0009)
+[2026-06-06 20:57:42,063][262198] Updated weights for policy 0, policy_version 3953 (0.0010)
+[2026-06-06 20:57:42,922][261293] Fps is (10 sec: 4710.0, 60 sec: 4505.5, 300 sec: 4408.4). Total num frames: 2027520. Throughput: 0: 4495.8. Samples: 2033736. Policy #0 lag: (min: 1.0, avg: 22.4, max: 37.0)
+[2026-06-06 20:57:42,926][261293] Avg episode reward: [(0, '725.235')]
+[2026-06-06 20:57:43,028][262198] Updated weights for policy 0, policy_version 3963 (0.0009)
+[2026-06-06 20:57:45,848][262198] Updated weights for policy 0, policy_version 3981 (0.0012)
+[2026-06-06 20:57:47,021][262198] Updated weights for policy 0, policy_version 3994 (0.0011)
+[2026-06-06 20:57:47,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4403.2, 300 sec: 4394.5). Total num frames: 2045952. Throughput: 0: 4504.5. Samples: 2047398. Policy #0 lag: (min: 3.0, avg: 20.4, max: 39.0)
+[2026-06-06 20:57:47,923][261293] Avg episode reward: [(0, '757.274')]
+[2026-06-06 20:57:48,889][262198] Updated weights for policy 0, policy_version 4007 (0.0011)
+[2026-06-06 20:57:50,098][262198] Updated weights for policy 0, policy_version 4019 (0.0009)
+[2026-06-06 20:57:52,372][262198] Updated weights for policy 0, policy_version 4029 (0.0010)
+[2026-06-06 20:57:52,920][261293] Fps is (10 sec: 3891.7, 60 sec: 4403.2, 300 sec: 4387.6). Total num frames: 2066432. Throughput: 0: 4439.8. Samples: 2071306. Policy #0 lag: (min: 7.0, avg: 20.5, max: 43.0)
+[2026-06-06 20:57:52,922][261293] Avg episode reward: [(0, '694.650')]
+[2026-06-06 20:57:53,480][262198] Updated weights for policy 0, policy_version 4043 (0.0010)
+[2026-06-06 20:57:54,451][262198] Updated weights for policy 0, policy_version 4056 (0.0011)
+[2026-06-06 20:57:56,768][262198] Updated weights for policy 0, policy_version 4068 (0.0011)
+[2026-06-06 20:57:57,856][262198] Updated weights for policy 0, policy_version 4078 (0.0012)
+[2026-06-06 20:57:57,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4437.3, 300 sec: 4394.5). Total num frames: 2088960. Throughput: 0: 4375.5. Samples: 2095552. Policy #0 lag: (min: 5.0, avg: 20.4, max: 35.0)
+[2026-06-06 20:57:57,922][261293] Avg episode reward: [(0, '666.424')]
+[2026-06-06 20:57:58,494][262198] Updated weights for policy 0, policy_version 4088 (0.0010)
+[2026-06-06 20:58:00,979][262198] Updated weights for policy 0, policy_version 4101 (0.0010)
+[2026-06-06 20:58:01,923][262115] Signal inference workers to stop experience collection... (550 times)
+[2026-06-06 20:58:01,944][262198] InferenceWorker_p0-w0: stopping experience collection (550 times)
+[2026-06-06 20:58:02,002][262115] Signal inference workers to resume experience collection... (550 times)
+[2026-06-06 20:58:02,003][262198] InferenceWorker_p0-w0: resuming experience collection (550 times)
+[2026-06-06 20:58:02,005][262198] Updated weights for policy 0, policy_version 4112 (0.0010)
+[2026-06-06 20:58:02,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.0, 300 sec: 4387.6). Total num frames: 2109440. Throughput: 0: 4334.9. Samples: 2108402. Policy #0 lag: (min: 7.0, avg: 24.1, max: 41.0)
+[2026-06-06 20:58:02,926][261293] Avg episode reward: [(0, '735.321')]
+[2026-06-06 20:58:03,315][262198] Updated weights for policy 0, policy_version 4124 (0.0010)
+[2026-06-06 20:58:04,998][262198] Updated weights for policy 0, policy_version 4136 (0.0009)
+[2026-06-06 20:58:06,152][262198] Updated weights for policy 0, policy_version 4148 (0.0009)
+[2026-06-06 20:58:07,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4334.9, 300 sec: 4373.7). Total num frames: 2127872. Throughput: 0: 4282.1. Samples: 2132682. Policy #0 lag: (min: 3.0, avg: 22.9, max: 35.0)
+[2026-06-06 20:58:07,924][261293] Avg episode reward: [(0, '716.292')]
+[2026-06-06 20:58:08,362][262198] Updated weights for policy 0, policy_version 4160 (0.0011)
+[2026-06-06 20:58:09,769][262198] Updated weights for policy 0, policy_version 4176 (0.0009)
+[2026-06-06 20:58:10,939][262198] Updated weights for policy 0, policy_version 4188 (0.0009)
+[2026-06-06 20:58:12,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4334.9, 300 sec: 4380.6). Total num frames: 2150400. Throughput: 0: 4285.0. Samples: 2159732. Policy #0 lag: (min: 4.0, avg: 21.4, max: 36.0)
+[2026-06-06 20:58:12,923][261293] Avg episode reward: [(0, '662.007')]
+[2026-06-06 20:58:13,047][262198] Updated weights for policy 0, policy_version 4204 (0.0010)
+[2026-06-06 20:58:13,874][262198] Updated weights for policy 0, policy_version 4215 (0.0009)
+[2026-06-06 20:58:16,134][262198] Updated weights for policy 0, policy_version 4230 (0.0008)
+[2026-06-06 20:58:17,182][262198] Updated weights for policy 0, policy_version 4242 (0.0010)
+[2026-06-06 20:58:17,921][261293] Fps is (10 sec: 4710.5, 60 sec: 4334.9, 300 sec: 4380.6). Total num frames: 2174976. Throughput: 0: 4288.5. Samples: 2173168. Policy #0 lag: (min: 6.0, avg: 24.3, max: 38.0)
+[2026-06-06 20:58:17,924][261293] Avg episode reward: [(0, '764.268')]
+[2026-06-06 20:58:18,157][262198] Updated weights for policy 0, policy_version 4252 (0.0008)
+[2026-06-06 20:58:20,068][262198] Updated weights for policy 0, policy_version 4266 (0.0009)
+[2026-06-06 20:58:21,040][262198] Updated weights for policy 0, policy_version 4278 (0.0010)
+[2026-06-06 20:58:22,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4335.0, 300 sec: 4380.6). Total num frames: 2195456. Throughput: 0: 4263.3. Samples: 2199506. Policy #0 lag: (min: 3.0, avg: 19.8, max: 37.0)
+[2026-06-06 20:58:22,922][261293] Avg episode reward: [(0, '895.051')]
+[2026-06-06 20:58:23,155][262198] Updated weights for policy 0, policy_version 4290 (0.0010)
+[2026-06-06 20:58:23,225][262115] Saving new best policy, reward=895.051!
+[2026-06-06 20:58:24,667][262198] Updated weights for policy 0, policy_version 4308 (0.0011)
+[2026-06-06 20:58:26,649][262198] Updated weights for policy 0, policy_version 4318 (0.0010)
+[2026-06-06 20:58:27,672][262198] Updated weights for policy 0, policy_version 4332 (0.0011)
+[2026-06-06 20:58:27,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4335.0, 300 sec: 4387.6). Total num frames: 2220032. Throughput: 0: 4283.4. Samples: 2226484. Policy #0 lag: (min: 7.0, avg: 21.2, max: 39.0)
+[2026-06-06 20:58:27,924][261293] Avg episode reward: [(0, '759.312')]
+[2026-06-06 20:58:28,700][262198] Updated weights for policy 0, policy_version 4344 (0.0009)
+[2026-06-06 20:58:30,760][262198] Updated weights for policy 0, policy_version 4356 (0.0009)
+[2026-06-06 20:58:32,147][262198] Updated weights for policy 0, policy_version 4372 (0.0009)
+[2026-06-06 20:58:32,920][261293] Fps is (10 sec: 4710.6, 60 sec: 4369.1, 300 sec: 4387.6). Total num frames: 2242560. Throughput: 0: 4282.4. Samples: 2240102. Policy #0 lag: (min: 3.0, avg: 24.3, max: 36.0)
+[2026-06-06 20:58:32,922][261293] Avg episode reward: [(0, '775.206')]
+[2026-06-06 20:58:33,953][262198] Updated weights for policy 0, policy_version 4382 (0.0009)
+[2026-06-06 20:58:34,286][262115] Signal inference workers to stop experience collection... (600 times)
+[2026-06-06 20:58:34,304][262198] InferenceWorker_p0-w0: stopping experience collection (600 times)
+[2026-06-06 20:58:34,380][262115] Signal inference workers to resume experience collection... (600 times)
+[2026-06-06 20:58:34,381][262198] InferenceWorker_p0-w0: resuming experience collection (600 times)
+[2026-06-06 20:58:35,192][262198] Updated weights for policy 0, policy_version 4398 (0.0011)
+[2026-06-06 20:58:36,495][262198] Updated weights for policy 0, policy_version 4409 (0.0010)
+[2026-06-06 20:58:37,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4300.8, 300 sec: 4380.6). Total num frames: 2260992. Throughput: 0: 4310.4. Samples: 2265276. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 20:58:37,923][261293] Avg episode reward: [(0, '851.648')]
+[2026-06-06 20:58:38,316][262198] Updated weights for policy 0, policy_version 4421 (0.0011)
+[2026-06-06 20:58:39,816][262198] Updated weights for policy 0, policy_version 4436 (0.0010)
+[2026-06-06 20:58:41,828][262198] Updated weights for policy 0, policy_version 4448 (0.0009)
+[2026-06-06 20:58:42,897][262198] Updated weights for policy 0, policy_version 4461 (0.0009)
+[2026-06-06 20:58:42,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4266.8, 300 sec: 4380.6). Total num frames: 2283520. Throughput: 0: 4351.3. Samples: 2291360. Policy #0 lag: (min: 7.0, avg: 20.5, max: 39.0)
+[2026-06-06 20:58:42,922][261293] Avg episode reward: [(0, '805.054')]
+[2026-06-06 20:58:43,846][262198] Updated weights for policy 0, policy_version 4472 (0.0009)
+[2026-06-06 20:58:45,992][262198] Updated weights for policy 0, policy_version 4484 (0.0008)
+[2026-06-06 20:58:47,210][262198] Updated weights for policy 0, policy_version 4498 (0.0009)
+[2026-06-06 20:58:47,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4335.0, 300 sec: 4373.7). Total num frames: 2306048. Throughput: 0: 4361.1. Samples: 2304652. Policy #0 lag: (min: 5.0, avg: 25.2, max: 39.0)
+[2026-06-06 20:58:47,923][261293] Avg episode reward: [(0, '781.927')]
+[2026-06-06 20:58:48,097][262198] Updated weights for policy 0, policy_version 4508 (0.0010)
+[2026-06-06 20:58:50,284][262198] Updated weights for policy 0, policy_version 4524 (0.0009)
+[2026-06-06 20:58:51,227][262198] Updated weights for policy 0, policy_version 4534 (0.0009)
+[2026-06-06 20:58:52,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4300.8, 300 sec: 4366.8). Total num frames: 2324480. Throughput: 0: 4395.4. Samples: 2330474. Policy #0 lag: (min: 5.0, avg: 25.2, max: 39.0)
+[2026-06-06 20:58:52,922][261293] Avg episode reward: [(0, '933.019')]
+[2026-06-06 20:58:53,250][262198] Updated weights for policy 0, policy_version 4547 (0.0009)
+[2026-06-06 20:58:53,373][262115] Saving new best policy, reward=933.019!
+[2026-06-06 20:58:54,608][262198] Updated weights for policy 0, policy_version 4562 (0.0010)
+[2026-06-06 20:58:55,511][262198] Updated weights for policy 0, policy_version 4572 (0.0009)
+[2026-06-06 20:58:57,157][262198] Updated weights for policy 0, policy_version 4584 (0.0009)
+[2026-06-06 20:58:57,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4369.1, 300 sec: 4373.7). Total num frames: 2351104. Throughput: 0: 4407.5. Samples: 2358070. Policy #0 lag: (min: 7.0, avg: 22.8, max: 42.0)
+[2026-06-06 20:58:57,924][261293] Avg episode reward: [(0, '835.697')]
+[2026-06-06 20:58:58,489][262198] Updated weights for policy 0, policy_version 4598 (0.0011)
+[2026-06-06 20:59:00,212][262198] Updated weights for policy 0, policy_version 4609 (0.0009)
+[2026-06-06 20:59:01,375][262198] Updated weights for policy 0, policy_version 4621 (0.0010)
+[2026-06-06 20:59:02,186][262198] Updated weights for policy 0, policy_version 4632 (0.0008)
+[2026-06-06 20:59:02,920][261293] Fps is (10 sec: 4915.3, 60 sec: 4403.2, 300 sec: 4387.6). Total num frames: 2373632. Throughput: 0: 4411.3. Samples: 2371674. Policy #0 lag: (min: 7.0, avg: 22.8, max: 42.0)
+[2026-06-06 20:59:02,923][261293] Avg episode reward: [(0, '864.338')]
+[2026-06-06 20:59:03,761][262198] Updated weights for policy 0, policy_version 4642 (0.0009)
+[2026-06-06 20:59:04,816][262198] Updated weights for policy 0, policy_version 4654 (0.0010)
+[2026-06-06 20:59:05,788][262198] Updated weights for policy 0, policy_version 4665 (0.0010)
+[2026-06-06 20:59:07,108][262198] Updated weights for policy 0, policy_version 4675 (0.0009)
+[2026-06-06 20:59:07,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.5, 300 sec: 4394.5). Total num frames: 2396160. Throughput: 0: 4460.4. Samples: 2400224. Policy #0 lag: (min: 3.0, avg: 19.0, max: 35.0)
+[2026-06-06 20:59:07,928][261293] Avg episode reward: [(0, '817.714')]
+[2026-06-06 20:59:08,039][262115] Signal inference workers to stop experience collection... (650 times)
+[2026-06-06 20:59:08,061][262198] InferenceWorker_p0-w0: stopping experience collection (650 times)
+[2026-06-06 20:59:08,146][262115] Signal inference workers to resume experience collection... (650 times)
+[2026-06-06 20:59:08,147][262198] InferenceWorker_p0-w0: resuming experience collection (650 times)
+[2026-06-06 20:59:08,478][262198] Updated weights for policy 0, policy_version 4688 (0.0009)
+[2026-06-06 20:59:09,514][262198] Updated weights for policy 0, policy_version 4700 (0.0008)
+[2026-06-06 20:59:11,173][262198] Updated weights for policy 0, policy_version 4710 (0.0009)
+[2026-06-06 20:59:12,254][262198] Updated weights for policy 0, policy_version 4723 (0.0011)
+[2026-06-06 20:59:12,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4505.6, 300 sec: 4394.5). Total num frames: 2420736. Throughput: 0: 4452.3. Samples: 2426836. Policy #0 lag: (min: 11.0, avg: 25.0, max: 43.0)
+[2026-06-06 20:59:12,923][261293] Avg episode reward: [(0, '965.497')]
+[2026-06-06 20:59:13,157][262115] Saving new best policy, reward=965.497!
+[2026-06-06 20:59:14,005][262198] Updated weights for policy 0, policy_version 4733 (0.0009)
+[2026-06-06 20:59:14,921][262198] Updated weights for policy 0, policy_version 4743 (0.0009)
+[2026-06-06 20:59:16,347][262198] Updated weights for policy 0, policy_version 4759 (0.0009)
+[2026-06-06 20:59:17,906][262198] Updated weights for policy 0, policy_version 4769 (0.0009)
+[2026-06-06 20:59:17,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4437.3, 300 sec: 4394.5). Total num frames: 2441216. Throughput: 0: 4460.2. Samples: 2440816. Policy #0 lag: (min: 3.0, avg: 18.9, max: 35.0)
+[2026-06-06 20:59:17,926][261293] Avg episode reward: [(0, '942.079')]
+[2026-06-06 20:59:18,025][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000004772_2443264.pth...
+[2026-06-06 20:59:19,013][262198] Updated weights for policy 0, policy_version 4780 (0.0010)
+[2026-06-06 20:59:20,418][262198] Updated weights for policy 0, policy_version 4793 (0.0011)
+[2026-06-06 20:59:21,712][262198] Updated weights for policy 0, policy_version 4803 (0.0009)
+[2026-06-06 20:59:22,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4471.4, 300 sec: 4394.5). Total num frames: 2463744. Throughput: 0: 4490.4. Samples: 2467344. Policy #0 lag: (min: 7.0, avg: 22.2, max: 39.0)
+[2026-06-06 20:59:22,924][261293] Avg episode reward: [(0, '784.960')]
+[2026-06-06 20:59:23,153][262198] Updated weights for policy 0, policy_version 4816 (0.0009)
+[2026-06-06 20:59:24,200][262198] Updated weights for policy 0, policy_version 4827 (0.0010)
+[2026-06-06 20:59:25,954][262198] Updated weights for policy 0, policy_version 4837 (0.0010)
+[2026-06-06 20:59:26,906][262198] Updated weights for policy 0, policy_version 4850 (0.0011)
+[2026-06-06 20:59:27,921][261293] Fps is (10 sec: 4505.8, 60 sec: 4437.3, 300 sec: 4394.5). Total num frames: 2486272. Throughput: 0: 4478.0. Samples: 2492870. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 20:59:27,925][261293] Avg episode reward: [(0, '942.338')]
+[2026-06-06 20:59:28,041][262198] Updated weights for policy 0, policy_version 4860 (0.0010)
+[2026-06-06 20:59:29,858][262198] Updated weights for policy 0, policy_version 4870 (0.0010)
+[2026-06-06 20:59:31,020][262198] Updated weights for policy 0, policy_version 4884 (0.0010)
+[2026-06-06 20:59:32,651][262198] Updated weights for policy 0, policy_version 4896 (0.0010)
+[2026-06-06 20:59:32,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4403.2, 300 sec: 4394.5). Total num frames: 2506752. Throughput: 0: 4488.9. Samples: 2506654. Policy #0 lag: (min: 5.0, avg: 18.8, max: 39.0)
+[2026-06-06 20:59:32,923][261293] Avg episode reward: [(0, '819.306')]
+[2026-06-06 20:59:33,978][262198] Updated weights for policy 0, policy_version 4908 (0.0008)
+[2026-06-06 20:59:35,303][262198] Updated weights for policy 0, policy_version 4923 (0.0009)
+[2026-06-06 20:59:37,029][262198] Updated weights for policy 0, policy_version 4933 (0.0010)
+[2026-06-06 20:59:37,922][261293] Fps is (10 sec: 4300.2, 60 sec: 4471.3, 300 sec: 4401.4). Total num frames: 2529280. Throughput: 0: 4510.6. Samples: 2533458. Policy #0 lag: (min: 2.0, avg: 21.3, max: 37.0)
+[2026-06-06 20:59:37,924][261293] Avg episode reward: [(0, '826.980')]
+[2026-06-06 20:59:38,284][262198] Updated weights for policy 0, policy_version 4948 (0.0010)
+[2026-06-06 20:59:40,063][262198] Updated weights for policy 0, policy_version 4960 (0.0010)
+[2026-06-06 20:59:41,467][262198] Updated weights for policy 0, policy_version 4972 (0.0009)
+[2026-06-06 20:59:42,725][262198] Updated weights for policy 0, policy_version 4984 (0.0010)
+[2026-06-06 20:59:42,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.4, 300 sec: 4401.5). Total num frames: 2551808. Throughput: 0: 4458.1. Samples: 2558686. Policy #0 lag: (min: 5.0, avg: 20.2, max: 39.0)
+[2026-06-06 20:59:42,923][261293] Avg episode reward: [(0, '865.977')]
+[2026-06-06 20:59:44,362][262198] Updated weights for policy 0, policy_version 4994 (0.0009)
+[2026-06-06 20:59:45,867][262198] Updated weights for policy 0, policy_version 5009 (0.0009)
+[2026-06-06 20:59:47,765][262198] Updated weights for policy 0, policy_version 5021 (0.0009)
+[2026-06-06 20:59:47,921][261293] Fps is (10 sec: 4096.6, 60 sec: 4403.2, 300 sec: 4387.6). Total num frames: 2570240. Throughput: 0: 4444.0. Samples: 2571656. Policy #0 lag: (min: 3.0, avg: 23.3, max: 39.0)
+[2026-06-06 20:59:47,925][261293] Avg episode reward: [(0, '1020.379')]
+[2026-06-06 20:59:48,023][262115] Saving new best policy, reward=1020.379!
+[2026-06-06 20:59:48,768][262198] Updated weights for policy 0, policy_version 5031 (0.0009)
+[2026-06-06 20:59:50,659][262198] Updated weights for policy 0, policy_version 5049 (0.0012)
+[2026-06-06 20:59:52,086][262198] Updated weights for policy 0, policy_version 5059 (0.0009)
+[2026-06-06 20:59:52,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4471.4, 300 sec: 4394.5). Total num frames: 2592768. Throughput: 0: 4366.3. Samples: 2596708. Policy #0 lag: (min: 6.0, avg: 19.3, max: 39.0)
+[2026-06-06 20:59:52,924][261293] Avg episode reward: [(0, '989.537')]
+[2026-06-06 20:59:53,598][262198] Updated weights for policy 0, policy_version 5072 (0.0011)
+[2026-06-06 20:59:54,718][262198] Updated weights for policy 0, policy_version 5083 (0.0010)
+[2026-06-06 20:59:55,949][262115] Signal inference workers to stop experience collection... (700 times)
+[2026-06-06 20:59:55,970][262198] InferenceWorker_p0-w0: stopping experience collection (700 times)
+[2026-06-06 20:59:56,052][262115] Signal inference workers to resume experience collection... (700 times)
+[2026-06-06 20:59:56,053][262198] InferenceWorker_p0-w0: resuming experience collection (700 times)
+[2026-06-06 20:59:56,811][262198] Updated weights for policy 0, policy_version 5094 (0.0010)
+[2026-06-06 20:59:57,914][262198] Updated weights for policy 0, policy_version 5108 (0.0009)
+[2026-06-06 20:59:57,920][261293] Fps is (10 sec: 4505.8, 60 sec: 4403.2, 300 sec: 4394.5). Total num frames: 2615296. Throughput: 0: 4323.8. Samples: 2621406. Policy #0 lag: (min: 7.0, avg: 21.2, max: 39.0)
+[2026-06-06 20:59:57,922][261293] Avg episode reward: [(0, '782.562')]
+[2026-06-06 20:59:59,938][262198] Updated weights for policy 0, policy_version 5120 (0.0010)
+[2026-06-06 21:00:01,187][262198] Updated weights for policy 0, policy_version 5132 (0.0010)
+[2026-06-06 21:00:02,541][262198] Updated weights for policy 0, policy_version 5146 (0.0009)
+[2026-06-06 21:00:02,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4369.1, 300 sec: 4387.6). Total num frames: 2635776. Throughput: 0: 4295.7. Samples: 2634118. Policy #0 lag: (min: 7.0, avg: 21.2, max: 39.0)
+[2026-06-06 21:00:02,924][261293] Avg episode reward: [(0, '851.734')]
+[2026-06-06 21:00:04,185][262198] Updated weights for policy 0, policy_version 5157 (0.0010)
+[2026-06-06 21:00:05,196][262198] Updated weights for policy 0, policy_version 5170 (0.0009)
+[2026-06-06 21:00:07,107][262198] Updated weights for policy 0, policy_version 5182 (0.0010)
+[2026-06-06 21:00:07,922][262198] Updated weights for policy 0, policy_version 5192 (0.0009)
+[2026-06-06 21:00:07,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4369.0, 300 sec: 4394.5). Total num frames: 2658304. Throughput: 0: 4289.1. Samples: 2660352. Policy #0 lag: (min: 3.0, avg: 17.4, max: 35.0)
+[2026-06-06 21:00:07,924][261293] Avg episode reward: [(0, '928.460')]
+[2026-06-06 21:00:09,087][262198] Updated weights for policy 0, policy_version 5204 (0.0009)
+[2026-06-06 21:00:10,708][262198] Updated weights for policy 0, policy_version 5215 (0.0009)
+[2026-06-06 21:00:11,765][262198] Updated weights for policy 0, policy_version 5225 (0.0009)
+[2026-06-06 21:00:12,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4334.9, 300 sec: 4394.5). Total num frames: 2680832. Throughput: 0: 4341.3. Samples: 2688226. Policy #0 lag: (min: 8.0, avg: 22.1, max: 40.0)
+[2026-06-06 21:00:12,923][261293] Avg episode reward: [(0, '1067.840')]
+[2026-06-06 21:00:13,156][262115] Saving new best policy, reward=1083.163!
+[2026-06-06 21:00:13,363][262198] Updated weights for policy 0, policy_version 5242 (0.0010)
+[2026-06-06 21:00:14,975][262198] Updated weights for policy 0, policy_version 5252 (0.0009)
+[2026-06-06 21:00:16,217][262198] Updated weights for policy 0, policy_version 5266 (0.0009)
+[2026-06-06 21:00:17,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4335.0, 300 sec: 4387.6). Total num frames: 2701312. Throughput: 0: 4332.1. Samples: 2701600. Policy #0 lag: (min: 3.0, avg: 17.4, max: 35.0)
+[2026-06-06 21:00:17,922][261293] Avg episode reward: [(0, '922.420')]
+[2026-06-06 21:00:18,031][262198] Updated weights for policy 0, policy_version 5278 (0.0009)
+[2026-06-06 21:00:18,805][262198] Updated weights for policy 0, policy_version 5288 (0.0009)
+[2026-06-06 21:00:20,425][262198] Updated weights for policy 0, policy_version 5304 (0.0010)
+[2026-06-06 21:00:22,110][262198] Updated weights for policy 0, policy_version 5314 (0.0008)
+[2026-06-06 21:00:22,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4334.9, 300 sec: 4401.5). Total num frames: 2723840. Throughput: 0: 4319.6. Samples: 2727836. Policy #0 lag: (min: 10.0, avg: 24.8, max: 42.0)
+[2026-06-06 21:00:22,923][261293] Avg episode reward: [(0, '956.697')]
+[2026-06-06 21:00:23,217][262198] Updated weights for policy 0, policy_version 5325 (0.0010)
+[2026-06-06 21:00:24,227][262198] Updated weights for policy 0, policy_version 5339 (0.0010)
+[2026-06-06 21:00:26,082][262198] Updated weights for policy 0, policy_version 5350 (0.0009)
+[2026-06-06 21:00:27,227][262198] Updated weights for policy 0, policy_version 5364 (0.0009)
+[2026-06-06 21:00:27,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4369.1, 300 sec: 4408.4). Total num frames: 2748416. Throughput: 0: 4350.3. Samples: 2754448. Policy #0 lag: (min: 3.0, avg: 17.8, max: 35.0)
+[2026-06-06 21:00:27,923][261293] Avg episode reward: [(0, '891.647')]
+[2026-06-06 21:00:29,195][262198] Updated weights for policy 0, policy_version 5376 (0.0009)
+[2026-06-06 21:00:30,991][262198] Updated weights for policy 0, policy_version 5393 (0.0010)
+[2026-06-06 21:00:31,896][262198] Updated weights for policy 0, policy_version 5404 (0.0010)
+[2026-06-06 21:00:32,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4334.9, 300 sec: 4394.5). Total num frames: 2766848. Throughput: 0: 4364.1. Samples: 2768038. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 21:00:32,922][261293] Avg episode reward: [(0, '794.853')]
+[2026-06-06 21:00:33,579][262198] Updated weights for policy 0, policy_version 5414 (0.0009)
+[2026-06-06 21:00:34,484][262115] Signal inference workers to stop experience collection... (750 times)
+[2026-06-06 21:00:34,509][262198] InferenceWorker_p0-w0: stopping experience collection (750 times)
+[2026-06-06 21:00:34,620][262115] Signal inference workers to resume experience collection... (750 times)
+[2026-06-06 21:00:34,621][262198] InferenceWorker_p0-w0: resuming experience collection (750 times)
+[2026-06-06 21:00:35,231][262198] Updated weights for policy 0, policy_version 5429 (0.0009)
+[2026-06-06 21:00:36,648][262198] Updated weights for policy 0, policy_version 5439 (0.0008)
+[2026-06-06 21:00:37,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4335.1, 300 sec: 4401.5). Total num frames: 2789376. Throughput: 0: 4383.2. Samples: 2793948. Policy #0 lag: (min: 2.0, avg: 20.9, max: 38.0)
+[2026-06-06 21:00:37,922][261293] Avg episode reward: [(0, '1080.845')]
+[2026-06-06 21:00:38,379][262198] Updated weights for policy 0, policy_version 5453 (0.0009)
+[2026-06-06 21:00:39,454][262198] Updated weights for policy 0, policy_version 5467 (0.0010)
+[2026-06-06 21:00:41,603][262198] Updated weights for policy 0, policy_version 5480 (0.0011)
+[2026-06-06 21:00:42,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4334.9, 300 sec: 4394.5). Total num frames: 2811904. Throughput: 0: 4377.7. Samples: 2818404. Policy #0 lag: (min: 8.0, avg: 20.1, max: 40.0)
+[2026-06-06 21:00:42,925][261293] Avg episode reward: [(0, '1052.046')]
+[2026-06-06 21:00:43,309][262198] Updated weights for policy 0, policy_version 5495 (0.0011)
+[2026-06-06 21:00:45,147][262198] Updated weights for policy 0, policy_version 5505 (0.0009)
+[2026-06-06 21:00:46,361][262198] Updated weights for policy 0, policy_version 5520 (0.0009)
+[2026-06-06 21:00:47,320][262198] Updated weights for policy 0, policy_version 5531 (0.0008)
+[2026-06-06 21:00:47,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.1, 300 sec: 4387.6). Total num frames: 2832384. Throughput: 0: 4394.0. Samples: 2831848. Policy #0 lag: (min: 8.0, avg: 20.1, max: 40.0)
+[2026-06-06 21:00:47,922][261293] Avg episode reward: [(0, '982.041')]
+[2026-06-06 21:00:49,110][262198] Updated weights for policy 0, policy_version 5543 (0.0009)
+[2026-06-06 21:00:50,395][262198] Updated weights for policy 0, policy_version 5557 (0.0009)
+[2026-06-06 21:00:51,850][262198] Updated weights for policy 0, policy_version 5567 (0.0009)
+[2026-06-06 21:00:52,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4369.1, 300 sec: 4394.5). Total num frames: 2854912. Throughput: 0: 4402.4. Samples: 2858458. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 21:00:52,922][261293] Avg episode reward: [(0, '956.736')]
+[2026-06-06 21:00:53,167][262198] Updated weights for policy 0, policy_version 5580 (0.0009)
+[2026-06-06 21:00:54,591][262198] Updated weights for policy 0, policy_version 5594 (0.0010)
+[2026-06-06 21:00:56,506][262198] Updated weights for policy 0, policy_version 5606 (0.0011)
+[2026-06-06 21:00:57,600][262198] Updated weights for policy 0, policy_version 5620 (0.0010)
+[2026-06-06 21:00:57,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4369.1, 300 sec: 4394.5). Total num frames: 2877440. Throughput: 0: 4351.4. Samples: 2884040. Policy #0 lag: (min: 12.0, avg: 25.4, max: 43.0)
+[2026-06-06 21:00:57,925][261293] Avg episode reward: [(0, '929.477')]
+[2026-06-06 21:00:59,413][262198] Updated weights for policy 0, policy_version 5630 (0.0009)
+[2026-06-06 21:01:00,841][262198] Updated weights for policy 0, policy_version 5642 (0.0011)
+[2026-06-06 21:01:01,767][262198] Updated weights for policy 0, policy_version 5655 (0.0009)
+[2026-06-06 21:01:02,921][261293] Fps is (10 sec: 4300.5, 60 sec: 4369.0, 300 sec: 4387.6). Total num frames: 2897920. Throughput: 0: 4359.1. Samples: 2897760. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:01:02,923][261293] Avg episode reward: [(0, '946.594')]
+[2026-06-06 21:01:03,862][262198] Updated weights for policy 0, policy_version 5665 (0.0010)
+[2026-06-06 21:01:04,924][262198] Updated weights for policy 0, policy_version 5680 (0.0010)
+[2026-06-06 21:01:06,100][262198] Updated weights for policy 0, policy_version 5692 (0.0009)
+[2026-06-06 21:01:07,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4369.1, 300 sec: 4394.5). Total num frames: 2920448. Throughput: 0: 4354.9. Samples: 2923806. Policy #0 lag: (min: 11.0, avg: 24.5, max: 43.0)
+[2026-06-06 21:01:07,923][261293] Avg episode reward: [(0, '905.694')]
+[2026-06-06 21:01:08,151][262198] Updated weights for policy 0, policy_version 5705 (0.0010)
+[2026-06-06 21:01:09,237][262198] Updated weights for policy 0, policy_version 5720 (0.0010)
+[2026-06-06 21:01:11,136][262198] Updated weights for policy 0, policy_version 5730 (0.0009)
+[2026-06-06 21:01:12,402][262198] Updated weights for policy 0, policy_version 5745 (0.0009)
+[2026-06-06 21:01:12,920][261293] Fps is (10 sec: 4710.7, 60 sec: 4403.2, 300 sec: 4401.5). Total num frames: 2945024. Throughput: 0: 4356.4. Samples: 2950486. Policy #0 lag: (min: 2.0, avg: 18.9, max: 38.0)
+[2026-06-06 21:01:12,922][261293] Avg episode reward: [(0, '938.665')]
+[2026-06-06 21:01:13,650][262198] Updated weights for policy 0, policy_version 5756 (0.0009)
+[2026-06-06 21:01:15,254][262198] Updated weights for policy 0, policy_version 5768 (0.0009)
+[2026-06-06 21:01:16,730][262198] Updated weights for policy 0, policy_version 5783 (0.0010)
+[2026-06-06 21:01:17,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.1, 300 sec: 4380.6). Total num frames: 2963456. Throughput: 0: 4340.3. Samples: 2963354. Policy #0 lag: (min: 11.0, avg: 24.8, max: 39.0)
+[2026-06-06 21:01:17,923][261293] Avg episode reward: [(0, '991.650')]
+[2026-06-06 21:01:18,648][262198] Updated weights for policy 0, policy_version 5793 (0.0009)
+[2026-06-06 21:01:19,783][262198] Updated weights for policy 0, policy_version 5806 (0.0009)
+[2026-06-06 21:01:20,998][262198] Updated weights for policy 0, policy_version 5817 (0.0010)
+[2026-06-06 21:01:22,385][262115] Signal inference workers to stop experience collection... (800 times)
+[2026-06-06 21:01:22,410][262198] InferenceWorker_p0-w0: stopping experience collection (800 times)
+[2026-06-06 21:01:22,456][262115] Signal inference workers to resume experience collection... (800 times)
+[2026-06-06 21:01:22,457][262198] InferenceWorker_p0-w0: resuming experience collection (800 times)
+[2026-06-06 21:01:22,459][262198] Updated weights for policy 0, policy_version 5828 (0.0009)
+[2026-06-06 21:01:22,922][261293] Fps is (10 sec: 4095.2, 60 sec: 4369.0, 300 sec: 4387.6). Total num frames: 2985984. Throughput: 0: 4338.2. Samples: 2989176. Policy #0 lag: (min: 0.0, avg: 16.7, max: 36.0)
+[2026-06-06 21:01:22,925][261293] Avg episode reward: [(0, '1052.483')]
+[2026-06-06 21:01:23,860][262198] Updated weights for policy 0, policy_version 5841 (0.0010)
+[2026-06-06 21:01:25,350][262198] Updated weights for policy 0, policy_version 5853 (0.0009)
+[2026-06-06 21:01:26,779][262198] Updated weights for policy 0, policy_version 5866 (0.0010)
+[2026-06-06 21:01:27,635][262198] Updated weights for policy 0, policy_version 5878 (0.0008)
+[2026-06-06 21:01:27,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4369.1, 300 sec: 4394.5). Total num frames: 3010560. Throughput: 0: 4402.4. Samples: 3016512. Policy #0 lag: (min: 0.0, avg: 16.7, max: 36.0)
+[2026-06-06 21:01:27,923][261293] Avg episode reward: [(0, '1118.208')]
+[2026-06-06 21:01:27,928][262115] Saving new best policy, reward=1118.208!
+[2026-06-06 21:01:29,011][262198] Updated weights for policy 0, policy_version 5888 (0.0009)
+[2026-06-06 21:01:30,608][262198] Updated weights for policy 0, policy_version 5901 (0.0010)
+[2026-06-06 21:01:31,846][262198] Updated weights for policy 0, policy_version 5914 (0.0010)
+[2026-06-06 21:01:32,921][261293] Fps is (10 sec: 4506.3, 60 sec: 4403.2, 300 sec: 4387.6). Total num frames: 3031040. Throughput: 0: 4421.6. Samples: 3030820. Policy #0 lag: (min: 3.0, avg: 22.1, max: 35.0)
+[2026-06-06 21:01:32,923][261293] Avg episode reward: [(0, '908.472')]
+[2026-06-06 21:01:33,472][262198] Updated weights for policy 0, policy_version 5926 (0.0010)
+[2026-06-06 21:01:34,790][262198] Updated weights for policy 0, policy_version 5939 (0.0010)
+[2026-06-06 21:01:36,087][262198] Updated weights for policy 0, policy_version 5949 (0.0010)
+[2026-06-06 21:01:37,286][262198] Updated weights for policy 0, policy_version 5960 (0.0008)
+[2026-06-06 21:01:37,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4403.2, 300 sec: 4394.5). Total num frames: 3053568. Throughput: 0: 4425.3. Samples: 3057596. Policy #0 lag: (min: 2.0, avg: 21.4, max: 42.0)
+[2026-06-06 21:01:37,922][261293] Avg episode reward: [(0, '1126.004')]
+[2026-06-06 21:01:38,060][262115] Saving new best policy, reward=1126.004!
+[2026-06-06 21:01:38,503][262198] Updated weights for policy 0, policy_version 5973 (0.0011)
+[2026-06-06 21:01:39,882][262198] Updated weights for policy 0, policy_version 5983 (0.0008)
+[2026-06-06 21:01:41,495][262198] Updated weights for policy 0, policy_version 5997 (0.0008)
+[2026-06-06 21:01:42,763][262198] Updated weights for policy 0, policy_version 6009 (0.0009)
+[2026-06-06 21:01:42,922][261293] Fps is (10 sec: 4710.0, 60 sec: 4437.3, 300 sec: 4394.5). Total num frames: 3078144. Throughput: 0: 4452.6. Samples: 3084414. Policy #0 lag: (min: 3.0, avg: 19.8, max: 35.0)
+[2026-06-06 21:01:42,924][261293] Avg episode reward: [(0, '993.931')]
+[2026-06-06 21:01:44,135][262198] Updated weights for policy 0, policy_version 6021 (0.0008)
+[2026-06-06 21:01:45,224][262198] Updated weights for policy 0, policy_version 6035 (0.0009)
+[2026-06-06 21:01:46,656][262198] Updated weights for policy 0, policy_version 6045 (0.0007)
+[2026-06-06 21:01:47,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4471.5, 300 sec: 4401.5). Total num frames: 3100672. Throughput: 0: 4466.6. Samples: 3098756. Policy #0 lag: (min: 2.0, avg: 23.1, max: 41.0)
+[2026-06-06 21:01:47,922][261293] Avg episode reward: [(0, '983.187')]
+[2026-06-06 21:01:47,942][262198] Updated weights for policy 0, policy_version 6058 (0.0008)
+[2026-06-06 21:01:48,904][262198] Updated weights for policy 0, policy_version 6070 (0.0008)
+[2026-06-06 21:01:50,088][262198] Updated weights for policy 0, policy_version 6080 (0.0008)
+[2026-06-06 21:01:51,834][262198] Updated weights for policy 0, policy_version 6094 (0.0010)
+[2026-06-06 21:01:52,920][261293] Fps is (10 sec: 4710.9, 60 sec: 4505.6, 300 sec: 4415.3). Total num frames: 3125248. Throughput: 0: 4545.7. Samples: 3128362. Policy #0 lag: (min: 4.0, avg: 21.9, max: 43.0)
+[2026-06-06 21:01:52,922][261293] Avg episode reward: [(0, '1074.051')]
+[2026-06-06 21:01:53,033][262198] Updated weights for policy 0, policy_version 6105 (0.0009)
+[2026-06-06 21:01:54,378][262198] Updated weights for policy 0, policy_version 6117 (0.0009)
+[2026-06-06 21:01:55,451][262198] Updated weights for policy 0, policy_version 6131 (0.0008)
+[2026-06-06 21:01:56,885][262198] Updated weights for policy 0, policy_version 6141 (0.0007)
+[2026-06-06 21:01:57,920][261293] Fps is (10 sec: 4710.4, 60 sec: 4505.6, 300 sec: 4408.4). Total num frames: 3147776. Throughput: 0: 4591.1. Samples: 3157086. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 21:01:57,922][261293] Avg episode reward: [(0, '1092.430')]
+[2026-06-06 21:01:58,434][262198] Updated weights for policy 0, policy_version 6153 (0.0010)
+[2026-06-06 21:01:59,381][262198] Updated weights for policy 0, policy_version 6167 (0.0009)
+[2026-06-06 21:02:01,086][262198] Updated weights for policy 0, policy_version 6177 (0.0008)
+[2026-06-06 21:02:01,878][262198] Updated weights for policy 0, policy_version 6188 (0.0008)
+[2026-06-06 21:02:02,816][262198] Updated weights for policy 0, policy_version 6199 (0.0008)
+[2026-06-06 21:02:02,920][261293] Fps is (10 sec: 4915.3, 60 sec: 4608.1, 300 sec: 4429.2). Total num frames: 3174400. Throughput: 0: 4608.1. Samples: 3170714. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 21:02:02,922][261293] Avg episode reward: [(0, '966.821')]
+[2026-06-06 21:02:04,596][262115] Signal inference workers to stop experience collection... (850 times)
+[2026-06-06 21:02:04,613][262198] InferenceWorker_p0-w0: stopping experience collection (850 times)
+[2026-06-06 21:02:04,615][262198] Updated weights for policy 0, policy_version 6209 (0.0008)
+[2026-06-06 21:02:04,732][262115] Signal inference workers to resume experience collection... (850 times)
+[2026-06-06 21:02:04,733][262198] InferenceWorker_p0-w0: resuming experience collection (850 times)
+[2026-06-06 21:02:05,711][262198] Updated weights for policy 0, policy_version 6221 (0.0009)
+[2026-06-06 21:02:06,426][262198] Updated weights for policy 0, policy_version 6232 (0.0009)
+[2026-06-06 21:02:07,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4573.9, 300 sec: 4422.3). Total num frames: 3194880. Throughput: 0: 4641.4. Samples: 3198034. Policy #0 lag: (min: 3.0, avg: 22.0, max: 35.0)
+[2026-06-06 21:02:07,922][261293] Avg episode reward: [(0, '1077.882')]
+[2026-06-06 21:02:08,502][262198] Updated weights for policy 0, policy_version 6244 (0.0009)
+[2026-06-06 21:02:09,771][262198] Updated weights for policy 0, policy_version 6259 (0.0008)
+[2026-06-06 21:02:11,651][262198] Updated weights for policy 0, policy_version 6269 (0.0009)
+[2026-06-06 21:02:12,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4505.6, 300 sec: 4408.4). Total num frames: 3215360. Throughput: 0: 4611.6. Samples: 3224032. Policy #0 lag: (min: 4.0, avg: 19.5, max: 36.0)
+[2026-06-06 21:02:12,925][261293] Avg episode reward: [(0, '1120.332')]
+[2026-06-06 21:02:13,221][262198] Updated weights for policy 0, policy_version 6281 (0.0011)
+[2026-06-06 21:02:14,296][262198] Updated weights for policy 0, policy_version 6294 (0.0011)
+[2026-06-06 21:02:15,893][262198] Updated weights for policy 0, policy_version 6304 (0.0010)
+[2026-06-06 21:02:17,502][262198] Updated weights for policy 0, policy_version 6320 (0.0011)
+[2026-06-06 21:02:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4573.9, 300 sec: 4415.4). Total num frames: 3237888. Throughput: 0: 4551.8. Samples: 3235650. Policy #0 lag: (min: 3.0, avg: 17.8, max: 35.0)
+[2026-06-06 21:02:17,922][261293] Avg episode reward: [(0, '1095.039')]
+[2026-06-06 21:02:19,002][262198] Updated weights for policy 0, policy_version 6330 (0.0009)
+[2026-06-06 21:02:20,237][262198] Updated weights for policy 0, policy_version 6340 (0.0009)
+[2026-06-06 21:02:21,473][262198] Updated weights for policy 0, policy_version 6352 (0.0009)
+[2026-06-06 21:02:22,716][262198] Updated weights for policy 0, policy_version 6364 (0.0009)
+[2026-06-06 21:02:22,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4539.8, 300 sec: 4401.5). Total num frames: 3258368. Throughput: 0: 4531.4. Samples: 3261510. Policy #0 lag: (min: 10.0, avg: 24.0, max: 42.0)
+[2026-06-06 21:02:22,923][261293] Avg episode reward: [(0, '1026.231')]
+[2026-06-06 21:02:24,160][262198] Updated weights for policy 0, policy_version 6376 (0.0010)
+[2026-06-06 21:02:25,568][262198] Updated weights for policy 0, policy_version 6390 (0.0009)
+[2026-06-06 21:02:27,178][262198] Updated weights for policy 0, policy_version 6400 (0.0008)
+[2026-06-06 21:02:27,922][261293] Fps is (10 sec: 4300.2, 60 sec: 4505.5, 300 sec: 4408.4). Total num frames: 3280896. Throughput: 0: 4547.2. Samples: 3289038. Policy #0 lag: (min: 9.0, avg: 21.2, max: 40.0)
+[2026-06-06 21:02:27,925][261293] Avg episode reward: [(0, '1183.389')]
+[2026-06-06 21:02:27,932][262115] Saving new best policy, reward=1183.389!
+[2026-06-06 21:02:28,720][262198] Updated weights for policy 0, policy_version 6416 (0.0008)
+[2026-06-06 21:02:30,063][262198] Updated weights for policy 0, policy_version 6426 (0.0009)
+[2026-06-06 21:02:31,725][262198] Updated weights for policy 0, policy_version 6438 (0.0009)
+[2026-06-06 21:02:32,674][262198] Updated weights for policy 0, policy_version 6450 (0.0011)
+[2026-06-06 21:02:32,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4539.7, 300 sec: 4408.4). Total num frames: 3303424. Throughput: 0: 4492.6. Samples: 3300924. Policy #0 lag: (min: 9.0, avg: 21.2, max: 40.0)
+[2026-06-06 21:02:32,925][261293] Avg episode reward: [(0, '1168.202')]
+[2026-06-06 21:02:33,856][262198] Updated weights for policy 0, policy_version 6460 (0.0009)
+[2026-06-06 21:02:35,854][262198] Updated weights for policy 0, policy_version 6474 (0.0009)
+[2026-06-06 21:02:36,964][262198] Updated weights for policy 0, policy_version 6487 (0.0009)
+[2026-06-06 21:02:37,921][261293] Fps is (10 sec: 4301.4, 60 sec: 4505.6, 300 sec: 4394.5). Total num frames: 3323904. Throughput: 0: 4418.2. Samples: 3327182. Policy #0 lag: (min: 3.0, avg: 24.0, max: 38.0)
+[2026-06-06 21:02:37,923][261293] Avg episode reward: [(0, '993.532')]
+[2026-06-06 21:02:39,012][262198] Updated weights for policy 0, policy_version 6500 (0.0008)
+[2026-06-06 21:02:39,885][262198] Updated weights for policy 0, policy_version 6510 (0.0009)
+[2026-06-06 21:02:41,216][262198] Updated weights for policy 0, policy_version 6521 (0.0010)
+[2026-06-06 21:02:42,815][262198] Updated weights for policy 0, policy_version 6534 (0.0009)
+[2026-06-06 21:02:42,922][261293] Fps is (10 sec: 4095.7, 60 sec: 4437.4, 300 sec: 4401.5). Total num frames: 3344384. Throughput: 0: 4380.8. Samples: 3354226. Policy #0 lag: (min: 7.0, avg: 22.6, max: 39.0)
+[2026-06-06 21:02:42,926][261293] Avg episode reward: [(0, '923.800')]
+[2026-06-06 21:02:43,936][262198] Updated weights for policy 0, policy_version 6546 (0.0009)
+[2026-06-06 21:02:45,103][262198] Updated weights for policy 0, policy_version 6556 (0.0007)
+[2026-06-06 21:02:46,962][262198] Updated weights for policy 0, policy_version 6571 (0.0008)
+[2026-06-06 21:02:47,924][261293] Fps is (10 sec: 4504.0, 60 sec: 4471.2, 300 sec: 4415.3). Total num frames: 3368960. Throughput: 0: 4352.5. Samples: 3366592. Policy #0 lag: (min: 1.0, avg: 19.2, max: 36.0)
+[2026-06-06 21:02:47,926][261293] Avg episode reward: [(0, '953.479')]
+[2026-06-06 21:02:48,016][262198] Updated weights for policy 0, policy_version 6583 (0.0008)
+[2026-06-06 21:02:49,923][262198] Updated weights for policy 0, policy_version 6596 (0.0008)
+[2026-06-06 21:02:50,526][262115] Signal inference workers to stop experience collection... (900 times)
+[2026-06-06 21:02:50,543][262198] InferenceWorker_p0-w0: stopping experience collection (900 times)
+[2026-06-06 21:02:50,578][262115] Signal inference workers to resume experience collection... (900 times)
+[2026-06-06 21:02:50,579][262198] InferenceWorker_p0-w0: resuming experience collection (900 times)
+[2026-06-06 21:02:50,835][262198] Updated weights for policy 0, policy_version 6606 (0.0008)
+[2026-06-06 21:02:52,064][262198] Updated weights for policy 0, policy_version 6617 (0.0007)
+[2026-06-06 21:02:52,920][261293] Fps is (10 sec: 4506.0, 60 sec: 4403.2, 300 sec: 4408.4). Total num frames: 3389440. Throughput: 0: 4347.5. Samples: 3393670. Policy #0 lag: (min: 6.0, avg: 24.1, max: 38.0)
+[2026-06-06 21:02:52,922][261293] Avg episode reward: [(0, '1132.649')]
+[2026-06-06 21:02:53,985][262198] Updated weights for policy 0, policy_version 6632 (0.0008)
+[2026-06-06 21:02:54,752][262198] Updated weights for policy 0, policy_version 6642 (0.0010)
+[2026-06-06 21:02:56,074][262198] Updated weights for policy 0, policy_version 6652 (0.0009)
+[2026-06-06 21:02:57,921][261293] Fps is (10 sec: 4302.2, 60 sec: 4403.2, 300 sec: 4415.3). Total num frames: 3411968. Throughput: 0: 4364.9. Samples: 3420452. Policy #0 lag: (min: 6.0, avg: 24.1, max: 38.0)
+[2026-06-06 21:02:57,923][261293] Avg episode reward: [(0, '976.638')]
+[2026-06-06 21:02:58,552][262198] Updated weights for policy 0, policy_version 6672 (0.0010)
+[2026-06-06 21:02:59,963][262198] Updated weights for policy 0, policy_version 6682 (0.0008)
+[2026-06-06 21:03:01,492][262198] Updated weights for policy 0, policy_version 6695 (0.0008)
+[2026-06-06 21:03:02,537][262198] Updated weights for policy 0, policy_version 6707 (0.0008)
+[2026-06-06 21:03:02,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4334.9, 300 sec: 4429.2). Total num frames: 3434496. Throughput: 0: 4384.3. Samples: 3432944. Policy #0 lag: (min: 8.0, avg: 20.1, max: 38.0)
+[2026-06-06 21:03:02,922][261293] Avg episode reward: [(0, '998.464')]
+[2026-06-06 21:03:04,419][262198] Updated weights for policy 0, policy_version 6717 (0.0008)
+[2026-06-06 21:03:05,429][262198] Updated weights for policy 0, policy_version 6731 (0.0007)
+[2026-06-06 21:03:06,386][262198] Updated weights for policy 0, policy_version 6743 (0.0008)
+[2026-06-06 21:03:07,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4334.9, 300 sec: 4422.3). Total num frames: 3454976. Throughput: 0: 4416.8. Samples: 3460268. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 21:03:07,923][261293] Avg episode reward: [(0, '1082.183')]
+[2026-06-06 21:03:08,467][262198] Updated weights for policy 0, policy_version 6758 (0.0008)
+[2026-06-06 21:03:09,435][262198] Updated weights for policy 0, policy_version 6770 (0.0008)
+[2026-06-06 21:03:10,595][262198] Updated weights for policy 0, policy_version 6780 (0.0008)
+[2026-06-06 21:03:12,539][262198] Updated weights for policy 0, policy_version 6796 (0.0010)
+[2026-06-06 21:03:12,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4437.4, 300 sec: 4429.2). Total num frames: 3481600. Throughput: 0: 4421.2. Samples: 3487986. Policy #0 lag: (min: 1.0, avg: 15.3, max: 33.0)
+[2026-06-06 21:03:12,921][261293] Avg episode reward: [(0, '869.770')]
+[2026-06-06 21:03:13,457][262198] Updated weights for policy 0, policy_version 6806 (0.0008)
+[2026-06-06 21:03:15,394][262198] Updated weights for policy 0, policy_version 6820 (0.0009)
+[2026-06-06 21:03:16,514][262198] Updated weights for policy 0, policy_version 6832 (0.0008)
+[2026-06-06 21:03:17,765][262198] Updated weights for policy 0, policy_version 6842 (0.0008)
+[2026-06-06 21:03:17,920][261293] Fps is (10 sec: 4915.4, 60 sec: 4437.3, 300 sec: 4436.2). Total num frames: 3504128. Throughput: 0: 4478.4. Samples: 3502450. Policy #0 lag: (min: 4.0, avg: 22.7, max: 36.0)
+[2026-06-06 21:03:17,922][261293] Avg episode reward: [(0, '916.590')]
+[2026-06-06 21:03:19,126][262198] Updated weights for policy 0, policy_version 6855 (0.0009)
+[2026-06-06 21:03:20,133][262198] Updated weights for policy 0, policy_version 6867 (0.0008)
+[2026-06-06 21:03:22,010][262198] Updated weights for policy 0, policy_version 6877 (0.0010)
+[2026-06-06 21:03:22,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.5, 300 sec: 4429.2). Total num frames: 3526656. Throughput: 0: 4493.2. Samples: 3529378. Policy #0 lag: (min: 4.0, avg: 22.7, max: 36.0)
+[2026-06-06 21:03:22,922][261293] Avg episode reward: [(0, '928.851')]
+[2026-06-06 21:03:23,310][262198] Updated weights for policy 0, policy_version 6892 (0.0009)
+[2026-06-06 21:03:24,241][262198] Updated weights for policy 0, policy_version 6902 (0.0009)
+[2026-06-06 21:03:26,054][262198] Updated weights for policy 0, policy_version 6913 (0.0009)
+[2026-06-06 21:03:27,134][262198] Updated weights for policy 0, policy_version 6927 (0.0009)
+[2026-06-06 21:03:27,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.6, 300 sec: 4429.2). Total num frames: 3549184. Throughput: 0: 4483.8. Samples: 3555994. Policy #0 lag: (min: 5.0, avg: 23.5, max: 39.0)
+[2026-06-06 21:03:27,922][261293] Avg episode reward: [(0, '1123.730')]
+[2026-06-06 21:03:28,439][262198] Updated weights for policy 0, policy_version 6937 (0.0007)
+[2026-06-06 21:03:29,565][262115] Signal inference workers to stop experience collection... (950 times)
+[2026-06-06 21:03:29,593][262198] InferenceWorker_p0-w0: stopping experience collection (950 times)
+[2026-06-06 21:03:29,701][262115] Signal inference workers to resume experience collection... (950 times)
+[2026-06-06 21:03:29,702][262198] InferenceWorker_p0-w0: resuming experience collection (950 times)
+[2026-06-06 21:03:29,862][262198] Updated weights for policy 0, policy_version 6950 (0.0008)
+[2026-06-06 21:03:30,889][262198] Updated weights for policy 0, policy_version 6961 (0.0009)
+[2026-06-06 21:03:32,107][262198] Updated weights for policy 0, policy_version 6971 (0.0007)
+[2026-06-06 21:03:32,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4437.3, 300 sec: 4436.2). Total num frames: 3569664. Throughput: 0: 4534.6. Samples: 3570636. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:03:32,923][261293] Avg episode reward: [(0, '1028.307')]
+[2026-06-06 21:03:33,486][262198] Updated weights for policy 0, policy_version 6983 (0.0008)
+[2026-06-06 21:03:34,659][262198] Updated weights for policy 0, policy_version 6995 (0.0009)
+[2026-06-06 21:03:36,435][262198] Updated weights for policy 0, policy_version 7005 (0.0008)
+[2026-06-06 21:03:37,446][262198] Updated weights for policy 0, policy_version 7017 (0.0008)
+[2026-06-06 21:03:37,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4539.7, 300 sec: 4450.1). Total num frames: 3596288. Throughput: 0: 4529.6. Samples: 3597506. Policy #0 lag: (min: 5.0, avg: 18.2, max: 38.0)
+[2026-06-06 21:03:37,923][261293] Avg episode reward: [(0, '1021.917')]
+[2026-06-06 21:03:38,285][262198] Updated weights for policy 0, policy_version 7027 (0.0008)
+[2026-06-06 21:03:40,073][262198] Updated weights for policy 0, policy_version 7037 (0.0008)
+[2026-06-06 21:03:41,216][262198] Updated weights for policy 0, policy_version 7049 (0.0008)
+[2026-06-06 21:03:42,118][262198] Updated weights for policy 0, policy_version 7061 (0.0010)
+[2026-06-06 21:03:42,920][261293] Fps is (10 sec: 4915.4, 60 sec: 4573.9, 300 sec: 4450.1). Total num frames: 3618816. Throughput: 0: 4535.7. Samples: 3624558. Policy #0 lag: (min: 5.0, avg: 18.2, max: 38.0)
+[2026-06-06 21:03:42,922][261293] Avg episode reward: [(0, '1151.050')]
+[2026-06-06 21:03:44,176][262198] Updated weights for policy 0, policy_version 7073 (0.0008)
+[2026-06-06 21:03:45,198][262198] Updated weights for policy 0, policy_version 7086 (0.0009)
+[2026-06-06 21:03:46,191][262198] Updated weights for policy 0, policy_version 7096 (0.0008)
+[2026-06-06 21:03:47,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4471.7, 300 sec: 4450.1). Total num frames: 3637248. Throughput: 0: 4552.9. Samples: 3637824. Policy #0 lag: (min: 1.0, avg: 15.6, max: 33.0)
+[2026-06-06 21:03:47,922][261293] Avg episode reward: [(0, '1157.611')]
+[2026-06-06 21:03:48,235][262198] Updated weights for policy 0, policy_version 7110 (0.0009)
+[2026-06-06 21:03:48,979][262198] Updated weights for policy 0, policy_version 7120 (0.0010)
+[2026-06-06 21:03:50,540][262198] Updated weights for policy 0, policy_version 7130 (0.0009)
+[2026-06-06 21:03:52,148][262198] Updated weights for policy 0, policy_version 7141 (0.0012)
+[2026-06-06 21:03:52,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4505.6, 300 sec: 4436.2). Total num frames: 3659776. Throughput: 0: 4514.7. Samples: 3663430. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 21:03:52,924][261293] Avg episode reward: [(0, '1053.863')]
+[2026-06-06 21:03:52,953][262198] Updated weights for policy 0, policy_version 7152 (0.0008)
+[2026-06-06 21:03:54,491][262198] Updated weights for policy 0, policy_version 7162 (0.0008)
+[2026-06-06 21:03:56,505][262198] Updated weights for policy 0, policy_version 7177 (0.0010)
+[2026-06-06 21:03:57,673][262198] Updated weights for policy 0, policy_version 7189 (0.0011)
+[2026-06-06 21:03:57,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4505.6, 300 sec: 4436.2). Total num frames: 3682304. Throughput: 0: 4441.6. Samples: 3687860. Policy #0 lag: (min: 4.0, avg: 22.2, max: 40.0)
+[2026-06-06 21:03:57,923][261293] Avg episode reward: [(0, '983.198')]
+[2026-06-06 21:03:59,621][262198] Updated weights for policy 0, policy_version 7200 (0.0009)
+[2026-06-06 21:04:00,584][262198] Updated weights for policy 0, policy_version 7212 (0.0010)
+[2026-06-06 21:04:01,686][262198] Updated weights for policy 0, policy_version 7222 (0.0009)
+[2026-06-06 21:04:02,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4437.3, 300 sec: 4422.3). Total num frames: 3700736. Throughput: 0: 4409.4. Samples: 3700874. Policy #0 lag: (min: 4.0, avg: 22.2, max: 40.0)
+[2026-06-06 21:04:02,922][261293] Avg episode reward: [(0, '951.820')]
+[2026-06-06 21:04:03,932][262198] Updated weights for policy 0, policy_version 7237 (0.0010)
+[2026-06-06 21:04:04,793][262198] Updated weights for policy 0, policy_version 7248 (0.0010)
+[2026-06-06 21:04:06,455][262198] Updated weights for policy 0, policy_version 7258 (0.0010)
+[2026-06-06 21:04:07,922][261293] Fps is (10 sec: 3686.0, 60 sec: 4403.1, 300 sec: 4401.4). Total num frames: 3719168. Throughput: 0: 4353.4. Samples: 3725286. Policy #0 lag: (min: 6.0, avg: 18.2, max: 38.0)
+[2026-06-06 21:04:07,924][261293] Avg episode reward: [(0, '1165.982')]
+[2026-06-06 21:04:08,601][262198] Updated weights for policy 0, policy_version 7276 (0.0009)
+[2026-06-06 21:04:09,683][262198] Updated weights for policy 0, policy_version 7286 (0.0011)
+[2026-06-06 21:04:11,795][262198] Updated weights for policy 0, policy_version 7300 (0.0010)
+[2026-06-06 21:04:12,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.0, 300 sec: 4415.4). Total num frames: 3743744. Throughput: 0: 4322.2. Samples: 3750496. Policy #0 lag: (min: 6.0, avg: 25.6, max: 38.0)
+[2026-06-06 21:04:12,923][261293] Avg episode reward: [(0, '1190.813')]
+[2026-06-06 21:04:12,998][262198] Updated weights for policy 0, policy_version 7314 (0.0009)
+[2026-06-06 21:04:13,067][262115] Saving new best policy, reward=1190.813!
+[2026-06-06 21:04:14,283][262198] Updated weights for policy 0, policy_version 7324 (0.0009)
+[2026-06-06 21:04:15,101][262115] Signal inference workers to stop experience collection... (1000 times)
+[2026-06-06 21:04:15,119][262198] InferenceWorker_p0-w0: stopping experience collection (1000 times)
+[2026-06-06 21:04:15,221][262115] Signal inference workers to resume experience collection... (1000 times)
+[2026-06-06 21:04:15,222][262198] InferenceWorker_p0-w0: resuming experience collection (1000 times)
+[2026-06-06 21:04:16,288][262198] Updated weights for policy 0, policy_version 7340 (0.0011)
+[2026-06-06 21:04:17,410][262198] Updated weights for policy 0, policy_version 7352 (0.0008)
+[2026-06-06 21:04:17,921][261293] Fps is (10 sec: 4506.0, 60 sec: 4334.9, 300 sec: 4408.4). Total num frames: 3764224. Throughput: 0: 4298.0. Samples: 3764044. Policy #0 lag: (min: 1.0, avg: 21.2, max: 33.0)
+[2026-06-06 21:04:17,924][261293] Avg episode reward: [(0, '1176.034')]
+[2026-06-06 21:04:19,722][262198] Updated weights for policy 0, policy_version 7366 (0.0010)
+[2026-06-06 21:04:20,621][262198] Updated weights for policy 0, policy_version 7379 (0.0009)
+[2026-06-06 21:04:22,920][261293] Fps is (10 sec: 3891.4, 60 sec: 4266.7, 300 sec: 4394.5). Total num frames: 3782656. Throughput: 0: 4243.1. Samples: 3788442. Policy #0 lag: (min: 1.0, avg: 21.2, max: 33.0)
+[2026-06-06 21:04:22,922][261293] Avg episode reward: [(0, '1190.067')]
+[2026-06-06 21:04:22,985][262198] Updated weights for policy 0, policy_version 7389 (0.0009)
+[2026-06-06 21:04:24,036][262198] Updated weights for policy 0, policy_version 7403 (0.0013)
+[2026-06-06 21:04:25,115][262198] Updated weights for policy 0, policy_version 7413 (0.0009)
+[2026-06-06 21:04:27,003][262198] Updated weights for policy 0, policy_version 7425 (0.0010)
+[2026-06-06 21:04:27,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4300.8, 300 sec: 4408.4). Total num frames: 3807232. Throughput: 0: 4233.5. Samples: 3815066. Policy #0 lag: (min: 6.0, avg: 20.9, max: 39.0)
+[2026-06-06 21:04:27,923][261293] Avg episode reward: [(0, '1137.580')]
+[2026-06-06 21:04:28,245][262198] Updated weights for policy 0, policy_version 7441 (0.0009)
+[2026-06-06 21:04:29,645][262198] Updated weights for policy 0, policy_version 7451 (0.0009)
+[2026-06-06 21:04:31,106][262198] Updated weights for policy 0, policy_version 7463 (0.0011)
+[2026-06-06 21:04:32,326][262198] Updated weights for policy 0, policy_version 7477 (0.0010)
+[2026-06-06 21:04:32,921][261293] Fps is (10 sec: 4710.1, 60 sec: 4334.9, 300 sec: 4408.4). Total num frames: 3829760. Throughput: 0: 4236.2. Samples: 3828456. Policy #0 lag: (min: 7.0, avg: 19.4, max: 39.0)
+[2026-06-06 21:04:32,924][261293] Avg episode reward: [(0, '1114.012')]
+[2026-06-06 21:04:34,112][262198] Updated weights for policy 0, policy_version 7488 (0.0010)
+[2026-06-06 21:04:35,617][262198] Updated weights for policy 0, policy_version 7505 (0.0010)
+[2026-06-06 21:04:36,783][262198] Updated weights for policy 0, policy_version 7515 (0.0007)
+[2026-06-06 21:04:37,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4232.6, 300 sec: 4401.5). Total num frames: 3850240. Throughput: 0: 4245.4. Samples: 3854474. Policy #0 lag: (min: 3.0, avg: 17.1, max: 35.0)
+[2026-06-06 21:04:37,922][261293] Avg episode reward: [(0, '1140.798')]
+[2026-06-06 21:04:38,303][262198] Updated weights for policy 0, policy_version 7525 (0.0008)
+[2026-06-06 21:04:39,316][262198] Updated weights for policy 0, policy_version 7539 (0.0010)
+[2026-06-06 21:04:41,417][262198] Updated weights for policy 0, policy_version 7549 (0.0009)
+[2026-06-06 21:04:42,565][262198] Updated weights for policy 0, policy_version 7564 (0.0010)
+[2026-06-06 21:04:42,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4266.6, 300 sec: 4422.3). Total num frames: 3874816. Throughput: 0: 4290.6. Samples: 3880938. Policy #0 lag: (min: 3.0, avg: 17.1, max: 35.0)
+[2026-06-06 21:04:42,922][261293] Avg episode reward: [(0, '1091.341')]
+[2026-06-06 21:04:43,504][262198] Updated weights for policy 0, policy_version 7575 (0.0009)
+[2026-06-06 21:04:45,394][262198] Updated weights for policy 0, policy_version 7586 (0.0009)
+[2026-06-06 21:04:46,463][262198] Updated weights for policy 0, policy_version 7599 (0.0010)
+[2026-06-06 21:04:47,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4300.8, 300 sec: 4415.4). Total num frames: 3895296. Throughput: 0: 4317.3. Samples: 3895152. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 21:04:47,923][261293] Avg episode reward: [(0, '1182.432')]
+[2026-06-06 21:04:48,063][262198] Updated weights for policy 0, policy_version 7609 (0.0009)
+[2026-06-06 21:04:49,366][262198] Updated weights for policy 0, policy_version 7621 (0.0009)
+[2026-06-06 21:04:50,507][262198] Updated weights for policy 0, policy_version 7635 (0.0009)
+[2026-06-06 21:04:52,517][262115] Signal inference workers to stop experience collection... (1050 times)
+[2026-06-06 21:04:52,536][262198] InferenceWorker_p0-w0: stopping experience collection (1050 times)
+[2026-06-06 21:04:52,665][262115] Signal inference workers to resume experience collection... (1050 times)
+[2026-06-06 21:04:52,665][262198] InferenceWorker_p0-w0: resuming experience collection (1050 times)
+[2026-06-06 21:04:52,668][262198] Updated weights for policy 0, policy_version 7648 (0.0010)
+[2026-06-06 21:04:52,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4266.6, 300 sec: 4408.4). Total num frames: 3915776. Throughput: 0: 4348.0. Samples: 3920942. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:04:52,925][261293] Avg episode reward: [(0, '1008.019')]
+[2026-06-06 21:04:53,764][262198] Updated weights for policy 0, policy_version 7661 (0.0058)
+[2026-06-06 21:04:54,668][262198] Updated weights for policy 0, policy_version 7671 (0.0008)
+[2026-06-06 21:04:56,863][262198] Updated weights for policy 0, policy_version 7684 (0.0010)
+[2026-06-06 21:04:57,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4300.8, 300 sec: 4422.3). Total num frames: 3940352. Throughput: 0: 4371.3. Samples: 3947202. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:04:57,922][261293] Avg episode reward: [(0, '1050.846')]
+[2026-06-06 21:04:58,107][262198] Updated weights for policy 0, policy_version 7699 (0.0011)
+[2026-06-06 21:05:00,669][262198] Updated weights for policy 0, policy_version 7709 (0.0013)
+[2026-06-06 21:05:01,455][262198] Updated weights for policy 0, policy_version 7719 (0.0013)
+[2026-06-06 21:05:02,695][262198] Updated weights for policy 0, policy_version 7731 (0.0010)
+[2026-06-06 21:05:02,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4300.8, 300 sec: 4408.4). Total num frames: 3958784. Throughput: 0: 4300.5. Samples: 3957568. Policy #0 lag: (min: 3.0, avg: 24.5, max: 35.0)
+[2026-06-06 21:05:02,923][261293] Avg episode reward: [(0, '1115.014')]
+[2026-06-06 21:05:04,812][262198] Updated weights for policy 0, policy_version 7741 (0.0009)
+[2026-06-06 21:05:06,345][262198] Updated weights for policy 0, policy_version 7757 (0.0012)
+[2026-06-06 21:05:07,921][261293] Fps is (10 sec: 3686.3, 60 sec: 4300.8, 300 sec: 4394.5). Total num frames: 3977216. Throughput: 0: 4288.6. Samples: 3981430. Policy #0 lag: (min: 8.0, avg: 18.2, max: 37.0)
+[2026-06-06 21:05:07,925][261293] Avg episode reward: [(0, '1038.849')]
+[2026-06-06 21:05:08,156][262198] Updated weights for policy 0, policy_version 7769 (0.0010)
+[2026-06-06 21:05:09,209][262198] Updated weights for policy 0, policy_version 7783 (0.0008)
+[2026-06-06 21:05:10,426][262198] Updated weights for policy 0, policy_version 7794 (0.0012)
+[2026-06-06 21:05:12,369][262198] Updated weights for policy 0, policy_version 7804 (0.0009)
+[2026-06-06 21:05:12,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4232.5, 300 sec: 4394.5). Total num frames: 3997696. Throughput: 0: 4260.5. Samples: 4006788. Policy #0 lag: (min: 1.0, avg: 21.7, max: 38.0)
+[2026-06-06 21:05:12,925][261293] Avg episode reward: [(0, '1404.392')]
+[2026-06-06 21:05:13,238][262115] Saving new best policy, reward=1404.392!
+[2026-06-06 21:05:13,625][262198] Updated weights for policy 0, policy_version 7816 (0.0009)
+[2026-06-06 21:05:14,788][262198] Updated weights for policy 0, policy_version 7829 (0.0010)
+[2026-06-06 21:05:16,945][262198] Updated weights for policy 0, policy_version 7841 (0.0009)
+[2026-06-06 21:05:17,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4266.6, 300 sec: 4394.5). Total num frames: 4020224. Throughput: 0: 4231.9. Samples: 4018890. Policy #0 lag: (min: 1.0, avg: 21.7, max: 38.0)
+[2026-06-06 21:05:17,924][261293] Avg episode reward: [(0, '1204.552')]
+[2026-06-06 21:05:17,958][262198] Updated weights for policy 0, policy_version 7855 (0.0011)
+[2026-06-06 21:05:19,832][262198] Updated weights for policy 0, policy_version 7865 (0.0009)
+[2026-06-06 21:05:21,087][262198] Updated weights for policy 0, policy_version 7878 (0.0010)
+[2026-06-06 21:05:22,171][262198] Updated weights for policy 0, policy_version 7891 (0.0010)
+[2026-06-06 21:05:22,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4334.9, 300 sec: 4387.6). Total num frames: 4042752. Throughput: 0: 4224.1. Samples: 4044558. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 21:05:22,923][261293] Avg episode reward: [(0, '1201.605')]
+[2026-06-06 21:05:24,300][262198] Updated weights for policy 0, policy_version 7901 (0.0010)
+[2026-06-06 21:05:25,344][262198] Updated weights for policy 0, policy_version 7914 (0.0009)
+[2026-06-06 21:05:26,258][262198] Updated weights for policy 0, policy_version 7926 (0.0010)
+[2026-06-06 21:05:27,920][261293] Fps is (10 sec: 4096.3, 60 sec: 4232.6, 300 sec: 4387.6). Total num frames: 4061184. Throughput: 0: 4221.5. Samples: 4070904. Policy #0 lag: (min: 7.0, avg: 27.0, max: 38.0)
+[2026-06-06 21:05:27,923][261293] Avg episode reward: [(0, '1369.735')]
+[2026-06-06 21:05:28,207][262198] Updated weights for policy 0, policy_version 7936 (0.0008)
+[2026-06-06 21:05:29,694][262198] Updated weights for policy 0, policy_version 7953 (0.0010)
+[2026-06-06 21:05:29,882][262115] Signal inference workers to stop experience collection... (1100 times)
+[2026-06-06 21:05:29,898][262198] InferenceWorker_p0-w0: stopping experience collection (1100 times)
+[2026-06-06 21:05:29,913][262115] Signal inference workers to resume experience collection... (1100 times)
+[2026-06-06 21:05:29,914][262198] InferenceWorker_p0-w0: resuming experience collection (1100 times)
+[2026-06-06 21:05:31,314][262198] Updated weights for policy 0, policy_version 7963 (0.0010)
+[2026-06-06 21:05:32,469][262198] Updated weights for policy 0, policy_version 7976 (0.0009)
+[2026-06-06 21:05:32,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4232.5, 300 sec: 4387.6). Total num frames: 4083712. Throughput: 0: 4182.1. Samples: 4083348. Policy #0 lag: (min: 7.0, avg: 27.0, max: 38.0)
+[2026-06-06 21:05:32,924][261293] Avg episode reward: [(0, '1183.681')]
+[2026-06-06 21:05:33,634][262198] Updated weights for policy 0, policy_version 7988 (0.0010)
+[2026-06-06 21:05:35,605][262198] Updated weights for policy 0, policy_version 8000 (0.0009)
+[2026-06-06 21:05:36,755][262198] Updated weights for policy 0, policy_version 8015 (0.0013)
+[2026-06-06 21:05:37,920][261293] Fps is (10 sec: 4710.4, 60 sec: 4300.8, 300 sec: 4394.5). Total num frames: 4108288. Throughput: 0: 4217.8. Samples: 4110742. Policy #0 lag: (min: 6.0, avg: 24.9, max: 38.0)
+[2026-06-06 21:05:37,922][261293] Avg episode reward: [(0, '1267.090')]
+[2026-06-06 21:05:38,906][262198] Updated weights for policy 0, policy_version 8027 (0.0009)
+[2026-06-06 21:05:40,169][262198] Updated weights for policy 0, policy_version 8041 (0.0009)
+[2026-06-06 21:05:41,182][262198] Updated weights for policy 0, policy_version 8055 (0.0010)
+[2026-06-06 21:05:42,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4198.4, 300 sec: 4387.6). Total num frames: 4126720. Throughput: 0: 4226.6. Samples: 4137396. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:05:42,922][261293] Avg episode reward: [(0, '1181.596')]
+[2026-06-06 21:05:43,404][262198] Updated weights for policy 0, policy_version 8068 (0.0010)
+[2026-06-06 21:05:44,325][262198] Updated weights for policy 0, policy_version 8080 (0.0010)
+[2026-06-06 21:05:46,829][262198] Updated weights for policy 0, policy_version 8096 (0.0011)
+[2026-06-06 21:05:47,922][261293] Fps is (10 sec: 4300.0, 60 sec: 4266.6, 300 sec: 4394.5). Total num frames: 4151296. Throughput: 0: 4257.0. Samples: 4149140. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:05:47,926][261293] Avg episode reward: [(0, '1044.547')]
+[2026-06-06 21:05:48,030][262198] Updated weights for policy 0, policy_version 8110 (0.0010)
+[2026-06-06 21:05:50,457][262198] Updated weights for policy 0, policy_version 8122 (0.0009)
+[2026-06-06 21:05:51,931][262198] Updated weights for policy 0, policy_version 8140 (0.0011)
+[2026-06-06 21:05:52,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4300.8, 300 sec: 4394.5). Total num frames: 4173824. Throughput: 0: 4311.0. Samples: 4175424. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:05:52,922][261293] Avg episode reward: [(0, '1175.141')]
+[2026-06-06 21:05:54,107][262198] Updated weights for policy 0, policy_version 8153 (0.0012)
+[2026-06-06 21:05:55,250][262198] Updated weights for policy 0, policy_version 8168 (0.0010)
+[2026-06-06 21:05:56,524][262198] Updated weights for policy 0, policy_version 8181 (0.0009)
+[2026-06-06 21:05:57,920][261293] Fps is (10 sec: 3891.9, 60 sec: 4164.3, 300 sec: 4380.6). Total num frames: 4190208. Throughput: 0: 4308.0. Samples: 4200648. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:05:57,922][261293] Avg episode reward: [(0, '1268.910')]
+[2026-06-06 21:05:58,805][262198] Updated weights for policy 0, policy_version 8196 (0.0010)
+[2026-06-06 21:05:59,831][262198] Updated weights for policy 0, policy_version 8208 (0.0010)
+[2026-06-06 21:06:02,484][262198] Updated weights for policy 0, policy_version 8222 (0.0011)
+[2026-06-06 21:06:02,912][262115] Signal inference workers to stop experience collection... (1150 times)
+[2026-06-06 21:06:02,919][262115] Signal inference workers to resume experience collection... (1150 times)
+[2026-06-06 21:06:02,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4232.5, 300 sec: 4380.6). Total num frames: 4212736. Throughput: 0: 4306.1. Samples: 4212666. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 21:06:02,926][261293] Avg episode reward: [(0, '1104.165')]
+[2026-06-06 21:06:02,931][262198] InferenceWorker_p0-w0: stopping experience collection (1150 times)
+[2026-06-06 21:06:02,948][262198] InferenceWorker_p0-w0: resuming experience collection (1150 times)
+[2026-06-06 21:06:03,434][262198] Updated weights for policy 0, policy_version 8233 (0.0009)
+[2026-06-06 21:06:04,662][262198] Updated weights for policy 0, policy_version 8246 (0.0011)
+[2026-06-06 21:06:06,927][262198] Updated weights for policy 0, policy_version 8259 (0.0010)
+[2026-06-06 21:06:07,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4300.8, 300 sec: 4373.7). Total num frames: 4235264. Throughput: 0: 4292.6. Samples: 4237724. Policy #0 lag: (min: 10.0, avg: 27.7, max: 42.0)
+[2026-06-06 21:06:07,922][261293] Avg episode reward: [(0, '1176.396')]
+[2026-06-06 21:06:08,423][262198] Updated weights for policy 0, policy_version 8273 (0.0011)
+[2026-06-06 21:06:10,483][262198] Updated weights for policy 0, policy_version 8288 (0.0011)
+[2026-06-06 21:06:11,653][262198] Updated weights for policy 0, policy_version 8301 (0.0009)
+[2026-06-06 21:06:12,746][262198] Updated weights for policy 0, policy_version 8312 (0.0009)
+[2026-06-06 21:06:12,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4300.8, 300 sec: 4380.6). Total num frames: 4255744. Throughput: 0: 4247.9. Samples: 4262062. Policy #0 lag: (min: 10.0, avg: 27.7, max: 42.0)
+[2026-06-06 21:06:12,924][261293] Avg episode reward: [(0, '1162.340')]
+[2026-06-06 21:06:15,286][262198] Updated weights for policy 0, policy_version 8328 (0.0010)
+[2026-06-06 21:06:16,596][262198] Updated weights for policy 0, policy_version 8342 (0.0011)
+[2026-06-06 21:06:17,921][261293] Fps is (10 sec: 3686.2, 60 sec: 4198.4, 300 sec: 4359.8). Total num frames: 4272128. Throughput: 0: 4249.5. Samples: 4274578. Policy #0 lag: (min: 1.0, avg: 14.7, max: 33.0)
+[2026-06-06 21:06:17,923][261293] Avg episode reward: [(0, '1241.594')]
+[2026-06-06 21:06:18,529][262198] Updated weights for policy 0, policy_version 8356 (0.0010)
+[2026-06-06 21:06:19,425][262198] Updated weights for policy 0, policy_version 8366 (0.0012)
+[2026-06-06 21:06:21,976][262198] Updated weights for policy 0, policy_version 8380 (0.0010)
+[2026-06-06 21:06:22,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4198.4, 300 sec: 4352.9). Total num frames: 4294656. Throughput: 0: 4205.6. Samples: 4299996. Policy #0 lag: (min: 3.0, avg: 15.3, max: 35.0)
+[2026-06-06 21:06:22,922][261293] Avg episode reward: [(0, '1142.425')]
+[2026-06-06 21:06:23,036][262198] Updated weights for policy 0, policy_version 8392 (0.0009)
+[2026-06-06 21:06:24,253][262198] Updated weights for policy 0, policy_version 8402 (0.0010)
+[2026-06-06 21:06:26,139][262198] Updated weights for policy 0, policy_version 8414 (0.0011)
+[2026-06-06 21:06:26,999][262198] Updated weights for policy 0, policy_version 8426 (0.0010)
+[2026-06-06 21:06:27,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4266.6, 300 sec: 4359.8). Total num frames: 4317184. Throughput: 0: 4163.0. Samples: 4324732. Policy #0 lag: (min: 3.0, avg: 15.3, max: 35.0)
+[2026-06-06 21:06:27,924][261293] Avg episode reward: [(0, '1354.773')]
+[2026-06-06 21:06:28,187][262198] Updated weights for policy 0, policy_version 8437 (0.0009)
+[2026-06-06 21:06:29,957][262198] Updated weights for policy 0, policy_version 8449 (0.0009)
+[2026-06-06 21:06:30,676][262198] Updated weights for policy 0, policy_version 8459 (0.0009)
+[2026-06-06 21:06:31,947][262198] Updated weights for policy 0, policy_version 8471 (0.0009)
+[2026-06-06 21:06:32,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4232.6, 300 sec: 4352.9). Total num frames: 4337664. Throughput: 0: 4202.3. Samples: 4338236. Policy #0 lag: (min: 1.0, avg: 14.3, max: 33.0)
+[2026-06-06 21:06:32,922][261293] Avg episode reward: [(0, '1189.437')]
+[2026-06-06 21:06:33,852][262198] Updated weights for policy 0, policy_version 8482 (0.0009)
+[2026-06-06 21:06:34,715][262198] Updated weights for policy 0, policy_version 8493 (0.0010)
+[2026-06-06 21:06:37,064][262198] Updated weights for policy 0, policy_version 8505 (0.0009)
+[2026-06-06 21:06:37,560][262115] Signal inference workers to stop experience collection... (1200 times)
+[2026-06-06 21:06:37,592][262198] InferenceWorker_p0-w0: stopping experience collection (1200 times)
+[2026-06-06 21:06:37,595][262115] Signal inference workers to resume experience collection... (1200 times)
+[2026-06-06 21:06:37,609][262198] InferenceWorker_p0-w0: resuming experience collection (1200 times)
+[2026-06-06 21:06:37,921][261293] Fps is (10 sec: 4096.2, 60 sec: 4164.3, 300 sec: 4339.0). Total num frames: 4358144. Throughput: 0: 4199.7. Samples: 4364408. Policy #0 lag: (min: 1.0, avg: 14.2, max: 33.0)
+[2026-06-06 21:06:37,924][261293] Avg episode reward: [(0, '1150.118')]
+[2026-06-06 21:06:38,473][262198] Updated weights for policy 0, policy_version 8522 (0.0011)
+[2026-06-06 21:06:39,736][262198] Updated weights for policy 0, policy_version 8533 (0.0010)
+[2026-06-06 21:06:41,605][262198] Updated weights for policy 0, policy_version 8545 (0.0009)
+[2026-06-06 21:06:42,666][262198] Updated weights for policy 0, policy_version 8557 (0.0010)
+[2026-06-06 21:06:42,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4266.7, 300 sec: 4345.9). Total num frames: 4382720. Throughput: 0: 4176.0. Samples: 4388568. Policy #0 lag: (min: 1.0, avg: 14.2, max: 33.0)
+[2026-06-06 21:06:42,922][261293] Avg episode reward: [(0, '1146.428')]
+[2026-06-06 21:06:43,748][262198] Updated weights for policy 0, policy_version 8568 (0.0010)
+[2026-06-06 21:06:46,219][262198] Updated weights for policy 0, policy_version 8584 (0.0011)
+[2026-06-06 21:06:47,660][262198] Updated weights for policy 0, policy_version 8596 (0.0013)
+[2026-06-06 21:06:47,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4198.5, 300 sec: 4332.0). Total num frames: 4403200. Throughput: 0: 4194.8. Samples: 4401430. Policy #0 lag: (min: 10.0, avg: 26.8, max: 42.0)
+[2026-06-06 21:06:47,923][261293] Avg episode reward: [(0, '1108.547')]
+[2026-06-06 21:06:49,589][262198] Updated weights for policy 0, policy_version 8608 (0.0010)
+[2026-06-06 21:06:50,787][262198] Updated weights for policy 0, policy_version 8622 (0.0009)
+[2026-06-06 21:06:52,921][261293] Fps is (10 sec: 3686.4, 60 sec: 4096.0, 300 sec: 4311.2). Total num frames: 4419584. Throughput: 0: 4175.4. Samples: 4425618. Policy #0 lag: (min: 10.0, avg: 26.8, max: 42.0)
+[2026-06-06 21:06:52,927][261293] Avg episode reward: [(0, '1096.135')]
+[2026-06-06 21:06:52,940][262198] Updated weights for policy 0, policy_version 8634 (0.0011)
+[2026-06-06 21:06:53,906][262198] Updated weights for policy 0, policy_version 8645 (0.0012)
+[2026-06-06 21:06:54,729][262198] Updated weights for policy 0, policy_version 8655 (0.0009)
+[2026-06-06 21:06:56,934][262198] Updated weights for policy 0, policy_version 8667 (0.0010)
+[2026-06-06 21:06:57,920][261293] Fps is (10 sec: 3891.1, 60 sec: 4198.4, 300 sec: 4297.3). Total num frames: 4442112. Throughput: 0: 4198.9. Samples: 4451012. Policy #0 lag: (min: 3.0, avg: 19.6, max: 43.0)
+[2026-06-06 21:06:57,922][261293] Avg episode reward: [(0, '1169.089')]
+[2026-06-06 21:06:57,956][262198] Updated weights for policy 0, policy_version 8680 (0.0008)
+[2026-06-06 21:06:59,152][262198] Updated weights for policy 0, policy_version 8690 (0.0008)
+[2026-06-06 21:07:01,022][262198] Updated weights for policy 0, policy_version 8704 (0.0008)
+[2026-06-06 21:07:02,190][262198] Updated weights for policy 0, policy_version 8718 (0.0009)
+[2026-06-06 21:07:02,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4198.4, 300 sec: 4304.3). Total num frames: 4464640. Throughput: 0: 4227.4. Samples: 4464808. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:07:02,924][261293] Avg episode reward: [(0, '1203.175')]
+[2026-06-06 21:07:04,641][262198] Updated weights for policy 0, policy_version 8732 (0.0009)
+[2026-06-06 21:07:05,625][262198] Updated weights for policy 0, policy_version 8744 (0.0008)
+[2026-06-06 21:07:06,862][262198] Updated weights for policy 0, policy_version 8756 (0.0008)
+[2026-06-06 21:07:07,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4164.3, 300 sec: 4304.3). Total num frames: 4485120. Throughput: 0: 4212.2. Samples: 4489542. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:07:07,922][261293] Avg episode reward: [(0, '1209.040')]
+[2026-06-06 21:07:08,854][262198] Updated weights for policy 0, policy_version 8769 (0.0009)
+[2026-06-06 21:07:09,800][262198] Updated weights for policy 0, policy_version 8781 (0.0008)
+[2026-06-06 21:07:11,979][262198] Updated weights for policy 0, policy_version 8793 (0.0008)
+[2026-06-06 21:07:12,685][262115] Signal inference workers to stop experience collection... (1250 times)
+[2026-06-06 21:07:12,702][262198] InferenceWorker_p0-w0: stopping experience collection (1250 times)
+[2026-06-06 21:07:12,736][262115] Signal inference workers to resume experience collection... (1250 times)
+[2026-06-06 21:07:12,737][262198] InferenceWorker_p0-w0: resuming experience collection (1250 times)
+[2026-06-06 21:07:12,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4198.4, 300 sec: 4304.3). Total num frames: 4507648. Throughput: 0: 4251.9. Samples: 4516064. Policy #0 lag: (min: 10.0, avg: 24.8, max: 41.0)
+[2026-06-06 21:07:12,921][261293] Avg episode reward: [(0, '1269.040')]
+[2026-06-06 21:07:12,967][262198] Updated weights for policy 0, policy_version 8806 (0.0008)
+[2026-06-06 21:07:14,059][262198] Updated weights for policy 0, policy_version 8817 (0.0007)
+[2026-06-06 21:07:16,012][262198] Updated weights for policy 0, policy_version 8832 (0.0007)
+[2026-06-06 21:07:17,351][262198] Updated weights for policy 0, policy_version 8845 (0.0010)
+[2026-06-06 21:07:17,920][261293] Fps is (10 sec: 4505.6, 60 sec: 4300.8, 300 sec: 4311.2). Total num frames: 4530176. Throughput: 0: 4278.1. Samples: 4530750. Policy #0 lag: (min: 7.0, avg: 19.1, max: 39.0)
+[2026-06-06 21:07:17,922][261293] Avg episode reward: [(0, '1309.965')]
+[2026-06-06 21:07:18,310][262198] Updated weights for policy 0, policy_version 8856 (0.0008)
+[2026-06-06 21:07:20,487][262198] Updated weights for policy 0, policy_version 8872 (0.0009)
+[2026-06-06 21:07:21,777][262198] Updated weights for policy 0, policy_version 8884 (0.0010)
+[2026-06-06 21:07:22,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4266.7, 300 sec: 4304.3). Total num frames: 4550656. Throughput: 0: 4253.3. Samples: 4555804. Policy #0 lag: (min: 7.0, avg: 19.1, max: 39.0)
+[2026-06-06 21:07:22,923][261293] Avg episode reward: [(0, '1270.775')]
+[2026-06-06 21:07:23,670][262198] Updated weights for policy 0, policy_version 8898 (0.0008)
+[2026-06-06 21:07:24,785][262198] Updated weights for policy 0, policy_version 8910 (0.0010)
+[2026-06-06 21:07:27,190][262198] Updated weights for policy 0, policy_version 8921 (0.0010)
+[2026-06-06 21:07:27,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4266.7, 300 sec: 4304.3). Total num frames: 4573184. Throughput: 0: 4285.5. Samples: 4581416. Policy #0 lag: (min: 7.0, avg: 19.8, max: 39.0)
+[2026-06-06 21:07:27,927][261293] Avg episode reward: [(0, '1138.689')]
+[2026-06-06 21:07:28,377][262198] Updated weights for policy 0, policy_version 8935 (0.0009)
+[2026-06-06 21:07:29,657][262198] Updated weights for policy 0, policy_version 8945 (0.0010)
+[2026-06-06 21:07:31,706][262198] Updated weights for policy 0, policy_version 8958 (0.0009)
+[2026-06-06 21:07:32,661][262198] Updated weights for policy 0, policy_version 8969 (0.0009)
+[2026-06-06 21:07:32,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4266.7, 300 sec: 4304.3). Total num frames: 4593664. Throughput: 0: 4255.3. Samples: 4592918. Policy #0 lag: (min: 10.0, avg: 25.9, max: 42.0)
+[2026-06-06 21:07:32,923][261293] Avg episode reward: [(0, '1246.498')]
+[2026-06-06 21:07:33,803][262198] Updated weights for policy 0, policy_version 8980 (0.0010)
+[2026-06-06 21:07:35,880][262198] Updated weights for policy 0, policy_version 8993 (0.0011)
+[2026-06-06 21:07:36,979][262198] Updated weights for policy 0, policy_version 9005 (0.0010)
+[2026-06-06 21:07:37,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4266.7, 300 sec: 4304.3). Total num frames: 4614144. Throughput: 0: 4248.2. Samples: 4616790. Policy #0 lag: (min: 10.0, avg: 25.9, max: 42.0)
+[2026-06-06 21:07:37,924][261293] Avg episode reward: [(0, '1274.008')]
+[2026-06-06 21:07:39,274][262198] Updated weights for policy 0, policy_version 9020 (0.0009)
+[2026-06-06 21:07:40,355][262198] Updated weights for policy 0, policy_version 9032 (0.0009)
+[2026-06-06 21:07:41,763][262198] Updated weights for policy 0, policy_version 9045 (0.0011)
+[2026-06-06 21:07:42,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4164.2, 300 sec: 4283.5). Total num frames: 4632576. Throughput: 0: 4267.2. Samples: 4643036. Policy #0 lag: (min: 3.0, avg: 21.3, max: 43.0)
+[2026-06-06 21:07:42,923][261293] Avg episode reward: [(0, '1270.506')]
+[2026-06-06 21:07:43,709][262198] Updated weights for policy 0, policy_version 9059 (0.0010)
+[2026-06-06 21:07:44,565][262198] Updated weights for policy 0, policy_version 9070 (0.0012)
+[2026-06-06 21:07:46,795][262198] Updated weights for policy 0, policy_version 9082 (0.0008)
+[2026-06-06 21:07:47,779][262115] Signal inference workers to stop experience collection... (1300 times)
+[2026-06-06 21:07:47,803][262198] InferenceWorker_p0-w0: stopping experience collection (1300 times)
+[2026-06-06 21:07:47,865][262115] Signal inference workers to resume experience collection... (1300 times)
+[2026-06-06 21:07:47,865][262198] InferenceWorker_p0-w0: resuming experience collection (1300 times)
+[2026-06-06 21:07:47,867][262198] Updated weights for policy 0, policy_version 9096 (0.0011)
+[2026-06-06 21:07:47,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4232.5, 300 sec: 4297.3). Total num frames: 4657152. Throughput: 0: 4245.3. Samples: 4655848. Policy #0 lag: (min: 3.0, avg: 22.2, max: 43.0)
+[2026-06-06 21:07:47,923][261293] Avg episode reward: [(0, '1190.561')]
+[2026-06-06 21:07:49,255][262198] Updated weights for policy 0, policy_version 9110 (0.0010)
+[2026-06-06 21:07:51,085][262198] Updated weights for policy 0, policy_version 9122 (0.0009)
+[2026-06-06 21:07:51,968][262198] Updated weights for policy 0, policy_version 9133 (0.0009)
+[2026-06-06 21:07:52,920][261293] Fps is (10 sec: 4710.6, 60 sec: 4334.9, 300 sec: 4297.3). Total num frames: 4679680. Throughput: 0: 4282.3. Samples: 4682244. Policy #0 lag: (min: 3.0, avg: 22.2, max: 43.0)
+[2026-06-06 21:07:52,922][261293] Avg episode reward: [(0, '1087.912')]
+[2026-06-06 21:07:54,222][262198] Updated weights for policy 0, policy_version 9145 (0.0010)
+[2026-06-06 21:07:55,454][262198] Updated weights for policy 0, policy_version 9160 (0.0010)
+[2026-06-06 21:07:56,893][262198] Updated weights for policy 0, policy_version 9173 (0.0010)
+[2026-06-06 21:07:57,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4266.7, 300 sec: 4283.4). Total num frames: 4698112. Throughput: 0: 4263.0. Samples: 4707900. Policy #0 lag: (min: 5.0, avg: 16.9, max: 37.0)
+[2026-06-06 21:07:57,923][261293] Avg episode reward: [(0, '1146.942')]
+[2026-06-06 21:07:58,920][262198] Updated weights for policy 0, policy_version 9189 (0.0011)
+[2026-06-06 21:07:59,754][262198] Updated weights for policy 0, policy_version 9200 (0.0009)
+[2026-06-06 21:08:02,209][262198] Updated weights for policy 0, policy_version 9213 (0.0010)
+[2026-06-06 21:08:02,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4266.7, 300 sec: 4290.4). Total num frames: 4720640. Throughput: 0: 4227.1. Samples: 4720972. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 21:08:02,923][261293] Avg episode reward: [(0, '1330.774')]
+[2026-06-06 21:08:03,263][262198] Updated weights for policy 0, policy_version 9225 (0.0008)
+[2026-06-06 21:08:04,448][262198] Updated weights for policy 0, policy_version 9237 (0.0009)
+[2026-06-06 21:08:06,133][262198] Updated weights for policy 0, policy_version 9250 (0.0009)
+[2026-06-06 21:08:06,951][262198] Updated weights for policy 0, policy_version 9261 (0.0009)
+[2026-06-06 21:08:07,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4334.9, 300 sec: 4283.4). Total num frames: 4745216. Throughput: 0: 4274.6. Samples: 4748164. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 21:08:07,922][261293] Avg episode reward: [(0, '1016.759')]
+[2026-06-06 21:08:09,048][262198] Updated weights for policy 0, policy_version 9273 (0.0010)
+[2026-06-06 21:08:10,232][262198] Updated weights for policy 0, policy_version 9288 (0.0009)
+[2026-06-06 21:08:11,563][262198] Updated weights for policy 0, policy_version 9300 (0.0009)
+[2026-06-06 21:08:12,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4266.6, 300 sec: 4269.5). Total num frames: 4763648. Throughput: 0: 4297.3. Samples: 4774794. Policy #0 lag: (min: 11.0, avg: 21.4, max: 40.0)
+[2026-06-06 21:08:12,923][261293] Avg episode reward: [(0, '991.429')]
+[2026-06-06 21:08:13,668][262198] Updated weights for policy 0, policy_version 9316 (0.0009)
+[2026-06-06 21:08:14,781][262198] Updated weights for policy 0, policy_version 9326 (0.0010)
+[2026-06-06 21:08:17,093][262198] Updated weights for policy 0, policy_version 9340 (0.0009)
+[2026-06-06 21:08:17,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4266.7, 300 sec: 4269.6). Total num frames: 4786176. Throughput: 0: 4294.6. Samples: 4786176. Policy #0 lag: (min: 11.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:08:17,922][261293] Avg episode reward: [(0, '1274.485')]
+[2026-06-06 21:08:18,514][262198] Updated weights for policy 0, policy_version 9354 (0.0009)
+[2026-06-06 21:08:19,839][262198] Updated weights for policy 0, policy_version 9366 (0.0008)
+[2026-06-06 21:08:21,933][262198] Updated weights for policy 0, policy_version 9382 (0.0010)
+[2026-06-06 21:08:22,902][262198] Updated weights for policy 0, policy_version 9392 (0.0010)
+[2026-06-06 21:08:22,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4300.8, 300 sec: 4269.6). Total num frames: 4808704. Throughput: 0: 4321.5. Samples: 4811258. Policy #0 lag: (min: 11.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:08:22,923][261293] Avg episode reward: [(0, '1157.842')]
+[2026-06-06 21:08:25,340][262115] Signal inference workers to stop experience collection... (1350 times)
+[2026-06-06 21:08:25,355][262198] InferenceWorker_p0-w0: stopping experience collection (1350 times)
+[2026-06-06 21:08:25,448][262115] Signal inference workers to resume experience collection... (1350 times)
+[2026-06-06 21:08:25,449][262198] InferenceWorker_p0-w0: resuming experience collection (1350 times)
+[2026-06-06 21:08:25,451][262198] Updated weights for policy 0, policy_version 9408 (0.0009)
+[2026-06-06 21:08:26,642][262198] Updated weights for policy 0, policy_version 9421 (0.0011)
+[2026-06-06 21:08:27,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4266.7, 300 sec: 4269.6). Total num frames: 4829184. Throughput: 0: 4289.9. Samples: 4836082. Policy #0 lag: (min: 11.0, avg: 21.9, max: 43.0)
+[2026-06-06 21:08:27,926][261293] Avg episode reward: [(0, '1281.573')]
+[2026-06-06 21:08:28,707][262198] Updated weights for policy 0, policy_version 9433 (0.0010)
+[2026-06-06 21:08:30,104][262198] Updated weights for policy 0, policy_version 9449 (0.0010)
+[2026-06-06 21:08:31,509][262198] Updated weights for policy 0, policy_version 9461 (0.0010)
+[2026-06-06 21:08:32,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4232.5, 300 sec: 4241.8). Total num frames: 4847616. Throughput: 0: 4286.2. Samples: 4848730. Policy #0 lag: (min: 11.0, avg: 21.9, max: 43.0)
+[2026-06-06 21:08:32,924][261293] Avg episode reward: [(0, '1220.046')]
+[2026-06-06 21:08:33,460][262198] Updated weights for policy 0, policy_version 9476 (0.0010)
+[2026-06-06 21:08:34,474][262198] Updated weights for policy 0, policy_version 9486 (0.0009)
+[2026-06-06 21:08:36,976][262198] Updated weights for policy 0, policy_version 9500 (0.0010)
+[2026-06-06 21:08:37,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4266.7, 300 sec: 4241.8). Total num frames: 4870144. Throughput: 0: 4241.3. Samples: 4873100. Policy #0 lag: (min: 11.0, avg: 23.3, max: 43.0)
+[2026-06-06 21:08:37,922][261293] Avg episode reward: [(0, '1365.504')]
+[2026-06-06 21:08:38,324][262198] Updated weights for policy 0, policy_version 9514 (0.0009)
+[2026-06-06 21:08:39,835][262198] Updated weights for policy 0, policy_version 9526 (0.0010)
+[2026-06-06 21:08:41,958][262198] Updated weights for policy 0, policy_version 9542 (0.0009)
+[2026-06-06 21:08:42,827][262198] Updated weights for policy 0, policy_version 9552 (0.0009)
+[2026-06-06 21:08:42,920][261293] Fps is (10 sec: 4301.1, 60 sec: 4300.8, 300 sec: 4248.7). Total num frames: 4890624. Throughput: 0: 4193.1. Samples: 4896590. Policy #0 lag: (min: 11.0, avg: 21.2, max: 43.0)
+[2026-06-06 21:08:42,922][261293] Avg episode reward: [(0, '1043.452')]
+[2026-06-06 21:08:45,486][262198] Updated weights for policy 0, policy_version 9568 (0.0011)
+[2026-06-06 21:08:46,681][262198] Updated weights for policy 0, policy_version 9580 (0.0009)
+[2026-06-06 21:08:47,920][261293] Fps is (10 sec: 3891.2, 60 sec: 4198.4, 300 sec: 4234.8). Total num frames: 4909056. Throughput: 0: 4175.7. Samples: 4908876. Policy #0 lag: (min: 11.0, avg: 21.2, max: 43.0)
+[2026-06-06 21:08:47,922][261293] Avg episode reward: [(0, '1237.655')]
+[2026-06-06 21:08:47,990][262198] Updated weights for policy 0, policy_version 9592 (0.0010)
+[2026-06-06 21:08:50,264][262198] Updated weights for policy 0, policy_version 9607 (0.0012)
+[2026-06-06 21:08:51,715][262198] Updated weights for policy 0, policy_version 9617 (0.0010)
+[2026-06-06 21:08:52,920][261293] Fps is (10 sec: 3686.4, 60 sec: 4130.1, 300 sec: 4221.0). Total num frames: 4927488. Throughput: 0: 4099.3. Samples: 4932630. Policy #0 lag: (min: 6.0, avg: 18.4, max: 38.0)
+[2026-06-06 21:08:52,922][261293] Avg episode reward: [(0, '1426.918')]
+[2026-06-06 21:08:52,928][262115] Saving new best policy, reward=1426.918!
+[2026-06-06 21:08:53,731][262198] Updated weights for policy 0, policy_version 9632 (0.0010)
+[2026-06-06 21:08:54,812][262198] Updated weights for policy 0, policy_version 9644 (0.0009)
+[2026-06-06 21:08:56,772][262198] Updated weights for policy 0, policy_version 9657 (0.0009)
+[2026-06-06 21:08:57,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4198.4, 300 sec: 4234.8). Total num frames: 4950016. Throughput: 0: 4089.1. Samples: 4958800. Policy #0 lag: (min: 10.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:08:57,923][261293] Avg episode reward: [(0, '1203.900')]
+[2026-06-06 21:08:57,979][262198] Updated weights for policy 0, policy_version 9672 (0.0009)
+[2026-06-06 21:08:59,650][262198] Updated weights for policy 0, policy_version 9685 (0.0011)
+[2026-06-06 21:09:01,502][262198] Updated weights for policy 0, policy_version 9700 (0.0010)
+[2026-06-06 21:09:02,362][262198] Updated weights for policy 0, policy_version 9710 (0.0008)
+[2026-06-06 21:09:02,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4198.4, 300 sec: 4248.7). Total num frames: 4972544. Throughput: 0: 4117.8. Samples: 4971476. Policy #0 lag: (min: 10.0, avg: 21.8, max: 43.0)
+[2026-06-06 21:09:02,924][261293] Avg episode reward: [(0, '1231.391')]
+[2026-06-06 21:09:04,157][262198] Updated weights for policy 0, policy_version 9721 (0.0010)
+[2026-06-06 21:09:04,372][262115] Signal inference workers to stop experience collection... (1400 times)
+[2026-06-06 21:09:04,399][262198] InferenceWorker_p0-w0: stopping experience collection (1400 times)
+[2026-06-06 21:09:04,424][262115] Signal inference workers to resume experience collection... (1400 times)
+[2026-06-06 21:09:04,425][262198] InferenceWorker_p0-w0: resuming experience collection (1400 times)
+[2026-06-06 21:09:05,342][262198] Updated weights for policy 0, policy_version 9735 (0.0008)
+[2026-06-06 21:09:06,679][262198] Updated weights for policy 0, policy_version 9748 (0.0008)
+[2026-06-06 21:09:07,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4130.1, 300 sec: 4234.9). Total num frames: 4993024. Throughput: 0: 4163.3. Samples: 4998604. Policy #0 lag: (min: 6.0, avg: 22.7, max: 38.0)
+[2026-06-06 21:09:07,922][261293] Avg episode reward: [(0, '1102.655')]
+[2026-06-06 21:09:08,319][262198] Updated weights for policy 0, policy_version 9758 (0.0009)
+[2026-06-06 21:09:09,249][262198] Updated weights for policy 0, policy_version 9771 (0.0010)
+[2026-06-06 21:09:10,549][262198] Updated weights for policy 0, policy_version 9783 (0.0008)
+[2026-06-06 21:09:12,301][262198] Updated weights for policy 0, policy_version 9798 (0.0010)
+[2026-06-06 21:09:12,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4266.7, 300 sec: 4255.7). Total num frames: 5019648. Throughput: 0: 4216.8. Samples: 5025836. Policy #0 lag: (min: 6.0, avg: 22.7, max: 38.0)
+[2026-06-06 21:09:12,922][261293] Avg episode reward: [(0, '1194.593')]
+[2026-06-06 21:09:13,346][262198] Updated weights for policy 0, policy_version 9808 (0.0010)
+[2026-06-06 21:09:15,886][262198] Updated weights for policy 0, policy_version 9821 (0.0011)
+[2026-06-06 21:09:17,215][262198] Updated weights for policy 0, policy_version 9835 (0.0010)
+[2026-06-06 21:09:17,926][261293] Fps is (10 sec: 4503.3, 60 sec: 4198.1, 300 sec: 4255.6). Total num frames: 5038080. Throughput: 0: 4211.3. Samples: 5038258. Policy #0 lag: (min: 1.0, avg: 18.7, max: 33.0)
+[2026-06-06 21:09:17,929][261293] Avg episode reward: [(0, '1381.841')]
+[2026-06-06 21:09:17,937][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000009840_5038080.pth...
+[2026-06-06 21:09:18,578][262198] Updated weights for policy 0, policy_version 9846 (0.0009)
+[2026-06-06 21:09:20,521][262198] Updated weights for policy 0, policy_version 9860 (0.0011)
+[2026-06-06 21:09:21,537][262198] Updated weights for policy 0, policy_version 9870 (0.0009)
+[2026-06-06 21:09:22,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4164.3, 300 sec: 4241.8). Total num frames: 5058560. Throughput: 0: 4186.5. Samples: 5061494. Policy #0 lag: (min: 3.0, avg: 20.5, max: 35.0)
+[2026-06-06 21:09:22,924][261293] Avg episode reward: [(0, '1170.035')]
+[2026-06-06 21:09:23,826][262198] Updated weights for policy 0, policy_version 9884 (0.0012)
+[2026-06-06 21:09:25,005][262198] Updated weights for policy 0, policy_version 9897 (0.0009)
+[2026-06-06 21:09:26,290][262198] Updated weights for policy 0, policy_version 9909 (0.0009)
+[2026-06-06 21:09:27,921][261293] Fps is (10 sec: 3893.0, 60 sec: 4130.1, 300 sec: 4227.9). Total num frames: 5076992. Throughput: 0: 4218.2. Samples: 5086412. Policy #0 lag: (min: 3.0, avg: 20.5, max: 35.0)
+[2026-06-06 21:09:27,923][261293] Avg episode reward: [(0, '1364.828')]
+[2026-06-06 21:09:28,614][262198] Updated weights for policy 0, policy_version 9925 (0.0011)
+[2026-06-06 21:09:29,627][262198] Updated weights for policy 0, policy_version 9935 (0.0008)
+[2026-06-06 21:09:31,660][262198] Updated weights for policy 0, policy_version 9947 (0.0011)
+[2026-06-06 21:09:32,827][262198] Updated weights for policy 0, policy_version 9960 (0.0010)
+[2026-06-06 21:09:32,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4198.4, 300 sec: 4234.8). Total num frames: 5099520. Throughput: 0: 4212.4. Samples: 5098438. Policy #0 lag: (min: 3.0, avg: 21.6, max: 41.0)
+[2026-06-06 21:09:32,922][261293] Avg episode reward: [(0, '1454.415')]
+[2026-06-06 21:09:32,927][262115] Saving new best policy, reward=1454.415!
+[2026-06-06 21:09:34,622][262198] Updated weights for policy 0, policy_version 9973 (0.0009)
+[2026-06-06 21:09:36,296][262198] Updated weights for policy 0, policy_version 9985 (0.0009)
+[2026-06-06 21:09:37,287][262198] Updated weights for policy 0, policy_version 9995 (0.0008)
+[2026-06-06 21:09:37,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4164.3, 300 sec: 4221.0). Total num frames: 5120000. Throughput: 0: 4215.8. Samples: 5122342. Policy #0 lag: (min: 3.0, avg: 21.6, max: 41.0)
+[2026-06-06 21:09:37,923][261293] Avg episode reward: [(0, '1109.319')]
+[2026-06-06 21:09:38,633][262198] Updated weights for policy 0, policy_version 10007 (0.0009)
+[2026-06-06 21:09:40,555][262198] Updated weights for policy 0, policy_version 10020 (0.0011)
+[2026-06-06 21:09:41,629][262198] Updated weights for policy 0, policy_version 10030 (0.0010)
+[2026-06-06 21:09:42,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4164.3, 300 sec: 4221.0). Total num frames: 5140480. Throughput: 0: 4185.5. Samples: 5147146. Policy #0 lag: (min: 5.0, avg: 20.9, max: 37.0)
+[2026-06-06 21:09:42,924][261293] Avg episode reward: [(0, '1331.339')]
+[2026-06-06 21:09:43,829][262198] Updated weights for policy 0, policy_version 10042 (0.0012)
+[2026-06-06 21:09:45,126][262198] Updated weights for policy 0, policy_version 10053 (0.0010)
+[2026-06-06 21:09:45,844][262115] Signal inference workers to stop experience collection... (1450 times)
+[2026-06-06 21:09:45,861][262198] InferenceWorker_p0-w0: stopping experience collection (1450 times)
+[2026-06-06 21:09:45,913][262115] Signal inference workers to resume experience collection... (1450 times)
+[2026-06-06 21:09:45,914][262198] InferenceWorker_p0-w0: resuming experience collection (1450 times)
+[2026-06-06 21:09:46,143][262198] Updated weights for policy 0, policy_version 10064 (0.0008)
+[2026-06-06 21:09:47,846][262198] Updated weights for policy 0, policy_version 10076 (0.0010)
+[2026-06-06 21:09:47,922][261293] Fps is (10 sec: 3890.9, 60 sec: 4164.2, 300 sec: 4214.0). Total num frames: 5158912. Throughput: 0: 4136.5. Samples: 5157622. Policy #0 lag: (min: 3.0, avg: 20.0, max: 35.0)
+[2026-06-06 21:09:47,924][261293] Avg episode reward: [(0, '1122.600')]
+[2026-06-06 21:09:48,896][262198] Updated weights for policy 0, policy_version 10088 (0.0010)
+[2026-06-06 21:09:50,474][262198] Updated weights for policy 0, policy_version 10100 (0.0010)
+[2026-06-06 21:09:51,926][262198] Updated weights for policy 0, policy_version 10110 (0.0009)
+[2026-06-06 21:09:52,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4232.5, 300 sec: 4207.1). Total num frames: 5181440. Throughput: 0: 4137.1. Samples: 5184776. Policy #0 lag: (min: 3.0, avg: 20.0, max: 35.0)
+[2026-06-06 21:09:52,923][261293] Avg episode reward: [(0, '1191.758')]
+[2026-06-06 21:09:53,019][262198] Updated weights for policy 0, policy_version 10122 (0.0010)
+[2026-06-06 21:09:54,266][262198] Updated weights for policy 0, policy_version 10133 (0.0009)
+[2026-06-06 21:09:56,014][262198] Updated weights for policy 0, policy_version 10145 (0.0011)
+[2026-06-06 21:09:57,294][262198] Updated weights for policy 0, policy_version 10157 (0.0010)
+[2026-06-06 21:09:57,921][261293] Fps is (10 sec: 4301.1, 60 sec: 4198.4, 300 sec: 4214.0). Total num frames: 5201920. Throughput: 0: 4083.7. Samples: 5209604. Policy #0 lag: (min: 3.0, avg: 18.4, max: 35.0)
+[2026-06-06 21:09:57,922][261293] Avg episode reward: [(0, '1351.388')]
+[2026-06-06 21:09:59,122][262198] Updated weights for policy 0, policy_version 10169 (0.0010)
+[2026-06-06 21:10:00,568][262198] Updated weights for policy 0, policy_version 10183 (0.0011)
+[2026-06-06 21:10:02,290][262198] Updated weights for policy 0, policy_version 10196 (0.0010)
+[2026-06-06 21:10:02,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4164.3, 300 sec: 4221.0). Total num frames: 5222400. Throughput: 0: 4091.8. Samples: 5222368. Policy #0 lag: (min: 3.0, avg: 18.4, max: 35.0)
+[2026-06-06 21:10:02,925][261293] Avg episode reward: [(0, '1251.523')]
+[2026-06-06 21:10:03,811][262198] Updated weights for policy 0, policy_version 10208 (0.0008)
+[2026-06-06 21:10:05,059][262198] Updated weights for policy 0, policy_version 10221 (0.0011)
+[2026-06-06 21:10:06,990][262198] Updated weights for policy 0, policy_version 10233 (0.0010)
+[2026-06-06 21:10:07,920][262198] Updated weights for policy 0, policy_version 10244 (0.0010)
+[2026-06-06 21:10:07,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4198.4, 300 sec: 4227.9). Total num frames: 5244928. Throughput: 0: 4136.0. Samples: 5247612. Policy #0 lag: (min: 5.0, avg: 19.3, max: 35.0)
+[2026-06-06 21:10:07,923][261293] Avg episode reward: [(0, '1306.678')]
+[2026-06-06 21:10:09,004][262198] Updated weights for policy 0, policy_version 10255 (0.0009)
+[2026-06-06 21:10:11,357][262198] Updated weights for policy 0, policy_version 10268 (0.0009)
+[2026-06-06 21:10:12,323][262198] Updated weights for policy 0, policy_version 10280 (0.0011)
+[2026-06-06 21:10:12,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4096.0, 300 sec: 4221.0). Total num frames: 5265408. Throughput: 0: 4123.8. Samples: 5271982. Policy #0 lag: (min: 10.0, avg: 22.9, max: 42.0)
+[2026-06-06 21:10:12,924][261293] Avg episode reward: [(0, '1288.162')]
+[2026-06-06 21:10:14,022][262198] Updated weights for policy 0, policy_version 10292 (0.0008)
+[2026-06-06 21:10:15,303][262198] Updated weights for policy 0, policy_version 10302 (0.0010)
+[2026-06-06 21:10:16,169][262198] Updated weights for policy 0, policy_version 10312 (0.0009)
+[2026-06-06 21:10:17,890][262198] Updated weights for policy 0, policy_version 10325 (0.0010)
+[2026-06-06 21:10:17,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4130.5, 300 sec: 4214.0). Total num frames: 5285888. Throughput: 0: 4139.0. Samples: 5284692. Policy #0 lag: (min: 10.0, avg: 22.9, max: 42.0)
+[2026-06-06 21:10:17,923][261293] Avg episode reward: [(0, '1396.334')]
+[2026-06-06 21:10:19,700][262198] Updated weights for policy 0, policy_version 10338 (0.0009)
+[2026-06-06 21:10:20,754][262198] Updated weights for policy 0, policy_version 10348 (0.0009)
+[2026-06-06 21:10:22,024][262198] Updated weights for policy 0, policy_version 10358 (0.0010)
+[2026-06-06 21:10:22,920][261293] Fps is (10 sec: 3891.2, 60 sec: 4096.0, 300 sec: 4214.0). Total num frames: 5304320. Throughput: 0: 4155.6. Samples: 5309344. Policy #0 lag: (min: 2.0, avg: 21.6, max: 38.0)
+[2026-06-06 21:10:22,922][261293] Avg episode reward: [(0, '1103.611')]
+[2026-06-06 21:10:23,616][262198] Updated weights for policy 0, policy_version 10369 (0.0011)
+[2026-06-06 21:10:24,679][262198] Updated weights for policy 0, policy_version 10379 (0.0009)
+[2026-06-06 21:10:25,766][262198] Updated weights for policy 0, policy_version 10389 (0.0009)
+[2026-06-06 21:10:27,443][262198] Updated weights for policy 0, policy_version 10401 (0.0010)
+[2026-06-06 21:10:27,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4164.3, 300 sec: 4214.0). Total num frames: 5326848. Throughput: 0: 4176.5. Samples: 5335088. Policy #0 lag: (min: 2.0, avg: 21.6, max: 38.0)
+[2026-06-06 21:10:27,928][261293] Avg episode reward: [(0, '1345.519')]
+[2026-06-06 21:10:28,626][262198] Updated weights for policy 0, policy_version 10413 (0.0008)
+[2026-06-06 21:10:29,811][262198] Updated weights for policy 0, policy_version 10424 (0.0008)
+[2026-06-06 21:10:30,937][262115] Signal inference workers to stop experience collection... (1500 times)
+[2026-06-06 21:10:30,969][262198] InferenceWorker_p0-w0: stopping experience collection (1500 times)
+[2026-06-06 21:10:31,071][262115] Signal inference workers to resume experience collection... (1500 times)
+[2026-06-06 21:10:31,071][262198] InferenceWorker_p0-w0: resuming experience collection (1500 times)
+[2026-06-06 21:10:31,278][262198] Updated weights for policy 0, policy_version 10436 (0.0009)
+[2026-06-06 21:10:32,700][262198] Updated weights for policy 0, policy_version 10448 (0.0009)
+[2026-06-06 21:10:32,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4164.3, 300 sec: 4207.1). Total num frames: 5349376. Throughput: 0: 4234.3. Samples: 5348160. Policy #0 lag: (min: 3.0, avg: 20.6, max: 35.0)
+[2026-06-06 21:10:32,922][261293] Avg episode reward: [(0, '1340.701')]
+[2026-06-06 21:10:34,607][262198] Updated weights for policy 0, policy_version 10458 (0.0009)
+[2026-06-06 21:10:35,699][262198] Updated weights for policy 0, policy_version 10471 (0.0009)
+[2026-06-06 21:10:37,065][262198] Updated weights for policy 0, policy_version 10482 (0.0010)
+[2026-06-06 21:10:37,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4164.3, 300 sec: 4214.0). Total num frames: 5369856. Throughput: 0: 4192.9. Samples: 5373456. Policy #0 lag: (min: 3.0, avg: 20.6, max: 35.0)
+[2026-06-06 21:10:37,923][261293] Avg episode reward: [(0, '1300.739')]
+[2026-06-06 21:10:38,511][262198] Updated weights for policy 0, policy_version 10492 (0.0009)
+[2026-06-06 21:10:39,570][262198] Updated weights for policy 0, policy_version 10504 (0.0009)
+[2026-06-06 21:10:40,693][262198] Updated weights for policy 0, policy_version 10514 (0.0009)
+[2026-06-06 21:10:42,450][262198] Updated weights for policy 0, policy_version 10526 (0.0009)
+[2026-06-06 21:10:42,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4198.4, 300 sec: 4207.1). Total num frames: 5392384. Throughput: 0: 4232.2. Samples: 5400050. Policy #0 lag: (min: 1.0, avg: 16.7, max: 33.0)
+[2026-06-06 21:10:42,922][261293] Avg episode reward: [(0, '1214.267')]
+[2026-06-06 21:10:43,557][262198] Updated weights for policy 0, policy_version 10537 (0.0010)
+[2026-06-06 21:10:44,411][262198] Updated weights for policy 0, policy_version 10547 (0.0008)
+[2026-06-06 21:10:46,210][262198] Updated weights for policy 0, policy_version 10557 (0.0010)
+[2026-06-06 21:10:47,210][262198] Updated weights for policy 0, policy_version 10567 (0.0011)
+[2026-06-06 21:10:47,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4198.4, 300 sec: 4193.2). Total num frames: 5410816. Throughput: 0: 4228.1. Samples: 5412632. Policy #0 lag: (min: 6.0, avg: 21.7, max: 42.0)
+[2026-06-06 21:10:47,923][261293] Avg episode reward: [(0, '1385.550')]
+[2026-06-06 21:10:48,640][262198] Updated weights for policy 0, policy_version 10577 (0.0013)
+[2026-06-06 21:10:50,369][262198] Updated weights for policy 0, policy_version 10587 (0.0013)
+[2026-06-06 21:10:51,500][262198] Updated weights for policy 0, policy_version 10597 (0.0010)
+[2026-06-06 21:10:52,383][262198] Updated weights for policy 0, policy_version 10607 (0.0009)
+[2026-06-06 21:10:52,924][261293] Fps is (10 sec: 3889.8, 60 sec: 4164.0, 300 sec: 4207.0). Total num frames: 5431296. Throughput: 0: 4168.0. Samples: 5435188. Policy #0 lag: (min: 6.0, avg: 21.7, max: 42.0)
+[2026-06-06 21:10:52,927][261293] Avg episode reward: [(0, '1163.179')]
+[2026-06-06 21:10:54,162][262198] Updated weights for policy 0, policy_version 10617 (0.0010)
+[2026-06-06 21:10:55,392][262198] Updated weights for policy 0, policy_version 10629 (0.0009)
+[2026-06-06 21:10:56,141][262198] Updated weights for policy 0, policy_version 10639 (0.0009)
+[2026-06-06 21:10:57,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4164.2, 300 sec: 4200.1). Total num frames: 5451776. Throughput: 0: 4211.7. Samples: 5461510. Policy #0 lag: (min: 6.0, avg: 19.4, max: 38.0)
+[2026-06-06 21:10:57,924][261293] Avg episode reward: [(0, '1345.952')]
+[2026-06-06 21:10:57,990][262198] Updated weights for policy 0, policy_version 10652 (0.0009)
+[2026-06-06 21:10:59,245][262198] Updated weights for policy 0, policy_version 10663 (0.0010)
+[2026-06-06 21:11:00,312][262198] Updated weights for policy 0, policy_version 10673 (0.0009)
+[2026-06-06 21:11:01,770][262198] Updated weights for policy 0, policy_version 10685 (0.0008)
+[2026-06-06 21:11:02,860][262198] Updated weights for policy 0, policy_version 10697 (0.0008)
+[2026-06-06 21:11:02,921][261293] Fps is (10 sec: 4507.0, 60 sec: 4232.5, 300 sec: 4207.1). Total num frames: 5476352. Throughput: 0: 4230.5. Samples: 5475064. Policy #0 lag: (min: 6.0, avg: 19.4, max: 38.0)
+[2026-06-06 21:11:02,924][261293] Avg episode reward: [(0, '1246.120')]
+[2026-06-06 21:11:03,891][262198] Updated weights for policy 0, policy_version 10708 (0.0009)
+[2026-06-06 21:11:05,288][262198] Updated weights for policy 0, policy_version 10720 (0.0008)
+[2026-06-06 21:11:06,585][262198] Updated weights for policy 0, policy_version 10732 (0.0010)
+[2026-06-06 21:11:07,449][262198] Updated weights for policy 0, policy_version 10742 (0.0008)
+[2026-06-06 21:11:07,921][261293] Fps is (10 sec: 4915.4, 60 sec: 4266.7, 300 sec: 4221.0). Total num frames: 5500928. Throughput: 0: 4321.7. Samples: 5503822. Policy #0 lag: (min: 6.0, avg: 21.9, max: 38.0)
+[2026-06-06 21:11:07,924][261293] Avg episode reward: [(0, '1301.488')]
+[2026-06-06 21:11:09,045][262198] Updated weights for policy 0, policy_version 10754 (0.0010)
+[2026-06-06 21:11:10,391][262198] Updated weights for policy 0, policy_version 10765 (0.0010)
+[2026-06-06 21:11:11,289][262198] Updated weights for policy 0, policy_version 10775 (0.0009)
+[2026-06-06 21:11:12,828][262198] Updated weights for policy 0, policy_version 10787 (0.0008)
+[2026-06-06 21:11:12,920][261293] Fps is (10 sec: 4710.7, 60 sec: 4300.8, 300 sec: 4241.8). Total num frames: 5523456. Throughput: 0: 4331.4. Samples: 5530000. Policy #0 lag: (min: 3.0, avg: 19.7, max: 35.0)
+[2026-06-06 21:11:12,922][261293] Avg episode reward: [(0, '1388.774')]
+[2026-06-06 21:11:14,355][262198] Updated weights for policy 0, policy_version 10800 (0.0010)
+[2026-06-06 21:11:14,539][262115] Signal inference workers to stop experience collection... (1550 times)
+[2026-06-06 21:11:14,540][262115] Signal inference workers to resume experience collection... (1550 times)
+[2026-06-06 21:11:14,554][262198] InferenceWorker_p0-w0: stopping experience collection (1550 times)
+[2026-06-06 21:11:14,555][262198] InferenceWorker_p0-w0: resuming experience collection (1550 times)
+[2026-06-06 21:11:15,796][262198] Updated weights for policy 0, policy_version 10810 (0.0009)
+[2026-06-06 21:11:17,393][262198] Updated weights for policy 0, policy_version 10824 (0.0009)
+[2026-06-06 21:11:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4300.8, 300 sec: 4234.9). Total num frames: 5543936. Throughput: 0: 4321.0. Samples: 5542604. Policy #0 lag: (min: 3.0, avg: 19.7, max: 35.0)
+[2026-06-06 21:11:17,922][261293] Avg episode reward: [(0, '1415.984')]
+[2026-06-06 21:11:18,530][262198] Updated weights for policy 0, policy_version 10834 (0.0010)
+[2026-06-06 21:11:19,713][262198] Updated weights for policy 0, policy_version 10844 (0.0010)
+[2026-06-06 21:11:20,994][262198] Updated weights for policy 0, policy_version 10856 (0.0009)
+[2026-06-06 21:11:21,951][262198] Updated weights for policy 0, policy_version 10866 (0.0010)
+[2026-06-06 21:11:22,920][261293] Fps is (10 sec: 4300.7, 60 sec: 4369.1, 300 sec: 4234.9). Total num frames: 5566464. Throughput: 0: 4360.3. Samples: 5569670. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:11:22,922][261293] Avg episode reward: [(0, '1255.723')]
+[2026-06-06 21:11:23,153][262198] Updated weights for policy 0, policy_version 10876 (0.0009)
+[2026-06-06 21:11:24,664][262198] Updated weights for policy 0, policy_version 10886 (0.0009)
+[2026-06-06 21:11:25,791][262198] Updated weights for policy 0, policy_version 10897 (0.0011)
+[2026-06-06 21:11:27,053][262198] Updated weights for policy 0, policy_version 10907 (0.0009)
+[2026-06-06 21:11:27,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4369.1, 300 sec: 4241.8). Total num frames: 5588992. Throughput: 0: 4340.4. Samples: 5595370. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:11:27,923][261293] Avg episode reward: [(0, '1137.250')]
+[2026-06-06 21:11:28,817][262198] Updated weights for policy 0, policy_version 10922 (0.0010)
+[2026-06-06 21:11:29,727][262198] Updated weights for policy 0, policy_version 10932 (0.0009)
+[2026-06-06 21:11:31,458][262198] Updated weights for policy 0, policy_version 10945 (0.0010)
+[2026-06-06 21:11:32,728][262198] Updated weights for policy 0, policy_version 10956 (0.0009)
+[2026-06-06 21:11:32,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4334.9, 300 sec: 4241.8). Total num frames: 5609472. Throughput: 0: 4355.5. Samples: 5608632. Policy #0 lag: (min: 5.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:11:32,924][261293] Avg episode reward: [(0, '1231.053')]
+[2026-06-06 21:11:33,658][262198] Updated weights for policy 0, policy_version 10966 (0.0008)
+[2026-06-06 21:11:35,038][262198] Updated weights for policy 0, policy_version 10978 (0.0008)
+[2026-06-06 21:11:36,363][262198] Updated weights for policy 0, policy_version 10990 (0.0010)
+[2026-06-06 21:11:37,343][262198] Updated weights for policy 0, policy_version 11000 (0.0008)
+[2026-06-06 21:11:37,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4403.2, 300 sec: 4241.8). Total num frames: 5634048. Throughput: 0: 4465.1. Samples: 5636104. Policy #0 lag: (min: 5.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:11:37,922][261293] Avg episode reward: [(0, '1295.458')]
+[2026-06-06 21:11:39,206][262198] Updated weights for policy 0, policy_version 11013 (0.0009)
+[2026-06-06 21:11:40,022][262198] Updated weights for policy 0, policy_version 11023 (0.0009)
+[2026-06-06 21:11:41,487][262198] Updated weights for policy 0, policy_version 11033 (0.0010)
+[2026-06-06 21:11:42,413][262198] Updated weights for policy 0, policy_version 11044 (0.0009)
+[2026-06-06 21:11:42,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4369.0, 300 sec: 4241.8). Total num frames: 5654528. Throughput: 0: 4481.2. Samples: 5663164. Policy #0 lag: (min: 6.0, avg: 22.3, max: 38.0)
+[2026-06-06 21:11:42,924][261293] Avg episode reward: [(0, '1351.553')]
+[2026-06-06 21:11:43,794][262198] Updated weights for policy 0, policy_version 11055 (0.0009)
+[2026-06-06 21:11:45,182][262198] Updated weights for policy 0, policy_version 11065 (0.0009)
+[2026-06-06 21:11:46,683][262198] Updated weights for policy 0, policy_version 11078 (0.0011)
+[2026-06-06 21:11:47,462][262198] Updated weights for policy 0, policy_version 11088 (0.0009)
+[2026-06-06 21:11:47,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4437.3, 300 sec: 4262.6). Total num frames: 5677056. Throughput: 0: 4471.4. Samples: 5676274. Policy #0 lag: (min: 3.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:11:47,922][261293] Avg episode reward: [(0, '1317.497')]
+[2026-06-06 21:11:48,918][262198] Updated weights for policy 0, policy_version 11098 (0.0008)
+[2026-06-06 21:11:50,291][262198] Updated weights for policy 0, policy_version 11109 (0.0009)
+[2026-06-06 21:11:51,116][262198] Updated weights for policy 0, policy_version 11120 (0.0010)
+[2026-06-06 21:11:52,567][262198] Updated weights for policy 0, policy_version 11130 (0.0008)
+[2026-06-06 21:11:52,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.7, 300 sec: 4262.6). Total num frames: 5699584. Throughput: 0: 4423.6. Samples: 5702884. Policy #0 lag: (min: 3.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:11:52,924][261293] Avg episode reward: [(0, '1349.056')]
+[2026-06-06 21:11:53,864][262198] Updated weights for policy 0, policy_version 11141 (0.0009)
+[2026-06-06 21:11:54,767][262198] Updated weights for policy 0, policy_version 11151 (0.0009)
+[2026-06-06 21:11:56,247][262198] Updated weights for policy 0, policy_version 11161 (0.0008)
+[2026-06-06 21:11:57,550][262198] Updated weights for policy 0, policy_version 11174 (0.0009)
+[2026-06-06 21:11:57,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4505.6, 300 sec: 4262.6). Total num frames: 5722112. Throughput: 0: 4447.5. Samples: 5730142. Policy #0 lag: (min: 6.0, avg: 21.5, max: 38.0)
+[2026-06-06 21:11:57,923][261293] Avg episode reward: [(0, '1292.787')]
+[2026-06-06 21:11:58,627][262198] Updated weights for policy 0, policy_version 11185 (0.0009)
+[2026-06-06 21:11:59,815][262198] Updated weights for policy 0, policy_version 11195 (0.0009)
+[2026-06-06 21:12:01,209][262198] Updated weights for policy 0, policy_version 11205 (0.0009)
+[2026-06-06 21:12:02,123][262198] Updated weights for policy 0, policy_version 11215 (0.0010)
+[2026-06-06 21:12:02,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4471.5, 300 sec: 4269.6). Total num frames: 5744640. Throughput: 0: 4468.2. Samples: 5743674. Policy #0 lag: (min: 6.0, avg: 21.5, max: 38.0)
+[2026-06-06 21:12:02,923][261293] Avg episode reward: [(0, '1226.905')]
+[2026-06-06 21:12:03,528][262198] Updated weights for policy 0, policy_version 11225 (0.0009)
+[2026-06-06 21:12:04,228][262198] Updated weights for policy 0, policy_version 11235 (0.0010)
+[2026-06-06 21:12:05,756][262198] Updated weights for policy 0, policy_version 11246 (0.0009)
+[2026-06-06 21:12:06,725][262198] Updated weights for policy 0, policy_version 11256 (0.0008)
+[2026-06-06 21:12:07,921][261293] Fps is (10 sec: 4505.9, 60 sec: 4437.3, 300 sec: 4269.6). Total num frames: 5767168. Throughput: 0: 4457.2. Samples: 5770242. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:12:07,924][261293] Avg episode reward: [(0, '1533.254')]
+[2026-06-06 21:12:08,013][262115] Saving new best policy, reward=1533.254!
+[2026-06-06 21:12:08,015][262198] Updated weights for policy 0, policy_version 11268 (0.0009)
+[2026-06-06 21:12:09,629][262198] Updated weights for policy 0, policy_version 11279 (0.0009)
+[2026-06-06 21:12:10,695][262198] Updated weights for policy 0, policy_version 11289 (0.0008)
+[2026-06-06 21:12:12,262][262198] Updated weights for policy 0, policy_version 11301 (0.0008)
+[2026-06-06 21:12:12,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4437.3, 300 sec: 4269.6). Total num frames: 5789696. Throughput: 0: 4493.9. Samples: 5797594. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:12:12,923][261293] Avg episode reward: [(0, '1013.084')]
+[2026-06-06 21:12:13,297][262198] Updated weights for policy 0, policy_version 11313 (0.0010)
+[2026-06-06 21:12:14,118][262115] Signal inference workers to stop experience collection... (1600 times)
+[2026-06-06 21:12:14,126][262115] Signal inference workers to resume experience collection... (1600 times)
+[2026-06-06 21:12:14,143][262198] InferenceWorker_p0-w0: stopping experience collection (1600 times)
+[2026-06-06 21:12:14,144][262198] InferenceWorker_p0-w0: resuming experience collection (1600 times)
+[2026-06-06 21:12:14,337][262198] Updated weights for policy 0, policy_version 11323 (0.0009)
+[2026-06-06 21:12:15,772][262198] Updated weights for policy 0, policy_version 11334 (0.0009)
+[2026-06-06 21:12:16,902][262198] Updated weights for policy 0, policy_version 11345 (0.0010)
+[2026-06-06 21:12:17,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4471.4, 300 sec: 4276.5). Total num frames: 5812224. Throughput: 0: 4497.6. Samples: 5811026. Policy #0 lag: (min: 6.0, avg: 22.2, max: 38.0)
+[2026-06-06 21:12:17,923][261293] Avg episode reward: [(0, '1441.261')]
+[2026-06-06 21:12:17,962][262198] Updated weights for policy 0, policy_version 11355 (0.0009)
+[2026-06-06 21:12:19,329][262198] Updated weights for policy 0, policy_version 11365 (0.0009)
+[2026-06-06 21:12:20,418][262198] Updated weights for policy 0, policy_version 11377 (0.0009)
+[2026-06-06 21:12:21,570][262198] Updated weights for policy 0, policy_version 11388 (0.0010)
+[2026-06-06 21:12:22,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4471.4, 300 sec: 4276.5). Total num frames: 5834752. Throughput: 0: 4505.0. Samples: 5838830. Policy #0 lag: (min: 6.0, avg: 22.2, max: 38.0)
+[2026-06-06 21:12:22,922][261293] Avg episode reward: [(0, '1406.244')]
+[2026-06-06 21:12:23,130][262198] Updated weights for policy 0, policy_version 11400 (0.0010)
+[2026-06-06 21:12:24,162][262198] Updated weights for policy 0, policy_version 11410 (0.0009)
+[2026-06-06 21:12:25,232][262198] Updated weights for policy 0, policy_version 11420 (0.0009)
+[2026-06-06 21:12:26,674][262198] Updated weights for policy 0, policy_version 11430 (0.0008)
+[2026-06-06 21:12:27,605][262198] Updated weights for policy 0, policy_version 11440 (0.0010)
+[2026-06-06 21:12:27,921][261293] Fps is (10 sec: 4710.5, 60 sec: 4505.6, 300 sec: 4290.4). Total num frames: 5859328. Throughput: 0: 4494.2. Samples: 5865402. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 21:12:27,923][261293] Avg episode reward: [(0, '1299.555')]
+[2026-06-06 21:12:28,645][262198] Updated weights for policy 0, policy_version 11450 (0.0009)
+[2026-06-06 21:12:29,609][262198] Updated weights for policy 0, policy_version 11460 (0.0009)
+[2026-06-06 21:12:31,215][262198] Updated weights for policy 0, policy_version 11473 (0.0011)
+[2026-06-06 21:12:32,238][262198] Updated weights for policy 0, policy_version 11483 (0.0009)
+[2026-06-06 21:12:32,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4539.7, 300 sec: 4297.3). Total num frames: 5881856. Throughput: 0: 4522.4. Samples: 5879782. Policy #0 lag: (min: 5.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:12:32,924][261293] Avg episode reward: [(0, '1274.524')]
+[2026-06-06 21:12:33,688][262198] Updated weights for policy 0, policy_version 11493 (0.0009)
+[2026-06-06 21:12:34,790][262198] Updated weights for policy 0, policy_version 11506 (0.0010)
+[2026-06-06 21:12:35,743][262198] Updated weights for policy 0, policy_version 11516 (0.0009)
+[2026-06-06 21:12:37,418][262198] Updated weights for policy 0, policy_version 11528 (0.0009)
+[2026-06-06 21:12:37,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4505.6, 300 sec: 4311.2). Total num frames: 5904384. Throughput: 0: 4541.7. Samples: 5907258. Policy #0 lag: (min: 5.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:12:37,922][261293] Avg episode reward: [(0, '1306.678')]
+[2026-06-06 21:12:38,628][262198] Updated weights for policy 0, policy_version 11538 (0.0009)
+[2026-06-06 21:12:39,627][262198] Updated weights for policy 0, policy_version 11548 (0.0008)
+[2026-06-06 21:12:41,111][262198] Updated weights for policy 0, policy_version 11560 (0.0009)
+[2026-06-06 21:12:42,326][262198] Updated weights for policy 0, policy_version 11570 (0.0013)
+[2026-06-06 21:12:42,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4505.6, 300 sec: 4297.3). Total num frames: 5924864. Throughput: 0: 4506.6. Samples: 5932936. Policy #0 lag: (min: 5.0, avg: 20.3, max: 37.0)
+[2026-06-06 21:12:42,923][261293] Avg episode reward: [(0, '1173.901')]
+[2026-06-06 21:12:43,590][262198] Updated weights for policy 0, policy_version 11580 (0.0011)
+[2026-06-06 21:12:45,663][262198] Updated weights for policy 0, policy_version 11590 (0.0014)
+[2026-06-06 21:12:46,993][262198] Updated weights for policy 0, policy_version 11601 (0.0010)
+[2026-06-06 21:12:47,920][261293] Fps is (10 sec: 3891.2, 60 sec: 4437.3, 300 sec: 4283.4). Total num frames: 5943296. Throughput: 0: 4435.9. Samples: 5943290. Policy #0 lag: (min: 5.0, avg: 20.3, max: 37.0)
+[2026-06-06 21:12:47,922][261293] Avg episode reward: [(0, '1428.513')]
+[2026-06-06 21:12:47,968][262198] Updated weights for policy 0, policy_version 11611 (0.0010)
+[2026-06-06 21:12:49,682][262198] Updated weights for policy 0, policy_version 11621 (0.0010)
+[2026-06-06 21:12:50,690][262198] Updated weights for policy 0, policy_version 11632 (0.0010)
+[2026-06-06 21:12:51,768][262198] Updated weights for policy 0, policy_version 11642 (0.0009)
+[2026-06-06 21:12:52,741][262198] Updated weights for policy 0, policy_version 11652 (0.0009)
+[2026-06-06 21:12:52,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4437.3, 300 sec: 4297.3). Total num frames: 5965824. Throughput: 0: 4398.7. Samples: 5968184. Policy #0 lag: (min: 3.0, avg: 20.6, max: 35.0)
+[2026-06-06 21:12:52,924][261293] Avg episode reward: [(0, '1212.148')]
+[2026-06-06 21:12:54,502][262198] Updated weights for policy 0, policy_version 11666 (0.0009)
+[2026-06-06 21:12:55,559][262198] Updated weights for policy 0, policy_version 11676 (0.0009)
+[2026-06-06 21:12:56,869][262198] Updated weights for policy 0, policy_version 11686 (0.0008)
+[2026-06-06 21:12:57,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4437.4, 300 sec: 4297.3). Total num frames: 5988352. Throughput: 0: 4406.6. Samples: 5995892. Policy #0 lag: (min: 3.0, avg: 20.6, max: 35.0)
+[2026-06-06 21:12:57,923][261293] Avg episode reward: [(0, '1167.099')]
+[2026-06-06 21:12:57,938][262198] Updated weights for policy 0, policy_version 11698 (0.0010)
+[2026-06-06 21:12:58,867][262198] Updated weights for policy 0, policy_version 11708 (0.0007)
+[2026-06-06 21:13:00,432][262198] Updated weights for policy 0, policy_version 11720 (0.0010)
+[2026-06-06 21:13:01,309][262198] Updated weights for policy 0, policy_version 11730 (0.0008)
+[2026-06-06 21:13:02,768][262115] Signal inference workers to stop experience collection... (1650 times)
+[2026-06-06 21:13:02,782][262115] Signal inference workers to resume experience collection... (1650 times)
+[2026-06-06 21:13:02,796][262198] InferenceWorker_p0-w0: stopping experience collection (1650 times)
+[2026-06-06 21:13:02,798][262198] Updated weights for policy 0, policy_version 11740 (0.0011)
+[2026-06-06 21:13:02,815][262198] InferenceWorker_p0-w0: resuming experience collection (1650 times)
+[2026-06-06 21:13:02,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4437.3, 300 sec: 4290.4). Total num frames: 6010880. Throughput: 0: 4436.8. Samples: 6010684. Policy #0 lag: (min: 3.0, avg: 20.3, max: 39.0)
+[2026-06-06 21:13:02,925][261293] Avg episode reward: [(0, '1279.066')]
+[2026-06-06 21:13:04,189][262198] Updated weights for policy 0, policy_version 11753 (0.0009)
+[2026-06-06 21:13:05,613][262198] Updated weights for policy 0, policy_version 11765 (0.0010)
+[2026-06-06 21:13:06,574][262198] Updated weights for policy 0, policy_version 11775 (0.0008)
+[2026-06-06 21:13:07,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4437.3, 300 sec: 4304.3). Total num frames: 6033408. Throughput: 0: 4404.8. Samples: 6037044. Policy #0 lag: (min: 3.0, avg: 20.3, max: 39.0)
+[2026-06-06 21:13:07,923][261293] Avg episode reward: [(0, '1386.090')]
+[2026-06-06 21:13:07,971][262198] Updated weights for policy 0, policy_version 11788 (0.0009)
+[2026-06-06 21:13:09,442][262198] Updated weights for policy 0, policy_version 11798 (0.0011)
+[2026-06-06 21:13:10,670][262198] Updated weights for policy 0, policy_version 11808 (0.0009)
+[2026-06-06 21:13:11,863][262198] Updated weights for policy 0, policy_version 11819 (0.0009)
+[2026-06-06 21:13:12,920][261293] Fps is (10 sec: 4505.8, 60 sec: 4437.4, 300 sec: 4304.3). Total num frames: 6055936. Throughput: 0: 4379.9. Samples: 6062496. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:13:12,922][261293] Avg episode reward: [(0, '1445.583')]
+[2026-06-06 21:13:13,219][262198] Updated weights for policy 0, policy_version 11829 (0.0010)
+[2026-06-06 21:13:14,529][262198] Updated weights for policy 0, policy_version 11839 (0.0009)
+[2026-06-06 21:13:15,906][262198] Updated weights for policy 0, policy_version 11851 (0.0010)
+[2026-06-06 21:13:17,083][262198] Updated weights for policy 0, policy_version 11861 (0.0009)
+[2026-06-06 21:13:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4403.2, 300 sec: 4297.3). Total num frames: 6076416. Throughput: 0: 4332.9. Samples: 6074758. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:13:17,922][261293] Avg episode reward: [(0, '1468.792')]
+[2026-06-06 21:13:18,437][262198] Updated weights for policy 0, policy_version 11871 (0.0009)
+[2026-06-06 21:13:19,623][262198] Updated weights for policy 0, policy_version 11882 (0.0010)
+[2026-06-06 21:13:20,389][262198] Updated weights for policy 0, policy_version 11892 (0.0009)
+[2026-06-06 21:13:21,983][262198] Updated weights for policy 0, policy_version 11902 (0.0011)
+[2026-06-06 21:13:22,920][261293] Fps is (10 sec: 4095.9, 60 sec: 4369.1, 300 sec: 4297.3). Total num frames: 6096896. Throughput: 0: 4306.4. Samples: 6101048. Policy #0 lag: (min: 3.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:13:22,922][261293] Avg episode reward: [(0, '1283.657')]
+[2026-06-06 21:13:23,376][262198] Updated weights for policy 0, policy_version 11914 (0.0009)
+[2026-06-06 21:13:24,761][262198] Updated weights for policy 0, policy_version 11926 (0.0010)
+[2026-06-06 21:13:25,907][262198] Updated weights for policy 0, policy_version 11936 (0.0010)
+[2026-06-06 21:13:27,378][262198] Updated weights for policy 0, policy_version 11946 (0.0009)
+[2026-06-06 21:13:27,922][261293] Fps is (10 sec: 4095.5, 60 sec: 4300.7, 300 sec: 4304.3). Total num frames: 6117376. Throughput: 0: 4294.4. Samples: 6126188. Policy #0 lag: (min: 3.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:13:27,926][261293] Avg episode reward: [(0, '1326.931')]
+[2026-06-06 21:13:28,747][262198] Updated weights for policy 0, policy_version 11958 (0.0010)
+[2026-06-06 21:13:29,938][262198] Updated weights for policy 0, policy_version 11968 (0.0009)
+[2026-06-06 21:13:31,200][262198] Updated weights for policy 0, policy_version 11979 (0.0011)
+[2026-06-06 21:13:32,514][262198] Updated weights for policy 0, policy_version 11991 (0.0008)
+[2026-06-06 21:13:32,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4300.9, 300 sec: 4304.3). Total num frames: 6139904. Throughput: 0: 4351.2. Samples: 6139094. Policy #0 lag: (min: 7.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:13:32,923][261293] Avg episode reward: [(0, '1345.271')]
+[2026-06-06 21:13:33,988][262198] Updated weights for policy 0, policy_version 12001 (0.0008)
+[2026-06-06 21:13:35,241][262198] Updated weights for policy 0, policy_version 12016 (0.0009)
+[2026-06-06 21:13:36,359][262198] Updated weights for policy 0, policy_version 12026 (0.0008)
+[2026-06-06 21:13:37,882][262198] Updated weights for policy 0, policy_version 12038 (0.0009)
+[2026-06-06 21:13:37,921][261293] Fps is (10 sec: 4506.0, 60 sec: 4300.8, 300 sec: 4311.2). Total num frames: 6162432. Throughput: 0: 4417.1. Samples: 6166954. Policy #0 lag: (min: 3.0, avg: 19.8, max: 39.0)
+[2026-06-06 21:13:37,925][261293] Avg episode reward: [(0, '1373.099')]
+[2026-06-06 21:13:39,139][262198] Updated weights for policy 0, policy_version 12052 (0.0011)
+[2026-06-06 21:13:40,793][262198] Updated weights for policy 0, policy_version 12062 (0.0009)
+[2026-06-06 21:13:41,963][262198] Updated weights for policy 0, policy_version 12074 (0.0011)
+[2026-06-06 21:13:42,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4334.9, 300 sec: 4325.1). Total num frames: 6184960. Throughput: 0: 4363.6. Samples: 6192252. Policy #0 lag: (min: 3.0, avg: 19.8, max: 39.0)
+[2026-06-06 21:13:42,922][261293] Avg episode reward: [(0, '1098.191')]
+[2026-06-06 21:13:43,435][262198] Updated weights for policy 0, policy_version 12086 (0.0010)
+[2026-06-06 21:13:44,754][262198] Updated weights for policy 0, policy_version 12096 (0.0009)
+[2026-06-06 21:13:46,683][262198] Updated weights for policy 0, policy_version 12109 (0.0013)
+[2026-06-06 21:13:47,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4369.0, 300 sec: 4332.0). Total num frames: 6205440. Throughput: 0: 4330.5. Samples: 6205556. Policy #0 lag: (min: 6.0, avg: 22.2, max: 38.0)
+[2026-06-06 21:13:47,923][261293] Avg episode reward: [(0, '1264.250')]
+[2026-06-06 21:13:48,111][262198] Updated weights for policy 0, policy_version 12121 (0.0013)
+[2026-06-06 21:13:49,371][262198] Updated weights for policy 0, policy_version 12131 (0.0010)
+[2026-06-06 21:13:50,863][262198] Updated weights for policy 0, policy_version 12145 (0.0011)
+[2026-06-06 21:13:52,064][262198] Updated weights for policy 0, policy_version 12155 (0.0008)
+[2026-06-06 21:13:52,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4334.9, 300 sec: 4325.1). Total num frames: 6225920. Throughput: 0: 4283.3. Samples: 6229790. Policy #0 lag: (min: 6.0, avg: 22.2, max: 38.0)
+[2026-06-06 21:13:52,923][261293] Avg episode reward: [(0, '1377.786')]
+[2026-06-06 21:13:53,909][262198] Updated weights for policy 0, policy_version 12165 (0.0013)
+[2026-06-06 21:13:55,519][262198] Updated weights for policy 0, policy_version 12178 (0.0013)
+[2026-06-06 21:13:56,953][262198] Updated weights for policy 0, policy_version 12188 (0.0012)
+[2026-06-06 21:13:57,921][261293] Fps is (10 sec: 3686.5, 60 sec: 4232.5, 300 sec: 4304.3). Total num frames: 6242304. Throughput: 0: 4181.5. Samples: 6250664. Policy #0 lag: (min: 4.0, avg: 21.4, max: 36.0)
+[2026-06-06 21:13:57,924][261293] Avg episode reward: [(0, '1271.757')]
+[2026-06-06 21:13:58,708][262198] Updated weights for policy 0, policy_version 12198 (0.0014)
+[2026-06-06 21:14:00,345][262198] Updated weights for policy 0, policy_version 12211 (0.0011)
+[2026-06-06 21:14:02,446][262198] Updated weights for policy 0, policy_version 12221 (0.0014)
+[2026-06-06 21:14:02,921][261293] Fps is (10 sec: 3276.8, 60 sec: 4130.1, 300 sec: 4290.4). Total num frames: 6258688. Throughput: 0: 4129.1. Samples: 6260570. Policy #0 lag: (min: 4.0, avg: 21.4, max: 36.0)
+[2026-06-06 21:14:02,922][261293] Avg episode reward: [(0, '1375.683')]
+[2026-06-06 21:14:03,817][262115] Signal inference workers to stop experience collection... (1700 times)
+[2026-06-06 21:14:03,852][262198] InferenceWorker_p0-w0: stopping experience collection (1700 times)
+[2026-06-06 21:14:03,854][262198] Updated weights for policy 0, policy_version 12231 (0.0012)
+[2026-06-06 21:14:03,880][262115] Signal inference workers to resume experience collection... (1700 times)
+[2026-06-06 21:14:03,881][262198] InferenceWorker_p0-w0: resuming experience collection (1700 times)
+[2026-06-06 21:14:05,302][262198] Updated weights for policy 0, policy_version 12243 (0.0013)
+[2026-06-06 21:14:07,039][262198] Updated weights for policy 0, policy_version 12253 (0.0010)
+[2026-06-06 21:14:07,921][261293] Fps is (10 sec: 3481.5, 60 sec: 4061.9, 300 sec: 4262.6). Total num frames: 6277120. Throughput: 0: 3999.8. Samples: 6281038. Policy #0 lag: (min: 3.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:14:07,923][261293] Avg episode reward: [(0, '1410.659')]
+[2026-06-06 21:14:08,326][262198] Updated weights for policy 0, policy_version 12265 (0.0010)
+[2026-06-06 21:14:09,936][262198] Updated weights for policy 0, policy_version 12278 (0.0013)
+[2026-06-06 21:14:11,476][262198] Updated weights for policy 0, policy_version 12288 (0.0010)
+[2026-06-06 21:14:12,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4027.7, 300 sec: 4269.6). Total num frames: 6297600. Throughput: 0: 3981.9. Samples: 6305370. Policy #0 lag: (min: 3.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:14:12,923][261293] Avg episode reward: [(0, '1420.148')]
+[2026-06-06 21:14:13,485][262198] Updated weights for policy 0, policy_version 12304 (0.0010)
+[2026-06-06 21:14:14,533][262198] Updated weights for policy 0, policy_version 12314 (0.0009)
+[2026-06-06 21:14:16,111][262198] Updated weights for policy 0, policy_version 12325 (0.0009)
+[2026-06-06 21:14:17,186][262198] Updated weights for policy 0, policy_version 12337 (0.0009)
+[2026-06-06 21:14:17,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4061.8, 300 sec: 4276.5). Total num frames: 6320128. Throughput: 0: 3972.7. Samples: 6317870. Policy #0 lag: (min: 7.0, avg: 20.9, max: 39.0)
+[2026-06-06 21:14:17,924][261293] Avg episode reward: [(0, '1268.182')]
+[2026-06-06 21:14:18,203][262198] Updated weights for policy 0, policy_version 12347 (0.0008)
+[2026-06-06 21:14:19,872][262198] Updated weights for policy 0, policy_version 12359 (0.0012)
+[2026-06-06 21:14:21,206][262198] Updated weights for policy 0, policy_version 12369 (0.0012)
+[2026-06-06 21:14:22,470][262198] Updated weights for policy 0, policy_version 12379 (0.0011)
+[2026-06-06 21:14:22,921][261293] Fps is (10 sec: 4096.0, 60 sec: 4027.7, 300 sec: 4276.5). Total num frames: 6338560. Throughput: 0: 3903.4. Samples: 6342604. Policy #0 lag: (min: 7.0, avg: 20.9, max: 39.0)
+[2026-06-06 21:14:22,923][261293] Avg episode reward: [(0, '1281.923')]
+[2026-06-06 21:14:24,424][262198] Updated weights for policy 0, policy_version 12389 (0.0013)
+[2026-06-06 21:14:26,048][262198] Updated weights for policy 0, policy_version 12401 (0.0012)
+[2026-06-06 21:14:27,423][262198] Updated weights for policy 0, policy_version 12411 (0.0013)
+[2026-06-06 21:14:27,920][261293] Fps is (10 sec: 3481.8, 60 sec: 3959.6, 300 sec: 4255.7). Total num frames: 6354944. Throughput: 0: 3790.7. Samples: 6362832. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:14:27,921][261293] Avg episode reward: [(0, '1366.003')]
+[2026-06-06 21:14:29,443][262198] Updated weights for policy 0, policy_version 12421 (0.0012)
+[2026-06-06 21:14:30,933][262198] Updated weights for policy 0, policy_version 12434 (0.0011)
+[2026-06-06 21:14:31,963][262198] Updated weights for policy 0, policy_version 12444 (0.0009)
+[2026-06-06 21:14:32,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3891.2, 300 sec: 4248.7). Total num frames: 6373376. Throughput: 0: 3730.7. Samples: 6373438. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:14:32,922][261293] Avg episode reward: [(0, '1450.500')]
+[2026-06-06 21:14:33,412][262198] Updated weights for policy 0, policy_version 12454 (0.0009)
+[2026-06-06 21:14:34,729][262198] Updated weights for policy 0, policy_version 12466 (0.0010)
+[2026-06-06 21:14:35,758][262198] Updated weights for policy 0, policy_version 12476 (0.0009)
+[2026-06-06 21:14:37,444][262198] Updated weights for policy 0, policy_version 12488 (0.0009)
+[2026-06-06 21:14:37,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3891.2, 300 sec: 4255.7). Total num frames: 6395904. Throughput: 0: 3769.1. Samples: 6399398. Policy #0 lag: (min: 3.0, avg: 23.2, max: 39.0)
+[2026-06-06 21:14:37,924][261293] Avg episode reward: [(0, '1392.351')]
+[2026-06-06 21:14:38,804][262198] Updated weights for policy 0, policy_version 12503 (0.0009)
+[2026-06-06 21:14:40,585][262198] Updated weights for policy 0, policy_version 12513 (0.0009)
+[2026-06-06 21:14:41,381][262198] Updated weights for policy 0, policy_version 12524 (0.0011)
+[2026-06-06 21:14:42,905][262198] Updated weights for policy 0, policy_version 12538 (0.0010)
+[2026-06-06 21:14:42,921][261293] Fps is (10 sec: 4505.6, 60 sec: 3891.2, 300 sec: 4269.6). Total num frames: 6418432. Throughput: 0: 3891.1. Samples: 6425762. Policy #0 lag: (min: 3.0, avg: 23.2, max: 39.0)
+[2026-06-06 21:14:42,923][261293] Avg episode reward: [(0, '1296.168')]
+[2026-06-06 21:14:44,523][262198] Updated weights for policy 0, policy_version 12549 (0.0010)
+[2026-06-06 21:14:45,926][262198] Updated weights for policy 0, policy_version 12561 (0.0013)
+[2026-06-06 21:14:46,953][262198] Updated weights for policy 0, policy_version 12571 (0.0010)
+[2026-06-06 21:14:47,921][261293] Fps is (10 sec: 4300.7, 60 sec: 3891.2, 300 sec: 4262.6). Total num frames: 6438912. Throughput: 0: 3965.9. Samples: 6439038. Policy #0 lag: (min: 2.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:14:47,924][261293] Avg episode reward: [(0, '1193.869')]
+[2026-06-06 21:14:48,830][262198] Updated weights for policy 0, policy_version 12582 (0.0010)
+[2026-06-06 21:14:49,994][262198] Updated weights for policy 0, policy_version 12594 (0.0009)
+[2026-06-06 21:14:50,858][262198] Updated weights for policy 0, policy_version 12604 (0.0008)
+[2026-06-06 21:14:52,516][262198] Updated weights for policy 0, policy_version 12614 (0.0010)
+[2026-06-06 21:14:52,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3891.2, 300 sec: 4262.6). Total num frames: 6459392. Throughput: 0: 4050.6. Samples: 6463316. Policy #0 lag: (min: 2.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:14:52,924][261293] Avg episode reward: [(0, '1485.070')]
+[2026-06-06 21:14:54,177][262198] Updated weights for policy 0, policy_version 12630 (0.0010)
+[2026-06-06 21:14:55,530][262198] Updated weights for policy 0, policy_version 12640 (0.0009)
+[2026-06-06 21:14:57,324][262198] Updated weights for policy 0, policy_version 12652 (0.0009)
+[2026-06-06 21:14:57,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3959.5, 300 sec: 4262.6). Total num frames: 6479872. Throughput: 0: 4082.0. Samples: 6489060. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:14:57,923][261293] Avg episode reward: [(0, '1413.237')]
+[2026-06-06 21:14:58,312][262198] Updated weights for policy 0, policy_version 12664 (0.0010)
+[2026-06-06 21:14:58,425][262115] Signal inference workers to stop experience collection... (1750 times)
+[2026-06-06 21:14:58,426][262115] Signal inference workers to resume experience collection... (1750 times)
+[2026-06-06 21:14:58,451][262198] InferenceWorker_p0-w0: stopping experience collection (1750 times)
+[2026-06-06 21:14:58,451][262198] InferenceWorker_p0-w0: resuming experience collection (1750 times)
+[2026-06-06 21:15:00,209][262198] Updated weights for policy 0, policy_version 12676 (0.0012)
+[2026-06-06 21:15:01,475][262198] Updated weights for policy 0, policy_version 12688 (0.0011)
+[2026-06-06 21:15:02,587][262198] Updated weights for policy 0, policy_version 12698 (0.0008)
+[2026-06-06 21:15:02,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4061.9, 300 sec: 4262.6). Total num frames: 6502400. Throughput: 0: 4083.5. Samples: 6501628. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:15:02,924][261293] Avg episode reward: [(0, '1531.970')]
+[2026-06-06 21:15:04,188][262198] Updated weights for policy 0, policy_version 12709 (0.0010)
+[2026-06-06 21:15:05,431][262198] Updated weights for policy 0, policy_version 12723 (0.0009)
+[2026-06-06 21:15:06,737][262198] Updated weights for policy 0, policy_version 12733 (0.0009)
+[2026-06-06 21:15:07,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4096.0, 300 sec: 4262.6). Total num frames: 6522880. Throughput: 0: 4096.1. Samples: 6526928. Policy #0 lag: (min: 5.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:15:07,924][261293] Avg episode reward: [(0, '1336.561')]
+[2026-06-06 21:15:08,836][262198] Updated weights for policy 0, policy_version 12748 (0.0009)
+[2026-06-06 21:15:10,077][262198] Updated weights for policy 0, policy_version 12762 (0.0009)
+[2026-06-06 21:15:11,745][262198] Updated weights for policy 0, policy_version 12773 (0.0009)
+[2026-06-06 21:15:12,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4096.0, 300 sec: 4262.6). Total num frames: 6543360. Throughput: 0: 4230.5. Samples: 6553208. Policy #0 lag: (min: 5.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:15:12,925][261293] Avg episode reward: [(0, '1382.857')]
+[2026-06-06 21:15:13,479][262198] Updated weights for policy 0, policy_version 12788 (0.0008)
+[2026-06-06 21:15:14,637][262198] Updated weights for policy 0, policy_version 12800 (0.0008)
+[2026-06-06 21:15:16,846][262198] Updated weights for policy 0, policy_version 12813 (0.0010)
+[2026-06-06 21:15:17,852][262198] Updated weights for policy 0, policy_version 12828 (0.0009)
+[2026-06-06 21:15:17,921][261293] Fps is (10 sec: 4505.7, 60 sec: 4130.2, 300 sec: 4283.4). Total num frames: 6567936. Throughput: 0: 4270.6. Samples: 6565614. Policy #0 lag: (min: 2.0, avg: 21.9, max: 36.0)
+[2026-06-06 21:15:17,923][261293] Avg episode reward: [(0, '1565.123')]
+[2026-06-06 21:15:17,928][262115] Saving new best policy, reward=1565.123!
+[2026-06-06 21:15:19,592][262198] Updated weights for policy 0, policy_version 12839 (0.0011)
+[2026-06-06 21:15:21,670][262198] Updated weights for policy 0, policy_version 12856 (0.0009)
+[2026-06-06 21:15:22,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4130.2, 300 sec: 4269.6). Total num frames: 6586368. Throughput: 0: 4236.0. Samples: 6590016. Policy #0 lag: (min: 2.0, avg: 21.9, max: 36.0)
+[2026-06-06 21:15:22,923][261293] Avg episode reward: [(0, '1346.305')]
+[2026-06-06 21:15:23,290][262198] Updated weights for policy 0, policy_version 12866 (0.0008)
+[2026-06-06 21:15:24,959][262198] Updated weights for policy 0, policy_version 12880 (0.0009)
+[2026-06-06 21:15:26,476][262198] Updated weights for policy 0, policy_version 12894 (0.0012)
+[2026-06-06 21:15:27,686][262198] Updated weights for policy 0, policy_version 12904 (0.0010)
+[2026-06-06 21:15:27,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4198.3, 300 sec: 4262.6). Total num frames: 6606848. Throughput: 0: 4196.0. Samples: 6614584. Policy #0 lag: (min: 8.0, avg: 21.1, max: 40.0)
+[2026-06-06 21:15:27,923][261293] Avg episode reward: [(0, '1392.704')]
+[2026-06-06 21:15:29,919][262198] Updated weights for policy 0, policy_version 12921 (0.0011)
+[2026-06-06 21:15:31,414][262198] Updated weights for policy 0, policy_version 12932 (0.0008)
+[2026-06-06 21:15:32,920][261293] Fps is (10 sec: 3891.1, 60 sec: 4198.4, 300 sec: 4255.7). Total num frames: 6625280. Throughput: 0: 4158.0. Samples: 6626146. Policy #0 lag: (min: 8.0, avg: 21.1, max: 40.0)
+[2026-06-06 21:15:32,923][261293] Avg episode reward: [(0, '1476.173')]
+[2026-06-06 21:15:33,117][262198] Updated weights for policy 0, policy_version 12944 (0.0010)
+[2026-06-06 21:15:34,501][262198] Updated weights for policy 0, policy_version 12958 (0.0009)
+[2026-06-06 21:15:36,540][262198] Updated weights for policy 0, policy_version 12969 (0.0009)
+[2026-06-06 21:15:37,920][261293] Fps is (10 sec: 4096.3, 60 sec: 4198.4, 300 sec: 4255.7). Total num frames: 6647808. Throughput: 0: 4195.8. Samples: 6652124. Policy #0 lag: (min: 5.0, avg: 24.4, max: 38.0)
+[2026-06-06 21:15:37,923][261293] Avg episode reward: [(0, '1437.297')]
+[2026-06-06 21:15:37,941][262198] Updated weights for policy 0, policy_version 12987 (0.0010)
+[2026-06-06 21:15:39,587][262198] Updated weights for policy 0, policy_version 12999 (0.0009)
+[2026-06-06 21:15:40,869][262115] Signal inference workers to stop experience collection... (1800 times)
+[2026-06-06 21:15:40,897][262198] InferenceWorker_p0-w0: stopping experience collection (1800 times)
+[2026-06-06 21:15:40,966][262115] Signal inference workers to resume experience collection... (1800 times)
+[2026-06-06 21:15:40,966][262198] InferenceWorker_p0-w0: resuming experience collection (1800 times)
+[2026-06-06 21:15:41,668][262198] Updated weights for policy 0, policy_version 13016 (0.0009)
+[2026-06-06 21:15:42,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4164.3, 300 sec: 4262.6). Total num frames: 6668288. Throughput: 0: 4160.9. Samples: 6676298. Policy #0 lag: (min: 5.0, avg: 24.4, max: 38.0)
+[2026-06-06 21:15:42,922][261293] Avg episode reward: [(0, '1420.785')]
+[2026-06-06 21:15:43,195][262198] Updated weights for policy 0, policy_version 13026 (0.0009)
+[2026-06-06 21:15:44,250][262198] Updated weights for policy 0, policy_version 13036 (0.0008)
+[2026-06-06 21:15:45,551][262198] Updated weights for policy 0, policy_version 13051 (0.0010)
+[2026-06-06 21:15:47,046][262198] Updated weights for policy 0, policy_version 13063 (0.0010)
+[2026-06-06 21:15:47,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4164.3, 300 sec: 4262.7). Total num frames: 6688768. Throughput: 0: 4201.2. Samples: 6690678. Policy #0 lag: (min: 14.0, avg: 25.6, max: 47.0)
+[2026-06-06 21:15:47,922][261293] Avg episode reward: [(0, '1552.053')]
+[2026-06-06 21:15:48,831][262198] Updated weights for policy 0, policy_version 13078 (0.0010)
+[2026-06-06 21:15:49,796][262198] Updated weights for policy 0, policy_version 13088 (0.0008)
+[2026-06-06 21:15:51,580][262198] Updated weights for policy 0, policy_version 13100 (0.0009)
+[2026-06-06 21:15:52,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4266.7, 300 sec: 4283.4). Total num frames: 6715392. Throughput: 0: 4242.6. Samples: 6717846. Policy #0 lag: (min: 14.0, avg: 25.6, max: 47.0)
+[2026-06-06 21:15:52,923][261293] Avg episode reward: [(0, '1381.754')]
+[2026-06-06 21:15:53,309][262198] Updated weights for policy 0, policy_version 13117 (0.0009)
+[2026-06-06 21:15:55,091][262198] Updated weights for policy 0, policy_version 13129 (0.0009)
+[2026-06-06 21:15:56,389][262198] Updated weights for policy 0, policy_version 13144 (0.0009)
+[2026-06-06 21:15:57,729][262198] Updated weights for policy 0, policy_version 13154 (0.0009)
+[2026-06-06 21:15:57,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4266.7, 300 sec: 4269.6). Total num frames: 6735872. Throughput: 0: 4240.6. Samples: 6744034. Policy #0 lag: (min: 3.0, avg: 17.8, max: 35.0)
+[2026-06-06 21:15:57,924][261293] Avg episode reward: [(0, '1358.276')]
+[2026-06-06 21:15:59,577][262198] Updated weights for policy 0, policy_version 13168 (0.0012)
+[2026-06-06 21:16:00,634][262198] Updated weights for policy 0, policy_version 13179 (0.0009)
+[2026-06-06 21:16:02,696][262198] Updated weights for policy 0, policy_version 13191 (0.0011)
+[2026-06-06 21:16:02,923][261293] Fps is (10 sec: 3890.4, 60 sec: 4198.3, 300 sec: 4248.7). Total num frames: 6754304. Throughput: 0: 4188.8. Samples: 6754120. Policy #0 lag: (min: 3.0, avg: 17.8, max: 35.0)
+[2026-06-06 21:16:02,927][261293] Avg episode reward: [(0, '1324.733')]
+[2026-06-06 21:16:04,781][262198] Updated weights for policy 0, policy_version 13208 (0.0011)
+[2026-06-06 21:16:06,452][262198] Updated weights for policy 0, policy_version 13218 (0.0009)
+[2026-06-06 21:16:07,921][261293] Fps is (10 sec: 3686.3, 60 sec: 4164.3, 300 sec: 4234.8). Total num frames: 6772736. Throughput: 0: 4184.1. Samples: 6778302. Policy #0 lag: (min: 3.0, avg: 21.2, max: 35.0)
+[2026-06-06 21:16:07,923][261293] Avg episode reward: [(0, '1581.662')]
+[2026-06-06 21:16:08,182][262115] Saving new best policy, reward=1581.662!
+[2026-06-06 21:16:08,903][262198] Updated weights for policy 0, policy_version 13240 (0.0009)
+[2026-06-06 21:16:10,967][262198] Updated weights for policy 0, policy_version 13250 (0.0012)
+[2026-06-06 21:16:12,857][262198] Updated weights for policy 0, policy_version 13264 (0.0013)
+[2026-06-06 21:16:12,920][261293] Fps is (10 sec: 3687.3, 60 sec: 4130.2, 300 sec: 4227.9). Total num frames: 6791168. Throughput: 0: 4123.0. Samples: 6800114. Policy #0 lag: (min: 3.0, avg: 21.2, max: 35.0)
+[2026-06-06 21:16:12,922][261293] Avg episode reward: [(0, '1473.687')]
+[2026-06-06 21:16:14,340][262198] Updated weights for policy 0, policy_version 13277 (0.0009)
+[2026-06-06 21:16:16,191][262198] Updated weights for policy 0, policy_version 13289 (0.0009)
+[2026-06-06 21:16:17,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4096.0, 300 sec: 4227.9). Total num frames: 6813696. Throughput: 0: 4145.3. Samples: 6812686. Policy #0 lag: (min: 3.0, avg: 16.5, max: 35.0)
+[2026-06-06 21:16:17,923][261293] Avg episode reward: [(0, '1330.880')]
+[2026-06-06 21:16:18,058][262198] Updated weights for policy 0, policy_version 13309 (0.0009)
+[2026-06-06 21:16:19,886][262198] Updated weights for policy 0, policy_version 13322 (0.0009)
+[2026-06-06 21:16:20,520][262115] Signal inference workers to stop experience collection... (1850 times)
+[2026-06-06 21:16:20,541][262198] InferenceWorker_p0-w0: stopping experience collection (1850 times)
+[2026-06-06 21:16:20,635][262115] Signal inference workers to resume experience collection... (1850 times)
+[2026-06-06 21:16:20,636][262198] InferenceWorker_p0-w0: resuming experience collection (1850 times)
+[2026-06-06 21:16:21,104][262198] Updated weights for policy 0, policy_version 13337 (0.0009)
+[2026-06-06 21:16:22,466][262198] Updated weights for policy 0, policy_version 13349 (0.0009)
+[2026-06-06 21:16:22,922][261293] Fps is (10 sec: 4505.3, 60 sec: 4164.2, 300 sec: 4227.9). Total num frames: 6836224. Throughput: 0: 4155.3. Samples: 6839116. Policy #0 lag: (min: 3.0, avg: 16.5, max: 35.0)
+[2026-06-06 21:16:22,926][261293] Avg episode reward: [(0, '1307.751')]
+[2026-06-06 21:16:24,184][262198] Updated weights for policy 0, policy_version 13364 (0.0009)
+[2026-06-06 21:16:25,229][262198] Updated weights for policy 0, policy_version 13376 (0.0008)
+[2026-06-06 21:16:27,310][262198] Updated weights for policy 0, policy_version 13389 (0.0008)
+[2026-06-06 21:16:27,923][261293] Fps is (10 sec: 4504.8, 60 sec: 4198.3, 300 sec: 4234.8). Total num frames: 6858752. Throughput: 0: 4250.0. Samples: 6867556. Policy #0 lag: (min: 7.0, avg: 22.6, max: 39.0)
+[2026-06-06 21:16:27,926][261293] Avg episode reward: [(0, '1402.577')]
+[2026-06-06 21:16:28,362][262198] Updated weights for policy 0, policy_version 13404 (0.0008)
+[2026-06-06 21:16:29,819][262198] Updated weights for policy 0, policy_version 13416 (0.0010)
+[2026-06-06 21:16:31,765][262198] Updated weights for policy 0, policy_version 13434 (0.0009)
+[2026-06-06 21:16:32,921][261293] Fps is (10 sec: 4505.8, 60 sec: 4266.7, 300 sec: 4227.9). Total num frames: 6881280. Throughput: 0: 4229.7. Samples: 6881014. Policy #0 lag: (min: 7.0, avg: 22.6, max: 39.0)
+[2026-06-06 21:16:32,924][261293] Avg episode reward: [(0, '1583.209')]
+[2026-06-06 21:16:32,931][262115] Saving new best policy, reward=1583.209!
+[2026-06-06 21:16:33,603][262198] Updated weights for policy 0, policy_version 13446 (0.0010)
+[2026-06-06 21:16:35,557][262198] Updated weights for policy 0, policy_version 13461 (0.0008)
+[2026-06-06 21:16:37,019][262198] Updated weights for policy 0, policy_version 13473 (0.0009)
+[2026-06-06 21:16:37,920][261293] Fps is (10 sec: 4301.7, 60 sec: 4232.5, 300 sec: 4227.9). Total num frames: 6901760. Throughput: 0: 4177.0. Samples: 6905812. Policy #0 lag: (min: 0.0, avg: 21.4, max: 34.0)
+[2026-06-06 21:16:37,922][261293] Avg episode reward: [(0, '1421.344')]
+[2026-06-06 21:16:38,818][262198] Updated weights for policy 0, policy_version 13488 (0.0012)
+[2026-06-06 21:16:40,338][262198] Updated weights for policy 0, policy_version 13502 (0.0009)
+[2026-06-06 21:16:41,397][262198] Updated weights for policy 0, policy_version 13512 (0.0011)
+[2026-06-06 21:16:42,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4198.4, 300 sec: 4214.0). Total num frames: 6920192. Throughput: 0: 4145.0. Samples: 6930558. Policy #0 lag: (min: 0.0, avg: 21.4, max: 34.0)
+[2026-06-06 21:16:42,923][261293] Avg episode reward: [(0, '1182.511')]
+[2026-06-06 21:16:43,678][262198] Updated weights for policy 0, policy_version 13526 (0.0009)
+[2026-06-06 21:16:45,125][262198] Updated weights for policy 0, policy_version 13537 (0.0009)
+[2026-06-06 21:16:46,341][262198] Updated weights for policy 0, policy_version 13548 (0.0009)
+[2026-06-06 21:16:47,462][262198] Updated weights for policy 0, policy_version 13561 (0.0008)
+[2026-06-06 21:16:47,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4266.7, 300 sec: 4221.0). Total num frames: 6944768. Throughput: 0: 4202.7. Samples: 6943230. Policy #0 lag: (min: 5.0, avg: 19.6, max: 39.0)
+[2026-06-06 21:16:47,923][261293] Avg episode reward: [(0, '1561.498')]
+[2026-06-06 21:16:48,961][262198] Updated weights for policy 0, policy_version 13573 (0.0008)
+[2026-06-06 21:16:50,691][262198] Updated weights for policy 0, policy_version 13585 (0.0009)
+[2026-06-06 21:16:51,377][262198] Updated weights for policy 0, policy_version 13596 (0.0010)
+[2026-06-06 21:16:52,922][261293] Fps is (10 sec: 4504.8, 60 sec: 4164.1, 300 sec: 4214.0). Total num frames: 6965248. Throughput: 0: 4233.8. Samples: 6968830. Policy #0 lag: (min: 5.0, avg: 19.6, max: 39.0)
+[2026-06-06 21:16:52,924][261293] Avg episode reward: [(0, '1382.873')]
+[2026-06-06 21:16:53,018][262198] Updated weights for policy 0, policy_version 13608 (0.0010)
+[2026-06-06 21:16:54,943][262198] Updated weights for policy 0, policy_version 13621 (0.0009)
+[2026-06-06 21:16:56,006][262198] Updated weights for policy 0, policy_version 13632 (0.0010)
+[2026-06-06 21:16:57,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4130.1, 300 sec: 4200.1). Total num frames: 6983680. Throughput: 0: 4298.0. Samples: 6993528. Policy #0 lag: (min: 7.0, avg: 22.9, max: 39.0)
+[2026-06-06 21:16:57,923][261293] Avg episode reward: [(0, '1405.505')]
+[2026-06-06 21:16:58,514][262198] Updated weights for policy 0, policy_version 13646 (0.0011)
+[2026-06-06 21:16:58,822][262115] Signal inference workers to stop experience collection... (1900 times)
+[2026-06-06 21:16:58,836][262198] InferenceWorker_p0-w0: stopping experience collection (1900 times)
+[2026-06-06 21:16:58,946][262115] Signal inference workers to resume experience collection... (1900 times)
+[2026-06-06 21:16:58,946][262198] InferenceWorker_p0-w0: resuming experience collection (1900 times)
+[2026-06-06 21:16:59,567][262198] Updated weights for policy 0, policy_version 13659 (0.0012)
+[2026-06-06 21:17:01,553][262198] Updated weights for policy 0, policy_version 13670 (0.0013)
+[2026-06-06 21:17:02,921][261293] Fps is (10 sec: 3482.2, 60 sec: 4096.1, 300 sec: 4179.3). Total num frames: 7000064. Throughput: 0: 4256.1. Samples: 7004212. Policy #0 lag: (min: 7.0, avg: 22.9, max: 39.0)
+[2026-06-06 21:17:02,923][261293] Avg episode reward: [(0, '1530.865')]
+[2026-06-06 21:17:03,436][262198] Updated weights for policy 0, policy_version 13680 (0.0013)
+[2026-06-06 21:17:04,278][262198] Updated weights for policy 0, policy_version 13691 (0.0010)
+[2026-06-06 21:17:06,206][262198] Updated weights for policy 0, policy_version 13703 (0.0015)
+[2026-06-06 21:17:07,920][261293] Fps is (10 sec: 3481.7, 60 sec: 4096.0, 300 sec: 4165.4). Total num frames: 7018496. Throughput: 0: 4126.0. Samples: 7024784. Policy #0 lag: (min: 4.0, avg: 22.4, max: 41.0)
+[2026-06-06 21:17:07,922][261293] Avg episode reward: [(0, '1505.613')]
+[2026-06-06 21:17:08,547][262198] Updated weights for policy 0, policy_version 13717 (0.0014)
+[2026-06-06 21:17:09,651][262198] Updated weights for policy 0, policy_version 13727 (0.0010)
+[2026-06-06 21:17:11,859][262198] Updated weights for policy 0, policy_version 13739 (0.0010)
+[2026-06-06 21:17:12,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4130.1, 300 sec: 4158.5). Total num frames: 7038976. Throughput: 0: 3987.7. Samples: 7046996. Policy #0 lag: (min: 4.0, avg: 22.4, max: 41.0)
+[2026-06-06 21:17:12,927][261293] Avg episode reward: [(0, '1388.384')]
+[2026-06-06 21:17:13,252][262198] Updated weights for policy 0, policy_version 13753 (0.0012)
+[2026-06-06 21:17:14,744][262198] Updated weights for policy 0, policy_version 13765 (0.0009)
+[2026-06-06 21:17:16,976][262198] Updated weights for policy 0, policy_version 13781 (0.0008)
+[2026-06-06 21:17:17,829][262198] Updated weights for policy 0, policy_version 13792 (0.0009)
+[2026-06-06 21:17:17,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4130.1, 300 sec: 4158.5). Total num frames: 7061504. Throughput: 0: 3978.3. Samples: 7060036. Policy #0 lag: (min: 4.0, avg: 22.4, max: 41.0)
+[2026-06-06 21:17:17,923][261293] Avg episode reward: [(0, '1426.562')]
+[2026-06-06 21:17:20,070][262198] Updated weights for policy 0, policy_version 13805 (0.0010)
+[2026-06-06 21:17:21,199][262198] Updated weights for policy 0, policy_version 13820 (0.0010)
+[2026-06-06 21:17:22,722][262198] Updated weights for policy 0, policy_version 13830 (0.0009)
+[2026-06-06 21:17:22,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4096.0, 300 sec: 4144.6). Total num frames: 7081984. Throughput: 0: 3965.4. Samples: 7084258. Policy #0 lag: (min: 3.0, avg: 20.5, max: 35.0)
+[2026-06-06 21:17:22,927][261293] Avg episode reward: [(0, '1393.929')]
+[2026-06-06 21:17:24,521][262198] Updated weights for policy 0, policy_version 13841 (0.0011)
+[2026-06-06 21:17:25,751][262198] Updated weights for policy 0, policy_version 13853 (0.0011)
+[2026-06-06 21:17:27,790][262198] Updated weights for policy 0, policy_version 13865 (0.0009)
+[2026-06-06 21:17:27,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3993.7, 300 sec: 4123.8). Total num frames: 7098368. Throughput: 0: 3966.9. Samples: 7109070. Policy #0 lag: (min: 3.0, avg: 20.5, max: 35.0)
+[2026-06-06 21:17:27,922][261293] Avg episode reward: [(0, '1473.247')]
+[2026-06-06 21:17:29,013][262198] Updated weights for policy 0, policy_version 13878 (0.0011)
+[2026-06-06 21:17:30,418][262198] Updated weights for policy 0, policy_version 13889 (0.0010)
+[2026-06-06 21:17:32,170][262198] Updated weights for policy 0, policy_version 13901 (0.0008)
+[2026-06-06 21:17:32,921][261293] Fps is (10 sec: 4096.1, 60 sec: 4027.7, 300 sec: 4130.7). Total num frames: 7122944. Throughput: 0: 3950.8. Samples: 7121014. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:17:32,922][261293] Avg episode reward: [(0, '1398.042')]
+[2026-06-06 21:17:33,239][262198] Updated weights for policy 0, policy_version 13914 (0.0010)
+[2026-06-06 21:17:34,639][262198] Updated weights for policy 0, policy_version 13925 (0.0009)
+[2026-06-06 21:17:36,322][262198] Updated weights for policy 0, policy_version 13937 (0.0009)
+[2026-06-06 21:17:37,457][262198] Updated weights for policy 0, policy_version 13949 (0.0009)
+[2026-06-06 21:17:37,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4027.7, 300 sec: 4130.7). Total num frames: 7143424. Throughput: 0: 3948.8. Samples: 7146518. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:17:37,924][261293] Avg episode reward: [(0, '1415.001')]
+[2026-06-06 21:17:38,379][262198] Updated weights for policy 0, policy_version 13959 (0.0009)
+[2026-06-06 21:17:40,533][262198] Updated weights for policy 0, policy_version 13974 (0.0009)
+[2026-06-06 21:17:41,657][262198] Updated weights for policy 0, policy_version 13984 (0.0009)
+[2026-06-06 21:17:42,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4061.9, 300 sec: 4137.6). Total num frames: 7163904. Throughput: 0: 3951.2. Samples: 7171332. Policy #0 lag: (min: 6.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:17:42,922][261293] Avg episode reward: [(0, '1206.548')]
+[2026-06-06 21:17:43,655][262115] Signal inference workers to stop experience collection... (1950 times)
+[2026-06-06 21:17:43,675][262198] InferenceWorker_p0-w0: stopping experience collection (1950 times)
+[2026-06-06 21:17:43,806][262115] Signal inference workers to resume experience collection... (1950 times)
+[2026-06-06 21:17:43,807][262198] InferenceWorker_p0-w0: resuming experience collection (1950 times)
+[2026-06-06 21:17:44,155][262198] Updated weights for policy 0, policy_version 14000 (0.0010)
+[2026-06-06 21:17:45,341][262198] Updated weights for policy 0, policy_version 14012 (0.0010)
+[2026-06-06 21:17:46,723][262198] Updated weights for policy 0, policy_version 14022 (0.0009)
+[2026-06-06 21:17:47,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3925.3, 300 sec: 4116.8). Total num frames: 7180288. Throughput: 0: 3968.7. Samples: 7182804. Policy #0 lag: (min: 6.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:17:47,922][261293] Avg episode reward: [(0, '1410.765')]
+[2026-06-06 21:17:48,880][262198] Updated weights for policy 0, policy_version 14038 (0.0011)
+[2026-06-06 21:17:49,865][262198] Updated weights for policy 0, policy_version 14048 (0.0010)
+[2026-06-06 21:17:52,158][262198] Updated weights for policy 0, policy_version 14061 (0.0010)
+[2026-06-06 21:17:52,920][261293] Fps is (10 sec: 4096.2, 60 sec: 3993.7, 300 sec: 4123.8). Total num frames: 7204864. Throughput: 0: 4065.6. Samples: 7207736. Policy #0 lag: (min: 7.0, avg: 24.5, max: 38.0)
+[2026-06-06 21:17:52,922][261293] Avg episode reward: [(0, '1393.395')]
+[2026-06-06 21:17:53,294][262198] Updated weights for policy 0, policy_version 14074 (0.0010)
+[2026-06-06 21:17:54,491][262198] Updated weights for policy 0, policy_version 14085 (0.0010)
+[2026-06-06 21:17:56,608][262198] Updated weights for policy 0, policy_version 14100 (0.0010)
+[2026-06-06 21:17:57,515][262198] Updated weights for policy 0, policy_version 14111 (0.0009)
+[2026-06-06 21:17:57,927][261293] Fps is (10 sec: 4502.8, 60 sec: 4027.3, 300 sec: 4116.7). Total num frames: 7225344. Throughput: 0: 4146.4. Samples: 7233610. Policy #0 lag: (min: 7.0, avg: 24.5, max: 38.0)
+[2026-06-06 21:17:57,929][261293] Avg episode reward: [(0, '1531.057')]
+[2026-06-06 21:17:59,611][262198] Updated weights for policy 0, policy_version 14124 (0.0011)
+[2026-06-06 21:18:00,561][262198] Updated weights for policy 0, policy_version 14136 (0.0011)
+[2026-06-06 21:18:02,252][262198] Updated weights for policy 0, policy_version 14148 (0.0010)
+[2026-06-06 21:18:02,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4096.0, 300 sec: 4109.9). Total num frames: 7245824. Throughput: 0: 4121.7. Samples: 7245512. Policy #0 lag: (min: 3.0, avg: 22.0, max: 42.0)
+[2026-06-06 21:18:02,923][261293] Avg episode reward: [(0, '1214.406')]
+[2026-06-06 21:18:04,322][262198] Updated weights for policy 0, policy_version 14165 (0.0010)
+[2026-06-06 21:18:05,273][262198] Updated weights for policy 0, policy_version 14175 (0.0010)
+[2026-06-06 21:18:07,632][262198] Updated weights for policy 0, policy_version 14192 (0.0010)
+[2026-06-06 21:18:07,921][261293] Fps is (10 sec: 4098.5, 60 sec: 4130.1, 300 sec: 4102.9). Total num frames: 7266304. Throughput: 0: 4150.4. Samples: 7271028. Policy #0 lag: (min: 3.0, avg: 22.0, max: 42.0)
+[2026-06-06 21:18:07,923][261293] Avg episode reward: [(0, '1500.115')]
+[2026-06-06 21:18:08,572][262198] Updated weights for policy 0, policy_version 14203 (0.0011)
+[2026-06-06 21:18:09,771][262198] Updated weights for policy 0, policy_version 14214 (0.0009)
+[2026-06-06 21:18:12,024][262198] Updated weights for policy 0, policy_version 14229 (0.0011)
+[2026-06-06 21:18:12,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4164.3, 300 sec: 4109.9). Total num frames: 7288832. Throughput: 0: 4153.6. Samples: 7295984. Policy #0 lag: (min: 8.0, avg: 19.9, max: 40.0)
+[2026-06-06 21:18:12,922][261293] Avg episode reward: [(0, '1388.093')]
+[2026-06-06 21:18:13,173][262198] Updated weights for policy 0, policy_version 14239 (0.0011)
+[2026-06-06 21:18:15,338][262198] Updated weights for policy 0, policy_version 14252 (0.0010)
+[2026-06-06 21:18:16,691][262198] Updated weights for policy 0, policy_version 14266 (0.0010)
+[2026-06-06 21:18:17,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4130.1, 300 sec: 4109.9). Total num frames: 7309312. Throughput: 0: 4186.0. Samples: 7309382. Policy #0 lag: (min: 8.0, avg: 19.9, max: 40.0)
+[2026-06-06 21:18:17,922][261293] Avg episode reward: [(0, '1465.138')]
+[2026-06-06 21:18:18,004][262198] Updated weights for policy 0, policy_version 14278 (0.0010)
+[2026-06-06 21:18:19,976][262198] Updated weights for policy 0, policy_version 14290 (0.0014)
+[2026-06-06 21:18:21,316][262198] Updated weights for policy 0, policy_version 14301 (0.0010)
+[2026-06-06 21:18:22,921][261293] Fps is (10 sec: 3891.2, 60 sec: 4096.0, 300 sec: 4103.0). Total num frames: 7327744. Throughput: 0: 4126.6. Samples: 7332216. Policy #0 lag: (min: 5.0, avg: 22.3, max: 37.0)
+[2026-06-06 21:18:22,923][261293] Avg episode reward: [(0, '1425.527')]
+[2026-06-06 21:18:23,696][262198] Updated weights for policy 0, policy_version 14313 (0.0010)
+[2026-06-06 21:18:24,670][262198] Updated weights for policy 0, policy_version 14326 (0.0009)
+[2026-06-06 21:18:25,628][262198] Updated weights for policy 0, policy_version 14336 (0.0010)
+[2026-06-06 21:18:27,831][262115] Signal inference workers to stop experience collection... (2000 times)
+[2026-06-06 21:18:27,863][262198] InferenceWorker_p0-w0: stopping experience collection (2000 times)
+[2026-06-06 21:18:27,921][261293] Fps is (10 sec: 3481.5, 60 sec: 4096.0, 300 sec: 4082.1). Total num frames: 7344128. Throughput: 0: 4092.6. Samples: 7355498. Policy #0 lag: (min: 5.0, avg: 22.3, max: 37.0)
+[2026-06-06 21:18:27,923][261293] Avg episode reward: [(0, '1429.988')]
+[2026-06-06 21:18:27,939][262115] Signal inference workers to resume experience collection... (2000 times)
+[2026-06-06 21:18:27,940][262198] InferenceWorker_p0-w0: resuming experience collection (2000 times)
+[2026-06-06 21:18:28,327][262198] Updated weights for policy 0, policy_version 14352 (0.0010)
+[2026-06-06 21:18:29,643][262198] Updated weights for policy 0, policy_version 14364 (0.0010)
+[2026-06-06 21:18:30,774][262198] Updated weights for policy 0, policy_version 14376 (0.0009)
+[2026-06-06 21:18:32,921][261293] Fps is (10 sec: 3686.2, 60 sec: 4027.7, 300 sec: 4075.2). Total num frames: 7364608. Throughput: 0: 4084.9. Samples: 7366626. Policy #0 lag: (min: 5.0, avg: 22.3, max: 37.0)
+[2026-06-06 21:18:32,925][261293] Avg episode reward: [(0, '1482.463')]
+[2026-06-06 21:18:33,108][262198] Updated weights for policy 0, policy_version 14391 (0.0014)
+[2026-06-06 21:18:34,697][262198] Updated weights for policy 0, policy_version 14401 (0.0009)
+[2026-06-06 21:18:36,977][262198] Updated weights for policy 0, policy_version 14417 (0.0011)
+[2026-06-06 21:18:37,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4027.7, 300 sec: 4068.2). Total num frames: 7385088. Throughput: 0: 4058.2. Samples: 7390354. Policy #0 lag: (min: 11.0, avg: 23.9, max: 43.0)
+[2026-06-06 21:18:37,922][261293] Avg episode reward: [(0, '1397.737')]
+[2026-06-06 21:18:38,404][262198] Updated weights for policy 0, policy_version 14429 (0.0009)
+[2026-06-06 21:18:40,295][262198] Updated weights for policy 0, policy_version 14442 (0.0011)
+[2026-06-06 21:18:41,257][262198] Updated weights for policy 0, policy_version 14456 (0.0011)
+[2026-06-06 21:18:42,920][261293] Fps is (10 sec: 3891.5, 60 sec: 3993.6, 300 sec: 4061.3). Total num frames: 7403520. Throughput: 0: 3997.9. Samples: 7413492. Policy #0 lag: (min: 11.0, avg: 23.9, max: 43.0)
+[2026-06-06 21:18:42,921][261293] Avg episode reward: [(0, '1344.584')]
+[2026-06-06 21:18:43,390][262198] Updated weights for policy 0, policy_version 14468 (0.0012)
+[2026-06-06 21:18:45,670][262198] Updated weights for policy 0, policy_version 14484 (0.0009)
+[2026-06-06 21:18:47,112][262198] Updated weights for policy 0, policy_version 14496 (0.0010)
+[2026-06-06 21:18:47,920][261293] Fps is (10 sec: 4095.9, 60 sec: 4096.0, 300 sec: 4068.2). Total num frames: 7426048. Throughput: 0: 4001.0. Samples: 7425554. Policy #0 lag: (min: 3.0, avg: 17.3, max: 35.0)
+[2026-06-06 21:18:47,922][261293] Avg episode reward: [(0, '1455.348')]
+[2026-06-06 21:18:49,182][262198] Updated weights for policy 0, policy_version 14509 (0.0012)
+[2026-06-06 21:18:50,632][262198] Updated weights for policy 0, policy_version 14521 (0.0013)
+[2026-06-06 21:18:51,899][262198] Updated weights for policy 0, policy_version 14533 (0.0011)
+[2026-06-06 21:18:52,920][261293] Fps is (10 sec: 3891.1, 60 sec: 3959.5, 300 sec: 4068.2). Total num frames: 7442432. Throughput: 0: 3933.9. Samples: 7448052. Policy #0 lag: (min: 3.0, avg: 17.3, max: 35.0)
+[2026-06-06 21:18:52,922][261293] Avg episode reward: [(0, '1441.194')]
+[2026-06-06 21:18:54,185][262198] Updated weights for policy 0, policy_version 14549 (0.0011)
+[2026-06-06 21:18:55,708][262198] Updated weights for policy 0, policy_version 14561 (0.0009)
+[2026-06-06 21:18:57,320][262198] Updated weights for policy 0, policy_version 14573 (0.0009)
+[2026-06-06 21:18:57,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3994.0, 300 sec: 4089.1). Total num frames: 7464960. Throughput: 0: 3936.6. Samples: 7473130. Policy #0 lag: (min: 6.0, avg: 21.9, max: 38.0)
+[2026-06-06 21:18:57,924][261293] Avg episode reward: [(0, '1535.524')]
+[2026-06-06 21:18:58,610][262198] Updated weights for policy 0, policy_version 14586 (0.0008)
+[2026-06-06 21:18:59,767][262198] Updated weights for policy 0, policy_version 14598 (0.0009)
+[2026-06-06 21:19:01,604][262198] Updated weights for policy 0, policy_version 14613 (0.0010)
+[2026-06-06 21:19:02,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3993.6, 300 sec: 4096.0). Total num frames: 7485440. Throughput: 0: 3928.3. Samples: 7486156. Policy #0 lag: (min: 6.0, avg: 21.9, max: 38.0)
+[2026-06-06 21:19:02,921][261293] Avg episode reward: [(0, '1571.766')]
+[2026-06-06 21:19:03,175][262198] Updated weights for policy 0, policy_version 14623 (0.0011)
+[2026-06-06 21:19:05,202][262198] Updated weights for policy 0, policy_version 14634 (0.0010)
+[2026-06-06 21:19:06,216][262198] Updated weights for policy 0, policy_version 14647 (0.0009)
+[2026-06-06 21:19:07,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3993.6, 300 sec: 4096.0). Total num frames: 7505920. Throughput: 0: 3917.8. Samples: 7508520. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:19:07,924][261293] Avg episode reward: [(0, '1565.428')]
+[2026-06-06 21:19:07,949][262198] Updated weights for policy 0, policy_version 14661 (0.0011)
+[2026-06-06 21:19:09,736][262115] Signal inference workers to stop experience collection... (2050 times)
+[2026-06-06 21:19:09,755][262198] InferenceWorker_p0-w0: stopping experience collection (2050 times)
+[2026-06-06 21:19:09,775][262115] Signal inference workers to resume experience collection... (2050 times)
+[2026-06-06 21:19:09,776][262198] InferenceWorker_p0-w0: resuming experience collection (2050 times)
+[2026-06-06 21:19:10,243][262198] Updated weights for policy 0, policy_version 14673 (0.0012)
+[2026-06-06 21:19:11,464][262198] Updated weights for policy 0, policy_version 14683 (0.0011)
+[2026-06-06 21:19:12,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3891.2, 300 sec: 4075.2). Total num frames: 7522304. Throughput: 0: 3883.4. Samples: 7530250. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:19:12,922][261293] Avg episode reward: [(0, '1478.680')]
+[2026-06-06 21:19:12,932][262198] Updated weights for policy 0, policy_version 14696 (0.0014)
+[2026-06-06 21:19:15,063][262198] Updated weights for policy 0, policy_version 14707 (0.0011)
+[2026-06-06 21:19:16,469][262198] Updated weights for policy 0, policy_version 14717 (0.0009)
+[2026-06-06 21:19:17,921][261293] Fps is (10 sec: 3481.9, 60 sec: 3857.1, 300 sec: 4075.2). Total num frames: 7540736. Throughput: 0: 3869.1. Samples: 7540732. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:19:17,923][261293] Avg episode reward: [(0, '1435.294')]
+[2026-06-06 21:19:18,236][262198] Updated weights for policy 0, policy_version 14731 (0.0010)
+[2026-06-06 21:19:18,286][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000014732_7542784.pth...
+[2026-06-06 21:19:19,345][262198] Updated weights for policy 0, policy_version 14744 (0.0010)
+[2026-06-06 21:19:21,179][262198] Updated weights for policy 0, policy_version 14757 (0.0010)
+[2026-06-06 21:19:22,859][262198] Updated weights for policy 0, policy_version 14767 (0.0014)
+[2026-06-06 21:19:22,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3891.2, 300 sec: 4089.1). Total num frames: 7561216. Throughput: 0: 3889.5. Samples: 7565382. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:19:22,924][261293] Avg episode reward: [(0, '1482.739')]
+[2026-06-06 21:19:24,225][262198] Updated weights for policy 0, policy_version 14777 (0.0012)
+[2026-06-06 21:19:25,750][262198] Updated weights for policy 0, policy_version 14791 (0.0011)
+[2026-06-06 21:19:27,361][262198] Updated weights for policy 0, policy_version 14802 (0.0010)
+[2026-06-06 21:19:27,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3959.4, 300 sec: 4096.0). Total num frames: 7581696. Throughput: 0: 3872.7. Samples: 7587766. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:19:27,924][261293] Avg episode reward: [(0, '1483.920')]
+[2026-06-06 21:19:28,321][262198] Updated weights for policy 0, policy_version 14812 (0.0009)
+[2026-06-06 21:19:30,878][262198] Updated weights for policy 0, policy_version 14827 (0.0010)
+[2026-06-06 21:19:31,703][262198] Updated weights for policy 0, policy_version 14837 (0.0011)
+[2026-06-06 21:19:32,922][261293] Fps is (10 sec: 3890.8, 60 sec: 3925.3, 300 sec: 4082.1). Total num frames: 7600128. Throughput: 0: 3869.2. Samples: 7599672. Policy #0 lag: (min: 2.0, avg: 21.0, max: 38.0)
+[2026-06-06 21:19:32,924][261293] Avg episode reward: [(0, '1404.343')]
+[2026-06-06 21:19:33,484][262198] Updated weights for policy 0, policy_version 14849 (0.0010)
+[2026-06-06 21:19:35,145][262198] Updated weights for policy 0, policy_version 14861 (0.0010)
+[2026-06-06 21:19:35,996][262198] Updated weights for policy 0, policy_version 14871 (0.0010)
+[2026-06-06 21:19:37,903][262198] Updated weights for policy 0, policy_version 14885 (0.0013)
+[2026-06-06 21:19:37,920][261293] Fps is (10 sec: 3891.4, 60 sec: 3925.3, 300 sec: 4075.2). Total num frames: 7620608. Throughput: 0: 3912.1. Samples: 7624094. Policy #0 lag: (min: 2.0, avg: 21.0, max: 38.0)
+[2026-06-06 21:19:37,922][261293] Avg episode reward: [(0, '1412.825')]
+[2026-06-06 21:19:39,609][262198] Updated weights for policy 0, policy_version 14896 (0.0015)
+[2026-06-06 21:19:40,773][262198] Updated weights for policy 0, policy_version 14906 (0.0009)
+[2026-06-06 21:19:42,163][262198] Updated weights for policy 0, policy_version 14918 (0.0010)
+[2026-06-06 21:19:42,922][261293] Fps is (10 sec: 3891.3, 60 sec: 3925.3, 300 sec: 4068.2). Total num frames: 7639040. Throughput: 0: 3876.5. Samples: 7647576. Policy #0 lag: (min: 4.0, avg: 20.5, max: 39.0)
+[2026-06-06 21:19:42,926][261293] Avg episode reward: [(0, '1561.162')]
+[2026-06-06 21:19:44,062][262198] Updated weights for policy 0, policy_version 14931 (0.0010)
+[2026-06-06 21:19:46,340][262198] Updated weights for policy 0, policy_version 14946 (0.0009)
+[2026-06-06 21:19:47,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3822.9, 300 sec: 4054.3). Total num frames: 7655424. Throughput: 0: 3830.0. Samples: 7658508. Policy #0 lag: (min: 4.0, avg: 20.5, max: 39.0)
+[2026-06-06 21:19:47,922][261293] Avg episode reward: [(0, '1332.998')]
+[2026-06-06 21:19:48,548][262198] Updated weights for policy 0, policy_version 14961 (0.0010)
+[2026-06-06 21:19:50,079][262198] Updated weights for policy 0, policy_version 14973 (0.0011)
+[2026-06-06 21:19:51,934][262115] Signal inference workers to stop experience collection... (2100 times)
+[2026-06-06 21:19:51,954][262198] InferenceWorker_p0-w0: stopping experience collection (2100 times)
+[2026-06-06 21:19:51,957][262198] Updated weights for policy 0, policy_version 14985 (0.0012)
+[2026-06-06 21:19:52,088][262115] Signal inference workers to resume experience collection... (2100 times)
+[2026-06-06 21:19:52,089][262198] InferenceWorker_p0-w0: resuming experience collection (2100 times)
+[2026-06-06 21:19:52,920][261293] Fps is (10 sec: 3891.6, 60 sec: 3925.3, 300 sec: 4061.3). Total num frames: 7677952. Throughput: 0: 3859.4. Samples: 7682188. Policy #0 lag: (min: 2.0, avg: 22.3, max: 40.0)
+[2026-06-06 21:19:52,922][261293] Avg episode reward: [(0, '1449.019')]
+[2026-06-06 21:19:53,054][262198] Updated weights for policy 0, policy_version 14998 (0.0010)
+[2026-06-06 21:19:54,443][262198] Updated weights for policy 0, policy_version 15010 (0.0010)
+[2026-06-06 21:19:56,265][262198] Updated weights for policy 0, policy_version 15023 (0.0008)
+[2026-06-06 21:19:57,346][262198] Updated weights for policy 0, policy_version 15035 (0.0009)
+[2026-06-06 21:19:57,920][261293] Fps is (10 sec: 4301.0, 60 sec: 3891.2, 300 sec: 4054.4). Total num frames: 7698432. Throughput: 0: 3917.1. Samples: 7706518. Policy #0 lag: (min: 2.0, avg: 22.3, max: 40.0)
+[2026-06-06 21:19:57,921][261293] Avg episode reward: [(0, '1439.351')]
+[2026-06-06 21:19:58,648][262198] Updated weights for policy 0, policy_version 15046 (0.0008)
+[2026-06-06 21:20:00,398][262198] Updated weights for policy 0, policy_version 15057 (0.0009)
+[2026-06-06 21:20:02,143][262198] Updated weights for policy 0, policy_version 15072 (0.0010)
+[2026-06-06 21:20:02,921][261293] Fps is (10 sec: 4300.5, 60 sec: 3925.3, 300 sec: 4061.3). Total num frames: 7720960. Throughput: 0: 3970.8. Samples: 7719422. Policy #0 lag: (min: 2.0, avg: 22.3, max: 40.0)
+[2026-06-06 21:20:02,923][261293] Avg episode reward: [(0, '1523.481')]
+[2026-06-06 21:20:03,877][262198] Updated weights for policy 0, policy_version 15085 (0.0009)
+[2026-06-06 21:20:04,854][262198] Updated weights for policy 0, policy_version 15098 (0.0012)
+[2026-06-06 21:20:06,170][262198] Updated weights for policy 0, policy_version 15108 (0.0012)
+[2026-06-06 21:20:07,690][262198] Updated weights for policy 0, policy_version 15118 (0.0013)
+[2026-06-06 21:20:07,921][261293] Fps is (10 sec: 4300.6, 60 sec: 3925.4, 300 sec: 4061.3). Total num frames: 7741440. Throughput: 0: 4002.9. Samples: 7745514. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 21:20:07,923][261293] Avg episode reward: [(0, '1640.836')]
+[2026-06-06 21:20:08,066][262115] Saving new best policy, reward=1640.836!
+[2026-06-06 21:20:08,514][262198] Updated weights for policy 0, policy_version 15128 (0.0011)
+[2026-06-06 21:20:10,364][262198] Updated weights for policy 0, policy_version 15138 (0.0014)
+[2026-06-06 21:20:12,061][262198] Updated weights for policy 0, policy_version 15148 (0.0012)
+[2026-06-06 21:20:12,920][261293] Fps is (10 sec: 3891.5, 60 sec: 3959.5, 300 sec: 4040.5). Total num frames: 7759872. Throughput: 0: 3979.6. Samples: 7766844. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 21:20:12,922][261293] Avg episode reward: [(0, '1496.456')]
+[2026-06-06 21:20:13,022][262198] Updated weights for policy 0, policy_version 15158 (0.0013)
+[2026-06-06 21:20:15,075][262198] Updated weights for policy 0, policy_version 15169 (0.0011)
+[2026-06-06 21:20:16,916][262198] Updated weights for policy 0, policy_version 15179 (0.0011)
+[2026-06-06 21:20:17,859][262198] Updated weights for policy 0, policy_version 15189 (0.0012)
+[2026-06-06 21:20:17,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3925.3, 300 sec: 4033.5). Total num frames: 7776256. Throughput: 0: 3940.5. Samples: 7776992. Policy #0 lag: (min: 0.0, avg: 19.0, max: 36.0)
+[2026-06-06 21:20:17,923][261293] Avg episode reward: [(0, '1372.705')]
+[2026-06-06 21:20:19,539][262198] Updated weights for policy 0, policy_version 15199 (0.0011)
+[2026-06-06 21:20:21,705][262198] Updated weights for policy 0, policy_version 15209 (0.0012)
+[2026-06-06 21:20:22,920][261293] Fps is (10 sec: 3481.6, 60 sec: 3891.2, 300 sec: 4026.6). Total num frames: 7794688. Throughput: 0: 3860.7. Samples: 7797828. Policy #0 lag: (min: 0.0, avg: 19.0, max: 36.0)
+[2026-06-06 21:20:22,922][261293] Avg episode reward: [(0, '1588.389')]
+[2026-06-06 21:20:23,231][262198] Updated weights for policy 0, policy_version 15225 (0.0012)
+[2026-06-06 21:20:24,638][262198] Updated weights for policy 0, policy_version 15236 (0.0011)
+[2026-06-06 21:20:26,490][262198] Updated weights for policy 0, policy_version 15249 (0.0009)
+[2026-06-06 21:20:27,288][262198] Updated weights for policy 0, policy_version 15260 (0.0009)
+[2026-06-06 21:20:27,920][261293] Fps is (10 sec: 3686.6, 60 sec: 3857.1, 300 sec: 4026.6). Total num frames: 7813120. Throughput: 0: 3871.3. Samples: 7821782. Policy #0 lag: (min: 2.0, avg: 20.2, max: 38.0)
+[2026-06-06 21:20:27,922][261293] Avg episode reward: [(0, '1537.428')]
+[2026-06-06 21:20:29,501][262198] Updated weights for policy 0, policy_version 15273 (0.0009)
+[2026-06-06 21:20:30,912][262198] Updated weights for policy 0, policy_version 15289 (0.0010)
+[2026-06-06 21:20:32,166][262198] Updated weights for policy 0, policy_version 15300 (0.0011)
+[2026-06-06 21:20:32,921][261293] Fps is (10 sec: 4095.9, 60 sec: 3925.4, 300 sec: 4026.6). Total num frames: 7835648. Throughput: 0: 3924.3. Samples: 7835100. Policy #0 lag: (min: 2.0, avg: 20.2, max: 38.0)
+[2026-06-06 21:20:32,922][261293] Avg episode reward: [(0, '1609.073')]
+[2026-06-06 21:20:34,149][262198] Updated weights for policy 0, policy_version 15310 (0.0014)
+[2026-06-06 21:20:35,363][262198] Updated weights for policy 0, policy_version 15321 (0.0009)
+[2026-06-06 21:20:36,539][262198] Updated weights for policy 0, policy_version 15331 (0.0011)
+[2026-06-06 21:20:37,912][262115] Signal inference workers to stop experience collection... (2150 times)
+[2026-06-06 21:20:37,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3857.0, 300 sec: 4012.7). Total num frames: 7852032. Throughput: 0: 3904.6. Samples: 7857898. Policy #0 lag: (min: 3.0, avg: 20.5, max: 41.0)
+[2026-06-06 21:20:37,924][261293] Avg episode reward: [(0, '1547.969')]
+[2026-06-06 21:20:37,936][262198] InferenceWorker_p0-w0: stopping experience collection (2150 times)
+[2026-06-06 21:20:38,033][262115] Signal inference workers to resume experience collection... (2150 times)
+[2026-06-06 21:20:38,034][262198] InferenceWorker_p0-w0: resuming experience collection (2150 times)
+[2026-06-06 21:20:38,853][262198] Updated weights for policy 0, policy_version 15348 (0.0010)
+[2026-06-06 21:20:40,361][262198] Updated weights for policy 0, policy_version 15358 (0.0014)
+[2026-06-06 21:20:41,244][262198] Updated weights for policy 0, policy_version 15368 (0.0012)
+[2026-06-06 21:20:42,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3891.2, 300 sec: 4012.7). Total num frames: 7872512. Throughput: 0: 3880.4. Samples: 7881136. Policy #0 lag: (min: 3.0, avg: 20.5, max: 41.0)
+[2026-06-06 21:20:42,923][261293] Avg episode reward: [(0, '1532.568')]
+[2026-06-06 21:20:43,471][262198] Updated weights for policy 0, policy_version 15383 (0.0009)
+[2026-06-06 21:20:44,916][262198] Updated weights for policy 0, policy_version 15395 (0.0009)
+[2026-06-06 21:20:46,923][262198] Updated weights for policy 0, policy_version 15409 (0.0008)
+[2026-06-06 21:20:47,921][261293] Fps is (10 sec: 4300.9, 60 sec: 3993.6, 300 sec: 3998.8). Total num frames: 7895040. Throughput: 0: 3879.2. Samples: 7893984. Policy #0 lag: (min: 3.0, avg: 20.5, max: 41.0)
+[2026-06-06 21:20:47,923][261293] Avg episode reward: [(0, '1516.111')]
+[2026-06-06 21:20:48,086][262198] Updated weights for policy 0, policy_version 15421 (0.0009)
+[2026-06-06 21:20:49,637][262198] Updated weights for policy 0, policy_version 15433 (0.0010)
+[2026-06-06 21:20:50,737][262198] Updated weights for policy 0, policy_version 15447 (0.0008)
+[2026-06-06 21:20:51,991][262198] Updated weights for policy 0, policy_version 15457 (0.0010)
+[2026-06-06 21:20:52,921][261293] Fps is (10 sec: 4505.5, 60 sec: 3993.6, 300 sec: 4005.7). Total num frames: 7917568. Throughput: 0: 3897.6. Samples: 7920908. Policy #0 lag: (min: 4.0, avg: 22.2, max: 36.0)
+[2026-06-06 21:20:52,923][261293] Avg episode reward: [(0, '1483.598')]
+[2026-06-06 21:20:53,681][262198] Updated weights for policy 0, policy_version 15467 (0.0011)
+[2026-06-06 21:20:55,179][262198] Updated weights for policy 0, policy_version 15479 (0.0011)
+[2026-06-06 21:20:56,646][262198] Updated weights for policy 0, policy_version 15489 (0.0012)
+[2026-06-06 21:20:57,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3925.3, 300 sec: 3998.8). Total num frames: 7933952. Throughput: 0: 3924.9. Samples: 7943466. Policy #0 lag: (min: 4.0, avg: 22.2, max: 36.0)
+[2026-06-06 21:20:57,921][261293] Avg episode reward: [(0, '1531.270')]
+[2026-06-06 21:20:58,043][262198] Updated weights for policy 0, policy_version 15500 (0.0010)
+[2026-06-06 21:20:59,174][262198] Updated weights for policy 0, policy_version 15512 (0.0008)
+[2026-06-06 21:21:00,520][262198] Updated weights for policy 0, policy_version 15523 (0.0009)
+[2026-06-06 21:21:02,197][262198] Updated weights for policy 0, policy_version 15537 (0.0009)
+[2026-06-06 21:21:02,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3925.4, 300 sec: 4012.7). Total num frames: 7956480. Throughput: 0: 3992.8. Samples: 7956666. Policy #0 lag: (min: 6.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:21:02,923][261293] Avg episode reward: [(0, '1446.558')]
+[2026-06-06 21:21:03,733][262198] Updated weights for policy 0, policy_version 15548 (0.0010)
+[2026-06-06 21:21:05,079][262198] Updated weights for policy 0, policy_version 15559 (0.0010)
+[2026-06-06 21:21:06,710][262198] Updated weights for policy 0, policy_version 15573 (0.0010)
+[2026-06-06 21:21:07,625][262198] Updated weights for policy 0, policy_version 15583 (0.0009)
+[2026-06-06 21:21:07,921][261293] Fps is (10 sec: 4505.6, 60 sec: 3959.5, 300 sec: 4026.6). Total num frames: 7979008. Throughput: 0: 4098.9. Samples: 7982280. Policy #0 lag: (min: 6.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:21:07,923][261293] Avg episode reward: [(0, '1557.047')]
+[2026-06-06 21:21:09,291][262198] Updated weights for policy 0, policy_version 15595 (0.0010)
+[2026-06-06 21:21:10,558][262198] Updated weights for policy 0, policy_version 15606 (0.0010)
+[2026-06-06 21:21:11,523][262198] Updated weights for policy 0, policy_version 15616 (0.0009)
+[2026-06-06 21:21:12,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3993.6, 300 sec: 4019.6). Total num frames: 7999488. Throughput: 0: 4134.4. Samples: 8007830. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:21:12,922][261293] Avg episode reward: [(0, '1444.721')]
+[2026-06-06 21:21:13,254][262198] Updated weights for policy 0, policy_version 15626 (0.0009)
+[2026-06-06 21:21:14,622][262198] Updated weights for policy 0, policy_version 15637 (0.0010)
+[2026-06-06 21:21:15,679][262198] Updated weights for policy 0, policy_version 15647 (0.0011)
+[2026-06-06 21:21:17,613][262115] Signal inference workers to stop experience collection... (2200 times)
+[2026-06-06 21:21:17,638][262198] InferenceWorker_p0-w0: stopping experience collection (2200 times)
+[2026-06-06 21:21:17,709][262115] Signal inference workers to resume experience collection... (2200 times)
+[2026-06-06 21:21:17,709][262198] InferenceWorker_p0-w0: resuming experience collection (2200 times)
+[2026-06-06 21:21:17,711][262198] Updated weights for policy 0, policy_version 15660 (0.0011)
+[2026-06-06 21:21:17,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4027.8, 300 sec: 4005.8). Total num frames: 8017920. Throughput: 0: 4093.9. Samples: 8019326. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:21:17,923][261293] Avg episode reward: [(0, '1492.652')]
+[2026-06-06 21:21:18,930][262198] Updated weights for policy 0, policy_version 15672 (0.0010)
+[2026-06-06 21:21:21,467][262198] Updated weights for policy 0, policy_version 15688 (0.0011)
+[2026-06-06 21:21:22,420][262198] Updated weights for policy 0, policy_version 15698 (0.0012)
+[2026-06-06 21:21:22,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4061.8, 300 sec: 3998.8). Total num frames: 8038400. Throughput: 0: 4113.0. Samples: 8042984. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:21:22,923][261293] Avg episode reward: [(0, '1286.731')]
+[2026-06-06 21:21:24,404][262198] Updated weights for policy 0, policy_version 15708 (0.0014)
+[2026-06-06 21:21:26,795][262198] Updated weights for policy 0, policy_version 15720 (0.0014)
+[2026-06-06 21:21:27,840][262198] Updated weights for policy 0, policy_version 15731 (0.0013)
+[2026-06-06 21:21:27,921][261293] Fps is (10 sec: 3686.4, 60 sec: 4027.7, 300 sec: 3978.0). Total num frames: 8054784. Throughput: 0: 4002.4. Samples: 8061246. Policy #0 lag: (min: 7.0, avg: 25.5, max: 42.0)
+[2026-06-06 21:21:27,923][261293] Avg episode reward: [(0, '1425.717')]
+[2026-06-06 21:21:29,783][262198] Updated weights for policy 0, policy_version 15741 (0.0014)
+[2026-06-06 21:21:31,833][262198] Updated weights for policy 0, policy_version 15752 (0.0012)
+[2026-06-06 21:21:32,775][262198] Updated weights for policy 0, policy_version 15763 (0.0013)
+[2026-06-06 21:21:32,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3925.4, 300 sec: 3964.1). Total num frames: 8071168. Throughput: 0: 3934.7. Samples: 8071044. Policy #0 lag: (min: 7.0, avg: 25.5, max: 42.0)
+[2026-06-06 21:21:32,921][261293] Avg episode reward: [(0, '1420.224')]
+[2026-06-06 21:21:34,814][262198] Updated weights for policy 0, policy_version 15773 (0.0014)
+[2026-06-06 21:21:37,050][262198] Updated weights for policy 0, policy_version 15785 (0.0010)
+[2026-06-06 21:21:37,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3925.4, 300 sec: 3957.2). Total num frames: 8087552. Throughput: 0: 3769.3. Samples: 8090524. Policy #0 lag: (min: 2.0, avg: 18.7, max: 37.0)
+[2026-06-06 21:21:37,924][261293] Avg episode reward: [(0, '1470.917')]
+[2026-06-06 21:21:38,441][262198] Updated weights for policy 0, policy_version 15797 (0.0013)
+[2026-06-06 21:21:39,609][262198] Updated weights for policy 0, policy_version 15807 (0.0011)
+[2026-06-06 21:21:41,558][262198] Updated weights for policy 0, policy_version 15819 (0.0011)
+[2026-06-06 21:21:42,803][262198] Updated weights for policy 0, policy_version 15830 (0.0010)
+[2026-06-06 21:21:42,924][261293] Fps is (10 sec: 3480.4, 60 sec: 3891.0, 300 sec: 3936.3). Total num frames: 8105984. Throughput: 0: 3760.6. Samples: 8112708. Policy #0 lag: (min: 2.0, avg: 18.7, max: 37.0)
+[2026-06-06 21:21:42,927][261293] Avg episode reward: [(0, '1485.345')]
+[2026-06-06 21:21:44,010][262198] Updated weights for policy 0, policy_version 15840 (0.0009)
+[2026-06-06 21:21:45,962][262198] Updated weights for policy 0, policy_version 15852 (0.0011)
+[2026-06-06 21:21:47,061][262198] Updated weights for policy 0, policy_version 15864 (0.0010)
+[2026-06-06 21:21:47,921][261293] Fps is (10 sec: 3686.2, 60 sec: 3822.9, 300 sec: 3929.4). Total num frames: 8124416. Throughput: 0: 3732.4. Samples: 8124624. Policy #0 lag: (min: 9.0, avg: 22.9, max: 40.0)
+[2026-06-06 21:21:47,924][261293] Avg episode reward: [(0, '1343.186')]
+[2026-06-06 21:21:48,616][262198] Updated weights for policy 0, policy_version 15875 (0.0009)
+[2026-06-06 21:21:50,637][262198] Updated weights for policy 0, policy_version 15891 (0.0010)
+[2026-06-06 21:21:52,083][262198] Updated weights for policy 0, policy_version 15902 (0.0009)
+[2026-06-06 21:21:52,920][261293] Fps is (10 sec: 3892.5, 60 sec: 3788.8, 300 sec: 3936.3). Total num frames: 8144896. Throughput: 0: 3683.3. Samples: 8148030. Policy #0 lag: (min: 9.0, avg: 22.9, max: 40.0)
+[2026-06-06 21:21:52,921][261293] Avg episode reward: [(0, '1462.110')]
+[2026-06-06 21:21:53,842][262198] Updated weights for policy 0, policy_version 15913 (0.0010)
+[2026-06-06 21:21:55,033][262198] Updated weights for policy 0, policy_version 15927 (0.0010)
+[2026-06-06 21:21:56,337][262198] Updated weights for policy 0, policy_version 15938 (0.0010)
+[2026-06-06 21:21:57,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3857.0, 300 sec: 3950.2). Total num frames: 8165376. Throughput: 0: 3695.2. Samples: 8174116. Policy #0 lag: (min: 9.0, avg: 22.9, max: 40.0)
+[2026-06-06 21:21:57,923][261293] Avg episode reward: [(0, '1477.591')]
+[2026-06-06 21:21:58,070][262198] Updated weights for policy 0, policy_version 15950 (0.0011)
+[2026-06-06 21:21:59,581][262198] Updated weights for policy 0, policy_version 15961 (0.0010)
+[2026-06-06 21:22:01,253][262198] Updated weights for policy 0, policy_version 15974 (0.0010)
+[2026-06-06 21:22:02,738][262198] Updated weights for policy 0, policy_version 15989 (0.0011)
+[2026-06-06 21:22:02,921][261293] Fps is (10 sec: 4300.4, 60 sec: 3857.0, 300 sec: 3964.1). Total num frames: 8187904. Throughput: 0: 3709.1. Samples: 8186236. Policy #0 lag: (min: 4.0, avg: 21.0, max: 42.0)
+[2026-06-06 21:22:02,924][261293] Avg episode reward: [(0, '1495.664')]
+[2026-06-06 21:22:04,104][262198] Updated weights for policy 0, policy_version 16001 (0.0008)
+[2026-06-06 21:22:06,319][262198] Updated weights for policy 0, policy_version 16018 (0.0011)
+[2026-06-06 21:22:07,577][262198] Updated weights for policy 0, policy_version 16029 (0.0008)
+[2026-06-06 21:22:07,922][261293] Fps is (10 sec: 4300.6, 60 sec: 3822.9, 300 sec: 3964.1). Total num frames: 8208384. Throughput: 0: 3732.1. Samples: 8210932. Policy #0 lag: (min: 4.0, avg: 21.0, max: 42.0)
+[2026-06-06 21:22:07,927][261293] Avg episode reward: [(0, '1448.450')]
+[2026-06-06 21:22:09,176][262198] Updated weights for policy 0, policy_version 16040 (0.0008)
+[2026-06-06 21:22:09,412][262115] Signal inference workers to stop experience collection... (2250 times)
+[2026-06-06 21:22:09,437][262198] InferenceWorker_p0-w0: stopping experience collection (2250 times)
+[2026-06-06 21:22:09,535][262115] Signal inference workers to resume experience collection... (2250 times)
+[2026-06-06 21:22:09,536][262198] InferenceWorker_p0-w0: resuming experience collection (2250 times)
+[2026-06-06 21:22:10,433][262198] Updated weights for policy 0, policy_version 16054 (0.0010)
+[2026-06-06 21:22:11,528][262198] Updated weights for policy 0, policy_version 16064 (0.0009)
+[2026-06-06 21:22:12,921][261293] Fps is (10 sec: 3891.5, 60 sec: 3788.8, 300 sec: 3950.2). Total num frames: 8226816. Throughput: 0: 3918.3. Samples: 8237568. Policy #0 lag: (min: 8.0, avg: 19.4, max: 41.0)
+[2026-06-06 21:22:12,923][261293] Avg episode reward: [(0, '1612.953')]
+[2026-06-06 21:22:13,712][262198] Updated weights for policy 0, policy_version 16080 (0.0009)
+[2026-06-06 21:22:14,801][262198] Updated weights for policy 0, policy_version 16090 (0.0008)
+[2026-06-06 21:22:15,714][262198] Updated weights for policy 0, policy_version 16100 (0.0009)
+[2026-06-06 21:22:17,687][262198] Updated weights for policy 0, policy_version 16114 (0.0009)
+[2026-06-06 21:22:17,922][261293] Fps is (10 sec: 4300.6, 60 sec: 3891.1, 300 sec: 3964.1). Total num frames: 8251392. Throughput: 0: 3978.0. Samples: 8250060. Policy #0 lag: (min: 8.0, avg: 19.4, max: 41.0)
+[2026-06-06 21:22:17,925][261293] Avg episode reward: [(0, '1599.824')]
+[2026-06-06 21:22:18,852][262198] Updated weights for policy 0, policy_version 16124 (0.0010)
+[2026-06-06 21:22:20,472][262198] Updated weights for policy 0, policy_version 16134 (0.0008)
+[2026-06-06 21:22:21,547][262198] Updated weights for policy 0, policy_version 16148 (0.0010)
+[2026-06-06 21:22:22,921][261293] Fps is (10 sec: 4505.6, 60 sec: 3891.2, 300 sec: 3978.0). Total num frames: 8271872. Throughput: 0: 4118.0. Samples: 8275836. Policy #0 lag: (min: 11.0, avg: 24.9, max: 45.0)
+[2026-06-06 21:22:22,927][261293] Avg episode reward: [(0, '1416.585')]
+[2026-06-06 21:22:22,930][262198] Updated weights for policy 0, policy_version 16158 (0.0010)
+[2026-06-06 21:22:24,605][262198] Updated weights for policy 0, policy_version 16168 (0.0012)
+[2026-06-06 21:22:26,455][262198] Updated weights for policy 0, policy_version 16182 (0.0011)
+[2026-06-06 21:22:27,697][262198] Updated weights for policy 0, policy_version 16192 (0.0012)
+[2026-06-06 21:22:27,921][261293] Fps is (10 sec: 3891.5, 60 sec: 3925.3, 300 sec: 3957.1). Total num frames: 8290304. Throughput: 0: 4108.1. Samples: 8297560. Policy #0 lag: (min: 11.0, avg: 24.9, max: 45.0)
+[2026-06-06 21:22:27,926][261293] Avg episode reward: [(0, '1595.007')]
+[2026-06-06 21:22:30,017][262198] Updated weights for policy 0, policy_version 16204 (0.0011)
+[2026-06-06 21:22:31,416][262198] Updated weights for policy 0, policy_version 16216 (0.0013)
+[2026-06-06 21:22:32,920][261293] Fps is (10 sec: 3481.6, 60 sec: 3925.3, 300 sec: 3943.3). Total num frames: 8306688. Throughput: 0: 4064.9. Samples: 8307542. Policy #0 lag: (min: 11.0, avg: 24.9, max: 45.0)
+[2026-06-06 21:22:32,922][261293] Avg episode reward: [(0, '1511.477')]
+[2026-06-06 21:22:32,951][262198] Updated weights for policy 0, policy_version 16226 (0.0011)
+[2026-06-06 21:22:34,992][262198] Updated weights for policy 0, policy_version 16240 (0.0012)
+[2026-06-06 21:22:36,680][262198] Updated weights for policy 0, policy_version 16250 (0.0012)
+[2026-06-06 21:22:37,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3959.5, 300 sec: 3936.3). Total num frames: 8325120. Throughput: 0: 4021.4. Samples: 8328992. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:22:37,922][261293] Avg episode reward: [(0, '1436.275')]
+[2026-06-06 21:22:38,258][262198] Updated weights for policy 0, policy_version 16261 (0.0010)
+[2026-06-06 21:22:39,336][262198] Updated weights for policy 0, policy_version 16274 (0.0010)
+[2026-06-06 21:22:40,880][262198] Updated weights for policy 0, policy_version 16285 (0.0009)
+[2026-06-06 21:22:42,179][262198] Updated weights for policy 0, policy_version 16295 (0.0009)
+[2026-06-06 21:22:42,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3993.8, 300 sec: 3950.2). Total num frames: 8345600. Throughput: 0: 4010.1. Samples: 8354570. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:22:42,923][261293] Avg episode reward: [(0, '1494.313')]
+[2026-06-06 21:22:43,756][262198] Updated weights for policy 0, policy_version 16309 (0.0010)
+[2026-06-06 21:22:44,765][262198] Updated weights for policy 0, policy_version 16319 (0.0010)
+[2026-06-06 21:22:46,381][262198] Updated weights for policy 0, policy_version 16330 (0.0008)
+[2026-06-06 21:22:47,594][262198] Updated weights for policy 0, policy_version 16343 (0.0010)
+[2026-06-06 21:22:47,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4061.9, 300 sec: 3943.3). Total num frames: 8368128. Throughput: 0: 4019.6. Samples: 8367116. Policy #0 lag: (min: 6.0, avg: 23.5, max: 38.0)
+[2026-06-06 21:22:47,925][261293] Avg episode reward: [(0, '1654.309')]
+[2026-06-06 21:22:47,932][262115] Saving new best policy, reward=1654.309!
+[2026-06-06 21:22:49,128][262198] Updated weights for policy 0, policy_version 16356 (0.0009)
+[2026-06-06 21:22:50,849][262198] Updated weights for policy 0, policy_version 16370 (0.0009)
+[2026-06-06 21:22:52,344][262198] Updated weights for policy 0, policy_version 16381 (0.0010)
+[2026-06-06 21:22:52,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4061.8, 300 sec: 3943.3). Total num frames: 8388608. Throughput: 0: 4034.1. Samples: 8392464. Policy #0 lag: (min: 6.0, avg: 23.5, max: 38.0)
+[2026-06-06 21:22:52,925][261293] Avg episode reward: [(0, '1462.109')]
+[2026-06-06 21:22:53,906][262198] Updated weights for policy 0, policy_version 16392 (0.0011)
+[2026-06-06 21:22:54,756][262198] Updated weights for policy 0, policy_version 16404 (0.0010)
+[2026-06-06 21:22:56,406][262198] Updated weights for policy 0, policy_version 16414 (0.0009)
+[2026-06-06 21:22:57,920][261293] Fps is (10 sec: 3891.4, 60 sec: 4027.7, 300 sec: 3936.3). Total num frames: 8407040. Throughput: 0: 3986.0. Samples: 8416938. Policy #0 lag: (min: 6.0, avg: 23.5, max: 38.0)
+[2026-06-06 21:22:57,923][261293] Avg episode reward: [(0, '1503.345')]
+[2026-06-06 21:22:57,986][262115] Signal inference workers to stop experience collection... (2300 times)
+[2026-06-06 21:22:58,014][262198] InferenceWorker_p0-w0: stopping experience collection (2300 times)
+[2026-06-06 21:22:58,020][262115] Signal inference workers to resume experience collection... (2300 times)
+[2026-06-06 21:22:58,031][262198] InferenceWorker_p0-w0: resuming experience collection (2300 times)
+[2026-06-06 21:22:58,033][262198] Updated weights for policy 0, policy_version 16424 (0.0009)
+[2026-06-06 21:22:59,403][262198] Updated weights for policy 0, policy_version 16438 (0.0010)
+[2026-06-06 21:23:01,059][262198] Updated weights for policy 0, policy_version 16450 (0.0010)
+[2026-06-06 21:23:02,505][262198] Updated weights for policy 0, policy_version 16462 (0.0010)
+[2026-06-06 21:23:02,922][261293] Fps is (10 sec: 4095.9, 60 sec: 4027.7, 300 sec: 3943.3). Total num frames: 8429568. Throughput: 0: 3972.8. Samples: 8428832. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:23:02,925][261293] Avg episode reward: [(0, '1391.035')]
+[2026-06-06 21:23:03,558][262198] Updated weights for policy 0, policy_version 16472 (0.0011)
+[2026-06-06 21:23:04,888][262198] Updated weights for policy 0, policy_version 16482 (0.0009)
+[2026-06-06 21:23:06,681][262198] Updated weights for policy 0, policy_version 16496 (0.0011)
+[2026-06-06 21:23:07,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4027.8, 300 sec: 3936.3). Total num frames: 8450048. Throughput: 0: 3982.1. Samples: 8455030. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:23:07,922][261293] Avg episode reward: [(0, '1274.951')]
+[2026-06-06 21:23:08,017][262198] Updated weights for policy 0, policy_version 16506 (0.0010)
+[2026-06-06 21:23:09,241][262198] Updated weights for policy 0, policy_version 16517 (0.0009)
+[2026-06-06 21:23:10,369][262198] Updated weights for policy 0, policy_version 16531 (0.0009)
+[2026-06-06 21:23:11,761][262198] Updated weights for policy 0, policy_version 16541 (0.0009)
+[2026-06-06 21:23:12,921][261293] Fps is (10 sec: 4301.1, 60 sec: 4096.0, 300 sec: 3943.3). Total num frames: 8472576. Throughput: 0: 4087.8. Samples: 8481508. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:23:12,923][261293] Avg episode reward: [(0, '1558.812')]
+[2026-06-06 21:23:12,983][262198] Updated weights for policy 0, policy_version 16551 (0.0010)
+[2026-06-06 21:23:14,245][262198] Updated weights for policy 0, policy_version 16564 (0.0009)
+[2026-06-06 21:23:15,596][262198] Updated weights for policy 0, policy_version 16574 (0.0010)
+[2026-06-06 21:23:16,772][262198] Updated weights for policy 0, policy_version 16584 (0.0009)
+[2026-06-06 21:23:17,889][262198] Updated weights for policy 0, policy_version 16596 (0.0009)
+[2026-06-06 21:23:17,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4096.1, 300 sec: 3964.1). Total num frames: 8497152. Throughput: 0: 4148.4. Samples: 8494220. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:23:17,922][261293] Avg episode reward: [(0, '1629.540')]
+[2026-06-06 21:23:19,118][262198] Updated weights for policy 0, policy_version 16606 (0.0009)
+[2026-06-06 21:23:20,385][262198] Updated weights for policy 0, policy_version 16616 (0.0008)
+[2026-06-06 21:23:21,698][262198] Updated weights for policy 0, policy_version 16629 (0.0010)
+[2026-06-06 21:23:22,601][262198] Updated weights for policy 0, policy_version 16639 (0.0008)
+[2026-06-06 21:23:22,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4130.2, 300 sec: 3984.9). Total num frames: 8519680. Throughput: 0: 4289.0. Samples: 8521998. Policy #0 lag: (min: 0.0, avg: 21.9, max: 40.0)
+[2026-06-06 21:23:22,922][261293] Avg episode reward: [(0, '1412.331')]
+[2026-06-06 21:23:24,048][262198] Updated weights for policy 0, policy_version 16649 (0.0009)
+[2026-06-06 21:23:25,090][262198] Updated weights for policy 0, policy_version 16661 (0.0010)
+[2026-06-06 21:23:26,090][262198] Updated weights for policy 0, policy_version 16671 (0.0009)
+[2026-06-06 21:23:27,686][262198] Updated weights for policy 0, policy_version 16681 (0.0009)
+[2026-06-06 21:23:27,921][261293] Fps is (10 sec: 4505.6, 60 sec: 4198.4, 300 sec: 3991.9). Total num frames: 8542208. Throughput: 0: 4361.1. Samples: 8550818. Policy #0 lag: (min: 0.0, avg: 21.9, max: 40.0)
+[2026-06-06 21:23:27,924][261293] Avg episode reward: [(0, '1602.439')]
+[2026-06-06 21:23:28,775][262198] Updated weights for policy 0, policy_version 16696 (0.0009)
+[2026-06-06 21:23:30,193][262198] Updated weights for policy 0, policy_version 16706 (0.0008)
+[2026-06-06 21:23:31,573][262198] Updated weights for policy 0, policy_version 16718 (0.0009)
+[2026-06-06 21:23:32,816][262198] Updated weights for policy 0, policy_version 16729 (0.0009)
+[2026-06-06 21:23:32,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4300.8, 300 sec: 3998.8). Total num frames: 8564736. Throughput: 0: 4373.1. Samples: 8563904. Policy #0 lag: (min: 0.0, avg: 21.9, max: 40.0)
+[2026-06-06 21:23:32,925][261293] Avg episode reward: [(0, '1400.106')]
+[2026-06-06 21:23:34,236][262198] Updated weights for policy 0, policy_version 16740 (0.0010)
+[2026-06-06 21:23:35,599][262198] Updated weights for policy 0, policy_version 16752 (0.0009)
+[2026-06-06 21:23:36,715][262198] Updated weights for policy 0, policy_version 16762 (0.0010)
+[2026-06-06 21:23:37,865][262198] Updated weights for policy 0, policy_version 16772 (0.0009)
+[2026-06-06 21:23:37,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4369.0, 300 sec: 4012.7). Total num frames: 8587264. Throughput: 0: 4391.6. Samples: 8590088. Policy #0 lag: (min: 3.0, avg: 20.3, max: 37.0)
+[2026-06-06 21:23:37,923][261293] Avg episode reward: [(0, '1586.383')]
+[2026-06-06 21:23:39,371][262198] Updated weights for policy 0, policy_version 16784 (0.0009)
+[2026-06-06 21:23:40,578][262198] Updated weights for policy 0, policy_version 16794 (0.0011)
+[2026-06-06 21:23:41,918][262198] Updated weights for policy 0, policy_version 16804 (0.0010)
+[2026-06-06 21:23:42,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4369.1, 300 sec: 4005.8). Total num frames: 8607744. Throughput: 0: 4420.7. Samples: 8615868. Policy #0 lag: (min: 3.0, avg: 20.3, max: 37.0)
+[2026-06-06 21:23:42,926][261293] Avg episode reward: [(0, '1645.925')]
+[2026-06-06 21:23:43,374][262198] Updated weights for policy 0, policy_version 16818 (0.0011)
+[2026-06-06 21:23:44,511][262198] Updated weights for policy 0, policy_version 16828 (0.0010)
+[2026-06-06 21:23:46,249][262198] Updated weights for policy 0, policy_version 16838 (0.0010)
+[2026-06-06 21:23:47,469][262198] Updated weights for policy 0, policy_version 16848 (0.0012)
+[2026-06-06 21:23:47,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4335.0, 300 sec: 4019.6). Total num frames: 8628224. Throughput: 0: 4417.9. Samples: 8627636. Policy #0 lag: (min: 1.0, avg: 18.7, max: 36.0)
+[2026-06-06 21:23:47,922][261293] Avg episode reward: [(0, '1375.740')]
+[2026-06-06 21:23:48,806][262198] Updated weights for policy 0, policy_version 16858 (0.0009)
+[2026-06-06 21:23:50,052][262198] Updated weights for policy 0, policy_version 16868 (0.0010)
+[2026-06-06 21:23:51,049][262115] Signal inference workers to stop experience collection... (2350 times)
+[2026-06-06 21:23:51,076][262198] InferenceWorker_p0-w0: stopping experience collection (2350 times)
+[2026-06-06 21:23:51,184][262115] Signal inference workers to resume experience collection... (2350 times)
+[2026-06-06 21:23:51,185][262198] InferenceWorker_p0-w0: resuming experience collection (2350 times)
+[2026-06-06 21:23:51,743][262198] Updated weights for policy 0, policy_version 16882 (0.0010)
+[2026-06-06 21:23:52,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4335.0, 300 sec: 4012.7). Total num frames: 8648704. Throughput: 0: 4365.1. Samples: 8651458. Policy #0 lag: (min: 1.0, avg: 18.7, max: 36.0)
+[2026-06-06 21:23:52,923][261293] Avg episode reward: [(0, '1462.826')]
+[2026-06-06 21:23:52,926][262198] Updated weights for policy 0, policy_version 16892 (0.0009)
+[2026-06-06 21:23:54,339][262198] Updated weights for policy 0, policy_version 16902 (0.0010)
+[2026-06-06 21:23:55,646][262198] Updated weights for policy 0, policy_version 16914 (0.0009)
+[2026-06-06 21:23:57,456][262198] Updated weights for policy 0, policy_version 16925 (0.0011)
+[2026-06-06 21:23:57,920][261293] Fps is (10 sec: 3891.2, 60 sec: 4334.9, 300 sec: 4005.7). Total num frames: 8667136. Throughput: 0: 4298.5. Samples: 8674942. Policy #0 lag: (min: 1.0, avg: 18.7, max: 36.0)
+[2026-06-06 21:23:57,922][261293] Avg episode reward: [(0, '1686.999')]
+[2026-06-06 21:23:57,927][262115] Saving new best policy, reward=1686.999!
+[2026-06-06 21:23:58,686][262198] Updated weights for policy 0, policy_version 16936 (0.0009)
+[2026-06-06 21:24:00,855][262198] Updated weights for policy 0, policy_version 16952 (0.0011)
+[2026-06-06 21:24:02,249][262198] Updated weights for policy 0, policy_version 16962 (0.0010)
+[2026-06-06 21:24:02,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4300.8, 300 sec: 4005.8). Total num frames: 8687616. Throughput: 0: 4284.7. Samples: 8687032. Policy #0 lag: (min: 3.0, avg: 20.2, max: 36.0)
+[2026-06-06 21:24:02,925][261293] Avg episode reward: [(0, '1445.322')]
+[2026-06-06 21:24:03,741][262198] Updated weights for policy 0, policy_version 16973 (0.0011)
+[2026-06-06 21:24:05,124][262198] Updated weights for policy 0, policy_version 16984 (0.0011)
+[2026-06-06 21:24:06,323][262198] Updated weights for policy 0, policy_version 16994 (0.0010)
+[2026-06-06 21:24:07,903][262198] Updated weights for policy 0, policy_version 17006 (0.0009)
+[2026-06-06 21:24:07,921][261293] Fps is (10 sec: 3891.0, 60 sec: 4266.7, 300 sec: 4012.7). Total num frames: 8706048. Throughput: 0: 4195.9. Samples: 8710814. Policy #0 lag: (min: 3.0, avg: 20.2, max: 36.0)
+[2026-06-06 21:24:07,923][261293] Avg episode reward: [(0, '1471.007')]
+[2026-06-06 21:24:09,414][262198] Updated weights for policy 0, policy_version 17017 (0.0011)
+[2026-06-06 21:24:10,522][262198] Updated weights for policy 0, policy_version 17027 (0.0009)
+[2026-06-06 21:24:11,846][262198] Updated weights for policy 0, policy_version 17038 (0.0009)
+[2026-06-06 21:24:12,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4232.5, 300 sec: 4019.6). Total num frames: 8726528. Throughput: 0: 4089.3. Samples: 8734840. Policy #0 lag: (min: 6.0, avg: 20.4, max: 35.0)
+[2026-06-06 21:24:12,924][261293] Avg episode reward: [(0, '1486.222')]
+[2026-06-06 21:24:13,349][262198] Updated weights for policy 0, policy_version 17050 (0.0010)
+[2026-06-06 21:24:14,567][262198] Updated weights for policy 0, policy_version 17060 (0.0008)
+[2026-06-06 21:24:16,031][262198] Updated weights for policy 0, policy_version 17074 (0.0010)
+[2026-06-06 21:24:17,196][262198] Updated weights for policy 0, policy_version 17084 (0.0009)
+[2026-06-06 21:24:17,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4198.4, 300 sec: 4026.6). Total num frames: 8749056. Throughput: 0: 4108.1. Samples: 8748766. Policy #0 lag: (min: 6.0, avg: 20.4, max: 35.0)
+[2026-06-06 21:24:17,922][261293] Avg episode reward: [(0, '1422.235')]
+[2026-06-06 21:24:18,425][262198] Updated weights for policy 0, policy_version 17094 (0.0009)
+[2026-06-06 21:24:19,756][262198] Updated weights for policy 0, policy_version 17105 (0.0009)
+[2026-06-06 21:24:21,249][262198] Updated weights for policy 0, policy_version 17117 (0.0010)
+[2026-06-06 21:24:22,405][262198] Updated weights for policy 0, policy_version 17127 (0.0009)
+[2026-06-06 21:24:22,921][261293] Fps is (10 sec: 4301.0, 60 sec: 4164.2, 300 sec: 4026.6). Total num frames: 8769536. Throughput: 0: 4088.8. Samples: 8774084. Policy #0 lag: (min: 6.0, avg: 20.4, max: 35.0)
+[2026-06-06 21:24:22,924][261293] Avg episode reward: [(0, '1484.349')]
+[2026-06-06 21:24:23,675][262198] Updated weights for policy 0, policy_version 17139 (0.0012)
+[2026-06-06 21:24:25,482][262198] Updated weights for policy 0, policy_version 17149 (0.0010)
+[2026-06-06 21:24:26,524][262198] Updated weights for policy 0, policy_version 17159 (0.0009)
+[2026-06-06 21:24:27,792][262198] Updated weights for policy 0, policy_version 17169 (0.0008)
+[2026-06-06 21:24:27,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4164.3, 300 sec: 4040.5). Total num frames: 8792064. Throughput: 0: 4050.2. Samples: 8798126. Policy #0 lag: (min: 3.0, avg: 20.4, max: 35.0)
+[2026-06-06 21:24:27,924][261293] Avg episode reward: [(0, '1550.983')]
+[2026-06-06 21:24:29,315][262198] Updated weights for policy 0, policy_version 17181 (0.0010)
+[2026-06-06 21:24:30,468][262198] Updated weights for policy 0, policy_version 17192 (0.0010)
+[2026-06-06 21:24:31,807][262198] Updated weights for policy 0, policy_version 17203 (0.0009)
+[2026-06-06 21:24:32,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4130.2, 300 sec: 4040.5). Total num frames: 8812544. Throughput: 0: 4073.2. Samples: 8810928. Policy #0 lag: (min: 3.0, avg: 20.4, max: 35.0)
+[2026-06-06 21:24:32,921][261293] Avg episode reward: [(0, '1410.349')]
+[2026-06-06 21:24:33,304][262198] Updated weights for policy 0, policy_version 17215 (0.0009)
+[2026-06-06 21:24:34,703][262198] Updated weights for policy 0, policy_version 17226 (0.0009)
+[2026-06-06 21:24:35,708][262198] Updated weights for policy 0, policy_version 17236 (0.0008)
+[2026-06-06 21:24:37,016][262198] Updated weights for policy 0, policy_version 17248 (0.0010)
+[2026-06-06 21:24:37,921][261293] Fps is (10 sec: 3891.3, 60 sec: 4061.9, 300 sec: 4040.5). Total num frames: 8830976. Throughput: 0: 4119.4. Samples: 8836832. Policy #0 lag: (min: 2.0, avg: 19.8, max: 37.0)
+[2026-06-06 21:24:37,923][261293] Avg episode reward: [(0, '1446.231')]
+[2026-06-06 21:24:38,693][262198] Updated weights for policy 0, policy_version 17260 (0.0011)
+[2026-06-06 21:24:39,889][262115] Signal inference workers to stop experience collection... (2400 times)
+[2026-06-06 21:24:39,913][262198] InferenceWorker_p0-w0: stopping experience collection (2400 times)
+[2026-06-06 21:24:40,036][262115] Signal inference workers to resume experience collection... (2400 times)
+[2026-06-06 21:24:40,037][262198] InferenceWorker_p0-w0: resuming experience collection (2400 times)
+[2026-06-06 21:24:40,039][262198] Updated weights for policy 0, policy_version 17272 (0.0009)
+[2026-06-06 21:24:41,826][262198] Updated weights for policy 0, policy_version 17284 (0.0010)
+[2026-06-06 21:24:42,874][262198] Updated weights for policy 0, policy_version 17295 (0.0009)
+[2026-06-06 21:24:42,920][261293] Fps is (10 sec: 4096.1, 60 sec: 4096.0, 300 sec: 4061.3). Total num frames: 8853504. Throughput: 0: 4169.1. Samples: 8862552. Policy #0 lag: (min: 2.0, avg: 19.8, max: 37.0)
+[2026-06-06 21:24:42,924][261293] Avg episode reward: [(0, '1544.207')]
+[2026-06-06 21:24:44,054][262198] Updated weights for policy 0, policy_version 17306 (0.0008)
+[2026-06-06 21:24:45,638][262198] Updated weights for policy 0, policy_version 17318 (0.0008)
+[2026-06-06 21:24:46,671][262198] Updated weights for policy 0, policy_version 17330 (0.0009)
+[2026-06-06 21:24:47,921][261293] Fps is (10 sec: 4710.1, 60 sec: 4164.2, 300 sec: 4068.2). Total num frames: 8878080. Throughput: 0: 4216.3. Samples: 8876766. Policy #0 lag: (min: 2.0, avg: 19.8, max: 37.0)
+[2026-06-06 21:24:47,924][261293] Avg episode reward: [(0, '1510.703')]
+[2026-06-06 21:24:48,139][262198] Updated weights for policy 0, policy_version 17342 (0.0010)
+[2026-06-06 21:24:49,550][262198] Updated weights for policy 0, policy_version 17354 (0.0010)
+[2026-06-06 21:24:50,935][262198] Updated weights for policy 0, policy_version 17366 (0.0009)
+[2026-06-06 21:24:52,417][262198] Updated weights for policy 0, policy_version 17377 (0.0008)
+[2026-06-06 21:24:52,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4164.3, 300 sec: 4068.2). Total num frames: 8898560. Throughput: 0: 4272.6. Samples: 8903080. Policy #0 lag: (min: 4.0, avg: 24.1, max: 42.0)
+[2026-06-06 21:24:52,922][261293] Avg episode reward: [(0, '1551.682')]
+[2026-06-06 21:24:53,549][262198] Updated weights for policy 0, policy_version 17389 (0.0009)
+[2026-06-06 21:24:54,717][262198] Updated weights for policy 0, policy_version 17402 (0.0010)
+[2026-06-06 21:24:56,458][262198] Updated weights for policy 0, policy_version 17413 (0.0009)
+[2026-06-06 21:24:57,291][262198] Updated weights for policy 0, policy_version 17424 (0.0009)
+[2026-06-06 21:24:57,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4232.5, 300 sec: 4068.2). Total num frames: 8921088. Throughput: 0: 4341.3. Samples: 8930196. Policy #0 lag: (min: 4.0, avg: 24.1, max: 42.0)
+[2026-06-06 21:24:57,925][261293] Avg episode reward: [(0, '1645.166')]
+[2026-06-06 21:24:58,681][262198] Updated weights for policy 0, policy_version 17435 (0.0010)
+[2026-06-06 21:25:00,129][262198] Updated weights for policy 0, policy_version 17446 (0.0009)
+[2026-06-06 21:25:01,973][262198] Updated weights for policy 0, policy_version 17461 (0.0009)
+[2026-06-06 21:25:02,790][262198] Updated weights for policy 0, policy_version 17472 (0.0009)
+[2026-06-06 21:25:02,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4300.8, 300 sec: 4082.1). Total num frames: 8945664. Throughput: 0: 4331.6. Samples: 8943686. Policy #0 lag: (min: 7.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:25:02,922][261293] Avg episode reward: [(0, '1637.392')]
+[2026-06-06 21:25:04,555][262198] Updated weights for policy 0, policy_version 17482 (0.0009)
+[2026-06-06 21:25:06,199][262198] Updated weights for policy 0, policy_version 17496 (0.0011)
+[2026-06-06 21:25:07,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4266.7, 300 sec: 4075.2). Total num frames: 8962048. Throughput: 0: 4297.1. Samples: 8967454. Policy #0 lag: (min: 7.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:25:07,922][261293] Avg episode reward: [(0, '1494.333')]
+[2026-06-06 21:25:08,095][262198] Updated weights for policy 0, policy_version 17506 (0.0010)
+[2026-06-06 21:25:09,503][262198] Updated weights for policy 0, policy_version 17518 (0.0012)
+[2026-06-06 21:25:11,102][262198] Updated weights for policy 0, policy_version 17532 (0.0010)
+[2026-06-06 21:25:12,920][261293] Fps is (10 sec: 3481.7, 60 sec: 4232.6, 300 sec: 4082.1). Total num frames: 8980480. Throughput: 0: 4241.0. Samples: 8988968. Policy #0 lag: (min: 7.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:25:12,922][261293] Avg episode reward: [(0, '1317.951')]
+[2026-06-06 21:25:13,067][262198] Updated weights for policy 0, policy_version 17542 (0.0012)
+[2026-06-06 21:25:15,086][262198] Updated weights for policy 0, policy_version 17554 (0.0012)
+[2026-06-06 21:25:15,674][262198] Updated weights for policy 0, policy_version 17564 (0.0013)
+[2026-06-06 21:25:17,845][262198] Updated weights for policy 0, policy_version 17574 (0.0014)
+[2026-06-06 21:25:17,921][261293] Fps is (10 sec: 3686.3, 60 sec: 4164.3, 300 sec: 4082.1). Total num frames: 8998912. Throughput: 0: 4185.5. Samples: 8999276. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 21:25:17,923][261293] Avg episode reward: [(0, '1434.366')]
+[2026-06-06 21:25:20,190][262198] Updated weights for policy 0, policy_version 17588 (0.0013)
+[2026-06-06 21:25:21,278][262198] Updated weights for policy 0, policy_version 17600 (0.0010)
+[2026-06-06 21:25:22,920][261293] Fps is (10 sec: 3481.6, 60 sec: 4096.0, 300 sec: 4075.2). Total num frames: 9015296. Throughput: 0: 4057.0. Samples: 9019396. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 21:25:22,922][261293] Avg episode reward: [(0, '1530.923')]
+[2026-06-06 21:25:23,093][262198] Updated weights for policy 0, policy_version 17610 (0.0011)
+[2026-06-06 21:25:24,355][262198] Updated weights for policy 0, policy_version 17621 (0.0013)
+[2026-06-06 21:25:25,429][262198] Updated weights for policy 0, policy_version 17632 (0.0014)
+[2026-06-06 21:25:27,571][262198] Updated weights for policy 0, policy_version 17642 (0.0013)
+[2026-06-06 21:25:27,924][261293] Fps is (10 sec: 3685.2, 60 sec: 4061.7, 300 sec: 4068.2). Total num frames: 9035776. Throughput: 0: 3987.6. Samples: 9042010. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:25:27,928][261293] Avg episode reward: [(0, '1432.050')]
+[2026-06-06 21:25:28,594][262115] Signal inference workers to stop experience collection... (2450 times)
+[2026-06-06 21:25:28,626][262198] InferenceWorker_p0-w0: stopping experience collection (2450 times)
+[2026-06-06 21:25:28,626][262115] Signal inference workers to resume experience collection... (2450 times)
+[2026-06-06 21:25:28,629][262198] Updated weights for policy 0, policy_version 17652 (0.0014)
+[2026-06-06 21:25:28,641][262198] InferenceWorker_p0-w0: resuming experience collection (2450 times)
+[2026-06-06 21:25:29,765][262198] Updated weights for policy 0, policy_version 17662 (0.0009)
+[2026-06-06 21:25:31,131][262198] Updated weights for policy 0, policy_version 17672 (0.0012)
+[2026-06-06 21:25:32,920][261293] Fps is (10 sec: 3686.3, 60 sec: 3993.6, 300 sec: 4068.2). Total num frames: 9052160. Throughput: 0: 3935.9. Samples: 9053882. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:25:32,921][261293] Avg episode reward: [(0, '1365.389')]
+[2026-06-06 21:25:33,103][262198] Updated weights for policy 0, policy_version 17682 (0.0009)
+[2026-06-06 21:25:33,555][262198] Updated weights for policy 0, policy_version 17692 (0.0011)
+[2026-06-06 21:25:35,459][262198] Updated weights for policy 0, policy_version 17702 (0.0010)
+[2026-06-06 21:25:36,407][262198] Updated weights for policy 0, policy_version 17712 (0.0012)
+[2026-06-06 21:25:37,921][261293] Fps is (10 sec: 3892.4, 60 sec: 4061.8, 300 sec: 4075.2). Total num frames: 9074688. Throughput: 0: 3855.7. Samples: 9076586. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:25:37,922][261293] Avg episode reward: [(0, '1595.937')]
+[2026-06-06 21:25:38,564][262198] Updated weights for policy 0, policy_version 17725 (0.0015)
+[2026-06-06 21:25:40,151][262198] Updated weights for policy 0, policy_version 17735 (0.0012)
+[2026-06-06 21:25:41,963][262198] Updated weights for policy 0, policy_version 17745 (0.0013)
+[2026-06-06 21:25:42,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3959.5, 300 sec: 4054.3). Total num frames: 9091072. Throughput: 0: 3720.4. Samples: 9097612. Policy #0 lag: (min: 0.0, avg: 19.4, max: 33.0)
+[2026-06-06 21:25:42,922][261293] Avg episode reward: [(0, '1453.661')]
+[2026-06-06 21:25:43,119][262198] Updated weights for policy 0, policy_version 17759 (0.0013)
+[2026-06-06 21:25:44,887][262198] Updated weights for policy 0, policy_version 17770 (0.0010)
+[2026-06-06 21:25:46,162][262198] Updated weights for policy 0, policy_version 17782 (0.0011)
+[2026-06-06 21:25:47,082][262198] Updated weights for policy 0, policy_version 17792 (0.0009)
+[2026-06-06 21:25:47,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3891.2, 300 sec: 4047.4). Total num frames: 9111552. Throughput: 0: 3691.1. Samples: 9109784. Policy #0 lag: (min: 0.0, avg: 19.4, max: 33.0)
+[2026-06-06 21:25:47,922][261293] Avg episode reward: [(0, '1516.542')]
+[2026-06-06 21:25:48,464][262198] Updated weights for policy 0, policy_version 17802 (0.0009)
+[2026-06-06 21:25:49,498][262198] Updated weights for policy 0, policy_version 17812 (0.0010)
+[2026-06-06 21:25:50,760][262198] Updated weights for policy 0, policy_version 17823 (0.0009)
+[2026-06-06 21:25:52,650][262198] Updated weights for policy 0, policy_version 17833 (0.0012)
+[2026-06-06 21:25:52,920][261293] Fps is (10 sec: 4096.0, 60 sec: 3891.2, 300 sec: 4061.3). Total num frames: 9132032. Throughput: 0: 3733.9. Samples: 9135478. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:25:52,922][261293] Avg episode reward: [(0, '1476.215')]
+[2026-06-06 21:25:54,197][262198] Updated weights for policy 0, policy_version 17846 (0.0010)
+[2026-06-06 21:25:55,544][262198] Updated weights for policy 0, policy_version 17857 (0.0010)
+[2026-06-06 21:25:56,331][262198] Updated weights for policy 0, policy_version 17867 (0.0008)
+[2026-06-06 21:25:57,714][262198] Updated weights for policy 0, policy_version 17881 (0.0010)
+[2026-06-06 21:25:57,920][261293] Fps is (10 sec: 4505.6, 60 sec: 3925.4, 300 sec: 4068.2). Total num frames: 9156608. Throughput: 0: 3848.2. Samples: 9162138. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:25:57,922][261293] Avg episode reward: [(0, '1567.837')]
+[2026-06-06 21:25:58,818][262198] Updated weights for policy 0, policy_version 17891 (0.0008)
+[2026-06-06 21:26:00,004][262198] Updated weights for policy 0, policy_version 17901 (0.0010)
+[2026-06-06 21:26:00,798][262198] Updated weights for policy 0, policy_version 17911 (0.0008)
+[2026-06-06 21:26:02,081][262198] Updated weights for policy 0, policy_version 17921 (0.0009)
+[2026-06-06 21:26:02,921][261293] Fps is (10 sec: 4710.3, 60 sec: 3891.2, 300 sec: 4068.2). Total num frames: 9179136. Throughput: 0: 3939.8. Samples: 9176566. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:26:02,927][261293] Avg episode reward: [(0, '1628.986')]
+[2026-06-06 21:26:03,088][262198] Updated weights for policy 0, policy_version 17932 (0.0011)
+[2026-06-06 21:26:04,175][262198] Updated weights for policy 0, policy_version 17942 (0.0010)
+[2026-06-06 21:26:04,926][262198] Updated weights for policy 0, policy_version 17952 (0.0010)
+[2026-06-06 21:26:06,445][262198] Updated weights for policy 0, policy_version 17962 (0.0009)
+[2026-06-06 21:26:07,446][262198] Updated weights for policy 0, policy_version 17974 (0.0010)
+[2026-06-06 21:26:07,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4061.8, 300 sec: 4089.1). Total num frames: 9205760. Throughput: 0: 4138.3. Samples: 9205622. Policy #0 lag: (min: 6.0, avg: 20.1, max: 39.0)
+[2026-06-06 21:26:07,922][261293] Avg episode reward: [(0, '1283.510')]
+[2026-06-06 21:26:08,034][262198] Updated weights for policy 0, policy_version 17984 (0.0009)
+[2026-06-06 21:26:09,523][262198] Updated weights for policy 0, policy_version 17994 (0.0008)
+[2026-06-06 21:26:10,443][262198] Updated weights for policy 0, policy_version 18004 (0.0009)
+[2026-06-06 21:26:11,131][262198] Updated weights for policy 0, policy_version 18014 (0.0008)
+[2026-06-06 21:26:12,791][262198] Updated weights for policy 0, policy_version 18025 (0.0011)
+[2026-06-06 21:26:12,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4164.2, 300 sec: 4109.9). Total num frames: 9230336. Throughput: 0: 4330.2. Samples: 9236858. Policy #0 lag: (min: 6.0, avg: 20.1, max: 39.0)
+[2026-06-06 21:26:12,924][261293] Avg episode reward: [(0, '1564.883')]
+[2026-06-06 21:26:13,815][262198] Updated weights for policy 0, policy_version 18036 (0.0010)
+[2026-06-06 21:26:14,402][262198] Updated weights for policy 0, policy_version 18046 (0.0009)
+[2026-06-06 21:26:15,668][262198] Updated weights for policy 0, policy_version 18056 (0.0009)
+[2026-06-06 21:26:16,849][262198] Updated weights for policy 0, policy_version 18066 (0.0009)
+[2026-06-06 21:26:17,210][262198] Updated weights for policy 0, policy_version 18076 (0.0007)
+[2026-06-06 21:26:17,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4300.8, 300 sec: 4130.7). Total num frames: 9256960. Throughput: 0: 4403.5. Samples: 9252040. Policy #0 lag: (min: 7.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:26:17,923][261293] Avg episode reward: [(0, '1527.993')]
+[2026-06-06 21:26:18,710][262198] Updated weights for policy 0, policy_version 18086 (0.0009)
+[2026-06-06 21:26:19,231][262198] Updated weights for policy 0, policy_version 18096 (0.0008)
+[2026-06-06 21:26:20,141][262198] Updated weights for policy 0, policy_version 18106 (0.0010)
+[2026-06-06 21:26:21,164][262198] Updated weights for policy 0, policy_version 18116 (0.0009)
+[2026-06-06 21:26:22,132][262198] Updated weights for policy 0, policy_version 18128 (0.0010)
+[2026-06-06 21:26:22,921][261293] Fps is (10 sec: 5529.5, 60 sec: 4505.5, 300 sec: 4172.4). Total num frames: 9285632. Throughput: 0: 4636.8. Samples: 9285246. Policy #0 lag: (min: 7.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:26:22,924][261293] Avg episode reward: [(0, '1384.127')]
+[2026-06-06 21:26:23,738][262198] Updated weights for policy 0, policy_version 18141 (0.0009)
+[2026-06-06 21:26:24,724][262198] Updated weights for policy 0, policy_version 18151 (0.0009)
+[2026-06-06 21:26:25,963][262198] Updated weights for policy 0, policy_version 18161 (0.0008)
+[2026-06-06 21:26:26,997][262198] Updated weights for policy 0, policy_version 18173 (0.0008)
+[2026-06-06 21:26:27,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4540.0, 300 sec: 4193.2). Total num frames: 9308160. Throughput: 0: 4839.3. Samples: 9315380. Policy #0 lag: (min: 7.0, avg: 20.6, max: 39.0)
+[2026-06-06 21:26:27,924][261293] Avg episode reward: [(0, '1692.028')]
+[2026-06-06 21:26:27,929][262115] Saving new best policy, reward=1692.028!
+[2026-06-06 21:26:28,102][262198] Updated weights for policy 0, policy_version 18183 (0.0008)
+[2026-06-06 21:26:29,411][262198] Updated weights for policy 0, policy_version 18195 (0.0010)
+[2026-06-06 21:26:30,522][262198] Updated weights for policy 0, policy_version 18207 (0.0010)
+[2026-06-06 21:26:32,261][262198] Updated weights for policy 0, policy_version 18217 (0.0010)
+[2026-06-06 21:26:32,920][261293] Fps is (10 sec: 4505.9, 60 sec: 4642.2, 300 sec: 4214.0). Total num frames: 9330688. Throughput: 0: 4883.3. Samples: 9329534. Policy #0 lag: (min: 5.0, avg: 21.1, max: 34.0)
+[2026-06-06 21:26:32,922][261293] Avg episode reward: [(0, '1546.925')]
+[2026-06-06 21:26:33,123][262198] Updated weights for policy 0, policy_version 18227 (0.0010)
+[2026-06-06 21:26:34,304][262198] Updated weights for policy 0, policy_version 18237 (0.0010)
+[2026-06-06 21:26:35,234][262198] Updated weights for policy 0, policy_version 18247 (0.0010)
+[2026-06-06 21:26:36,561][262198] Updated weights for policy 0, policy_version 18257 (0.0011)
+[2026-06-06 21:26:37,126][262198] Updated weights for policy 0, policy_version 18267 (0.0012)
+[2026-06-06 21:26:37,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4642.1, 300 sec: 4227.9). Total num frames: 9353216. Throughput: 0: 4910.5. Samples: 9356450. Policy #0 lag: (min: 5.0, avg: 21.1, max: 34.0)
+[2026-06-06 21:26:37,923][261293] Avg episode reward: [(0, '1641.354')]
+[2026-06-06 21:26:39,259][262198] Updated weights for policy 0, policy_version 18278 (0.0014)
+[2026-06-06 21:26:40,017][262198] Updated weights for policy 0, policy_version 18288 (0.0010)
+[2026-06-06 21:26:41,169][262198] Updated weights for policy 0, policy_version 18298 (0.0012)
+[2026-06-06 21:26:42,851][262198] Updated weights for policy 0, policy_version 18308 (0.0009)
+[2026-06-06 21:26:42,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4710.4, 300 sec: 4234.9). Total num frames: 9373696. Throughput: 0: 4851.5. Samples: 9380456. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:26:42,923][261293] Avg episode reward: [(0, '1476.896')]
+[2026-06-06 21:26:44,119][262198] Updated weights for policy 0, policy_version 18318 (0.0010)
+[2026-06-06 21:26:45,256][262198] Updated weights for policy 0, policy_version 18330 (0.0009)
+[2026-06-06 21:26:46,513][262198] Updated weights for policy 0, policy_version 18340 (0.0010)
+[2026-06-06 21:26:47,549][262198] Updated weights for policy 0, policy_version 18351 (0.0008)
+[2026-06-06 21:26:47,920][261293] Fps is (10 sec: 4301.1, 60 sec: 4744.5, 300 sec: 4241.8). Total num frames: 9396224. Throughput: 0: 4823.2. Samples: 9393608. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:26:47,922][261293] Avg episode reward: [(0, '1447.656')]
+[2026-06-06 21:26:48,553][262198] Updated weights for policy 0, policy_version 18363 (0.0009)
+[2026-06-06 21:26:50,177][262198] Updated weights for policy 0, policy_version 18373 (0.0008)
+[2026-06-06 21:26:50,969][262198] Updated weights for policy 0, policy_version 18383 (0.0009)
+[2026-06-06 21:26:51,979][262198] Updated weights for policy 0, policy_version 18396 (0.0009)
+[2026-06-06 21:26:52,920][261293] Fps is (10 sec: 4710.6, 60 sec: 4812.8, 300 sec: 4255.7). Total num frames: 9420800. Throughput: 0: 4828.4. Samples: 9422900. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:26:52,922][261293] Avg episode reward: [(0, '1477.697')]
+[2026-06-06 21:26:53,361][262198] Updated weights for policy 0, policy_version 18406 (0.0008)
+[2026-06-06 21:26:54,194][262198] Updated weights for policy 0, policy_version 18416 (0.0009)
+[2026-06-06 21:26:55,604][262198] Updated weights for policy 0, policy_version 18430 (0.0008)
+[2026-06-06 21:26:56,432][262198] Updated weights for policy 0, policy_version 18440 (0.0008)
+[2026-06-06 21:26:57,779][262198] Updated weights for policy 0, policy_version 18455 (0.0009)
+[2026-06-06 21:26:57,921][261293] Fps is (10 sec: 5324.7, 60 sec: 4881.1, 300 sec: 4276.5). Total num frames: 9449472. Throughput: 0: 4864.0. Samples: 9455738. Policy #0 lag: (min: 3.0, avg: 23.0, max: 40.0)
+[2026-06-06 21:26:57,923][261293] Avg episode reward: [(0, '1385.646')]
+[2026-06-06 21:26:59,088][262198] Updated weights for policy 0, policy_version 18465 (0.0009)
+[2026-06-06 21:26:59,875][262198] Updated weights for policy 0, policy_version 18475 (0.0009)
+[2026-06-06 21:27:01,112][262198] Updated weights for policy 0, policy_version 18492 (0.0009)
+[2026-06-06 21:27:02,506][262198] Updated weights for policy 0, policy_version 18502 (0.0010)
+[2026-06-06 21:27:02,921][261293] Fps is (10 sec: 5529.5, 60 sec: 4949.3, 300 sec: 4297.3). Total num frames: 9476096. Throughput: 0: 4883.0. Samples: 9471774. Policy #0 lag: (min: 3.0, avg: 23.0, max: 40.0)
+[2026-06-06 21:27:02,922][261293] Avg episode reward: [(0, '1445.118')]
+[2026-06-06 21:27:03,595][262115] Signal inference workers to stop experience collection... (2500 times)
+[2026-06-06 21:27:03,617][262198] InferenceWorker_p0-w0: stopping experience collection (2500 times)
+[2026-06-06 21:27:03,652][262115] Signal inference workers to resume experience collection... (2500 times)
+[2026-06-06 21:27:03,652][262198] InferenceWorker_p0-w0: resuming experience collection (2500 times)
+[2026-06-06 21:27:03,908][262198] Updated weights for policy 0, policy_version 18516 (0.0010)
+[2026-06-06 21:27:05,008][262198] Updated weights for policy 0, policy_version 18528 (0.0010)
+[2026-06-06 21:27:05,979][262198] Updated weights for policy 0, policy_version 18538 (0.0008)
+[2026-06-06 21:27:07,225][262198] Updated weights for policy 0, policy_version 18555 (0.0008)
+[2026-06-06 21:27:07,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4915.2, 300 sec: 4318.2). Total num frames: 9500672. Throughput: 0: 4826.1. Samples: 9502420. Policy #0 lag: (min: 3.0, avg: 23.0, max: 40.0)
+[2026-06-06 21:27:07,922][261293] Avg episode reward: [(0, '1388.147')]
+[2026-06-06 21:27:08,801][262198] Updated weights for policy 0, policy_version 18565 (0.0010)
+[2026-06-06 21:27:09,871][262198] Updated weights for policy 0, policy_version 18577 (0.0010)
+[2026-06-06 21:27:11,122][262198] Updated weights for policy 0, policy_version 18590 (0.0009)
+[2026-06-06 21:27:12,127][262198] Updated weights for policy 0, policy_version 18600 (0.0008)
+[2026-06-06 21:27:12,920][261293] Fps is (10 sec: 5120.1, 60 sec: 4949.4, 300 sec: 4325.1). Total num frames: 9527296. Throughput: 0: 4870.2. Samples: 9534538. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 21:27:12,922][261293] Avg episode reward: [(0, '1601.907')]
+[2026-06-06 21:27:13,042][262198] Updated weights for policy 0, policy_version 18612 (0.0009)
+[2026-06-06 21:27:14,268][262198] Updated weights for policy 0, policy_version 18622 (0.0009)
+[2026-06-06 21:27:15,344][262198] Updated weights for policy 0, policy_version 18634 (0.0009)
+[2026-06-06 21:27:16,835][262198] Updated weights for policy 0, policy_version 18649 (0.0010)
+[2026-06-06 21:27:17,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4915.2, 300 sec: 4339.0). Total num frames: 9551872. Throughput: 0: 4911.3. Samples: 9550544. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 21:27:17,923][261293] Avg episode reward: [(0, '1615.108')]
+[2026-06-06 21:27:17,934][262198] Updated weights for policy 0, policy_version 18659 (0.0009)
+[2026-06-06 21:27:19,485][262198] Updated weights for policy 0, policy_version 18674 (0.0009)
+[2026-06-06 21:27:20,167][262198] Updated weights for policy 0, policy_version 18684 (0.0010)
+[2026-06-06 21:27:22,001][262198] Updated weights for policy 0, policy_version 18696 (0.0009)
+[2026-06-06 21:27:22,874][262198] Updated weights for policy 0, policy_version 18708 (0.0010)
+[2026-06-06 21:27:22,921][261293] Fps is (10 sec: 5119.8, 60 sec: 4881.1, 300 sec: 4366.8). Total num frames: 9578496. Throughput: 0: 4953.7. Samples: 9579368. Policy #0 lag: (min: 2.0, avg: 21.0, max: 39.0)
+[2026-06-06 21:27:22,923][261293] Avg episode reward: [(0, '1466.015')]
+[2026-06-06 21:27:24,206][262198] Updated weights for policy 0, policy_version 18718 (0.0009)
+[2026-06-06 21:27:25,251][262198] Updated weights for policy 0, policy_version 18728 (0.0011)
+[2026-06-06 21:27:26,175][262198] Updated weights for policy 0, policy_version 18738 (0.0010)
+[2026-06-06 21:27:26,911][262198] Updated weights for policy 0, policy_version 18748 (0.0010)
+[2026-06-06 21:27:27,920][261293] Fps is (10 sec: 5120.2, 60 sec: 4915.2, 300 sec: 4394.5). Total num frames: 9603072. Throughput: 0: 5066.8. Samples: 9608458. Policy #0 lag: (min: 2.0, avg: 21.0, max: 39.0)
+[2026-06-06 21:27:27,922][261293] Avg episode reward: [(0, '1688.920')]
+[2026-06-06 21:27:28,942][262198] Updated weights for policy 0, policy_version 18761 (0.0010)
+[2026-06-06 21:27:29,978][262198] Updated weights for policy 0, policy_version 18773 (0.0011)
+[2026-06-06 21:27:31,056][262198] Updated weights for policy 0, policy_version 18783 (0.0009)
+[2026-06-06 21:27:32,356][262198] Updated weights for policy 0, policy_version 18796 (0.0009)
+[2026-06-06 21:27:32,920][261293] Fps is (10 sec: 4710.6, 60 sec: 4915.2, 300 sec: 4408.4). Total num frames: 9625600. Throughput: 0: 5079.2. Samples: 9622172. Policy #0 lag: (min: 2.0, avg: 21.0, max: 39.0)
+[2026-06-06 21:27:32,922][261293] Avg episode reward: [(0, '1747.029')]
+[2026-06-06 21:27:33,148][262115] Saving new best policy, reward=1747.029!
+[2026-06-06 21:27:33,342][262198] Updated weights for policy 0, policy_version 18808 (0.0009)
+[2026-06-06 21:27:34,746][262198] Updated weights for policy 0, policy_version 18818 (0.0009)
+[2026-06-06 21:27:36,026][262198] Updated weights for policy 0, policy_version 18830 (0.0011)
+[2026-06-06 21:27:37,030][262198] Updated weights for policy 0, policy_version 18842 (0.0009)
+[2026-06-06 21:27:37,920][261293] Fps is (10 sec: 4710.4, 60 sec: 4949.4, 300 sec: 4422.3). Total num frames: 9650176. Throughput: 0: 5088.9. Samples: 9651902. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:27:37,922][261293] Avg episode reward: [(0, '1518.712')]
+[2026-06-06 21:27:38,316][262198] Updated weights for policy 0, policy_version 18852 (0.0009)
+[2026-06-06 21:27:39,439][262198] Updated weights for policy 0, policy_version 18864 (0.0008)
+[2026-06-06 21:27:40,317][262198] Updated weights for policy 0, policy_version 18876 (0.0010)
+[2026-06-06 21:27:42,187][262198] Updated weights for policy 0, policy_version 18888 (0.0010)
+[2026-06-06 21:27:42,921][261293] Fps is (10 sec: 4915.0, 60 sec: 5017.6, 300 sec: 4429.2). Total num frames: 9674752. Throughput: 0: 5020.5. Samples: 9681662. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:27:42,924][261293] Avg episode reward: [(0, '1412.141')]
+[2026-06-06 21:27:43,185][262198] Updated weights for policy 0, policy_version 18898 (0.0008)
+[2026-06-06 21:27:44,307][262198] Updated weights for policy 0, policy_version 18909 (0.0009)
+[2026-06-06 21:27:45,506][262198] Updated weights for policy 0, policy_version 18922 (0.0009)
+[2026-06-06 21:27:46,851][262198] Updated weights for policy 0, policy_version 18938 (0.0012)
+[2026-06-06 21:27:47,747][262198] Updated weights for policy 0, policy_version 18948 (0.0009)
+[2026-06-06 21:27:47,921][261293] Fps is (10 sec: 5119.7, 60 sec: 5085.8, 300 sec: 4450.1). Total num frames: 9701376. Throughput: 0: 5008.8. Samples: 9697172. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:27:47,926][261293] Avg episode reward: [(0, '1417.691')]
+[2026-06-06 21:27:49,072][262198] Updated weights for policy 0, policy_version 18960 (0.0009)
+[2026-06-06 21:27:49,919][262198] Updated weights for policy 0, policy_version 18971 (0.0010)
+[2026-06-06 21:27:51,541][262198] Updated weights for policy 0, policy_version 18981 (0.0009)
+[2026-06-06 21:27:52,312][262198] Updated weights for policy 0, policy_version 18991 (0.0008)
+[2026-06-06 21:27:52,921][261293] Fps is (10 sec: 5120.0, 60 sec: 5085.8, 300 sec: 4470.9). Total num frames: 9725952. Throughput: 0: 5005.8. Samples: 9727684. Policy #0 lag: (min: 7.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:27:52,923][261293] Avg episode reward: [(0, '1657.337')]
+[2026-06-06 21:27:53,152][262198] Updated weights for policy 0, policy_version 19001 (0.0008)
+[2026-06-06 21:27:54,264][262198] Updated weights for policy 0, policy_version 19011 (0.0009)
+[2026-06-06 21:27:55,454][262198] Updated weights for policy 0, policy_version 19021 (0.0008)
+[2026-06-06 21:27:56,124][262198] Updated weights for policy 0, policy_version 19031 (0.0008)
+[2026-06-06 21:27:57,402][262198] Updated weights for policy 0, policy_version 19041 (0.0009)
+[2026-06-06 21:27:57,920][261293] Fps is (10 sec: 4915.4, 60 sec: 5017.6, 300 sec: 4477.8). Total num frames: 9750528. Throughput: 0: 4966.5. Samples: 9758030. Policy #0 lag: (min: 7.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:27:57,922][261293] Avg episode reward: [(0, '1552.117')]
+[2026-06-06 21:27:58,418][262198] Updated weights for policy 0, policy_version 19051 (0.0010)
+[2026-06-06 21:27:59,567][262198] Updated weights for policy 0, policy_version 19061 (0.0010)
+[2026-06-06 21:28:00,897][262198] Updated weights for policy 0, policy_version 19071 (0.0011)
+[2026-06-06 21:28:02,036][262198] Updated weights for policy 0, policy_version 19081 (0.0009)
+[2026-06-06 21:28:02,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4949.3, 300 sec: 4484.8). Total num frames: 9773056. Throughput: 0: 4903.3. Samples: 9771194. Policy #0 lag: (min: 3.0, avg: 19.2, max: 39.0)
+[2026-06-06 21:28:02,923][261293] Avg episode reward: [(0, '1578.308')]
+[2026-06-06 21:28:02,973][262198] Updated weights for policy 0, policy_version 19091 (0.0009)
+[2026-06-06 21:28:04,416][262198] Updated weights for policy 0, policy_version 19104 (0.0009)
+[2026-06-06 21:28:05,228][262198] Updated weights for policy 0, policy_version 19114 (0.0008)
+[2026-06-06 21:28:06,200][262198] Updated weights for policy 0, policy_version 19126 (0.0009)
+[2026-06-06 21:28:07,234][262198] Updated weights for policy 0, policy_version 19136 (0.0009)
+[2026-06-06 21:28:07,921][261293] Fps is (10 sec: 5120.0, 60 sec: 5017.6, 300 sec: 4505.6). Total num frames: 9801728. Throughput: 0: 4948.7. Samples: 9802058. Policy #0 lag: (min: 3.0, avg: 19.2, max: 39.0)
+[2026-06-06 21:28:07,926][261293] Avg episode reward: [(0, '1567.768')]
+[2026-06-06 21:28:08,238][262198] Updated weights for policy 0, policy_version 19146 (0.0009)
+[2026-06-06 21:28:08,814][262198] Updated weights for policy 0, policy_version 19156 (0.0008)
+[2026-06-06 21:28:09,207][262115] Signal inference workers to stop experience collection... (2550 times)
+[2026-06-06 21:28:09,229][262198] InferenceWorker_p0-w0: stopping experience collection (2550 times)
+[2026-06-06 21:28:09,262][262115] Signal inference workers to resume experience collection... (2550 times)
+[2026-06-06 21:28:09,262][262198] InferenceWorker_p0-w0: resuming experience collection (2550 times)
+[2026-06-06 21:28:09,858][262198] Updated weights for policy 0, policy_version 19167 (0.0008)
+[2026-06-06 21:28:11,157][262198] Updated weights for policy 0, policy_version 19177 (0.0007)
+[2026-06-06 21:28:12,253][262198] Updated weights for policy 0, policy_version 19189 (0.0008)
+[2026-06-06 21:28:12,921][261293] Fps is (10 sec: 5734.6, 60 sec: 5051.7, 300 sec: 4519.5). Total num frames: 9830400. Throughput: 0: 5043.9. Samples: 9835434. Policy #0 lag: (min: 3.0, avg: 19.2, max: 39.0)
+[2026-06-06 21:28:12,923][261293] Avg episode reward: [(0, '1590.892')]
+[2026-06-06 21:28:13,096][262198] Updated weights for policy 0, policy_version 19201 (0.0008)
+[2026-06-06 21:28:14,178][262198] Updated weights for policy 0, policy_version 19211 (0.0008)
+[2026-06-06 21:28:14,799][262198] Updated weights for policy 0, policy_version 19221 (0.0008)
+[2026-06-06 21:28:15,571][262198] Updated weights for policy 0, policy_version 19231 (0.0008)
+[2026-06-06 21:28:17,324][262198] Updated weights for policy 0, policy_version 19241 (0.0009)
+[2026-06-06 21:28:17,921][261293] Fps is (10 sec: 5529.5, 60 sec: 5085.9, 300 sec: 4533.4). Total num frames: 9857024. Throughput: 0: 5135.3. Samples: 9853262. Policy #0 lag: (min: 6.0, avg: 24.9, max: 44.0)
+[2026-06-06 21:28:17,922][261293] Avg episode reward: [(0, '1441.230')]
+[2026-06-06 21:28:18,035][262198] Updated weights for policy 0, policy_version 19254 (0.0008)
+[2026-06-06 21:28:18,815][262198] Updated weights for policy 0, policy_version 19264 (0.0010)
+[2026-06-06 21:28:20,401][262198] Updated weights for policy 0, policy_version 19276 (0.0010)
+[2026-06-06 21:28:21,253][262198] Updated weights for policy 0, policy_version 19286 (0.0009)
+[2026-06-06 21:28:22,303][262198] Updated weights for policy 0, policy_version 19296 (0.0008)
+[2026-06-06 21:28:22,921][261293] Fps is (10 sec: 5119.9, 60 sec: 5051.7, 300 sec: 4540.3). Total num frames: 9881600. Throughput: 0: 5151.7. Samples: 9883732. Policy #0 lag: (min: 6.0, avg: 24.9, max: 44.0)
+[2026-06-06 21:28:22,923][261293] Avg episode reward: [(0, '1528.670')]
+[2026-06-06 21:28:23,731][262198] Updated weights for policy 0, policy_version 19308 (0.0009)
+[2026-06-06 21:28:24,970][262198] Updated weights for policy 0, policy_version 19321 (0.0011)
+[2026-06-06 21:28:25,613][262198] Updated weights for policy 0, policy_version 19331 (0.0009)
+[2026-06-06 21:28:26,840][262198] Updated weights for policy 0, policy_version 19343 (0.0008)
+[2026-06-06 21:28:27,870][262198] Updated weights for policy 0, policy_version 19353 (0.0009)
+[2026-06-06 21:28:27,921][261293] Fps is (10 sec: 5120.0, 60 sec: 5085.8, 300 sec: 4554.2). Total num frames: 9908224. Throughput: 0: 5194.0. Samples: 9915390. Policy #0 lag: (min: 6.0, avg: 24.9, max: 44.0)
+[2026-06-06 21:28:27,922][261293] Avg episode reward: [(0, '1661.005')]
+[2026-06-06 21:28:29,225][262198] Updated weights for policy 0, policy_version 19365 (0.0011)
+[2026-06-06 21:28:30,275][262198] Updated weights for policy 0, policy_version 19378 (0.0010)
+[2026-06-06 21:28:31,062][262198] Updated weights for policy 0, policy_version 19388 (0.0008)
+[2026-06-06 21:28:32,446][262198] Updated weights for policy 0, policy_version 19398 (0.0010)
+[2026-06-06 21:28:32,921][261293] Fps is (10 sec: 5120.1, 60 sec: 5120.0, 300 sec: 4561.1). Total num frames: 9932800. Throughput: 0: 5194.5. Samples: 9930926. Policy #0 lag: (min: 1.0, avg: 17.1, max: 33.0)
+[2026-06-06 21:28:32,924][261293] Avg episode reward: [(0, '1472.510')]
+[2026-06-06 21:28:33,595][262198] Updated weights for policy 0, policy_version 19410 (0.0009)
+[2026-06-06 21:28:34,460][262198] Updated weights for policy 0, policy_version 19420 (0.0010)
+[2026-06-06 21:28:35,911][262198] Updated weights for policy 0, policy_version 19430 (0.0010)
+[2026-06-06 21:28:36,803][262198] Updated weights for policy 0, policy_version 19443 (0.0009)
+[2026-06-06 21:28:37,921][261293] Fps is (10 sec: 5120.1, 60 sec: 5154.1, 300 sec: 4582.0). Total num frames: 9959424. Throughput: 0: 5185.5. Samples: 9961032. Policy #0 lag: (min: 1.0, avg: 17.1, max: 33.0)
+[2026-06-06 21:28:37,925][261293] Avg episode reward: [(0, '1582.776')]
+[2026-06-06 21:28:38,071][262198] Updated weights for policy 0, policy_version 19453 (0.0009)
+[2026-06-06 21:28:39,056][262198] Updated weights for policy 0, policy_version 19463 (0.0009)
+[2026-06-06 21:28:40,117][262198] Updated weights for policy 0, policy_version 19475 (0.0009)
+[2026-06-06 21:28:41,303][262198] Updated weights for policy 0, policy_version 19485 (0.0009)
+[2026-06-06 21:28:42,237][262198] Updated weights for policy 0, policy_version 19495 (0.0009)
+[2026-06-06 21:28:42,920][261293] Fps is (10 sec: 5120.3, 60 sec: 5154.2, 300 sec: 4595.9). Total num frames: 9984000. Throughput: 0: 5188.6. Samples: 9991514. Policy #0 lag: (min: 1.0, avg: 17.3, max: 36.0)
+[2026-06-06 21:28:42,922][261293] Avg episode reward: [(0, '1498.859')]
+[2026-06-06 21:28:43,337][262198] Updated weights for policy 0, policy_version 19507 (0.0009)
+[2026-06-06 21:28:44,414][262198] Updated weights for policy 0, policy_version 19517 (0.0009)
+[2026-06-06 21:28:45,366][262198] Updated weights for policy 0, policy_version 19527 (0.0009)
+[2026-06-06 21:28:46,365][262198] Updated weights for policy 0, policy_version 19540 (0.0010)
+[2026-06-06 21:28:47,921][261293] Fps is (10 sec: 5120.1, 60 sec: 5154.2, 300 sec: 4616.7). Total num frames: 10010624. Throughput: 0: 5245.3. Samples: 10007230. Policy #0 lag: (min: 1.0, avg: 17.3, max: 36.0)
+[2026-06-06 21:28:47,924][261293] Avg episode reward: [(0, '1692.668')]
+[2026-06-06 21:28:47,983][262198] Updated weights for policy 0, policy_version 19553 (0.0009)
+[2026-06-06 21:28:49,319][262198] Updated weights for policy 0, policy_version 19565 (0.0009)
+[2026-06-06 21:28:50,047][262198] Updated weights for policy 0, policy_version 19575 (0.0009)
+[2026-06-06 21:28:51,436][262198] Updated weights for policy 0, policy_version 19588 (0.0009)
+[2026-06-06 21:28:52,820][262198] Updated weights for policy 0, policy_version 19601 (0.0010)
+[2026-06-06 21:28:52,921][261293] Fps is (10 sec: 5324.5, 60 sec: 5188.3, 300 sec: 4644.4). Total num frames: 10037248. Throughput: 0: 5221.1. Samples: 10037008. Policy #0 lag: (min: 1.0, avg: 17.3, max: 36.0)
+[2026-06-06 21:28:52,923][261293] Avg episode reward: [(0, '1437.620')]
+[2026-06-06 21:28:54,271][262198] Updated weights for policy 0, policy_version 19613 (0.0008)
+[2026-06-06 21:28:55,251][262198] Updated weights for policy 0, policy_version 19623 (0.0009)
+[2026-06-06 21:28:56,152][262198] Updated weights for policy 0, policy_version 19635 (0.0010)
+[2026-06-06 21:28:57,223][262198] Updated weights for policy 0, policy_version 19645 (0.0009)
+[2026-06-06 21:28:57,921][261293] Fps is (10 sec: 5119.7, 60 sec: 5188.2, 300 sec: 4658.3). Total num frames: 10061824. Throughput: 0: 5177.3. Samples: 10068414. Policy #0 lag: (min: 0.0, avg: 18.9, max: 32.0)
+[2026-06-06 21:28:57,925][261293] Avg episode reward: [(0, '1603.835')]
+[2026-06-06 21:28:58,063][262198] Updated weights for policy 0, policy_version 19655 (0.0008)
+[2026-06-06 21:28:59,010][262198] Updated weights for policy 0, policy_version 19665 (0.0009)
+[2026-06-06 21:28:59,807][262198] Updated weights for policy 0, policy_version 19675 (0.0008)
+[2026-06-06 21:29:00,861][262198] Updated weights for policy 0, policy_version 19685 (0.0009)
+[2026-06-06 21:29:02,089][262198] Updated weights for policy 0, policy_version 19700 (0.0010)
+[2026-06-06 21:29:02,922][261293] Fps is (10 sec: 5324.1, 60 sec: 5290.6, 300 sec: 4693.0). Total num frames: 10090496. Throughput: 0: 5162.4. Samples: 10085578. Policy #0 lag: (min: 0.0, avg: 18.9, max: 32.0)
+[2026-06-06 21:29:02,924][261293] Avg episode reward: [(0, '1589.176')]
+[2026-06-06 21:29:03,108][262198] Updated weights for policy 0, policy_version 19710 (0.0009)
+[2026-06-06 21:29:04,408][262198] Updated weights for policy 0, policy_version 19725 (0.0010)
+[2026-06-06 21:29:05,827][262198] Updated weights for policy 0, policy_version 19738 (0.0010)
+[2026-06-06 21:29:06,846][262198] Updated weights for policy 0, policy_version 19748 (0.0010)
+[2026-06-06 21:29:07,921][261293] Fps is (10 sec: 5529.5, 60 sec: 5256.5, 300 sec: 4713.9). Total num frames: 10117120. Throughput: 0: 5195.4. Samples: 10117526. Policy #0 lag: (min: 0.0, avg: 18.9, max: 32.0)
+[2026-06-06 21:29:07,927][261293] Avg episode reward: [(0, '1493.772')]
+[2026-06-06 21:29:08,545][262198] Updated weights for policy 0, policy_version 19761 (0.0010)
+[2026-06-06 21:29:09,725][262198] Updated weights for policy 0, policy_version 19773 (0.0010)
+[2026-06-06 21:29:10,635][262198] Updated weights for policy 0, policy_version 19783 (0.0009)
+[2026-06-06 21:29:11,952][262198] Updated weights for policy 0, policy_version 19793 (0.0009)
+[2026-06-06 21:29:12,920][261293] Fps is (10 sec: 4916.0, 60 sec: 5154.1, 300 sec: 4713.9). Total num frames: 10139648. Throughput: 0: 5132.5. Samples: 10146352. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:29:12,922][261293] Avg episode reward: [(0, '1607.853')]
+[2026-06-06 21:29:13,157][262198] Updated weights for policy 0, policy_version 19805 (0.0010)
+[2026-06-06 21:29:14,068][262198] Updated weights for policy 0, policy_version 19815 (0.0009)
+[2026-06-06 21:29:15,400][262198] Updated weights for policy 0, policy_version 19828 (0.0009)
+[2026-06-06 21:29:16,396][262198] Updated weights for policy 0, policy_version 19838 (0.0009)
+[2026-06-06 21:29:17,490][262198] Updated weights for policy 0, policy_version 19848 (0.0008)
+[2026-06-06 21:29:17,921][261293] Fps is (10 sec: 4710.4, 60 sec: 5120.0, 300 sec: 4727.8). Total num frames: 10164224. Throughput: 0: 5112.2. Samples: 10160976. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:29:17,923][261293] Avg episode reward: [(0, '1581.775')]
+[2026-06-06 21:29:17,969][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000019856_10166272.pth...
+[2026-06-06 21:29:18,742][262198] Updated weights for policy 0, policy_version 19861 (0.0011)
+[2026-06-06 21:29:19,738][262198] Updated weights for policy 0, policy_version 19871 (0.0009)
+[2026-06-06 21:29:21,184][262198] Updated weights for policy 0, policy_version 19884 (0.0011)
+[2026-06-06 21:29:22,138][262198] Updated weights for policy 0, policy_version 19896 (0.0009)
+[2026-06-06 21:29:22,921][261293] Fps is (10 sec: 4915.1, 60 sec: 5120.0, 300 sec: 4734.7). Total num frames: 10188800. Throughput: 0: 5107.0. Samples: 10190848. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:29:22,924][261293] Avg episode reward: [(0, '1784.997')]
+[2026-06-06 21:29:23,000][262115] Saving new best policy, reward=1784.997!
+[2026-06-06 21:29:23,587][262198] Updated weights for policy 0, policy_version 19906 (0.0009)
+[2026-06-06 21:29:24,475][262198] Updated weights for policy 0, policy_version 19919 (0.0007)
+[2026-06-06 21:29:25,541][262198] Updated weights for policy 0, policy_version 19929 (0.0009)
+[2026-06-06 21:29:26,673][262198] Updated weights for policy 0, policy_version 19939 (0.0008)
+[2026-06-06 21:29:27,137][262115] Signal inference workers to stop experience collection... (2600 times)
+[2026-06-06 21:29:27,148][262115] Signal inference workers to resume experience collection... (2600 times)
+[2026-06-06 21:29:27,159][262198] InferenceWorker_p0-w0: stopping experience collection (2600 times)
+[2026-06-06 21:29:27,159][262198] InferenceWorker_p0-w0: resuming experience collection (2600 times)
+[2026-06-06 21:29:27,572][262198] Updated weights for policy 0, policy_version 19952 (0.0008)
+[2026-06-06 21:29:27,921][261293] Fps is (10 sec: 5120.0, 60 sec: 5120.0, 300 sec: 4755.5). Total num frames: 10215424. Throughput: 0: 5141.0. Samples: 10222862. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:29:27,925][261293] Avg episode reward: [(0, '1562.456')]
+[2026-06-06 21:29:28,737][262198] Updated weights for policy 0, policy_version 19964 (0.0009)
+[2026-06-06 21:29:30,145][262198] Updated weights for policy 0, policy_version 19974 (0.0008)
+[2026-06-06 21:29:31,233][262198] Updated weights for policy 0, policy_version 19987 (0.0009)
+[2026-06-06 21:29:32,302][262198] Updated weights for policy 0, policy_version 19997 (0.0009)
+[2026-06-06 21:29:32,920][261293] Fps is (10 sec: 5120.1, 60 sec: 5120.0, 300 sec: 4776.4). Total num frames: 10240000. Throughput: 0: 5125.7. Samples: 10237888. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:29:32,922][261293] Avg episode reward: [(0, '1726.462')]
+[2026-06-06 21:29:33,247][262198] Updated weights for policy 0, policy_version 20007 (0.0010)
+[2026-06-06 21:29:34,330][262198] Updated weights for policy 0, policy_version 20021 (0.0009)
+[2026-06-06 21:29:35,306][262198] Updated weights for policy 0, policy_version 20031 (0.0009)
+[2026-06-06 21:29:36,533][262198] Updated weights for policy 0, policy_version 20041 (0.0008)
+[2026-06-06 21:29:37,278][262198] Updated weights for policy 0, policy_version 20056 (0.0010)
+[2026-06-06 21:29:37,921][261293] Fps is (10 sec: 5529.8, 60 sec: 5188.3, 300 sec: 4804.1). Total num frames: 10270720. Throughput: 0: 5196.9. Samples: 10270868. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 21:29:37,922][261293] Avg episode reward: [(0, '1603.996')]
+[2026-06-06 21:29:38,772][262198] Updated weights for policy 0, policy_version 20066 (0.0010)
+[2026-06-06 21:29:39,724][262198] Updated weights for policy 0, policy_version 20079 (0.0009)
+[2026-06-06 21:29:40,781][262198] Updated weights for policy 0, policy_version 20092 (0.0009)
+[2026-06-06 21:29:42,217][262198] Updated weights for policy 0, policy_version 20102 (0.0011)
+[2026-06-06 21:29:42,920][261293] Fps is (10 sec: 5529.7, 60 sec: 5188.3, 300 sec: 4804.1). Total num frames: 10295296. Throughput: 0: 5197.4. Samples: 10302296. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 21:29:42,922][261293] Avg episode reward: [(0, '1451.018')]
+[2026-06-06 21:29:42,969][262198] Updated weights for policy 0, policy_version 20112 (0.0008)
+[2026-06-06 21:29:44,199][262198] Updated weights for policy 0, policy_version 20124 (0.0009)
+[2026-06-06 21:29:45,474][262198] Updated weights for policy 0, policy_version 20134 (0.0009)
+[2026-06-06 21:29:46,505][262198] Updated weights for policy 0, policy_version 20145 (0.0009)
+[2026-06-06 21:29:47,613][262198] Updated weights for policy 0, policy_version 20157 (0.0009)
+[2026-06-06 21:29:47,920][261293] Fps is (10 sec: 5120.2, 60 sec: 5188.3, 300 sec: 4825.0). Total num frames: 10321920. Throughput: 0: 5161.1. Samples: 10317818. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 21:29:47,922][261293] Avg episode reward: [(0, '1712.529')]
+[2026-06-06 21:29:48,987][262198] Updated weights for policy 0, policy_version 20169 (0.0008)
+[2026-06-06 21:29:49,714][262198] Updated weights for policy 0, policy_version 20183 (0.0009)
+[2026-06-06 21:29:50,887][262198] Updated weights for policy 0, policy_version 20193 (0.0008)
+[2026-06-06 21:29:52,102][262198] Updated weights for policy 0, policy_version 20208 (0.0009)
+[2026-06-06 21:29:52,921][261293] Fps is (10 sec: 5529.1, 60 sec: 5222.4, 300 sec: 4845.8). Total num frames: 10350592. Throughput: 0: 5185.1. Samples: 10350856. Policy #0 lag: (min: 3.0, avg: 19.1, max: 39.0)
+[2026-06-06 21:29:52,925][261293] Avg episode reward: [(0, '1602.599')]
+[2026-06-06 21:29:53,658][262198] Updated weights for policy 0, policy_version 20221 (0.0009)
+[2026-06-06 21:29:54,916][262198] Updated weights for policy 0, policy_version 20234 (0.0009)
+[2026-06-06 21:29:56,088][262198] Updated weights for policy 0, policy_version 20250 (0.0009)
+[2026-06-06 21:29:57,079][262198] Updated weights for policy 0, policy_version 20260 (0.0008)
+[2026-06-06 21:29:57,921][261293] Fps is (10 sec: 5324.4, 60 sec: 5222.4, 300 sec: 4845.8). Total num frames: 10375168. Throughput: 0: 5258.9. Samples: 10383004. Policy #0 lag: (min: 3.0, avg: 19.1, max: 39.0)
+[2026-06-06 21:29:57,924][261293] Avg episode reward: [(0, '1602.296')]
+[2026-06-06 21:29:58,464][262198] Updated weights for policy 0, policy_version 20273 (0.0010)
+[2026-06-06 21:29:59,042][262198] Updated weights for policy 0, policy_version 20283 (0.0008)
+[2026-06-06 21:30:00,399][262198] Updated weights for policy 0, policy_version 20293 (0.0008)
+[2026-06-06 21:30:01,064][262198] Updated weights for policy 0, policy_version 20304 (0.0008)
+[2026-06-06 21:30:02,586][262198] Updated weights for policy 0, policy_version 20314 (0.0012)
+[2026-06-06 21:30:02,921][261293] Fps is (10 sec: 5120.2, 60 sec: 5188.4, 300 sec: 4880.5). Total num frames: 10401792. Throughput: 0: 5325.7. Samples: 10400632. Policy #0 lag: (min: 3.0, avg: 19.1, max: 39.0)
+[2026-06-06 21:30:02,925][261293] Avg episode reward: [(0, '1635.845')]
+[2026-06-06 21:30:05,169][262198] Updated weights for policy 0, policy_version 20325 (0.0012)
+[2026-06-06 21:30:06,809][262198] Updated weights for policy 0, policy_version 20337 (0.0013)
+[2026-06-06 21:30:07,439][262198] Updated weights for policy 0, policy_version 20347 (0.0008)
+[2026-06-06 21:30:07,921][261293] Fps is (10 sec: 4505.8, 60 sec: 5051.8, 300 sec: 4880.5). Total num frames: 10420224. Throughput: 0: 5098.5. Samples: 10420280. Policy #0 lag: (min: 3.0, avg: 18.4, max: 39.0)
+[2026-06-06 21:30:07,922][261293] Avg episode reward: [(0, '1355.467')]
+[2026-06-06 21:30:08,517][262198] Updated weights for policy 0, policy_version 20357 (0.0008)
+[2026-06-06 21:30:09,101][262198] Updated weights for policy 0, policy_version 20367 (0.0007)
+[2026-06-06 21:30:09,948][262198] Updated weights for policy 0, policy_version 20377 (0.0007)
+[2026-06-06 21:30:10,805][262198] Updated weights for policy 0, policy_version 20387 (0.0007)
+[2026-06-06 21:30:11,868][262198] Updated weights for policy 0, policy_version 20397 (0.0009)
+[2026-06-06 21:30:12,341][262198] Updated weights for policy 0, policy_version 20407 (0.0007)
+[2026-06-06 21:30:12,921][261293] Fps is (10 sec: 4915.3, 60 sec: 5188.3, 300 sec: 4922.1). Total num frames: 10450944. Throughput: 0: 5175.6. Samples: 10455762. Policy #0 lag: (min: 3.0, avg: 18.4, max: 39.0)
+[2026-06-06 21:30:12,925][261293] Avg episode reward: [(0, '1508.500')]
+[2026-06-06 21:30:13,622][262198] Updated weights for policy 0, policy_version 20417 (0.0008)
+[2026-06-06 21:30:14,464][262198] Updated weights for policy 0, policy_version 20427 (0.0009)
+[2026-06-06 21:30:15,335][262198] Updated weights for policy 0, policy_version 20439 (0.0009)
+[2026-06-06 21:30:16,630][262198] Updated weights for policy 0, policy_version 20452 (0.0008)
+[2026-06-06 21:30:17,691][262198] Updated weights for policy 0, policy_version 20463 (0.0007)
+[2026-06-06 21:30:17,920][261293] Fps is (10 sec: 5734.5, 60 sec: 5222.4, 300 sec: 4956.9). Total num frames: 10477568. Throughput: 0: 5221.1. Samples: 10472838. Policy #0 lag: (min: 3.0, avg: 18.4, max: 39.0)
+[2026-06-06 21:30:17,922][261293] Avg episode reward: [(0, '1596.388')]
+[2026-06-06 21:30:18,477][262198] Updated weights for policy 0, policy_version 20473 (0.0008)
+[2026-06-06 21:30:19,326][262198] Updated weights for policy 0, policy_version 20483 (0.0009)
+[2026-06-06 21:30:20,511][262198] Updated weights for policy 0, policy_version 20493 (0.0008)
+[2026-06-06 21:30:21,382][262198] Updated weights for policy 0, policy_version 20505 (0.0010)
+[2026-06-06 21:30:22,332][262198] Updated weights for policy 0, policy_version 20515 (0.0009)
+[2026-06-06 21:30:22,922][261293] Fps is (10 sec: 5324.6, 60 sec: 5256.5, 300 sec: 4977.7). Total num frames: 10504192. Throughput: 0: 5235.1. Samples: 10506448. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:30:22,925][261293] Avg episode reward: [(0, '1448.570')]
+[2026-06-06 21:30:23,641][262198] Updated weights for policy 0, policy_version 20525 (0.0011)
+[2026-06-06 21:30:24,195][262198] Updated weights for policy 0, policy_version 20535 (0.0009)
+[2026-06-06 21:30:25,363][262198] Updated weights for policy 0, policy_version 20545 (0.0010)
+[2026-06-06 21:30:26,373][262198] Updated weights for policy 0, policy_version 20555 (0.0009)
+[2026-06-06 21:30:27,321][262198] Updated weights for policy 0, policy_version 20565 (0.0009)
+[2026-06-06 21:30:27,920][261293] Fps is (10 sec: 5529.7, 60 sec: 5290.7, 300 sec: 5019.3). Total num frames: 10532864. Throughput: 0: 5244.4. Samples: 10538294. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:30:27,922][261293] Avg episode reward: [(0, '1544.583')]
+[2026-06-06 21:30:28,351][262198] Updated weights for policy 0, policy_version 20575 (0.0008)
+[2026-06-06 21:30:29,463][262198] Updated weights for policy 0, policy_version 20585 (0.0008)
+[2026-06-06 21:30:30,387][262198] Updated weights for policy 0, policy_version 20596 (0.0009)
+[2026-06-06 21:30:31,357][262198] Updated weights for policy 0, policy_version 20606 (0.0008)
+[2026-06-06 21:30:32,214][262198] Updated weights for policy 0, policy_version 20616 (0.0009)
+[2026-06-06 21:30:32,922][261293] Fps is (10 sec: 5324.3, 60 sec: 5290.5, 300 sec: 5026.3). Total num frames: 10557440. Throughput: 0: 5255.3. Samples: 10554314. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:30:32,924][261293] Avg episode reward: [(0, '1579.265')]
+[2026-06-06 21:30:33,223][262198] Updated weights for policy 0, policy_version 20626 (0.0008)
+[2026-06-06 21:30:33,945][262198] Updated weights for policy 0, policy_version 20636 (0.0007)
+[2026-06-06 21:30:35,081][262198] Updated weights for policy 0, policy_version 20646 (0.0009)
+[2026-06-06 21:30:36,416][262198] Updated weights for policy 0, policy_version 20661 (0.0010)
+[2026-06-06 21:30:37,294][262198] Updated weights for policy 0, policy_version 20671 (0.0008)
+[2026-06-06 21:30:37,921][261293] Fps is (10 sec: 5119.8, 60 sec: 5222.4, 300 sec: 5061.0). Total num frames: 10584064. Throughput: 0: 5252.7. Samples: 10587224. Policy #0 lag: (min: 1.0, avg: 20.3, max: 39.0)
+[2026-06-06 21:30:37,926][261293] Avg episode reward: [(0, '1543.901')]
+[2026-06-06 21:30:38,573][262198] Updated weights for policy 0, policy_version 20683 (0.0009)
+[2026-06-06 21:30:39,423][262198] Updated weights for policy 0, policy_version 20695 (0.0009)
+[2026-06-06 21:30:40,921][262198] Updated weights for policy 0, policy_version 20705 (0.0009)
+[2026-06-06 21:30:41,908][262198] Updated weights for policy 0, policy_version 20717 (0.0009)
+[2026-06-06 21:30:42,796][262198] Updated weights for policy 0, policy_version 20729 (0.0009)
+[2026-06-06 21:30:42,921][261293] Fps is (10 sec: 5735.2, 60 sec: 5324.8, 300 sec: 5095.7). Total num frames: 10614784. Throughput: 0: 5268.0. Samples: 10620062. Policy #0 lag: (min: 1.0, avg: 20.3, max: 39.0)
+[2026-06-06 21:30:42,923][261293] Avg episode reward: [(0, '1479.395')]
+[2026-06-06 21:30:43,976][262198] Updated weights for policy 0, policy_version 20739 (0.0009)
+[2026-06-06 21:30:44,925][262115] Signal inference workers to stop experience collection... (2650 times)
+[2026-06-06 21:30:44,946][262198] InferenceWorker_p0-w0: stopping experience collection (2650 times)
+[2026-06-06 21:30:44,947][262198] Updated weights for policy 0, policy_version 20749 (0.0009)
+[2026-06-06 21:30:45,003][262115] Signal inference workers to resume experience collection... (2650 times)
+[2026-06-06 21:30:45,003][262198] InferenceWorker_p0-w0: resuming experience collection (2650 times)
+[2026-06-06 21:30:45,939][262198] Updated weights for policy 0, policy_version 20761 (0.0008)
+[2026-06-06 21:30:47,082][262198] Updated weights for policy 0, policy_version 20771 (0.0008)
+[2026-06-06 21:30:47,921][261293] Fps is (10 sec: 5529.6, 60 sec: 5290.6, 300 sec: 5109.6). Total num frames: 10639360. Throughput: 0: 5227.4. Samples: 10635862. Policy #0 lag: (min: 1.0, avg: 20.3, max: 39.0)
+[2026-06-06 21:30:47,927][261293] Avg episode reward: [(0, '1701.805')]
+[2026-06-06 21:30:47,994][262198] Updated weights for policy 0, policy_version 20784 (0.0010)
+[2026-06-06 21:30:49,139][262198] Updated weights for policy 0, policy_version 20794 (0.0009)
+[2026-06-06 21:30:50,412][262198] Updated weights for policy 0, policy_version 20804 (0.0010)
+[2026-06-06 21:30:51,214][262198] Updated weights for policy 0, policy_version 20814 (0.0009)
+[2026-06-06 21:30:51,919][262198] Updated weights for policy 0, policy_version 20824 (0.0009)
+[2026-06-06 21:30:52,921][261293] Fps is (10 sec: 5119.9, 60 sec: 5256.6, 300 sec: 5116.5). Total num frames: 10665984. Throughput: 0: 5474.8. Samples: 10666646. Policy #0 lag: (min: 3.0, avg: 21.5, max: 39.0)
+[2026-06-06 21:30:52,924][261293] Avg episode reward: [(0, '1458.045')]
+[2026-06-06 21:30:53,604][262198] Updated weights for policy 0, policy_version 20836 (0.0010)
+[2026-06-06 21:30:54,490][262198] Updated weights for policy 0, policy_version 20846 (0.0009)
+[2026-06-06 21:30:55,303][262198] Updated weights for policy 0, policy_version 20856 (0.0010)
+[2026-06-06 21:30:56,994][262198] Updated weights for policy 0, policy_version 20872 (0.0010)
+[2026-06-06 21:30:57,921][261293] Fps is (10 sec: 5119.9, 60 sec: 5256.5, 300 sec: 5123.5). Total num frames: 10690560. Throughput: 0: 5348.1. Samples: 10696426. Policy #0 lag: (min: 3.0, avg: 21.5, max: 39.0)
+[2026-06-06 21:30:57,924][261293] Avg episode reward: [(0, '1468.190')]
+[2026-06-06 21:30:58,129][262198] Updated weights for policy 0, policy_version 20882 (0.0010)
+[2026-06-06 21:30:59,319][262198] Updated weights for policy 0, policy_version 20893 (0.0010)
+[2026-06-06 21:31:00,402][262198] Updated weights for policy 0, policy_version 20903 (0.0010)
+[2026-06-06 21:31:01,411][262198] Updated weights for policy 0, policy_version 20913 (0.0008)
+[2026-06-06 21:31:02,422][262198] Updated weights for policy 0, policy_version 20923 (0.0009)
+[2026-06-06 21:31:02,920][261293] Fps is (10 sec: 4915.4, 60 sec: 5222.4, 300 sec: 5116.5). Total num frames: 10715136. Throughput: 0: 5284.9. Samples: 10710658. Policy #0 lag: (min: 6.0, avg: 21.5, max: 39.0)
+[2026-06-06 21:31:02,923][261293] Avg episode reward: [(0, '1525.684')]
+[2026-06-06 21:31:03,714][262198] Updated weights for policy 0, policy_version 20933 (0.0012)
+[2026-06-06 21:31:04,710][262198] Updated weights for policy 0, policy_version 20944 (0.0010)
+[2026-06-06 21:31:05,742][262198] Updated weights for policy 0, policy_version 20954 (0.0009)
+[2026-06-06 21:31:06,851][262198] Updated weights for policy 0, policy_version 20964 (0.0010)
+[2026-06-06 21:31:07,920][261293] Fps is (10 sec: 4505.8, 60 sec: 5256.5, 300 sec: 5102.6). Total num frames: 10735616. Throughput: 0: 5188.2. Samples: 10739916. Policy #0 lag: (min: 6.0, avg: 21.5, max: 39.0)
+[2026-06-06 21:31:07,922][261293] Avg episode reward: [(0, '1449.428')]
+[2026-06-06 21:31:08,278][262198] Updated weights for policy 0, policy_version 20979 (0.0010)
+[2026-06-06 21:31:09,317][262198] Updated weights for policy 0, policy_version 20989 (0.0008)
+[2026-06-06 21:31:10,482][262198] Updated weights for policy 0, policy_version 20999 (0.0009)
+[2026-06-06 21:31:11,641][262198] Updated weights for policy 0, policy_version 21012 (0.0011)
+[2026-06-06 21:31:12,765][262198] Updated weights for policy 0, policy_version 21022 (0.0008)
+[2026-06-06 21:31:12,921][261293] Fps is (10 sec: 4914.8, 60 sec: 5222.3, 300 sec: 5109.6). Total num frames: 10764288. Throughput: 0: 5126.2. Samples: 10768980. Policy #0 lag: (min: 6.0, avg: 21.5, max: 39.0)
+[2026-06-06 21:31:12,923][261293] Avg episode reward: [(0, '1432.995')]
+[2026-06-06 21:31:13,993][262198] Updated weights for policy 0, policy_version 21032 (0.0009)
+[2026-06-06 21:31:15,212][262198] Updated weights for policy 0, policy_version 21043 (0.0012)
+[2026-06-06 21:31:16,581][262198] Updated weights for policy 0, policy_version 21053 (0.0009)
+[2026-06-06 21:31:17,837][262198] Updated weights for policy 0, policy_version 21063 (0.0009)
+[2026-06-06 21:31:17,921][261293] Fps is (10 sec: 4915.2, 60 sec: 5120.0, 300 sec: 5081.8). Total num frames: 10784768. Throughput: 0: 5056.7. Samples: 10781858. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:31:17,924][261293] Avg episode reward: [(0, '1450.383')]
+[2026-06-06 21:31:18,800][262198] Updated weights for policy 0, policy_version 21076 (0.0009)
+[2026-06-06 21:31:19,886][262198] Updated weights for policy 0, policy_version 21086 (0.0009)
+[2026-06-06 21:31:21,149][262198] Updated weights for policy 0, policy_version 21096 (0.0009)
+[2026-06-06 21:31:22,545][262198] Updated weights for policy 0, policy_version 21111 (0.0012)
+[2026-06-06 21:31:22,921][261293] Fps is (10 sec: 4505.9, 60 sec: 5085.9, 300 sec: 5088.8). Total num frames: 10809344. Throughput: 0: 4976.1. Samples: 10811148. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:31:22,924][261293] Avg episode reward: [(0, '1596.965')]
+[2026-06-06 21:31:24,130][262198] Updated weights for policy 0, policy_version 21121 (0.0009)
+[2026-06-06 21:31:24,955][262198] Updated weights for policy 0, policy_version 21131 (0.0009)
+[2026-06-06 21:31:25,800][262198] Updated weights for policy 0, policy_version 21143 (0.0009)
+[2026-06-06 21:31:27,469][262198] Updated weights for policy 0, policy_version 21153 (0.0011)
+[2026-06-06 21:31:27,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4983.4, 300 sec: 5088.7). Total num frames: 10831872. Throughput: 0: 4878.6. Samples: 10839602. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 21:31:27,922][261293] Avg episode reward: [(0, '1538.404')]
+[2026-06-06 21:31:28,792][262198] Updated weights for policy 0, policy_version 21163 (0.0009)
+[2026-06-06 21:31:29,848][262198] Updated weights for policy 0, policy_version 21175 (0.0010)
+[2026-06-06 21:31:31,136][262198] Updated weights for policy 0, policy_version 21185 (0.0009)
+[2026-06-06 21:31:32,167][262198] Updated weights for policy 0, policy_version 21195 (0.0009)
+[2026-06-06 21:31:32,921][261293] Fps is (10 sec: 4710.5, 60 sec: 4983.6, 300 sec: 5095.7). Total num frames: 10856448. Throughput: 0: 4806.6. Samples: 10852160. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 21:31:32,923][261293] Avg episode reward: [(0, '1661.224')]
+[2026-06-06 21:31:32,938][262198] Updated weights for policy 0, policy_version 21207 (0.0009)
+[2026-06-06 21:31:34,219][262198] Updated weights for policy 0, policy_version 21218 (0.0009)
+[2026-06-06 21:31:35,156][262198] Updated weights for policy 0, policy_version 21228 (0.0008)
+[2026-06-06 21:31:35,966][262198] Updated weights for policy 0, policy_version 21238 (0.0008)
+[2026-06-06 21:31:36,595][262198] Updated weights for policy 0, policy_version 21248 (0.0008)
+[2026-06-06 21:31:37,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4983.4, 300 sec: 5116.5). Total num frames: 10883072. Throughput: 0: 4836.4. Samples: 10884284. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 21:31:37,924][261293] Avg episode reward: [(0, '1509.268')]
+[2026-06-06 21:31:38,097][262198] Updated weights for policy 0, policy_version 21258 (0.0008)
+[2026-06-06 21:31:39,244][262198] Updated weights for policy 0, policy_version 21272 (0.0011)
+[2026-06-06 21:31:40,574][262198] Updated weights for policy 0, policy_version 21283 (0.0012)
+[2026-06-06 21:31:42,250][262198] Updated weights for policy 0, policy_version 21293 (0.0012)
+[2026-06-06 21:31:42,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4846.9, 300 sec: 5116.5). Total num frames: 10905600. Throughput: 0: 4799.6. Samples: 10912408. Policy #0 lag: (min: 3.0, avg: 16.9, max: 35.0)
+[2026-06-06 21:31:42,922][261293] Avg episode reward: [(0, '1544.609')]
+[2026-06-06 21:31:43,094][262198] Updated weights for policy 0, policy_version 21304 (0.0012)
+[2026-06-06 21:31:44,289][262198] Updated weights for policy 0, policy_version 21314 (0.0009)
+[2026-06-06 21:31:45,457][262198] Updated weights for policy 0, policy_version 21325 (0.0012)
+[2026-06-06 21:31:46,467][262198] Updated weights for policy 0, policy_version 21335 (0.0010)
+[2026-06-06 21:31:47,711][262198] Updated weights for policy 0, policy_version 21345 (0.0010)
+[2026-06-06 21:31:47,921][261293] Fps is (10 sec: 4710.6, 60 sec: 4847.0, 300 sec: 5116.5). Total num frames: 10930176. Throughput: 0: 4819.3. Samples: 10927526. Policy #0 lag: (min: 3.0, avg: 18.0, max: 36.0)
+[2026-06-06 21:31:47,923][261293] Avg episode reward: [(0, '1403.100')]
+[2026-06-06 21:31:48,736][262198] Updated weights for policy 0, policy_version 21355 (0.0009)
+[2026-06-06 21:31:50,007][262198] Updated weights for policy 0, policy_version 21368 (0.0009)
+[2026-06-06 21:31:51,117][262198] Updated weights for policy 0, policy_version 21380 (0.0013)
+[2026-06-06 21:31:52,594][262198] Updated weights for policy 0, policy_version 21390 (0.0009)
+[2026-06-06 21:31:52,920][261293] Fps is (10 sec: 4915.4, 60 sec: 4812.8, 300 sec: 5102.6). Total num frames: 10954752. Throughput: 0: 4800.4. Samples: 10955932. Policy #0 lag: (min: 3.0, avg: 18.0, max: 36.0)
+[2026-06-06 21:31:52,922][261293] Avg episode reward: [(0, '1545.149')]
+[2026-06-06 21:31:53,592][262198] Updated weights for policy 0, policy_version 21401 (0.0009)
+[2026-06-06 21:31:54,553][262198] Updated weights for policy 0, policy_version 21411 (0.0009)
+[2026-06-06 21:31:55,767][262198] Updated weights for policy 0, policy_version 21422 (0.0011)
+[2026-06-06 21:31:56,594][262198] Updated weights for policy 0, policy_version 21435 (0.0009)
+[2026-06-06 21:31:57,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4812.8, 300 sec: 5095.7). Total num frames: 10979328. Throughput: 0: 4829.8. Samples: 10986318. Policy #0 lag: (min: 3.0, avg: 18.0, max: 36.0)
+[2026-06-06 21:31:57,924][261293] Avg episode reward: [(0, '1488.851')]
+[2026-06-06 21:31:58,062][262198] Updated weights for policy 0, policy_version 21445 (0.0009)
+[2026-06-06 21:31:58,514][262115] Signal inference workers to stop experience collection... (2700 times)
+[2026-06-06 21:31:58,530][262198] InferenceWorker_p0-w0: stopping experience collection (2700 times)
+[2026-06-06 21:31:58,589][262115] Signal inference workers to resume experience collection... (2700 times)
+[2026-06-06 21:31:58,590][262198] InferenceWorker_p0-w0: resuming experience collection (2700 times)
+[2026-06-06 21:31:58,900][262198] Updated weights for policy 0, policy_version 21459 (0.0007)
+[2026-06-06 21:31:59,979][262198] Updated weights for policy 0, policy_version 21471 (0.0010)
+[2026-06-06 21:32:01,625][262198] Updated weights for policy 0, policy_version 21481 (0.0008)
+[2026-06-06 21:32:02,531][262198] Updated weights for policy 0, policy_version 21497 (0.0008)
+[2026-06-06 21:32:02,920][261293] Fps is (10 sec: 5324.9, 60 sec: 4881.1, 300 sec: 5109.6). Total num frames: 11008000. Throughput: 0: 4913.2. Samples: 11002950. Policy #0 lag: (min: 3.0, avg: 17.4, max: 36.0)
+[2026-06-06 21:32:02,922][261293] Avg episode reward: [(0, '1464.612')]
+[2026-06-06 21:32:03,354][262198] Updated weights for policy 0, policy_version 21507 (0.0008)
+[2026-06-06 21:32:04,841][262198] Updated weights for policy 0, policy_version 21523 (0.0009)
+[2026-06-06 21:32:05,764][262198] Updated weights for policy 0, policy_version 21535 (0.0009)
+[2026-06-06 21:32:07,509][262198] Updated weights for policy 0, policy_version 21548 (0.0008)
+[2026-06-06 21:32:07,921][261293] Fps is (10 sec: 5734.5, 60 sec: 5017.6, 300 sec: 5116.5). Total num frames: 11036672. Throughput: 0: 5008.2. Samples: 11036514. Policy #0 lag: (min: 3.0, avg: 17.4, max: 36.0)
+[2026-06-06 21:32:07,923][261293] Avg episode reward: [(0, '1639.084')]
+[2026-06-06 21:32:08,495][262198] Updated weights for policy 0, policy_version 21564 (0.0009)
+[2026-06-06 21:32:09,945][262198] Updated weights for policy 0, policy_version 21575 (0.0010)
+[2026-06-06 21:32:10,964][262198] Updated weights for policy 0, policy_version 21586 (0.0009)
+[2026-06-06 21:32:12,221][262198] Updated weights for policy 0, policy_version 21597 (0.0010)
+[2026-06-06 21:32:12,921][261293] Fps is (10 sec: 5324.6, 60 sec: 4949.4, 300 sec: 5116.5). Total num frames: 11061248. Throughput: 0: 5069.6. Samples: 11067734. Policy #0 lag: (min: 3.0, avg: 17.4, max: 36.0)
+[2026-06-06 21:32:12,922][261293] Avg episode reward: [(0, '1539.071')]
+[2026-06-06 21:32:13,052][262198] Updated weights for policy 0, policy_version 21607 (0.0007)
+[2026-06-06 21:32:13,941][262198] Updated weights for policy 0, policy_version 21617 (0.0008)
+[2026-06-06 21:32:14,501][262198] Updated weights for policy 0, policy_version 21627 (0.0007)
+[2026-06-06 21:32:15,984][262198] Updated weights for policy 0, policy_version 21637 (0.0008)
+[2026-06-06 21:32:16,671][262198] Updated weights for policy 0, policy_version 21647 (0.0008)
+[2026-06-06 21:32:17,623][262198] Updated weights for policy 0, policy_version 21659 (0.0009)
+[2026-06-06 21:32:17,921][261293] Fps is (10 sec: 5324.7, 60 sec: 5085.9, 300 sec: 5123.5). Total num frames: 11089920. Throughput: 0: 5175.2. Samples: 11085044. Policy #0 lag: (min: 3.0, avg: 17.4, max: 35.0)
+[2026-06-06 21:32:17,925][261293] Avg episode reward: [(0, '1416.700')]
+[2026-06-06 21:32:19,151][262198] Updated weights for policy 0, policy_version 21672 (0.0009)
+[2026-06-06 21:32:20,507][262198] Updated weights for policy 0, policy_version 21685 (0.0012)
+[2026-06-06 21:32:21,890][262198] Updated weights for policy 0, policy_version 21696 (0.0012)
+[2026-06-06 21:32:22,921][261293] Fps is (10 sec: 5119.9, 60 sec: 5051.7, 300 sec: 5116.5). Total num frames: 11112448. Throughput: 0: 5148.5. Samples: 11115964. Policy #0 lag: (min: 3.0, avg: 17.4, max: 35.0)
+[2026-06-06 21:32:22,924][261293] Avg episode reward: [(0, '1651.503')]
+[2026-06-06 21:32:23,236][262198] Updated weights for policy 0, policy_version 21706 (0.0010)
+[2026-06-06 21:32:24,393][262198] Updated weights for policy 0, policy_version 21716 (0.0012)
+[2026-06-06 21:32:26,086][262198] Updated weights for policy 0, policy_version 21728 (0.0012)
+[2026-06-06 21:32:27,176][262198] Updated weights for policy 0, policy_version 21738 (0.0009)
+[2026-06-06 21:32:27,922][261293] Fps is (10 sec: 4300.4, 60 sec: 5017.5, 300 sec: 5109.6). Total num frames: 11132928. Throughput: 0: 5090.6. Samples: 11141488. Policy #0 lag: (min: 3.0, avg: 17.4, max: 35.0)
+[2026-06-06 21:32:27,927][261293] Avg episode reward: [(0, '1692.122')]
+[2026-06-06 21:32:27,994][262198] Updated weights for policy 0, policy_version 21749 (0.0009)
+[2026-06-06 21:32:29,386][262198] Updated weights for policy 0, policy_version 21761 (0.0009)
+[2026-06-06 21:32:30,515][262198] Updated weights for policy 0, policy_version 21771 (0.0008)
+[2026-06-06 21:32:31,314][262198] Updated weights for policy 0, policy_version 21781 (0.0009)
+[2026-06-06 21:32:32,573][262198] Updated weights for policy 0, policy_version 21793 (0.0009)
+[2026-06-06 21:32:32,920][261293] Fps is (10 sec: 4710.7, 60 sec: 5051.8, 300 sec: 5116.5). Total num frames: 11159552. Throughput: 0: 5073.1. Samples: 11155816. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:32:32,921][261293] Avg episode reward: [(0, '1736.343')]
+[2026-06-06 21:32:33,582][262198] Updated weights for policy 0, policy_version 21803 (0.0009)
+[2026-06-06 21:32:34,562][262198] Updated weights for policy 0, policy_version 21813 (0.0008)
+[2026-06-06 21:32:35,751][262198] Updated weights for policy 0, policy_version 21825 (0.0009)
+[2026-06-06 21:32:36,614][262198] Updated weights for policy 0, policy_version 21836 (0.0009)
+[2026-06-06 21:32:37,718][262198] Updated weights for policy 0, policy_version 21846 (0.0009)
+[2026-06-06 21:32:37,921][261293] Fps is (10 sec: 5325.2, 60 sec: 5051.8, 300 sec: 5123.5). Total num frames: 11186176. Throughput: 0: 5137.4. Samples: 11187118. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:32:37,927][261293] Avg episode reward: [(0, '1484.207')]
+[2026-06-06 21:32:38,761][262198] Updated weights for policy 0, policy_version 21856 (0.0008)
+[2026-06-06 21:32:39,658][262198] Updated weights for policy 0, policy_version 21866 (0.0009)
+[2026-06-06 21:32:40,568][262198] Updated weights for policy 0, policy_version 21876 (0.0008)
+[2026-06-06 21:32:41,757][262198] Updated weights for policy 0, policy_version 21886 (0.0009)
+[2026-06-06 21:32:42,559][262198] Updated weights for policy 0, policy_version 21896 (0.0009)
+[2026-06-06 21:32:42,921][261293] Fps is (10 sec: 5119.8, 60 sec: 5085.9, 300 sec: 5116.5). Total num frames: 11210752. Throughput: 0: 5140.8. Samples: 11217652. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:32:42,923][261293] Avg episode reward: [(0, '1730.077')]
+[2026-06-06 21:32:43,998][262198] Updated weights for policy 0, policy_version 21906 (0.0009)
+[2026-06-06 21:32:44,661][262198] Updated weights for policy 0, policy_version 21916 (0.0008)
+[2026-06-06 21:32:46,519][262198] Updated weights for policy 0, policy_version 21929 (0.0009)
+[2026-06-06 21:32:47,404][262198] Updated weights for policy 0, policy_version 21941 (0.0007)
+[2026-06-06 21:32:47,920][261293] Fps is (10 sec: 5120.3, 60 sec: 5120.0, 300 sec: 5123.5). Total num frames: 11237376. Throughput: 0: 5107.4. Samples: 11232784. Policy #0 lag: (min: 0.0, avg: 18.2, max: 33.0)
+[2026-06-06 21:32:47,921][261293] Avg episode reward: [(0, '1673.230')]
+[2026-06-06 21:32:48,406][262198] Updated weights for policy 0, policy_version 21951 (0.0009)
+[2026-06-06 21:32:49,641][262198] Updated weights for policy 0, policy_version 21961 (0.0009)
+[2026-06-06 21:32:50,487][262198] Updated weights for policy 0, policy_version 21972 (0.0009)
+[2026-06-06 21:32:51,719][262198] Updated weights for policy 0, policy_version 21982 (0.0009)
+[2026-06-06 21:32:52,499][262198] Updated weights for policy 0, policy_version 21992 (0.0010)
+[2026-06-06 21:32:52,920][261293] Fps is (10 sec: 4915.3, 60 sec: 5085.9, 300 sec: 5116.5). Total num frames: 11259904. Throughput: 0: 5015.6. Samples: 11262216. Policy #0 lag: (min: 0.0, avg: 18.2, max: 33.0)
+[2026-06-06 21:32:52,922][261293] Avg episode reward: [(0, '1712.251')]
+[2026-06-06 21:32:53,933][262198] Updated weights for policy 0, policy_version 22002 (0.0011)
+[2026-06-06 21:32:55,012][262198] Updated weights for policy 0, policy_version 22012 (0.0009)
+[2026-06-06 21:32:56,462][262198] Updated weights for policy 0, policy_version 22022 (0.0011)
+[2026-06-06 21:32:57,396][262198] Updated weights for policy 0, policy_version 22032 (0.0011)
+[2026-06-06 21:32:57,921][261293] Fps is (10 sec: 4505.3, 60 sec: 5051.7, 300 sec: 5116.5). Total num frames: 11282432. Throughput: 0: 4895.4. Samples: 11288028. Policy #0 lag: (min: 0.0, avg: 18.2, max: 33.0)
+[2026-06-06 21:32:57,923][261293] Avg episode reward: [(0, '1340.287')]
+[2026-06-06 21:32:58,545][262198] Updated weights for policy 0, policy_version 22042 (0.0008)
+[2026-06-06 21:32:59,468][262198] Updated weights for policy 0, policy_version 22052 (0.0009)
+[2026-06-06 21:33:00,649][262198] Updated weights for policy 0, policy_version 22063 (0.0011)
+[2026-06-06 21:33:01,699][262198] Updated weights for policy 0, policy_version 22073 (0.0009)
+[2026-06-06 21:33:02,691][262198] Updated weights for policy 0, policy_version 22083 (0.0008)
+[2026-06-06 21:33:02,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4983.5, 300 sec: 5102.6). Total num frames: 11307008. Throughput: 0: 4860.9. Samples: 11303782. Policy #0 lag: (min: 3.0, avg: 20.0, max: 35.0)
+[2026-06-06 21:33:02,921][261293] Avg episode reward: [(0, '1654.760')]
+[2026-06-06 21:33:03,896][262198] Updated weights for policy 0, policy_version 22096 (0.0008)
+[2026-06-06 21:33:05,057][262198] Updated weights for policy 0, policy_version 22106 (0.0009)
+[2026-06-06 21:33:05,883][262198] Updated weights for policy 0, policy_version 22116 (0.0008)
+[2026-06-06 21:33:07,119][262198] Updated weights for policy 0, policy_version 22128 (0.0010)
+[2026-06-06 21:33:07,920][261293] Fps is (10 sec: 4915.4, 60 sec: 4915.2, 300 sec: 5088.8). Total num frames: 11331584. Throughput: 0: 4847.7. Samples: 11334108. Policy #0 lag: (min: 3.0, avg: 20.0, max: 35.0)
+[2026-06-06 21:33:07,924][261293] Avg episode reward: [(0, '1538.077')]
+[2026-06-06 21:33:08,305][262198] Updated weights for policy 0, policy_version 22138 (0.0009)
+[2026-06-06 21:33:09,355][262198] Updated weights for policy 0, policy_version 22148 (0.0011)
+[2026-06-06 21:33:10,691][262198] Updated weights for policy 0, policy_version 22158 (0.0010)
+[2026-06-06 21:33:11,896][262198] Updated weights for policy 0, policy_version 22168 (0.0011)
+[2026-06-06 21:33:12,920][261293] Fps is (10 sec: 4710.4, 60 sec: 4881.1, 300 sec: 5074.9). Total num frames: 11354112. Throughput: 0: 4857.1. Samples: 11360054. Policy #0 lag: (min: 3.0, avg: 20.0, max: 35.0)
+[2026-06-06 21:33:12,923][261293] Avg episode reward: [(0, '1486.432')]
+[2026-06-06 21:33:13,267][262198] Updated weights for policy 0, policy_version 22178 (0.0012)
+[2026-06-06 21:33:14,455][262198] Updated weights for policy 0, policy_version 22188 (0.0011)
+[2026-06-06 21:33:16,054][262198] Updated weights for policy 0, policy_version 22198 (0.0008)
+[2026-06-06 21:33:17,154][262198] Updated weights for policy 0, policy_version 22208 (0.0009)
+[2026-06-06 21:33:17,924][261293] Fps is (10 sec: 4095.0, 60 sec: 4710.2, 300 sec: 5054.0). Total num frames: 11372544. Throughput: 0: 4799.7. Samples: 11371814. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:33:17,928][261293] Avg episode reward: [(0, '1757.780')]
+[2026-06-06 21:33:18,727][262198] Updated weights for policy 0, policy_version 22219 (0.0012)
+[2026-06-06 21:33:20,243][262198] Updated weights for policy 0, policy_version 22229 (0.0011)
+[2026-06-06 21:33:20,970][262198] Updated weights for policy 0, policy_version 22239 (0.0008)
+[2026-06-06 21:33:22,383][262198] Updated weights for policy 0, policy_version 22249 (0.0008)
+[2026-06-06 21:33:22,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4710.4, 300 sec: 5040.2). Total num frames: 11395072. Throughput: 0: 4643.5. Samples: 11396076. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:33:22,923][261293] Avg episode reward: [(0, '1704.250')]
+[2026-06-06 21:33:23,254][262198] Updated weights for policy 0, policy_version 22259 (0.0010)
+[2026-06-06 21:33:24,890][262198] Updated weights for policy 0, policy_version 22269 (0.0010)
+[2026-06-06 21:33:26,684][262198] Updated weights for policy 0, policy_version 22281 (0.0011)
+[2026-06-06 21:33:27,680][262198] Updated weights for policy 0, policy_version 22291 (0.0010)
+[2026-06-06 21:33:27,921][261293] Fps is (10 sec: 4096.9, 60 sec: 4676.3, 300 sec: 5019.3). Total num frames: 11413504. Throughput: 0: 4498.3. Samples: 11420078. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:33:27,924][261293] Avg episode reward: [(0, '1576.959')]
+[2026-06-06 21:33:29,154][262198] Updated weights for policy 0, policy_version 22301 (0.0009)
+[2026-06-06 21:33:30,356][262198] Updated weights for policy 0, policy_version 22312 (0.0011)
+[2026-06-06 21:33:32,261][262198] Updated weights for policy 0, policy_version 22322 (0.0010)
+[2026-06-06 21:33:32,921][261293] Fps is (10 sec: 3481.6, 60 sec: 4505.6, 300 sec: 4984.6). Total num frames: 11429888. Throughput: 0: 4425.5. Samples: 11431932. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:33:32,924][261293] Avg episode reward: [(0, '1693.793')]
+[2026-06-06 21:33:33,513][262198] Updated weights for policy 0, policy_version 22332 (0.0009)
+[2026-06-06 21:33:34,662][262198] Updated weights for policy 0, policy_version 22342 (0.0008)
+[2026-06-06 21:33:35,403][262198] Updated weights for policy 0, policy_version 22352 (0.0009)
+[2026-06-06 21:33:36,842][262198] Updated weights for policy 0, policy_version 22362 (0.0009)
+[2026-06-06 21:33:37,482][262198] Updated weights for policy 0, policy_version 22372 (0.0008)
+[2026-06-06 21:33:37,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4471.4, 300 sec: 4984.6). Total num frames: 11454464. Throughput: 0: 4320.9. Samples: 11456658. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:33:37,925][261293] Avg episode reward: [(0, '1609.346')]
+[2026-06-06 21:33:39,248][262198] Updated weights for policy 0, policy_version 22385 (0.0009)
+[2026-06-06 21:33:40,116][262198] Updated weights for policy 0, policy_version 22395 (0.0008)
+[2026-06-06 21:33:41,302][262198] Updated weights for policy 0, policy_version 22405 (0.0009)
+[2026-06-06 21:33:42,097][262198] Updated weights for policy 0, policy_version 22415 (0.0011)
+[2026-06-06 21:33:42,922][261293] Fps is (10 sec: 4914.6, 60 sec: 4471.4, 300 sec: 4977.7). Total num frames: 11479040. Throughput: 0: 4395.5. Samples: 11485830. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:33:42,925][261293] Avg episode reward: [(0, '1566.205')]
+[2026-06-06 21:33:43,552][262198] Updated weights for policy 0, policy_version 22425 (0.0009)
+[2026-06-06 21:33:44,246][262198] Updated weights for policy 0, policy_version 22436 (0.0008)
+[2026-06-06 21:33:45,423][262198] Updated weights for policy 0, policy_version 22446 (0.0009)
+[2026-06-06 21:33:46,535][262198] Updated weights for policy 0, policy_version 22456 (0.0008)
+[2026-06-06 21:33:47,663][262198] Updated weights for policy 0, policy_version 22467 (0.0012)
+[2026-06-06 21:33:47,920][261293] Fps is (10 sec: 4915.6, 60 sec: 4437.3, 300 sec: 4970.7). Total num frames: 11503616. Throughput: 0: 4377.3. Samples: 11500762. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:33:47,922][261293] Avg episode reward: [(0, '1571.893')]
+[2026-06-06 21:33:48,910][262198] Updated weights for policy 0, policy_version 22477 (0.0009)
+[2026-06-06 21:33:50,166][262198] Updated weights for policy 0, policy_version 22487 (0.0010)
+[2026-06-06 21:33:51,222][262198] Updated weights for policy 0, policy_version 22498 (0.0009)
+[2026-06-06 21:33:52,435][262198] Updated weights for policy 0, policy_version 22510 (0.0009)
+[2026-06-06 21:33:52,921][261293] Fps is (10 sec: 4915.6, 60 sec: 4471.4, 300 sec: 4970.7). Total num frames: 11528192. Throughput: 0: 4314.7. Samples: 11528270. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:33:52,922][261293] Avg episode reward: [(0, '1573.630')]
+[2026-06-06 21:33:53,265][262198] Updated weights for policy 0, policy_version 22520 (0.0009)
+[2026-06-06 21:33:54,302][262198] Updated weights for policy 0, policy_version 22531 (0.0009)
+[2026-06-06 21:33:55,394][262115] Signal inference workers to stop experience collection... (2750 times)
+[2026-06-06 21:33:55,395][262115] Signal inference workers to resume experience collection... (2750 times)
+[2026-06-06 21:33:55,416][262198] InferenceWorker_p0-w0: stopping experience collection (2750 times)
+[2026-06-06 21:33:55,416][262198] InferenceWorker_p0-w0: resuming experience collection (2750 times)
+[2026-06-06 21:33:55,453][262198] Updated weights for policy 0, policy_version 22541 (0.0008)
+[2026-06-06 21:33:56,376][262198] Updated weights for policy 0, policy_version 22551 (0.0009)
+[2026-06-06 21:33:57,423][262198] Updated weights for policy 0, policy_version 22562 (0.0008)
+[2026-06-06 21:33:57,920][261293] Fps is (10 sec: 4915.2, 60 sec: 4505.6, 300 sec: 4956.9). Total num frames: 11552768. Throughput: 0: 4440.0. Samples: 11559856. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:33:57,923][261293] Avg episode reward: [(0, '1636.071')]
+[2026-06-06 21:33:58,623][262198] Updated weights for policy 0, policy_version 22572 (0.0012)
+[2026-06-06 21:34:00,293][262198] Updated weights for policy 0, policy_version 22583 (0.0011)
+[2026-06-06 21:34:01,789][262198] Updated weights for policy 0, policy_version 22593 (0.0009)
+[2026-06-06 21:34:02,906][262198] Updated weights for policy 0, policy_version 22603 (0.0010)
+[2026-06-06 21:34:02,920][261293] Fps is (10 sec: 4301.0, 60 sec: 4403.2, 300 sec: 4929.1). Total num frames: 11571200. Throughput: 0: 4433.6. Samples: 11571316. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:34:02,922][261293] Avg episode reward: [(0, '1546.772')]
+[2026-06-06 21:34:04,325][262198] Updated weights for policy 0, policy_version 22613 (0.0010)
+[2026-06-06 21:34:05,359][262198] Updated weights for policy 0, policy_version 22623 (0.0010)
+[2026-06-06 21:34:06,778][262198] Updated weights for policy 0, policy_version 22634 (0.0009)
+[2026-06-06 21:34:07,859][262198] Updated weights for policy 0, policy_version 22644 (0.0010)
+[2026-06-06 21:34:07,921][261293] Fps is (10 sec: 4095.6, 60 sec: 4369.0, 300 sec: 4929.1). Total num frames: 11593728. Throughput: 0: 4438.9. Samples: 11595828. Policy #0 lag: (min: 1.0, avg: 17.9, max: 36.0)
+[2026-06-06 21:34:07,922][261293] Avg episode reward: [(0, '1639.613')]
+[2026-06-06 21:34:09,247][262198] Updated weights for policy 0, policy_version 22654 (0.0009)
+[2026-06-06 21:34:10,442][262198] Updated weights for policy 0, policy_version 22664 (0.0009)
+[2026-06-06 21:34:12,129][262198] Updated weights for policy 0, policy_version 22675 (0.0009)
+[2026-06-06 21:34:12,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4300.8, 300 sec: 4908.3). Total num frames: 11612160. Throughput: 0: 4429.1. Samples: 11619384. Policy #0 lag: (min: 1.0, avg: 17.9, max: 36.0)
+[2026-06-06 21:34:12,922][261293] Avg episode reward: [(0, '1516.565')]
+[2026-06-06 21:34:13,613][262198] Updated weights for policy 0, policy_version 22686 (0.0009)
+[2026-06-06 21:34:14,763][262198] Updated weights for policy 0, policy_version 22696 (0.0012)
+[2026-06-06 21:34:16,394][262198] Updated weights for policy 0, policy_version 22706 (0.0011)
+[2026-06-06 21:34:17,920][261293] Fps is (10 sec: 3686.8, 60 sec: 4301.0, 300 sec: 4887.4). Total num frames: 11630592. Throughput: 0: 4423.1. Samples: 11630972. Policy #0 lag: (min: 1.0, avg: 17.9, max: 36.0)
+[2026-06-06 21:34:17,921][262198] Updated weights for policy 0, policy_version 22717 (0.0010)
+[2026-06-06 21:34:17,921][261293] Avg episode reward: [(0, '1629.797')]
+[2026-06-06 21:34:18,988][262198] Updated weights for policy 0, policy_version 22727 (0.0010)
+[2026-06-06 21:34:20,804][262198] Updated weights for policy 0, policy_version 22739 (0.0012)
+[2026-06-06 21:34:22,310][262198] Updated weights for policy 0, policy_version 22749 (0.0011)
+[2026-06-06 21:34:22,921][261293] Fps is (10 sec: 3891.1, 60 sec: 4266.7, 300 sec: 4866.6). Total num frames: 11651072. Throughput: 0: 4371.8. Samples: 11653386. Policy #0 lag: (min: 1.0, avg: 18.1, max: 33.0)
+[2026-06-06 21:34:22,923][261293] Avg episode reward: [(0, '1655.199')]
+[2026-06-06 21:34:23,386][262198] Updated weights for policy 0, policy_version 22759 (0.0011)
+[2026-06-06 21:34:25,214][262198] Updated weights for policy 0, policy_version 22769 (0.0010)
+[2026-06-06 21:34:26,334][262198] Updated weights for policy 0, policy_version 22779 (0.0009)
+[2026-06-06 21:34:27,816][262198] Updated weights for policy 0, policy_version 22789 (0.0010)
+[2026-06-06 21:34:27,920][261293] Fps is (10 sec: 3891.1, 60 sec: 4266.7, 300 sec: 4845.8). Total num frames: 11669504. Throughput: 0: 4219.8. Samples: 11675714. Policy #0 lag: (min: 1.0, avg: 18.1, max: 33.0)
+[2026-06-06 21:34:27,922][261293] Avg episode reward: [(0, '1407.484')]
+[2026-06-06 21:34:28,659][262198] Updated weights for policy 0, policy_version 22799 (0.0009)
+[2026-06-06 21:34:30,620][262198] Updated weights for policy 0, policy_version 22809 (0.0009)
+[2026-06-06 21:34:31,432][262198] Updated weights for policy 0, policy_version 22819 (0.0010)
+[2026-06-06 21:34:32,920][261293] Fps is (10 sec: 3686.4, 60 sec: 4300.8, 300 sec: 4804.1). Total num frames: 11687936. Throughput: 0: 4141.5. Samples: 11687130. Policy #0 lag: (min: 1.0, avg: 18.1, max: 33.0)
+[2026-06-06 21:34:32,922][261293] Avg episode reward: [(0, '1595.374')]
+[2026-06-06 21:34:32,935][262198] Updated weights for policy 0, policy_version 22829 (0.0008)
+[2026-06-06 21:34:34,131][262198] Updated weights for policy 0, policy_version 22839 (0.0009)
+[2026-06-06 21:34:35,293][262198] Updated weights for policy 0, policy_version 22851 (0.0009)
+[2026-06-06 21:34:36,439][262198] Updated weights for policy 0, policy_version 22861 (0.0008)
+[2026-06-06 21:34:37,641][262198] Updated weights for policy 0, policy_version 22871 (0.0008)
+[2026-06-06 21:34:37,921][261293] Fps is (10 sec: 4095.8, 60 sec: 4266.7, 300 sec: 4797.2). Total num frames: 11710464. Throughput: 0: 4102.7. Samples: 11712892. Policy #0 lag: (min: 2.0, avg: 20.9, max: 34.0)
+[2026-06-06 21:34:37,924][261293] Avg episode reward: [(0, '1535.384')]
+[2026-06-06 21:34:38,688][262198] Updated weights for policy 0, policy_version 22881 (0.0009)
+[2026-06-06 21:34:39,615][262198] Updated weights for policy 0, policy_version 22891 (0.0009)
+[2026-06-06 21:34:41,060][262198] Updated weights for policy 0, policy_version 22901 (0.0009)
+[2026-06-06 21:34:41,940][262198] Updated weights for policy 0, policy_version 22911 (0.0008)
+[2026-06-06 21:34:42,903][262198] Updated weights for policy 0, policy_version 22921 (0.0008)
+[2026-06-06 21:34:42,920][261293] Fps is (10 sec: 4710.4, 60 sec: 4266.8, 300 sec: 4790.2). Total num frames: 11735040. Throughput: 0: 4044.2. Samples: 11741846. Policy #0 lag: (min: 2.0, avg: 20.9, max: 34.0)
+[2026-06-06 21:34:42,923][261293] Avg episode reward: [(0, '1569.400')]
+[2026-06-06 21:34:43,784][262198] Updated weights for policy 0, policy_version 22931 (0.0009)
+[2026-06-06 21:34:44,879][262198] Updated weights for policy 0, policy_version 22941 (0.0009)
+[2026-06-06 21:34:45,802][262198] Updated weights for policy 0, policy_version 22951 (0.0009)
+[2026-06-06 21:34:46,947][262198] Updated weights for policy 0, policy_version 22961 (0.0008)
+[2026-06-06 21:34:47,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4266.6, 300 sec: 4776.4). Total num frames: 11759616. Throughput: 0: 4139.5. Samples: 11757598. Policy #0 lag: (min: 2.0, avg: 20.9, max: 34.0)
+[2026-06-06 21:34:47,926][261293] Avg episode reward: [(0, '1620.713')]
+[2026-06-06 21:34:47,930][262198] Updated weights for policy 0, policy_version 22971 (0.0008)
+[2026-06-06 21:34:49,167][262198] Updated weights for policy 0, policy_version 22983 (0.0009)
+[2026-06-06 21:34:50,577][262198] Updated weights for policy 0, policy_version 22993 (0.0008)
+[2026-06-06 21:34:51,888][262198] Updated weights for policy 0, policy_version 23003 (0.0009)
+[2026-06-06 21:34:52,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4232.5, 300 sec: 4769.4). Total num frames: 11782144. Throughput: 0: 4206.7. Samples: 11785126. Policy #0 lag: (min: 3.0, avg: 19.7, max: 35.0)
+[2026-06-06 21:34:52,924][261293] Avg episode reward: [(0, '1494.343')]
+[2026-06-06 21:34:53,213][262198] Updated weights for policy 0, policy_version 23014 (0.0012)
+[2026-06-06 21:34:53,788][262198] Updated weights for policy 0, policy_version 23024 (0.0007)
+[2026-06-06 21:34:55,238][262198] Updated weights for policy 0, policy_version 23034 (0.0010)
+[2026-06-06 21:34:56,377][262198] Updated weights for policy 0, policy_version 23047 (0.0009)
+[2026-06-06 21:34:57,586][262198] Updated weights for policy 0, policy_version 23059 (0.0008)
+[2026-06-06 21:34:57,921][261293] Fps is (10 sec: 4710.6, 60 sec: 4232.5, 300 sec: 4762.5). Total num frames: 11806720. Throughput: 0: 4312.0. Samples: 11813426. Policy #0 lag: (min: 3.0, avg: 19.7, max: 35.0)
+[2026-06-06 21:34:57,922][261293] Avg episode reward: [(0, '1724.587')]
+[2026-06-06 21:34:58,936][262198] Updated weights for policy 0, policy_version 23069 (0.0009)
+[2026-06-06 21:34:59,722][262198] Updated weights for policy 0, policy_version 23080 (0.0010)
+[2026-06-06 21:35:00,881][262198] Updated weights for policy 0, policy_version 23090 (0.0009)
+[2026-06-06 21:35:01,894][262198] Updated weights for policy 0, policy_version 23100 (0.0009)
+[2026-06-06 21:35:02,885][262198] Updated weights for policy 0, policy_version 23111 (0.0008)
+[2026-06-06 21:35:02,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4369.0, 300 sec: 4790.2). Total num frames: 11833344. Throughput: 0: 4386.1. Samples: 11828350. Policy #0 lag: (min: 3.0, avg: 19.7, max: 35.0)
+[2026-06-06 21:35:02,924][261293] Avg episode reward: [(0, '1610.596')]
+[2026-06-06 21:35:04,160][262198] Updated weights for policy 0, policy_version 23121 (0.0008)
+[2026-06-06 21:35:05,279][262198] Updated weights for policy 0, policy_version 23131 (0.0008)
+[2026-06-06 21:35:06,488][262198] Updated weights for policy 0, policy_version 23142 (0.0008)
+[2026-06-06 21:35:07,486][262198] Updated weights for policy 0, policy_version 23152 (0.0009)
+[2026-06-06 21:35:07,921][261293] Fps is (10 sec: 4710.5, 60 sec: 4335.0, 300 sec: 4755.5). Total num frames: 11853824. Throughput: 0: 4520.6. Samples: 11856810. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:35:07,923][261293] Avg episode reward: [(0, '1510.557')]
+[2026-06-06 21:35:09,042][262198] Updated weights for policy 0, policy_version 23162 (0.0009)
+[2026-06-06 21:35:09,997][262198] Updated weights for policy 0, policy_version 23174 (0.0009)
+[2026-06-06 21:35:10,985][262198] Updated weights for policy 0, policy_version 23184 (0.0009)
+[2026-06-06 21:35:12,523][262198] Updated weights for policy 0, policy_version 23196 (0.0010)
+[2026-06-06 21:35:12,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4403.2, 300 sec: 4741.6). Total num frames: 11876352. Throughput: 0: 4645.1. Samples: 11884746. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:35:12,923][261293] Avg episode reward: [(0, '1694.782')]
+[2026-06-06 21:35:13,615][262198] Updated weights for policy 0, policy_version 23208 (0.0009)
+[2026-06-06 21:35:14,886][262198] Updated weights for policy 0, policy_version 23218 (0.0011)
+[2026-06-06 21:35:16,004][262198] Updated weights for policy 0, policy_version 23228 (0.0010)
+[2026-06-06 21:35:17,357][262198] Updated weights for policy 0, policy_version 23240 (0.0009)
+[2026-06-06 21:35:17,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4505.6, 300 sec: 4734.7). Total num frames: 11900928. Throughput: 0: 4695.5. Samples: 11898428. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:35:17,922][261293] Avg episode reward: [(0, '1560.005')]
+[2026-06-06 21:35:18,478][262198] Updated weights for policy 0, policy_version 23250 (0.0009)
+[2026-06-06 21:35:19,584][262198] Updated weights for policy 0, policy_version 23260 (0.0008)
+[2026-06-06 21:35:20,632][262198] Updated weights for policy 0, policy_version 23271 (0.0009)
+[2026-06-06 21:35:21,684][262198] Updated weights for policy 0, policy_version 23281 (0.0009)
+[2026-06-06 21:35:22,846][262198] Updated weights for policy 0, policy_version 23291 (0.0009)
+[2026-06-06 21:35:22,921][261293] Fps is (10 sec: 4915.2, 60 sec: 4573.9, 300 sec: 4720.8). Total num frames: 11925504. Throughput: 0: 4764.2. Samples: 11927280. Policy #0 lag: (min: 4.0, avg: 19.7, max: 36.0)
+[2026-06-06 21:35:22,924][261293] Avg episode reward: [(0, '1486.953')]
+[2026-06-06 21:35:23,948][262198] Updated weights for policy 0, policy_version 23301 (0.0009)
+[2026-06-06 21:35:24,784][262198] Updated weights for policy 0, policy_version 23311 (0.0009)
+[2026-06-06 21:35:26,273][262198] Updated weights for policy 0, policy_version 23321 (0.0008)
+[2026-06-06 21:35:27,074][262198] Updated weights for policy 0, policy_version 23335 (0.0010)
+[2026-06-06 21:35:27,921][261293] Fps is (10 sec: 4915.0, 60 sec: 4676.3, 300 sec: 4720.8). Total num frames: 11950080. Throughput: 0: 4795.0. Samples: 11957622. Policy #0 lag: (min: 4.0, avg: 19.7, max: 36.0)
+[2026-06-06 21:35:27,926][261293] Avg episode reward: [(0, '1357.293')]
+[2026-06-06 21:35:28,483][262198] Updated weights for policy 0, policy_version 23346 (0.0014)
+[2026-06-06 21:35:29,945][262198] Updated weights for policy 0, policy_version 23360 (0.0009)
+[2026-06-06 21:35:30,873][262198] Updated weights for policy 0, policy_version 23370 (0.0009)
+[2026-06-06 21:35:31,845][262198] Updated weights for policy 0, policy_version 23380 (0.0009)
+[2026-06-06 21:35:32,920][261293] Fps is (10 sec: 4710.6, 60 sec: 4744.5, 300 sec: 4706.9). Total num frames: 11972608. Throughput: 0: 4747.6. Samples: 11971238. Policy #0 lag: (min: 4.0, avg: 19.7, max: 36.0)
+[2026-06-06 21:35:32,922][261293] Avg episode reward: [(0, '1595.069')]
+[2026-06-06 21:35:33,371][262198] Updated weights for policy 0, policy_version 23393 (0.0010)
+[2026-06-06 21:35:34,238][262198] Updated weights for policy 0, policy_version 23403 (0.0008)
+[2026-06-06 21:35:35,698][262198] Updated weights for policy 0, policy_version 23416 (0.0009)
+[2026-06-06 21:35:36,907][262198] Updated weights for policy 0, policy_version 23428 (0.0009)
+[2026-06-06 21:35:37,920][261293] Fps is (10 sec: 4915.3, 60 sec: 4812.8, 300 sec: 4693.0). Total num frames: 11999232. Throughput: 0: 4782.1. Samples: 12000318. Policy #0 lag: (min: 7.0, avg: 19.4, max: 39.0)
+[2026-06-06 21:35:37,923][261293] Avg episode reward: [(0, '1566.900')]
+[2026-06-06 21:35:38,477][262198] Updated weights for policy 0, policy_version 23440 (0.0010)
+[2026-06-06 21:35:39,645][262198] Updated weights for policy 0, policy_version 23450 (0.0009)
+[2026-06-06 21:35:40,653][262198] Updated weights for policy 0, policy_version 23463 (0.0008)
+[2026-06-06 21:35:41,722][262198] Updated weights for policy 0, policy_version 23475 (0.0009)
+[2026-06-06 21:35:42,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4812.8, 300 sec: 4693.0). Total num frames: 12023808. Throughput: 0: 4841.1. Samples: 12031276. Policy #0 lag: (min: 7.0, avg: 19.4, max: 39.0)
+[2026-06-06 21:35:42,925][261293] Avg episode reward: [(0, '1692.532')]
+[2026-06-06 21:35:43,168][262198] Updated weights for policy 0, policy_version 23488 (0.0009)
+[2026-06-06 21:35:44,024][262198] Updated weights for policy 0, policy_version 23498 (0.0009)
+[2026-06-06 21:35:45,039][262198] Updated weights for policy 0, policy_version 23510 (0.0008)
+[2026-06-06 21:35:46,224][262198] Updated weights for policy 0, policy_version 23523 (0.0009)
+[2026-06-06 21:35:47,573][262198] Updated weights for policy 0, policy_version 23533 (0.0008)
+[2026-06-06 21:35:47,920][261293] Fps is (10 sec: 5324.9, 60 sec: 4881.1, 300 sec: 4700.0). Total num frames: 12052480. Throughput: 0: 4869.9. Samples: 12047492. Policy #0 lag: (min: 7.0, avg: 19.4, max: 39.0)
+[2026-06-06 21:35:47,922][261293] Avg episode reward: [(0, '1718.234')]
+[2026-06-06 21:35:48,166][262198] Updated weights for policy 0, policy_version 23543 (0.0008)
+[2026-06-06 21:35:49,372][262198] Updated weights for policy 0, policy_version 23556 (0.0008)
+[2026-06-06 21:35:50,727][262198] Updated weights for policy 0, policy_version 23569 (0.0009)
+[2026-06-06 21:35:51,906][262198] Updated weights for policy 0, policy_version 23579 (0.0008)
+[2026-06-06 21:35:52,921][261293] Fps is (10 sec: 5324.7, 60 sec: 4915.2, 300 sec: 4700.0). Total num frames: 12077056. Throughput: 0: 4947.7. Samples: 12079456. Policy #0 lag: (min: 7.0, avg: 19.4, max: 39.0)
+[2026-06-06 21:35:52,923][261293] Avg episode reward: [(0, '1608.418')]
+[2026-06-06 21:35:53,133][262198] Updated weights for policy 0, policy_version 23592 (0.0010)
+[2026-06-06 21:35:53,803][262198] Updated weights for policy 0, policy_version 23602 (0.0008)
+[2026-06-06 21:35:54,941][262198] Updated weights for policy 0, policy_version 23612 (0.0008)
+[2026-06-06 21:35:56,410][262198] Updated weights for policy 0, policy_version 23625 (0.0009)
+[2026-06-06 21:35:57,054][262198] Updated weights for policy 0, policy_version 23635 (0.0010)
+[2026-06-06 21:35:57,921][261293] Fps is (10 sec: 5119.7, 60 sec: 4949.3, 300 sec: 4706.9). Total num frames: 12103680. Throughput: 0: 5006.3. Samples: 12110030. Policy #0 lag: (min: 0.0, avg: 19.3, max: 40.0)
+[2026-06-06 21:35:57,924][261293] Avg episode reward: [(0, '1484.809')]
+[2026-06-06 21:35:58,926][262198] Updated weights for policy 0, policy_version 23648 (0.0011)
+[2026-06-06 21:35:59,948][262198] Updated weights for policy 0, policy_version 23658 (0.0009)
+[2026-06-06 21:36:00,673][262198] Updated weights for policy 0, policy_version 23668 (0.0009)
+[2026-06-06 21:36:02,316][262198] Updated weights for policy 0, policy_version 23678 (0.0008)
+[2026-06-06 21:36:02,922][261293] Fps is (10 sec: 4914.9, 60 sec: 4881.0, 300 sec: 4713.9). Total num frames: 12126208. Throughput: 0: 5010.3. Samples: 12123894. Policy #0 lag: (min: 0.0, avg: 19.3, max: 40.0)
+[2026-06-06 21:36:02,928][261293] Avg episode reward: [(0, '1662.229')]
+[2026-06-06 21:36:03,068][262198] Updated weights for policy 0, policy_version 23688 (0.0009)
+[2026-06-06 21:36:03,933][262198] Updated weights for policy 0, policy_version 23698 (0.0008)
+[2026-06-06 21:36:05,180][262198] Updated weights for policy 0, policy_version 23708 (0.0007)
+[2026-06-06 21:36:06,382][262198] Updated weights for policy 0, policy_version 23720 (0.0010)
+[2026-06-06 21:36:06,877][262115] Signal inference workers to stop experience collection... (2800 times)
+[2026-06-06 21:36:06,896][262198] InferenceWorker_p0-w0: stopping experience collection (2800 times)
+[2026-06-06 21:36:06,957][262115] Signal inference workers to resume experience collection... (2800 times)
+[2026-06-06 21:36:06,958][262198] InferenceWorker_p0-w0: resuming experience collection (2800 times)
+[2026-06-06 21:36:07,075][262198] Updated weights for policy 0, policy_version 23732 (0.0009)
+[2026-06-06 21:36:07,921][261293] Fps is (10 sec: 4915.4, 60 sec: 4983.5, 300 sec: 4706.9). Total num frames: 12152832. Throughput: 0: 5043.9. Samples: 12154256. Policy #0 lag: (min: 0.0, avg: 19.3, max: 40.0)
+[2026-06-06 21:36:07,922][261293] Avg episode reward: [(0, '1454.808')]
+[2026-06-06 21:36:08,765][262198] Updated weights for policy 0, policy_version 23742 (0.0009)
+[2026-06-06 21:36:09,789][262198] Updated weights for policy 0, policy_version 23754 (0.0008)
+[2026-06-06 21:36:10,816][262198] Updated weights for policy 0, policy_version 23767 (0.0010)
+[2026-06-06 21:36:12,504][262198] Updated weights for policy 0, policy_version 23777 (0.0008)
+[2026-06-06 21:36:12,921][261293] Fps is (10 sec: 4915.6, 60 sec: 4983.5, 300 sec: 4713.9). Total num frames: 12175360. Throughput: 0: 5002.8. Samples: 12182748. Policy #0 lag: (min: 1.0, avg: 16.8, max: 36.0)
+[2026-06-06 21:36:12,924][261293] Avg episode reward: [(0, '1608.185')]
+[2026-06-06 21:36:13,948][262198] Updated weights for policy 0, policy_version 23789 (0.0010)
+[2026-06-06 21:36:15,519][262198] Updated weights for policy 0, policy_version 23801 (0.0008)
+[2026-06-06 21:36:16,258][262198] Updated weights for policy 0, policy_version 23811 (0.0008)
+[2026-06-06 21:36:17,665][262198] Updated weights for policy 0, policy_version 23825 (0.0009)
+[2026-06-06 21:36:17,921][261293] Fps is (10 sec: 4710.3, 60 sec: 4983.4, 300 sec: 4713.9). Total num frames: 12199936. Throughput: 0: 4996.6. Samples: 12196088. Policy #0 lag: (min: 1.0, avg: 16.8, max: 36.0)
+[2026-06-06 21:36:17,923][261293] Avg episode reward: [(0, '1648.977')]
+[2026-06-06 21:36:19,057][262198] Updated weights for policy 0, policy_version 23835 (0.0010)
+[2026-06-06 21:36:20,150][262198] Updated weights for policy 0, policy_version 23845 (0.0009)
+[2026-06-06 21:36:21,189][262198] Updated weights for policy 0, policy_version 23859 (0.0010)
+[2026-06-06 21:36:22,478][262198] Updated weights for policy 0, policy_version 23869 (0.0008)
+[2026-06-06 21:36:22,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4983.5, 300 sec: 4720.8). Total num frames: 12224512. Throughput: 0: 4967.7. Samples: 12223866. Policy #0 lag: (min: 1.0, avg: 16.8, max: 36.0)
+[2026-06-06 21:36:22,923][261293] Avg episode reward: [(0, '1489.837')]
+[2026-06-06 21:36:23,225][262198] Updated weights for policy 0, policy_version 23879 (0.0008)
+[2026-06-06 21:36:23,973][262198] Updated weights for policy 0, policy_version 23889 (0.0007)
+[2026-06-06 21:36:24,814][262198] Updated weights for policy 0, policy_version 23899 (0.0007)
+[2026-06-06 21:36:26,921][262198] Updated weights for policy 0, policy_version 23921 (0.0008)
+[2026-06-06 21:36:27,820][262198] Updated weights for policy 0, policy_version 23931 (0.0010)
+[2026-06-06 21:36:27,921][261293] Fps is (10 sec: 5324.7, 60 sec: 5051.7, 300 sec: 4734.7). Total num frames: 12253184. Throughput: 0: 5061.3. Samples: 12259038. Policy #0 lag: (min: 0.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:36:27,924][261293] Avg episode reward: [(0, '1595.840')]
+[2026-06-06 21:36:29,272][262198] Updated weights for policy 0, policy_version 23941 (0.0008)
+[2026-06-06 21:36:30,308][262198] Updated weights for policy 0, policy_version 23956 (0.0009)
+[2026-06-06 21:36:31,534][262198] Updated weights for policy 0, policy_version 23966 (0.0008)
+[2026-06-06 21:36:32,433][262198] Updated weights for policy 0, policy_version 23976 (0.0009)
+[2026-06-06 21:36:32,921][261293] Fps is (10 sec: 5324.8, 60 sec: 5085.9, 300 sec: 4727.8). Total num frames: 12277760. Throughput: 0: 5040.3. Samples: 12274304. Policy #0 lag: (min: 0.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:36:32,924][261293] Avg episode reward: [(0, '1581.126')]
+[2026-06-06 21:36:33,714][262198] Updated weights for policy 0, policy_version 23989 (0.0009)
+[2026-06-06 21:36:34,694][262198] Updated weights for policy 0, policy_version 23999 (0.0008)
+[2026-06-06 21:36:35,987][262198] Updated weights for policy 0, policy_version 24009 (0.0009)
+[2026-06-06 21:36:37,131][262198] Updated weights for policy 0, policy_version 24021 (0.0008)
+[2026-06-06 21:36:37,921][261293] Fps is (10 sec: 4915.2, 60 sec: 5051.7, 300 sec: 4734.7). Total num frames: 12302336. Throughput: 0: 4987.2. Samples: 12303882. Policy #0 lag: (min: 0.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:36:37,924][261293] Avg episode reward: [(0, '1598.681')]
+[2026-06-06 21:36:38,160][262198] Updated weights for policy 0, policy_version 24031 (0.0008)
+[2026-06-06 21:36:40,076][262198] Updated weights for policy 0, policy_version 24048 (0.0009)
+[2026-06-06 21:36:41,011][262198] Updated weights for policy 0, policy_version 24058 (0.0008)
+[2026-06-06 21:36:41,785][262198] Updated weights for policy 0, policy_version 24068 (0.0008)
+[2026-06-06 21:36:42,921][261293] Fps is (10 sec: 4915.1, 60 sec: 5051.7, 300 sec: 4734.7). Total num frames: 12326912. Throughput: 0: 4990.0. Samples: 12334578. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 21:36:42,922][261293] Avg episode reward: [(0, '1597.961')]
+[2026-06-06 21:36:43,042][262198] Updated weights for policy 0, policy_version 24082 (0.0011)
+[2026-06-06 21:36:44,832][262198] Updated weights for policy 0, policy_version 24093 (0.0009)
+[2026-06-06 21:36:46,199][262198] Updated weights for policy 0, policy_version 24106 (0.0009)
+[2026-06-06 21:36:47,235][262198] Updated weights for policy 0, policy_version 24118 (0.0009)
+[2026-06-06 21:36:47,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4949.3, 300 sec: 4727.7). Total num frames: 12349440. Throughput: 0: 4982.8. Samples: 12348120. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 21:36:47,924][261293] Avg episode reward: [(0, '1368.171')]
+[2026-06-06 21:36:48,358][262198] Updated weights for policy 0, policy_version 24128 (0.0009)
+[2026-06-06 21:36:50,256][262198] Updated weights for policy 0, policy_version 24145 (0.0010)
+[2026-06-06 21:36:51,533][262198] Updated weights for policy 0, policy_version 24155 (0.0010)
+[2026-06-06 21:36:52,921][261293] Fps is (10 sec: 4505.4, 60 sec: 4915.2, 300 sec: 4720.8). Total num frames: 12371968. Throughput: 0: 4906.8. Samples: 12375066. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 21:36:52,924][261293] Avg episode reward: [(0, '1516.300')]
+[2026-06-06 21:36:52,963][262198] Updated weights for policy 0, policy_version 24165 (0.0010)
+[2026-06-06 21:36:53,748][262198] Updated weights for policy 0, policy_version 24178 (0.0009)
+[2026-06-06 21:36:55,018][262198] Updated weights for policy 0, policy_version 24188 (0.0009)
+[2026-06-06 21:36:56,309][262198] Updated weights for policy 0, policy_version 24198 (0.0008)
+[2026-06-06 21:36:57,223][262198] Updated weights for policy 0, policy_version 24211 (0.0008)
+[2026-06-06 21:36:57,921][261293] Fps is (10 sec: 4915.5, 60 sec: 4915.2, 300 sec: 4713.9). Total num frames: 12398592. Throughput: 0: 4931.5. Samples: 12404668. Policy #0 lag: (min: 1.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:36:57,923][261293] Avg episode reward: [(0, '1602.665')]
+[2026-06-06 21:36:58,502][262198] Updated weights for policy 0, policy_version 24221 (0.0008)
+[2026-06-06 21:36:59,503][262198] Updated weights for policy 0, policy_version 24231 (0.0008)
+[2026-06-06 21:37:00,345][262198] Updated weights for policy 0, policy_version 24243 (0.0008)
+[2026-06-06 21:37:01,433][262198] Updated weights for policy 0, policy_version 24253 (0.0009)
+[2026-06-06 21:37:02,357][262198] Updated weights for policy 0, policy_version 24263 (0.0008)
+[2026-06-06 21:37:02,921][261293] Fps is (10 sec: 5324.7, 60 sec: 4983.5, 300 sec: 4706.9). Total num frames: 12425216. Throughput: 0: 4990.7. Samples: 12420670. Policy #0 lag: (min: 1.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:37:02,923][261293] Avg episode reward: [(0, '1723.216')]
+[2026-06-06 21:37:03,205][262198] Updated weights for policy 0, policy_version 24275 (0.0008)
+[2026-06-06 21:37:04,428][262198] Updated weights for policy 0, policy_version 24285 (0.0009)
+[2026-06-06 21:37:05,311][262198] Updated weights for policy 0, policy_version 24295 (0.0009)
+[2026-06-06 21:37:06,338][262198] Updated weights for policy 0, policy_version 24305 (0.0010)
+[2026-06-06 21:37:07,380][262198] Updated weights for policy 0, policy_version 24315 (0.0010)
+[2026-06-06 21:37:07,921][261293] Fps is (10 sec: 5119.8, 60 sec: 4949.3, 300 sec: 4706.9). Total num frames: 12449792. Throughput: 0: 5098.8. Samples: 12453314. Policy #0 lag: (min: 1.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:37:07,922][261293] Avg episode reward: [(0, '1580.657')]
+[2026-06-06 21:37:08,602][262198] Updated weights for policy 0, policy_version 24325 (0.0009)
+[2026-06-06 21:37:09,892][262198] Updated weights for policy 0, policy_version 24337 (0.0010)
+[2026-06-06 21:37:11,233][262198] Updated weights for policy 0, policy_version 24347 (0.0008)
+[2026-06-06 21:37:12,847][262198] Updated weights for policy 0, policy_version 24357 (0.0010)
+[2026-06-06 21:37:12,921][261293] Fps is (10 sec: 4710.6, 60 sec: 4949.3, 300 sec: 4686.1). Total num frames: 12472320. Throughput: 0: 4886.3. Samples: 12478920. Policy #0 lag: (min: 1.0, avg: 21.0, max: 40.0)
+[2026-06-06 21:37:12,922][261293] Avg episode reward: [(0, '1635.537')]
+[2026-06-06 21:37:13,912][262198] Updated weights for policy 0, policy_version 24369 (0.0009)
+[2026-06-06 21:37:15,231][262198] Updated weights for policy 0, policy_version 24379 (0.0012)
+[2026-06-06 21:37:16,849][262198] Updated weights for policy 0, policy_version 24389 (0.0010)
+[2026-06-06 21:37:17,505][262198] Updated weights for policy 0, policy_version 24399 (0.0009)
+[2026-06-06 21:37:17,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4881.1, 300 sec: 4679.2). Total num frames: 12492800. Throughput: 0: 4827.6. Samples: 12491548. Policy #0 lag: (min: 7.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:37:17,923][261293] Avg episode reward: [(0, '1656.562')]
+[2026-06-06 21:37:19,363][262198] Updated weights for policy 0, policy_version 24409 (0.0010)
+[2026-06-06 21:37:20,357][262198] Updated weights for policy 0, policy_version 24419 (0.0010)
+[2026-06-06 21:37:21,551][262198] Updated weights for policy 0, policy_version 24430 (0.0011)
+[2026-06-06 21:37:22,776][262198] Updated weights for policy 0, policy_version 24440 (0.0012)
+[2026-06-06 21:37:22,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4812.8, 300 sec: 4679.2). Total num frames: 12513280. Throughput: 0: 4705.2. Samples: 12515614. Policy #0 lag: (min: 7.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:37:22,922][261293] Avg episode reward: [(0, '1690.025')]
+[2026-06-06 21:37:24,459][262198] Updated weights for policy 0, policy_version 24450 (0.0009)
+[2026-06-06 21:37:25,208][262198] Updated weights for policy 0, policy_version 24460 (0.0010)
+[2026-06-06 21:37:26,555][262198] Updated weights for policy 0, policy_version 24470 (0.0009)
+[2026-06-06 21:37:27,894][262198] Updated weights for policy 0, policy_version 24480 (0.0010)
+[2026-06-06 21:37:27,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4676.3, 300 sec: 4658.3). Total num frames: 12533760. Throughput: 0: 4575.3. Samples: 12540466. Policy #0 lag: (min: 7.0, avg: 22.2, max: 39.0)
+[2026-06-06 21:37:27,922][261293] Avg episode reward: [(0, '1532.584')]
+[2026-06-06 21:37:29,205][262198] Updated weights for policy 0, policy_version 24492 (0.0012)
+[2026-06-06 21:37:30,703][262198] Updated weights for policy 0, policy_version 24502 (0.0008)
+[2026-06-06 21:37:32,162][262198] Updated weights for policy 0, policy_version 24512 (0.0008)
+[2026-06-06 21:37:32,920][261293] Fps is (10 sec: 3891.2, 60 sec: 4573.9, 300 sec: 4630.6). Total num frames: 12552192. Throughput: 0: 4537.8. Samples: 12552320. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:37:32,922][261293] Avg episode reward: [(0, '1655.194')]
+[2026-06-06 21:37:33,271][262198] Updated weights for policy 0, policy_version 24522 (0.0008)
+[2026-06-06 21:37:34,229][262198] Updated weights for policy 0, policy_version 24532 (0.0009)
+[2026-06-06 21:37:35,987][262198] Updated weights for policy 0, policy_version 24542 (0.0010)
+[2026-06-06 21:37:37,166][262198] Updated weights for policy 0, policy_version 24552 (0.0008)
+[2026-06-06 21:37:37,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4539.8, 300 sec: 4623.6). Total num frames: 12574720. Throughput: 0: 4480.2. Samples: 12576672. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:37:37,923][261293] Avg episode reward: [(0, '1582.902')]
+[2026-06-06 21:37:38,178][262198] Updated weights for policy 0, policy_version 24562 (0.0008)
+[2026-06-06 21:37:39,557][262198] Updated weights for policy 0, policy_version 24572 (0.0009)
+[2026-06-06 21:37:41,141][262198] Updated weights for policy 0, policy_version 24582 (0.0010)
+[2026-06-06 21:37:41,906][262198] Updated weights for policy 0, policy_version 24592 (0.0009)
+[2026-06-06 21:37:42,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4437.3, 300 sec: 4595.8). Total num frames: 12593152. Throughput: 0: 4351.8. Samples: 12600498. Policy #0 lag: (min: 3.0, avg: 21.3, max: 35.0)
+[2026-06-06 21:37:42,921][261293] Avg episode reward: [(0, '1623.908')]
+[2026-06-06 21:37:43,602][262198] Updated weights for policy 0, policy_version 24602 (0.0010)
+[2026-06-06 21:37:44,817][262198] Updated weights for policy 0, policy_version 24612 (0.0009)
+[2026-06-06 21:37:45,961][262198] Updated weights for policy 0, policy_version 24622 (0.0010)
+[2026-06-06 21:37:46,953][262198] Updated weights for policy 0, policy_version 24632 (0.0011)
+[2026-06-06 21:37:47,920][261293] Fps is (10 sec: 3891.3, 60 sec: 4403.2, 300 sec: 4588.9). Total num frames: 12613632. Throughput: 0: 4260.6. Samples: 12612396. Policy #0 lag: (min: 3.0, avg: 19.8, max: 35.0)
+[2026-06-06 21:37:47,922][261293] Avg episode reward: [(0, '1631.530')]
+[2026-06-06 21:37:48,315][262198] Updated weights for policy 0, policy_version 24642 (0.0009)
+[2026-06-06 21:37:48,842][262115] Signal inference workers to stop experience collection... (2850 times)
+[2026-06-06 21:37:48,862][262198] InferenceWorker_p0-w0: stopping experience collection (2850 times)
+[2026-06-06 21:37:48,927][262115] Signal inference workers to resume experience collection... (2850 times)
+[2026-06-06 21:37:48,927][262198] InferenceWorker_p0-w0: resuming experience collection (2850 times)
+[2026-06-06 21:37:49,346][262198] Updated weights for policy 0, policy_version 24655 (0.0008)
+[2026-06-06 21:37:50,595][262198] Updated weights for policy 0, policy_version 24665 (0.0008)
+[2026-06-06 21:37:51,467][262198] Updated weights for policy 0, policy_version 24675 (0.0009)
+[2026-06-06 21:37:52,924][261293] Fps is (10 sec: 4503.9, 60 sec: 4437.1, 300 sec: 4595.8). Total num frames: 12638208. Throughput: 0: 4191.9. Samples: 12641962. Policy #0 lag: (min: 3.0, avg: 19.8, max: 35.0)
+[2026-06-06 21:37:52,927][261293] Avg episode reward: [(0, '1677.672')]
+[2026-06-06 21:37:52,944][262198] Updated weights for policy 0, policy_version 24687 (0.0010)
+[2026-06-06 21:37:54,718][262198] Updated weights for policy 0, policy_version 24698 (0.0010)
+[2026-06-06 21:37:56,311][262198] Updated weights for policy 0, policy_version 24712 (0.0011)
+[2026-06-06 21:37:57,469][262198] Updated weights for policy 0, policy_version 24723 (0.0011)
+[2026-06-06 21:37:57,921][261293] Fps is (10 sec: 4505.5, 60 sec: 4334.9, 300 sec: 4582.0). Total num frames: 12658688. Throughput: 0: 4151.3. Samples: 12665730. Policy #0 lag: (min: 3.0, avg: 19.8, max: 35.0)
+[2026-06-06 21:37:57,922][261293] Avg episode reward: [(0, '1571.157')]
+[2026-06-06 21:37:59,688][262198] Updated weights for policy 0, policy_version 24735 (0.0009)
+[2026-06-06 21:38:01,341][262198] Updated weights for policy 0, policy_version 24748 (0.0012)
+[2026-06-06 21:38:02,920][261293] Fps is (10 sec: 3687.7, 60 sec: 4164.3, 300 sec: 4554.2). Total num frames: 12675072. Throughput: 0: 4089.0. Samples: 12675554. Policy #0 lag: (min: 3.0, avg: 19.8, max: 35.0)
+[2026-06-06 21:38:02,922][261293] Avg episode reward: [(0, '1449.835')]
+[2026-06-06 21:38:03,224][262198] Updated weights for policy 0, policy_version 24758 (0.0008)
+[2026-06-06 21:38:04,505][262198] Updated weights for policy 0, policy_version 24768 (0.0011)
+[2026-06-06 21:38:06,301][262198] Updated weights for policy 0, policy_version 24781 (0.0010)
+[2026-06-06 21:38:07,409][262198] Updated weights for policy 0, policy_version 24791 (0.0009)
+[2026-06-06 21:38:07,920][261293] Fps is (10 sec: 3481.8, 60 sec: 4061.9, 300 sec: 4540.3). Total num frames: 12693504. Throughput: 0: 4035.8. Samples: 12697226. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 21:38:07,921][261293] Avg episode reward: [(0, '1575.621')]
+[2026-06-06 21:38:09,282][262198] Updated weights for policy 0, policy_version 24802 (0.0011)
+[2026-06-06 21:38:10,871][262198] Updated weights for policy 0, policy_version 24814 (0.0011)
+[2026-06-06 21:38:11,833][262198] Updated weights for policy 0, policy_version 24824 (0.0011)
+[2026-06-06 21:38:12,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3993.6, 300 sec: 4540.4). Total num frames: 12711936. Throughput: 0: 3989.1. Samples: 12719974. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 21:38:12,923][261293] Avg episode reward: [(0, '1523.093')]
+[2026-06-06 21:38:13,411][262198] Updated weights for policy 0, policy_version 24834 (0.0009)
+[2026-06-06 21:38:14,674][262198] Updated weights for policy 0, policy_version 24845 (0.0009)
+[2026-06-06 21:38:15,480][262198] Updated weights for policy 0, policy_version 24855 (0.0007)
+[2026-06-06 21:38:16,580][262198] Updated weights for policy 0, policy_version 24865 (0.0008)
+[2026-06-06 21:38:17,856][262198] Updated weights for policy 0, policy_version 24878 (0.0009)
+[2026-06-06 21:38:17,921][261293] Fps is (10 sec: 4505.3, 60 sec: 4096.0, 300 sec: 4554.2). Total num frames: 12738560. Throughput: 0: 4025.9. Samples: 12733488. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 21:38:17,924][261293] Avg episode reward: [(0, '1497.449')]
+[2026-06-06 21:38:18,686][262198] Updated weights for policy 0, policy_version 24888 (0.0009)
+[2026-06-06 21:38:19,780][262198] Updated weights for policy 0, policy_version 24899 (0.0009)
+[2026-06-06 21:38:20,900][262198] Updated weights for policy 0, policy_version 24909 (0.0008)
+[2026-06-06 21:38:21,839][262198] Updated weights for policy 0, policy_version 24919 (0.0008)
+[2026-06-06 21:38:22,921][261293] Fps is (10 sec: 4915.0, 60 sec: 4130.1, 300 sec: 4568.1). Total num frames: 12761088. Throughput: 0: 4179.3. Samples: 12764742. Policy #0 lag: (min: 1.0, avg: 19.9, max: 41.0)
+[2026-06-06 21:38:22,924][261293] Avg episode reward: [(0, '1786.261')]
+[2026-06-06 21:38:23,007][262115] Saving new best policy, reward=1786.261!
+[2026-06-06 21:38:23,167][262198] Updated weights for policy 0, policy_version 24931 (0.0009)
+[2026-06-06 21:38:24,140][262198] Updated weights for policy 0, policy_version 24941 (0.0008)
+[2026-06-06 21:38:25,012][262198] Updated weights for policy 0, policy_version 24951 (0.0008)
+[2026-06-06 21:38:26,271][262198] Updated weights for policy 0, policy_version 24962 (0.0009)
+[2026-06-06 21:38:27,162][262198] Updated weights for policy 0, policy_version 24972 (0.0009)
+[2026-06-06 21:38:27,920][261293] Fps is (10 sec: 4915.5, 60 sec: 4232.5, 300 sec: 4602.8). Total num frames: 12787712. Throughput: 0: 4309.0. Samples: 12794402. Policy #0 lag: (min: 1.0, avg: 19.9, max: 41.0)
+[2026-06-06 21:38:27,922][261293] Avg episode reward: [(0, '1585.674')]
+[2026-06-06 21:38:28,459][262198] Updated weights for policy 0, policy_version 24982 (0.0009)
+[2026-06-06 21:38:29,599][262198] Updated weights for policy 0, policy_version 24995 (0.0009)
+[2026-06-06 21:38:30,746][262198] Updated weights for policy 0, policy_version 25005 (0.0008)
+[2026-06-06 21:38:31,535][262198] Updated weights for policy 0, policy_version 25015 (0.0009)
+[2026-06-06 21:38:32,809][262198] Updated weights for policy 0, policy_version 25025 (0.0008)
+[2026-06-06 21:38:32,920][261293] Fps is (10 sec: 5325.1, 60 sec: 4369.1, 300 sec: 4609.8). Total num frames: 12814336. Throughput: 0: 4385.5. Samples: 12809744. Policy #0 lag: (min: 1.0, avg: 19.9, max: 41.0)
+[2026-06-06 21:38:32,922][261293] Avg episode reward: [(0, '1496.591')]
+[2026-06-06 21:38:33,619][262198] Updated weights for policy 0, policy_version 25035 (0.0011)
+[2026-06-06 21:38:34,778][262198] Updated weights for policy 0, policy_version 25045 (0.0009)
+[2026-06-06 21:38:35,984][262198] Updated weights for policy 0, policy_version 25058 (0.0009)
+[2026-06-06 21:38:36,873][262198] Updated weights for policy 0, policy_version 25068 (0.0009)
+[2026-06-06 21:38:37,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4369.1, 300 sec: 4602.8). Total num frames: 12836864. Throughput: 0: 4405.2. Samples: 12840180. Policy #0 lag: (min: 1.0, avg: 19.9, max: 41.0)
+[2026-06-06 21:38:37,924][261293] Avg episode reward: [(0, '1710.613')]
+[2026-06-06 21:38:38,172][262198] Updated weights for policy 0, policy_version 25080 (0.0011)
+[2026-06-06 21:38:39,325][262198] Updated weights for policy 0, policy_version 25092 (0.0009)
+[2026-06-06 21:38:40,436][262198] Updated weights for policy 0, policy_version 25102 (0.0008)
+[2026-06-06 21:38:41,160][262198] Updated weights for policy 0, policy_version 25112 (0.0008)
+[2026-06-06 21:38:42,306][262198] Updated weights for policy 0, policy_version 25124 (0.0011)
+[2026-06-06 21:38:42,920][261293] Fps is (10 sec: 5120.0, 60 sec: 4539.7, 300 sec: 4616.7). Total num frames: 12865536. Throughput: 0: 4578.7. Samples: 12871768. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 21:38:42,922][261293] Avg episode reward: [(0, '1582.863')]
+[2026-06-06 21:38:43,584][262198] Updated weights for policy 0, policy_version 25134 (0.0010)
+[2026-06-06 21:38:44,635][262198] Updated weights for policy 0, policy_version 25144 (0.0012)
+[2026-06-06 21:38:45,774][262198] Updated weights for policy 0, policy_version 25155 (0.0008)
+[2026-06-06 21:38:47,102][262198] Updated weights for policy 0, policy_version 25165 (0.0009)
+[2026-06-06 21:38:47,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4573.9, 300 sec: 4609.7). Total num frames: 12888064. Throughput: 0: 4658.3. Samples: 12885176. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 21:38:47,924][261293] Avg episode reward: [(0, '1706.264')]
+[2026-06-06 21:38:48,053][262198] Updated weights for policy 0, policy_version 25175 (0.0008)
+[2026-06-06 21:38:49,283][262198] Updated weights for policy 0, policy_version 25187 (0.0009)
+[2026-06-06 21:38:50,695][262198] Updated weights for policy 0, policy_version 25197 (0.0011)
+[2026-06-06 21:38:51,845][262198] Updated weights for policy 0, policy_version 25207 (0.0010)
+[2026-06-06 21:38:52,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4505.9, 300 sec: 4595.8). Total num frames: 12908544. Throughput: 0: 4785.6. Samples: 12912578. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 21:38:52,924][261293] Avg episode reward: [(0, '1381.707')]
+[2026-06-06 21:38:53,248][262198] Updated weights for policy 0, policy_version 25217 (0.0008)
+[2026-06-06 21:38:54,362][262198] Updated weights for policy 0, policy_version 25227 (0.0009)
+[2026-06-06 21:38:55,737][262198] Updated weights for policy 0, policy_version 25237 (0.0010)
+[2026-06-06 21:38:56,764][262198] Updated weights for policy 0, policy_version 25247 (0.0008)
+[2026-06-06 21:38:57,922][261293] Fps is (10 sec: 4300.4, 60 sec: 4539.7, 300 sec: 4609.7). Total num frames: 12931072. Throughput: 0: 4837.6. Samples: 12937670. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 21:38:57,925][261293] Avg episode reward: [(0, '1552.733')]
+[2026-06-06 21:38:58,029][262198] Updated weights for policy 0, policy_version 25257 (0.0010)
+[2026-06-06 21:38:59,199][262198] Updated weights for policy 0, policy_version 25267 (0.0008)
+[2026-06-06 21:39:00,080][262198] Updated weights for policy 0, policy_version 25277 (0.0010)
+[2026-06-06 21:39:00,679][262198] Updated weights for policy 0, policy_version 25287 (0.0008)
+[2026-06-06 21:39:01,912][262198] Updated weights for policy 0, policy_version 25297 (0.0008)
+[2026-06-06 21:39:02,782][262198] Updated weights for policy 0, policy_version 25307 (0.0008)
+[2026-06-06 21:39:02,921][261293] Fps is (10 sec: 4915.0, 60 sec: 4710.4, 300 sec: 4623.6). Total num frames: 12957696. Throughput: 0: 4873.5. Samples: 12952796. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 21:39:02,924][261293] Avg episode reward: [(0, '1582.245')]
+[2026-06-06 21:39:03,600][262198] Updated weights for policy 0, policy_version 25319 (0.0007)
+[2026-06-06 21:39:05,008][262198] Updated weights for policy 0, policy_version 25329 (0.0008)
+[2026-06-06 21:39:05,862][262198] Updated weights for policy 0, policy_version 25340 (0.0008)
+[2026-06-06 21:39:06,743][262198] Updated weights for policy 0, policy_version 25352 (0.0009)
+[2026-06-06 21:39:07,921][261293] Fps is (10 sec: 5325.3, 60 sec: 4846.9, 300 sec: 4651.4). Total num frames: 12984320. Throughput: 0: 4912.2. Samples: 12985792. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 21:39:07,922][261293] Avg episode reward: [(0, '1761.818')]
+[2026-06-06 21:39:08,284][262198] Updated weights for policy 0, policy_version 25364 (0.0008)
+[2026-06-06 21:39:09,335][262198] Updated weights for policy 0, policy_version 25374 (0.0008)
+[2026-06-06 21:39:10,159][262198] Updated weights for policy 0, policy_version 25386 (0.0008)
+[2026-06-06 21:39:11,523][262198] Updated weights for policy 0, policy_version 25398 (0.0009)
+[2026-06-06 21:39:12,451][262198] Updated weights for policy 0, policy_version 25408 (0.0007)
+[2026-06-06 21:39:12,921][261293] Fps is (10 sec: 5324.7, 60 sec: 4983.4, 300 sec: 4679.1). Total num frames: 13010944. Throughput: 0: 4954.9. Samples: 13017378. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:39:12,924][261293] Avg episode reward: [(0, '1639.965')]
+[2026-06-06 21:39:13,167][262198] Updated weights for policy 0, policy_version 25418 (0.0010)
+[2026-06-06 21:39:14,552][262198] Updated weights for policy 0, policy_version 25430 (0.0009)
+[2026-06-06 21:39:15,530][262198] Updated weights for policy 0, policy_version 25440 (0.0008)
+[2026-06-06 21:39:16,442][262198] Updated weights for policy 0, policy_version 25450 (0.0014)
+[2026-06-06 21:39:17,754][262198] Updated weights for policy 0, policy_version 25463 (0.0009)
+[2026-06-06 21:39:17,921][261293] Fps is (10 sec: 5324.7, 60 sec: 4983.5, 300 sec: 4700.0). Total num frames: 13037568. Throughput: 0: 4967.0. Samples: 13033262. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:39:17,924][261293] Avg episode reward: [(0, '1828.769')]
+[2026-06-06 21:39:17,931][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000025464_13037568.pth...
+[2026-06-06 21:39:17,987][262115] Saving new best policy, reward=1828.769!
+[2026-06-06 21:39:18,832][262198] Updated weights for policy 0, policy_version 25473 (0.0009)
+[2026-06-06 21:39:19,939][262198] Updated weights for policy 0, policy_version 25485 (0.0009)
+[2026-06-06 21:39:21,570][262198] Updated weights for policy 0, policy_version 25498 (0.0009)
+[2026-06-06 21:39:22,921][261293] Fps is (10 sec: 4915.4, 60 sec: 4983.5, 300 sec: 4713.9). Total num frames: 13060096. Throughput: 0: 4966.2. Samples: 13063660. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:39:22,924][261293] Avg episode reward: [(0, '1419.562')]
+[2026-06-06 21:39:23,113][262198] Updated weights for policy 0, policy_version 25513 (0.0009)
+[2026-06-06 21:39:24,046][262198] Updated weights for policy 0, policy_version 25523 (0.0008)
+[2026-06-06 21:39:25,161][262198] Updated weights for policy 0, policy_version 25533 (0.0010)
+[2026-06-06 21:39:26,096][262198] Updated weights for policy 0, policy_version 25546 (0.0009)
+[2026-06-06 21:39:27,301][262198] Updated weights for policy 0, policy_version 25558 (0.0009)
+[2026-06-06 21:39:27,920][261293] Fps is (10 sec: 5120.3, 60 sec: 5017.6, 300 sec: 4748.6). Total num frames: 13088768. Throughput: 0: 4956.7. Samples: 13094818. Policy #0 lag: (min: 1.0, avg: 17.0, max: 33.0)
+[2026-06-06 21:39:27,922][261293] Avg episode reward: [(0, '1584.395')]
+[2026-06-06 21:39:28,291][262198] Updated weights for policy 0, policy_version 25568 (0.0008)
+[2026-06-06 21:39:29,663][262198] Updated weights for policy 0, policy_version 25582 (0.0009)
+[2026-06-06 21:39:31,084][262198] Updated weights for policy 0, policy_version 25594 (0.0009)
+[2026-06-06 21:39:31,798][262198] Updated weights for policy 0, policy_version 25604 (0.0007)
+[2026-06-06 21:39:32,647][262198] Updated weights for policy 0, policy_version 25616 (0.0009)
+[2026-06-06 21:39:32,921][261293] Fps is (10 sec: 5529.5, 60 sec: 5017.6, 300 sec: 4762.5). Total num frames: 13115392. Throughput: 0: 4995.0. Samples: 13109954. Policy #0 lag: (min: 1.0, avg: 17.2, max: 33.0)
+[2026-06-06 21:39:32,925][261293] Avg episode reward: [(0, '1575.823')]
+[2026-06-06 21:39:34,394][262198] Updated weights for policy 0, policy_version 25628 (0.0009)
+[2026-06-06 21:39:35,475][262198] Updated weights for policy 0, policy_version 25641 (0.0008)
+[2026-06-06 21:39:36,835][262198] Updated weights for policy 0, policy_version 25653 (0.0008)
+[2026-06-06 21:39:37,921][261293] Fps is (10 sec: 5119.8, 60 sec: 5051.7, 300 sec: 4762.5). Total num frames: 13139968. Throughput: 0: 5082.0. Samples: 13141270. Policy #0 lag: (min: 1.0, avg: 17.2, max: 33.0)
+[2026-06-06 21:39:37,923][261293] Avg episode reward: [(0, '1587.348')]
+[2026-06-06 21:39:38,066][262198] Updated weights for policy 0, policy_version 25665 (0.0009)
+[2026-06-06 21:39:38,840][262198] Updated weights for policy 0, policy_version 25677 (0.0008)
+[2026-06-06 21:39:40,717][262198] Updated weights for policy 0, policy_version 25693 (0.0009)
+[2026-06-06 21:39:41,434][262115] Signal inference workers to stop experience collection... (2900 times)
+[2026-06-06 21:39:41,458][262198] InferenceWorker_p0-w0: stopping experience collection (2900 times)
+[2026-06-06 21:39:41,496][262115] Signal inference workers to resume experience collection... (2900 times)
+[2026-06-06 21:39:41,497][262198] InferenceWorker_p0-w0: resuming experience collection (2900 times)
+[2026-06-06 21:39:41,836][262198] Updated weights for policy 0, policy_version 25707 (0.0009)
+[2026-06-06 21:39:42,920][261293] Fps is (10 sec: 4915.5, 60 sec: 4983.5, 300 sec: 4762.5). Total num frames: 13164544. Throughput: 0: 5216.0. Samples: 13172382. Policy #0 lag: (min: 1.0, avg: 17.2, max: 33.0)
+[2026-06-06 21:39:42,922][261293] Avg episode reward: [(0, '1573.951')]
+[2026-06-06 21:39:43,227][262198] Updated weights for policy 0, policy_version 25720 (0.0010)
+[2026-06-06 21:39:44,584][262198] Updated weights for policy 0, policy_version 25732 (0.0009)
+[2026-06-06 21:39:45,818][262198] Updated weights for policy 0, policy_version 25744 (0.0009)
+[2026-06-06 21:39:47,806][262198] Updated weights for policy 0, policy_version 25756 (0.0010)
+[2026-06-06 21:39:47,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4983.5, 300 sec: 4762.5). Total num frames: 13187072. Throughput: 0: 5189.3. Samples: 13186312. Policy #0 lag: (min: 2.0, avg: 19.8, max: 34.0)
+[2026-06-06 21:39:47,924][261293] Avg episode reward: [(0, '1696.935')]
+[2026-06-06 21:39:48,954][262198] Updated weights for policy 0, policy_version 25766 (0.0009)
+[2026-06-06 21:39:50,460][262198] Updated weights for policy 0, policy_version 25780 (0.0010)
+[2026-06-06 21:39:51,506][262198] Updated weights for policy 0, policy_version 25792 (0.0009)
+[2026-06-06 21:39:52,534][262198] Updated weights for policy 0, policy_version 25802 (0.0009)
+[2026-06-06 21:39:52,922][261293] Fps is (10 sec: 4709.7, 60 sec: 5051.6, 300 sec: 4762.4). Total num frames: 13211648. Throughput: 0: 5062.2. Samples: 13213598. Policy #0 lag: (min: 2.0, avg: 19.8, max: 34.0)
+[2026-06-06 21:39:52,926][261293] Avg episode reward: [(0, '1483.043')]
+[2026-06-06 21:39:54,216][262198] Updated weights for policy 0, policy_version 25814 (0.0012)
+[2026-06-06 21:39:55,442][262198] Updated weights for policy 0, policy_version 25824 (0.0010)
+[2026-06-06 21:39:56,502][262198] Updated weights for policy 0, policy_version 25835 (0.0009)
+[2026-06-06 21:39:57,910][262198] Updated weights for policy 0, policy_version 25845 (0.0009)
+[2026-06-06 21:39:57,921][261293] Fps is (10 sec: 4505.5, 60 sec: 5017.7, 300 sec: 4741.6). Total num frames: 13232128. Throughput: 0: 4934.1. Samples: 13239410. Policy #0 lag: (min: 2.0, avg: 19.8, max: 34.0)
+[2026-06-06 21:39:57,924][261293] Avg episode reward: [(0, '1561.199')]
+[2026-06-06 21:39:59,040][262198] Updated weights for policy 0, policy_version 25856 (0.0011)
+[2026-06-06 21:40:00,206][262198] Updated weights for policy 0, policy_version 25866 (0.0012)
+[2026-06-06 21:40:01,726][262198] Updated weights for policy 0, policy_version 25876 (0.0009)
+[2026-06-06 21:40:02,921][261293] Fps is (10 sec: 4096.4, 60 sec: 4915.2, 300 sec: 4741.6). Total num frames: 13252608. Throughput: 0: 4851.6. Samples: 13251584. Policy #0 lag: (min: 2.0, avg: 19.8, max: 34.0)
+[2026-06-06 21:40:02,924][261293] Avg episode reward: [(0, '1653.016')]
+[2026-06-06 21:40:03,017][262198] Updated weights for policy 0, policy_version 25886 (0.0010)
+[2026-06-06 21:40:04,217][262198] Updated weights for policy 0, policy_version 25897 (0.0010)
+[2026-06-06 21:40:05,591][262198] Updated weights for policy 0, policy_version 25907 (0.0009)
+[2026-06-06 21:40:06,755][262198] Updated weights for policy 0, policy_version 25917 (0.0010)
+[2026-06-06 21:40:07,793][262198] Updated weights for policy 0, policy_version 25927 (0.0009)
+[2026-06-06 21:40:07,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4846.9, 300 sec: 4741.6). Total num frames: 13275136. Throughput: 0: 4756.9. Samples: 13277718. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 21:40:07,922][261293] Avg episode reward: [(0, '1559.698')]
+[2026-06-06 21:40:09,307][262198] Updated weights for policy 0, policy_version 25937 (0.0009)
+[2026-06-06 21:40:10,247][262198] Updated weights for policy 0, policy_version 25947 (0.0010)
+[2026-06-06 21:40:11,617][262198] Updated weights for policy 0, policy_version 25957 (0.0009)
+[2026-06-06 21:40:12,364][262198] Updated weights for policy 0, policy_version 25967 (0.0009)
+[2026-06-06 21:40:12,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4744.6, 300 sec: 4727.7). Total num frames: 13295616. Throughput: 0: 4618.9. Samples: 13302672. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 21:40:12,923][261293] Avg episode reward: [(0, '1671.511')]
+[2026-06-06 21:40:13,972][262198] Updated weights for policy 0, policy_version 25979 (0.0009)
+[2026-06-06 21:40:15,175][262198] Updated weights for policy 0, policy_version 25989 (0.0010)
+[2026-06-06 21:40:15,925][262198] Updated weights for policy 0, policy_version 25999 (0.0011)
+[2026-06-06 21:40:17,778][262198] Updated weights for policy 0, policy_version 26012 (0.0009)
+[2026-06-06 21:40:17,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4676.3, 300 sec: 4720.8). Total num frames: 13318144. Throughput: 0: 4599.9. Samples: 13316948. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 21:40:17,923][261293] Avg episode reward: [(0, '1773.620')]
+[2026-06-06 21:40:19,344][262198] Updated weights for policy 0, policy_version 26023 (0.0013)
+[2026-06-06 21:40:20,989][262198] Updated weights for policy 0, policy_version 26033 (0.0010)
+[2026-06-06 21:40:22,232][262198] Updated weights for policy 0, policy_version 26045 (0.0010)
+[2026-06-06 21:40:22,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4642.2, 300 sec: 4706.9). Total num frames: 13338624. Throughput: 0: 4446.2. Samples: 13341348. Policy #0 lag: (min: 3.0, avg: 20.2, max: 35.0)
+[2026-06-06 21:40:22,922][261293] Avg episode reward: [(0, '1631.663')]
+[2026-06-06 21:40:23,210][262198] Updated weights for policy 0, policy_version 26055 (0.0009)
+[2026-06-06 21:40:24,620][262198] Updated weights for policy 0, policy_version 26065 (0.0010)
+[2026-06-06 21:40:25,302][262198] Updated weights for policy 0, policy_version 26075 (0.0009)
+[2026-06-06 21:40:26,946][262198] Updated weights for policy 0, policy_version 26085 (0.0009)
+[2026-06-06 21:40:27,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4539.7, 300 sec: 4706.9). Total num frames: 13361152. Throughput: 0: 4341.4. Samples: 13367748. Policy #0 lag: (min: 3.0, avg: 20.2, max: 35.0)
+[2026-06-06 21:40:27,925][261293] Avg episode reward: [(0, '1778.209')]
+[2026-06-06 21:40:28,125][262198] Updated weights for policy 0, policy_version 26097 (0.0009)
+[2026-06-06 21:40:28,802][262198] Updated weights for policy 0, policy_version 26107 (0.0008)
+[2026-06-06 21:40:29,947][262198] Updated weights for policy 0, policy_version 26117 (0.0008)
+[2026-06-06 21:40:30,753][262198] Updated weights for policy 0, policy_version 26127 (0.0009)
+[2026-06-06 21:40:32,003][262198] Updated weights for policy 0, policy_version 26137 (0.0008)
+[2026-06-06 21:40:32,880][262198] Updated weights for policy 0, policy_version 26148 (0.0009)
+[2026-06-06 21:40:32,920][261293] Fps is (10 sec: 4915.2, 60 sec: 4539.8, 300 sec: 4706.9). Total num frames: 13387776. Throughput: 0: 4371.0. Samples: 13383006. Policy #0 lag: (min: 3.0, avg: 20.2, max: 35.0)
+[2026-06-06 21:40:32,922][261293] Avg episode reward: [(0, '1517.247')]
+[2026-06-06 21:40:33,936][262198] Updated weights for policy 0, policy_version 26158 (0.0008)
+[2026-06-06 21:40:34,890][262198] Updated weights for policy 0, policy_version 26168 (0.0009)
+[2026-06-06 21:40:36,104][262198] Updated weights for policy 0, policy_version 26180 (0.0009)
+[2026-06-06 21:40:37,265][262198] Updated weights for policy 0, policy_version 26190 (0.0009)
+[2026-06-06 21:40:37,921][261293] Fps is (10 sec: 4915.2, 60 sec: 4505.6, 300 sec: 4700.0). Total num frames: 13410304. Throughput: 0: 4439.3. Samples: 13413362. Policy #0 lag: (min: 3.0, avg: 20.2, max: 35.0)
+[2026-06-06 21:40:37,924][261293] Avg episode reward: [(0, '1545.031')]
+[2026-06-06 21:40:38,645][262198] Updated weights for policy 0, policy_version 26201 (0.0008)
+[2026-06-06 21:40:39,811][262198] Updated weights for policy 0, policy_version 26213 (0.0009)
+[2026-06-06 21:40:40,586][262198] Updated weights for policy 0, policy_version 26224 (0.0009)
+[2026-06-06 21:40:41,954][262198] Updated weights for policy 0, policy_version 26234 (0.0009)
+[2026-06-06 21:40:42,920][261293] Fps is (10 sec: 4915.4, 60 sec: 4539.7, 300 sec: 4693.0). Total num frames: 13436928. Throughput: 0: 4536.2. Samples: 13443538. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:40:42,921][261293] Avg episode reward: [(0, '1708.386')]
+[2026-06-06 21:40:43,148][262198] Updated weights for policy 0, policy_version 26246 (0.0010)
+[2026-06-06 21:40:43,891][262198] Updated weights for policy 0, policy_version 26256 (0.0008)
+[2026-06-06 21:40:45,333][262198] Updated weights for policy 0, policy_version 26266 (0.0009)
+[2026-06-06 21:40:46,477][262198] Updated weights for policy 0, policy_version 26278 (0.0010)
+[2026-06-06 21:40:47,206][262198] Updated weights for policy 0, policy_version 26288 (0.0009)
+[2026-06-06 21:40:47,921][261293] Fps is (10 sec: 4915.3, 60 sec: 4539.7, 300 sec: 4686.1). Total num frames: 13459456. Throughput: 0: 4586.9. Samples: 13457994. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:40:47,925][261293] Avg episode reward: [(0, '1642.763')]
+[2026-06-06 21:40:48,735][262198] Updated weights for policy 0, policy_version 26300 (0.0009)
+[2026-06-06 21:40:50,137][262198] Updated weights for policy 0, policy_version 26311 (0.0010)
+[2026-06-06 21:40:51,330][262198] Updated weights for policy 0, policy_version 26321 (0.0009)
+[2026-06-06 21:40:52,506][262198] Updated weights for policy 0, policy_version 26333 (0.0011)
+[2026-06-06 21:40:52,921][261293] Fps is (10 sec: 4914.9, 60 sec: 4573.9, 300 sec: 4686.1). Total num frames: 13486080. Throughput: 0: 4658.0. Samples: 13487330. Policy #0 lag: (min: 3.0, avg: 19.1, max: 35.0)
+[2026-06-06 21:40:52,923][261293] Avg episode reward: [(0, '1533.928')]
+[2026-06-06 21:40:53,670][262198] Updated weights for policy 0, policy_version 26344 (0.0009)
+[2026-06-06 21:40:54,915][262198] Updated weights for policy 0, policy_version 26356 (0.0010)
+[2026-06-06 21:40:56,108][262198] Updated weights for policy 0, policy_version 26367 (0.0012)
+[2026-06-06 21:40:57,351][262198] Updated weights for policy 0, policy_version 26377 (0.0009)
+[2026-06-06 21:40:57,921][261293] Fps is (10 sec: 4915.2, 60 sec: 4608.0, 300 sec: 4686.1). Total num frames: 13508608. Throughput: 0: 4704.3. Samples: 13514364. Policy #0 lag: (min: 3.0, avg: 18.2, max: 35.0)
+[2026-06-06 21:40:57,923][261293] Avg episode reward: [(0, '1692.486')]
+[2026-06-06 21:40:58,738][262198] Updated weights for policy 0, policy_version 26390 (0.0010)
+[2026-06-06 21:40:59,868][262198] Updated weights for policy 0, policy_version 26400 (0.0008)
+[2026-06-06 21:41:01,405][262198] Updated weights for policy 0, policy_version 26412 (0.0010)
+[2026-06-06 21:41:02,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4608.0, 300 sec: 4665.3). Total num frames: 13529088. Throughput: 0: 4681.7. Samples: 13527624. Policy #0 lag: (min: 3.0, avg: 18.2, max: 35.0)
+[2026-06-06 21:41:02,923][261293] Avg episode reward: [(0, '1678.849')]
+[2026-06-06 21:41:02,942][262198] Updated weights for policy 0, policy_version 26425 (0.0008)
+[2026-06-06 21:41:03,141][262115] Signal inference workers to stop experience collection... (2950 times)
+[2026-06-06 21:41:03,161][262115] Signal inference workers to resume experience collection... (2950 times)
+[2026-06-06 21:41:03,166][262198] InferenceWorker_p0-w0: stopping experience collection (2950 times)
+[2026-06-06 21:41:03,182][262198] InferenceWorker_p0-w0: resuming experience collection (2950 times)
+[2026-06-06 21:41:04,111][262198] Updated weights for policy 0, policy_version 26437 (0.0009)
+[2026-06-06 21:41:05,431][262198] Updated weights for policy 0, policy_version 26453 (0.0009)
+[2026-06-06 21:41:06,647][262198] Updated weights for policy 0, policy_version 26463 (0.0008)
+[2026-06-06 21:41:07,805][262198] Updated weights for policy 0, policy_version 26473 (0.0009)
+[2026-06-06 21:41:07,921][261293] Fps is (10 sec: 4710.4, 60 sec: 4676.3, 300 sec: 4679.2). Total num frames: 13555712. Throughput: 0: 4805.0. Samples: 13557576. Policy #0 lag: (min: 3.0, avg: 18.2, max: 35.0)
+[2026-06-06 21:41:07,923][261293] Avg episode reward: [(0, '1691.379')]
+[2026-06-06 21:41:08,972][262198] Updated weights for policy 0, policy_version 26486 (0.0009)
+[2026-06-06 21:41:10,242][262198] Updated weights for policy 0, policy_version 26497 (0.0009)
+[2026-06-06 21:41:11,251][262198] Updated weights for policy 0, policy_version 26507 (0.0009)
+[2026-06-06 21:41:12,393][262198] Updated weights for policy 0, policy_version 26519 (0.0009)
+[2026-06-06 21:41:12,921][261293] Fps is (10 sec: 4915.2, 60 sec: 4710.4, 300 sec: 4672.2). Total num frames: 13578240. Throughput: 0: 4848.5. Samples: 13585930. Policy #0 lag: (min: 3.0, avg: 18.2, max: 35.0)
+[2026-06-06 21:41:12,923][261293] Avg episode reward: [(0, '1697.765')]
+[2026-06-06 21:41:13,739][262198] Updated weights for policy 0, policy_version 26531 (0.0009)
+[2026-06-06 21:41:15,050][262198] Updated weights for policy 0, policy_version 26541 (0.0010)
+[2026-06-06 21:41:16,601][262198] Updated weights for policy 0, policy_version 26553 (0.0009)
+[2026-06-06 21:41:17,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4710.4, 300 sec: 4665.3). Total num frames: 13600768. Throughput: 0: 4817.4. Samples: 13599790. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:41:17,922][261293] Avg episode reward: [(0, '1550.243')]
+[2026-06-06 21:41:18,099][262198] Updated weights for policy 0, policy_version 26565 (0.0011)
+[2026-06-06 21:41:18,739][262198] Updated weights for policy 0, policy_version 26575 (0.0009)
+[2026-06-06 21:41:20,200][262198] Updated weights for policy 0, policy_version 26585 (0.0010)
+[2026-06-06 21:41:21,640][262198] Updated weights for policy 0, policy_version 26597 (0.0010)
+[2026-06-06 21:41:22,371][262198] Updated weights for policy 0, policy_version 26607 (0.0011)
+[2026-06-06 21:41:22,920][261293] Fps is (10 sec: 4505.7, 60 sec: 4744.5, 300 sec: 4644.5). Total num frames: 13623296. Throughput: 0: 4745.1. Samples: 13626888. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:41:22,922][261293] Avg episode reward: [(0, '1696.514')]
+[2026-06-06 21:41:23,927][262198] Updated weights for policy 0, policy_version 26617 (0.0010)
+[2026-06-06 21:41:25,044][262198] Updated weights for policy 0, policy_version 26630 (0.0008)
+[2026-06-06 21:41:25,626][262198] Updated weights for policy 0, policy_version 26640 (0.0008)
+[2026-06-06 21:41:27,175][262198] Updated weights for policy 0, policy_version 26652 (0.0009)
+[2026-06-06 21:41:27,920][261293] Fps is (10 sec: 4915.2, 60 sec: 4812.8, 300 sec: 4651.4). Total num frames: 13649920. Throughput: 0: 4738.1. Samples: 13656752. Policy #0 lag: (min: 3.0, avg: 19.6, max: 36.0)
+[2026-06-06 21:41:27,922][261293] Avg episode reward: [(0, '1605.442')]
+[2026-06-06 21:41:27,988][262198] Updated weights for policy 0, policy_version 26662 (0.0009)
+[2026-06-06 21:41:28,709][262198] Updated weights for policy 0, policy_version 26672 (0.0012)
+[2026-06-06 21:41:29,938][262198] Updated weights for policy 0, policy_version 26684 (0.0008)
+[2026-06-06 21:41:31,475][262198] Updated weights for policy 0, policy_version 26697 (0.0008)
+[2026-06-06 21:41:32,524][262198] Updated weights for policy 0, policy_version 26710 (0.0011)
+[2026-06-06 21:41:32,922][261293] Fps is (10 sec: 5324.5, 60 sec: 4812.8, 300 sec: 4658.3). Total num frames: 13676544. Throughput: 0: 4793.6. Samples: 13673708. Policy #0 lag: (min: 9.0, avg: 22.5, max: 39.0)
+[2026-06-06 21:41:32,926][261293] Avg episode reward: [(0, '1759.798')]
+[2026-06-06 21:41:34,101][262198] Updated weights for policy 0, policy_version 26723 (0.0011)
+[2026-06-06 21:41:35,231][262198] Updated weights for policy 0, policy_version 26733 (0.0008)
+[2026-06-06 21:41:36,369][262198] Updated weights for policy 0, policy_version 26745 (0.0011)
+[2026-06-06 21:41:37,488][262198] Updated weights for policy 0, policy_version 26757 (0.0009)
+[2026-06-06 21:41:37,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4847.0, 300 sec: 4658.3). Total num frames: 13701120. Throughput: 0: 4790.5. Samples: 13702902. Policy #0 lag: (min: 9.0, avg: 22.5, max: 39.0)
+[2026-06-06 21:41:37,923][261293] Avg episode reward: [(0, '1484.794')]
+[2026-06-06 21:41:38,437][262198] Updated weights for policy 0, policy_version 26767 (0.0009)
+[2026-06-06 21:41:39,495][262198] Updated weights for policy 0, policy_version 26777 (0.0008)
+[2026-06-06 21:41:40,572][262198] Updated weights for policy 0, policy_version 26789 (0.0009)
+[2026-06-06 21:41:41,608][262198] Updated weights for policy 0, policy_version 26799 (0.0009)
+[2026-06-06 21:41:42,740][262198] Updated weights for policy 0, policy_version 26809 (0.0009)
+[2026-06-06 21:41:42,921][261293] Fps is (10 sec: 5120.3, 60 sec: 4846.9, 300 sec: 4672.2). Total num frames: 13727744. Throughput: 0: 4875.0. Samples: 13733740. Policy #0 lag: (min: 9.0, avg: 22.5, max: 39.0)
+[2026-06-06 21:41:42,924][261293] Avg episode reward: [(0, '1752.341')]
+[2026-06-06 21:41:43,572][262198] Updated weights for policy 0, policy_version 26819 (0.0009)
+[2026-06-06 21:41:44,992][262198] Updated weights for policy 0, policy_version 26829 (0.0009)
+[2026-06-06 21:41:45,612][262198] Updated weights for policy 0, policy_version 26839 (0.0008)
+[2026-06-06 21:41:47,107][262198] Updated weights for policy 0, policy_version 26852 (0.0011)
+[2026-06-06 21:41:47,920][261293] Fps is (10 sec: 4915.2, 60 sec: 4846.9, 300 sec: 4672.2). Total num frames: 13750272. Throughput: 0: 4902.2. Samples: 13748224. Policy #0 lag: (min: 9.0, avg: 22.5, max: 39.0)
+[2026-06-06 21:41:47,923][261293] Avg episode reward: [(0, '1567.000')]
+[2026-06-06 21:41:48,386][262198] Updated weights for policy 0, policy_version 26862 (0.0009)
+[2026-06-06 21:41:49,067][262198] Updated weights for policy 0, policy_version 26872 (0.0009)
+[2026-06-06 21:41:50,518][262198] Updated weights for policy 0, policy_version 26884 (0.0011)
+[2026-06-06 21:41:51,742][262198] Updated weights for policy 0, policy_version 26894 (0.0008)
+[2026-06-06 21:41:52,748][262198] Updated weights for policy 0, policy_version 26906 (0.0009)
+[2026-06-06 21:41:52,921][261293] Fps is (10 sec: 4915.0, 60 sec: 4846.9, 300 sec: 4672.2). Total num frames: 13776896. Throughput: 0: 4881.8. Samples: 13777260. Policy #0 lag: (min: 2.0, avg: 21.8, max: 38.0)
+[2026-06-06 21:41:52,924][261293] Avg episode reward: [(0, '1631.128')]
+[2026-06-06 21:41:53,562][262198] Updated weights for policy 0, policy_version 26916 (0.0009)
+[2026-06-06 21:41:54,786][262198] Updated weights for policy 0, policy_version 26926 (0.0008)
+[2026-06-06 21:41:55,416][262198] Updated weights for policy 0, policy_version 26936 (0.0008)
+[2026-06-06 21:41:56,449][262198] Updated weights for policy 0, policy_version 26946 (0.0008)
+[2026-06-06 21:41:57,591][262198] Updated weights for policy 0, policy_version 26956 (0.0008)
+[2026-06-06 21:41:57,921][261293] Fps is (10 sec: 5119.9, 60 sec: 4881.1, 300 sec: 4665.3). Total num frames: 13801472. Throughput: 0: 4963.3. Samples: 13809278. Policy #0 lag: (min: 2.0, avg: 21.8, max: 38.0)
+[2026-06-06 21:41:57,923][261293] Avg episode reward: [(0, '1562.643')]
+[2026-06-06 21:41:58,881][262198] Updated weights for policy 0, policy_version 26969 (0.0008)
+[2026-06-06 21:41:59,792][262198] Updated weights for policy 0, policy_version 26979 (0.0008)
+[2026-06-06 21:42:01,120][262198] Updated weights for policy 0, policy_version 26989 (0.0009)
+[2026-06-06 21:42:02,161][262198] Updated weights for policy 0, policy_version 27003 (0.0009)
+[2026-06-06 21:42:02,921][261293] Fps is (10 sec: 5120.1, 60 sec: 4983.4, 300 sec: 4672.2). Total num frames: 13828096. Throughput: 0: 4994.7. Samples: 13824552. Policy #0 lag: (min: 2.0, avg: 21.8, max: 38.0)
+[2026-06-06 21:42:02,926][261293] Avg episode reward: [(0, '1661.950')]
+[2026-06-06 21:42:03,319][262198] Updated weights for policy 0, policy_version 27013 (0.0008)
+[2026-06-06 21:42:04,384][262198] Updated weights for policy 0, policy_version 27023 (0.0008)
+[2026-06-06 21:42:05,165][262198] Updated weights for policy 0, policy_version 27033 (0.0008)
+[2026-06-06 21:42:06,303][262198] Updated weights for policy 0, policy_version 27043 (0.0008)
+[2026-06-06 21:42:07,655][262198] Updated weights for policy 0, policy_version 27053 (0.0008)
+[2026-06-06 21:42:07,921][261293] Fps is (10 sec: 4915.2, 60 sec: 4915.2, 300 sec: 4672.2). Total num frames: 13850624. Throughput: 0: 5078.6. Samples: 13855428. Policy #0 lag: (min: 2.0, avg: 21.8, max: 38.0)
+[2026-06-06 21:42:07,922][261293] Avg episode reward: [(0, '1720.947')]
+[2026-06-06 21:42:09,210][262198] Updated weights for policy 0, policy_version 27070 (0.0009)
+[2026-06-06 21:42:10,292][262198] Updated weights for policy 0, policy_version 27081 (0.0009)
+[2026-06-06 21:42:11,697][262198] Updated weights for policy 0, policy_version 27097 (0.0009)
+[2026-06-06 21:42:12,853][262198] Updated weights for policy 0, policy_version 27107 (0.0008)
+[2026-06-06 21:42:12,921][261293] Fps is (10 sec: 5120.0, 60 sec: 5017.6, 300 sec: 4700.0). Total num frames: 13879296. Throughput: 0: 5078.6. Samples: 13885288. Policy #0 lag: (min: 7.0, avg: 18.8, max: 36.0)
+[2026-06-06 21:42:12,924][261293] Avg episode reward: [(0, '1649.504')]
+[2026-06-06 21:42:14,343][262198] Updated weights for policy 0, policy_version 27117 (0.0011)
+[2026-06-06 21:42:14,942][262115] Signal inference workers to stop experience collection... (3000 times)
+[2026-06-06 21:42:14,948][262115] Signal inference workers to resume experience collection... (3000 times)
+[2026-06-06 21:42:14,961][262198] InferenceWorker_p0-w0: stopping experience collection (3000 times)
+[2026-06-06 21:42:14,978][262198] InferenceWorker_p0-w0: resuming experience collection (3000 times)
+[2026-06-06 21:42:15,028][262198] Updated weights for policy 0, policy_version 27130 (0.0009)
+[2026-06-06 21:42:16,187][262198] Updated weights for policy 0, policy_version 27140 (0.0007)
+[2026-06-06 21:42:17,506][262198] Updated weights for policy 0, policy_version 27150 (0.0008)
+[2026-06-06 21:42:17,920][261293] Fps is (10 sec: 5120.1, 60 sec: 5017.6, 300 sec: 4706.9). Total num frames: 13901824. Throughput: 0: 5017.3. Samples: 13899482. Policy #0 lag: (min: 7.0, avg: 18.8, max: 36.0)
+[2026-06-06 21:42:17,922][261293] Avg episode reward: [(0, '1682.290')]
+[2026-06-06 21:42:18,669][262198] Updated weights for policy 0, policy_version 27165 (0.0009)
+[2026-06-06 21:42:19,649][262198] Updated weights for policy 0, policy_version 27175 (0.0009)
+[2026-06-06 21:42:21,040][262198] Updated weights for policy 0, policy_version 27185 (0.0007)
+[2026-06-06 21:42:21,839][262198] Updated weights for policy 0, policy_version 27197 (0.0010)
+[2026-06-06 21:42:22,758][262198] Updated weights for policy 0, policy_version 27207 (0.0008)
+[2026-06-06 21:42:22,921][261293] Fps is (10 sec: 5119.9, 60 sec: 5120.0, 300 sec: 4734.7). Total num frames: 13930496. Throughput: 0: 5068.5. Samples: 13930984. Policy #0 lag: (min: 7.0, avg: 18.8, max: 36.0)
+[2026-06-06 21:42:22,923][261293] Avg episode reward: [(0, '1617.089')]
+[2026-06-06 21:42:24,172][262198] Updated weights for policy 0, policy_version 27220 (0.0008)
+[2026-06-06 21:42:24,902][262198] Updated weights for policy 0, policy_version 27230 (0.0007)
+[2026-06-06 21:42:25,830][262198] Updated weights for policy 0, policy_version 27240 (0.0007)
+[2026-06-06 21:42:27,135][262198] Updated weights for policy 0, policy_version 27252 (0.0008)
+[2026-06-06 21:42:27,806][262198] Updated weights for policy 0, policy_version 27262 (0.0008)
+[2026-06-06 21:42:27,921][261293] Fps is (10 sec: 5734.3, 60 sec: 5154.1, 300 sec: 4769.4). Total num frames: 13959168. Throughput: 0: 5116.6. Samples: 13963988. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 21:42:27,923][261293] Avg episode reward: [(0, '1598.752')]
+[2026-06-06 21:42:28,695][262198] Updated weights for policy 0, policy_version 27272 (0.0008)
+[2026-06-06 21:42:30,430][262198] Updated weights for policy 0, policy_version 27282 (0.0011)
+[2026-06-06 21:42:31,496][262198] Updated weights for policy 0, policy_version 27293 (0.0012)
+[2026-06-06 21:42:32,755][262198] Updated weights for policy 0, policy_version 27303 (0.0010)
+[2026-06-06 21:42:32,922][261293] Fps is (10 sec: 4915.0, 60 sec: 5051.7, 300 sec: 4762.5). Total num frames: 13979648. Throughput: 0: 5114.7. Samples: 13978388. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 21:42:32,932][261293] Avg episode reward: [(0, '1722.582')]
+[2026-06-06 21:42:34,533][262198] Updated weights for policy 0, policy_version 27313 (0.0009)
+[2026-06-06 21:42:35,344][262198] Updated weights for policy 0, policy_version 27324 (0.0011)
+[2026-06-06 21:42:37,121][262198] Updated weights for policy 0, policy_version 27334 (0.0010)
+[2026-06-06 21:42:37,920][261293] Fps is (10 sec: 3891.3, 60 sec: 4949.3, 300 sec: 4762.5). Total num frames: 13998080. Throughput: 0: 4974.4. Samples: 14001104. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 21:42:37,923][261293] Avg episode reward: [(0, '1685.732')]
+[2026-06-06 21:42:38,418][262198] Updated weights for policy 0, policy_version 27344 (0.0010)
+[2026-06-06 21:42:39,524][262198] Updated weights for policy 0, policy_version 27354 (0.0010)
+[2026-06-06 21:42:41,027][262198] Updated weights for policy 0, policy_version 27364 (0.0011)
+[2026-06-06 21:42:42,836][262198] Updated weights for policy 0, policy_version 27374 (0.0012)
+[2026-06-06 21:42:42,921][261293] Fps is (10 sec: 3686.6, 60 sec: 4812.8, 300 sec: 4755.5). Total num frames: 14016512. Throughput: 0: 4767.7. Samples: 14023826. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 21:42:42,923][261293] Avg episode reward: [(0, '1748.796')]
+[2026-06-06 21:42:44,171][262198] Updated weights for policy 0, policy_version 27385 (0.0011)
+[2026-06-06 21:42:45,684][262198] Updated weights for policy 0, policy_version 27395 (0.0010)
+[2026-06-06 21:42:47,355][262198] Updated weights for policy 0, policy_version 27405 (0.0011)
+[2026-06-06 21:42:47,920][261293] Fps is (10 sec: 3481.6, 60 sec: 4710.4, 300 sec: 4727.8). Total num frames: 14032896. Throughput: 0: 4664.8. Samples: 14034466. Policy #0 lag: (min: 1.0, avg: 19.1, max: 34.0)
+[2026-06-06 21:42:47,922][261293] Avg episode reward: [(0, '1700.016')]
+[2026-06-06 21:42:48,161][262198] Updated weights for policy 0, policy_version 27416 (0.0010)
+[2026-06-06 21:42:49,988][262198] Updated weights for policy 0, policy_version 27426 (0.0009)
+[2026-06-06 21:42:50,807][262198] Updated weights for policy 0, policy_version 27436 (0.0009)
+[2026-06-06 21:42:52,342][262198] Updated weights for policy 0, policy_version 27446 (0.0010)
+[2026-06-06 21:42:52,921][261293] Fps is (10 sec: 3686.3, 60 sec: 4608.0, 300 sec: 4727.8). Total num frames: 14053376. Throughput: 0: 4485.4. Samples: 14057274. Policy #0 lag: (min: 1.0, avg: 19.1, max: 34.0)
+[2026-06-06 21:42:52,925][261293] Avg episode reward: [(0, '1741.641')]
+[2026-06-06 21:42:54,322][262198] Updated weights for policy 0, policy_version 27457 (0.0010)
+[2026-06-06 21:42:55,099][262198] Updated weights for policy 0, policy_version 27468 (0.0011)
+[2026-06-06 21:42:56,411][262198] Updated weights for policy 0, policy_version 27480 (0.0009)
+[2026-06-06 21:42:57,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4539.7, 300 sec: 4741.6). Total num frames: 14073856. Throughput: 0: 4373.3. Samples: 14082084. Policy #0 lag: (min: 1.0, avg: 19.1, max: 34.0)
+[2026-06-06 21:42:57,922][261293] Avg episode reward: [(0, '1490.955')]
+[2026-06-06 21:42:57,986][262198] Updated weights for policy 0, policy_version 27491 (0.0010)
+[2026-06-06 21:42:59,518][262198] Updated weights for policy 0, policy_version 27501 (0.0010)
+[2026-06-06 21:43:00,665][262198] Updated weights for policy 0, policy_version 27513 (0.0010)
+[2026-06-06 21:43:02,168][262198] Updated weights for policy 0, policy_version 27523 (0.0012)
+[2026-06-06 21:43:02,920][261293] Fps is (10 sec: 4096.2, 60 sec: 4437.3, 300 sec: 4748.6). Total num frames: 14094336. Throughput: 0: 4338.1. Samples: 14094696. Policy #0 lag: (min: 1.0, avg: 19.1, max: 34.0)
+[2026-06-06 21:43:02,922][261293] Avg episode reward: [(0, '1637.538')]
+[2026-06-06 21:43:04,080][262198] Updated weights for policy 0, policy_version 27533 (0.0012)
+[2026-06-06 21:43:04,928][262198] Updated weights for policy 0, policy_version 27544 (0.0010)
+[2026-06-06 21:43:07,249][262198] Updated weights for policy 0, policy_version 27554 (0.0016)
+[2026-06-06 21:43:07,921][261293] Fps is (10 sec: 3686.2, 60 sec: 4334.9, 300 sec: 4741.6). Total num frames: 14110720. Throughput: 0: 4080.8. Samples: 14114620. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:43:07,925][261293] Avg episode reward: [(0, '1478.386')]
+[2026-06-06 21:43:08,232][262198] Updated weights for policy 0, policy_version 27564 (0.0012)
+[2026-06-06 21:43:10,303][262198] Updated weights for policy 0, policy_version 27575 (0.0012)
+[2026-06-06 21:43:12,115][262198] Updated weights for policy 0, policy_version 27586 (0.0010)
+[2026-06-06 21:43:12,907][262198] Updated weights for policy 0, policy_version 27596 (0.0010)
+[2026-06-06 21:43:12,920][261293] Fps is (10 sec: 3481.6, 60 sec: 4164.3, 300 sec: 4713.9). Total num frames: 14129152. Throughput: 0: 3794.8. Samples: 14134754. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:43:12,922][261293] Avg episode reward: [(0, '1602.797')]
+[2026-06-06 21:43:14,178][262198] Updated weights for policy 0, policy_version 27606 (0.0009)
+[2026-06-06 21:43:15,123][262198] Updated weights for policy 0, policy_version 27616 (0.0009)
+[2026-06-06 21:43:16,850][262198] Updated weights for policy 0, policy_version 27626 (0.0008)
+[2026-06-06 21:43:17,921][261293] Fps is (10 sec: 3686.4, 60 sec: 4096.0, 300 sec: 4700.0). Total num frames: 14147584. Throughput: 0: 3758.2. Samples: 14147506. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:43:17,923][261293] Avg episode reward: [(0, '1663.441')]
+[2026-06-06 21:43:18,229][262198] Updated weights for policy 0, policy_version 27638 (0.0010)
+[2026-06-06 21:43:18,958][262198] Updated weights for policy 0, policy_version 27648 (0.0009)
+[2026-06-06 21:43:21,068][262198] Updated weights for policy 0, policy_version 27660 (0.0012)
+[2026-06-06 21:43:22,730][262198] Updated weights for policy 0, policy_version 27671 (0.0010)
+[2026-06-06 21:43:22,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3959.5, 300 sec: 4679.1). Total num frames: 14168064. Throughput: 0: 3772.7. Samples: 14170876. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:43:22,924][261293] Avg episode reward: [(0, '1711.185')]
+[2026-06-06 21:43:24,360][262198] Updated weights for policy 0, policy_version 27681 (0.0010)
+[2026-06-06 21:43:25,329][262198] Updated weights for policy 0, policy_version 27691 (0.0008)
+[2026-06-06 21:43:26,700][262198] Updated weights for policy 0, policy_version 27701 (0.0008)
+[2026-06-06 21:43:27,787][262198] Updated weights for policy 0, policy_version 27711 (0.0009)
+[2026-06-06 21:43:27,920][261293] Fps is (10 sec: 4096.2, 60 sec: 3822.9, 300 sec: 4658.3). Total num frames: 14188544. Throughput: 0: 3793.9. Samples: 14194552. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:43:27,923][261293] Avg episode reward: [(0, '1665.613')]
+[2026-06-06 21:43:29,366][262198] Updated weights for policy 0, policy_version 27721 (0.0009)
+[2026-06-06 21:43:30,708][262198] Updated weights for policy 0, policy_version 27731 (0.0010)
+[2026-06-06 21:43:31,921][262198] Updated weights for policy 0, policy_version 27741 (0.0010)
+[2026-06-06 21:43:32,921][261293] Fps is (10 sec: 3891.4, 60 sec: 3788.8, 300 sec: 4644.4). Total num frames: 14206976. Throughput: 0: 3829.4. Samples: 14206790. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:43:32,923][261293] Avg episode reward: [(0, '1617.741')]
+[2026-06-06 21:43:33,245][262198] Updated weights for policy 0, policy_version 27751 (0.0011)
+[2026-06-06 21:43:35,048][262198] Updated weights for policy 0, policy_version 27761 (0.0011)
+[2026-06-06 21:43:36,013][262198] Updated weights for policy 0, policy_version 27772 (0.0009)
+[2026-06-06 21:43:37,855][262198] Updated weights for policy 0, policy_version 27782 (0.0011)
+[2026-06-06 21:43:37,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3788.8, 300 sec: 4609.7). Total num frames: 14225408. Throughput: 0: 3809.2. Samples: 14228684. Policy #0 lag: (min: 3.0, avg: 17.5, max: 35.0)
+[2026-06-06 21:43:37,923][261293] Avg episode reward: [(0, '1673.276')]
+[2026-06-06 21:43:39,287][262198] Updated weights for policy 0, policy_version 27792 (0.0008)
+[2026-06-06 21:43:40,652][262198] Updated weights for policy 0, policy_version 27803 (0.0011)
+[2026-06-06 21:43:42,683][262198] Updated weights for policy 0, policy_version 27813 (0.0011)
+[2026-06-06 21:43:42,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3754.7, 300 sec: 4588.9). Total num frames: 14241792. Throughput: 0: 3716.4. Samples: 14249320. Policy #0 lag: (min: 3.0, avg: 21.3, max: 38.0)
+[2026-06-06 21:43:42,924][261293] Avg episode reward: [(0, '1599.611')]
+[2026-06-06 21:43:43,920][262198] Updated weights for policy 0, policy_version 27823 (0.0012)
+[2026-06-06 21:43:45,379][262198] Updated weights for policy 0, policy_version 27833 (0.0012)
+[2026-06-06 21:43:46,932][262198] Updated weights for policy 0, policy_version 27843 (0.0010)
+[2026-06-06 21:43:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3754.6, 300 sec: 4575.0). Total num frames: 14258176. Throughput: 0: 3667.1. Samples: 14259718. Policy #0 lag: (min: 3.0, avg: 21.3, max: 38.0)
+[2026-06-06 21:43:47,924][261293] Avg episode reward: [(0, '1793.956')]
+[2026-06-06 21:43:48,876][262198] Updated weights for policy 0, policy_version 27853 (0.0014)
+[2026-06-06 21:43:49,809][262198] Updated weights for policy 0, policy_version 27863 (0.0011)
+[2026-06-06 21:43:51,755][262198] Updated weights for policy 0, policy_version 27873 (0.0011)
+[2026-06-06 21:43:52,921][261293] Fps is (10 sec: 3276.6, 60 sec: 3686.4, 300 sec: 4554.2). Total num frames: 14274560. Throughput: 0: 3666.7. Samples: 14279622. Policy #0 lag: (min: 3.0, avg: 21.3, max: 38.0)
+[2026-06-06 21:43:52,924][261293] Avg episode reward: [(0, '1618.407')]
+[2026-06-06 21:43:53,034][262198] Updated weights for policy 0, policy_version 27883 (0.0010)
+[2026-06-06 21:43:54,711][262198] Updated weights for policy 0, policy_version 27894 (0.0013)
+[2026-06-06 21:43:55,860][262198] Updated weights for policy 0, policy_version 27904 (0.0010)
+[2026-06-06 21:43:57,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3618.1, 300 sec: 4519.5). Total num frames: 14290944. Throughput: 0: 3658.6. Samples: 14299396. Policy #0 lag: (min: 3.0, avg: 21.3, max: 38.0)
+[2026-06-06 21:43:57,923][261293] Avg episode reward: [(0, '1603.063')]
+[2026-06-06 21:43:57,980][262198] Updated weights for policy 0, policy_version 27914 (0.0009)
+[2026-06-06 21:43:59,600][262198] Updated weights for policy 0, policy_version 27924 (0.0014)
+[2026-06-06 21:44:01,075][262198] Updated weights for policy 0, policy_version 27935 (0.0012)
+[2026-06-06 21:44:02,920][261293] Fps is (10 sec: 3072.2, 60 sec: 3515.7, 300 sec: 4477.8). Total num frames: 14305280. Throughput: 0: 3579.5. Samples: 14308580. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:44:02,922][261293] Avg episode reward: [(0, '1458.059')]
+[2026-06-06 21:44:04,127][262198] Updated weights for policy 0, policy_version 27945 (0.0012)
+[2026-06-06 21:44:05,965][262198] Updated weights for policy 0, policy_version 27955 (0.0016)
+[2026-06-06 21:44:07,676][262198] Updated weights for policy 0, policy_version 27965 (0.0012)
+[2026-06-06 21:44:07,920][261293] Fps is (10 sec: 2867.5, 60 sec: 3481.6, 300 sec: 4436.2). Total num frames: 14319616. Throughput: 0: 3389.9. Samples: 14323418. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:44:07,922][261293] Avg episode reward: [(0, '1867.607')]
+[2026-06-06 21:44:07,927][262115] Saving new best policy, reward=1867.607!
+[2026-06-06 21:44:08,965][262198] Updated weights for policy 0, policy_version 27975 (0.0009)
+[2026-06-06 21:44:10,519][262198] Updated weights for policy 0, policy_version 27985 (0.0010)
+[2026-06-06 21:44:11,853][262198] Updated weights for policy 0, policy_version 27997 (0.0013)
+[2026-06-06 21:44:12,921][261293] Fps is (10 sec: 3071.9, 60 sec: 3447.4, 300 sec: 4401.5). Total num frames: 14336000. Throughput: 0: 3337.7. Samples: 14344748. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:44:12,923][261293] Avg episode reward: [(0, '1483.774')]
+[2026-06-06 21:44:13,503][262198] Updated weights for policy 0, policy_version 28007 (0.0012)
+[2026-06-06 21:44:15,091][262198] Updated weights for policy 0, policy_version 28017 (0.0011)
+[2026-06-06 21:44:16,142][262198] Updated weights for policy 0, policy_version 28028 (0.0010)
+[2026-06-06 21:44:17,440][262198] Updated weights for policy 0, policy_version 28038 (0.0009)
+[2026-06-06 21:44:17,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3481.6, 300 sec: 4394.5). Total num frames: 14356480. Throughput: 0: 3324.3. Samples: 14356384. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 21:44:17,924][261293] Avg episode reward: [(0, '1616.432')]
+[2026-06-06 21:44:19,152][262198] Updated weights for policy 0, policy_version 28049 (0.0011)
+[2026-06-06 21:44:20,311][262198] Updated weights for policy 0, policy_version 28061 (0.0008)
+[2026-06-06 21:44:21,784][262198] Updated weights for policy 0, policy_version 28071 (0.0010)
+[2026-06-06 21:44:22,923][261293] Fps is (10 sec: 3890.3, 60 sec: 3447.3, 300 sec: 4359.8). Total num frames: 14374912. Throughput: 0: 3358.9. Samples: 14379844. Policy #0 lag: (min: 3.0, avg: 21.3, max: 39.0)
+[2026-06-06 21:44:22,926][261293] Avg episode reward: [(0, '1569.758')]
+[2026-06-06 21:44:23,543][262198] Updated weights for policy 0, policy_version 28081 (0.0010)
+[2026-06-06 21:44:24,622][262198] Updated weights for policy 0, policy_version 28093 (0.0010)
+[2026-06-06 21:44:26,040][262198] Updated weights for policy 0, policy_version 28103 (0.0010)
+[2026-06-06 21:44:27,888][262198] Updated weights for policy 0, policy_version 28113 (0.0008)
+[2026-06-06 21:44:27,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3413.3, 300 sec: 4332.1). Total num frames: 14393344. Throughput: 0: 3411.4. Samples: 14402834. Policy #0 lag: (min: 3.0, avg: 21.3, max: 39.0)
+[2026-06-06 21:44:27,923][261293] Avg episode reward: [(0, '1578.541')]
+[2026-06-06 21:44:29,163][262198] Updated weights for policy 0, policy_version 28126 (0.0009)
+[2026-06-06 21:44:30,534][262198] Updated weights for policy 0, policy_version 28136 (0.0008)
+[2026-06-06 21:44:32,288][262198] Updated weights for policy 0, policy_version 28146 (0.0011)
+[2026-06-06 21:44:32,921][261293] Fps is (10 sec: 3892.0, 60 sec: 3447.4, 300 sec: 4318.2). Total num frames: 14413824. Throughput: 0: 3424.4. Samples: 14413816. Policy #0 lag: (min: 3.0, avg: 21.3, max: 39.0)
+[2026-06-06 21:44:32,925][261293] Avg episode reward: [(0, '1625.136')]
+[2026-06-06 21:44:33,507][262198] Updated weights for policy 0, policy_version 28157 (0.0011)
+[2026-06-06 21:44:34,618][262198] Updated weights for policy 0, policy_version 28167 (0.0008)
+[2026-06-06 21:44:36,543][262198] Updated weights for policy 0, policy_version 28177 (0.0010)
+[2026-06-06 21:44:37,752][262198] Updated weights for policy 0, policy_version 28189 (0.0009)
+[2026-06-06 21:44:37,921][261293] Fps is (10 sec: 4095.9, 60 sec: 3481.6, 300 sec: 4304.3). Total num frames: 14434304. Throughput: 0: 3501.5. Samples: 14437188. Policy #0 lag: (min: 2.0, avg: 18.9, max: 37.0)
+[2026-06-06 21:44:37,924][261293] Avg episode reward: [(0, '1702.841')]
+[2026-06-06 21:44:38,775][262198] Updated weights for policy 0, policy_version 28199 (0.0008)
+[2026-06-06 21:44:40,393][262198] Updated weights for policy 0, policy_version 28209 (0.0008)
+[2026-06-06 21:44:41,282][262198] Updated weights for policy 0, policy_version 28220 (0.0008)
+[2026-06-06 21:44:42,742][262198] Updated weights for policy 0, policy_version 28230 (0.0008)
+[2026-06-06 21:44:42,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3549.9, 300 sec: 4297.3). Total num frames: 14454784. Throughput: 0: 3598.9. Samples: 14461342. Policy #0 lag: (min: 2.0, avg: 18.9, max: 37.0)
+[2026-06-06 21:44:42,922][261293] Avg episode reward: [(0, '1745.411')]
+[2026-06-06 21:44:44,163][262198] Updated weights for policy 0, policy_version 28240 (0.0010)
+[2026-06-06 21:44:45,703][262198] Updated weights for policy 0, policy_version 28250 (0.0011)
+[2026-06-06 21:44:47,073][262198] Updated weights for policy 0, policy_version 28260 (0.0012)
+[2026-06-06 21:44:47,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3549.9, 300 sec: 4269.6). Total num frames: 14471168. Throughput: 0: 3625.7. Samples: 14471738. Policy #0 lag: (min: 2.0, avg: 18.9, max: 37.0)
+[2026-06-06 21:44:47,923][261293] Avg episode reward: [(0, '1715.435')]
+[2026-06-06 21:44:48,905][262198] Updated weights for policy 0, policy_version 28270 (0.0010)
+[2026-06-06 21:44:50,013][262198] Updated weights for policy 0, policy_version 28280 (0.0010)
+[2026-06-06 21:44:51,859][262198] Updated weights for policy 0, policy_version 28290 (0.0011)
+[2026-06-06 21:44:52,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3549.9, 300 sec: 4255.7). Total num frames: 14487552. Throughput: 0: 3744.2. Samples: 14491906. Policy #0 lag: (min: 2.0, avg: 18.9, max: 37.0)
+[2026-06-06 21:44:52,923][261293] Avg episode reward: [(0, '1816.986')]
+[2026-06-06 21:44:53,847][262198] Updated weights for policy 0, policy_version 28301 (0.0013)
+[2026-06-06 21:44:54,984][262198] Updated weights for policy 0, policy_version 28311 (0.0012)
+[2026-06-06 21:44:56,898][262198] Updated weights for policy 0, policy_version 28321 (0.0010)
+[2026-06-06 21:44:57,922][261293] Fps is (10 sec: 3276.5, 60 sec: 3549.8, 300 sec: 4241.8). Total num frames: 14503936. Throughput: 0: 3704.3. Samples: 14511444. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:44:57,924][261293] Avg episode reward: [(0, '1725.295')]
+[2026-06-06 21:44:58,348][262198] Updated weights for policy 0, policy_version 28331 (0.0011)
+[2026-06-06 21:45:00,128][262198] Updated weights for policy 0, policy_version 28341 (0.0009)
+[2026-06-06 21:45:01,296][262198] Updated weights for policy 0, policy_version 28351 (0.0009)
+[2026-06-06 21:45:02,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3584.0, 300 sec: 4221.0). Total num frames: 14520320. Throughput: 0: 3660.0. Samples: 14521082. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:45:02,923][261293] Avg episode reward: [(0, '1643.858')]
+[2026-06-06 21:45:03,125][262198] Updated weights for policy 0, policy_version 28361 (0.0011)
+[2026-06-06 21:45:04,455][262198] Updated weights for policy 0, policy_version 28371 (0.0010)
+[2026-06-06 21:45:05,716][262198] Updated weights for policy 0, policy_version 28381 (0.0010)
+[2026-06-06 21:45:07,131][262198] Updated weights for policy 0, policy_version 28391 (0.0009)
+[2026-06-06 21:45:07,921][261293] Fps is (10 sec: 3482.0, 60 sec: 3652.3, 300 sec: 4214.0). Total num frames: 14538752. Throughput: 0: 3611.0. Samples: 14542332. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:45:07,923][261293] Avg episode reward: [(0, '1666.083')]
+[2026-06-06 21:45:09,237][262198] Updated weights for policy 0, policy_version 28401 (0.0009)
+[2026-06-06 21:45:09,336][262115] Signal inference workers to stop experience collection... (3050 times)
+[2026-06-06 21:45:09,374][262198] InferenceWorker_p0-w0: stopping experience collection (3050 times)
+[2026-06-06 21:45:09,434][262115] Signal inference workers to resume experience collection... (3050 times)
+[2026-06-06 21:45:09,435][262198] InferenceWorker_p0-w0: resuming experience collection (3050 times)
+[2026-06-06 21:45:09,991][262198] Updated weights for policy 0, policy_version 28412 (0.0009)
+[2026-06-06 21:45:11,809][262198] Updated weights for policy 0, policy_version 28422 (0.0009)
+[2026-06-06 21:45:12,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3652.3, 300 sec: 4193.2). Total num frames: 14555136. Throughput: 0: 3559.4. Samples: 14563010. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 21:45:12,924][261293] Avg episode reward: [(0, '1613.716')]
+[2026-06-06 21:45:13,320][262198] Updated weights for policy 0, policy_version 28432 (0.0010)
+[2026-06-06 21:45:14,668][262198] Updated weights for policy 0, policy_version 28442 (0.0009)
+[2026-06-06 21:45:16,057][262198] Updated weights for policy 0, policy_version 28452 (0.0009)
+[2026-06-06 21:45:17,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3584.0, 300 sec: 4179.3). Total num frames: 14571520. Throughput: 0: 3550.4. Samples: 14573582. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 21:45:17,924][261293] Avg episode reward: [(0, '1649.253')]
+[2026-06-06 21:45:18,006][262198] Updated weights for policy 0, policy_version 28463 (0.0011)
+[2026-06-06 21:45:19,373][262198] Updated weights for policy 0, policy_version 28473 (0.0010)
+[2026-06-06 21:45:20,933][262198] Updated weights for policy 0, policy_version 28483 (0.0011)
+[2026-06-06 21:45:22,920][261293] Fps is (10 sec: 3277.0, 60 sec: 3550.0, 300 sec: 4158.5). Total num frames: 14587904. Throughput: 0: 3473.6. Samples: 14593502. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 21:45:22,922][261293] Avg episode reward: [(0, '1699.676')]
+[2026-06-06 21:45:23,026][262198] Updated weights for policy 0, policy_version 28493 (0.0011)
+[2026-06-06 21:45:24,134][262198] Updated weights for policy 0, policy_version 28503 (0.0010)
+[2026-06-06 21:45:25,991][262198] Updated weights for policy 0, policy_version 28513 (0.0010)
+[2026-06-06 21:45:27,335][262198] Updated weights for policy 0, policy_version 28523 (0.0009)
+[2026-06-06 21:45:27,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3515.7, 300 sec: 4123.8). Total num frames: 14604288. Throughput: 0: 3377.9. Samples: 14613348. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 21:45:27,923][261293] Avg episode reward: [(0, '1755.524')]
+[2026-06-06 21:45:28,900][262198] Updated weights for policy 0, policy_version 28534 (0.0009)
+[2026-06-06 21:45:30,177][262198] Updated weights for policy 0, policy_version 28545 (0.0009)
+[2026-06-06 21:45:31,443][262198] Updated weights for policy 0, policy_version 28555 (0.0008)
+[2026-06-06 21:45:32,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3515.8, 300 sec: 4116.8). Total num frames: 14624768. Throughput: 0: 3414.4. Samples: 14625384. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 21:45:32,923][261293] Avg episode reward: [(0, '1709.032')]
+[2026-06-06 21:45:33,071][262198] Updated weights for policy 0, policy_version 28568 (0.0012)
+[2026-06-06 21:45:34,453][262198] Updated weights for policy 0, policy_version 28578 (0.0009)
+[2026-06-06 21:45:35,747][262198] Updated weights for policy 0, policy_version 28588 (0.0009)
+[2026-06-06 21:45:37,153][262198] Updated weights for policy 0, policy_version 28600 (0.0009)
+[2026-06-06 21:45:37,921][261293] Fps is (10 sec: 4300.7, 60 sec: 3549.9, 300 sec: 4102.9). Total num frames: 14647296. Throughput: 0: 3493.8. Samples: 14649128. Policy #0 lag: (min: 3.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:45:37,923][261293] Avg episode reward: [(0, '1842.058')]
+[2026-06-06 21:45:38,404][262198] Updated weights for policy 0, policy_version 28610 (0.0009)
+[2026-06-06 21:45:39,875][262198] Updated weights for policy 0, policy_version 28620 (0.0009)
+[2026-06-06 21:45:41,513][262198] Updated weights for policy 0, policy_version 28630 (0.0011)
+[2026-06-06 21:45:42,402][262198] Updated weights for policy 0, policy_version 28640 (0.0011)
+[2026-06-06 21:45:42,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3481.6, 300 sec: 4082.1). Total num frames: 14663680. Throughput: 0: 3556.8. Samples: 14671496. Policy #0 lag: (min: 3.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:45:42,922][261293] Avg episode reward: [(0, '1729.576')]
+[2026-06-06 21:45:44,683][262198] Updated weights for policy 0, policy_version 28650 (0.0011)
+[2026-06-06 21:45:46,104][262198] Updated weights for policy 0, policy_version 28660 (0.0008)
+[2026-06-06 21:45:46,997][262198] Updated weights for policy 0, policy_version 28670 (0.0007)
+[2026-06-06 21:45:47,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3515.7, 300 sec: 4054.3). Total num frames: 14682112. Throughput: 0: 3569.3. Samples: 14681702. Policy #0 lag: (min: 3.0, avg: 21.6, max: 39.0)
+[2026-06-06 21:45:47,922][261293] Avg episode reward: [(0, '1654.052')]
+[2026-06-06 21:45:48,138][262198] Updated weights for policy 0, policy_version 28680 (0.0011)
+[2026-06-06 21:45:50,213][262198] Updated weights for policy 0, policy_version 28690 (0.0010)
+[2026-06-06 21:45:51,003][262198] Updated weights for policy 0, policy_version 28700 (0.0009)
+[2026-06-06 21:45:52,549][262198] Updated weights for policy 0, policy_version 28710 (0.0010)
+[2026-06-06 21:45:52,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3549.9, 300 sec: 4040.5). Total num frames: 14700544. Throughput: 0: 3588.1. Samples: 14703796. Policy #0 lag: (min: 3.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:45:52,923][261293] Avg episode reward: [(0, '1621.567')]
+[2026-06-06 21:45:54,452][262198] Updated weights for policy 0, policy_version 28720 (0.0011)
+[2026-06-06 21:45:55,741][262198] Updated weights for policy 0, policy_version 28730 (0.0011)
+[2026-06-06 21:45:57,259][262198] Updated weights for policy 0, policy_version 28740 (0.0013)
+[2026-06-06 21:45:57,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3549.9, 300 sec: 4026.6). Total num frames: 14716928. Throughput: 0: 3587.9. Samples: 14724466. Policy #0 lag: (min: 3.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:45:57,923][261293] Avg episode reward: [(0, '1417.874')]
+[2026-06-06 21:45:59,036][262198] Updated weights for policy 0, policy_version 28750 (0.0008)
+[2026-06-06 21:46:00,206][262198] Updated weights for policy 0, policy_version 28764 (0.0009)
+[2026-06-06 21:46:01,363][262198] Updated weights for policy 0, policy_version 28776 (0.0008)
+[2026-06-06 21:46:02,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3618.1, 300 sec: 4005.8). Total num frames: 14737408. Throughput: 0: 3635.7. Samples: 14737186. Policy #0 lag: (min: 3.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:46:02,923][261293] Avg episode reward: [(0, '1563.777')]
+[2026-06-06 21:46:03,122][262198] Updated weights for policy 0, policy_version 28786 (0.0009)
+[2026-06-06 21:46:04,001][262198] Updated weights for policy 0, policy_version 28796 (0.0009)
+[2026-06-06 21:46:05,194][262198] Updated weights for policy 0, policy_version 28807 (0.0009)
+[2026-06-06 21:46:07,146][262198] Updated weights for policy 0, policy_version 28821 (0.0008)
+[2026-06-06 21:46:07,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3686.4, 300 sec: 4005.7). Total num frames: 14759936. Throughput: 0: 3781.5. Samples: 14763670. Policy #0 lag: (min: 3.0, avg: 19.8, max: 36.0)
+[2026-06-06 21:46:07,923][261293] Avg episode reward: [(0, '1594.821')]
+[2026-06-06 21:46:08,643][262198] Updated weights for policy 0, policy_version 28833 (0.0008)
+[2026-06-06 21:46:10,335][262198] Updated weights for policy 0, policy_version 28845 (0.0010)
+[2026-06-06 21:46:11,093][262198] Updated weights for policy 0, policy_version 28855 (0.0007)
+[2026-06-06 21:46:12,280][262198] Updated weights for policy 0, policy_version 28866 (0.0008)
+[2026-06-06 21:46:12,921][261293] Fps is (10 sec: 4505.3, 60 sec: 3788.8, 300 sec: 4005.7). Total num frames: 14782464. Throughput: 0: 3900.3. Samples: 14788866. Policy #0 lag: (min: 4.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:46:12,925][261293] Avg episode reward: [(0, '1566.971')]
+[2026-06-06 21:46:13,519][262198] Updated weights for policy 0, policy_version 28876 (0.0010)
+[2026-06-06 21:46:15,467][262198] Updated weights for policy 0, policy_version 28886 (0.0011)
+[2026-06-06 21:46:16,450][262198] Updated weights for policy 0, policy_version 28896 (0.0011)
+[2026-06-06 21:46:17,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3788.8, 300 sec: 3984.9). Total num frames: 14798848. Throughput: 0: 3858.2. Samples: 14799004. Policy #0 lag: (min: 4.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:46:17,922][261293] Avg episode reward: [(0, '1690.118')]
+[2026-06-06 21:46:18,441][262198] Updated weights for policy 0, policy_version 28906 (0.0009)
+[2026-06-06 21:46:19,890][262198] Updated weights for policy 0, policy_version 28916 (0.0010)
+[2026-06-06 21:46:21,219][262198] Updated weights for policy 0, policy_version 28926 (0.0011)
+[2026-06-06 21:46:22,421][262198] Updated weights for policy 0, policy_version 28936 (0.0010)
+[2026-06-06 21:46:22,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3788.8, 300 sec: 3950.2). Total num frames: 14815232. Throughput: 0: 3781.0. Samples: 14819274. Policy #0 lag: (min: 4.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:46:22,923][261293] Avg episode reward: [(0, '1787.245')]
+[2026-06-06 21:46:24,661][262198] Updated weights for policy 0, policy_version 28946 (0.0012)
+[2026-06-06 21:46:25,974][262198] Updated weights for policy 0, policy_version 28956 (0.0013)
+[2026-06-06 21:46:27,392][262115] Signal inference workers to stop experience collection... (3100 times)
+[2026-06-06 21:46:27,393][262115] Signal inference workers to resume experience collection... (3100 times)
+[2026-06-06 21:46:27,428][262198] InferenceWorker_p0-w0: stopping experience collection (3100 times)
+[2026-06-06 21:46:27,428][262198] InferenceWorker_p0-w0: resuming experience collection (3100 times)
+[2026-06-06 21:46:27,539][262198] Updated weights for policy 0, policy_version 28966 (0.0009)
+[2026-06-06 21:46:27,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3788.8, 300 sec: 3915.5). Total num frames: 14831616. Throughput: 0: 3717.5. Samples: 14838786. Policy #0 lag: (min: 4.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:46:27,925][261293] Avg episode reward: [(0, '1721.229')]
+[2026-06-06 21:46:29,707][262198] Updated weights for policy 0, policy_version 28976 (0.0012)
+[2026-06-06 21:46:31,513][262198] Updated weights for policy 0, policy_version 28987 (0.0011)
+[2026-06-06 21:46:32,920][261293] Fps is (10 sec: 2867.4, 60 sec: 3652.3, 300 sec: 3873.8). Total num frames: 14843904. Throughput: 0: 3682.6. Samples: 14847420. Policy #0 lag: (min: 1.0, avg: 16.9, max: 33.0)
+[2026-06-06 21:46:32,921][261293] Avg episode reward: [(0, '1487.588')]
+[2026-06-06 21:46:33,593][262198] Updated weights for policy 0, policy_version 28997 (0.0012)
+[2026-06-06 21:46:36,346][262198] Updated weights for policy 0, policy_version 29008 (0.0011)
+[2026-06-06 21:46:37,920][261293] Fps is (10 sec: 2457.6, 60 sec: 3481.6, 300 sec: 3825.2). Total num frames: 14856192. Throughput: 0: 3520.0. Samples: 14862194. Policy #0 lag: (min: 1.0, avg: 16.9, max: 33.0)
+[2026-06-06 21:46:37,922][261293] Avg episode reward: [(0, '1708.814')]
+[2026-06-06 21:46:38,138][262198] Updated weights for policy 0, policy_version 29018 (0.0012)
+[2026-06-06 21:46:39,668][262198] Updated weights for policy 0, policy_version 29028 (0.0013)
+[2026-06-06 21:46:42,920][261293] Fps is (10 sec: 2252.8, 60 sec: 3379.2, 300 sec: 3783.6). Total num frames: 14866432. Throughput: 0: 3401.6. Samples: 14877538. Policy #0 lag: (min: 1.0, avg: 16.9, max: 33.0)
+[2026-06-06 21:46:42,925][261293] Avg episode reward: [(0, '1575.236')]
+[2026-06-06 21:46:42,926][262198] Updated weights for policy 0, policy_version 29038 (0.0011)
+[2026-06-06 21:46:44,166][262198] Updated weights for policy 0, policy_version 29048 (0.0013)
+[2026-06-06 21:46:46,189][262198] Updated weights for policy 0, policy_version 29058 (0.0012)
+[2026-06-06 21:46:47,921][261293] Fps is (10 sec: 2457.4, 60 sec: 3310.9, 300 sec: 3741.9). Total num frames: 14880768. Throughput: 0: 3276.8. Samples: 14884642. Policy #0 lag: (min: 1.0, avg: 16.9, max: 33.0)
+[2026-06-06 21:46:47,925][261293] Avg episode reward: [(0, '1585.254')]
+[2026-06-06 21:46:48,163][262198] Updated weights for policy 0, policy_version 29068 (0.0014)
+[2026-06-06 21:46:50,506][262198] Updated weights for policy 0, policy_version 29078 (0.0011)
+[2026-06-06 21:46:51,778][262198] Updated weights for policy 0, policy_version 29088 (0.0013)
+[2026-06-06 21:46:52,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3242.7, 300 sec: 3707.2). Total num frames: 14895104. Throughput: 0: 3023.9. Samples: 14899746. Policy #0 lag: (min: 3.0, avg: 15.7, max: 36.0)
+[2026-06-06 21:46:52,922][261293] Avg episode reward: [(0, '1721.486')]
+[2026-06-06 21:46:54,645][262198] Updated weights for policy 0, policy_version 29099 (0.0013)
+[2026-06-06 21:46:56,978][262198] Updated weights for policy 0, policy_version 29109 (0.0013)
+[2026-06-06 21:46:57,921][261293] Fps is (10 sec: 2662.5, 60 sec: 3174.4, 300 sec: 3658.6). Total num frames: 14907392. Throughput: 0: 2802.4. Samples: 14914972. Policy #0 lag: (min: 3.0, avg: 15.7, max: 36.0)
+[2026-06-06 21:46:57,923][261293] Avg episode reward: [(0, '1868.785')]
+[2026-06-06 21:46:57,937][262115] Saving new best policy, reward=1868.785!
+[2026-06-06 21:46:58,239][262198] Updated weights for policy 0, policy_version 29119 (0.0012)
+[2026-06-06 21:47:01,125][262198] Updated weights for policy 0, policy_version 29129 (0.0015)
+[2026-06-06 21:47:02,648][262198] Updated weights for policy 0, policy_version 29139 (0.0009)
+[2026-06-06 21:47:02,920][261293] Fps is (10 sec: 2457.7, 60 sec: 3037.9, 300 sec: 3623.9). Total num frames: 14919680. Throughput: 0: 2737.8. Samples: 14922204. Policy #0 lag: (min: 3.0, avg: 15.7, max: 36.0)
+[2026-06-06 21:47:02,923][261293] Avg episode reward: [(0, '1566.116')]
+[2026-06-06 21:47:04,137][262198] Updated weights for policy 0, policy_version 29149 (0.0011)
+[2026-06-06 21:47:05,487][262198] Updated weights for policy 0, policy_version 29159 (0.0011)
+[2026-06-06 21:47:07,448][262198] Updated weights for policy 0, policy_version 29169 (0.0009)
+[2026-06-06 21:47:07,921][261293] Fps is (10 sec: 2867.3, 60 sec: 2935.5, 300 sec: 3582.3). Total num frames: 14936064. Throughput: 0: 2712.5. Samples: 14941336. Policy #0 lag: (min: 3.0, avg: 15.7, max: 36.0)
+[2026-06-06 21:47:07,924][261293] Avg episode reward: [(0, '1541.504')]
+[2026-06-06 21:47:08,515][262198] Updated weights for policy 0, policy_version 29180 (0.0010)
+[2026-06-06 21:47:09,592][262198] Updated weights for policy 0, policy_version 29190 (0.0009)
+[2026-06-06 21:47:11,653][262198] Updated weights for policy 0, policy_version 29201 (0.0009)
+[2026-06-06 21:47:12,917][262198] Updated weights for policy 0, policy_version 29215 (0.0012)
+[2026-06-06 21:47:12,921][261293] Fps is (10 sec: 3686.2, 60 sec: 2901.3, 300 sec: 3575.3). Total num frames: 14956544. Throughput: 0: 2804.6. Samples: 14964992. Policy #0 lag: (min: 7.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:47:12,923][261293] Avg episode reward: [(0, '1628.619')]
+[2026-06-06 21:47:15,092][262198] Updated weights for policy 0, policy_version 29225 (0.0010)
+[2026-06-06 21:47:16,740][262198] Updated weights for policy 0, policy_version 29235 (0.0009)
+[2026-06-06 21:47:17,921][261293] Fps is (10 sec: 3686.3, 60 sec: 2901.3, 300 sec: 3533.7). Total num frames: 14972928. Throughput: 0: 2815.3. Samples: 14974110. Policy #0 lag: (min: 7.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:47:17,926][261293] Avg episode reward: [(0, '1794.570')]
+[2026-06-06 21:47:18,135][262198] Updated weights for policy 0, policy_version 29245 (0.0011)
+[2026-06-06 21:47:19,518][262198] Updated weights for policy 0, policy_version 29255 (0.0010)
+[2026-06-06 21:47:22,432][262198] Updated weights for policy 0, policy_version 29265 (0.0012)
+[2026-06-06 21:47:22,921][261293] Fps is (10 sec: 2867.2, 60 sec: 2833.1, 300 sec: 3478.1). Total num frames: 14985216. Throughput: 0: 2878.4. Samples: 14991726. Policy #0 lag: (min: 7.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:47:22,923][261293] Avg episode reward: [(0, '1631.713')]
+[2026-06-06 21:47:23,428][262198] Updated weights for policy 0, policy_version 29275 (0.0011)
+[2026-06-06 21:47:25,200][262198] Updated weights for policy 0, policy_version 29285 (0.0011)
+[2026-06-06 21:47:27,562][262198] Updated weights for policy 0, policy_version 29295 (0.0014)
+[2026-06-06 21:47:27,921][261293] Fps is (10 sec: 2662.4, 60 sec: 2798.9, 300 sec: 3457.3). Total num frames: 14999552. Throughput: 0: 2920.8. Samples: 15008976. Policy #0 lag: (min: 7.0, avg: 22.4, max: 38.0)
+[2026-06-06 21:47:27,925][261293] Avg episode reward: [(0, '1570.452')]
+[2026-06-06 21:47:29,005][262198] Updated weights for policy 0, policy_version 29305 (0.0011)
+[2026-06-06 21:47:30,316][262198] Updated weights for policy 0, policy_version 29315 (0.0015)
+[2026-06-06 21:47:32,920][261293] Fps is (10 sec: 2867.4, 60 sec: 2833.1, 300 sec: 3443.4). Total num frames: 15013888. Throughput: 0: 2957.8. Samples: 15017740. Policy #0 lag: (min: 3.0, avg: 18.6, max: 38.0)
+[2026-06-06 21:47:32,922][261293] Avg episode reward: [(0, '1524.130')]
+[2026-06-06 21:47:32,992][262198] Updated weights for policy 0, policy_version 29325 (0.0017)
+[2026-06-06 21:47:34,508][262198] Updated weights for policy 0, policy_version 29343 (0.0012)
+[2026-06-06 21:47:36,275][262198] Updated weights for policy 0, policy_version 29353 (0.0009)
+[2026-06-06 21:47:37,545][262198] Updated weights for policy 0, policy_version 29363 (0.0009)
+[2026-06-06 21:47:37,921][261293] Fps is (10 sec: 3481.4, 60 sec: 2969.6, 300 sec: 3450.4). Total num frames: 15034368. Throughput: 0: 3094.1. Samples: 15038982. Policy #0 lag: (min: 3.0, avg: 18.6, max: 38.0)
+[2026-06-06 21:47:37,924][261293] Avg episode reward: [(0, '1515.899')]
+[2026-06-06 21:47:39,000][262198] Updated weights for policy 0, policy_version 29376 (0.0010)
+[2026-06-06 21:47:40,594][262198] Updated weights for policy 0, policy_version 29386 (0.0008)
+[2026-06-06 21:47:41,893][262198] Updated weights for policy 0, policy_version 29396 (0.0009)
+[2026-06-06 21:47:42,921][261293] Fps is (10 sec: 4095.9, 60 sec: 3140.3, 300 sec: 3464.2). Total num frames: 15054848. Throughput: 0: 3273.9. Samples: 15062296. Policy #0 lag: (min: 3.0, avg: 18.6, max: 38.0)
+[2026-06-06 21:47:42,922][261293] Avg episode reward: [(0, '1730.715')]
+[2026-06-06 21:47:43,085][262198] Updated weights for policy 0, policy_version 29407 (0.0009)
+[2026-06-06 21:47:44,666][262198] Updated weights for policy 0, policy_version 29417 (0.0009)
+[2026-06-06 21:47:46,156][262198] Updated weights for policy 0, policy_version 29428 (0.0009)
+[2026-06-06 21:47:47,034][262198] Updated weights for policy 0, policy_version 29438 (0.0008)
+[2026-06-06 21:47:47,921][261293] Fps is (10 sec: 4096.2, 60 sec: 3242.7, 300 sec: 3464.2). Total num frames: 15075328. Throughput: 0: 3377.9. Samples: 15074212. Policy #0 lag: (min: 3.0, avg: 18.6, max: 38.0)
+[2026-06-06 21:47:47,923][261293] Avg episode reward: [(0, '1642.853')]
+[2026-06-06 21:47:48,521][262198] Updated weights for policy 0, policy_version 29448 (0.0009)
+[2026-06-06 21:47:50,286][262198] Updated weights for policy 0, policy_version 29461 (0.0009)
+[2026-06-06 21:47:51,429][262198] Updated weights for policy 0, policy_version 29471 (0.0009)
+[2026-06-06 21:47:52,792][262198] Updated weights for policy 0, policy_version 29481 (0.0009)
+[2026-06-06 21:47:52,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3310.9, 300 sec: 3457.3). Total num frames: 15093760. Throughput: 0: 3488.1. Samples: 15098300. Policy #0 lag: (min: 3.0, avg: 21.7, max: 35.0)
+[2026-06-06 21:47:52,923][261293] Avg episode reward: [(0, '1619.282')]
+[2026-06-06 21:47:54,330][262198] Updated weights for policy 0, policy_version 29491 (0.0010)
+[2026-06-06 21:47:55,445][262198] Updated weights for policy 0, policy_version 29502 (0.0011)
+[2026-06-06 21:47:56,969][262198] Updated weights for policy 0, policy_version 29512 (0.0012)
+[2026-06-06 21:47:57,922][261293] Fps is (10 sec: 3686.2, 60 sec: 3413.3, 300 sec: 3450.4). Total num frames: 15112192. Throughput: 0: 3429.0. Samples: 15119298. Policy #0 lag: (min: 3.0, avg: 21.7, max: 35.0)
+[2026-06-06 21:47:57,927][261293] Avg episode reward: [(0, '1630.163')]
+[2026-06-06 21:47:59,538][262198] Updated weights for policy 0, policy_version 29522 (0.0013)
+[2026-06-06 21:48:00,987][262198] Updated weights for policy 0, policy_version 29532 (0.0012)
+[2026-06-06 21:48:02,568][262198] Updated weights for policy 0, policy_version 29542 (0.0008)
+[2026-06-06 21:48:02,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3447.5, 300 sec: 3443.4). Total num frames: 15126528. Throughput: 0: 3402.7. Samples: 15127230. Policy #0 lag: (min: 3.0, avg: 21.7, max: 35.0)
+[2026-06-06 21:48:02,922][261293] Avg episode reward: [(0, '1722.281')]
+[2026-06-06 21:48:04,621][262198] Updated weights for policy 0, policy_version 29552 (0.0009)
+[2026-06-06 21:48:05,792][262198] Updated weights for policy 0, policy_version 29562 (0.0009)
+[2026-06-06 21:48:06,823][262198] Updated weights for policy 0, policy_version 29572 (0.0012)
+[2026-06-06 21:48:07,921][261293] Fps is (10 sec: 3072.1, 60 sec: 3447.4, 300 sec: 3436.5). Total num frames: 15142912. Throughput: 0: 3460.4. Samples: 15147442. Policy #0 lag: (min: 3.0, avg: 21.7, max: 35.0)
+[2026-06-06 21:48:07,923][261293] Avg episode reward: [(0, '1649.321')]
+[2026-06-06 21:48:09,262][262198] Updated weights for policy 0, policy_version 29582 (0.0011)
+[2026-06-06 21:48:10,385][262198] Updated weights for policy 0, policy_version 29592 (0.0011)
+[2026-06-06 21:48:11,742][262198] Updated weights for policy 0, policy_version 29602 (0.0009)
+[2026-06-06 21:48:12,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3379.2, 300 sec: 3429.5). Total num frames: 15159296. Throughput: 0: 3518.2. Samples: 15167294. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:48:12,925][261293] Avg episode reward: [(0, '1641.986')]
+[2026-06-06 21:48:13,120][262198] Updated weights for policy 0, policy_version 29612 (0.0011)
+[2026-06-06 21:48:15,258][262198] Updated weights for policy 0, policy_version 29622 (0.0011)
+[2026-06-06 21:48:16,130][262198] Updated weights for policy 0, policy_version 29632 (0.0010)
+[2026-06-06 21:48:17,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3379.2, 300 sec: 3415.7). Total num frames: 15175680. Throughput: 0: 3533.1. Samples: 15176732. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:48:17,923][261293] Avg episode reward: [(0, '1685.808')]
+[2026-06-06 21:48:18,186][262198] Updated weights for policy 0, policy_version 29642 (0.0010)
+[2026-06-06 21:48:19,886][262198] Updated weights for policy 0, policy_version 29652 (0.0012)
+[2026-06-06 21:48:21,166][262198] Updated weights for policy 0, policy_version 29662 (0.0011)
+[2026-06-06 21:48:22,532][262198] Updated weights for policy 0, policy_version 29672 (0.0009)
+[2026-06-06 21:48:22,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3447.5, 300 sec: 3401.8). Total num frames: 15192064. Throughput: 0: 3491.2. Samples: 15196082. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:48:22,922][261293] Avg episode reward: [(0, '1720.273')]
+[2026-06-06 21:48:24,713][262198] Updated weights for policy 0, policy_version 29682 (0.0010)
+[2026-06-06 21:48:25,939][262198] Updated weights for policy 0, policy_version 29692 (0.0012)
+[2026-06-06 21:48:27,498][262198] Updated weights for policy 0, policy_version 29702 (0.0011)
+[2026-06-06 21:48:27,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3481.6, 300 sec: 3394.8). Total num frames: 15208448. Throughput: 0: 3411.6. Samples: 15215820. Policy #0 lag: (min: 3.0, avg: 20.9, max: 35.0)
+[2026-06-06 21:48:27,923][261293] Avg episode reward: [(0, '1658.121')]
+[2026-06-06 21:48:29,689][262198] Updated weights for policy 0, policy_version 29712 (0.0012)
+[2026-06-06 21:48:31,011][262198] Updated weights for policy 0, policy_version 29722 (0.0010)
+[2026-06-06 21:48:32,295][262198] Updated weights for policy 0, policy_version 29732 (0.0012)
+[2026-06-06 21:48:32,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3515.7, 300 sec: 3387.9). Total num frames: 15224832. Throughput: 0: 3349.7. Samples: 15224948. Policy #0 lag: (min: 1.0, avg: 18.3, max: 33.0)
+[2026-06-06 21:48:32,922][261293] Avg episode reward: [(0, '1887.144')]
+[2026-06-06 21:48:32,927][262115] Saving new best policy, reward=1887.144!
+[2026-06-06 21:48:35,053][262198] Updated weights for policy 0, policy_version 29742 (0.0012)
+[2026-06-06 21:48:36,172][262198] Updated weights for policy 0, policy_version 29752 (0.0011)
+[2026-06-06 21:48:36,786][262115] Signal inference workers to stop experience collection... (3150 times)
+[2026-06-06 21:48:36,788][262115] Signal inference workers to resume experience collection... (3150 times)
+[2026-06-06 21:48:36,818][262198] InferenceWorker_p0-w0: stopping experience collection (3150 times)
+[2026-06-06 21:48:36,818][262198] InferenceWorker_p0-w0: resuming experience collection (3150 times)
+[2026-06-06 21:48:37,740][262198] Updated weights for policy 0, policy_version 29762 (0.0010)
+[2026-06-06 21:48:37,921][261293] Fps is (10 sec: 3072.0, 60 sec: 3413.4, 300 sec: 3380.9). Total num frames: 15239168. Throughput: 0: 3213.5. Samples: 15242908. Policy #0 lag: (min: 1.0, avg: 18.3, max: 33.0)
+[2026-06-06 21:48:37,923][261293] Avg episode reward: [(0, '1727.824')]
+[2026-06-06 21:48:39,246][262198] Updated weights for policy 0, policy_version 29772 (0.0011)
+[2026-06-06 21:48:41,478][262198] Updated weights for policy 0, policy_version 29782 (0.0012)
+[2026-06-06 21:48:42,410][262198] Updated weights for policy 0, policy_version 29792 (0.0011)
+[2026-06-06 21:48:42,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3310.9, 300 sec: 3374.0). Total num frames: 15253504. Throughput: 0: 3154.2. Samples: 15261236. Policy #0 lag: (min: 1.0, avg: 18.3, max: 33.0)
+[2026-06-06 21:48:42,923][261293] Avg episode reward: [(0, '1597.364')]
+[2026-06-06 21:48:44,040][262198] Updated weights for policy 0, policy_version 29802 (0.0009)
+[2026-06-06 21:48:45,387][262198] Updated weights for policy 0, policy_version 29812 (0.0010)
+[2026-06-06 21:48:47,087][262198] Updated weights for policy 0, policy_version 29825 (0.0009)
+[2026-06-06 21:48:47,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3310.9, 300 sec: 3387.9). Total num frames: 15273984. Throughput: 0: 3247.1. Samples: 15273350. Policy #0 lag: (min: 1.0, avg: 18.3, max: 33.0)
+[2026-06-06 21:48:47,922][261293] Avg episode reward: [(0, '1635.744')]
+[2026-06-06 21:48:48,231][262198] Updated weights for policy 0, policy_version 29835 (0.0008)
+[2026-06-06 21:48:50,280][262198] Updated weights for policy 0, policy_version 29845 (0.0009)
+[2026-06-06 21:48:51,198][262198] Updated weights for policy 0, policy_version 29855 (0.0014)
+[2026-06-06 21:48:52,923][261293] Fps is (10 sec: 3685.6, 60 sec: 3276.7, 300 sec: 3387.9). Total num frames: 15290368. Throughput: 0: 3262.7. Samples: 15294272. Policy #0 lag: (min: 3.0, avg: 21.5, max: 35.0)
+[2026-06-06 21:48:52,927][261293] Avg episode reward: [(0, '1751.385')]
+[2026-06-06 21:48:53,345][262198] Updated weights for policy 0, policy_version 29865 (0.0013)
+[2026-06-06 21:48:55,026][262198] Updated weights for policy 0, policy_version 29875 (0.0010)
+[2026-06-06 21:48:56,284][262198] Updated weights for policy 0, policy_version 29885 (0.0009)
+[2026-06-06 21:48:57,664][262198] Updated weights for policy 0, policy_version 29895 (0.0010)
+[2026-06-06 21:48:57,922][261293] Fps is (10 sec: 3276.4, 60 sec: 3242.6, 300 sec: 3394.8). Total num frames: 15306752. Throughput: 0: 3247.7. Samples: 15313444. Policy #0 lag: (min: 3.0, avg: 21.5, max: 35.0)
+[2026-06-06 21:48:57,924][261293] Avg episode reward: [(0, '1530.097')]
+[2026-06-06 21:49:00,297][262198] Updated weights for policy 0, policy_version 29905 (0.0011)
+[2026-06-06 21:49:01,361][262198] Updated weights for policy 0, policy_version 29915 (0.0010)
+[2026-06-06 21:49:02,921][261293] Fps is (10 sec: 3072.7, 60 sec: 3242.6, 300 sec: 3394.8). Total num frames: 15321088. Throughput: 0: 3241.3. Samples: 15322592. Policy #0 lag: (min: 3.0, avg: 21.5, max: 35.0)
+[2026-06-06 21:49:02,923][261293] Avg episode reward: [(0, '1472.232')]
+[2026-06-06 21:49:02,966][262198] Updated weights for policy 0, policy_version 29925 (0.0010)
+[2026-06-06 21:49:05,253][262198] Updated weights for policy 0, policy_version 29935 (0.0011)
+[2026-06-06 21:49:06,917][262198] Updated weights for policy 0, policy_version 29945 (0.0009)
+[2026-06-06 21:49:07,921][261293] Fps is (10 sec: 2867.5, 60 sec: 3208.5, 300 sec: 3387.9). Total num frames: 15335424. Throughput: 0: 3203.5. Samples: 15340242. Policy #0 lag: (min: 3.0, avg: 21.5, max: 35.0)
+[2026-06-06 21:49:07,923][261293] Avg episode reward: [(0, '1609.401')]
+[2026-06-06 21:49:08,115][262198] Updated weights for policy 0, policy_version 29955 (0.0012)
+[2026-06-06 21:49:10,731][262198] Updated weights for policy 0, policy_version 29965 (0.0011)
+[2026-06-06 21:49:11,904][262198] Updated weights for policy 0, policy_version 29975 (0.0011)
+[2026-06-06 21:49:12,920][261293] Fps is (10 sec: 3072.1, 60 sec: 3208.5, 300 sec: 3374.0). Total num frames: 15351808. Throughput: 0: 3162.9. Samples: 15358152. Policy #0 lag: (min: 6.0, avg: 21.3, max: 37.0)
+[2026-06-06 21:49:12,922][261293] Avg episode reward: [(0, '1764.772')]
+[2026-06-06 21:49:13,503][262198] Updated weights for policy 0, policy_version 29985 (0.0011)
+[2026-06-06 21:49:15,157][262198] Updated weights for policy 0, policy_version 29995 (0.0011)
+[2026-06-06 21:49:17,511][262198] Updated weights for policy 0, policy_version 30005 (0.0012)
+[2026-06-06 21:49:17,920][261293] Fps is (10 sec: 2867.3, 60 sec: 3140.3, 300 sec: 3353.2). Total num frames: 15364096. Throughput: 0: 3150.6. Samples: 15366726. Policy #0 lag: (min: 6.0, avg: 21.3, max: 37.0)
+[2026-06-06 21:49:17,922][261293] Avg episode reward: [(0, '1703.186')]
+[2026-06-06 21:49:18,230][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000030012_15366144.pth...
+[2026-06-06 21:49:18,293][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000004772_2443264.pth
+[2026-06-06 21:49:18,549][262198] Updated weights for policy 0, policy_version 30016 (0.0012)
+[2026-06-06 21:49:20,720][262198] Updated weights for policy 0, policy_version 30026 (0.0012)
+[2026-06-06 21:49:22,511][262198] Updated weights for policy 0, policy_version 30036 (0.0010)
+[2026-06-06 21:49:22,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3106.1, 300 sec: 3339.3). Total num frames: 15378432. Throughput: 0: 3147.0. Samples: 15384522. Policy #0 lag: (min: 6.0, avg: 21.3, max: 37.0)
+[2026-06-06 21:49:22,923][261293] Avg episode reward: [(0, '1744.394')]
+[2026-06-06 21:49:23,675][262198] Updated weights for policy 0, policy_version 30046 (0.0011)
+[2026-06-06 21:49:25,117][262198] Updated weights for policy 0, policy_version 30056 (0.0011)
+[2026-06-06 21:49:27,900][262198] Updated weights for policy 0, policy_version 30066 (0.0011)
+[2026-06-06 21:49:27,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3072.0, 300 sec: 3318.5). Total num frames: 15392768. Throughput: 0: 3144.1. Samples: 15402718. Policy #0 lag: (min: 6.0, avg: 21.3, max: 37.0)
+[2026-06-06 21:49:27,923][261293] Avg episode reward: [(0, '1630.931')]
+[2026-06-06 21:49:29,404][262198] Updated weights for policy 0, policy_version 30076 (0.0013)
+[2026-06-06 21:49:31,965][262198] Updated weights for policy 0, policy_version 30089 (0.0011)
+[2026-06-06 21:49:32,920][261293] Fps is (10 sec: 2867.2, 60 sec: 3037.9, 300 sec: 3297.6). Total num frames: 15407104. Throughput: 0: 3039.7. Samples: 15410138. Policy #0 lag: (min: 3.0, avg: 17.8, max: 38.0)
+[2026-06-06 21:49:32,922][261293] Avg episode reward: [(0, '1593.036')]
+[2026-06-06 21:49:33,894][262198] Updated weights for policy 0, policy_version 30099 (0.0012)
+[2026-06-06 21:49:35,386][262198] Updated weights for policy 0, policy_version 30111 (0.0012)
+[2026-06-06 21:49:37,659][262198] Updated weights for policy 0, policy_version 30121 (0.0011)
+[2026-06-06 21:49:37,920][261293] Fps is (10 sec: 3072.1, 60 sec: 3072.0, 300 sec: 3283.7). Total num frames: 15423488. Throughput: 0: 2966.2. Samples: 15427744. Policy #0 lag: (min: 3.0, avg: 17.8, max: 38.0)
+[2026-06-06 21:49:37,921][261293] Avg episode reward: [(0, '1594.328')]
+[2026-06-06 21:49:39,411][262198] Updated weights for policy 0, policy_version 30131 (0.0012)
+[2026-06-06 21:49:41,003][262198] Updated weights for policy 0, policy_version 30142 (0.0012)
+[2026-06-06 21:49:42,434][262198] Updated weights for policy 0, policy_version 30152 (0.0011)
+[2026-06-06 21:49:42,920][261293] Fps is (10 sec: 3072.0, 60 sec: 3072.0, 300 sec: 3276.8). Total num frames: 15437824. Throughput: 0: 2928.5. Samples: 15445224. Policy #0 lag: (min: 3.0, avg: 17.8, max: 38.0)
+[2026-06-06 21:49:42,922][261293] Avg episode reward: [(0, '1476.967')]
+[2026-06-06 21:49:45,032][262198] Updated weights for policy 0, policy_version 30162 (0.0012)
+[2026-06-06 21:49:46,410][262198] Updated weights for policy 0, policy_version 30172 (0.0011)
+[2026-06-06 21:49:47,790][262198] Updated weights for policy 0, policy_version 30182 (0.0011)
+[2026-06-06 21:49:47,921][261293] Fps is (10 sec: 2867.0, 60 sec: 2969.6, 300 sec: 3269.9). Total num frames: 15452160. Throughput: 0: 2921.2. Samples: 15454046. Policy #0 lag: (min: 3.0, avg: 17.8, max: 38.0)
+[2026-06-06 21:49:47,925][261293] Avg episode reward: [(0, '1568.918')]
+[2026-06-06 21:49:50,067][262198] Updated weights for policy 0, policy_version 30192 (0.0011)
+[2026-06-06 21:49:51,456][262198] Updated weights for policy 0, policy_version 30202 (0.0011)
+[2026-06-06 21:49:52,722][262198] Updated weights for policy 0, policy_version 30212 (0.0012)
+[2026-06-06 21:49:52,920][261293] Fps is (10 sec: 3072.0, 60 sec: 2969.7, 300 sec: 3269.9). Total num frames: 15468544. Throughput: 0: 2940.6. Samples: 15472566. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:49:52,922][261293] Avg episode reward: [(0, '1553.879')]
+[2026-06-06 21:49:55,302][262198] Updated weights for policy 0, policy_version 30222 (0.0013)
+[2026-06-06 21:49:56,805][262198] Updated weights for policy 0, policy_version 30232 (0.0011)
+[2026-06-06 21:49:57,921][261293] Fps is (10 sec: 3072.1, 60 sec: 2935.5, 300 sec: 3262.9). Total num frames: 15482880. Throughput: 0: 2944.2. Samples: 15490644. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:49:57,925][261293] Avg episode reward: [(0, '1511.122')]
+[2026-06-06 21:49:57,969][262198] Updated weights for policy 0, policy_version 30242 (0.0011)
+[2026-06-06 21:49:59,523][262198] Updated weights for policy 0, policy_version 30252 (0.0012)
+[2026-06-06 21:50:01,888][262198] Updated weights for policy 0, policy_version 30262 (0.0010)
+[2026-06-06 21:50:02,757][262198] Updated weights for policy 0, policy_version 30272 (0.0009)
+[2026-06-06 21:50:02,921][261293] Fps is (10 sec: 3071.8, 60 sec: 2969.6, 300 sec: 3256.0). Total num frames: 15499264. Throughput: 0: 2962.6. Samples: 15500044. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:50:02,923][261293] Avg episode reward: [(0, '1583.933')]
+[2026-06-06 21:50:04,977][262198] Updated weights for policy 0, policy_version 30282 (0.0012)
+[2026-06-06 21:50:06,785][262198] Updated weights for policy 0, policy_version 30292 (0.0011)
+[2026-06-06 21:50:07,920][261293] Fps is (10 sec: 2867.3, 60 sec: 2935.5, 300 sec: 3242.1). Total num frames: 15511552. Throughput: 0: 2963.0. Samples: 15517856. Policy #0 lag: (min: 3.0, avg: 21.6, max: 35.0)
+[2026-06-06 21:50:07,922][261293] Avg episode reward: [(0, '1722.477')]
+[2026-06-06 21:50:08,121][262198] Updated weights for policy 0, policy_version 30302 (0.0011)
+[2026-06-06 21:50:09,546][262198] Updated weights for policy 0, policy_version 30312 (0.0012)
+[2026-06-06 21:50:12,125][262198] Updated weights for policy 0, policy_version 30322 (0.0011)
+[2026-06-06 21:50:12,921][261293] Fps is (10 sec: 2867.4, 60 sec: 2935.5, 300 sec: 3242.1). Total num frames: 15527936. Throughput: 0: 2960.4. Samples: 15535936. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:50:12,925][261293] Avg episode reward: [(0, '1468.191')]
+[2026-06-06 21:50:13,503][262198] Updated weights for policy 0, policy_version 30332 (0.0011)
+[2026-06-06 21:50:15,033][262198] Updated weights for policy 0, policy_version 30342 (0.0010)
+[2026-06-06 21:50:17,184][262198] Updated weights for policy 0, policy_version 30352 (0.0010)
+[2026-06-06 21:50:17,921][261293] Fps is (10 sec: 3071.9, 60 sec: 2969.6, 300 sec: 3235.1). Total num frames: 15542272. Throughput: 0: 2994.2. Samples: 15544876. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:50:17,923][261293] Avg episode reward: [(0, '1507.170')]
+[2026-06-06 21:50:18,901][262198] Updated weights for policy 0, policy_version 30364 (0.0010)
+[2026-06-06 21:50:20,136][262198] Updated weights for policy 0, policy_version 30375 (0.0010)
+[2026-06-06 21:50:22,025][262198] Updated weights for policy 0, policy_version 30385 (0.0008)
+[2026-06-06 21:50:22,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3037.9, 300 sec: 3242.1). Total num frames: 15560704. Throughput: 0: 3057.0. Samples: 15565310. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:50:22,923][261293] Avg episode reward: [(0, '1644.490')]
+[2026-06-06 21:50:23,402][262198] Updated weights for policy 0, policy_version 30398 (0.0009)
+[2026-06-06 21:50:24,372][262198] Updated weights for policy 0, policy_version 30408 (0.0008)
+[2026-06-06 21:50:26,418][262198] Updated weights for policy 0, policy_version 30419 (0.0009)
+[2026-06-06 21:50:27,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3106.1, 300 sec: 3235.1). Total num frames: 15579136. Throughput: 0: 3191.9. Samples: 15588860. Policy #0 lag: (min: 4.0, avg: 19.6, max: 37.0)
+[2026-06-06 21:50:27,922][261293] Avg episode reward: [(0, '1600.028')]
+[2026-06-06 21:50:28,089][262198] Updated weights for policy 0, policy_version 30431 (0.0009)
+[2026-06-06 21:50:29,826][262198] Updated weights for policy 0, policy_version 30443 (0.0010)
+[2026-06-06 21:50:32,028][262198] Updated weights for policy 0, policy_version 30453 (0.0012)
+[2026-06-06 21:50:32,858][262198] Updated weights for policy 0, policy_version 30464 (0.0013)
+[2026-06-06 21:50:32,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3174.4, 300 sec: 3221.3). Total num frames: 15597568. Throughput: 0: 3199.0. Samples: 15597998. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:50:32,922][261293] Avg episode reward: [(0, '1735.449')]
+[2026-06-06 21:50:35,455][262198] Updated weights for policy 0, policy_version 30474 (0.0011)
+[2026-06-06 21:50:36,965][262198] Updated weights for policy 0, policy_version 30484 (0.0009)
+[2026-06-06 21:50:37,922][261293] Fps is (10 sec: 2866.7, 60 sec: 3071.9, 300 sec: 3200.4). Total num frames: 15607808. Throughput: 0: 3171.2. Samples: 15615276. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:50:37,925][261293] Avg episode reward: [(0, '1679.025')]
+[2026-06-06 21:50:39,238][262198] Updated weights for policy 0, policy_version 30497 (0.0011)
+[2026-06-06 21:50:40,915][262198] Updated weights for policy 0, policy_version 30507 (0.0011)
+[2026-06-06 21:50:42,921][261293] Fps is (10 sec: 2662.2, 60 sec: 3106.1, 300 sec: 3193.5). Total num frames: 15624192. Throughput: 0: 3175.7. Samples: 15633552. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:50:42,922][261293] Avg episode reward: [(0, '1536.808')]
+[2026-06-06 21:50:43,180][262198] Updated weights for policy 0, policy_version 30517 (0.0011)
+[2026-06-06 21:50:44,509][262198] Updated weights for policy 0, policy_version 30530 (0.0015)
+[2026-06-06 21:50:46,182][262198] Updated weights for policy 0, policy_version 30540 (0.0009)
+[2026-06-06 21:50:47,237][262115] Signal inference workers to stop experience collection... (3200 times)
+[2026-06-06 21:50:47,240][262115] Signal inference workers to resume experience collection... (3200 times)
+[2026-06-06 21:50:47,280][262198] InferenceWorker_p0-w0: stopping experience collection (3200 times)
+[2026-06-06 21:50:47,281][262198] InferenceWorker_p0-w0: resuming experience collection (3200 times)
+[2026-06-06 21:50:47,921][261293] Fps is (10 sec: 3277.4, 60 sec: 3140.3, 300 sec: 3186.6). Total num frames: 15640576. Throughput: 0: 3175.0. Samples: 15642916. Policy #0 lag: (min: 11.0, avg: 23.8, max: 43.0)
+[2026-06-06 21:50:47,923][261293] Avg episode reward: [(0, '1748.942')]
+[2026-06-06 21:50:48,453][262198] Updated weights for policy 0, policy_version 30552 (0.0009)
+[2026-06-06 21:50:49,728][262198] Updated weights for policy 0, policy_version 30564 (0.0010)
+[2026-06-06 21:50:51,684][262198] Updated weights for policy 0, policy_version 30574 (0.0009)
+[2026-06-06 21:50:52,930][261293] Fps is (10 sec: 3274.2, 60 sec: 3139.8, 300 sec: 3186.5). Total num frames: 15656960. Throughput: 0: 3233.4. Samples: 15663388. Policy #0 lag: (min: 3.0, avg: 17.6, max: 35.0)
+[2026-06-06 21:50:52,935][261293] Avg episode reward: [(0, '1669.837')]
+[2026-06-06 21:50:53,268][262198] Updated weights for policy 0, policy_version 30585 (0.0010)
+[2026-06-06 21:50:54,512][262198] Updated weights for policy 0, policy_version 30597 (0.0010)
+[2026-06-06 21:50:56,340][262198] Updated weights for policy 0, policy_version 30607 (0.0010)
+[2026-06-06 21:50:57,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3208.5, 300 sec: 3179.6). Total num frames: 15675392. Throughput: 0: 3313.6. Samples: 15685048. Policy #0 lag: (min: 3.0, avg: 17.6, max: 35.0)
+[2026-06-06 21:50:57,923][261293] Avg episode reward: [(0, '1810.677')]
+[2026-06-06 21:50:58,055][262198] Updated weights for policy 0, policy_version 30620 (0.0009)
+[2026-06-06 21:50:59,026][262198] Updated weights for policy 0, policy_version 30630 (0.0010)
+[2026-06-06 21:51:01,161][262198] Updated weights for policy 0, policy_version 30640 (0.0010)
+[2026-06-06 21:51:02,922][261293] Fps is (10 sec: 3689.1, 60 sec: 3242.7, 300 sec: 3165.7). Total num frames: 15693824. Throughput: 0: 3347.1. Samples: 15695496. Policy #0 lag: (min: 3.0, avg: 17.6, max: 35.0)
+[2026-06-06 21:51:02,926][261293] Avg episode reward: [(0, '1705.389')]
+[2026-06-06 21:51:03,083][262198] Updated weights for policy 0, policy_version 30656 (0.0010)
+[2026-06-06 21:51:04,472][262198] Updated weights for policy 0, policy_version 30666 (0.0008)
+[2026-06-06 21:51:05,807][262198] Updated weights for policy 0, policy_version 30676 (0.0008)
+[2026-06-06 21:51:07,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3345.0, 300 sec: 3151.8). Total num frames: 15712256. Throughput: 0: 3397.4. Samples: 15718194. Policy #0 lag: (min: 3.0, avg: 17.6, max: 35.0)
+[2026-06-06 21:51:07,924][261293] Avg episode reward: [(0, '1681.327')]
+[2026-06-06 21:51:08,065][262198] Updated weights for policy 0, policy_version 30694 (0.0009)
+[2026-06-06 21:51:09,567][262198] Updated weights for policy 0, policy_version 30704 (0.0008)
+[2026-06-06 21:51:11,628][262198] Updated weights for policy 0, policy_version 30717 (0.0008)
+[2026-06-06 21:51:12,374][262198] Updated weights for policy 0, policy_version 30728 (0.0008)
+[2026-06-06 21:51:12,921][261293] Fps is (10 sec: 3891.5, 60 sec: 3413.3, 300 sec: 3165.7). Total num frames: 15732736. Throughput: 0: 3365.8. Samples: 15740322. Policy #0 lag: (min: 3.0, avg: 21.9, max: 35.0)
+[2026-06-06 21:51:12,924][261293] Avg episode reward: [(0, '1568.675')]
+[2026-06-06 21:51:14,093][262198] Updated weights for policy 0, policy_version 30738 (0.0008)
+[2026-06-06 21:51:15,750][262198] Updated weights for policy 0, policy_version 30752 (0.0008)
+[2026-06-06 21:51:17,184][262198] Updated weights for policy 0, policy_version 30764 (0.0008)
+[2026-06-06 21:51:17,921][261293] Fps is (10 sec: 4096.2, 60 sec: 3515.7, 300 sec: 3179.6). Total num frames: 15753216. Throughput: 0: 3453.6. Samples: 15753410. Policy #0 lag: (min: 3.0, avg: 21.9, max: 35.0)
+[2026-06-06 21:51:17,925][261293] Avg episode reward: [(0, '1686.727')]
+[2026-06-06 21:51:18,688][262198] Updated weights for policy 0, policy_version 30776 (0.0007)
+[2026-06-06 21:51:19,589][262198] Updated weights for policy 0, policy_version 30787 (0.0008)
+[2026-06-06 21:51:21,446][262198] Updated weights for policy 0, policy_version 30797 (0.0008)
+[2026-06-06 21:51:22,645][262198] Updated weights for policy 0, policy_version 30808 (0.0009)
+[2026-06-06 21:51:22,921][261293] Fps is (10 sec: 4095.9, 60 sec: 3549.8, 300 sec: 3193.5). Total num frames: 15773696. Throughput: 0: 3629.0. Samples: 15778578. Policy #0 lag: (min: 3.0, avg: 21.9, max: 35.0)
+[2026-06-06 21:51:22,924][261293] Avg episode reward: [(0, '1675.831')]
+[2026-06-06 21:51:23,899][262198] Updated weights for policy 0, policy_version 30821 (0.0010)
+[2026-06-06 21:51:25,376][262198] Updated weights for policy 0, policy_version 30831 (0.0008)
+[2026-06-06 21:51:26,757][262198] Updated weights for policy 0, policy_version 30841 (0.0008)
+[2026-06-06 21:51:27,901][262198] Updated weights for policy 0, policy_version 30856 (0.0010)
+[2026-06-06 21:51:27,923][261293] Fps is (10 sec: 4504.3, 60 sec: 3652.1, 300 sec: 3235.1). Total num frames: 15798272. Throughput: 0: 3784.5. Samples: 15803862. Policy #0 lag: (min: 3.0, avg: 21.9, max: 35.0)
+[2026-06-06 21:51:27,927][261293] Avg episode reward: [(0, '1550.245')]
+[2026-06-06 21:51:30,087][262198] Updated weights for policy 0, policy_version 30866 (0.0008)
+[2026-06-06 21:51:31,368][262198] Updated weights for policy 0, policy_version 30880 (0.0010)
+[2026-06-06 21:51:32,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3618.1, 300 sec: 3249.0). Total num frames: 15814656. Throughput: 0: 3883.3. Samples: 15817666. Policy #0 lag: (min: 3.0, avg: 20.0, max: 36.0)
+[2026-06-06 21:51:32,923][261293] Avg episode reward: [(0, '1555.155')]
+[2026-06-06 21:51:33,019][262198] Updated weights for policy 0, policy_version 30891 (0.0012)
+[2026-06-06 21:51:34,947][262198] Updated weights for policy 0, policy_version 30901 (0.0014)
+[2026-06-06 21:51:36,263][262198] Updated weights for policy 0, policy_version 30913 (0.0012)
+[2026-06-06 21:51:37,921][261293] Fps is (10 sec: 3277.6, 60 sec: 3720.6, 300 sec: 3269.9). Total num frames: 15831040. Throughput: 0: 3845.4. Samples: 15836398. Policy #0 lag: (min: 3.0, avg: 20.0, max: 36.0)
+[2026-06-06 21:51:37,923][261293] Avg episode reward: [(0, '1642.167')]
+[2026-06-06 21:51:38,226][262198] Updated weights for policy 0, policy_version 30923 (0.0010)
+[2026-06-06 21:51:40,368][262198] Updated weights for policy 0, policy_version 30933 (0.0013)
+[2026-06-06 21:51:41,820][262115] Signal inference workers to stop experience collection... (3250 times)
+[2026-06-06 21:51:41,849][262198] InferenceWorker_p0-w0: stopping experience collection (3250 times)
+[2026-06-06 21:51:41,932][262115] Signal inference workers to resume experience collection... (3250 times)
+[2026-06-06 21:51:41,933][262198] InferenceWorker_p0-w0: resuming experience collection (3250 times)
+[2026-06-06 21:51:41,935][262198] Updated weights for policy 0, policy_version 30944 (0.0010)
+[2026-06-06 21:51:42,922][261293] Fps is (10 sec: 3276.8, 60 sec: 3720.5, 300 sec: 3276.8). Total num frames: 15847424. Throughput: 0: 3765.5. Samples: 15854496. Policy #0 lag: (min: 3.0, avg: 20.0, max: 36.0)
+[2026-06-06 21:51:42,926][261293] Avg episode reward: [(0, '1710.821')]
+[2026-06-06 21:51:43,466][262198] Updated weights for policy 0, policy_version 30956 (0.0009)
+[2026-06-06 21:51:44,995][262198] Updated weights for policy 0, policy_version 30966 (0.0009)
+[2026-06-06 21:51:46,856][262198] Updated weights for policy 0, policy_version 30982 (0.0009)
+[2026-06-06 21:51:47,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3754.6, 300 sec: 3290.7). Total num frames: 15865856. Throughput: 0: 3818.1. Samples: 15867308. Policy #0 lag: (min: 3.0, avg: 20.0, max: 36.0)
+[2026-06-06 21:51:47,924][261293] Avg episode reward: [(0, '1700.604')]
+[2026-06-06 21:51:48,012][262198] Updated weights for policy 0, policy_version 30992 (0.0008)
+[2026-06-06 21:51:49,586][262198] Updated weights for policy 0, policy_version 31002 (0.0008)
+[2026-06-06 21:51:50,687][262198] Updated weights for policy 0, policy_version 31016 (0.0008)
+[2026-06-06 21:51:52,534][262198] Updated weights for policy 0, policy_version 31026 (0.0008)
+[2026-06-06 21:51:52,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3823.5, 300 sec: 3318.5). Total num frames: 15886336. Throughput: 0: 3831.9. Samples: 15890628. Policy #0 lag: (min: 5.0, avg: 21.1, max: 37.0)
+[2026-06-06 21:51:52,924][261293] Avg episode reward: [(0, '1624.506')]
+[2026-06-06 21:51:54,221][262198] Updated weights for policy 0, policy_version 31037 (0.0009)
+[2026-06-06 21:51:55,730][262198] Updated weights for policy 0, policy_version 31050 (0.0009)
+[2026-06-06 21:51:56,843][262198] Updated weights for policy 0, policy_version 31060 (0.0008)
+[2026-06-06 21:51:57,922][261293] Fps is (10 sec: 3891.0, 60 sec: 3822.9, 300 sec: 3339.3). Total num frames: 15904768. Throughput: 0: 3877.2. Samples: 15914798. Policy #0 lag: (min: 5.0, avg: 21.1, max: 37.0)
+[2026-06-06 21:51:57,926][261293] Avg episode reward: [(0, '1599.831')]
+[2026-06-06 21:51:58,763][262198] Updated weights for policy 0, policy_version 31073 (0.0009)
+[2026-06-06 21:51:59,827][262198] Updated weights for policy 0, policy_version 31084 (0.0009)
+[2026-06-06 21:52:01,531][262198] Updated weights for policy 0, policy_version 31094 (0.0008)
+[2026-06-06 21:52:02,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3857.1, 300 sec: 3353.2). Total num frames: 15925248. Throughput: 0: 3844.5. Samples: 15926416. Policy #0 lag: (min: 5.0, avg: 21.1, max: 37.0)
+[2026-06-06 21:52:02,925][261293] Avg episode reward: [(0, '1700.430')]
+[2026-06-06 21:52:03,299][262198] Updated weights for policy 0, policy_version 31110 (0.0010)
+[2026-06-06 21:52:04,753][262198] Updated weights for policy 0, policy_version 31120 (0.0011)
+[2026-06-06 21:52:07,212][262198] Updated weights for policy 0, policy_version 31130 (0.0014)
+[2026-06-06 21:52:07,921][261293] Fps is (10 sec: 3686.6, 60 sec: 3822.9, 300 sec: 3339.3). Total num frames: 15941632. Throughput: 0: 3748.1. Samples: 15947242. Policy #0 lag: (min: 5.0, avg: 21.1, max: 37.0)
+[2026-06-06 21:52:07,925][261293] Avg episode reward: [(0, '1727.464')]
+[2026-06-06 21:52:08,137][262198] Updated weights for policy 0, policy_version 31142 (0.0012)
+[2026-06-06 21:52:10,248][262198] Updated weights for policy 0, policy_version 31152 (0.0011)
+[2026-06-06 21:52:12,871][262198] Updated weights for policy 0, policy_version 31162 (0.0013)
+[2026-06-06 21:52:12,921][261293] Fps is (10 sec: 2867.3, 60 sec: 3686.4, 300 sec: 3325.4). Total num frames: 15953920. Throughput: 0: 3564.8. Samples: 15964266. Policy #0 lag: (min: 3.0, avg: 20.4, max: 39.0)
+[2026-06-06 21:52:12,925][261293] Avg episode reward: [(0, '1841.122')]
+[2026-06-06 21:52:14,008][262198] Updated weights for policy 0, policy_version 31174 (0.0010)
+[2026-06-06 21:52:15,806][262198] Updated weights for policy 0, policy_version 31184 (0.0011)
+[2026-06-06 21:52:17,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3618.1, 300 sec: 3339.3). Total num frames: 15970304. Throughput: 0: 3458.1. Samples: 15973282. Policy #0 lag: (min: 3.0, avg: 20.4, max: 39.0)
+[2026-06-06 21:52:17,926][261293] Avg episode reward: [(0, '1639.135')]
+[2026-06-06 21:52:18,179][262198] Updated weights for policy 0, policy_version 31194 (0.0012)
+[2026-06-06 21:52:19,321][262198] Updated weights for policy 0, policy_version 31205 (0.0010)
+[2026-06-06 21:52:21,001][262198] Updated weights for policy 0, policy_version 31215 (0.0010)
+[2026-06-06 21:52:22,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3549.9, 300 sec: 3346.2). Total num frames: 15986688. Throughput: 0: 3458.2. Samples: 15992014. Policy #0 lag: (min: 3.0, avg: 20.4, max: 39.0)
+[2026-06-06 21:52:22,922][261293] Avg episode reward: [(0, '1767.607')]
+[2026-06-06 21:52:23,141][262198] Updated weights for policy 0, policy_version 31225 (0.0009)
+[2026-06-06 21:52:24,146][262198] Updated weights for policy 0, policy_version 31239 (0.0009)
+[2026-06-06 21:52:25,921][262198] Updated weights for policy 0, policy_version 31249 (0.0009)
+[2026-06-06 21:52:27,508][262198] Updated weights for policy 0, policy_version 31261 (0.0008)
+[2026-06-06 21:52:27,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3481.7, 300 sec: 3367.0). Total num frames: 16007168. Throughput: 0: 3589.9. Samples: 16016044. Policy #0 lag: (min: 3.0, avg: 20.4, max: 39.0)
+[2026-06-06 21:52:27,925][261293] Avg episode reward: [(0, '1604.475')]
+[2026-06-06 21:52:29,450][262198] Updated weights for policy 0, policy_version 31273 (0.0014)
+[2026-06-06 21:52:30,589][262198] Updated weights for policy 0, policy_version 31283 (0.0011)
+[2026-06-06 21:52:32,458][262198] Updated weights for policy 0, policy_version 31295 (0.0011)
+[2026-06-06 21:52:32,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3515.7, 300 sec: 3360.1). Total num frames: 16025600. Throughput: 0: 3514.3. Samples: 16025450. Policy #0 lag: (min: 3.0, avg: 21.0, max: 36.0)
+[2026-06-06 21:52:32,923][261293] Avg episode reward: [(0, '1706.183')]
+[2026-06-06 21:52:34,282][262198] Updated weights for policy 0, policy_version 31306 (0.0010)
+[2026-06-06 21:52:35,457][262198] Updated weights for policy 0, policy_version 31316 (0.0010)
+[2026-06-06 21:52:37,263][262198] Updated weights for policy 0, policy_version 31326 (0.0009)
+[2026-06-06 21:52:37,923][261293] Fps is (10 sec: 3481.0, 60 sec: 3515.6, 300 sec: 3346.2). Total num frames: 16041984. Throughput: 0: 3453.6. Samples: 16046046. Policy #0 lag: (min: 3.0, avg: 21.0, max: 36.0)
+[2026-06-06 21:52:37,927][261293] Avg episode reward: [(0, '1651.713')]
+[2026-06-06 21:52:38,926][262198] Updated weights for policy 0, policy_version 31337 (0.0011)
+[2026-06-06 21:52:40,014][262198] Updated weights for policy 0, policy_version 31347 (0.0009)
+[2026-06-06 21:52:41,701][262198] Updated weights for policy 0, policy_version 31357 (0.0011)
+[2026-06-06 21:52:42,574][262198] Updated weights for policy 0, policy_version 31368 (0.0007)
+[2026-06-06 21:52:42,921][261293] Fps is (10 sec: 3481.7, 60 sec: 3549.9, 300 sec: 3339.3). Total num frames: 16060416. Throughput: 0: 3391.8. Samples: 16067428. Policy #0 lag: (min: 3.0, avg: 21.0, max: 36.0)
+[2026-06-06 21:52:42,926][261293] Avg episode reward: [(0, '1608.520')]
+[2026-06-06 21:52:43,891][262198] Updated weights for policy 0, policy_version 31378 (0.0008)
+[2026-06-06 21:52:45,871][262198] Updated weights for policy 0, policy_version 31392 (0.0009)
+[2026-06-06 21:52:47,279][262198] Updated weights for policy 0, policy_version 31402 (0.0009)
+[2026-06-06 21:52:47,921][261293] Fps is (10 sec: 3892.0, 60 sec: 3584.0, 300 sec: 3346.2). Total num frames: 16080896. Throughput: 0: 3432.1. Samples: 16080858. Policy #0 lag: (min: 3.0, avg: 21.0, max: 36.0)
+[2026-06-06 21:52:47,923][261293] Avg episode reward: [(0, '1621.888')]
+[2026-06-06 21:52:48,323][262198] Updated weights for policy 0, policy_version 31412 (0.0008)
+[2026-06-06 21:52:49,720][262115] Signal inference workers to stop experience collection... (3300 times)
+[2026-06-06 21:52:49,760][262198] InferenceWorker_p0-w0: stopping experience collection (3300 times)
+[2026-06-06 21:52:49,866][262115] Signal inference workers to resume experience collection... (3300 times)
+[2026-06-06 21:52:49,866][262198] InferenceWorker_p0-w0: resuming experience collection (3300 times)
+[2026-06-06 21:52:50,354][262198] Updated weights for policy 0, policy_version 31425 (0.0010)
+[2026-06-06 21:52:51,703][262198] Updated weights for policy 0, policy_version 31435 (0.0008)
+[2026-06-06 21:52:52,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3549.9, 300 sec: 3346.2). Total num frames: 16099328. Throughput: 0: 3449.3. Samples: 16102458. Policy #0 lag: (min: 10.0, avg: 25.0, max: 46.0)
+[2026-06-06 21:52:52,923][261293] Avg episode reward: [(0, '1806.157')]
+[2026-06-06 21:52:53,393][262198] Updated weights for policy 0, policy_version 31445 (0.0008)
+[2026-06-06 21:52:54,948][262198] Updated weights for policy 0, policy_version 31459 (0.0010)
+[2026-06-06 21:52:56,408][262198] Updated weights for policy 0, policy_version 31469 (0.0008)
+[2026-06-06 21:52:57,922][261293] Fps is (10 sec: 3481.3, 60 sec: 3515.7, 300 sec: 3353.2). Total num frames: 16115712. Throughput: 0: 3568.3. Samples: 16124842. Policy #0 lag: (min: 10.0, avg: 25.0, max: 46.0)
+[2026-06-06 21:52:57,926][261293] Avg episode reward: [(0, '1619.275')]
+[2026-06-06 21:52:58,034][262198] Updated weights for policy 0, policy_version 31479 (0.0007)
+[2026-06-06 21:52:59,580][262198] Updated weights for policy 0, policy_version 31491 (0.0009)
+[2026-06-06 21:53:00,897][262198] Updated weights for policy 0, policy_version 31501 (0.0007)
+[2026-06-06 21:53:02,254][262198] Updated weights for policy 0, policy_version 31511 (0.0010)
+[2026-06-06 21:53:02,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3481.6, 300 sec: 3360.1). Total num frames: 16134144. Throughput: 0: 3616.6. Samples: 16136026. Policy #0 lag: (min: 10.0, avg: 25.0, max: 46.0)
+[2026-06-06 21:53:02,923][261293] Avg episode reward: [(0, '1685.526')]
+[2026-06-06 21:53:03,546][262198] Updated weights for policy 0, policy_version 31523 (0.0008)
+[2026-06-06 21:53:04,962][262198] Updated weights for policy 0, policy_version 31533 (0.0008)
+[2026-06-06 21:53:06,214][262198] Updated weights for policy 0, policy_version 31543 (0.0008)
+[2026-06-06 21:53:07,582][262198] Updated weights for policy 0, policy_version 31556 (0.0009)
+[2026-06-06 21:53:07,921][261293] Fps is (10 sec: 4096.4, 60 sec: 3584.0, 300 sec: 3380.9). Total num frames: 16156672. Throughput: 0: 3744.6. Samples: 16160522. Policy #0 lag: (min: 10.0, avg: 25.0, max: 46.0)
+[2026-06-06 21:53:07,923][261293] Avg episode reward: [(0, '1512.530')]
+[2026-06-06 21:53:09,368][262198] Updated weights for policy 0, policy_version 31568 (0.0009)
+[2026-06-06 21:53:11,303][262198] Updated weights for policy 0, policy_version 31578 (0.0009)
+[2026-06-06 21:53:12,320][262198] Updated weights for policy 0, policy_version 31590 (0.0008)
+[2026-06-06 21:53:12,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3686.4, 300 sec: 3387.9). Total num frames: 16175104. Throughput: 0: 3706.3. Samples: 16182828. Policy #0 lag: (min: 7.0, avg: 20.2, max: 39.0)
+[2026-06-06 21:53:12,924][261293] Avg episode reward: [(0, '1479.090')]
+[2026-06-06 21:53:13,747][262198] Updated weights for policy 0, policy_version 31600 (0.0008)
+[2026-06-06 21:53:15,875][262198] Updated weights for policy 0, policy_version 31612 (0.0009)
+[2026-06-06 21:53:16,803][262198] Updated weights for policy 0, policy_version 31624 (0.0009)
+[2026-06-06 21:53:17,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3720.6, 300 sec: 3394.8). Total num frames: 16193536. Throughput: 0: 3762.4. Samples: 16194758. Policy #0 lag: (min: 7.0, avg: 20.2, max: 39.0)
+[2026-06-06 21:53:17,924][261293] Avg episode reward: [(0, '1624.389')]
+[2026-06-06 21:53:18,266][262198] Updated weights for policy 0, policy_version 31635 (0.0011)
+[2026-06-06 21:53:20,227][262198] Updated weights for policy 0, policy_version 31646 (0.0009)
+[2026-06-06 21:53:21,567][262198] Updated weights for policy 0, policy_version 31657 (0.0009)
+[2026-06-06 21:53:22,529][262198] Updated weights for policy 0, policy_version 31668 (0.0009)
+[2026-06-06 21:53:22,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3788.8, 300 sec: 3408.7). Total num frames: 16214016. Throughput: 0: 3806.9. Samples: 16217348. Policy #0 lag: (min: 7.0, avg: 20.2, max: 39.0)
+[2026-06-06 21:53:22,924][261293] Avg episode reward: [(0, '1654.538')]
+[2026-06-06 21:53:24,697][262198] Updated weights for policy 0, policy_version 31681 (0.0009)
+[2026-06-06 21:53:25,866][262198] Updated weights for policy 0, policy_version 31691 (0.0008)
+[2026-06-06 21:53:27,545][262198] Updated weights for policy 0, policy_version 31702 (0.0010)
+[2026-06-06 21:53:27,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3754.7, 300 sec: 3415.6). Total num frames: 16232448. Throughput: 0: 3853.3. Samples: 16240828. Policy #0 lag: (min: 7.0, avg: 20.2, max: 39.0)
+[2026-06-06 21:53:27,922][261293] Avg episode reward: [(0, '1610.951')]
+[2026-06-06 21:53:28,974][262198] Updated weights for policy 0, policy_version 31715 (0.0009)
+[2026-06-06 21:53:30,569][262198] Updated weights for policy 0, policy_version 31725 (0.0009)
+[2026-06-06 21:53:31,918][262198] Updated weights for policy 0, policy_version 31735 (0.0009)
+[2026-06-06 21:53:32,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3754.7, 300 sec: 3429.5). Total num frames: 16250880. Throughput: 0: 3798.2. Samples: 16251778. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:53:32,923][261293] Avg episode reward: [(0, '1790.434')]
+[2026-06-06 21:53:34,188][262198] Updated weights for policy 0, policy_version 31751 (0.0013)
+[2026-06-06 21:53:35,752][262198] Updated weights for policy 0, policy_version 31761 (0.0011)
+[2026-06-06 21:53:37,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3720.7, 300 sec: 3429.5). Total num frames: 16265216. Throughput: 0: 3758.0. Samples: 16271570. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:53:37,923][261293] Avg episode reward: [(0, '1665.684')]
+[2026-06-06 21:53:38,253][262198] Updated weights for policy 0, policy_version 31772 (0.0013)
+[2026-06-06 21:53:39,731][262198] Updated weights for policy 0, policy_version 31784 (0.0011)
+[2026-06-06 21:53:41,347][262198] Updated weights for policy 0, policy_version 31795 (0.0011)
+[2026-06-06 21:53:42,921][261293] Fps is (10 sec: 3072.0, 60 sec: 3686.4, 300 sec: 3415.6). Total num frames: 16281600. Throughput: 0: 3670.3. Samples: 16290002. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:53:42,923][261293] Avg episode reward: [(0, '1700.785')]
+[2026-06-06 21:53:44,215][262198] Updated weights for policy 0, policy_version 31809 (0.0012)
+[2026-06-06 21:53:45,915][262198] Updated weights for policy 0, policy_version 31819 (0.0013)
+[2026-06-06 21:53:47,922][261293] Fps is (10 sec: 3071.9, 60 sec: 3584.0, 300 sec: 3408.7). Total num frames: 16295936. Throughput: 0: 3604.5. Samples: 16298230. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:53:47,928][261293] Avg episode reward: [(0, '1508.506')]
+[2026-06-06 21:53:48,302][262198] Updated weights for policy 0, policy_version 31831 (0.0013)
+[2026-06-06 21:53:49,958][262198] Updated weights for policy 0, policy_version 31843 (0.0011)
+[2026-06-06 21:53:52,058][262198] Updated weights for policy 0, policy_version 31853 (0.0009)
+[2026-06-06 21:53:52,131][262115] Signal inference workers to stop experience collection... (3350 times)
+[2026-06-06 21:53:52,153][262198] InferenceWorker_p0-w0: stopping experience collection (3350 times)
+[2026-06-06 21:53:52,181][262115] Signal inference workers to resume experience collection... (3350 times)
+[2026-06-06 21:53:52,182][262198] InferenceWorker_p0-w0: resuming experience collection (3350 times)
+[2026-06-06 21:53:52,920][261293] Fps is (10 sec: 3072.1, 60 sec: 3549.9, 300 sec: 3408.7). Total num frames: 16312320. Throughput: 0: 3454.4. Samples: 16315970. Policy #0 lag: (min: 3.0, avg: 17.2, max: 35.0)
+[2026-06-06 21:53:52,922][261293] Avg episode reward: [(0, '1604.360')]
+[2026-06-06 21:53:53,586][262198] Updated weights for policy 0, policy_version 31864 (0.0011)
+[2026-06-06 21:53:55,755][262198] Updated weights for policy 0, policy_version 31877 (0.0011)
+[2026-06-06 21:53:57,400][262198] Updated weights for policy 0, policy_version 31887 (0.0011)
+[2026-06-06 21:53:57,927][261293] Fps is (10 sec: 3275.0, 60 sec: 3549.6, 300 sec: 3415.6). Total num frames: 16328704. Throughput: 0: 3377.0. Samples: 16334814. Policy #0 lag: (min: 2.0, avg: 20.1, max: 34.0)
+[2026-06-06 21:53:57,931][261293] Avg episode reward: [(0, '1751.104')]
+[2026-06-06 21:53:59,538][262198] Updated weights for policy 0, policy_version 31897 (0.0009)
+[2026-06-06 21:54:00,897][262198] Updated weights for policy 0, policy_version 31909 (0.0010)
+[2026-06-06 21:54:02,381][262198] Updated weights for policy 0, policy_version 31920 (0.0008)
+[2026-06-06 21:54:02,921][261293] Fps is (10 sec: 3276.5, 60 sec: 3515.7, 300 sec: 3422.6). Total num frames: 16345088. Throughput: 0: 3345.2. Samples: 16345294. Policy #0 lag: (min: 2.0, avg: 20.1, max: 34.0)
+[2026-06-06 21:54:02,923][261293] Avg episode reward: [(0, '1600.681')]
+[2026-06-06 21:54:04,624][262198] Updated weights for policy 0, policy_version 31932 (0.0010)
+[2026-06-06 21:54:05,802][262198] Updated weights for policy 0, policy_version 31943 (0.0009)
+[2026-06-06 21:54:07,058][262198] Updated weights for policy 0, policy_version 31954 (0.0009)
+[2026-06-06 21:54:07,921][261293] Fps is (10 sec: 3278.8, 60 sec: 3413.3, 300 sec: 3422.6). Total num frames: 16361472. Throughput: 0: 3301.0. Samples: 16365892. Policy #0 lag: (min: 2.0, avg: 20.1, max: 34.0)
+[2026-06-06 21:54:07,923][261293] Avg episode reward: [(0, '1579.786')]
+[2026-06-06 21:54:09,046][262198] Updated weights for policy 0, policy_version 31964 (0.0008)
+[2026-06-06 21:54:10,271][262198] Updated weights for policy 0, policy_version 31974 (0.0008)
+[2026-06-06 21:54:11,583][262198] Updated weights for policy 0, policy_version 31986 (0.0009)
+[2026-06-06 21:54:12,921][261293] Fps is (10 sec: 3481.8, 60 sec: 3413.4, 300 sec: 3443.4). Total num frames: 16379904. Throughput: 0: 3257.3. Samples: 16387406. Policy #0 lag: (min: 2.0, avg: 20.1, max: 34.0)
+[2026-06-06 21:54:12,923][261293] Avg episode reward: [(0, '1398.611')]
+[2026-06-06 21:54:13,748][262198] Updated weights for policy 0, policy_version 31998 (0.0009)
+[2026-06-06 21:54:14,951][262198] Updated weights for policy 0, policy_version 32008 (0.0008)
+[2026-06-06 21:54:16,175][262198] Updated weights for policy 0, policy_version 32020 (0.0010)
+[2026-06-06 21:54:17,924][261293] Fps is (10 sec: 3480.3, 60 sec: 3379.0, 300 sec: 3450.3). Total num frames: 16396288. Throughput: 0: 3259.6. Samples: 16398472. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:54:17,927][261293] Avg episode reward: [(0, '1547.126')]
+[2026-06-06 21:54:18,153][262198] Updated weights for policy 0, policy_version 32030 (0.0010)
+[2026-06-06 21:54:19,305][262198] Updated weights for policy 0, policy_version 32040 (0.0008)
+[2026-06-06 21:54:20,547][262198] Updated weights for policy 0, policy_version 32052 (0.0008)
+[2026-06-06 21:54:22,527][262198] Updated weights for policy 0, policy_version 32064 (0.0008)
+[2026-06-06 21:54:22,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3379.2, 300 sec: 3471.2). Total num frames: 16416768. Throughput: 0: 3322.1. Samples: 16421064. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:54:22,925][261293] Avg episode reward: [(0, '1554.760')]
+[2026-06-06 21:54:24,114][262198] Updated weights for policy 0, policy_version 32074 (0.0009)
+[2026-06-06 21:54:24,868][262198] Updated weights for policy 0, policy_version 32084 (0.0008)
+[2026-06-06 21:54:27,000][262198] Updated weights for policy 0, policy_version 32095 (0.0009)
+[2026-06-06 21:54:27,921][261293] Fps is (10 sec: 3892.3, 60 sec: 3379.1, 300 sec: 3485.1). Total num frames: 16435200. Throughput: 0: 3406.3. Samples: 16443286. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:54:27,925][261293] Avg episode reward: [(0, '1481.557')]
+[2026-06-06 21:54:28,844][262198] Updated weights for policy 0, policy_version 32105 (0.0008)
+[2026-06-06 21:54:30,504][262198] Updated weights for policy 0, policy_version 32118 (0.0010)
+[2026-06-06 21:54:31,821][262198] Updated weights for policy 0, policy_version 32130 (0.0009)
+[2026-06-06 21:54:32,920][261293] Fps is (10 sec: 3481.8, 60 sec: 3345.1, 300 sec: 3485.1). Total num frames: 16451584. Throughput: 0: 3462.0. Samples: 16454016. Policy #0 lag: (min: 3.0, avg: 20.3, max: 35.0)
+[2026-06-06 21:54:32,922][261293] Avg episode reward: [(0, '1772.837')]
+[2026-06-06 21:54:33,772][262198] Updated weights for policy 0, policy_version 32141 (0.0010)
+[2026-06-06 21:54:35,120][262198] Updated weights for policy 0, policy_version 32152 (0.0008)
+[2026-06-06 21:54:36,499][262198] Updated weights for policy 0, policy_version 32164 (0.0012)
+[2026-06-06 21:54:37,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3413.3, 300 sec: 3498.9). Total num frames: 16470016. Throughput: 0: 3555.9. Samples: 16475988. Policy #0 lag: (min: 4.0, avg: 18.1, max: 36.0)
+[2026-06-06 21:54:37,925][261293] Avg episode reward: [(0, '1722.668')]
+[2026-06-06 21:54:38,542][262198] Updated weights for policy 0, policy_version 32174 (0.0010)
+[2026-06-06 21:54:39,870][262198] Updated weights for policy 0, policy_version 32184 (0.0010)
+[2026-06-06 21:54:41,347][262198] Updated weights for policy 0, policy_version 32196 (0.0010)
+[2026-06-06 21:54:42,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3447.5, 300 sec: 3512.8). Total num frames: 16488448. Throughput: 0: 3590.7. Samples: 16496372. Policy #0 lag: (min: 4.0, avg: 18.1, max: 36.0)
+[2026-06-06 21:54:42,922][261293] Avg episode reward: [(0, '1676.488')]
+[2026-06-06 21:54:42,930][262198] Updated weights for policy 0, policy_version 32206 (0.0009)
+[2026-06-06 21:54:44,346][262198] Updated weights for policy 0, policy_version 32216 (0.0009)
+[2026-06-06 21:54:46,842][262198] Updated weights for policy 0, policy_version 32229 (0.0009)
+[2026-06-06 21:54:47,856][262198] Updated weights for policy 0, policy_version 32242 (0.0010)
+[2026-06-06 21:54:47,921][261293] Fps is (10 sec: 3686.6, 60 sec: 3515.8, 300 sec: 3519.8). Total num frames: 16506880. Throughput: 0: 3608.6. Samples: 16507678. Policy #0 lag: (min: 4.0, avg: 18.1, max: 36.0)
+[2026-06-06 21:54:47,923][261293] Avg episode reward: [(0, '1731.459')]
+[2026-06-06 21:54:49,926][262198] Updated weights for policy 0, policy_version 32254 (0.0010)
+[2026-06-06 21:54:52,079][262198] Updated weights for policy 0, policy_version 32265 (0.0010)
+[2026-06-06 21:54:52,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3481.6, 300 sec: 3519.8). Total num frames: 16521216. Throughput: 0: 3591.9. Samples: 16527530. Policy #0 lag: (min: 4.0, avg: 18.1, max: 36.0)
+[2026-06-06 21:54:52,923][261293] Avg episode reward: [(0, '1584.142')]
+[2026-06-06 21:54:53,106][262198] Updated weights for policy 0, policy_version 32275 (0.0010)
+[2026-06-06 21:54:55,466][262198] Updated weights for policy 0, policy_version 32285 (0.0010)
+[2026-06-06 21:54:57,247][262198] Updated weights for policy 0, policy_version 32296 (0.0010)
+[2026-06-06 21:54:57,563][262115] Signal inference workers to stop experience collection... (3400 times)
+[2026-06-06 21:54:57,597][262198] InferenceWorker_p0-w0: stopping experience collection (3400 times)
+[2026-06-06 21:54:57,615][262115] Signal inference workers to resume experience collection... (3400 times)
+[2026-06-06 21:54:57,615][262198] InferenceWorker_p0-w0: resuming experience collection (3400 times)
+[2026-06-06 21:54:57,922][261293] Fps is (10 sec: 3071.9, 60 sec: 3481.9, 300 sec: 3519.8). Total num frames: 16537600. Throughput: 0: 3535.1. Samples: 16546488. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:54:57,926][261293] Avg episode reward: [(0, '1683.958')]
+[2026-06-06 21:54:58,162][262198] Updated weights for policy 0, policy_version 32306 (0.0010)
+[2026-06-06 21:55:00,354][262198] Updated weights for policy 0, policy_version 32316 (0.0011)
+[2026-06-06 21:55:02,483][262198] Updated weights for policy 0, policy_version 32328 (0.0011)
+[2026-06-06 21:55:02,922][261293] Fps is (10 sec: 3276.4, 60 sec: 3481.6, 300 sec: 3533.7). Total num frames: 16553984. Throughput: 0: 3498.7. Samples: 16555904. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:55:02,927][261293] Avg episode reward: [(0, '1683.219')]
+[2026-06-06 21:55:03,652][262198] Updated weights for policy 0, policy_version 32338 (0.0012)
+[2026-06-06 21:55:05,971][262198] Updated weights for policy 0, policy_version 32349 (0.0010)
+[2026-06-06 21:55:07,537][262198] Updated weights for policy 0, policy_version 32360 (0.0010)
+[2026-06-06 21:55:07,921][261293] Fps is (10 sec: 3072.1, 60 sec: 3447.5, 300 sec: 3526.7). Total num frames: 16568320. Throughput: 0: 3396.6. Samples: 16573908. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:55:07,924][261293] Avg episode reward: [(0, '1630.131')]
+[2026-06-06 21:55:08,814][262198] Updated weights for policy 0, policy_version 32370 (0.0010)
+[2026-06-06 21:55:11,259][262198] Updated weights for policy 0, policy_version 32385 (0.0011)
+[2026-06-06 21:55:12,921][261293] Fps is (10 sec: 3072.3, 60 sec: 3413.3, 300 sec: 3533.7). Total num frames: 16584704. Throughput: 0: 3348.0. Samples: 16593944. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:55:12,922][261293] Avg episode reward: [(0, '1676.765')]
+[2026-06-06 21:55:13,022][262198] Updated weights for policy 0, policy_version 32396 (0.0011)
+[2026-06-06 21:55:14,915][262198] Updated weights for policy 0, policy_version 32406 (0.0010)
+[2026-06-06 21:55:16,539][262198] Updated weights for policy 0, policy_version 32419 (0.0009)
+[2026-06-06 21:55:17,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3413.5, 300 sec: 3526.7). Total num frames: 16601088. Throughput: 0: 3326.8. Samples: 16603726. Policy #0 lag: (min: 3.0, avg: 19.6, max: 35.0)
+[2026-06-06 21:55:17,924][261293] Avg episode reward: [(0, '1789.038')]
+[2026-06-06 21:55:18,212][262198] Updated weights for policy 0, policy_version 32431 (0.0010)
+[2026-06-06 21:55:20,457][262198] Updated weights for policy 0, policy_version 32441 (0.0009)
+[2026-06-06 21:55:22,019][262198] Updated weights for policy 0, policy_version 32453 (0.0010)
+[2026-06-06 21:55:22,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3379.2, 300 sec: 3526.7). Total num frames: 16619520. Throughput: 0: 3291.4. Samples: 16624098. Policy #0 lag: (min: 7.0, avg: 21.7, max: 39.0)
+[2026-06-06 21:55:22,925][261293] Avg episode reward: [(0, '1852.842')]
+[2026-06-06 21:55:22,937][262198] Updated weights for policy 0, policy_version 32463 (0.0010)
+[2026-06-06 21:55:25,155][262198] Updated weights for policy 0, policy_version 32473 (0.0009)
+[2026-06-06 21:55:26,951][262198] Updated weights for policy 0, policy_version 32485 (0.0013)
+[2026-06-06 21:55:27,920][261293] Fps is (10 sec: 3481.8, 60 sec: 3345.1, 300 sec: 3519.8). Total num frames: 16635904. Throughput: 0: 3287.2. Samples: 16644296. Policy #0 lag: (min: 7.0, avg: 21.7, max: 39.0)
+[2026-06-06 21:55:27,922][261293] Avg episode reward: [(0, '1633.435')]
+[2026-06-06 21:55:28,243][262198] Updated weights for policy 0, policy_version 32497 (0.0010)
+[2026-06-06 21:55:30,293][262198] Updated weights for policy 0, policy_version 32508 (0.0010)
+[2026-06-06 21:55:31,673][262198] Updated weights for policy 0, policy_version 32519 (0.0010)
+[2026-06-06 21:55:32,764][262198] Updated weights for policy 0, policy_version 32530 (0.0011)
+[2026-06-06 21:55:32,920][261293] Fps is (10 sec: 3686.6, 60 sec: 3413.3, 300 sec: 3554.5). Total num frames: 16656384. Throughput: 0: 3274.8. Samples: 16655044. Policy #0 lag: (min: 7.0, avg: 21.7, max: 39.0)
+[2026-06-06 21:55:32,921][261293] Avg episode reward: [(0, '1829.317')]
+[2026-06-06 21:55:35,106][262198] Updated weights for policy 0, policy_version 32541 (0.0010)
+[2026-06-06 21:55:36,860][262198] Updated weights for policy 0, policy_version 32553 (0.0009)
+[2026-06-06 21:55:37,632][262198] Updated weights for policy 0, policy_version 32563 (0.0009)
+[2026-06-06 21:55:37,921][261293] Fps is (10 sec: 3686.1, 60 sec: 3379.2, 300 sec: 3554.5). Total num frames: 16672768. Throughput: 0: 3292.8. Samples: 16675708. Policy #0 lag: (min: 7.0, avg: 21.7, max: 39.0)
+[2026-06-06 21:55:37,923][261293] Avg episode reward: [(0, '1480.085')]
+[2026-06-06 21:55:39,779][262198] Updated weights for policy 0, policy_version 32573 (0.0009)
+[2026-06-06 21:55:41,781][262198] Updated weights for policy 0, policy_version 32585 (0.0014)
+[2026-06-06 21:55:42,731][262198] Updated weights for policy 0, policy_version 32595 (0.0015)
+[2026-06-06 21:55:42,920][261293] Fps is (10 sec: 3276.7, 60 sec: 3345.0, 300 sec: 3554.5). Total num frames: 16689152. Throughput: 0: 3297.0. Samples: 16694850. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:55:42,923][261293] Avg episode reward: [(0, '1725.264')]
+[2026-06-06 21:55:44,899][262198] Updated weights for policy 0, policy_version 32605 (0.0010)
+[2026-06-06 21:55:47,042][262198] Updated weights for policy 0, policy_version 32620 (0.0010)
+[2026-06-06 21:55:47,921][261293] Fps is (10 sec: 3277.0, 60 sec: 3310.9, 300 sec: 3554.6). Total num frames: 16705536. Throughput: 0: 3321.5. Samples: 16705368. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:55:47,924][261293] Avg episode reward: [(0, '1464.903')]
+[2026-06-06 21:55:48,373][262198] Updated weights for policy 0, policy_version 32631 (0.0010)
+[2026-06-06 21:55:49,885][262198] Updated weights for policy 0, policy_version 32642 (0.0011)
+[2026-06-06 21:55:51,679][262198] Updated weights for policy 0, policy_version 32652 (0.0012)
+[2026-06-06 21:55:52,923][261293] Fps is (10 sec: 3276.1, 60 sec: 3345.0, 300 sec: 3547.5). Total num frames: 16721920. Throughput: 0: 3375.2. Samples: 16725798. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:55:52,924][261293] Avg episode reward: [(0, '1332.616')]
+[2026-06-06 21:55:53,506][262198] Updated weights for policy 0, policy_version 32662 (0.0011)
+[2026-06-06 21:55:55,113][262198] Updated weights for policy 0, policy_version 32675 (0.0012)
+[2026-06-06 21:55:57,171][262198] Updated weights for policy 0, policy_version 32687 (0.0009)
+[2026-06-06 21:55:57,922][261293] Fps is (10 sec: 3276.2, 60 sec: 3345.0, 300 sec: 3540.6). Total num frames: 16738304. Throughput: 0: 3363.6. Samples: 16745312. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 21:55:57,925][261293] Avg episode reward: [(0, '1458.590')]
+[2026-06-06 21:55:59,349][262198] Updated weights for policy 0, policy_version 32697 (0.0009)
+[2026-06-06 21:55:59,359][262115] Signal inference workers to stop experience collection... (3450 times)
+[2026-06-06 21:55:59,382][262198] InferenceWorker_p0-w0: stopping experience collection (3450 times)
+[2026-06-06 21:55:59,450][262115] Signal inference workers to resume experience collection... (3450 times)
+[2026-06-06 21:55:59,451][262198] InferenceWorker_p0-w0: resuming experience collection (3450 times)
+[2026-06-06 21:56:01,555][262198] Updated weights for policy 0, policy_version 32712 (0.0011)
+[2026-06-06 21:56:02,920][261293] Fps is (10 sec: 3277.5, 60 sec: 3345.1, 300 sec: 3533.7). Total num frames: 16754688. Throughput: 0: 3370.3. Samples: 16755388. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 21:56:02,922][261293] Avg episode reward: [(0, '1512.430')]
+[2026-06-06 21:56:03,782][262198] Updated weights for policy 0, policy_version 32726 (0.0012)
+[2026-06-06 21:56:05,330][262198] Updated weights for policy 0, policy_version 32738 (0.0010)
+[2026-06-06 21:56:07,173][262198] Updated weights for policy 0, policy_version 32749 (0.0009)
+[2026-06-06 21:56:07,920][261293] Fps is (10 sec: 3277.4, 60 sec: 3379.2, 300 sec: 3519.8). Total num frames: 16771072. Throughput: 0: 3336.3. Samples: 16774230. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 21:56:07,923][261293] Avg episode reward: [(0, '1742.059')]
+[2026-06-06 21:56:08,587][262198] Updated weights for policy 0, policy_version 32759 (0.0012)
+[2026-06-06 21:56:10,078][262198] Updated weights for policy 0, policy_version 32771 (0.0010)
+[2026-06-06 21:56:12,366][262198] Updated weights for policy 0, policy_version 32785 (0.0009)
+[2026-06-06 21:56:12,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3379.2, 300 sec: 3505.9). Total num frames: 16787456. Throughput: 0: 3335.1. Samples: 16794378. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 21:56:12,923][261293] Avg episode reward: [(0, '1648.879')]
+[2026-06-06 21:56:14,955][262198] Updated weights for policy 0, policy_version 32797 (0.0009)
+[2026-06-06 21:56:16,415][262198] Updated weights for policy 0, policy_version 32807 (0.0008)
+[2026-06-06 21:56:17,474][262198] Updated weights for policy 0, policy_version 32818 (0.0010)
+[2026-06-06 21:56:17,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3379.2, 300 sec: 3492.0). Total num frames: 16803840. Throughput: 0: 3291.1. Samples: 16803142. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 21:56:17,923][261293] Avg episode reward: [(0, '1678.335')]
+[2026-06-06 21:56:19,572][262198] Updated weights for policy 0, policy_version 32829 (0.0011)
+[2026-06-06 21:56:21,323][262198] Updated weights for policy 0, policy_version 32839 (0.0009)
+[2026-06-06 21:56:22,370][262198] Updated weights for policy 0, policy_version 32850 (0.0010)
+[2026-06-06 21:56:22,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3345.1, 300 sec: 3464.3). Total num frames: 16820224. Throughput: 0: 3302.5. Samples: 16824320. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 21:56:22,922][261293] Avg episode reward: [(0, '1671.042')]
+[2026-06-06 21:56:23,954][262198] Updated weights for policy 0, policy_version 32861 (0.0009)
+[2026-06-06 21:56:25,315][262198] Updated weights for policy 0, policy_version 32871 (0.0009)
+[2026-06-06 21:56:26,631][262198] Updated weights for policy 0, policy_version 32883 (0.0009)
+[2026-06-06 21:56:27,925][261293] Fps is (10 sec: 3480.0, 60 sec: 3378.9, 300 sec: 3471.1). Total num frames: 16838656. Throughput: 0: 3374.5. Samples: 16846718. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:56:27,931][261293] Avg episode reward: [(0, '1834.575')]
+[2026-06-06 21:56:29,127][262198] Updated weights for policy 0, policy_version 32897 (0.0012)
+[2026-06-06 21:56:30,904][262198] Updated weights for policy 0, policy_version 32908 (0.0010)
+[2026-06-06 21:56:32,312][262198] Updated weights for policy 0, policy_version 32918 (0.0009)
+[2026-06-06 21:56:32,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3310.9, 300 sec: 3471.2). Total num frames: 16855040. Throughput: 0: 3358.6. Samples: 16856506. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:56:32,925][261293] Avg episode reward: [(0, '1458.085')]
+[2026-06-06 21:56:33,922][262198] Updated weights for policy 0, policy_version 32928 (0.0014)
+[2026-06-06 21:56:36,214][262198] Updated weights for policy 0, policy_version 32938 (0.0011)
+[2026-06-06 21:56:37,921][261293] Fps is (10 sec: 3073.2, 60 sec: 3276.8, 300 sec: 3464.2). Total num frames: 16869376. Throughput: 0: 3331.5. Samples: 16875710. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:56:37,923][261293] Avg episode reward: [(0, '1699.966')]
+[2026-06-06 21:56:37,927][262198] Updated weights for policy 0, policy_version 32949 (0.0011)
+[2026-06-06 21:56:39,512][262198] Updated weights for policy 0, policy_version 32963 (0.0013)
+[2026-06-06 21:56:42,353][262198] Updated weights for policy 0, policy_version 32974 (0.0013)
+[2026-06-06 21:56:42,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3242.7, 300 sec: 3450.4). Total num frames: 16883712. Throughput: 0: 3255.8. Samples: 16891818. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:56:42,923][261293] Avg episode reward: [(0, '1846.896')]
+[2026-06-06 21:56:44,453][262198] Updated weights for policy 0, policy_version 32984 (0.0015)
+[2026-06-06 21:56:46,038][262198] Updated weights for policy 0, policy_version 32994 (0.0012)
+[2026-06-06 21:56:47,920][261293] Fps is (10 sec: 2662.5, 60 sec: 3174.4, 300 sec: 3422.6). Total num frames: 16896000. Throughput: 0: 3198.4. Samples: 16899316. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:56:47,923][261293] Avg episode reward: [(0, '1783.659')]
+[2026-06-06 21:56:48,616][262198] Updated weights for policy 0, policy_version 33005 (0.0011)
+[2026-06-06 21:56:50,595][262198] Updated weights for policy 0, policy_version 33015 (0.0012)
+[2026-06-06 21:56:52,921][261293] Fps is (10 sec: 2662.3, 60 sec: 3140.3, 300 sec: 3408.7). Total num frames: 16910336. Throughput: 0: 3125.6. Samples: 16914886. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:56:52,924][261293] Avg episode reward: [(0, '1750.035')]
+[2026-06-06 21:56:53,498][262198] Updated weights for policy 0, policy_version 33029 (0.0013)
+[2026-06-06 21:56:54,781][262198] Updated weights for policy 0, policy_version 33039 (0.0012)
+[2026-06-06 21:56:57,922][261293] Fps is (10 sec: 2457.5, 60 sec: 3037.9, 300 sec: 3374.0). Total num frames: 16920576. Throughput: 0: 3028.7. Samples: 16930672. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:56:57,926][261293] Avg episode reward: [(0, '1547.299')]
+[2026-06-06 21:56:57,990][262198] Updated weights for policy 0, policy_version 33049 (0.0012)
+[2026-06-06 21:56:58,798][262198] Updated weights for policy 0, policy_version 33060 (0.0013)
+[2026-06-06 21:57:01,550][262198] Updated weights for policy 0, policy_version 33072 (0.0013)
+[2026-06-06 21:57:02,920][261293] Fps is (10 sec: 2457.8, 60 sec: 3003.7, 300 sec: 3367.1). Total num frames: 16934912. Throughput: 0: 2997.8. Samples: 16938044. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 21:57:02,922][261293] Avg episode reward: [(0, '1677.968')]
+[2026-06-06 21:57:04,350][262198] Updated weights for policy 0, policy_version 33084 (0.0012)
+[2026-06-06 21:57:06,231][262198] Updated weights for policy 0, policy_version 33094 (0.0010)
+[2026-06-06 21:57:07,921][261293] Fps is (10 sec: 2867.3, 60 sec: 2969.6, 300 sec: 3374.0). Total num frames: 16949248. Throughput: 0: 2887.7. Samples: 16954266. Policy #0 lag: (min: 1.0, avg: 16.3, max: 33.0)
+[2026-06-06 21:57:07,923][261293] Avg episode reward: [(0, '1577.566')]
+[2026-06-06 21:57:08,194][262198] Updated weights for policy 0, policy_version 33106 (0.0013)
+[2026-06-06 21:57:10,552][262198] Updated weights for policy 0, policy_version 33116 (0.0011)
+[2026-06-06 21:57:12,196][262198] Updated weights for policy 0, policy_version 33126 (0.0010)
+[2026-06-06 21:57:12,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2901.3, 300 sec: 3360.1). Total num frames: 16961536. Throughput: 0: 2759.8. Samples: 16970898. Policy #0 lag: (min: 1.0, avg: 16.3, max: 33.0)
+[2026-06-06 21:57:12,923][261293] Avg episode reward: [(0, '1655.771')]
+[2026-06-06 21:57:13,724][262198] Updated weights for policy 0, policy_version 33137 (0.0009)
+[2026-06-06 21:57:15,566][262198] Updated weights for policy 0, policy_version 33149 (0.0011)
+[2026-06-06 21:57:17,107][262198] Updated weights for policy 0, policy_version 33160 (0.0010)
+[2026-06-06 21:57:17,921][261293] Fps is (10 sec: 2867.1, 60 sec: 2901.3, 300 sec: 3360.1). Total num frames: 16977920. Throughput: 0: 2776.7. Samples: 16981460. Policy #0 lag: (min: 1.0, avg: 16.3, max: 33.0)
+[2026-06-06 21:57:17,924][261293] Avg episode reward: [(0, '1716.277')]
+[2026-06-06 21:57:18,969][262198] Updated weights for policy 0, policy_version 33172 (0.0011)
+[2026-06-06 21:57:20,885][262115] Signal inference workers to stop experience collection... (3500 times)
+[2026-06-06 21:57:20,914][262198] InferenceWorker_p0-w0: stopping experience collection (3500 times)
+[2026-06-06 21:57:20,964][262115] Signal inference workers to resume experience collection... (3500 times)
+[2026-06-06 21:57:20,964][262198] InferenceWorker_p0-w0: resuming experience collection (3500 times)
+[2026-06-06 21:57:20,966][262198] Updated weights for policy 0, policy_version 33184 (0.0008)
+[2026-06-06 21:57:22,822][262198] Updated weights for policy 0, policy_version 33194 (0.0009)
+[2026-06-06 21:57:22,920][261293] Fps is (10 sec: 3481.6, 60 sec: 2935.5, 300 sec: 3353.2). Total num frames: 16996352. Throughput: 0: 2774.8. Samples: 17000574. Policy #0 lag: (min: 1.0, avg: 16.3, max: 33.0)
+[2026-06-06 21:57:22,921][261293] Avg episode reward: [(0, '1610.666')]
+[2026-06-06 21:57:24,159][262198] Updated weights for policy 0, policy_version 33205 (0.0008)
+[2026-06-06 21:57:25,447][262198] Updated weights for policy 0, policy_version 33216 (0.0009)
+[2026-06-06 21:57:26,966][262198] Updated weights for policy 0, policy_version 33226 (0.0008)
+[2026-06-06 21:57:27,921][261293] Fps is (10 sec: 3891.5, 60 sec: 2969.8, 300 sec: 3360.1). Total num frames: 17016832. Throughput: 0: 2924.0. Samples: 17023396. Policy #0 lag: (min: 1.0, avg: 16.3, max: 33.0)
+[2026-06-06 21:57:27,923][261293] Avg episode reward: [(0, '1530.706')]
+[2026-06-06 21:57:28,399][262198] Updated weights for policy 0, policy_version 33237 (0.0008)
+[2026-06-06 21:57:29,617][262198] Updated weights for policy 0, policy_version 33248 (0.0009)
+[2026-06-06 21:57:31,001][262198] Updated weights for policy 0, policy_version 33258 (0.0008)
+[2026-06-06 21:57:32,550][262198] Updated weights for policy 0, policy_version 33269 (0.0010)
+[2026-06-06 21:57:32,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3003.7, 300 sec: 3367.1). Total num frames: 17035264. Throughput: 0: 3032.9. Samples: 17035796. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:57:32,924][261293] Avg episode reward: [(0, '1734.934')]
+[2026-06-06 21:57:33,502][262198] Updated weights for policy 0, policy_version 33279 (0.0010)
+[2026-06-06 21:57:35,387][262198] Updated weights for policy 0, policy_version 33289 (0.0009)
+[2026-06-06 21:57:36,196][262198] Updated weights for policy 0, policy_version 33299 (0.0008)
+[2026-06-06 21:57:37,843][262198] Updated weights for policy 0, policy_version 33309 (0.0009)
+[2026-06-06 21:57:37,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3072.0, 300 sec: 3367.1). Total num frames: 17053696. Throughput: 0: 3183.3. Samples: 17058132. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:57:37,923][261293] Avg episode reward: [(0, '1648.755')]
+[2026-06-06 21:57:39,009][262198] Updated weights for policy 0, policy_version 33319 (0.0009)
+[2026-06-06 21:57:40,531][262198] Updated weights for policy 0, policy_version 33329 (0.0010)
+[2026-06-06 21:57:42,106][262198] Updated weights for policy 0, policy_version 33339 (0.0010)
+[2026-06-06 21:57:42,920][261293] Fps is (10 sec: 3686.6, 60 sec: 3140.3, 300 sec: 3360.1). Total num frames: 17072128. Throughput: 0: 3316.3. Samples: 17079904. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:57:42,922][261293] Avg episode reward: [(0, '1722.232')]
+[2026-06-06 21:57:43,502][262198] Updated weights for policy 0, policy_version 33349 (0.0010)
+[2026-06-06 21:57:45,251][262198] Updated weights for policy 0, policy_version 33362 (0.0009)
+[2026-06-06 21:57:47,022][262198] Updated weights for policy 0, policy_version 33373 (0.0010)
+[2026-06-06 21:57:47,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3208.5, 300 sec: 3353.2). Total num frames: 17088512. Throughput: 0: 3391.7. Samples: 17090672. Policy #0 lag: (min: 6.0, avg: 20.4, max: 38.0)
+[2026-06-06 21:57:47,921][261293] Avg episode reward: [(0, '1575.821')]
+[2026-06-06 21:57:48,228][262198] Updated weights for policy 0, policy_version 33383 (0.0008)
+[2026-06-06 21:57:49,807][262198] Updated weights for policy 0, policy_version 33394 (0.0012)
+[2026-06-06 21:57:51,569][262198] Updated weights for policy 0, policy_version 33404 (0.0010)
+[2026-06-06 21:57:52,889][262198] Updated weights for policy 0, policy_version 33414 (0.0010)
+[2026-06-06 21:57:52,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3276.8, 300 sec: 3360.1). Total num frames: 17106944. Throughput: 0: 3498.9. Samples: 17111716. Policy #0 lag: (min: 0.0, avg: 20.4, max: 40.0)
+[2026-06-06 21:57:52,923][261293] Avg episode reward: [(0, '1526.968')]
+[2026-06-06 21:57:54,620][262198] Updated weights for policy 0, policy_version 33425 (0.0010)
+[2026-06-06 21:57:56,037][262198] Updated weights for policy 0, policy_version 33435 (0.0008)
+[2026-06-06 21:57:57,666][262198] Updated weights for policy 0, policy_version 33445 (0.0008)
+[2026-06-06 21:57:57,920][261293] Fps is (10 sec: 3686.3, 60 sec: 3413.4, 300 sec: 3360.1). Total num frames: 17125376. Throughput: 0: 3587.9. Samples: 17132356. Policy #0 lag: (min: 0.0, avg: 20.4, max: 40.0)
+[2026-06-06 21:57:57,922][261293] Avg episode reward: [(0, '1848.416')]
+[2026-06-06 21:57:59,422][262198] Updated weights for policy 0, policy_version 33457 (0.0010)
+[2026-06-06 21:58:01,040][262198] Updated weights for policy 0, policy_version 33468 (0.0008)
+[2026-06-06 21:58:02,057][262198] Updated weights for policy 0, policy_version 33478 (0.0008)
+[2026-06-06 21:58:02,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3447.4, 300 sec: 3339.3). Total num frames: 17141760. Throughput: 0: 3583.1. Samples: 17142700. Policy #0 lag: (min: 0.0, avg: 20.4, max: 40.0)
+[2026-06-06 21:58:02,926][261293] Avg episode reward: [(0, '1701.689')]
+[2026-06-06 21:58:03,693][262198] Updated weights for policy 0, policy_version 33489 (0.0008)
+[2026-06-06 21:58:05,402][262198] Updated weights for policy 0, policy_version 33501 (0.0010)
+[2026-06-06 21:58:06,504][262198] Updated weights for policy 0, policy_version 33511 (0.0010)
+[2026-06-06 21:58:07,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3549.9, 300 sec: 3346.2). Total num frames: 17162240. Throughput: 0: 3679.1. Samples: 17166134. Policy #0 lag: (min: 0.0, avg: 20.4, max: 40.0)
+[2026-06-06 21:58:07,923][261293] Avg episode reward: [(0, '1739.547')]
+[2026-06-06 21:58:08,221][262198] Updated weights for policy 0, policy_version 33523 (0.0008)
+[2026-06-06 21:58:09,911][262198] Updated weights for policy 0, policy_version 33533 (0.0009)
+[2026-06-06 21:58:10,967][262198] Updated weights for policy 0, policy_version 33544 (0.0010)
+[2026-06-06 21:58:12,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3618.1, 300 sec: 3339.3). Total num frames: 17178624. Throughput: 0: 3639.6. Samples: 17187180. Policy #0 lag: (min: 0.0, avg: 20.4, max: 40.0)
+[2026-06-06 21:58:12,922][261293] Avg episode reward: [(0, '1624.350')]
+[2026-06-06 21:58:12,941][262198] Updated weights for policy 0, policy_version 33555 (0.0011)
+[2026-06-06 21:58:14,948][262198] Updated weights for policy 0, policy_version 33566 (0.0012)
+[2026-06-06 21:58:16,189][262198] Updated weights for policy 0, policy_version 33576 (0.0012)
+[2026-06-06 21:58:17,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3618.2, 300 sec: 3325.4). Total num frames: 17195008. Throughput: 0: 3572.1. Samples: 17196542. Policy #0 lag: (min: 3.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:58:17,923][261293] Avg episode reward: [(0, '1533.602')]
+[2026-06-06 21:58:17,950][262198] Updated weights for policy 0, policy_version 33587 (0.0010)
+[2026-06-06 21:58:19,777][262198] Updated weights for policy 0, policy_version 33597 (0.0010)
+[2026-06-06 21:58:21,149][262198] Updated weights for policy 0, policy_version 33607 (0.0011)
+[2026-06-06 21:58:22,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3584.0, 300 sec: 3318.5). Total num frames: 17211392. Throughput: 0: 3510.8. Samples: 17216118. Policy #0 lag: (min: 3.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:58:22,923][261293] Avg episode reward: [(0, '1645.993')]
+[2026-06-06 21:58:23,087][262198] Updated weights for policy 0, policy_version 33619 (0.0011)
+[2026-06-06 21:58:25,091][262198] Updated weights for policy 0, policy_version 33629 (0.0010)
+[2026-06-06 21:58:26,344][262198] Updated weights for policy 0, policy_version 33639 (0.0009)
+[2026-06-06 21:58:27,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3515.7, 300 sec: 3311.5). Total num frames: 17227776. Throughput: 0: 3454.5. Samples: 17235360. Policy #0 lag: (min: 3.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:58:27,922][261293] Avg episode reward: [(0, '1735.985')]
+[2026-06-06 21:58:28,067][262198] Updated weights for policy 0, policy_version 33650 (0.0012)
+[2026-06-06 21:58:29,711][262198] Updated weights for policy 0, policy_version 33660 (0.0012)
+[2026-06-06 21:58:31,191][262198] Updated weights for policy 0, policy_version 33670 (0.0010)
+[2026-06-06 21:58:32,406][262198] Updated weights for policy 0, policy_version 33680 (0.0009)
+[2026-06-06 21:58:32,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3515.7, 300 sec: 3325.4). Total num frames: 17246208. Throughput: 0: 3435.6. Samples: 17245276. Policy #0 lag: (min: 3.0, avg: 22.1, max: 43.0)
+[2026-06-06 21:58:32,924][261293] Avg episode reward: [(0, '1711.718')]
+[2026-06-06 21:58:34,054][262115] Signal inference workers to stop experience collection... (3550 times)
+[2026-06-06 21:58:34,055][262115] Signal inference workers to resume experience collection... (3550 times)
+[2026-06-06 21:58:34,102][262198] InferenceWorker_p0-w0: stopping experience collection (3550 times)
+[2026-06-06 21:58:34,103][262198] InferenceWorker_p0-w0: resuming experience collection (3550 times)
+[2026-06-06 21:58:34,201][262198] Updated weights for policy 0, policy_version 33690 (0.0011)
+[2026-06-06 21:58:36,050][262198] Updated weights for policy 0, policy_version 33701 (0.0009)
+[2026-06-06 21:58:37,403][262198] Updated weights for policy 0, policy_version 33712 (0.0015)
+[2026-06-06 21:58:37,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3447.5, 300 sec: 3318.5). Total num frames: 17260544. Throughput: 0: 3420.2. Samples: 17265626. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:58:37,921][261293] Avg episode reward: [(0, '1745.961')]
+[2026-06-06 21:58:39,504][262198] Updated weights for policy 0, policy_version 33722 (0.0011)
+[2026-06-06 21:58:41,328][262198] Updated weights for policy 0, policy_version 33734 (0.0012)
+[2026-06-06 21:58:42,921][261293] Fps is (10 sec: 3072.1, 60 sec: 3413.3, 300 sec: 3325.4). Total num frames: 17276928. Throughput: 0: 3369.1. Samples: 17283966. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:58:42,926][261293] Avg episode reward: [(0, '1814.967')]
+[2026-06-06 21:58:43,394][262198] Updated weights for policy 0, policy_version 33745 (0.0011)
+[2026-06-06 21:58:45,092][262198] Updated weights for policy 0, policy_version 33755 (0.0011)
+[2026-06-06 21:58:46,953][262198] Updated weights for policy 0, policy_version 33765 (0.0011)
+[2026-06-06 21:58:47,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3345.0, 300 sec: 3311.5). Total num frames: 17289216. Throughput: 0: 3325.9. Samples: 17292364. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:58:47,925][261293] Avg episode reward: [(0, '1658.824')]
+[2026-06-06 21:58:48,401][262198] Updated weights for policy 0, policy_version 33775 (0.0011)
+[2026-06-06 21:58:50,544][262198] Updated weights for policy 0, policy_version 33785 (0.0010)
+[2026-06-06 21:58:52,111][262198] Updated weights for policy 0, policy_version 33795 (0.0010)
+[2026-06-06 21:58:52,921][261293] Fps is (10 sec: 2867.0, 60 sec: 3310.9, 300 sec: 3311.6). Total num frames: 17305600. Throughput: 0: 3201.6. Samples: 17310210. Policy #0 lag: (min: 4.0, avg: 20.8, max: 39.0)
+[2026-06-06 21:58:52,924][261293] Avg episode reward: [(0, '1700.277')]
+[2026-06-06 21:58:53,719][262198] Updated weights for policy 0, policy_version 33805 (0.0010)
+[2026-06-06 21:58:55,837][262198] Updated weights for policy 0, policy_version 33815 (0.0012)
+[2026-06-06 21:58:57,643][262198] Updated weights for policy 0, policy_version 33825 (0.0010)
+[2026-06-06 21:58:57,925][261293] Fps is (10 sec: 3070.7, 60 sec: 3242.4, 300 sec: 3304.5). Total num frames: 17319936. Throughput: 0: 3125.1. Samples: 17327822. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:58:57,928][261293] Avg episode reward: [(0, '1793.285')]
+[2026-06-06 21:58:59,046][262198] Updated weights for policy 0, policy_version 33835 (0.0011)
+[2026-06-06 21:59:01,254][262198] Updated weights for policy 0, policy_version 33845 (0.0009)
+[2026-06-06 21:59:02,174][262198] Updated weights for policy 0, policy_version 33856 (0.0011)
+[2026-06-06 21:59:02,922][261293] Fps is (10 sec: 2867.3, 60 sec: 3208.5, 300 sec: 3297.6). Total num frames: 17334272. Throughput: 0: 3113.5. Samples: 17336648. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:59:02,926][261293] Avg episode reward: [(0, '1851.160')]
+[2026-06-06 21:59:04,411][262198] Updated weights for policy 0, policy_version 33866 (0.0008)
+[2026-06-06 21:59:05,455][262198] Updated weights for policy 0, policy_version 33876 (0.0011)
+[2026-06-06 21:59:07,661][262198] Updated weights for policy 0, policy_version 33886 (0.0009)
+[2026-06-06 21:59:07,921][261293] Fps is (10 sec: 3073.4, 60 sec: 3140.2, 300 sec: 3290.7). Total num frames: 17350656. Throughput: 0: 3084.0. Samples: 17354898. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:59:07,922][261293] Avg episode reward: [(0, '1697.671')]
+[2026-06-06 21:59:10,316][262198] Updated weights for policy 0, policy_version 33898 (0.0011)
+[2026-06-06 21:59:12,424][262198] Updated weights for policy 0, policy_version 33909 (0.0012)
+[2026-06-06 21:59:12,920][261293] Fps is (10 sec: 2867.3, 60 sec: 3072.0, 300 sec: 3276.8). Total num frames: 17362944. Throughput: 0: 3042.2. Samples: 17372258. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:59:12,923][261293] Avg episode reward: [(0, '1885.861')]
+[2026-06-06 21:59:13,376][262198] Updated weights for policy 0, policy_version 33919 (0.0010)
+[2026-06-06 21:59:15,567][262198] Updated weights for policy 0, policy_version 33929 (0.0010)
+[2026-06-06 21:59:16,536][262198] Updated weights for policy 0, policy_version 33939 (0.0010)
+[2026-06-06 21:59:17,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3072.0, 300 sec: 3262.9). Total num frames: 17379328. Throughput: 0: 3025.1. Samples: 17381404. Policy #0 lag: (min: 3.0, avg: 20.1, max: 35.0)
+[2026-06-06 21:59:17,923][261293] Avg episode reward: [(0, '1637.015')]
+[2026-06-06 21:59:17,930][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000033944_17379328.pth...
+[2026-06-06 21:59:18,136][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000009840_5038080.pth
+[2026-06-06 21:59:18,626][262198] Updated weights for policy 0, policy_version 33949 (0.0010)
+[2026-06-06 21:59:19,970][262198] Updated weights for policy 0, policy_version 33960 (0.0016)
+[2026-06-06 21:59:21,970][262198] Updated weights for policy 0, policy_version 33970 (0.0020)
+[2026-06-06 21:59:22,920][261293] Fps is (10 sec: 3072.0, 60 sec: 3037.9, 300 sec: 3249.0). Total num frames: 17393664. Throughput: 0: 2972.9. Samples: 17399406. Policy #0 lag: (min: 4.0, avg: 18.6, max: 36.0)
+[2026-06-06 21:59:22,922][261293] Avg episode reward: [(0, '1811.721')]
+[2026-06-06 21:59:23,980][262198] Updated weights for policy 0, policy_version 33980 (0.0011)
+[2026-06-06 21:59:25,452][262198] Updated weights for policy 0, policy_version 33992 (0.0009)
+[2026-06-06 21:59:26,974][262198] Updated weights for policy 0, policy_version 34003 (0.0008)
+[2026-06-06 21:59:27,921][261293] Fps is (10 sec: 3072.1, 60 sec: 3037.9, 300 sec: 3249.0). Total num frames: 17410048. Throughput: 0: 3012.9. Samples: 17419548. Policy #0 lag: (min: 4.0, avg: 18.6, max: 36.0)
+[2026-06-06 21:59:27,925][261293] Avg episode reward: [(0, '1778.294')]
+[2026-06-06 21:59:28,654][262198] Updated weights for policy 0, policy_version 34013 (0.0009)
+[2026-06-06 21:59:30,749][262198] Updated weights for policy 0, policy_version 34025 (0.0009)
+[2026-06-06 21:59:32,383][262198] Updated weights for policy 0, policy_version 34037 (0.0010)
+[2026-06-06 21:59:32,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3037.8, 300 sec: 3249.0). Total num frames: 17428480. Throughput: 0: 3067.7. Samples: 17430410. Policy #0 lag: (min: 4.0, avg: 18.6, max: 36.0)
+[2026-06-06 21:59:32,930][261293] Avg episode reward: [(0, '1843.389')]
+[2026-06-06 21:59:33,313][262198] Updated weights for policy 0, policy_version 34048 (0.0008)
+[2026-06-06 21:59:35,223][262198] Updated weights for policy 0, policy_version 34060 (0.0010)
+[2026-06-06 21:59:36,668][262198] Updated weights for policy 0, policy_version 34070 (0.0009)
+[2026-06-06 21:59:37,624][262198] Updated weights for policy 0, policy_version 34080 (0.0008)
+[2026-06-06 21:59:37,920][261293] Fps is (10 sec: 3891.1, 60 sec: 3140.3, 300 sec: 3256.0). Total num frames: 17448960. Throughput: 0: 3165.4. Samples: 17452650. Policy #0 lag: (min: 4.0, avg: 18.6, max: 36.0)
+[2026-06-06 21:59:37,923][261293] Avg episode reward: [(0, '1529.049')]
+[2026-06-06 21:59:39,531][262198] Updated weights for policy 0, policy_version 34092 (0.0010)
+[2026-06-06 21:59:41,162][262198] Updated weights for policy 0, policy_version 34103 (0.0009)
+[2026-06-06 21:59:42,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3140.2, 300 sec: 3249.0). Total num frames: 17465344. Throughput: 0: 3250.8. Samples: 17474092. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:59:42,923][261293] Avg episode reward: [(0, '1465.996')]
+[2026-06-06 21:59:43,253][262198] Updated weights for policy 0, policy_version 34113 (0.0011)
+[2026-06-06 21:59:44,404][262198] Updated weights for policy 0, policy_version 34125 (0.0010)
+[2026-06-06 21:59:45,967][262198] Updated weights for policy 0, policy_version 34135 (0.0009)
+[2026-06-06 21:59:47,598][262198] Updated weights for policy 0, policy_version 34145 (0.0008)
+[2026-06-06 21:59:47,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3242.7, 300 sec: 3262.9). Total num frames: 17483776. Throughput: 0: 3283.9. Samples: 17484422. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:59:47,925][261293] Avg episode reward: [(0, '1816.981')]
+[2026-06-06 21:59:48,937][262198] Updated weights for policy 0, policy_version 34157 (0.0009)
+[2026-06-06 21:59:50,439][262198] Updated weights for policy 0, policy_version 34167 (0.0010)
+[2026-06-06 21:59:51,914][262198] Updated weights for policy 0, policy_version 34177 (0.0008)
+[2026-06-06 21:59:52,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3276.8, 300 sec: 3269.9). Total num frames: 17502208. Throughput: 0: 3385.2. Samples: 17507234. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:59:52,924][261293] Avg episode reward: [(0, '1692.489')]
+[2026-06-06 21:59:53,093][262198] Updated weights for policy 0, policy_version 34189 (0.0009)
+[2026-06-06 21:59:53,319][262115] Signal inference workers to stop experience collection... (3600 times)
+[2026-06-06 21:59:53,320][262115] Signal inference workers to resume experience collection... (3600 times)
+[2026-06-06 21:59:53,342][262198] InferenceWorker_p0-w0: stopping experience collection (3600 times)
+[2026-06-06 21:59:53,342][262198] InferenceWorker_p0-w0: resuming experience collection (3600 times)
+[2026-06-06 21:59:54,487][262198] Updated weights for policy 0, policy_version 34199 (0.0009)
+[2026-06-06 21:59:56,285][262198] Updated weights for policy 0, policy_version 34209 (0.0013)
+[2026-06-06 21:59:57,466][262198] Updated weights for policy 0, policy_version 34220 (0.0011)
+[2026-06-06 21:59:57,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3379.4, 300 sec: 3283.8). Total num frames: 17522688. Throughput: 0: 3496.8. Samples: 17529614. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 21:59:57,923][261293] Avg episode reward: [(0, '1880.997')]
+[2026-06-06 21:59:59,204][262198] Updated weights for policy 0, policy_version 34230 (0.0008)
+[2026-06-06 22:00:00,361][262198] Updated weights for policy 0, policy_version 34240 (0.0009)
+[2026-06-06 22:00:02,082][262198] Updated weights for policy 0, policy_version 34251 (0.0010)
+[2026-06-06 22:00:02,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3447.5, 300 sec: 3297.6). Total num frames: 17541120. Throughput: 0: 3525.1. Samples: 17540034. Policy #0 lag: (min: 3.0, avg: 18.7, max: 35.0)
+[2026-06-06 22:00:02,923][261293] Avg episode reward: [(0, '1510.304')]
+[2026-06-06 22:00:03,662][262198] Updated weights for policy 0, policy_version 34261 (0.0009)
+[2026-06-06 22:00:04,895][262198] Updated weights for policy 0, policy_version 34271 (0.0009)
+[2026-06-06 22:00:06,541][262198] Updated weights for policy 0, policy_version 34283 (0.0009)
+[2026-06-06 22:00:07,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3447.5, 300 sec: 3297.6). Total num frames: 17557504. Throughput: 0: 3612.6. Samples: 17561976. Policy #0 lag: (min: 3.0, avg: 18.9, max: 35.0)
+[2026-06-06 22:00:07,924][261293] Avg episode reward: [(0, '1579.856')]
+[2026-06-06 22:00:08,053][262198] Updated weights for policy 0, policy_version 34293 (0.0009)
+[2026-06-06 22:00:09,004][262198] Updated weights for policy 0, policy_version 34303 (0.0008)
+[2026-06-06 22:00:10,390][262198] Updated weights for policy 0, policy_version 34313 (0.0008)
+[2026-06-06 22:00:11,453][262198] Updated weights for policy 0, policy_version 34323 (0.0008)
+[2026-06-06 22:00:12,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3549.8, 300 sec: 3304.6). Total num frames: 17575936. Throughput: 0: 3688.3. Samples: 17585526. Policy #0 lag: (min: 3.0, avg: 18.9, max: 35.0)
+[2026-06-06 22:00:12,924][261293] Avg episode reward: [(0, '1577.126')]
+[2026-06-06 22:00:13,112][262198] Updated weights for policy 0, policy_version 34333 (0.0009)
+[2026-06-06 22:00:14,317][262198] Updated weights for policy 0, policy_version 34343 (0.0008)
+[2026-06-06 22:00:15,869][262198] Updated weights for policy 0, policy_version 34353 (0.0008)
+[2026-06-06 22:00:17,283][262198] Updated weights for policy 0, policy_version 34363 (0.0009)
+[2026-06-06 22:00:17,921][261293] Fps is (10 sec: 3891.4, 60 sec: 3618.1, 300 sec: 3311.5). Total num frames: 17596416. Throughput: 0: 3691.0. Samples: 17596504. Policy #0 lag: (min: 3.0, avg: 18.9, max: 35.0)
+[2026-06-06 22:00:17,922][261293] Avg episode reward: [(0, '1678.759')]
+[2026-06-06 22:00:18,646][262198] Updated weights for policy 0, policy_version 34373 (0.0010)
+[2026-06-06 22:00:19,567][262198] Updated weights for policy 0, policy_version 34383 (0.0007)
+[2026-06-06 22:00:21,332][262198] Updated weights for policy 0, policy_version 34393 (0.0008)
+[2026-06-06 22:00:22,455][262198] Updated weights for policy 0, policy_version 34403 (0.0009)
+[2026-06-06 22:00:22,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3686.4, 300 sec: 3318.5). Total num frames: 17614848. Throughput: 0: 3712.2. Samples: 17619698. Policy #0 lag: (min: 3.0, avg: 18.9, max: 35.0)
+[2026-06-06 22:00:22,925][261293] Avg episode reward: [(0, '1613.839')]
+[2026-06-06 22:00:23,629][262198] Updated weights for policy 0, policy_version 34413 (0.0009)
+[2026-06-06 22:00:25,173][262198] Updated weights for policy 0, policy_version 34423 (0.0008)
+[2026-06-06 22:00:26,685][262198] Updated weights for policy 0, policy_version 34433 (0.0008)
+[2026-06-06 22:00:27,512][262198] Updated weights for policy 0, policy_version 34443 (0.0010)
+[2026-06-06 22:00:27,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3754.6, 300 sec: 3318.4). Total num frames: 17635328. Throughput: 0: 3746.7. Samples: 17642694. Policy #0 lag: (min: 4.0, avg: 19.4, max: 36.0)
+[2026-06-06 22:00:27,923][261293] Avg episode reward: [(0, '1516.970')]
+[2026-06-06 22:00:29,288][262198] Updated weights for policy 0, policy_version 34453 (0.0009)
+[2026-06-06 22:00:30,550][262198] Updated weights for policy 0, policy_version 34463 (0.0008)
+[2026-06-06 22:00:31,840][262198] Updated weights for policy 0, policy_version 34473 (0.0009)
+[2026-06-06 22:00:32,869][262198] Updated weights for policy 0, policy_version 34483 (0.0008)
+[2026-06-06 22:00:32,921][261293] Fps is (10 sec: 4096.0, 60 sec: 3788.8, 300 sec: 3332.3). Total num frames: 17655808. Throughput: 0: 3772.2. Samples: 17654170. Policy #0 lag: (min: 4.0, avg: 19.4, max: 36.0)
+[2026-06-06 22:00:32,927][261293] Avg episode reward: [(0, '1765.363')]
+[2026-06-06 22:00:34,771][262198] Updated weights for policy 0, policy_version 34493 (0.0010)
+[2026-06-06 22:00:35,871][262198] Updated weights for policy 0, policy_version 34503 (0.0009)
+[2026-06-06 22:00:37,340][262198] Updated weights for policy 0, policy_version 34513 (0.0008)
+[2026-06-06 22:00:37,920][261293] Fps is (10 sec: 3686.6, 60 sec: 3720.5, 300 sec: 3332.3). Total num frames: 17672192. Throughput: 0: 3758.6. Samples: 17676368. Policy #0 lag: (min: 4.0, avg: 19.4, max: 36.0)
+[2026-06-06 22:00:37,922][261293] Avg episode reward: [(0, '1693.671')]
+[2026-06-06 22:00:39,076][262198] Updated weights for policy 0, policy_version 34523 (0.0010)
+[2026-06-06 22:00:40,350][262198] Updated weights for policy 0, policy_version 34533 (0.0009)
+[2026-06-06 22:00:41,227][262198] Updated weights for policy 0, policy_version 34543 (0.0009)
+[2026-06-06 22:00:42,921][261293] Fps is (10 sec: 3481.7, 60 sec: 3754.7, 300 sec: 3339.3). Total num frames: 17690624. Throughput: 0: 3755.2. Samples: 17698596. Policy #0 lag: (min: 4.0, avg: 19.4, max: 36.0)
+[2026-06-06 22:00:42,923][261293] Avg episode reward: [(0, '1769.654')]
+[2026-06-06 22:00:43,252][262198] Updated weights for policy 0, policy_version 34553 (0.0008)
+[2026-06-06 22:00:44,213][262198] Updated weights for policy 0, policy_version 34563 (0.0010)
+[2026-06-06 22:00:45,601][262198] Updated weights for policy 0, policy_version 34573 (0.0009)
+[2026-06-06 22:00:47,051][262198] Updated weights for policy 0, policy_version 34583 (0.0009)
+[2026-06-06 22:00:47,926][261293] Fps is (10 sec: 3684.6, 60 sec: 3754.4, 300 sec: 3346.2). Total num frames: 17709056. Throughput: 0: 3764.2. Samples: 17709440. Policy #0 lag: (min: 4.0, avg: 19.4, max: 36.0)
+[2026-06-06 22:00:47,931][261293] Avg episode reward: [(0, '1587.337')]
+[2026-06-06 22:00:48,871][262198] Updated weights for policy 0, policy_version 34593 (0.0009)
+[2026-06-06 22:00:50,085][262198] Updated weights for policy 0, policy_version 34603 (0.0010)
+[2026-06-06 22:00:51,490][262198] Updated weights for policy 0, policy_version 34613 (0.0009)
+[2026-06-06 22:00:52,857][262198] Updated weights for policy 0, policy_version 34623 (0.0008)
+[2026-06-06 22:00:52,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3754.7, 300 sec: 3353.2). Total num frames: 17727488. Throughput: 0: 3758.3. Samples: 17731100. Policy #0 lag: (min: 4.0, avg: 18.3, max: 37.0)
+[2026-06-06 22:00:52,924][261293] Avg episode reward: [(0, '1719.842')]
+[2026-06-06 22:00:54,300][262198] Updated weights for policy 0, policy_version 34633 (0.0008)
+[2026-06-06 22:00:55,363][262198] Updated weights for policy 0, policy_version 34644 (0.0008)
+[2026-06-06 22:00:57,149][262198] Updated weights for policy 0, policy_version 34654 (0.0009)
+[2026-06-06 22:00:57,921][261293] Fps is (10 sec: 3688.1, 60 sec: 3720.5, 300 sec: 3360.1). Total num frames: 17745920. Throughput: 0: 3730.1. Samples: 17753378. Policy #0 lag: (min: 4.0, avg: 18.3, max: 37.0)
+[2026-06-06 22:00:57,924][261293] Avg episode reward: [(0, '1525.866')]
+[2026-06-06 22:00:58,656][262198] Updated weights for policy 0, policy_version 34664 (0.0010)
+[2026-06-06 22:00:59,948][262198] Updated weights for policy 0, policy_version 34676 (0.0010)
+[2026-06-06 22:01:01,708][262198] Updated weights for policy 0, policy_version 34686 (0.0009)
+[2026-06-06 22:01:02,920][261293] Fps is (10 sec: 3481.8, 60 sec: 3686.4, 300 sec: 3360.1). Total num frames: 17762304. Throughput: 0: 3713.1. Samples: 17763590. Policy #0 lag: (min: 4.0, avg: 18.3, max: 37.0)
+[2026-06-06 22:01:02,921][261293] Avg episode reward: [(0, '1622.619')]
+[2026-06-06 22:01:03,181][262198] Updated weights for policy 0, policy_version 34696 (0.0009)
+[2026-06-06 22:01:04,266][262198] Updated weights for policy 0, policy_version 34707 (0.0009)
+[2026-06-06 22:01:06,154][262198] Updated weights for policy 0, policy_version 34717 (0.0010)
+[2026-06-06 22:01:07,289][262198] Updated weights for policy 0, policy_version 34727 (0.0009)
+[2026-06-06 22:01:07,922][261293] Fps is (10 sec: 3481.1, 60 sec: 3720.5, 300 sec: 3367.0). Total num frames: 17780736. Throughput: 0: 3698.8. Samples: 17786150. Policy #0 lag: (min: 4.0, avg: 18.3, max: 37.0)
+[2026-06-06 22:01:07,925][261293] Avg episode reward: [(0, '1604.856')]
+[2026-06-06 22:01:08,894][262198] Updated weights for policy 0, policy_version 34739 (0.0009)
+[2026-06-06 22:01:10,734][262198] Updated weights for policy 0, policy_version 34749 (0.0010)
+[2026-06-06 22:01:11,827][262198] Updated weights for policy 0, policy_version 34759 (0.0009)
+[2026-06-06 22:01:12,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3754.7, 300 sec: 3380.9). Total num frames: 17801216. Throughput: 0: 3667.1. Samples: 17807714. Policy #0 lag: (min: 3.0, avg: 19.6, max: 37.0)
+[2026-06-06 22:01:12,922][261293] Avg episode reward: [(0, '1780.100')]
+[2026-06-06 22:01:13,216][262198] Updated weights for policy 0, policy_version 34769 (0.0011)
+[2026-06-06 22:01:14,895][262198] Updated weights for policy 0, policy_version 34779 (0.0009)
+[2026-06-06 22:01:16,732][262198] Updated weights for policy 0, policy_version 34789 (0.0009)
+[2026-06-06 22:01:17,786][262198] Updated weights for policy 0, policy_version 34801 (0.0009)
+[2026-06-06 22:01:17,921][261293] Fps is (10 sec: 3891.8, 60 sec: 3720.5, 300 sec: 3387.9). Total num frames: 17819648. Throughput: 0: 3653.4. Samples: 17818570. Policy #0 lag: (min: 3.0, avg: 19.6, max: 37.0)
+[2026-06-06 22:01:17,923][261293] Avg episode reward: [(0, '1549.152')]
+[2026-06-06 22:01:19,581][262115] Signal inference workers to stop experience collection... (3650 times)
+[2026-06-06 22:01:19,610][262198] InferenceWorker_p0-w0: stopping experience collection (3650 times)
+[2026-06-06 22:01:19,613][262198] Updated weights for policy 0, policy_version 34811 (0.0009)
+[2026-06-06 22:01:19,629][262115] Signal inference workers to resume experience collection... (3650 times)
+[2026-06-06 22:01:19,644][262198] InferenceWorker_p0-w0: resuming experience collection (3650 times)
+[2026-06-06 22:01:21,130][262198] Updated weights for policy 0, policy_version 34821 (0.0009)
+[2026-06-06 22:01:22,744][262198] Updated weights for policy 0, policy_version 34833 (0.0013)
+[2026-06-06 22:01:22,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3686.4, 300 sec: 3381.0). Total num frames: 17836032. Throughput: 0: 3639.6. Samples: 17840152. Policy #0 lag: (min: 3.0, avg: 19.6, max: 37.0)
+[2026-06-06 22:01:22,924][261293] Avg episode reward: [(0, '1606.465')]
+[2026-06-06 22:01:25,489][262198] Updated weights for policy 0, policy_version 34843 (0.0012)
+[2026-06-06 22:01:27,824][262198] Updated weights for policy 0, policy_version 34854 (0.0012)
+[2026-06-06 22:01:27,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3515.7, 300 sec: 3360.1). Total num frames: 17846272. Throughput: 0: 3462.8. Samples: 17854420. Policy #0 lag: (min: 3.0, avg: 19.6, max: 37.0)
+[2026-06-06 22:01:27,925][261293] Avg episode reward: [(0, '1648.201')]
+[2026-06-06 22:01:28,613][262198] Updated weights for policy 0, policy_version 34864 (0.0011)
+[2026-06-06 22:01:31,400][262198] Updated weights for policy 0, policy_version 34874 (0.0011)
+[2026-06-06 22:01:32,660][262198] Updated weights for policy 0, policy_version 34884 (0.0012)
+[2026-06-06 22:01:32,920][261293] Fps is (10 sec: 2457.7, 60 sec: 3413.4, 300 sec: 3360.1). Total num frames: 17860608. Throughput: 0: 3407.0. Samples: 17862738. Policy #0 lag: (min: 3.0, avg: 19.6, max: 37.0)
+[2026-06-06 22:01:32,923][261293] Avg episode reward: [(0, '1760.457')]
+[2026-06-06 22:01:34,277][262198] Updated weights for policy 0, policy_version 34894 (0.0011)
+[2026-06-06 22:01:36,150][262198] Updated weights for policy 0, policy_version 34904 (0.0010)
+[2026-06-06 22:01:37,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3379.2, 300 sec: 3360.1). Total num frames: 17874944. Throughput: 0: 3310.6. Samples: 17880078. Policy #0 lag: (min: 6.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:01:37,924][261293] Avg episode reward: [(0, '1795.425')]
+[2026-06-06 22:01:38,373][262198] Updated weights for policy 0, policy_version 34914 (0.0010)
+[2026-06-06 22:01:39,607][262198] Updated weights for policy 0, policy_version 34924 (0.0010)
+[2026-06-06 22:01:41,865][262198] Updated weights for policy 0, policy_version 34934 (0.0010)
+[2026-06-06 22:01:42,920][261293] Fps is (10 sec: 2662.3, 60 sec: 3276.8, 300 sec: 3360.1). Total num frames: 17887232. Throughput: 0: 3190.8. Samples: 17896962. Policy #0 lag: (min: 6.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:01:42,922][261293] Avg episode reward: [(0, '1688.511')]
+[2026-06-06 22:01:43,543][262198] Updated weights for policy 0, policy_version 34944 (0.0011)
+[2026-06-06 22:01:45,519][262198] Updated weights for policy 0, policy_version 34954 (0.0011)
+[2026-06-06 22:01:46,903][262198] Updated weights for policy 0, policy_version 34964 (0.0010)
+[2026-06-06 22:01:47,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3242.9, 300 sec: 3367.1). Total num frames: 17903616. Throughput: 0: 3148.6. Samples: 17905278. Policy #0 lag: (min: 6.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:01:47,923][261293] Avg episode reward: [(0, '1688.245')]
+[2026-06-06 22:01:49,385][262198] Updated weights for policy 0, policy_version 34974 (0.0011)
+[2026-06-06 22:01:51,270][262198] Updated weights for policy 0, policy_version 34984 (0.0012)
+[2026-06-06 22:01:52,648][262198] Updated weights for policy 0, policy_version 34994 (0.0010)
+[2026-06-06 22:01:52,920][261293] Fps is (10 sec: 3072.0, 60 sec: 3174.4, 300 sec: 3380.9). Total num frames: 17917952. Throughput: 0: 3018.5. Samples: 17921976. Policy #0 lag: (min: 6.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:01:52,925][261293] Avg episode reward: [(0, '1797.410')]
+[2026-06-06 22:01:54,863][262198] Updated weights for policy 0, policy_version 35004 (0.0012)
+[2026-06-06 22:01:56,949][262198] Updated weights for policy 0, policy_version 35014 (0.0011)
+[2026-06-06 22:01:57,920][261293] Fps is (10 sec: 2867.2, 60 sec: 3106.2, 300 sec: 3380.9). Total num frames: 17932288. Throughput: 0: 2901.6. Samples: 17938284. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:01:57,922][261293] Avg episode reward: [(0, '1781.080')]
+[2026-06-06 22:01:57,934][262198] Updated weights for policy 0, policy_version 35024 (0.0011)
+[2026-06-06 22:02:00,825][262198] Updated weights for policy 0, policy_version 35034 (0.0011)
+[2026-06-06 22:02:02,130][262198] Updated weights for policy 0, policy_version 35044 (0.0012)
+[2026-06-06 22:02:02,920][261293] Fps is (10 sec: 2457.6, 60 sec: 3003.7, 300 sec: 3367.1). Total num frames: 17942528. Throughput: 0: 2843.7. Samples: 17946538. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:02:02,923][261293] Avg episode reward: [(0, '1706.560')]
+[2026-06-06 22:02:03,928][262198] Updated weights for policy 0, policy_version 35056 (0.0010)
+[2026-06-06 22:02:06,322][262198] Updated weights for policy 0, policy_version 35066 (0.0011)
+[2026-06-06 22:02:07,402][262198] Updated weights for policy 0, policy_version 35076 (0.0009)
+[2026-06-06 22:02:07,920][261293] Fps is (10 sec: 2662.3, 60 sec: 2969.7, 300 sec: 3380.9). Total num frames: 17958912. Throughput: 0: 2759.8. Samples: 17964344. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:02:07,922][261293] Avg episode reward: [(0, '1724.860')]
+[2026-06-06 22:02:08,825][262198] Updated weights for policy 0, policy_version 35086 (0.0010)
+[2026-06-06 22:02:10,389][262198] Updated weights for policy 0, policy_version 35096 (0.0010)
+[2026-06-06 22:02:12,178][262198] Updated weights for policy 0, policy_version 35106 (0.0011)
+[2026-06-06 22:02:12,920][261293] Fps is (10 sec: 3276.9, 60 sec: 2901.4, 300 sec: 3380.9). Total num frames: 17975296. Throughput: 0: 2895.6. Samples: 17984722. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:02:12,922][261293] Avg episode reward: [(0, '1639.502')]
+[2026-06-06 22:02:13,502][262198] Updated weights for policy 0, policy_version 35116 (0.0011)
+[2026-06-06 22:02:15,107][262198] Updated weights for policy 0, policy_version 35126 (0.0010)
+[2026-06-06 22:02:16,655][262198] Updated weights for policy 0, policy_version 35136 (0.0010)
+[2026-06-06 22:02:17,921][261293] Fps is (10 sec: 3276.8, 60 sec: 2867.2, 300 sec: 3374.0). Total num frames: 17991680. Throughput: 0: 2930.9. Samples: 17994628. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:02:17,922][261293] Avg episode reward: [(0, '1697.453')]
+[2026-06-06 22:02:18,495][262198] Updated weights for policy 0, policy_version 35146 (0.0010)
+[2026-06-06 22:02:19,508][262198] Updated weights for policy 0, policy_version 35156 (0.0009)
+[2026-06-06 22:02:21,648][262198] Updated weights for policy 0, policy_version 35166 (0.0011)
+[2026-06-06 22:02:22,921][261293] Fps is (10 sec: 3276.6, 60 sec: 2867.2, 300 sec: 3360.1). Total num frames: 18008064. Throughput: 0: 2977.5. Samples: 18014068. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:02:22,923][261293] Avg episode reward: [(0, '1578.959')]
+[2026-06-06 22:02:23,219][262198] Updated weights for policy 0, policy_version 35176 (0.0011)
+[2026-06-06 22:02:24,467][262198] Updated weights for policy 0, policy_version 35186 (0.0011)
+[2026-06-06 22:02:26,302][262198] Updated weights for policy 0, policy_version 35196 (0.0009)
+[2026-06-06 22:02:27,859][262198] Updated weights for policy 0, policy_version 35206 (0.0010)
+[2026-06-06 22:02:27,921][261293] Fps is (10 sec: 3276.8, 60 sec: 2969.6, 300 sec: 3353.2). Total num frames: 18024448. Throughput: 0: 3046.2. Samples: 18034040. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:02:27,923][261293] Avg episode reward: [(0, '1759.204')]
+[2026-06-06 22:02:29,140][262198] Updated weights for policy 0, policy_version 35217 (0.0012)
+[2026-06-06 22:02:30,841][262198] Updated weights for policy 0, policy_version 35227 (0.0010)
+[2026-06-06 22:02:32,344][262198] Updated weights for policy 0, policy_version 35237 (0.0009)
+[2026-06-06 22:02:32,922][261293] Fps is (10 sec: 3481.3, 60 sec: 3037.8, 300 sec: 3353.2). Total num frames: 18042880. Throughput: 0: 3088.3. Samples: 18044254. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:02:32,926][261293] Avg episode reward: [(0, '1839.763')]
+[2026-06-06 22:02:33,648][262198] Updated weights for policy 0, policy_version 35249 (0.0011)
+[2026-06-06 22:02:35,688][262198] Updated weights for policy 0, policy_version 35259 (0.0012)
+[2026-06-06 22:02:37,527][262198] Updated weights for policy 0, policy_version 35269 (0.0011)
+[2026-06-06 22:02:37,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3072.0, 300 sec: 3346.2). Total num frames: 18059264. Throughput: 0: 3182.1. Samples: 18065168. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:02:37,922][261293] Avg episode reward: [(0, '1855.616')]
+[2026-06-06 22:02:38,637][262198] Updated weights for policy 0, policy_version 35282 (0.0010)
+[2026-06-06 22:02:40,160][262198] Updated weights for policy 0, policy_version 35292 (0.0010)
+[2026-06-06 22:02:41,967][262198] Updated weights for policy 0, policy_version 35304 (0.0009)
+[2026-06-06 22:02:42,924][261293] Fps is (10 sec: 3685.6, 60 sec: 3208.3, 300 sec: 3360.1). Total num frames: 18079744. Throughput: 0: 3321.5. Samples: 18087764. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:02:42,929][261293] Avg episode reward: [(0, '1734.326')]
+[2026-06-06 22:02:43,055][262198] Updated weights for policy 0, policy_version 35315 (0.0009)
+[2026-06-06 22:02:44,658][262198] Updated weights for policy 0, policy_version 35325 (0.0009)
+[2026-06-06 22:02:45,919][262198] Updated weights for policy 0, policy_version 35335 (0.0009)
+[2026-06-06 22:02:47,082][262198] Updated weights for policy 0, policy_version 35345 (0.0008)
+[2026-06-06 22:02:47,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3276.8, 300 sec: 3367.0). Total num frames: 18100224. Throughput: 0: 3402.2. Samples: 18099638. Policy #0 lag: (min: 6.0, avg: 22.8, max: 41.0)
+[2026-06-06 22:02:47,923][261293] Avg episode reward: [(0, '1757.646')]
+[2026-06-06 22:02:48,469][262198] Updated weights for policy 0, policy_version 35355 (0.0009)
+[2026-06-06 22:02:49,973][262198] Updated weights for policy 0, policy_version 35365 (0.0008)
+[2026-06-06 22:02:51,151][262198] Updated weights for policy 0, policy_version 35377 (0.0009)
+[2026-06-06 22:02:52,645][262198] Updated weights for policy 0, policy_version 35387 (0.0008)
+[2026-06-06 22:02:52,921][261293] Fps is (10 sec: 3892.5, 60 sec: 3345.1, 300 sec: 3367.0). Total num frames: 18118656. Throughput: 0: 3524.7. Samples: 18122954. Policy #0 lag: (min: 6.0, avg: 22.8, max: 41.0)
+[2026-06-06 22:02:52,926][261293] Avg episode reward: [(0, '1711.050')]
+[2026-06-06 22:02:54,221][262198] Updated weights for policy 0, policy_version 35397 (0.0009)
+[2026-06-06 22:02:55,168][262198] Updated weights for policy 0, policy_version 35407 (0.0008)
+[2026-06-06 22:02:56,789][262198] Updated weights for policy 0, policy_version 35417 (0.0009)
+[2026-06-06 22:02:57,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3413.3, 300 sec: 3374.0). Total num frames: 18137088. Throughput: 0: 3579.8. Samples: 18145814. Policy #0 lag: (min: 6.0, avg: 22.8, max: 41.0)
+[2026-06-06 22:02:57,924][261293] Avg episode reward: [(0, '1724.678')]
+[2026-06-06 22:02:57,939][262198] Updated weights for policy 0, policy_version 35427 (0.0009)
+[2026-06-06 22:02:59,468][262198] Updated weights for policy 0, policy_version 35437 (0.0010)
+[2026-06-06 22:03:00,849][262198] Updated weights for policy 0, policy_version 35447 (0.0013)
+[2026-06-06 22:03:02,722][262198] Updated weights for policy 0, policy_version 35457 (0.0010)
+[2026-06-06 22:03:02,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3549.8, 300 sec: 3367.0). Total num frames: 18155520. Throughput: 0: 3585.3. Samples: 18155968. Policy #0 lag: (min: 6.0, avg: 22.8, max: 41.0)
+[2026-06-06 22:03:02,925][261293] Avg episode reward: [(0, '1736.777')]
+[2026-06-06 22:03:04,071][262198] Updated weights for policy 0, policy_version 35467 (0.0009)
+[2026-06-06 22:03:05,262][262198] Updated weights for policy 0, policy_version 35477 (0.0008)
+[2026-06-06 22:03:06,606][262198] Updated weights for policy 0, policy_version 35487 (0.0009)
+[2026-06-06 22:03:07,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3549.8, 300 sec: 3367.0). Total num frames: 18171904. Throughput: 0: 3627.3. Samples: 18177298. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:03:07,926][261293] Avg episode reward: [(0, '1634.896')]
+[2026-06-06 22:03:08,128][262198] Updated weights for policy 0, policy_version 35497 (0.0010)
+[2026-06-06 22:03:09,135][262198] Updated weights for policy 0, policy_version 35507 (0.0008)
+[2026-06-06 22:03:10,913][262198] Updated weights for policy 0, policy_version 35517 (0.0010)
+[2026-06-06 22:03:12,347][262198] Updated weights for policy 0, policy_version 35527 (0.0008)
+[2026-06-06 22:03:12,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3618.1, 300 sec: 3380.9). Total num frames: 18192384. Throughput: 0: 3687.4. Samples: 18199972. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:03:12,922][261293] Avg episode reward: [(0, '1614.961')]
+[2026-06-06 22:03:13,457][262198] Updated weights for policy 0, policy_version 35537 (0.0008)
+[2026-06-06 22:03:14,886][262198] Updated weights for policy 0, policy_version 35547 (0.0009)
+[2026-06-06 22:03:16,371][262198] Updated weights for policy 0, policy_version 35557 (0.0008)
+[2026-06-06 22:03:17,267][262198] Updated weights for policy 0, policy_version 35567 (0.0011)
+[2026-06-06 22:03:17,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3686.4, 300 sec: 3394.8). Total num frames: 18212864. Throughput: 0: 3713.2. Samples: 18211342. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:03:17,924][261293] Avg episode reward: [(0, '1682.468')]
+[2026-06-06 22:03:19,153][262198] Updated weights for policy 0, policy_version 35577 (0.0010)
+[2026-06-06 22:03:20,214][262198] Updated weights for policy 0, policy_version 35587 (0.0008)
+[2026-06-06 22:03:21,706][262198] Updated weights for policy 0, policy_version 35597 (0.0009)
+[2026-06-06 22:03:22,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3686.4, 300 sec: 3394.8). Total num frames: 18229248. Throughput: 0: 3751.6. Samples: 18233990. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:03:22,923][261293] Avg episode reward: [(0, '1621.778')]
+[2026-06-06 22:03:22,926][262198] Updated weights for policy 0, policy_version 35607 (0.0009)
+[2026-06-06 22:03:24,530][262198] Updated weights for policy 0, policy_version 35617 (0.0008)
+[2026-06-06 22:03:25,723][262198] Updated weights for policy 0, policy_version 35627 (0.0010)
+[2026-06-06 22:03:27,215][262198] Updated weights for policy 0, policy_version 35637 (0.0009)
+[2026-06-06 22:03:27,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3720.5, 300 sec: 3394.8). Total num frames: 18247680. Throughput: 0: 3741.8. Samples: 18256132. Policy #0 lag: (min: 3.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:03:27,925][261293] Avg episode reward: [(0, '1621.721')]
+[2026-06-06 22:03:28,524][262198] Updated weights for policy 0, policy_version 35647 (0.0010)
+[2026-06-06 22:03:30,072][262198] Updated weights for policy 0, policy_version 35657 (0.0009)
+[2026-06-06 22:03:31,179][262198] Updated weights for policy 0, policy_version 35668 (0.0010)
+[2026-06-06 22:03:32,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3720.6, 300 sec: 3408.7). Total num frames: 18266112. Throughput: 0: 3723.1. Samples: 18267178. Policy #0 lag: (min: 1.0, avg: 17.8, max: 36.0)
+[2026-06-06 22:03:32,922][261293] Avg episode reward: [(0, '1403.994')]
+[2026-06-06 22:03:33,147][262198] Updated weights for policy 0, policy_version 35678 (0.0010)
+[2026-06-06 22:03:34,468][262198] Updated weights for policy 0, policy_version 35688 (0.0009)
+[2026-06-06 22:03:34,480][262115] Signal inference workers to stop experience collection... (3700 times)
+[2026-06-06 22:03:34,481][262115] Signal inference workers to resume experience collection... (3700 times)
+[2026-06-06 22:03:34,499][262198] InferenceWorker_p0-w0: stopping experience collection (3700 times)
+[2026-06-06 22:03:34,499][262198] InferenceWorker_p0-w0: resuming experience collection (3700 times)
+[2026-06-06 22:03:35,615][262198] Updated weights for policy 0, policy_version 35699 (0.0008)
+[2026-06-06 22:03:37,424][262198] Updated weights for policy 0, policy_version 35709 (0.0009)
+[2026-06-06 22:03:37,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3754.6, 300 sec: 3415.6). Total num frames: 18284544. Throughput: 0: 3688.8. Samples: 18288952. Policy #0 lag: (min: 1.0, avg: 17.8, max: 36.0)
+[2026-06-06 22:03:37,923][261293] Avg episode reward: [(0, '1627.048')]
+[2026-06-06 22:03:38,632][262198] Updated weights for policy 0, policy_version 35719 (0.0008)
+[2026-06-06 22:03:40,125][262198] Updated weights for policy 0, policy_version 35731 (0.0008)
+[2026-06-06 22:03:41,814][262198] Updated weights for policy 0, policy_version 35741 (0.0009)
+[2026-06-06 22:03:42,922][261293] Fps is (10 sec: 3685.9, 60 sec: 3720.7, 300 sec: 3436.5). Total num frames: 18302976. Throughput: 0: 3673.7. Samples: 18311136. Policy #0 lag: (min: 1.0, avg: 17.8, max: 36.0)
+[2026-06-06 22:03:42,928][261293] Avg episode reward: [(0, '1586.939')]
+[2026-06-06 22:03:43,156][262198] Updated weights for policy 0, policy_version 35751 (0.0009)
+[2026-06-06 22:03:44,630][262198] Updated weights for policy 0, policy_version 35763 (0.0011)
+[2026-06-06 22:03:46,191][262198] Updated weights for policy 0, policy_version 35773 (0.0009)
+[2026-06-06 22:03:47,345][262198] Updated weights for policy 0, policy_version 35783 (0.0008)
+[2026-06-06 22:03:47,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3720.5, 300 sec: 3450.4). Total num frames: 18323456. Throughput: 0: 3690.7. Samples: 18322050. Policy #0 lag: (min: 1.0, avg: 17.8, max: 36.0)
+[2026-06-06 22:03:47,925][261293] Avg episode reward: [(0, '1626.573')]
+[2026-06-06 22:03:48,648][262198] Updated weights for policy 0, policy_version 35793 (0.0008)
+[2026-06-06 22:03:50,002][262198] Updated weights for policy 0, policy_version 35803 (0.0007)
+[2026-06-06 22:03:51,348][262198] Updated weights for policy 0, policy_version 35813 (0.0008)
+[2026-06-06 22:03:52,440][262198] Updated weights for policy 0, policy_version 35823 (0.0009)
+[2026-06-06 22:03:52,921][261293] Fps is (10 sec: 3891.7, 60 sec: 3720.5, 300 sec: 3464.3). Total num frames: 18341888. Throughput: 0: 3758.8. Samples: 18346444. Policy #0 lag: (min: 4.0, avg: 20.6, max: 36.0)
+[2026-06-06 22:03:52,923][261293] Avg episode reward: [(0, '1860.211')]
+[2026-06-06 22:03:54,195][262198] Updated weights for policy 0, policy_version 35833 (0.0007)
+[2026-06-06 22:03:55,157][262198] Updated weights for policy 0, policy_version 35843 (0.0008)
+[2026-06-06 22:03:56,725][262198] Updated weights for policy 0, policy_version 35853 (0.0008)
+[2026-06-06 22:03:57,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3754.7, 300 sec: 3485.1). Total num frames: 18362368. Throughput: 0: 3764.5. Samples: 18369376. Policy #0 lag: (min: 4.0, avg: 20.6, max: 36.0)
+[2026-06-06 22:03:57,923][261293] Avg episode reward: [(0, '1752.039')]
+[2026-06-06 22:03:58,359][262198] Updated weights for policy 0, policy_version 35865 (0.0008)
+[2026-06-06 22:03:59,300][262198] Updated weights for policy 0, policy_version 35875 (0.0008)
+[2026-06-06 22:04:00,714][262198] Updated weights for policy 0, policy_version 35885 (0.0007)
+[2026-06-06 22:04:01,798][262198] Updated weights for policy 0, policy_version 35896 (0.0009)
+[2026-06-06 22:04:02,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3754.7, 300 sec: 3492.0). Total num frames: 18380800. Throughput: 0: 3783.5. Samples: 18381598. Policy #0 lag: (min: 4.0, avg: 20.6, max: 36.0)
+[2026-06-06 22:04:02,922][261293] Avg episode reward: [(0, '1818.268')]
+[2026-06-06 22:04:03,604][262198] Updated weights for policy 0, policy_version 35906 (0.0008)
+[2026-06-06 22:04:04,702][262198] Updated weights for policy 0, policy_version 35916 (0.0008)
+[2026-06-06 22:04:06,196][262198] Updated weights for policy 0, policy_version 35928 (0.0009)
+[2026-06-06 22:04:07,735][262198] Updated weights for policy 0, policy_version 35938 (0.0008)
+[2026-06-06 22:04:07,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3823.0, 300 sec: 3519.8). Total num frames: 18401280. Throughput: 0: 3787.4. Samples: 18404424. Policy #0 lag: (min: 4.0, avg: 20.6, max: 36.0)
+[2026-06-06 22:04:07,925][261293] Avg episode reward: [(0, '1612.140')]
+[2026-06-06 22:04:08,937][262198] Updated weights for policy 0, policy_version 35948 (0.0009)
+[2026-06-06 22:04:10,384][262198] Updated weights for policy 0, policy_version 35959 (0.0010)
+[2026-06-06 22:04:12,366][262198] Updated weights for policy 0, policy_version 35969 (0.0010)
+[2026-06-06 22:04:12,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3788.8, 300 sec: 3526.7). Total num frames: 18419712. Throughput: 0: 3785.1. Samples: 18426462. Policy #0 lag: (min: 4.0, avg: 20.6, max: 36.0)
+[2026-06-06 22:04:12,922][261293] Avg episode reward: [(0, '1699.929')]
+[2026-06-06 22:04:13,361][262198] Updated weights for policy 0, policy_version 35979 (0.0008)
+[2026-06-06 22:04:14,995][262198] Updated weights for policy 0, policy_version 35990 (0.0010)
+[2026-06-06 22:04:16,210][262198] Updated weights for policy 0, policy_version 36000 (0.0011)
+[2026-06-06 22:04:17,887][262198] Updated weights for policy 0, policy_version 36010 (0.0009)
+[2026-06-06 22:04:17,921][261293] Fps is (10 sec: 3481.7, 60 sec: 3720.5, 300 sec: 3533.7). Total num frames: 18436096. Throughput: 0: 3792.4. Samples: 18437836. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 22:04:17,924][261293] Avg episode reward: [(0, '1528.593')]
+[2026-06-06 22:04:19,307][262198] Updated weights for policy 0, policy_version 36022 (0.0011)
+[2026-06-06 22:04:20,625][262198] Updated weights for policy 0, policy_version 36032 (0.0009)
+[2026-06-06 22:04:21,807][262198] Updated weights for policy 0, policy_version 36042 (0.0009)
+[2026-06-06 22:04:22,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3788.8, 300 sec: 3547.5). Total num frames: 18456576. Throughput: 0: 3837.1. Samples: 18461620. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 22:04:22,924][261293] Avg episode reward: [(0, '1781.127')]
+[2026-06-06 22:04:23,228][262198] Updated weights for policy 0, policy_version 36052 (0.0008)
+[2026-06-06 22:04:24,834][262198] Updated weights for policy 0, policy_version 36062 (0.0011)
+[2026-06-06 22:04:25,814][262198] Updated weights for policy 0, policy_version 36072 (0.0009)
+[2026-06-06 22:04:27,482][262198] Updated weights for policy 0, policy_version 36082 (0.0010)
+[2026-06-06 22:04:27,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3788.8, 300 sec: 3547.6). Total num frames: 18475008. Throughput: 0: 3825.4. Samples: 18483272. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 22:04:27,923][261293] Avg episode reward: [(0, '1513.644')]
+[2026-06-06 22:04:28,634][262198] Updated weights for policy 0, policy_version 36092 (0.0009)
+[2026-06-06 22:04:30,117][262198] Updated weights for policy 0, policy_version 36102 (0.0008)
+[2026-06-06 22:04:31,190][262198] Updated weights for policy 0, policy_version 36112 (0.0009)
+[2026-06-06 22:04:32,789][262198] Updated weights for policy 0, policy_version 36122 (0.0008)
+[2026-06-06 22:04:32,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3822.9, 300 sec: 3547.6). Total num frames: 18495488. Throughput: 0: 3838.4. Samples: 18494776. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 22:04:32,922][261293] Avg episode reward: [(0, '1655.393')]
+[2026-06-06 22:04:34,268][262198] Updated weights for policy 0, policy_version 36134 (0.0009)
+[2026-06-06 22:04:35,435][262198] Updated weights for policy 0, policy_version 36144 (0.0008)
+[2026-06-06 22:04:37,023][262198] Updated weights for policy 0, policy_version 36154 (0.0008)
+[2026-06-06 22:04:37,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3822.9, 300 sec: 3554.5). Total num frames: 18513920. Throughput: 0: 3812.4. Samples: 18518004. Policy #0 lag: (min: 3.0, avg: 21.1, max: 39.0)
+[2026-06-06 22:04:37,927][261293] Avg episode reward: [(0, '1815.640')]
+[2026-06-06 22:04:38,024][262198] Updated weights for policy 0, policy_version 36164 (0.0010)
+[2026-06-06 22:04:39,561][262198] Updated weights for policy 0, policy_version 36174 (0.0008)
+[2026-06-06 22:04:41,331][262198] Updated weights for policy 0, policy_version 36186 (0.0010)
+[2026-06-06 22:04:42,503][262198] Updated weights for policy 0, policy_version 36196 (0.0009)
+[2026-06-06 22:04:42,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3857.1, 300 sec: 3561.4). Total num frames: 18534400. Throughput: 0: 3818.2. Samples: 18541196. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 22:04:42,923][261293] Avg episode reward: [(0, '1593.358')]
+[2026-06-06 22:04:43,831][262198] Updated weights for policy 0, policy_version 36206 (0.0009)
+[2026-06-06 22:04:45,642][262198] Updated weights for policy 0, policy_version 36219 (0.0010)
+[2026-06-06 22:04:47,616][262198] Updated weights for policy 0, policy_version 36232 (0.0010)
+[2026-06-06 22:04:47,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3788.8, 300 sec: 3554.5). Total num frames: 18550784. Throughput: 0: 3793.8. Samples: 18552320. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 22:04:47,923][261293] Avg episode reward: [(0, '1790.680')]
+[2026-06-06 22:04:49,651][262198] Updated weights for policy 0, policy_version 36244 (0.0011)
+[2026-06-06 22:04:51,848][262198] Updated weights for policy 0, policy_version 36254 (0.0012)
+[2026-06-06 22:04:52,685][262198] Updated weights for policy 0, policy_version 36264 (0.0012)
+[2026-06-06 22:04:52,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3754.7, 300 sec: 3540.6). Total num frames: 18567168. Throughput: 0: 3710.0. Samples: 18571372. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 22:04:52,922][261293] Avg episode reward: [(0, '1588.920')]
+[2026-06-06 22:04:55,709][262198] Updated weights for policy 0, policy_version 36277 (0.0014)
+[2026-06-06 22:04:57,654][262198] Updated weights for policy 0, policy_version 36287 (0.0011)
+[2026-06-06 22:04:57,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3618.1, 300 sec: 3519.8). Total num frames: 18579456. Throughput: 0: 3594.4. Samples: 18588210. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 22:04:57,925][261293] Avg episode reward: [(0, '1663.706')]
+[2026-06-06 22:04:59,353][262198] Updated weights for policy 0, policy_version 36297 (0.0011)
+[2026-06-06 22:05:01,303][262198] Updated weights for policy 0, policy_version 36309 (0.0011)
+[2026-06-06 22:05:02,921][261293] Fps is (10 sec: 2662.3, 60 sec: 3549.9, 300 sec: 3512.8). Total num frames: 18593792. Throughput: 0: 3534.9. Samples: 18596908. Policy #0 lag: (min: 3.0, avg: 21.8, max: 43.0)
+[2026-06-06 22:05:02,922][261293] Avg episode reward: [(0, '1637.488')]
+[2026-06-06 22:05:03,327][262198] Updated weights for policy 0, policy_version 36319 (0.0012)
+[2026-06-06 22:05:05,062][262198] Updated weights for policy 0, policy_version 36329 (0.0011)
+[2026-06-06 22:05:07,273][262198] Updated weights for policy 0, policy_version 36341 (0.0014)
+[2026-06-06 22:05:07,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3447.5, 300 sec: 3499.0). Total num frames: 18608128. Throughput: 0: 3373.6. Samples: 18613432. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:05:07,923][261293] Avg episode reward: [(0, '1546.767')]
+[2026-06-06 22:05:09,339][262198] Updated weights for policy 0, policy_version 36351 (0.0010)
+[2026-06-06 22:05:10,745][262198] Updated weights for policy 0, policy_version 36361 (0.0011)
+[2026-06-06 22:05:12,390][262198] Updated weights for policy 0, policy_version 36373 (0.0010)
+[2026-06-06 22:05:12,921][261293] Fps is (10 sec: 3071.9, 60 sec: 3413.3, 300 sec: 3485.1). Total num frames: 18624512. Throughput: 0: 3320.4. Samples: 18632692. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:05:12,923][261293] Avg episode reward: [(0, '1756.485')]
+[2026-06-06 22:05:14,455][262198] Updated weights for policy 0, policy_version 36384 (0.0010)
+[2026-06-06 22:05:15,733][262198] Updated weights for policy 0, policy_version 36396 (0.0012)
+[2026-06-06 22:05:17,428][262198] Updated weights for policy 0, policy_version 36408 (0.0009)
+[2026-06-06 22:05:17,920][261293] Fps is (10 sec: 3277.0, 60 sec: 3413.3, 300 sec: 3478.1). Total num frames: 18640896. Throughput: 0: 3278.9. Samples: 18642326. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:05:17,923][261293] Avg episode reward: [(0, '1886.352')]
+[2026-06-06 22:05:19,875][262198] Updated weights for policy 0, policy_version 36420 (0.0009)
+[2026-06-06 22:05:21,072][262198] Updated weights for policy 0, policy_version 36430 (0.0009)
+[2026-06-06 22:05:21,886][262115] Signal inference workers to stop experience collection... (3750 times)
+[2026-06-06 22:05:21,923][262198] InferenceWorker_p0-w0: stopping experience collection (3750 times)
+[2026-06-06 22:05:21,988][262115] Signal inference workers to resume experience collection... (3750 times)
+[2026-06-06 22:05:21,989][262198] InferenceWorker_p0-w0: resuming experience collection (3750 times)
+[2026-06-06 22:05:22,738][262198] Updated weights for policy 0, policy_version 36441 (0.0010)
+[2026-06-06 22:05:22,920][261293] Fps is (10 sec: 3481.9, 60 sec: 3379.2, 300 sec: 3471.2). Total num frames: 18659328. Throughput: 0: 3221.5. Samples: 18662968. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:05:22,922][261293] Avg episode reward: [(0, '1707.322')]
+[2026-06-06 22:05:23,969][262198] Updated weights for policy 0, policy_version 36451 (0.0009)
+[2026-06-06 22:05:25,741][262198] Updated weights for policy 0, policy_version 36461 (0.0009)
+[2026-06-06 22:05:26,962][262198] Updated weights for policy 0, policy_version 36472 (0.0011)
+[2026-06-06 22:05:27,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3345.0, 300 sec: 3457.3). Total num frames: 18675712. Throughput: 0: 3158.4. Samples: 18683326. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 22:05:27,924][261293] Avg episode reward: [(0, '1881.984')]
+[2026-06-06 22:05:29,033][262198] Updated weights for policy 0, policy_version 36483 (0.0010)
+[2026-06-06 22:05:30,519][262198] Updated weights for policy 0, policy_version 36493 (0.0010)
+[2026-06-06 22:05:31,840][262198] Updated weights for policy 0, policy_version 36504 (0.0011)
+[2026-06-06 22:05:32,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3276.8, 300 sec: 3457.3). Total num frames: 18692096. Throughput: 0: 3147.7. Samples: 18693964. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 22:05:32,922][261293] Avg episode reward: [(0, '1535.418')]
+[2026-06-06 22:05:34,261][262198] Updated weights for policy 0, policy_version 36515 (0.0013)
+[2026-06-06 22:05:36,011][262198] Updated weights for policy 0, policy_version 36525 (0.0010)
+[2026-06-06 22:05:37,117][262198] Updated weights for policy 0, policy_version 36536 (0.0010)
+[2026-06-06 22:05:37,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3242.7, 300 sec: 3450.4). Total num frames: 18708480. Throughput: 0: 3127.7. Samples: 18712120. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 22:05:37,925][261293] Avg episode reward: [(0, '1803.904')]
+[2026-06-06 22:05:39,227][262198] Updated weights for policy 0, policy_version 36547 (0.0011)
+[2026-06-06 22:05:40,676][262198] Updated weights for policy 0, policy_version 36557 (0.0009)
+[2026-06-06 22:05:42,503][262198] Updated weights for policy 0, policy_version 36569 (0.0008)
+[2026-06-06 22:05:42,921][261293] Fps is (10 sec: 3276.6, 60 sec: 3174.4, 300 sec: 3443.5). Total num frames: 18724864. Throughput: 0: 3212.6. Samples: 18732776. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 22:05:42,923][261293] Avg episode reward: [(0, '1845.745')]
+[2026-06-06 22:05:43,946][262198] Updated weights for policy 0, policy_version 36580 (0.0010)
+[2026-06-06 22:05:45,608][262198] Updated weights for policy 0, policy_version 36590 (0.0009)
+[2026-06-06 22:05:47,799][262198] Updated weights for policy 0, policy_version 36602 (0.0012)
+[2026-06-06 22:05:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3174.4, 300 sec: 3436.5). Total num frames: 18741248. Throughput: 0: 3236.9. Samples: 18742568. Policy #0 lag: (min: 7.0, avg: 21.9, max: 39.0)
+[2026-06-06 22:05:47,923][261293] Avg episode reward: [(0, '1845.671')]
+[2026-06-06 22:05:49,296][262198] Updated weights for policy 0, policy_version 36612 (0.0010)
+[2026-06-06 22:05:51,172][262198] Updated weights for policy 0, policy_version 36622 (0.0013)
+[2026-06-06 22:05:52,921][261293] Fps is (10 sec: 3071.9, 60 sec: 3140.2, 300 sec: 3422.6). Total num frames: 18755584. Throughput: 0: 3267.5. Samples: 18760472. Policy #0 lag: (min: 4.0, avg: 21.6, max: 36.0)
+[2026-06-06 22:05:52,924][261293] Avg episode reward: [(0, '1758.881')]
+[2026-06-06 22:05:53,155][262198] Updated weights for policy 0, policy_version 36633 (0.0010)
+[2026-06-06 22:05:54,634][262198] Updated weights for policy 0, policy_version 36644 (0.0010)
+[2026-06-06 22:05:56,158][262198] Updated weights for policy 0, policy_version 36654 (0.0008)
+[2026-06-06 22:05:57,882][262198] Updated weights for policy 0, policy_version 36666 (0.0009)
+[2026-06-06 22:05:57,922][261293] Fps is (10 sec: 3072.0, 60 sec: 3208.5, 300 sec: 3422.6). Total num frames: 18771968. Throughput: 0: 3293.8. Samples: 18780910. Policy #0 lag: (min: 4.0, avg: 21.6, max: 36.0)
+[2026-06-06 22:05:57,927][261293] Avg episode reward: [(0, '1720.487')]
+[2026-06-06 22:05:59,614][262198] Updated weights for policy 0, policy_version 36677 (0.0011)
+[2026-06-06 22:06:00,907][262198] Updated weights for policy 0, policy_version 36687 (0.0009)
+[2026-06-06 22:06:02,634][262198] Updated weights for policy 0, policy_version 36698 (0.0007)
+[2026-06-06 22:06:02,921][261293] Fps is (10 sec: 3481.8, 60 sec: 3276.8, 300 sec: 3422.6). Total num frames: 18790400. Throughput: 0: 3300.6. Samples: 18790856. Policy #0 lag: (min: 4.0, avg: 21.6, max: 36.0)
+[2026-06-06 22:06:02,922][261293] Avg episode reward: [(0, '1640.029')]
+[2026-06-06 22:06:04,185][262198] Updated weights for policy 0, policy_version 36710 (0.0009)
+[2026-06-06 22:06:05,665][262198] Updated weights for policy 0, policy_version 36720 (0.0009)
+[2026-06-06 22:06:07,699][262198] Updated weights for policy 0, policy_version 36732 (0.0013)
+[2026-06-06 22:06:07,920][261293] Fps is (10 sec: 3481.8, 60 sec: 3311.0, 300 sec: 3408.7). Total num frames: 18806784. Throughput: 0: 3321.6. Samples: 18812440. Policy #0 lag: (min: 4.0, avg: 21.6, max: 36.0)
+[2026-06-06 22:06:07,922][261293] Avg episode reward: [(0, '1867.490')]
+[2026-06-06 22:06:09,672][262198] Updated weights for policy 0, policy_version 36742 (0.0013)
+[2026-06-06 22:06:11,556][262198] Updated weights for policy 0, policy_version 36752 (0.0012)
+[2026-06-06 22:06:12,921][261293] Fps is (10 sec: 3071.8, 60 sec: 3276.8, 300 sec: 3394.8). Total num frames: 18821120. Throughput: 0: 3213.7. Samples: 18827942. Policy #0 lag: (min: 4.0, avg: 21.6, max: 36.0)
+[2026-06-06 22:06:12,925][261293] Avg episode reward: [(0, '1711.246')]
+[2026-06-06 22:06:13,930][262198] Updated weights for policy 0, policy_version 36762 (0.0013)
+[2026-06-06 22:06:15,712][262198] Updated weights for policy 0, policy_version 36772 (0.0013)
+[2026-06-06 22:06:17,599][262198] Updated weights for policy 0, policy_version 36782 (0.0011)
+[2026-06-06 22:06:17,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3208.5, 300 sec: 3380.9). Total num frames: 18833408. Throughput: 0: 3142.3. Samples: 18835366. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 22:06:17,922][261293] Avg episode reward: [(0, '1748.704')]
+[2026-06-06 22:06:19,771][262198] Updated weights for policy 0, policy_version 36793 (0.0013)
+[2026-06-06 22:06:21,390][262198] Updated weights for policy 0, policy_version 36803 (0.0013)
+[2026-06-06 22:06:22,921][261293] Fps is (10 sec: 2662.5, 60 sec: 3140.3, 300 sec: 3394.8). Total num frames: 18847744. Throughput: 0: 3116.5. Samples: 18852364. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 22:06:22,923][261293] Avg episode reward: [(0, '1651.011')]
+[2026-06-06 22:06:23,462][262198] Updated weights for policy 0, policy_version 36813 (0.0012)
+[2026-06-06 22:06:24,700][262198] Updated weights for policy 0, policy_version 36823 (0.0012)
+[2026-06-06 22:06:27,562][262198] Updated weights for policy 0, policy_version 36834 (0.0013)
+[2026-06-06 22:06:27,921][261293] Fps is (10 sec: 2662.3, 60 sec: 3072.0, 300 sec: 3387.9). Total num frames: 18860032. Throughput: 0: 3007.3. Samples: 18868102. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 22:06:27,922][261293] Avg episode reward: [(0, '1879.911')]
+[2026-06-06 22:06:29,081][262198] Updated weights for policy 0, policy_version 36844 (0.0013)
+[2026-06-06 22:06:31,292][262198] Updated weights for policy 0, policy_version 36855 (0.0011)
+[2026-06-06 22:06:32,921][261293] Fps is (10 sec: 2457.5, 60 sec: 3003.7, 300 sec: 3380.9). Total num frames: 18872320. Throughput: 0: 2968.8. Samples: 18876164. Policy #0 lag: (min: 3.0, avg: 19.2, max: 35.0)
+[2026-06-06 22:06:32,924][261293] Avg episode reward: [(0, '1489.368')]
+[2026-06-06 22:06:34,117][262198] Updated weights for policy 0, policy_version 36865 (0.0012)
+[2026-06-06 22:06:35,925][262198] Updated weights for policy 0, policy_version 36875 (0.0013)
+[2026-06-06 22:06:37,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2935.5, 300 sec: 3380.9). Total num frames: 18884608. Throughput: 0: 2887.0. Samples: 18890386. Policy #0 lag: (min: 3.0, avg: 20.3, max: 38.0)
+[2026-06-06 22:06:37,923][261293] Avg episode reward: [(0, '1626.580')]
+[2026-06-06 22:06:37,974][262198] Updated weights for policy 0, policy_version 36886 (0.0012)
+[2026-06-06 22:06:40,290][262198] Updated weights for policy 0, policy_version 36896 (0.0013)
+[2026-06-06 22:06:42,483][262198] Updated weights for policy 0, policy_version 36906 (0.0013)
+[2026-06-06 22:06:42,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2867.2, 300 sec: 3367.0). Total num frames: 18896896. Throughput: 0: 2762.8. Samples: 18905236. Policy #0 lag: (min: 3.0, avg: 20.3, max: 38.0)
+[2026-06-06 22:06:42,923][261293] Avg episode reward: [(0, '1510.506')]
+[2026-06-06 22:06:44,279][262198] Updated weights for policy 0, policy_version 36916 (0.0012)
+[2026-06-06 22:06:46,672][262198] Updated weights for policy 0, policy_version 36926 (0.0011)
+[2026-06-06 22:06:47,659][262198] Updated weights for policy 0, policy_version 36936 (0.0011)
+[2026-06-06 22:06:47,922][261293] Fps is (10 sec: 2662.2, 60 sec: 2833.0, 300 sec: 3367.0). Total num frames: 18911232. Throughput: 0: 2723.8. Samples: 18913428. Policy #0 lag: (min: 3.0, avg: 20.3, max: 38.0)
+[2026-06-06 22:06:47,925][261293] Avg episode reward: [(0, '1759.967')]
+[2026-06-06 22:06:50,458][262198] Updated weights for policy 0, policy_version 36946 (0.0012)
+[2026-06-06 22:06:51,775][262198] Updated weights for policy 0, policy_version 36956 (0.0012)
+[2026-06-06 22:06:52,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2764.8, 300 sec: 3353.2). Total num frames: 18921472. Throughput: 0: 2589.5. Samples: 18928968. Policy #0 lag: (min: 3.0, avg: 20.3, max: 38.0)
+[2026-06-06 22:06:52,923][261293] Avg episode reward: [(0, '1707.531')]
+[2026-06-06 22:06:53,790][262198] Updated weights for policy 0, policy_version 36966 (0.0011)
+[2026-06-06 22:06:55,692][262198] Updated weights for policy 0, policy_version 36976 (0.0012)
+[2026-06-06 22:06:57,921][261293] Fps is (10 sec: 2457.7, 60 sec: 2730.7, 300 sec: 3367.0). Total num frames: 18935808. Throughput: 0: 2590.6. Samples: 18944520. Policy #0 lag: (min: 3.0, avg: 20.3, max: 38.0)
+[2026-06-06 22:06:57,924][261293] Avg episode reward: [(0, '1831.308')]
+[2026-06-06 22:06:58,060][262198] Updated weights for policy 0, policy_version 36986 (0.0013)
+[2026-06-06 22:06:59,962][262198] Updated weights for policy 0, policy_version 36996 (0.0014)
+[2026-06-06 22:07:02,490][262198] Updated weights for policy 0, policy_version 37006 (0.0013)
+[2026-06-06 22:07:02,921][261293] Fps is (10 sec: 2662.4, 60 sec: 2628.3, 300 sec: 3353.2). Total num frames: 18948096. Throughput: 0: 2584.9. Samples: 18951686. Policy #0 lag: (min: 3.0, avg: 20.4, max: 37.0)
+[2026-06-06 22:07:02,923][261293] Avg episode reward: [(0, '1722.692')]
+[2026-06-06 22:07:03,893][262198] Updated weights for policy 0, policy_version 37016 (0.0012)
+[2026-06-06 22:07:06,514][262198] Updated weights for policy 0, policy_version 37026 (0.0014)
+[2026-06-06 22:07:07,921][261293] Fps is (10 sec: 2457.5, 60 sec: 2560.0, 300 sec: 3339.3). Total num frames: 18960384. Throughput: 0: 2538.2. Samples: 18966584. Policy #0 lag: (min: 3.0, avg: 20.4, max: 37.0)
+[2026-06-06 22:07:07,923][261293] Avg episode reward: [(0, '1692.775')]
+[2026-06-06 22:07:08,254][262198] Updated weights for policy 0, policy_version 37036 (0.0013)
+[2026-06-06 22:07:10,331][262198] Updated weights for policy 0, policy_version 37048 (0.0015)
+[2026-06-06 22:07:12,859][262198] Updated weights for policy 0, policy_version 37058 (0.0011)
+[2026-06-06 22:07:12,920][261293] Fps is (10 sec: 2662.5, 60 sec: 2560.0, 300 sec: 3332.3). Total num frames: 18974720. Throughput: 0: 2533.0. Samples: 18982088. Policy #0 lag: (min: 3.0, avg: 20.4, max: 37.0)
+[2026-06-06 22:07:12,921][261293] Avg episode reward: [(0, '1697.886')]
+[2026-06-06 22:07:14,458][262198] Updated weights for policy 0, policy_version 37068 (0.0012)
+[2026-06-06 22:07:16,453][262198] Updated weights for policy 0, policy_version 37078 (0.0012)
+[2026-06-06 22:07:17,921][261293] Fps is (10 sec: 2662.5, 60 sec: 2560.0, 300 sec: 3318.5). Total num frames: 18987008. Throughput: 0: 2527.2. Samples: 18989890. Policy #0 lag: (min: 3.0, avg: 20.4, max: 37.0)
+[2026-06-06 22:07:17,923][261293] Avg episode reward: [(0, '1744.014')]
+[2026-06-06 22:07:18,701][262198] Updated weights for policy 0, policy_version 37088 (0.0011)
+[2026-06-06 22:07:20,572][262198] Updated weights for policy 0, policy_version 37098 (0.0011)
+[2026-06-06 22:07:22,416][262198] Updated weights for policy 0, policy_version 37108 (0.0012)
+[2026-06-06 22:07:22,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2560.0, 300 sec: 3311.5). Total num frames: 19001344. Throughput: 0: 2569.4. Samples: 19006008. Policy #0 lag: (min: 3.0, avg: 20.4, max: 37.0)
+[2026-06-06 22:07:22,922][261293] Avg episode reward: [(0, '1788.370')]
+[2026-06-06 22:07:24,593][262198] Updated weights for policy 0, policy_version 37118 (0.0011)
+[2026-06-06 22:07:25,726][262198] Updated weights for policy 0, policy_version 37128 (0.0012)
+[2026-06-06 22:07:27,921][261293] Fps is (10 sec: 2662.4, 60 sec: 2560.0, 300 sec: 3290.7). Total num frames: 19013632. Throughput: 0: 2598.2. Samples: 19022154. Policy #0 lag: (min: 2.0, avg: 18.6, max: 36.0)
+[2026-06-06 22:07:27,923][261293] Avg episode reward: [(0, '1692.460')]
+[2026-06-06 22:07:28,255][262198] Updated weights for policy 0, policy_version 37138 (0.0013)
+[2026-06-06 22:07:29,757][262198] Updated weights for policy 0, policy_version 37148 (0.0011)
+[2026-06-06 22:07:31,714][262198] Updated weights for policy 0, policy_version 37158 (0.0010)
+[2026-06-06 22:07:32,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2594.2, 300 sec: 3283.7). Total num frames: 19027968. Throughput: 0: 2592.6. Samples: 19030090. Policy #0 lag: (min: 2.0, avg: 18.6, max: 36.0)
+[2026-06-06 22:07:32,921][261293] Avg episode reward: [(0, '1787.820')]
+[2026-06-06 22:07:33,738][262198] Updated weights for policy 0, policy_version 37168 (0.0012)
+[2026-06-06 22:07:35,737][262198] Updated weights for policy 0, policy_version 37178 (0.0012)
+[2026-06-06 22:07:37,561][262198] Updated weights for policy 0, policy_version 37188 (0.0012)
+[2026-06-06 22:07:37,920][261293] Fps is (10 sec: 2662.5, 60 sec: 2594.1, 300 sec: 3256.0). Total num frames: 19040256. Throughput: 0: 2602.6. Samples: 19046086. Policy #0 lag: (min: 2.0, avg: 18.6, max: 36.0)
+[2026-06-06 22:07:37,921][261293] Avg episode reward: [(0, '1615.770')]
+[2026-06-06 22:07:39,676][262198] Updated weights for policy 0, policy_version 37198 (0.0011)
+[2026-06-06 22:07:41,819][262198] Updated weights for policy 0, policy_version 37209 (0.0012)
+[2026-06-06 22:07:42,920][261293] Fps is (10 sec: 2457.5, 60 sec: 2594.1, 300 sec: 3228.2). Total num frames: 19052544. Throughput: 0: 2617.3. Samples: 19062298. Policy #0 lag: (min: 2.0, avg: 18.6, max: 36.0)
+[2026-06-06 22:07:42,921][261293] Avg episode reward: [(0, '1681.338')]
+[2026-06-06 22:07:43,427][262198] Updated weights for policy 0, policy_version 37219 (0.0011)
+[2026-06-06 22:07:46,055][262198] Updated weights for policy 0, policy_version 37229 (0.0014)
+[2026-06-06 22:07:47,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2594.2, 300 sec: 3214.3). Total num frames: 19066880. Throughput: 0: 2623.0. Samples: 19069720. Policy #0 lag: (min: 2.0, avg: 18.6, max: 36.0)
+[2026-06-06 22:07:47,923][261293] Avg episode reward: [(0, '1846.579')]
+[2026-06-06 22:07:48,099][262198] Updated weights for policy 0, policy_version 37241 (0.0010)
+[2026-06-06 22:07:49,537][262198] Updated weights for policy 0, policy_version 37251 (0.0009)
+[2026-06-06 22:07:50,988][262198] Updated weights for policy 0, policy_version 37261 (0.0011)
+[2026-06-06 22:07:52,679][262198] Updated weights for policy 0, policy_version 37273 (0.0011)
+[2026-06-06 22:07:52,921][261293] Fps is (10 sec: 3276.8, 60 sec: 2730.7, 300 sec: 3214.3). Total num frames: 19085312. Throughput: 0: 2721.7. Samples: 19089056. Policy #0 lag: (min: 3.0, avg: 17.5, max: 37.0)
+[2026-06-06 22:07:52,924][261293] Avg episode reward: [(0, '1677.807')]
+[2026-06-06 22:07:54,322][262198] Updated weights for policy 0, policy_version 37285 (0.0011)
+[2026-06-06 22:07:55,883][262198] Updated weights for policy 0, policy_version 37295 (0.0009)
+[2026-06-06 22:07:57,771][262198] Updated weights for policy 0, policy_version 37307 (0.0011)
+[2026-06-06 22:07:57,926][261293] Fps is (10 sec: 3479.8, 60 sec: 2764.6, 300 sec: 3207.3). Total num frames: 19101696. Throughput: 0: 2822.2. Samples: 19109104. Policy #0 lag: (min: 3.0, avg: 17.5, max: 37.0)
+[2026-06-06 22:07:57,929][261293] Avg episode reward: [(0, '1630.975')]
+[2026-06-06 22:07:59,090][262115] Signal inference workers to stop experience collection... (3800 times)
+[2026-06-06 22:07:59,115][262198] InferenceWorker_p0-w0: stopping experience collection (3800 times)
+[2026-06-06 22:07:59,122][262115] Signal inference workers to resume experience collection... (3800 times)
+[2026-06-06 22:07:59,134][262198] InferenceWorker_p0-w0: resuming experience collection (3800 times)
+[2026-06-06 22:07:59,216][262198] Updated weights for policy 0, policy_version 37317 (0.0009)
+[2026-06-06 22:08:00,581][262198] Updated weights for policy 0, policy_version 37327 (0.0008)
+[2026-06-06 22:08:02,390][262198] Updated weights for policy 0, policy_version 37338 (0.0011)
+[2026-06-06 22:08:02,921][261293] Fps is (10 sec: 3276.7, 60 sec: 2833.1, 300 sec: 3207.4). Total num frames: 19118080. Throughput: 0: 2887.3. Samples: 19119816. Policy #0 lag: (min: 3.0, avg: 17.5, max: 37.0)
+[2026-06-06 22:08:02,923][261293] Avg episode reward: [(0, '1739.454')]
+[2026-06-06 22:08:03,916][262198] Updated weights for policy 0, policy_version 37348 (0.0008)
+[2026-06-06 22:08:05,411][262198] Updated weights for policy 0, policy_version 37358 (0.0009)
+[2026-06-06 22:08:07,088][262198] Updated weights for policy 0, policy_version 37370 (0.0010)
+[2026-06-06 22:08:07,921][261293] Fps is (10 sec: 3278.3, 60 sec: 2901.3, 300 sec: 3193.5). Total num frames: 19134464. Throughput: 0: 2991.1. Samples: 19140610. Policy #0 lag: (min: 3.0, avg: 17.5, max: 37.0)
+[2026-06-06 22:08:07,923][261293] Avg episode reward: [(0, '1764.414')]
+[2026-06-06 22:08:08,642][262198] Updated weights for policy 0, policy_version 37380 (0.0009)
+[2026-06-06 22:08:10,206][262198] Updated weights for policy 0, policy_version 37390 (0.0009)
+[2026-06-06 22:08:12,025][262198] Updated weights for policy 0, policy_version 37403 (0.0010)
+[2026-06-06 22:08:12,921][261293] Fps is (10 sec: 3276.8, 60 sec: 2935.4, 300 sec: 3179.6). Total num frames: 19150848. Throughput: 0: 3075.5. Samples: 19160550. Policy #0 lag: (min: 3.0, avg: 17.5, max: 37.0)
+[2026-06-06 22:08:12,926][261293] Avg episode reward: [(0, '1804.701')]
+[2026-06-06 22:08:13,712][262198] Updated weights for policy 0, policy_version 37414 (0.0009)
+[2026-06-06 22:08:15,232][262198] Updated weights for policy 0, policy_version 37424 (0.0011)
+[2026-06-06 22:08:16,974][262198] Updated weights for policy 0, policy_version 37436 (0.0012)
+[2026-06-06 22:08:17,920][261293] Fps is (10 sec: 3277.1, 60 sec: 3003.8, 300 sec: 3179.6). Total num frames: 19167232. Throughput: 0: 3127.8. Samples: 19170842. Policy #0 lag: (min: 7.0, avg: 20.3, max: 39.0)
+[2026-06-06 22:08:17,922][261293] Avg episode reward: [(0, '1675.786')]
+[2026-06-06 22:08:18,543][262198] Updated weights for policy 0, policy_version 37446 (0.0009)
+[2026-06-06 22:08:20,045][262198] Updated weights for policy 0, policy_version 37456 (0.0010)
+[2026-06-06 22:08:22,139][262198] Updated weights for policy 0, policy_version 37467 (0.0013)
+[2026-06-06 22:08:22,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3037.9, 300 sec: 3172.7). Total num frames: 19183616. Throughput: 0: 3198.8. Samples: 19190030. Policy #0 lag: (min: 7.0, avg: 20.3, max: 39.0)
+[2026-06-06 22:08:22,921][261293] Avg episode reward: [(0, '1718.037')]
+[2026-06-06 22:08:23,632][262198] Updated weights for policy 0, policy_version 37478 (0.0010)
+[2026-06-06 22:08:25,022][262198] Updated weights for policy 0, policy_version 37488 (0.0010)
+[2026-06-06 22:08:26,897][262198] Updated weights for policy 0, policy_version 37500 (0.0009)
+[2026-06-06 22:08:27,920][261293] Fps is (10 sec: 3481.6, 60 sec: 3140.3, 300 sec: 3172.7). Total num frames: 19202048. Throughput: 0: 3293.1. Samples: 19210488. Policy #0 lag: (min: 7.0, avg: 20.3, max: 39.0)
+[2026-06-06 22:08:27,922][261293] Avg episode reward: [(0, '1746.391')]
+[2026-06-06 22:08:28,412][262198] Updated weights for policy 0, policy_version 37510 (0.0010)
+[2026-06-06 22:08:29,823][262198] Updated weights for policy 0, policy_version 37520 (0.0009)
+[2026-06-06 22:08:31,588][262198] Updated weights for policy 0, policy_version 37531 (0.0011)
+[2026-06-06 22:08:32,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3174.4, 300 sec: 3165.7). Total num frames: 19218432. Throughput: 0: 3363.5. Samples: 19221076. Policy #0 lag: (min: 7.0, avg: 20.3, max: 39.0)
+[2026-06-06 22:08:32,923][261293] Avg episode reward: [(0, '1741.392')]
+[2026-06-06 22:08:33,221][262198] Updated weights for policy 0, policy_version 37541 (0.0010)
+[2026-06-06 22:08:34,701][262198] Updated weights for policy 0, policy_version 37551 (0.0011)
+[2026-06-06 22:08:36,723][262198] Updated weights for policy 0, policy_version 37563 (0.0011)
+[2026-06-06 22:08:37,922][261293] Fps is (10 sec: 3276.1, 60 sec: 3242.6, 300 sec: 3158.8). Total num frames: 19234816. Throughput: 0: 3360.6. Samples: 19240290. Policy #0 lag: (min: 6.0, avg: 21.8, max: 38.0)
+[2026-06-06 22:08:37,925][261293] Avg episode reward: [(0, '1733.450')]
+[2026-06-06 22:08:38,367][262198] Updated weights for policy 0, policy_version 37573 (0.0011)
+[2026-06-06 22:08:39,823][262198] Updated weights for policy 0, policy_version 37583 (0.0011)
+[2026-06-06 22:08:41,575][262198] Updated weights for policy 0, policy_version 37594 (0.0009)
+[2026-06-06 22:08:42,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3310.9, 300 sec: 3144.9). Total num frames: 19251200. Throughput: 0: 3363.0. Samples: 19260422. Policy #0 lag: (min: 6.0, avg: 21.8, max: 38.0)
+[2026-06-06 22:08:42,923][261293] Avg episode reward: [(0, '1635.658')]
+[2026-06-06 22:08:43,080][262198] Updated weights for policy 0, policy_version 37604 (0.0011)
+[2026-06-06 22:08:44,415][262198] Updated weights for policy 0, policy_version 37614 (0.0008)
+[2026-06-06 22:08:46,463][262198] Updated weights for policy 0, policy_version 37626 (0.0012)
+[2026-06-06 22:08:47,921][261293] Fps is (10 sec: 3277.3, 60 sec: 3345.1, 300 sec: 3138.0). Total num frames: 19267584. Throughput: 0: 3358.4. Samples: 19270944. Policy #0 lag: (min: 6.0, avg: 21.8, max: 38.0)
+[2026-06-06 22:08:47,922][261293] Avg episode reward: [(0, '1737.674')]
+[2026-06-06 22:08:48,047][262198] Updated weights for policy 0, policy_version 37636 (0.0011)
+[2026-06-06 22:08:49,602][262198] Updated weights for policy 0, policy_version 37646 (0.0010)
+[2026-06-06 22:08:51,478][262198] Updated weights for policy 0, policy_version 37657 (0.0011)
+[2026-06-06 22:08:52,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3310.9, 300 sec: 3124.1). Total num frames: 19283968. Throughput: 0: 3319.9. Samples: 19290004. Policy #0 lag: (min: 6.0, avg: 21.8, max: 38.0)
+[2026-06-06 22:08:52,923][261293] Avg episode reward: [(0, '1764.446')]
+[2026-06-06 22:08:53,104][262198] Updated weights for policy 0, policy_version 37667 (0.0009)
+[2026-06-06 22:08:54,907][262198] Updated weights for policy 0, policy_version 37677 (0.0010)
+[2026-06-06 22:08:57,023][262198] Updated weights for policy 0, policy_version 37689 (0.0011)
+[2026-06-06 22:08:57,921][261293] Fps is (10 sec: 3072.1, 60 sec: 3277.1, 300 sec: 3110.2). Total num frames: 19298304. Throughput: 0: 3273.1. Samples: 19307838. Policy #0 lag: (min: 6.0, avg: 21.8, max: 38.0)
+[2026-06-06 22:08:57,923][261293] Avg episode reward: [(0, '1815.225')]
+[2026-06-06 22:08:58,514][262198] Updated weights for policy 0, policy_version 37699 (0.0010)
+[2026-06-06 22:09:00,421][262198] Updated weights for policy 0, policy_version 37710 (0.0013)
+[2026-06-06 22:09:02,516][262198] Updated weights for policy 0, policy_version 37722 (0.0011)
+[2026-06-06 22:09:02,920][261293] Fps is (10 sec: 3072.1, 60 sec: 3276.8, 300 sec: 3096.3). Total num frames: 19314688. Throughput: 0: 3251.1. Samples: 19317140. Policy #0 lag: (min: 8.0, avg: 21.2, max: 40.0)
+[2026-06-06 22:09:02,922][261293] Avg episode reward: [(0, '1599.627')]
+[2026-06-06 22:09:04,359][262198] Updated weights for policy 0, policy_version 37735 (0.0010)
+[2026-06-06 22:09:06,573][262198] Updated weights for policy 0, policy_version 37745 (0.0011)
+[2026-06-06 22:09:07,749][262198] Updated weights for policy 0, policy_version 37756 (0.0011)
+[2026-06-06 22:09:07,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3276.8, 300 sec: 3089.4). Total num frames: 19331072. Throughput: 0: 3237.5. Samples: 19335716. Policy #0 lag: (min: 8.0, avg: 21.2, max: 40.0)
+[2026-06-06 22:09:07,922][261293] Avg episode reward: [(0, '1898.935')]
+[2026-06-06 22:09:07,927][262115] Saving new best policy, reward=1898.935!
+[2026-06-06 22:09:09,287][262198] Updated weights for policy 0, policy_version 37768 (0.0010)
+[2026-06-06 22:09:11,600][262198] Updated weights for policy 0, policy_version 37780 (0.0011)
+[2026-06-06 22:09:12,920][261293] Fps is (10 sec: 3276.8, 60 sec: 3276.8, 300 sec: 3089.4). Total num frames: 19347456. Throughput: 0: 3220.1. Samples: 19355394. Policy #0 lag: (min: 8.0, avg: 21.2, max: 40.0)
+[2026-06-06 22:09:12,922][261293] Avg episode reward: [(0, '1656.278')]
+[2026-06-06 22:09:13,340][262198] Updated weights for policy 0, policy_version 37790 (0.0010)
+[2026-06-06 22:09:14,781][262198] Updated weights for policy 0, policy_version 37801 (0.0012)
+[2026-06-06 22:09:16,391][262198] Updated weights for policy 0, policy_version 37813 (0.0009)
+[2026-06-06 22:09:17,840][262198] Updated weights for policy 0, policy_version 37823 (0.0008)
+[2026-06-06 22:09:17,922][261293] Fps is (10 sec: 3481.3, 60 sec: 3310.9, 300 sec: 3082.4). Total num frames: 19365888. Throughput: 0: 3217.9. Samples: 19365884. Policy #0 lag: (min: 8.0, avg: 21.2, max: 40.0)
+[2026-06-06 22:09:17,928][261293] Avg episode reward: [(0, '1796.910')]
+[2026-06-06 22:09:17,935][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000037824_19365888.pth...
+[2026-06-06 22:09:18,036][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000014732_7542784.pth
+[2026-06-06 22:09:19,178][262198] Updated weights for policy 0, policy_version 37835 (0.0010)
+[2026-06-06 22:09:20,779][262198] Updated weights for policy 0, policy_version 37846 (0.0010)
+[2026-06-06 22:09:22,420][262198] Updated weights for policy 0, policy_version 37856 (0.0010)
+[2026-06-06 22:09:22,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3310.9, 300 sec: 3075.5). Total num frames: 19382272. Throughput: 0: 3283.8. Samples: 19388054. Policy #0 lag: (min: 8.0, avg: 21.2, max: 40.0)
+[2026-06-06 22:09:22,924][261293] Avg episode reward: [(0, '1813.149')]
+[2026-06-06 22:09:23,934][262198] Updated weights for policy 0, policy_version 37868 (0.0011)
+[2026-06-06 22:09:25,582][262198] Updated weights for policy 0, policy_version 37880 (0.0011)
+[2026-06-06 22:09:27,778][262198] Updated weights for policy 0, policy_version 37890 (0.0009)
+[2026-06-06 22:09:27,921][261293] Fps is (10 sec: 3481.7, 60 sec: 3310.9, 300 sec: 3068.5). Total num frames: 19400704. Throughput: 0: 3310.4. Samples: 19409390. Policy #0 lag: (min: 3.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:09:27,925][261293] Avg episode reward: [(0, '1780.401')]
+[2026-06-06 22:09:28,889][262198] Updated weights for policy 0, policy_version 37901 (0.0009)
+[2026-06-06 22:09:29,629][262115] Signal inference workers to stop experience collection... (3850 times)
+[2026-06-06 22:09:29,648][262198] InferenceWorker_p0-w0: stopping experience collection (3850 times)
+[2026-06-06 22:09:29,732][262115] Signal inference workers to resume experience collection... (3850 times)
+[2026-06-06 22:09:29,734][262198] InferenceWorker_p0-w0: resuming experience collection (3850 times)
+[2026-06-06 22:09:29,978][262198] Updated weights for policy 0, policy_version 37912 (0.0009)
+[2026-06-06 22:09:32,050][262198] Updated weights for policy 0, policy_version 37924 (0.0008)
+[2026-06-06 22:09:32,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3379.2, 300 sec: 3075.5). Total num frames: 19421184. Throughput: 0: 3323.2. Samples: 19420488. Policy #0 lag: (min: 3.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:09:32,924][261293] Avg episode reward: [(0, '1847.911')]
+[2026-06-06 22:09:33,159][262198] Updated weights for policy 0, policy_version 37934 (0.0008)
+[2026-06-06 22:09:34,884][262198] Updated weights for policy 0, policy_version 37946 (0.0009)
+[2026-06-06 22:09:36,496][262198] Updated weights for policy 0, policy_version 37957 (0.0010)
+[2026-06-06 22:09:37,667][262198] Updated weights for policy 0, policy_version 37967 (0.0009)
+[2026-06-06 22:09:37,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3413.4, 300 sec: 3068.5). Total num frames: 19439616. Throughput: 0: 3416.9. Samples: 19443766. Policy #0 lag: (min: 3.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:09:37,923][261293] Avg episode reward: [(0, '1755.541')]
+[2026-06-06 22:09:39,379][262198] Updated weights for policy 0, policy_version 37978 (0.0013)
+[2026-06-06 22:09:41,034][262198] Updated weights for policy 0, policy_version 37989 (0.0011)
+[2026-06-06 22:09:42,149][262198] Updated weights for policy 0, policy_version 37999 (0.0009)
+[2026-06-06 22:09:42,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3447.5, 300 sec: 3075.5). Total num frames: 19458048. Throughput: 0: 3506.8. Samples: 19465646. Policy #0 lag: (min: 3.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:09:42,922][261293] Avg episode reward: [(0, '1594.944')]
+[2026-06-06 22:09:43,721][262198] Updated weights for policy 0, policy_version 38011 (0.0011)
+[2026-06-06 22:09:45,537][262198] Updated weights for policy 0, policy_version 38022 (0.0012)
+[2026-06-06 22:09:46,493][262198] Updated weights for policy 0, policy_version 38032 (0.0009)
+[2026-06-06 22:09:47,922][261293] Fps is (10 sec: 3686.1, 60 sec: 3481.6, 300 sec: 3082.4). Total num frames: 19476480. Throughput: 0: 3555.6. Samples: 19477144. Policy #0 lag: (min: 3.0, avg: 20.2, max: 39.0)
+[2026-06-06 22:09:47,927][261293] Avg episode reward: [(0, '1636.654')]
+[2026-06-06 22:09:48,010][262198] Updated weights for policy 0, policy_version 38042 (0.0009)
+[2026-06-06 22:09:49,640][262198] Updated weights for policy 0, policy_version 38053 (0.0010)
+[2026-06-06 22:09:50,698][262198] Updated weights for policy 0, policy_version 38063 (0.0009)
+[2026-06-06 22:09:52,063][262198] Updated weights for policy 0, policy_version 38073 (0.0008)
+[2026-06-06 22:09:52,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3549.9, 300 sec: 3110.2). Total num frames: 19496960. Throughput: 0: 3656.5. Samples: 19500258. Policy #0 lag: (min: 5.0, avg: 22.1, max: 39.0)
+[2026-06-06 22:09:52,923][261293] Avg episode reward: [(0, '1732.593')]
+[2026-06-06 22:09:53,571][262198] Updated weights for policy 0, policy_version 38084 (0.0008)
+[2026-06-06 22:09:54,794][262198] Updated weights for policy 0, policy_version 38094 (0.0008)
+[2026-06-06 22:09:55,745][262198] Updated weights for policy 0, policy_version 38104 (0.0008)
+[2026-06-06 22:09:57,882][262198] Updated weights for policy 0, policy_version 38116 (0.0010)
+[2026-06-06 22:09:57,920][261293] Fps is (10 sec: 3891.5, 60 sec: 3618.1, 300 sec: 3124.1). Total num frames: 19515392. Throughput: 0: 3736.1. Samples: 19523520. Policy #0 lag: (min: 5.0, avg: 22.1, max: 39.0)
+[2026-06-06 22:09:57,922][261293] Avg episode reward: [(0, '1725.235')]
+[2026-06-06 22:09:59,038][262198] Updated weights for policy 0, policy_version 38126 (0.0008)
+[2026-06-06 22:10:00,670][262198] Updated weights for policy 0, policy_version 38137 (0.0009)
+[2026-06-06 22:10:02,339][262198] Updated weights for policy 0, policy_version 38149 (0.0008)
+[2026-06-06 22:10:02,922][261293] Fps is (10 sec: 3890.8, 60 sec: 3686.3, 300 sec: 3144.9). Total num frames: 19535872. Throughput: 0: 3752.0. Samples: 19534724. Policy #0 lag: (min: 5.0, avg: 22.1, max: 39.0)
+[2026-06-06 22:10:02,925][261293] Avg episode reward: [(0, '1666.110')]
+[2026-06-06 22:10:03,443][262198] Updated weights for policy 0, policy_version 38159 (0.0008)
+[2026-06-06 22:10:04,994][262198] Updated weights for policy 0, policy_version 38171 (0.0009)
+[2026-06-06 22:10:06,690][262198] Updated weights for policy 0, policy_version 38183 (0.0009)
+[2026-06-06 22:10:07,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3720.5, 300 sec: 3151.8). Total num frames: 19554304. Throughput: 0: 3780.1. Samples: 19558158. Policy #0 lag: (min: 5.0, avg: 22.1, max: 39.0)
+[2026-06-06 22:10:07,923][261293] Avg episode reward: [(0, '1622.654')]
+[2026-06-06 22:10:08,313][262198] Updated weights for policy 0, policy_version 38193 (0.0009)
+[2026-06-06 22:10:09,543][262198] Updated weights for policy 0, policy_version 38204 (0.0011)
+[2026-06-06 22:10:11,369][262198] Updated weights for policy 0, policy_version 38216 (0.0013)
+[2026-06-06 22:10:12,921][261293] Fps is (10 sec: 3277.1, 60 sec: 3686.4, 300 sec: 3144.9). Total num frames: 19568640. Throughput: 0: 3744.1. Samples: 19577876. Policy #0 lag: (min: 5.0, avg: 22.1, max: 39.0)
+[2026-06-06 22:10:12,922][261293] Avg episode reward: [(0, '1702.187')]
+[2026-06-06 22:10:13,607][262198] Updated weights for policy 0, policy_version 38226 (0.0011)
+[2026-06-06 22:10:15,281][262198] Updated weights for policy 0, policy_version 38236 (0.0012)
+[2026-06-06 22:10:17,235][262198] Updated weights for policy 0, policy_version 38246 (0.0013)
+[2026-06-06 22:10:17,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3618.2, 300 sec: 3131.0). Total num frames: 19582976. Throughput: 0: 3675.3. Samples: 19585876. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 22:10:17,923][261293] Avg episode reward: [(0, '1763.751')]
+[2026-06-06 22:10:18,793][262198] Updated weights for policy 0, policy_version 38256 (0.0011)
+[2026-06-06 22:10:20,827][262198] Updated weights for policy 0, policy_version 38266 (0.0010)
+[2026-06-06 22:10:22,750][262198] Updated weights for policy 0, policy_version 38276 (0.0013)
+[2026-06-06 22:10:22,920][261293] Fps is (10 sec: 2867.3, 60 sec: 3584.0, 300 sec: 3124.1). Total num frames: 19597312. Throughput: 0: 3546.2. Samples: 19603344. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 22:10:22,922][261293] Avg episode reward: [(0, '1713.324')]
+[2026-06-06 22:10:24,476][262198] Updated weights for policy 0, policy_version 38286 (0.0011)
+[2026-06-06 22:10:25,629][262198] Updated weights for policy 0, policy_version 38296 (0.0011)
+[2026-06-06 22:10:27,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3515.7, 300 sec: 3117.1). Total num frames: 19611648. Throughput: 0: 3433.6. Samples: 19620160. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 22:10:27,923][261293] Avg episode reward: [(0, '1709.190')]
+[2026-06-06 22:10:28,207][262198] Updated weights for policy 0, policy_version 38306 (0.0012)
+[2026-06-06 22:10:29,394][262198] Updated weights for policy 0, policy_version 38316 (0.0012)
+[2026-06-06 22:10:31,207][262198] Updated weights for policy 0, policy_version 38326 (0.0011)
+[2026-06-06 22:10:32,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3413.3, 300 sec: 3110.2). Total num frames: 19625984. Throughput: 0: 3369.9. Samples: 19628786. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 22:10:32,923][261293] Avg episode reward: [(0, '1736.429')]
+[2026-06-06 22:10:32,986][262198] Updated weights for policy 0, policy_version 38336 (0.0010)
+[2026-06-06 22:10:34,725][262198] Updated weights for policy 0, policy_version 38347 (0.0009)
+[2026-06-06 22:10:36,379][262198] Updated weights for policy 0, policy_version 38357 (0.0010)
+[2026-06-06 22:10:37,600][262198] Updated weights for policy 0, policy_version 38367 (0.0009)
+[2026-06-06 22:10:37,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3413.3, 300 sec: 3117.1). Total num frames: 19644416. Throughput: 0: 3280.1. Samples: 19647862. Policy #0 lag: (min: 3.0, avg: 18.3, max: 35.0)
+[2026-06-06 22:10:37,922][261293] Avg episode reward: [(0, '1768.088')]
+[2026-06-06 22:10:39,304][262198] Updated weights for policy 0, policy_version 38378 (0.0009)
+[2026-06-06 22:10:40,733][262198] Updated weights for policy 0, policy_version 38388 (0.0010)
+[2026-06-06 22:10:42,343][262198] Updated weights for policy 0, policy_version 38399 (0.0010)
+[2026-06-06 22:10:42,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3379.2, 300 sec: 3117.1). Total num frames: 19660800. Throughput: 0: 3239.9. Samples: 19669314. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:10:42,922][261293] Avg episode reward: [(0, '1581.993')]
+[2026-06-06 22:10:44,206][262198] Updated weights for policy 0, policy_version 38411 (0.0008)
+[2026-06-06 22:10:46,018][262198] Updated weights for policy 0, policy_version 38421 (0.0009)
+[2026-06-06 22:10:47,289][262198] Updated weights for policy 0, policy_version 38431 (0.0009)
+[2026-06-06 22:10:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3345.1, 300 sec: 3124.1). Total num frames: 19677184. Throughput: 0: 3196.8. Samples: 19678576. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:10:47,923][261293] Avg episode reward: [(0, '1573.768')]
+[2026-06-06 22:10:49,184][262198] Updated weights for policy 0, policy_version 38441 (0.0010)
+[2026-06-06 22:10:50,607][262198] Updated weights for policy 0, policy_version 38451 (0.0009)
+[2026-06-06 22:10:52,317][262198] Updated weights for policy 0, policy_version 38463 (0.0010)
+[2026-06-06 22:10:52,921][261293] Fps is (10 sec: 3276.6, 60 sec: 3276.8, 300 sec: 3124.1). Total num frames: 19693568. Throughput: 0: 3125.3. Samples: 19698796. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:10:52,924][261293] Avg episode reward: [(0, '1861.287')]
+[2026-06-06 22:10:54,146][262198] Updated weights for policy 0, policy_version 38473 (0.0012)
+[2026-06-06 22:10:55,393][262198] Updated weights for policy 0, policy_version 38483 (0.0009)
+[2026-06-06 22:10:56,912][262198] Updated weights for policy 0, policy_version 38493 (0.0009)
+[2026-06-06 22:10:57,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3242.6, 300 sec: 3117.1). Total num frames: 19709952. Throughput: 0: 3141.6. Samples: 19719250. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:10:57,924][261293] Avg episode reward: [(0, '1814.564')]
+[2026-06-06 22:10:58,264][262198] Updated weights for policy 0, policy_version 38503 (0.0010)
+[2026-06-06 22:11:00,173][262198] Updated weights for policy 0, policy_version 38513 (0.0009)
+[2026-06-06 22:11:01,459][262198] Updated weights for policy 0, policy_version 38523 (0.0010)
+[2026-06-06 22:11:02,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3174.4, 300 sec: 3117.1). Total num frames: 19726336. Throughput: 0: 3186.6. Samples: 19729274. Policy #0 lag: (min: 3.0, avg: 19.9, max: 35.0)
+[2026-06-06 22:11:02,923][261293] Avg episode reward: [(0, '1832.889')]
+[2026-06-06 22:11:03,122][262198] Updated weights for policy 0, policy_version 38534 (0.0011)
+[2026-06-06 22:11:04,554][262198] Updated weights for policy 0, policy_version 38544 (0.0009)
+[2026-06-06 22:11:06,340][262198] Updated weights for policy 0, policy_version 38554 (0.0009)
+[2026-06-06 22:11:07,903][262115] Signal inference workers to stop experience collection... (3900 times)
+[2026-06-06 22:11:07,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3140.3, 300 sec: 3124.1). Total num frames: 19742720. Throughput: 0: 3245.8. Samples: 19749408. Policy #0 lag: (min: 4.0, avg: 19.0, max: 36.0)
+[2026-06-06 22:11:07,923][261293] Avg episode reward: [(0, '1622.030')]
+[2026-06-06 22:11:07,934][262198] InferenceWorker_p0-w0: stopping experience collection (3900 times)
+[2026-06-06 22:11:07,984][262115] Signal inference workers to resume experience collection... (3900 times)
+[2026-06-06 22:11:07,985][262198] InferenceWorker_p0-w0: resuming experience collection (3900 times)
+[2026-06-06 22:11:08,090][262198] Updated weights for policy 0, policy_version 38565 (0.0009)
+[2026-06-06 22:11:09,392][262198] Updated weights for policy 0, policy_version 38575 (0.0008)
+[2026-06-06 22:11:11,042][262198] Updated weights for policy 0, policy_version 38585 (0.0009)
+[2026-06-06 22:11:12,483][262198] Updated weights for policy 0, policy_version 38595 (0.0009)
+[2026-06-06 22:11:12,920][261293] Fps is (10 sec: 3686.7, 60 sec: 3242.7, 300 sec: 3151.8). Total num frames: 19763200. Throughput: 0: 3340.5. Samples: 19770482. Policy #0 lag: (min: 4.0, avg: 19.0, max: 36.0)
+[2026-06-06 22:11:12,923][261293] Avg episode reward: [(0, '1737.204')]
+[2026-06-06 22:11:13,792][262198] Updated weights for policy 0, policy_version 38605 (0.0009)
+[2026-06-06 22:11:14,844][262198] Updated weights for policy 0, policy_version 38615 (0.0009)
+[2026-06-06 22:11:17,078][262198] Updated weights for policy 0, policy_version 38627 (0.0009)
+[2026-06-06 22:11:17,921][261293] Fps is (10 sec: 3891.3, 60 sec: 3310.9, 300 sec: 3165.7). Total num frames: 19781632. Throughput: 0: 3389.3. Samples: 19781306. Policy #0 lag: (min: 4.0, avg: 19.0, max: 36.0)
+[2026-06-06 22:11:17,923][261293] Avg episode reward: [(0, '1717.704')]
+[2026-06-06 22:11:18,457][262198] Updated weights for policy 0, policy_version 38637 (0.0009)
+[2026-06-06 22:11:19,619][262198] Updated weights for policy 0, policy_version 38647 (0.0010)
+[2026-06-06 22:11:21,429][262198] Updated weights for policy 0, policy_version 38658 (0.0009)
+[2026-06-06 22:11:22,417][262198] Updated weights for policy 0, policy_version 38668 (0.0009)
+[2026-06-06 22:11:22,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3345.0, 300 sec: 3179.6). Total num frames: 19798016. Throughput: 0: 3439.4. Samples: 19802638. Policy #0 lag: (min: 4.0, avg: 19.0, max: 36.0)
+[2026-06-06 22:11:22,924][261293] Avg episode reward: [(0, '1625.237')]
+[2026-06-06 22:11:24,048][262198] Updated weights for policy 0, policy_version 38678 (0.0009)
+[2026-06-06 22:11:25,340][262198] Updated weights for policy 0, policy_version 38688 (0.0009)
+[2026-06-06 22:11:27,166][262198] Updated weights for policy 0, policy_version 38700 (0.0010)
+[2026-06-06 22:11:27,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3413.3, 300 sec: 3200.4). Total num frames: 19816448. Throughput: 0: 3423.2. Samples: 19823358. Policy #0 lag: (min: 4.0, avg: 19.0, max: 36.0)
+[2026-06-06 22:11:27,923][261293] Avg episode reward: [(0, '1584.017')]
+[2026-06-06 22:11:28,698][262198] Updated weights for policy 0, policy_version 38710 (0.0009)
+[2026-06-06 22:11:29,877][262198] Updated weights for policy 0, policy_version 38720 (0.0009)
+[2026-06-06 22:11:31,520][262198] Updated weights for policy 0, policy_version 38730 (0.0012)
+[2026-06-06 22:11:32,921][261293] Fps is (10 sec: 3481.7, 60 sec: 3447.5, 300 sec: 3214.3). Total num frames: 19832832. Throughput: 0: 3473.0. Samples: 19834862. Policy #0 lag: (min: 3.0, avg: 19.4, max: 40.0)
+[2026-06-06 22:11:32,924][261293] Avg episode reward: [(0, '1575.812')]
+[2026-06-06 22:11:32,948][262198] Updated weights for policy 0, policy_version 38740 (0.0008)
+[2026-06-06 22:11:34,462][262198] Updated weights for policy 0, policy_version 38750 (0.0010)
+[2026-06-06 22:11:35,540][262198] Updated weights for policy 0, policy_version 38760 (0.0010)
+[2026-06-06 22:11:37,304][262198] Updated weights for policy 0, policy_version 38770 (0.0009)
+[2026-06-06 22:11:37,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3481.6, 300 sec: 3242.1). Total num frames: 19853312. Throughput: 0: 3500.3. Samples: 19856310. Policy #0 lag: (min: 3.0, avg: 19.4, max: 40.0)
+[2026-06-06 22:11:37,924][261293] Avg episode reward: [(0, '1754.344')]
+[2026-06-06 22:11:38,760][262198] Updated weights for policy 0, policy_version 38780 (0.0008)
+[2026-06-06 22:11:40,464][262198] Updated weights for policy 0, policy_version 38792 (0.0009)
+[2026-06-06 22:11:41,947][262198] Updated weights for policy 0, policy_version 38802 (0.0008)
+[2026-06-06 22:11:42,921][261293] Fps is (10 sec: 3686.5, 60 sec: 3481.6, 300 sec: 3249.0). Total num frames: 19869696. Throughput: 0: 3513.9. Samples: 19877374. Policy #0 lag: (min: 3.0, avg: 19.4, max: 40.0)
+[2026-06-06 22:11:42,924][261293] Avg episode reward: [(0, '1711.931')]
+[2026-06-06 22:11:43,415][262198] Updated weights for policy 0, policy_version 38812 (0.0009)
+[2026-06-06 22:11:44,893][262198] Updated weights for policy 0, policy_version 38822 (0.0011)
+[2026-06-06 22:11:46,636][262198] Updated weights for policy 0, policy_version 38832 (0.0011)
+[2026-06-06 22:11:47,925][261293] Fps is (10 sec: 3275.5, 60 sec: 3481.4, 300 sec: 3269.8). Total num frames: 19886080. Throughput: 0: 3503.9. Samples: 19886960. Policy #0 lag: (min: 3.0, avg: 19.4, max: 40.0)
+[2026-06-06 22:11:47,928][261293] Avg episode reward: [(0, '1691.546')]
+[2026-06-06 22:11:48,812][262198] Updated weights for policy 0, policy_version 38842 (0.0011)
+[2026-06-06 22:11:50,443][262198] Updated weights for policy 0, policy_version 38852 (0.0009)
+[2026-06-06 22:11:52,260][262198] Updated weights for policy 0, policy_version 38862 (0.0011)
+[2026-06-06 22:11:52,921][261293] Fps is (10 sec: 3071.9, 60 sec: 3447.5, 300 sec: 3269.9). Total num frames: 19900416. Throughput: 0: 3443.1. Samples: 19904350. Policy #0 lag: (min: 3.0, avg: 19.4, max: 40.0)
+[2026-06-06 22:11:52,924][261293] Avg episode reward: [(0, '1809.414')]
+[2026-06-06 22:11:53,603][262198] Updated weights for policy 0, policy_version 38872 (0.0011)
+[2026-06-06 22:11:56,008][262198] Updated weights for policy 0, policy_version 38882 (0.0011)
+[2026-06-06 22:11:57,307][262198] Updated weights for policy 0, policy_version 38892 (0.0012)
+[2026-06-06 22:11:57,924][261293] Fps is (10 sec: 2662.7, 60 sec: 3379.1, 300 sec: 3269.8). Total num frames: 19912704. Throughput: 0: 3356.8. Samples: 19921548. Policy #0 lag: (min: 6.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:11:57,929][261293] Avg episode reward: [(0, '1677.847')]
+[2026-06-06 22:11:59,213][262198] Updated weights for policy 0, policy_version 38902 (0.0011)
+[2026-06-06 22:12:00,665][262198] Updated weights for policy 0, policy_version 38912 (0.0010)
+[2026-06-06 22:12:02,569][262198] Updated weights for policy 0, policy_version 38922 (0.0010)
+[2026-06-06 22:12:02,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3379.2, 300 sec: 3283.7). Total num frames: 19929088. Throughput: 0: 3306.9. Samples: 19930118. Policy #0 lag: (min: 6.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:12:02,926][261293] Avg episode reward: [(0, '1686.049')]
+[2026-06-06 22:12:03,823][262198] Updated weights for policy 0, policy_version 38932 (0.0008)
+[2026-06-06 22:12:05,285][262198] Updated weights for policy 0, policy_version 38942 (0.0010)
+[2026-06-06 22:12:06,432][262198] Updated weights for policy 0, policy_version 38952 (0.0008)
+[2026-06-06 22:12:07,922][261293] Fps is (10 sec: 3482.3, 60 sec: 3413.3, 300 sec: 3297.6). Total num frames: 19947520. Throughput: 0: 3317.0. Samples: 19951906. Policy #0 lag: (min: 6.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:12:07,927][261293] Avg episode reward: [(0, '1806.398')]
+[2026-06-06 22:12:08,018][262198] Updated weights for policy 0, policy_version 38962 (0.0011)
+[2026-06-06 22:12:09,400][262198] Updated weights for policy 0, policy_version 38972 (0.0008)
+[2026-06-06 22:12:10,813][262198] Updated weights for policy 0, policy_version 38982 (0.0009)
+[2026-06-06 22:12:12,236][262198] Updated weights for policy 0, policy_version 38992 (0.0013)
+[2026-06-06 22:12:12,920][261293] Fps is (10 sec: 3686.6, 60 sec: 3379.2, 300 sec: 3318.5). Total num frames: 19965952. Throughput: 0: 3331.7. Samples: 19973282. Policy #0 lag: (min: 6.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:12:12,924][261293] Avg episode reward: [(0, '1710.910')]
+[2026-06-06 22:12:14,583][262198] Updated weights for policy 0, policy_version 39002 (0.0012)
+[2026-06-06 22:12:16,129][262198] Updated weights for policy 0, policy_version 39012 (0.0011)
+[2026-06-06 22:12:17,920][261293] Fps is (10 sec: 3072.4, 60 sec: 3276.8, 300 sec: 3311.5). Total num frames: 19978240. Throughput: 0: 3249.3. Samples: 19981080. Policy #0 lag: (min: 6.0, avg: 20.0, max: 39.0)
+[2026-06-06 22:12:17,923][261293] Avg episode reward: [(0, '1614.228')]
+[2026-06-06 22:12:18,087][262198] Updated weights for policy 0, policy_version 39022 (0.0011)
+[2026-06-06 22:12:19,123][262198] Updated weights for policy 0, policy_version 39032 (0.0012)
+[2026-06-06 22:12:21,842][262198] Updated weights for policy 0, policy_version 39042 (0.0013)
+[2026-06-06 22:12:22,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3242.7, 300 sec: 3318.5). Total num frames: 19992576. Throughput: 0: 3153.7. Samples: 19998224. Policy #0 lag: (min: 4.0, avg: 17.2, max: 34.0)
+[2026-06-06 22:12:22,923][261293] Avg episode reward: [(0, '1721.079')]
+[2026-06-06 22:12:23,129][262198] Updated weights for policy 0, policy_version 39052 (0.0012)
+[2026-06-06 22:12:25,063][262198] Updated weights for policy 0, policy_version 39062 (0.0011)
+[2026-06-06 22:12:27,002][262198] Updated weights for policy 0, policy_version 39072 (0.0012)
+[2026-06-06 22:12:27,921][261293] Fps is (10 sec: 2867.0, 60 sec: 3174.4, 300 sec: 3318.4). Total num frames: 20006912. Throughput: 0: 3059.5. Samples: 20015052. Policy #0 lag: (min: 4.0, avg: 17.2, max: 34.0)
+[2026-06-06 22:12:27,924][261293] Avg episode reward: [(0, '1727.095')]
+[2026-06-06 22:12:28,837][262198] Updated weights for policy 0, policy_version 39082 (0.0011)
+[2026-06-06 22:12:30,277][262198] Updated weights for policy 0, policy_version 39092 (0.0013)
+[2026-06-06 22:12:32,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3106.1, 300 sec: 3318.5). Total num frames: 20019200. Throughput: 0: 3037.2. Samples: 20023622. Policy #0 lag: (min: 4.0, avg: 17.2, max: 34.0)
+[2026-06-06 22:12:32,922][261293] Avg episode reward: [(0, '1866.185')]
+[2026-06-06 22:12:32,992][262198] Updated weights for policy 0, policy_version 39102 (0.0013)
+[2026-06-06 22:12:35,301][262198] Updated weights for policy 0, policy_version 39113 (0.0024)
+[2026-06-06 22:12:37,003][262198] Updated weights for policy 0, policy_version 39124 (0.0012)
+[2026-06-06 22:12:37,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3003.7, 300 sec: 3325.4). Total num frames: 20033536. Throughput: 0: 2971.5. Samples: 20038068. Policy #0 lag: (min: 4.0, avg: 17.2, max: 34.0)
+[2026-06-06 22:12:37,926][261293] Avg episode reward: [(0, '1574.654')]
+[2026-06-06 22:12:39,711][262198] Updated weights for policy 0, policy_version 39134 (0.0011)
+[2026-06-06 22:12:40,832][262198] Updated weights for policy 0, policy_version 39144 (0.0013)
+[2026-06-06 22:12:42,922][261293] Fps is (10 sec: 2662.3, 60 sec: 2935.4, 300 sec: 3318.4). Total num frames: 20045824. Throughput: 0: 2952.0. Samples: 20054380. Policy #0 lag: (min: 4.0, avg: 17.2, max: 34.0)
+[2026-06-06 22:12:42,925][261293] Avg episode reward: [(0, '1645.168')]
+[2026-06-06 22:12:42,940][262198] Updated weights for policy 0, policy_version 39154 (0.0011)
+[2026-06-06 22:12:44,961][262198] Updated weights for policy 0, policy_version 39164 (0.0013)
+[2026-06-06 22:12:46,994][262198] Updated weights for policy 0, policy_version 39174 (0.0010)
+[2026-06-06 22:12:47,921][261293] Fps is (10 sec: 2662.5, 60 sec: 2901.5, 300 sec: 3304.6). Total num frames: 20060160. Throughput: 0: 2924.9. Samples: 20061736. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:12:47,923][261293] Avg episode reward: [(0, '1864.486')]
+[2026-06-06 22:12:48,965][262198] Updated weights for policy 0, policy_version 39184 (0.0012)
+[2026-06-06 22:12:50,866][262198] Updated weights for policy 0, policy_version 39194 (0.0011)
+[2026-06-06 22:12:52,537][262198] Updated weights for policy 0, policy_version 39204 (0.0013)
+[2026-06-06 22:12:52,921][261293] Fps is (10 sec: 2662.5, 60 sec: 2867.2, 300 sec: 3290.7). Total num frames: 20072448. Throughput: 0: 2807.4. Samples: 20078238. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:12:52,924][261293] Avg episode reward: [(0, '1778.944')]
+[2026-06-06 22:12:54,597][262198] Updated weights for policy 0, policy_version 39214 (0.0012)
+[2026-06-06 22:12:55,744][262198] Updated weights for policy 0, policy_version 39224 (0.0012)
+[2026-06-06 22:12:57,920][261293] Fps is (10 sec: 2662.5, 60 sec: 2901.5, 300 sec: 3283.7). Total num frames: 20086784. Throughput: 0: 2713.4. Samples: 20095384. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:12:57,922][261293] Avg episode reward: [(0, '1793.122')]
+[2026-06-06 22:12:58,171][262198] Updated weights for policy 0, policy_version 39234 (0.0011)
+[2026-06-06 22:12:59,488][262198] Updated weights for policy 0, policy_version 39244 (0.0012)
+[2026-06-06 22:13:01,505][262198] Updated weights for policy 0, policy_version 39254 (0.0011)
+[2026-06-06 22:13:02,920][261293] Fps is (10 sec: 2867.2, 60 sec: 2867.2, 300 sec: 3276.8). Total num frames: 20101120. Throughput: 0: 2719.3. Samples: 20103448. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:13:02,923][261293] Avg episode reward: [(0, '1808.100')]
+[2026-06-06 22:13:03,993][262198] Updated weights for policy 0, policy_version 39265 (0.0011)
+[2026-06-06 22:13:05,468][262198] Updated weights for policy 0, policy_version 39275 (0.0011)
+[2026-06-06 22:13:07,476][262198] Updated weights for policy 0, policy_version 39285 (0.0012)
+[2026-06-06 22:13:07,922][261293] Fps is (10 sec: 2867.1, 60 sec: 2799.0, 300 sec: 3269.9). Total num frames: 20115456. Throughput: 0: 2719.4. Samples: 20120596. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:13:07,926][261293] Avg episode reward: [(0, '1714.429')]
+[2026-06-06 22:13:09,841][262198] Updated weights for policy 0, policy_version 39295 (0.0011)
+[2026-06-06 22:13:11,847][262198] Updated weights for policy 0, policy_version 39305 (0.0013)
+[2026-06-06 22:13:12,920][261293] Fps is (10 sec: 2457.6, 60 sec: 2662.4, 300 sec: 3249.0). Total num frames: 20125696. Throughput: 0: 2661.2. Samples: 20134804. Policy #0 lag: (min: 3.0, avg: 19.9, max: 38.0)
+[2026-06-06 22:13:12,922][261293] Avg episode reward: [(0, '1755.666')]
+[2026-06-06 22:13:13,683][262198] Updated weights for policy 0, policy_version 39315 (0.0012)
+[2026-06-06 22:13:16,218][262198] Updated weights for policy 0, policy_version 39325 (0.0011)
+[2026-06-06 22:13:17,398][262198] Updated weights for policy 0, policy_version 39335 (0.0012)
+[2026-06-06 22:13:17,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2696.5, 300 sec: 3242.1). Total num frames: 20140032. Throughput: 0: 2648.4. Samples: 20142802. Policy #0 lag: (min: 1.0, avg: 16.5, max: 33.0)
+[2026-06-06 22:13:17,923][261293] Avg episode reward: [(0, '1705.137')]
+[2026-06-06 22:13:19,815][262198] Updated weights for policy 0, policy_version 39345 (0.0012)
+[2026-06-06 22:13:21,688][262198] Updated weights for policy 0, policy_version 39356 (0.0012)
+[2026-06-06 22:13:22,920][261293] Fps is (10 sec: 2662.5, 60 sec: 2662.4, 300 sec: 3221.3). Total num frames: 20152320. Throughput: 0: 2674.4. Samples: 20158412. Policy #0 lag: (min: 1.0, avg: 16.5, max: 33.0)
+[2026-06-06 22:13:22,921][261293] Avg episode reward: [(0, '1551.979')]
+[2026-06-06 22:13:23,463][262198] Updated weights for policy 0, policy_version 39366 (0.0012)
+[2026-06-06 22:13:25,325][262198] Updated weights for policy 0, policy_version 39376 (0.0011)
+[2026-06-06 22:13:27,522][262198] Updated weights for policy 0, policy_version 39386 (0.0012)
+[2026-06-06 22:13:27,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2662.4, 300 sec: 3214.3). Total num frames: 20166656. Throughput: 0: 2680.0. Samples: 20174980. Policy #0 lag: (min: 1.0, avg: 16.5, max: 33.0)
+[2026-06-06 22:13:27,922][261293] Avg episode reward: [(0, '1663.241')]
+[2026-06-06 22:13:29,093][262198] Updated weights for policy 0, policy_version 39396 (0.0010)
+[2026-06-06 22:13:30,757][262198] Updated weights for policy 0, policy_version 39406 (0.0010)
+[2026-06-06 22:13:31,617][262198] Updated weights for policy 0, policy_version 39416 (0.0009)
+[2026-06-06 22:13:32,921][261293] Fps is (10 sec: 3071.9, 60 sec: 2730.7, 300 sec: 3214.3). Total num frames: 20183040. Throughput: 0: 2726.7. Samples: 20184436. Policy #0 lag: (min: 1.0, avg: 16.5, max: 33.0)
+[2026-06-06 22:13:32,925][261293] Avg episode reward: [(0, '1675.641')]
+[2026-06-06 22:13:33,492][262198] Updated weights for policy 0, policy_version 39426 (0.0010)
+[2026-06-06 22:13:34,538][262198] Updated weights for policy 0, policy_version 39436 (0.0009)
+[2026-06-06 22:13:36,037][262198] Updated weights for policy 0, policy_version 39446 (0.0010)
+[2026-06-06 22:13:37,630][262198] Updated weights for policy 0, policy_version 39456 (0.0010)
+[2026-06-06 22:13:37,921][261293] Fps is (10 sec: 3481.5, 60 sec: 2798.9, 300 sec: 3221.3). Total num frames: 20201472. Throughput: 0: 2850.0. Samples: 20206488. Policy #0 lag: (min: 1.0, avg: 16.5, max: 33.0)
+[2026-06-06 22:13:37,923][261293] Avg episode reward: [(0, '1746.635')]
+[2026-06-06 22:13:39,228][262198] Updated weights for policy 0, policy_version 39466 (0.0010)
+[2026-06-06 22:13:40,653][262198] Updated weights for policy 0, policy_version 39476 (0.0010)
+[2026-06-06 22:13:42,373][262198] Updated weights for policy 0, policy_version 39486 (0.0008)
+[2026-06-06 22:13:42,922][261293] Fps is (10 sec: 3686.1, 60 sec: 2901.3, 300 sec: 3228.2). Total num frames: 20219904. Throughput: 0: 2937.5. Samples: 20227574. Policy #0 lag: (min: 3.0, avg: 19.7, max: 39.0)
+[2026-06-06 22:13:42,925][261293] Avg episode reward: [(0, '1664.826')]
+[2026-06-06 22:13:43,151][262198] Updated weights for policy 0, policy_version 39496 (0.0008)
+[2026-06-06 22:13:44,647][262198] Updated weights for policy 0, policy_version 39506 (0.0008)
+[2026-06-06 22:13:45,992][262198] Updated weights for policy 0, policy_version 39516 (0.0010)
+[2026-06-06 22:13:47,584][262198] Updated weights for policy 0, policy_version 39527 (0.0009)
+[2026-06-06 22:13:47,921][261293] Fps is (10 sec: 3686.4, 60 sec: 2969.6, 300 sec: 3235.1). Total num frames: 20238336. Throughput: 0: 3010.1. Samples: 20238904. Policy #0 lag: (min: 3.0, avg: 19.7, max: 39.0)
+[2026-06-06 22:13:47,923][261293] Avg episode reward: [(0, '1757.262')]
+[2026-06-06 22:13:49,706][262198] Updated weights for policy 0, policy_version 39540 (0.0009)
+[2026-06-06 22:13:51,557][262198] Updated weights for policy 0, policy_version 39550 (0.0012)
+[2026-06-06 22:13:52,293][262198] Updated weights for policy 0, policy_version 39560 (0.0007)
+[2026-06-06 22:13:52,921][261293] Fps is (10 sec: 3686.8, 60 sec: 3072.0, 300 sec: 3249.0). Total num frames: 20256768. Throughput: 0: 3105.1. Samples: 20260326. Policy #0 lag: (min: 3.0, avg: 19.7, max: 39.0)
+[2026-06-06 22:13:52,923][261293] Avg episode reward: [(0, '1624.467')]
+[2026-06-06 22:13:53,840][262198] Updated weights for policy 0, policy_version 39570 (0.0010)
+[2026-06-06 22:13:55,232][262198] Updated weights for policy 0, policy_version 39580 (0.0010)
+[2026-06-06 22:13:57,042][262198] Updated weights for policy 0, policy_version 39591 (0.0010)
+[2026-06-06 22:13:57,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3106.1, 300 sec: 3249.0). Total num frames: 20273152. Throughput: 0: 3232.4. Samples: 20280262. Policy #0 lag: (min: 3.0, avg: 19.7, max: 39.0)
+[2026-06-06 22:13:57,923][261293] Avg episode reward: [(0, '1772.433')]
+[2026-06-06 22:13:58,933][262198] Updated weights for policy 0, policy_version 39601 (0.0008)
+[2026-06-06 22:14:00,123][262198] Updated weights for policy 0, policy_version 39612 (0.0010)
+[2026-06-06 22:14:01,441][262115] Signal inference workers to stop experience collection... (3950 times)
+[2026-06-06 22:14:01,447][262115] Signal inference workers to resume experience collection... (3950 times)
+[2026-06-06 22:14:01,480][262198] InferenceWorker_p0-w0: stopping experience collection (3950 times)
+[2026-06-06 22:14:01,480][262198] InferenceWorker_p0-w0: resuming experience collection (3950 times)
+[2026-06-06 22:14:01,592][262198] Updated weights for policy 0, policy_version 39622 (0.0008)
+[2026-06-06 22:14:02,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3140.3, 300 sec: 3249.0). Total num frames: 20289536. Throughput: 0: 3298.7. Samples: 20291244. Policy #0 lag: (min: 3.0, avg: 19.7, max: 39.0)
+[2026-06-06 22:14:02,923][261293] Avg episode reward: [(0, '1620.370')]
+[2026-06-06 22:14:03,040][262198] Updated weights for policy 0, policy_version 39632 (0.0009)
+[2026-06-06 22:14:04,491][262198] Updated weights for policy 0, policy_version 39643 (0.0011)
+[2026-06-06 22:14:06,192][262198] Updated weights for policy 0, policy_version 39655 (0.0011)
+[2026-06-06 22:14:07,924][261293] Fps is (10 sec: 3480.6, 60 sec: 3208.4, 300 sec: 3255.9). Total num frames: 20307968. Throughput: 0: 3418.0. Samples: 20312232. Policy #0 lag: (min: 7.0, avg: 22.3, max: 39.0)
+[2026-06-06 22:14:07,931][261293] Avg episode reward: [(0, '1712.980')]
+[2026-06-06 22:14:08,393][262198] Updated weights for policy 0, policy_version 39665 (0.0009)
+[2026-06-06 22:14:10,506][262198] Updated weights for policy 0, policy_version 39677 (0.0009)
+[2026-06-06 22:14:11,717][262198] Updated weights for policy 0, policy_version 39691 (0.0011)
+[2026-06-06 22:14:12,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3345.1, 300 sec: 3256.0). Total num frames: 20326400. Throughput: 0: 3522.6. Samples: 20333496. Policy #0 lag: (min: 7.0, avg: 22.3, max: 39.0)
+[2026-06-06 22:14:12,924][261293] Avg episode reward: [(0, '1724.114')]
+[2026-06-06 22:14:12,966][262198] Updated weights for policy 0, policy_version 39701 (0.0009)
+[2026-06-06 22:14:14,691][262198] Updated weights for policy 0, policy_version 39711 (0.0010)
+[2026-06-06 22:14:16,103][262198] Updated weights for policy 0, policy_version 39721 (0.0008)
+[2026-06-06 22:14:17,100][262198] Updated weights for policy 0, policy_version 39731 (0.0010)
+[2026-06-06 22:14:17,920][261293] Fps is (10 sec: 3687.5, 60 sec: 3413.3, 300 sec: 3262.9). Total num frames: 20344832. Throughput: 0: 3561.9. Samples: 20344722. Policy #0 lag: (min: 7.0, avg: 22.3, max: 39.0)
+[2026-06-06 22:14:17,922][261293] Avg episode reward: [(0, '1622.213')]
+[2026-06-06 22:14:18,950][262198] Updated weights for policy 0, policy_version 39741 (0.0008)
+[2026-06-06 22:14:19,778][262198] Updated weights for policy 0, policy_version 39751 (0.0009)
+[2026-06-06 22:14:21,386][262198] Updated weights for policy 0, policy_version 39761 (0.0009)
+[2026-06-06 22:14:22,575][262198] Updated weights for policy 0, policy_version 39772 (0.0010)
+[2026-06-06 22:14:22,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3515.7, 300 sec: 3262.9). Total num frames: 20363264. Throughput: 0: 3585.9. Samples: 20367854. Policy #0 lag: (min: 7.0, avg: 22.3, max: 39.0)
+[2026-06-06 22:14:22,923][261293] Avg episode reward: [(0, '1937.002')]
+[2026-06-06 22:14:22,930][262115] Saving new best policy, reward=1937.002!
+[2026-06-06 22:14:24,559][262198] Updated weights for policy 0, policy_version 39785 (0.0009)
+[2026-06-06 22:14:25,789][262198] Updated weights for policy 0, policy_version 39796 (0.0008)
+[2026-06-06 22:14:27,356][262198] Updated weights for policy 0, policy_version 39806 (0.0007)
+[2026-06-06 22:14:27,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3584.0, 300 sec: 3256.0). Total num frames: 20381696. Throughput: 0: 3634.8. Samples: 20391136. Policy #0 lag: (min: 7.0, avg: 22.3, max: 39.0)
+[2026-06-06 22:14:27,925][261293] Avg episode reward: [(0, '1720.338')]
+[2026-06-06 22:14:28,219][262198] Updated weights for policy 0, policy_version 39816 (0.0009)
+[2026-06-06 22:14:29,732][262198] Updated weights for policy 0, policy_version 39826 (0.0008)
+[2026-06-06 22:14:31,398][262198] Updated weights for policy 0, policy_version 39837 (0.0008)
+[2026-06-06 22:14:32,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3652.3, 300 sec: 3262.9). Total num frames: 20402176. Throughput: 0: 3650.2. Samples: 20403162. Policy #0 lag: (min: 4.0, avg: 20.2, max: 35.0)
+[2026-06-06 22:14:32,924][261293] Avg episode reward: [(0, '1711.242')]
+[2026-06-06 22:14:33,289][262198] Updated weights for policy 0, policy_version 39849 (0.0014)
+[2026-06-06 22:14:35,160][262198] Updated weights for policy 0, policy_version 39859 (0.0018)
+[2026-06-06 22:14:37,920][261293] Fps is (10 sec: 3072.1, 60 sec: 3515.8, 300 sec: 3235.1). Total num frames: 20412416. Throughput: 0: 3536.8. Samples: 20419484. Policy #0 lag: (min: 4.0, avg: 20.2, max: 35.0)
+[2026-06-06 22:14:37,922][261293] Avg episode reward: [(0, '1650.581')]
+[2026-06-06 22:14:38,187][262198] Updated weights for policy 0, policy_version 39869 (0.0014)
+[2026-06-06 22:14:40,959][262198] Updated weights for policy 0, policy_version 39881 (0.0017)
+[2026-06-06 22:14:42,811][262198] Updated weights for policy 0, policy_version 39891 (0.0013)
+[2026-06-06 22:14:42,921][261293] Fps is (10 sec: 2252.7, 60 sec: 3413.4, 300 sec: 3214.3). Total num frames: 20424704. Throughput: 0: 3376.8. Samples: 20432220. Policy #0 lag: (min: 4.0, avg: 20.2, max: 35.0)
+[2026-06-06 22:14:42,923][261293] Avg episode reward: [(0, '1861.557')]
+[2026-06-06 22:14:45,713][262198] Updated weights for policy 0, policy_version 39901 (0.0014)
+[2026-06-06 22:14:46,656][262198] Updated weights for policy 0, policy_version 39911 (0.0012)
+[2026-06-06 22:14:47,921][261293] Fps is (10 sec: 2252.6, 60 sec: 3276.8, 300 sec: 3179.6). Total num frames: 20434944. Throughput: 0: 3289.2. Samples: 20439258. Policy #0 lag: (min: 4.0, avg: 20.2, max: 35.0)
+[2026-06-06 22:14:47,924][261293] Avg episode reward: [(0, '1700.834')]
+[2026-06-06 22:14:49,621][262198] Updated weights for policy 0, policy_version 39921 (0.0015)
+[2026-06-06 22:14:51,606][262198] Updated weights for policy 0, policy_version 39931 (0.0013)
+[2026-06-06 22:14:52,921][261293] Fps is (10 sec: 2048.1, 60 sec: 3140.3, 300 sec: 3151.8). Total num frames: 20445184. Throughput: 0: 3131.5. Samples: 20453142. Policy #0 lag: (min: 4.0, avg: 20.2, max: 35.0)
+[2026-06-06 22:14:52,923][261293] Avg episode reward: [(0, '1764.276')]
+[2026-06-06 22:14:53,828][262198] Updated weights for policy 0, policy_version 39941 (0.0014)
+[2026-06-06 22:14:56,347][262198] Updated weights for policy 0, policy_version 39951 (0.0016)
+[2026-06-06 22:14:57,921][261293] Fps is (10 sec: 2457.7, 60 sec: 3106.1, 300 sec: 3131.0). Total num frames: 20459520. Throughput: 0: 2951.8. Samples: 20466326. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 22:14:57,923][261293] Avg episode reward: [(0, '1788.786')]
+[2026-06-06 22:14:58,941][262198] Updated weights for policy 0, policy_version 39961 (0.0014)
+[2026-06-06 22:15:00,716][262198] Updated weights for policy 0, policy_version 39972 (0.0012)
+[2026-06-06 22:15:02,921][261293] Fps is (10 sec: 2457.4, 60 sec: 3003.7, 300 sec: 3103.2). Total num frames: 20469760. Throughput: 0: 2851.6. Samples: 20473046. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 22:15:02,923][261293] Avg episode reward: [(0, '1705.900')]
+[2026-06-06 22:15:03,674][262198] Updated weights for policy 0, policy_version 39982 (0.0015)
+[2026-06-06 22:15:05,024][262198] Updated weights for policy 0, policy_version 39992 (0.0013)
+[2026-06-06 22:15:07,925][261293] Fps is (10 sec: 2047.1, 60 sec: 2867.1, 300 sec: 3089.3). Total num frames: 20480000. Throughput: 0: 2628.6. Samples: 20486154. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 22:15:07,928][261293] Avg episode reward: [(0, '1772.041')]
+[2026-06-06 22:15:08,196][262198] Updated weights for policy 0, policy_version 40002 (0.0014)
+[2026-06-06 22:15:11,596][262198] Updated weights for policy 0, policy_version 40013 (0.0015)
+[2026-06-06 22:15:12,921][261293] Fps is (10 sec: 2048.1, 60 sec: 2730.6, 300 sec: 3075.5). Total num frames: 20490240. Throughput: 0: 2376.0. Samples: 20498058. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 22:15:12,923][261293] Avg episode reward: [(0, '1573.161')]
+[2026-06-06 22:15:13,069][262198] Updated weights for policy 0, policy_version 40023 (0.0013)
+[2026-06-06 22:15:15,895][262198] Updated weights for policy 0, policy_version 40033 (0.0013)
+[2026-06-06 22:15:17,920][261293] Fps is (10 sec: 2048.9, 60 sec: 2594.1, 300 sec: 3061.6). Total num frames: 20500480. Throughput: 0: 2268.1. Samples: 20505226. Policy #0 lag: (min: 3.0, avg: 17.9, max: 35.0)
+[2026-06-06 22:15:17,921][261293] Avg episode reward: [(0, '1638.547')]
+[2026-06-06 22:15:18,000][262198] Updated weights for policy 0, policy_version 40043 (0.0015)
+[2026-06-06 22:15:20,047][262198] Updated weights for policy 0, policy_version 40053 (0.0013)
+[2026-06-06 22:15:22,921][261293] Fps is (10 sec: 2048.0, 60 sec: 2457.6, 300 sec: 3047.7). Total num frames: 20510720. Throughput: 0: 2202.6. Samples: 20518602. Policy #0 lag: (min: 1.0, avg: 16.4, max: 33.0)
+[2026-06-06 22:15:22,924][261293] Avg episode reward: [(0, '1668.192')]
+[2026-06-06 22:15:22,995][262198] Updated weights for policy 0, policy_version 40063 (0.0012)
+[2026-06-06 22:15:25,443][262198] Updated weights for policy 0, policy_version 40073 (0.0011)
+[2026-06-06 22:15:27,163][262198] Updated weights for policy 0, policy_version 40083 (0.0012)
+[2026-06-06 22:15:27,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2389.3, 300 sec: 3047.7). Total num frames: 20525056. Throughput: 0: 2211.2. Samples: 20531722. Policy #0 lag: (min: 1.0, avg: 16.4, max: 33.0)
+[2026-06-06 22:15:27,923][261293] Avg episode reward: [(0, '1745.188')]
+[2026-06-06 22:15:30,379][262198] Updated weights for policy 0, policy_version 40093 (0.0014)
+[2026-06-06 22:15:31,476][262198] Updated weights for policy 0, policy_version 40103 (0.0014)
+[2026-06-06 22:15:32,921][261293] Fps is (10 sec: 2252.9, 60 sec: 2184.5, 300 sec: 3013.0). Total num frames: 20533248. Throughput: 0: 2201.3. Samples: 20538314. Policy #0 lag: (min: 1.0, avg: 16.4, max: 33.0)
+[2026-06-06 22:15:32,923][261293] Avg episode reward: [(0, '1706.970')]
+[2026-06-06 22:15:34,451][262198] Updated weights for policy 0, policy_version 40113 (0.0013)
+[2026-06-06 22:15:36,319][262198] Updated weights for policy 0, policy_version 40123 (0.0014)
+[2026-06-06 22:15:37,920][261293] Fps is (10 sec: 2048.0, 60 sec: 2218.7, 300 sec: 2999.1). Total num frames: 20545536. Throughput: 0: 2195.6. Samples: 20551944. Policy #0 lag: (min: 1.0, avg: 16.4, max: 33.0)
+[2026-06-06 22:15:37,922][261293] Avg episode reward: [(0, '1571.451')]
+[2026-06-06 22:15:38,689][262198] Updated weights for policy 0, policy_version 40133 (0.0015)
+[2026-06-06 22:15:41,315][262198] Updated weights for policy 0, policy_version 40143 (0.0013)
+[2026-06-06 22:15:42,920][261293] Fps is (10 sec: 2457.7, 60 sec: 2218.7, 300 sec: 2985.2). Total num frames: 20557824. Throughput: 0: 2191.3. Samples: 20564936. Policy #0 lag: (min: 1.0, avg: 16.4, max: 33.0)
+[2026-06-06 22:15:42,922][261293] Avg episode reward: [(0, '1845.728')]
+[2026-06-06 22:15:43,712][262198] Updated weights for policy 0, policy_version 40153 (0.0013)
+[2026-06-06 22:15:45,530][262198] Updated weights for policy 0, policy_version 40164 (0.0013)
+[2026-06-06 22:15:47,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2218.7, 300 sec: 2964.4). Total num frames: 20568064. Throughput: 0: 2196.6. Samples: 20571890. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:15:47,924][261293] Avg episode reward: [(0, '1700.502')]
+[2026-06-06 22:15:48,302][262198] Updated weights for policy 0, policy_version 40174 (0.0015)
+[2026-06-06 22:15:49,601][262198] Updated weights for policy 0, policy_version 40184 (0.0012)
+[2026-06-06 22:15:52,537][262198] Updated weights for policy 0, policy_version 40194 (0.0013)
+[2026-06-06 22:15:52,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2252.8, 300 sec: 2950.5). Total num frames: 20580352. Throughput: 0: 2212.7. Samples: 20585714. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:15:52,922][261293] Avg episode reward: [(0, '1718.781')]
+[2026-06-06 22:15:55,291][262198] Updated weights for policy 0, policy_version 40205 (0.0013)
+[2026-06-06 22:15:56,567][262198] Updated weights for policy 0, policy_version 40215 (0.0011)
+[2026-06-06 22:15:57,921][261293] Fps is (10 sec: 2252.6, 60 sec: 2184.5, 300 sec: 2929.7). Total num frames: 20590592. Throughput: 0: 2257.9. Samples: 20599666. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:15:57,924][261293] Avg episode reward: [(0, '1721.763')]
+[2026-06-06 22:15:59,556][262198] Updated weights for policy 0, policy_version 40225 (0.0012)
+[2026-06-06 22:16:01,540][262198] Updated weights for policy 0, policy_version 40235 (0.0012)
+[2026-06-06 22:16:02,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2218.7, 300 sec: 2915.8). Total num frames: 20602880. Throughput: 0: 2255.1. Samples: 20606706. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:16:02,922][261293] Avg episode reward: [(0, '1725.174')]
+[2026-06-06 22:16:03,246][262198] Updated weights for policy 0, policy_version 40245 (0.0012)
+[2026-06-06 22:16:05,782][262198] Updated weights for policy 0, policy_version 40255 (0.0013)
+[2026-06-06 22:16:07,677][262198] Updated weights for policy 0, policy_version 40265 (0.0012)
+[2026-06-06 22:16:07,921][261293] Fps is (10 sec: 2662.5, 60 sec: 2287.1, 300 sec: 2895.0). Total num frames: 20617216. Throughput: 0: 2307.7. Samples: 20622446. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:16:07,926][261293] Avg episode reward: [(0, '1671.252')]
+[2026-06-06 22:16:09,238][262198] Updated weights for policy 0, policy_version 40275 (0.0012)
+[2026-06-06 22:16:11,781][262198] Updated weights for policy 0, policy_version 40285 (0.0012)
+[2026-06-06 22:16:12,684][262198] Updated weights for policy 0, policy_version 40296 (0.0012)
+[2026-06-06 22:16:12,920][261293] Fps is (10 sec: 2867.2, 60 sec: 2355.2, 300 sec: 2881.1). Total num frames: 20631552. Throughput: 0: 2371.2. Samples: 20638426. Policy #0 lag: (min: 1.0, avg: 16.0, max: 33.0)
+[2026-06-06 22:16:12,922][261293] Avg episode reward: [(0, '1836.184')]
+[2026-06-06 22:16:15,188][262198] Updated weights for policy 0, policy_version 40306 (0.0013)
+[2026-06-06 22:16:17,099][262198] Updated weights for policy 0, policy_version 40316 (0.0012)
+[2026-06-06 22:16:17,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2355.2, 300 sec: 2860.3). Total num frames: 20641792. Throughput: 0: 2400.4. Samples: 20646332. Policy #0 lag: (min: 0.0, avg: 18.4, max: 35.0)
+[2026-06-06 22:16:17,924][261293] Avg episode reward: [(0, '1622.214')]
+[2026-06-06 22:16:19,277][262198] Updated weights for policy 0, policy_version 40326 (0.0013)
+[2026-06-06 22:16:21,474][262198] Updated weights for policy 0, policy_version 40336 (0.0012)
+[2026-06-06 22:16:22,921][261293] Fps is (10 sec: 2252.7, 60 sec: 2389.4, 300 sec: 2839.4). Total num frames: 20654080. Throughput: 0: 2407.5. Samples: 20660280. Policy #0 lag: (min: 0.0, avg: 18.4, max: 35.0)
+[2026-06-06 22:16:22,923][261293] Avg episode reward: [(0, '1662.033')]
+[2026-06-06 22:16:24,055][262198] Updated weights for policy 0, policy_version 40346 (0.0012)
+[2026-06-06 22:16:26,248][262198] Updated weights for policy 0, policy_version 40358 (0.0012)
+[2026-06-06 22:16:27,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2355.2, 300 sec: 2825.5). Total num frames: 20666368. Throughput: 0: 2447.5. Samples: 20675074. Policy #0 lag: (min: 0.0, avg: 18.4, max: 35.0)
+[2026-06-06 22:16:27,923][261293] Avg episode reward: [(0, '1728.187')]
+[2026-06-06 22:16:28,245][262198] Updated weights for policy 0, policy_version 40368 (0.0010)
+[2026-06-06 22:16:30,506][262198] Updated weights for policy 0, policy_version 40378 (0.0011)
+[2026-06-06 22:16:32,512][262198] Updated weights for policy 0, policy_version 40388 (0.0013)
+[2026-06-06 22:16:32,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2423.5, 300 sec: 2797.8). Total num frames: 20678656. Throughput: 0: 2459.0. Samples: 20682544. Policy #0 lag: (min: 0.0, avg: 18.4, max: 35.0)
+[2026-06-06 22:16:32,923][261293] Avg episode reward: [(0, '1684.347')]
+[2026-06-06 22:16:35,846][262198] Updated weights for policy 0, policy_version 40398 (0.0015)
+[2026-06-06 22:16:36,923][262198] Updated weights for policy 0, policy_version 40408 (0.0010)
+[2026-06-06 22:16:37,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2389.3, 300 sec: 2776.9). Total num frames: 20688896. Throughput: 0: 2423.1. Samples: 20694752. Policy #0 lag: (min: 0.0, avg: 18.4, max: 35.0)
+[2026-06-06 22:16:37,923][261293] Avg episode reward: [(0, '1636.328')]
+[2026-06-06 22:16:39,624][262198] Updated weights for policy 0, policy_version 40419 (0.0011)
+[2026-06-06 22:16:41,438][262198] Updated weights for policy 0, policy_version 40429 (0.0010)
+[2026-06-06 22:16:42,343][262198] Updated weights for policy 0, policy_version 40440 (0.0009)
+[2026-06-06 22:16:42,921][261293] Fps is (10 sec: 2662.4, 60 sec: 2457.6, 300 sec: 2777.0). Total num frames: 20705280. Throughput: 0: 2518.9. Samples: 20713016. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 22:16:42,922][261293] Avg episode reward: [(0, '1789.074')]
+[2026-06-06 22:16:44,456][262198] Updated weights for policy 0, policy_version 40452 (0.0012)
+[2026-06-06 22:16:46,440][262198] Updated weights for policy 0, policy_version 40462 (0.0009)
+[2026-06-06 22:16:47,251][262198] Updated weights for policy 0, policy_version 40472 (0.0008)
+[2026-06-06 22:16:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 2560.0, 300 sec: 2783.9). Total num frames: 20721664. Throughput: 0: 2584.1. Samples: 20722992. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 22:16:47,924][261293] Avg episode reward: [(0, '1751.759')]
+[2026-06-06 22:16:49,280][262198] Updated weights for policy 0, policy_version 40484 (0.0010)
+[2026-06-06 22:16:50,774][262198] Updated weights for policy 0, policy_version 40494 (0.0008)
+[2026-06-06 22:16:52,431][262198] Updated weights for policy 0, policy_version 40505 (0.0010)
+[2026-06-06 22:16:52,921][261293] Fps is (10 sec: 3481.5, 60 sec: 2662.4, 300 sec: 2804.7). Total num frames: 20740096. Throughput: 0: 2714.2. Samples: 20744586. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 22:16:52,922][261293] Avg episode reward: [(0, '1757.417')]
+[2026-06-06 22:16:53,749][262198] Updated weights for policy 0, policy_version 40516 (0.0009)
+[2026-06-06 22:16:53,771][262115] Signal inference workers to stop experience collection... (4000 times)
+[2026-06-06 22:16:53,773][262115] Signal inference workers to resume experience collection... (4000 times)
+[2026-06-06 22:16:53,796][262198] InferenceWorker_p0-w0: stopping experience collection (4000 times)
+[2026-06-06 22:16:53,797][262198] InferenceWorker_p0-w0: resuming experience collection (4000 times)
+[2026-06-06 22:16:55,491][262198] Updated weights for policy 0, policy_version 40526 (0.0008)
+[2026-06-06 22:16:56,573][262198] Updated weights for policy 0, policy_version 40537 (0.0009)
+[2026-06-06 22:16:57,835][262198] Updated weights for policy 0, policy_version 40547 (0.0010)
+[2026-06-06 22:16:57,921][261293] Fps is (10 sec: 3891.3, 60 sec: 2833.1, 300 sec: 2818.6). Total num frames: 20760576. Throughput: 0: 2864.5. Samples: 20767330. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 22:16:57,924][261293] Avg episode reward: [(0, '1775.182')]
+[2026-06-06 22:16:59,558][262198] Updated weights for policy 0, policy_version 40558 (0.0010)
+[2026-06-06 22:17:01,405][262198] Updated weights for policy 0, policy_version 40569 (0.0011)
+[2026-06-06 22:17:02,713][262198] Updated weights for policy 0, policy_version 40581 (0.0009)
+[2026-06-06 22:17:02,921][261293] Fps is (10 sec: 3891.3, 60 sec: 2935.5, 300 sec: 2818.6). Total num frames: 20779008. Throughput: 0: 2918.0. Samples: 20777642. Policy #0 lag: (min: 6.0, avg: 18.5, max: 38.0)
+[2026-06-06 22:17:02,923][261293] Avg episode reward: [(0, '1762.369')]
+[2026-06-06 22:17:03,832][262198] Updated weights for policy 0, policy_version 40591 (0.0008)
+[2026-06-06 22:17:05,399][262198] Updated weights for policy 0, policy_version 40601 (0.0009)
+[2026-06-06 22:17:06,569][262198] Updated weights for policy 0, policy_version 40611 (0.0010)
+[2026-06-06 22:17:07,920][261293] Fps is (10 sec: 3481.7, 60 sec: 2969.6, 300 sec: 2811.7). Total num frames: 20795392. Throughput: 0: 3140.7. Samples: 20801610. Policy #0 lag: (min: 3.0, avg: 21.7, max: 39.0)
+[2026-06-06 22:17:07,922][261293] Avg episode reward: [(0, '1398.805')]
+[2026-06-06 22:17:08,063][262198] Updated weights for policy 0, policy_version 40621 (0.0010)
+[2026-06-06 22:17:09,532][262198] Updated weights for policy 0, policy_version 40631 (0.0009)
+[2026-06-06 22:17:11,324][262198] Updated weights for policy 0, policy_version 40642 (0.0009)
+[2026-06-06 22:17:12,634][262198] Updated weights for policy 0, policy_version 40655 (0.0012)
+[2026-06-06 22:17:12,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3072.0, 300 sec: 2839.4). Total num frames: 20815872. Throughput: 0: 3303.0. Samples: 20823710. Policy #0 lag: (min: 3.0, avg: 21.7, max: 39.0)
+[2026-06-06 22:17:12,923][261293] Avg episode reward: [(0, '1594.360')]
+[2026-06-06 22:17:14,302][262198] Updated weights for policy 0, policy_version 40665 (0.0008)
+[2026-06-06 22:17:15,464][262198] Updated weights for policy 0, policy_version 40676 (0.0010)
+[2026-06-06 22:17:17,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3140.3, 300 sec: 2839.4). Total num frames: 20830208. Throughput: 0: 3382.4. Samples: 20834752. Policy #0 lag: (min: 3.0, avg: 21.7, max: 39.0)
+[2026-06-06 22:17:17,922][261293] Avg episode reward: [(0, '1754.152')]
+[2026-06-06 22:17:18,246][262198] Updated weights for policy 0, policy_version 40687 (0.0014)
+[2026-06-06 22:17:21,632][262198] Updated weights for policy 0, policy_version 40697 (0.0014)
+[2026-06-06 22:17:22,920][261293] Fps is (10 sec: 2252.9, 60 sec: 3072.0, 300 sec: 2818.6). Total num frames: 20838400. Throughput: 0: 3367.6. Samples: 20846292. Policy #0 lag: (min: 3.0, avg: 21.7, max: 39.0)
+[2026-06-06 22:17:22,921][261293] Avg episode reward: [(0, '1769.139')]
+[2026-06-06 22:17:23,962][262198] Updated weights for policy 0, policy_version 40707 (0.0012)
+[2026-06-06 22:17:27,301][262198] Updated weights for policy 0, policy_version 40717 (0.0020)
+[2026-06-06 22:17:27,921][261293] Fps is (10 sec: 1843.2, 60 sec: 3037.9, 300 sec: 2811.7). Total num frames: 20848640. Throughput: 0: 3194.7. Samples: 20856778. Policy #0 lag: (min: 3.0, avg: 21.7, max: 39.0)
+[2026-06-06 22:17:27,923][261293] Avg episode reward: [(0, '1784.543')]
+[2026-06-06 22:17:29,681][262198] Updated weights for policy 0, policy_version 40727 (0.0013)
+[2026-06-06 22:17:32,287][262198] Updated weights for policy 0, policy_version 40737 (0.0013)
+[2026-06-06 22:17:32,920][261293] Fps is (10 sec: 2252.8, 60 sec: 3037.9, 300 sec: 2804.7). Total num frames: 20860928. Throughput: 0: 3116.7. Samples: 20863242. Policy #0 lag: (min: 6.0, avg: 20.7, max: 38.0)
+[2026-06-06 22:17:32,922][261293] Avg episode reward: [(0, '1666.051')]
+[2026-06-06 22:17:34,095][262198] Updated weights for policy 0, policy_version 40748 (0.0012)
+[2026-06-06 22:17:35,653][262198] Updated weights for policy 0, policy_version 40758 (0.0009)
+[2026-06-06 22:17:37,091][262198] Updated weights for policy 0, policy_version 40768 (0.0009)
+[2026-06-06 22:17:37,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3140.3, 300 sec: 2818.6). Total num frames: 20877312. Throughput: 0: 3023.5. Samples: 20880642. Policy #0 lag: (min: 6.0, avg: 20.7, max: 38.0)
+[2026-06-06 22:17:37,922][261293] Avg episode reward: [(0, '1657.377')]
+[2026-06-06 22:17:38,645][262198] Updated weights for policy 0, policy_version 40779 (0.0010)
+[2026-06-06 22:17:40,209][262198] Updated weights for policy 0, policy_version 40789 (0.0011)
+[2026-06-06 22:17:42,065][262198] Updated weights for policy 0, policy_version 40801 (0.0010)
+[2026-06-06 22:17:42,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3140.3, 300 sec: 2825.5). Total num frames: 20893696. Throughput: 0: 2987.6. Samples: 20901772. Policy #0 lag: (min: 6.0, avg: 20.7, max: 38.0)
+[2026-06-06 22:17:42,922][261293] Avg episode reward: [(0, '1685.899')]
+[2026-06-06 22:17:43,654][262198] Updated weights for policy 0, policy_version 40815 (0.0010)
+[2026-06-06 22:17:45,381][262198] Updated weights for policy 0, policy_version 40825 (0.0010)
+[2026-06-06 22:17:46,927][262198] Updated weights for policy 0, policy_version 40838 (0.0011)
+[2026-06-06 22:17:47,922][261293] Fps is (10 sec: 3481.4, 60 sec: 3174.4, 300 sec: 2846.4). Total num frames: 20912128. Throughput: 0: 3003.2. Samples: 20912790. Policy #0 lag: (min: 6.0, avg: 20.7, max: 38.0)
+[2026-06-06 22:17:47,925][261293] Avg episode reward: [(0, '1710.042')]
+[2026-06-06 22:17:47,963][262198] Updated weights for policy 0, policy_version 40848 (0.0009)
+[2026-06-06 22:17:49,606][262198] Updated weights for policy 0, policy_version 40858 (0.0008)
+[2026-06-06 22:17:51,275][262198] Updated weights for policy 0, policy_version 40869 (0.0012)
+[2026-06-06 22:17:52,921][261293] Fps is (10 sec: 3686.1, 60 sec: 3174.4, 300 sec: 2860.2). Total num frames: 20930560. Throughput: 0: 2969.0. Samples: 20935216. Policy #0 lag: (min: 6.0, avg: 20.7, max: 38.0)
+[2026-06-06 22:17:52,925][261293] Avg episode reward: [(0, '1775.136')]
+[2026-06-06 22:17:53,025][262198] Updated weights for policy 0, policy_version 40881 (0.0010)
+[2026-06-06 22:17:54,199][262198] Updated weights for policy 0, policy_version 40891 (0.0010)
+[2026-06-06 22:17:55,502][262198] Updated weights for policy 0, policy_version 40902 (0.0009)
+[2026-06-06 22:17:56,810][262198] Updated weights for policy 0, policy_version 40912 (0.0008)
+[2026-06-06 22:17:57,921][261293] Fps is (10 sec: 3891.6, 60 sec: 3174.4, 300 sec: 2881.1). Total num frames: 20951040. Throughput: 0: 2977.9. Samples: 20957714. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:17:57,925][261293] Avg episode reward: [(0, '1641.539')]
+[2026-06-06 22:17:58,358][262198] Updated weights for policy 0, policy_version 40922 (0.0008)
+[2026-06-06 22:17:59,676][262198] Updated weights for policy 0, policy_version 40934 (0.0008)
+[2026-06-06 22:18:00,939][262198] Updated weights for policy 0, policy_version 40944 (0.0008)
+[2026-06-06 22:18:02,444][262198] Updated weights for policy 0, policy_version 40954 (0.0009)
+[2026-06-06 22:18:02,921][261293] Fps is (10 sec: 3891.6, 60 sec: 3174.4, 300 sec: 2895.0). Total num frames: 20969472. Throughput: 0: 3002.2. Samples: 20969850. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:18:02,923][261293] Avg episode reward: [(0, '1800.386')]
+[2026-06-06 22:18:03,726][262198] Updated weights for policy 0, policy_version 40966 (0.0010)
+[2026-06-06 22:18:05,188][262198] Updated weights for policy 0, policy_version 40976 (0.0008)
+[2026-06-06 22:18:06,813][262198] Updated weights for policy 0, policy_version 40986 (0.0009)
+[2026-06-06 22:18:07,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3208.5, 300 sec: 2922.7). Total num frames: 20987904. Throughput: 0: 3276.8. Samples: 20993750. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:18:07,923][261293] Avg episode reward: [(0, '1812.108')]
+[2026-06-06 22:18:08,020][262198] Updated weights for policy 0, policy_version 40997 (0.0010)
+[2026-06-06 22:18:09,324][262198] Updated weights for policy 0, policy_version 41007 (0.0009)
+[2026-06-06 22:18:11,399][262198] Updated weights for policy 0, policy_version 41017 (0.0013)
+[2026-06-06 22:18:12,838][262198] Updated weights for policy 0, policy_version 41027 (0.0011)
+[2026-06-06 22:18:12,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3174.4, 300 sec: 2936.6). Total num frames: 21006336. Throughput: 0: 3496.6. Samples: 21014126. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:18:12,922][261293] Avg episode reward: [(0, '1777.071')]
+[2026-06-06 22:18:14,614][262198] Updated weights for policy 0, policy_version 41037 (0.0010)
+[2026-06-06 22:18:15,905][262198] Updated weights for policy 0, policy_version 41047 (0.0012)
+[2026-06-06 22:18:17,765][262198] Updated weights for policy 0, policy_version 41057 (0.0011)
+[2026-06-06 22:18:17,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3174.4, 300 sec: 2943.6). Total num frames: 21020672. Throughput: 0: 3563.4. Samples: 21023596. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:18:17,924][261293] Avg episode reward: [(0, '1736.995')]
+[2026-06-06 22:18:19,036][262198] Updated weights for policy 0, policy_version 41067 (0.0012)
+[2026-06-06 22:18:20,951][262198] Updated weights for policy 0, policy_version 41077 (0.0010)
+[2026-06-06 22:18:22,468][262198] Updated weights for policy 0, policy_version 41087 (0.0012)
+[2026-06-06 22:18:22,920][261293] Fps is (10 sec: 3072.0, 60 sec: 3310.9, 300 sec: 2950.5). Total num frames: 21037056. Throughput: 0: 3599.2. Samples: 21042604. Policy #0 lag: (min: 1.0, avg: 19.1, max: 37.0)
+[2026-06-06 22:18:22,922][261293] Avg episode reward: [(0, '1662.847')]
+[2026-06-06 22:18:24,289][262198] Updated weights for policy 0, policy_version 41097 (0.0012)
+[2026-06-06 22:18:25,721][262198] Updated weights for policy 0, policy_version 41108 (0.0011)
+[2026-06-06 22:18:27,308][262198] Updated weights for policy 0, policy_version 41118 (0.0008)
+[2026-06-06 22:18:27,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3447.4, 300 sec: 2957.4). Total num frames: 21055488. Throughput: 0: 3577.0. Samples: 21062740. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 22:18:27,923][261293] Avg episode reward: [(0, '1786.677')]
+[2026-06-06 22:18:28,667][262198] Updated weights for policy 0, policy_version 41129 (0.0009)
+[2026-06-06 22:18:29,800][262198] Updated weights for policy 0, policy_version 41139 (0.0009)
+[2026-06-06 22:18:31,403][262198] Updated weights for policy 0, policy_version 41149 (0.0009)
+[2026-06-06 22:18:32,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3549.9, 300 sec: 2957.5). Total num frames: 21073920. Throughput: 0: 3596.4. Samples: 21074626. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 22:18:32,924][261293] Avg episode reward: [(0, '1735.836')]
+[2026-06-06 22:18:32,994][262198] Updated weights for policy 0, policy_version 41161 (0.0010)
+[2026-06-06 22:18:34,238][262198] Updated weights for policy 0, policy_version 41172 (0.0010)
+[2026-06-06 22:18:35,863][262198] Updated weights for policy 0, policy_version 41182 (0.0009)
+[2026-06-06 22:18:36,825][262198] Updated weights for policy 0, policy_version 41192 (0.0008)
+[2026-06-06 22:18:37,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3584.0, 300 sec: 2957.5). Total num frames: 21092352. Throughput: 0: 3590.5. Samples: 21096786. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 22:18:37,924][261293] Avg episode reward: [(0, '1873.908')]
+[2026-06-06 22:18:38,128][262115] Signal inference workers to stop experience collection... (4050 times)
+[2026-06-06 22:18:38,149][262198] InferenceWorker_p0-w0: stopping experience collection (4050 times)
+[2026-06-06 22:18:38,251][262115] Signal inference workers to resume experience collection... (4050 times)
+[2026-06-06 22:18:38,252][262198] InferenceWorker_p0-w0: resuming experience collection (4050 times)
+[2026-06-06 22:18:38,486][262198] Updated weights for policy 0, policy_version 41204 (0.0010)
+[2026-06-06 22:18:40,394][262198] Updated weights for policy 0, policy_version 41214 (0.0010)
+[2026-06-06 22:18:41,362][262198] Updated weights for policy 0, policy_version 41224 (0.0010)
+[2026-06-06 22:18:42,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3584.0, 300 sec: 2950.5). Total num frames: 21108736. Throughput: 0: 3561.1. Samples: 21117964. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 22:18:42,924][261293] Avg episode reward: [(0, '1858.042')]
+[2026-06-06 22:18:43,563][262198] Updated weights for policy 0, policy_version 41235 (0.0012)
+[2026-06-06 22:18:45,551][262198] Updated weights for policy 0, policy_version 41245 (0.0011)
+[2026-06-06 22:18:46,472][262198] Updated weights for policy 0, policy_version 41255 (0.0010)
+[2026-06-06 22:18:47,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3549.9, 300 sec: 2943.6). Total num frames: 21125120. Throughput: 0: 3498.7. Samples: 21127294. Policy #0 lag: (min: 3.0, avg: 20.6, max: 39.0)
+[2026-06-06 22:18:47,924][261293] Avg episode reward: [(0, '1783.580')]
+[2026-06-06 22:18:48,475][262198] Updated weights for policy 0, policy_version 41265 (0.0011)
+[2026-06-06 22:18:50,006][262198] Updated weights for policy 0, policy_version 41275 (0.0011)
+[2026-06-06 22:18:51,419][262198] Updated weights for policy 0, policy_version 41285 (0.0009)
+[2026-06-06 22:18:52,921][261293] Fps is (10 sec: 3277.0, 60 sec: 3515.8, 300 sec: 2943.6). Total num frames: 21141504. Throughput: 0: 3407.7. Samples: 21147096. Policy #0 lag: (min: 1.0, avg: 17.6, max: 36.0)
+[2026-06-06 22:18:52,924][261293] Avg episode reward: [(0, '1804.275')]
+[2026-06-06 22:18:52,953][262198] Updated weights for policy 0, policy_version 41295 (0.0011)
+[2026-06-06 22:18:54,862][262198] Updated weights for policy 0, policy_version 41305 (0.0010)
+[2026-06-06 22:18:55,942][262198] Updated weights for policy 0, policy_version 41315 (0.0009)
+[2026-06-06 22:18:57,771][262198] Updated weights for policy 0, policy_version 41325 (0.0010)
+[2026-06-06 22:18:57,921][261293] Fps is (10 sec: 3277.0, 60 sec: 3447.5, 300 sec: 2943.6). Total num frames: 21157888. Throughput: 0: 3403.1. Samples: 21167268. Policy #0 lag: (min: 1.0, avg: 17.6, max: 36.0)
+[2026-06-06 22:18:57,923][261293] Avg episode reward: [(0, '1798.364')]
+[2026-06-06 22:18:59,120][262198] Updated weights for policy 0, policy_version 41335 (0.0010)
+[2026-06-06 22:19:00,734][262198] Updated weights for policy 0, policy_version 41345 (0.0009)
+[2026-06-06 22:19:01,985][262198] Updated weights for policy 0, policy_version 41355 (0.0010)
+[2026-06-06 22:19:02,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3447.5, 300 sec: 2943.6). Total num frames: 21176320. Throughput: 0: 3416.7. Samples: 21177346. Policy #0 lag: (min: 1.0, avg: 17.6, max: 36.0)
+[2026-06-06 22:19:02,921][261293] Avg episode reward: [(0, '1769.845')]
+[2026-06-06 22:19:03,946][262198] Updated weights for policy 0, policy_version 41365 (0.0009)
+[2026-06-06 22:19:05,231][262198] Updated weights for policy 0, policy_version 41375 (0.0008)
+[2026-06-06 22:19:06,809][262198] Updated weights for policy 0, policy_version 41385 (0.0008)
+[2026-06-06 22:19:07,900][262198] Updated weights for policy 0, policy_version 41395 (0.0009)
+[2026-06-06 22:19:07,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3413.3, 300 sec: 2936.6). Total num frames: 21192704. Throughput: 0: 3446.2. Samples: 21197682. Policy #0 lag: (min: 1.0, avg: 17.6, max: 36.0)
+[2026-06-06 22:19:07,925][261293] Avg episode reward: [(0, '1727.344')]
+[2026-06-06 22:19:09,675][262198] Updated weights for policy 0, policy_version 41405 (0.0009)
+[2026-06-06 22:19:10,594][262198] Updated weights for policy 0, policy_version 41415 (0.0008)
+[2026-06-06 22:19:12,208][262198] Updated weights for policy 0, policy_version 41425 (0.0008)
+[2026-06-06 22:19:12,921][261293] Fps is (10 sec: 3481.3, 60 sec: 3413.3, 300 sec: 2936.6). Total num frames: 21211136. Throughput: 0: 3480.7. Samples: 21219370. Policy #0 lag: (min: 1.0, avg: 17.6, max: 36.0)
+[2026-06-06 22:19:12,924][261293] Avg episode reward: [(0, '1700.992')]
+[2026-06-06 22:19:13,924][262198] Updated weights for policy 0, policy_version 41435 (0.0009)
+[2026-06-06 22:19:15,350][262198] Updated weights for policy 0, policy_version 41445 (0.0011)
+[2026-06-06 22:19:16,944][262198] Updated weights for policy 0, policy_version 41455 (0.0008)
+[2026-06-06 22:19:17,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3447.5, 300 sec: 2929.7). Total num frames: 21227520. Throughput: 0: 3435.7. Samples: 21229232. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:17,923][261293] Avg episode reward: [(0, '1777.215')]
+[2026-06-06 22:19:18,046][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000041464_21229568.pth...
+[2026-06-06 22:19:18,098][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000019856_10166272.pth
+[2026-06-06 22:19:18,754][262198] Updated weights for policy 0, policy_version 41465 (0.0011)
+[2026-06-06 22:19:19,745][262198] Updated weights for policy 0, policy_version 41475 (0.0008)
+[2026-06-06 22:19:21,627][262198] Updated weights for policy 0, policy_version 41485 (0.0010)
+[2026-06-06 22:19:22,813][262198] Updated weights for policy 0, policy_version 41495 (0.0009)
+[2026-06-06 22:19:22,921][261293] Fps is (10 sec: 3481.8, 60 sec: 3481.6, 300 sec: 2929.7). Total num frames: 21245952. Throughput: 0: 3394.8. Samples: 21249550. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:22,922][261293] Avg episode reward: [(0, '1724.103')]
+[2026-06-06 22:19:24,477][262198] Updated weights for policy 0, policy_version 41505 (0.0010)
+[2026-06-06 22:19:25,727][262198] Updated weights for policy 0, policy_version 41515 (0.0010)
+[2026-06-06 22:19:27,484][262198] Updated weights for policy 0, policy_version 41525 (0.0008)
+[2026-06-06 22:19:27,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3447.5, 300 sec: 2915.8). Total num frames: 21262336. Throughput: 0: 3398.0. Samples: 21270870. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:27,923][261293] Avg episode reward: [(0, '1727.006')]
+[2026-06-06 22:19:29,200][262198] Updated weights for policy 0, policy_version 41537 (0.0010)
+[2026-06-06 22:19:30,352][262198] Updated weights for policy 0, policy_version 41547 (0.0008)
+[2026-06-06 22:19:32,197][262198] Updated weights for policy 0, policy_version 41557 (0.0009)
+[2026-06-06 22:19:32,923][261293] Fps is (10 sec: 3276.1, 60 sec: 3413.2, 300 sec: 2936.6). Total num frames: 21278720. Throughput: 0: 3418.2. Samples: 21281120. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:32,924][261293] Avg episode reward: [(0, '1610.350')]
+[2026-06-06 22:19:33,945][262198] Updated weights for policy 0, policy_version 41569 (0.0009)
+[2026-06-06 22:19:35,103][262198] Updated weights for policy 0, policy_version 41579 (0.0011)
+[2026-06-06 22:19:36,785][262198] Updated weights for policy 0, policy_version 41589 (0.0011)
+[2026-06-06 22:19:37,920][261293] Fps is (10 sec: 3481.6, 60 sec: 3413.4, 300 sec: 2957.5). Total num frames: 21297152. Throughput: 0: 3443.5. Samples: 21302054. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:37,922][261293] Avg episode reward: [(0, '1794.367')]
+[2026-06-06 22:19:38,321][262198] Updated weights for policy 0, policy_version 41600 (0.0008)
+[2026-06-06 22:19:39,517][262198] Updated weights for policy 0, policy_version 41610 (0.0009)
+[2026-06-06 22:19:40,680][262198] Updated weights for policy 0, policy_version 41620 (0.0008)
+[2026-06-06 22:19:42,495][262198] Updated weights for policy 0, policy_version 41630 (0.0008)
+[2026-06-06 22:19:42,920][261293] Fps is (10 sec: 3687.3, 60 sec: 3447.5, 300 sec: 2985.2). Total num frames: 21315584. Throughput: 0: 3495.1. Samples: 21324548. Policy #0 lag: (min: 3.0, avg: 20.5, max: 37.0)
+[2026-06-06 22:19:42,922][261293] Avg episode reward: [(0, '1672.752')]
+[2026-06-06 22:19:43,514][262198] Updated weights for policy 0, policy_version 41640 (0.0008)
+[2026-06-06 22:19:45,010][262198] Updated weights for policy 0, policy_version 41650 (0.0010)
+[2026-06-06 22:19:46,634][262198] Updated weights for policy 0, policy_version 41660 (0.0012)
+[2026-06-06 22:19:47,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3481.6, 300 sec: 3013.0). Total num frames: 21334016. Throughput: 0: 3503.6. Samples: 21335010. Policy #0 lag: (min: 4.0, avg: 21.3, max: 38.0)
+[2026-06-06 22:19:47,921][261293] Avg episode reward: [(0, '1782.080')]
+[2026-06-06 22:19:48,028][262198] Updated weights for policy 0, policy_version 41672 (0.0009)
+[2026-06-06 22:19:49,480][262198] Updated weights for policy 0, policy_version 41682 (0.0008)
+[2026-06-06 22:19:51,130][262198] Updated weights for policy 0, policy_version 41692 (0.0010)
+[2026-06-06 22:19:52,442][262198] Updated weights for policy 0, policy_version 41703 (0.0009)
+[2026-06-06 22:19:52,921][261293] Fps is (10 sec: 3686.1, 60 sec: 3515.7, 300 sec: 3026.9). Total num frames: 21352448. Throughput: 0: 3541.3. Samples: 21357040. Policy #0 lag: (min: 4.0, avg: 21.3, max: 38.0)
+[2026-06-06 22:19:52,924][261293] Avg episode reward: [(0, '1724.043')]
+[2026-06-06 22:19:53,853][262198] Updated weights for policy 0, policy_version 41713 (0.0009)
+[2026-06-06 22:19:55,581][262198] Updated weights for policy 0, policy_version 41723 (0.0009)
+[2026-06-06 22:19:56,757][262198] Updated weights for policy 0, policy_version 41733 (0.0011)
+[2026-06-06 22:19:57,923][261293] Fps is (10 sec: 3685.6, 60 sec: 3549.8, 300 sec: 3054.6). Total num frames: 21370880. Throughput: 0: 3529.0. Samples: 21378182. Policy #0 lag: (min: 4.0, avg: 21.3, max: 38.0)
+[2026-06-06 22:19:57,925][261293] Avg episode reward: [(0, '1757.862')]
+[2026-06-06 22:19:58,127][262198] Updated weights for policy 0, policy_version 41743 (0.0008)
+[2026-06-06 22:19:59,978][262198] Updated weights for policy 0, policy_version 41753 (0.0007)
+[2026-06-06 22:20:01,272][262198] Updated weights for policy 0, policy_version 41765 (0.0009)
+[2026-06-06 22:20:02,343][262198] Updated weights for policy 0, policy_version 41775 (0.0009)
+[2026-06-06 22:20:02,921][261293] Fps is (10 sec: 3891.4, 60 sec: 3584.0, 300 sec: 3089.4). Total num frames: 21391360. Throughput: 0: 3580.1. Samples: 21390336. Policy #0 lag: (min: 4.0, avg: 21.3, max: 38.0)
+[2026-06-06 22:20:02,922][261293] Avg episode reward: [(0, '1810.651')]
+[2026-06-06 22:20:04,183][262198] Updated weights for policy 0, policy_version 41785 (0.0010)
+[2026-06-06 22:20:05,091][262198] Updated weights for policy 0, policy_version 41795 (0.0011)
+[2026-06-06 22:20:06,550][262198] Updated weights for policy 0, policy_version 41805 (0.0008)
+[2026-06-06 22:20:07,862][262198] Updated weights for policy 0, policy_version 41816 (0.0009)
+[2026-06-06 22:20:07,921][261293] Fps is (10 sec: 3891.9, 60 sec: 3618.1, 300 sec: 3117.1). Total num frames: 21409792. Throughput: 0: 3637.9. Samples: 21413256. Policy #0 lag: (min: 4.0, avg: 21.3, max: 38.0)
+[2026-06-06 22:20:07,923][261293] Avg episode reward: [(0, '1871.833')]
+[2026-06-06 22:20:09,244][262198] Updated weights for policy 0, policy_version 41826 (0.0009)
+[2026-06-06 22:20:10,309][262198] Updated weights for policy 0, policy_version 41836 (0.0010)
+[2026-06-06 22:20:11,868][262198] Updated weights for policy 0, policy_version 41848 (0.0008)
+[2026-06-06 22:20:12,921][261293] Fps is (10 sec: 3686.4, 60 sec: 3618.2, 300 sec: 3144.9). Total num frames: 21428224. Throughput: 0: 3696.4. Samples: 21437208. Policy #0 lag: (min: 6.0, avg: 23.3, max: 42.0)
+[2026-06-06 22:20:12,923][261293] Avg episode reward: [(0, '1537.359')]
+[2026-06-06 22:20:13,391][262198] Updated weights for policy 0, policy_version 41858 (0.0009)
+[2026-06-06 22:20:14,833][262198] Updated weights for policy 0, policy_version 41869 (0.0008)
+[2026-06-06 22:20:16,053][262198] Updated weights for policy 0, policy_version 41880 (0.0009)
+[2026-06-06 22:20:17,830][262198] Updated weights for policy 0, policy_version 41891 (0.0010)
+[2026-06-06 22:20:17,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3686.4, 300 sec: 3179.6). Total num frames: 21448704. Throughput: 0: 3718.7. Samples: 21448454. Policy #0 lag: (min: 6.0, avg: 23.3, max: 42.0)
+[2026-06-06 22:20:17,924][261293] Avg episode reward: [(0, '1646.102')]
+[2026-06-06 22:20:19,636][262198] Updated weights for policy 0, policy_version 41901 (0.0013)
+[2026-06-06 22:20:21,083][262198] Updated weights for policy 0, policy_version 41912 (0.0011)
+[2026-06-06 22:20:22,696][262198] Updated weights for policy 0, policy_version 41922 (0.0009)
+[2026-06-06 22:20:22,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3652.3, 300 sec: 3186.5). Total num frames: 21465088. Throughput: 0: 3707.1. Samples: 21468872. Policy #0 lag: (min: 6.0, avg: 23.3, max: 42.0)
+[2026-06-06 22:20:22,922][261293] Avg episode reward: [(0, '1867.999')]
+[2026-06-06 22:20:24,626][262198] Updated weights for policy 0, policy_version 41933 (0.0013)
+[2026-06-06 22:20:26,393][262198] Updated weights for policy 0, policy_version 41945 (0.0010)
+[2026-06-06 22:20:27,514][262198] Updated weights for policy 0, policy_version 41955 (0.0012)
+[2026-06-06 22:20:27,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3652.3, 300 sec: 3214.3). Total num frames: 21481472. Throughput: 0: 3670.8. Samples: 21489736. Policy #0 lag: (min: 6.0, avg: 23.3, max: 42.0)
+[2026-06-06 22:20:27,922][261293] Avg episode reward: [(0, '1607.906')]
+[2026-06-06 22:20:29,376][262198] Updated weights for policy 0, policy_version 41965 (0.0011)
+[2026-06-06 22:20:31,208][262198] Updated weights for policy 0, policy_version 41977 (0.0009)
+[2026-06-06 22:20:32,247][262198] Updated weights for policy 0, policy_version 41987 (0.0008)
+[2026-06-06 22:20:32,920][261293] Fps is (10 sec: 3481.7, 60 sec: 3686.5, 300 sec: 3235.1). Total num frames: 21499904. Throughput: 0: 3665.7. Samples: 21499966. Policy #0 lag: (min: 6.0, avg: 23.3, max: 42.0)
+[2026-06-06 22:20:32,922][261293] Avg episode reward: [(0, '1739.459')]
+[2026-06-06 22:20:33,634][262198] Updated weights for policy 0, policy_version 41997 (0.0010)
+[2026-06-06 22:20:35,083][262198] Updated weights for policy 0, policy_version 42007 (0.0011)
+[2026-06-06 22:20:36,846][262198] Updated weights for policy 0, policy_version 42017 (0.0008)
+[2026-06-06 22:20:37,833][262198] Updated weights for policy 0, policy_version 42027 (0.0010)
+[2026-06-06 22:20:37,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3686.4, 300 sec: 3256.0). Total num frames: 21518336. Throughput: 0: 3646.0. Samples: 21521108. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:20:37,922][261293] Avg episode reward: [(0, '1690.820')]
+[2026-06-06 22:20:38,644][262115] Signal inference workers to stop experience collection... (4100 times)
+[2026-06-06 22:20:38,675][262198] InferenceWorker_p0-w0: stopping experience collection (4100 times)
+[2026-06-06 22:20:38,703][262115] Signal inference workers to resume experience collection... (4100 times)
+[2026-06-06 22:20:38,704][262198] InferenceWorker_p0-w0: resuming experience collection (4100 times)
+[2026-06-06 22:20:39,648][262198] Updated weights for policy 0, policy_version 42037 (0.0009)
+[2026-06-06 22:20:41,049][262198] Updated weights for policy 0, policy_version 42047 (0.0009)
+[2026-06-06 22:20:42,409][262198] Updated weights for policy 0, policy_version 42058 (0.0010)
+[2026-06-06 22:20:42,921][261293] Fps is (10 sec: 3481.4, 60 sec: 3652.2, 300 sec: 3276.8). Total num frames: 21534720. Throughput: 0: 3652.3. Samples: 21542528. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:20:42,923][261293] Avg episode reward: [(0, '1808.803')]
+[2026-06-06 22:20:43,490][262198] Updated weights for policy 0, policy_version 42068 (0.0009)
+[2026-06-06 22:20:45,618][262198] Updated weights for policy 0, policy_version 42078 (0.0009)
+[2026-06-06 22:20:47,293][262198] Updated weights for policy 0, policy_version 42089 (0.0010)
+[2026-06-06 22:20:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 3618.1, 300 sec: 3290.7). Total num frames: 21551104. Throughput: 0: 3621.5. Samples: 21553306. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:20:47,922][261293] Avg episode reward: [(0, '1796.582')]
+[2026-06-06 22:20:48,372][262198] Updated weights for policy 0, policy_version 42099 (0.0009)
+[2026-06-06 22:20:50,815][262198] Updated weights for policy 0, policy_version 42109 (0.0010)
+[2026-06-06 22:20:52,786][262198] Updated weights for policy 0, policy_version 42121 (0.0012)
+[2026-06-06 22:20:52,921][261293] Fps is (10 sec: 3277.0, 60 sec: 3584.0, 300 sec: 3311.5). Total num frames: 21567488. Throughput: 0: 3512.3. Samples: 21571306. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:20:52,924][261293] Avg episode reward: [(0, '1664.790')]
+[2026-06-06 22:20:54,315][262198] Updated weights for policy 0, policy_version 42131 (0.0013)
+[2026-06-06 22:20:56,924][262198] Updated weights for policy 0, policy_version 42141 (0.0011)
+[2026-06-06 22:20:57,921][261293] Fps is (10 sec: 2662.4, 60 sec: 3447.6, 300 sec: 3304.6). Total num frames: 21577728. Throughput: 0: 3332.3. Samples: 21587164. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:20:57,922][261293] Avg episode reward: [(0, '1678.211')]
+[2026-06-06 22:20:59,024][262198] Updated weights for policy 0, policy_version 42151 (0.0010)
+[2026-06-06 22:21:00,397][262198] Updated weights for policy 0, policy_version 42161 (0.0009)
+[2026-06-06 22:21:02,869][262198] Updated weights for policy 0, policy_version 42172 (0.0012)
+[2026-06-06 22:21:02,921][261293] Fps is (10 sec: 2457.5, 60 sec: 3345.1, 300 sec: 3304.6). Total num frames: 21592064. Throughput: 0: 3258.7. Samples: 21595096. Policy #0 lag: (min: 3.0, avg: 18.6, max: 35.0)
+[2026-06-06 22:21:02,923][261293] Avg episode reward: [(0, '1663.453')]
+[2026-06-06 22:21:04,338][262198] Updated weights for policy 0, policy_version 42182 (0.0010)
+[2026-06-06 22:21:05,634][262198] Updated weights for policy 0, policy_version 42193 (0.0011)
+[2026-06-06 22:21:07,500][262198] Updated weights for policy 0, policy_version 42204 (0.0010)
+[2026-06-06 22:21:07,921][261293] Fps is (10 sec: 3276.9, 60 sec: 3345.1, 300 sec: 3318.5). Total num frames: 21610496. Throughput: 0: 3233.2. Samples: 21614366. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 22:21:07,922][261293] Avg episode reward: [(0, '1798.815')]
+[2026-06-06 22:21:08,586][262198] Updated weights for policy 0, policy_version 42214 (0.0008)
+[2026-06-06 22:21:09,791][262198] Updated weights for policy 0, policy_version 42224 (0.0008)
+[2026-06-06 22:21:11,282][262198] Updated weights for policy 0, policy_version 42234 (0.0008)
+[2026-06-06 22:21:12,277][262198] Updated weights for policy 0, policy_version 42245 (0.0008)
+[2026-06-06 22:21:12,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3379.2, 300 sec: 3353.2). Total num frames: 21630976. Throughput: 0: 3326.5. Samples: 21639428. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 22:21:12,922][261293] Avg episode reward: [(0, '1822.395')]
+[2026-06-06 22:21:13,669][262198] Updated weights for policy 0, policy_version 42257 (0.0009)
+[2026-06-06 22:21:15,520][262198] Updated weights for policy 0, policy_version 42269 (0.0008)
+[2026-06-06 22:21:16,433][262198] Updated weights for policy 0, policy_version 42279 (0.0007)
+[2026-06-06 22:21:17,648][262198] Updated weights for policy 0, policy_version 42289 (0.0007)
+[2026-06-06 22:21:17,921][261293] Fps is (10 sec: 4300.6, 60 sec: 3413.3, 300 sec: 3387.9). Total num frames: 21653504. Throughput: 0: 3373.8. Samples: 21651788. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 22:21:17,923][261293] Avg episode reward: [(0, '1655.994')]
+[2026-06-06 22:21:19,184][262198] Updated weights for policy 0, policy_version 42300 (0.0008)
+[2026-06-06 22:21:20,028][262198] Updated weights for policy 0, policy_version 42310 (0.0007)
+[2026-06-06 22:21:21,129][262198] Updated weights for policy 0, policy_version 42320 (0.0007)
+[2026-06-06 22:21:22,887][262198] Updated weights for policy 0, policy_version 42330 (0.0008)
+[2026-06-06 22:21:22,922][261293] Fps is (10 sec: 4095.8, 60 sec: 3447.4, 300 sec: 3408.7). Total num frames: 21671936. Throughput: 0: 3486.8. Samples: 21678014. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 22:21:22,927][261293] Avg episode reward: [(0, '1879.347')]
+[2026-06-06 22:21:23,899][262198] Updated weights for policy 0, policy_version 42343 (0.0036)
+[2026-06-06 22:21:25,086][262198] Updated weights for policy 0, policy_version 42353 (0.0008)
+[2026-06-06 22:21:26,330][262198] Updated weights for policy 0, policy_version 42363 (0.0008)
+[2026-06-06 22:21:27,316][262198] Updated weights for policy 0, policy_version 42373 (0.0008)
+[2026-06-06 22:21:27,922][261293] Fps is (10 sec: 4300.7, 60 sec: 3584.0, 300 sec: 3450.4). Total num frames: 21696512. Throughput: 0: 3590.8. Samples: 21704116. Policy #0 lag: (min: 3.0, avg: 18.8, max: 35.0)
+[2026-06-06 22:21:27,926][261293] Avg episode reward: [(0, '1688.110')]
+[2026-06-06 22:21:28,432][262198] Updated weights for policy 0, policy_version 42383 (0.0008)
+[2026-06-06 22:21:30,198][262198] Updated weights for policy 0, policy_version 42393 (0.0008)
+[2026-06-06 22:21:31,092][262198] Updated weights for policy 0, policy_version 42405 (0.0007)
+[2026-06-06 22:21:32,266][262198] Updated weights for policy 0, policy_version 42415 (0.0009)
+[2026-06-06 22:21:32,921][261293] Fps is (10 sec: 4710.6, 60 sec: 3652.3, 300 sec: 3492.0). Total num frames: 21719040. Throughput: 0: 3639.0. Samples: 21717058. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:21:32,923][261293] Avg episode reward: [(0, '1818.037')]
+[2026-06-06 22:21:33,898][262198] Updated weights for policy 0, policy_version 42426 (0.0008)
+[2026-06-06 22:21:35,003][262198] Updated weights for policy 0, policy_version 42438 (0.0008)
+[2026-06-06 22:21:36,576][262198] Updated weights for policy 0, policy_version 42449 (0.0009)
+[2026-06-06 22:21:37,922][261293] Fps is (10 sec: 4095.8, 60 sec: 3652.2, 300 sec: 3498.9). Total num frames: 21737472. Throughput: 0: 3805.3. Samples: 21742552. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:21:37,926][261293] Avg episode reward: [(0, '1629.450')]
+[2026-06-06 22:21:38,134][262198] Updated weights for policy 0, policy_version 42461 (0.0008)
+[2026-06-06 22:21:39,246][262198] Updated weights for policy 0, policy_version 42471 (0.0008)
+[2026-06-06 22:21:40,334][262198] Updated weights for policy 0, policy_version 42481 (0.0007)
+[2026-06-06 22:21:41,829][262198] Updated weights for policy 0, policy_version 42493 (0.0008)
+[2026-06-06 22:21:42,734][262198] Updated weights for policy 0, policy_version 42503 (0.0007)
+[2026-06-06 22:21:42,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3788.8, 300 sec: 3526.7). Total num frames: 21762048. Throughput: 0: 4026.7. Samples: 21768364. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:21:42,923][261293] Avg episode reward: [(0, '1811.972')]
+[2026-06-06 22:21:43,952][262198] Updated weights for policy 0, policy_version 42513 (0.0007)
+[2026-06-06 22:21:45,193][262198] Updated weights for policy 0, policy_version 42524 (0.0007)
+[2026-06-06 22:21:46,253][262198] Updated weights for policy 0, policy_version 42534 (0.0007)
+[2026-06-06 22:21:47,149][262115] Signal inference workers to stop experience collection... (4150 times)
+[2026-06-06 22:21:47,162][262198] InferenceWorker_p0-w0: stopping experience collection (4150 times)
+[2026-06-06 22:21:47,240][262115] Signal inference workers to resume experience collection... (4150 times)
+[2026-06-06 22:21:47,240][262198] InferenceWorker_p0-w0: resuming experience collection (4150 times)
+[2026-06-06 22:21:47,773][262198] Updated weights for policy 0, policy_version 42545 (0.0008)
+[2026-06-06 22:21:47,921][261293] Fps is (10 sec: 4711.0, 60 sec: 3891.2, 300 sec: 3540.6). Total num frames: 21784576. Throughput: 0: 4162.5. Samples: 21782406. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:21:47,923][261293] Avg episode reward: [(0, '1705.920')]
+[2026-06-06 22:21:49,205][262198] Updated weights for policy 0, policy_version 42557 (0.0007)
+[2026-06-06 22:21:50,042][262198] Updated weights for policy 0, policy_version 42567 (0.0007)
+[2026-06-06 22:21:51,528][262198] Updated weights for policy 0, policy_version 42578 (0.0008)
+[2026-06-06 22:21:52,868][262198] Updated weights for policy 0, policy_version 42589 (0.0007)
+[2026-06-06 22:21:52,920][261293] Fps is (10 sec: 4300.8, 60 sec: 3959.5, 300 sec: 3540.6). Total num frames: 21805056. Throughput: 0: 4306.1. Samples: 21808140. Policy #0 lag: (min: 4.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:21:52,922][261293] Avg episode reward: [(0, '1764.572')]
+[2026-06-06 22:21:53,696][262198] Updated weights for policy 0, policy_version 42599 (0.0007)
+[2026-06-06 22:21:55,490][262198] Updated weights for policy 0, policy_version 42611 (0.0008)
+[2026-06-06 22:21:56,673][262198] Updated weights for policy 0, policy_version 42623 (0.0008)
+[2026-06-06 22:21:57,921][261293] Fps is (10 sec: 4300.9, 60 sec: 4164.3, 300 sec: 3554.5). Total num frames: 21827584. Throughput: 0: 4324.0. Samples: 21834008. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:21:57,922][261293] Avg episode reward: [(0, '1558.840')]
+[2026-06-06 22:21:58,497][262198] Updated weights for policy 0, policy_version 42633 (0.0008)
+[2026-06-06 22:21:59,259][262198] Updated weights for policy 0, policy_version 42644 (0.0008)
+[2026-06-06 22:22:00,504][262198] Updated weights for policy 0, policy_version 42655 (0.0007)
+[2026-06-06 22:22:02,216][262198] Updated weights for policy 0, policy_version 42665 (0.0007)
+[2026-06-06 22:22:02,920][261293] Fps is (10 sec: 4300.9, 60 sec: 4266.7, 300 sec: 3568.4). Total num frames: 21848064. Throughput: 0: 4342.9. Samples: 21847214. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:22:02,922][261293] Avg episode reward: [(0, '1795.918')]
+[2026-06-06 22:22:03,436][262198] Updated weights for policy 0, policy_version 42676 (0.0009)
+[2026-06-06 22:22:04,790][262198] Updated weights for policy 0, policy_version 42690 (0.0010)
+[2026-06-06 22:22:07,920][261293] Fps is (10 sec: 3276.8, 60 sec: 4164.3, 300 sec: 3540.6). Total num frames: 21860352. Throughput: 0: 4243.2. Samples: 21868954. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:22:07,921][261293] Avg episode reward: [(0, '1827.994')]
+[2026-06-06 22:22:08,041][262198] Updated weights for policy 0, policy_version 42700 (0.0014)
+[2026-06-06 22:22:10,815][262198] Updated weights for policy 0, policy_version 42711 (0.0014)
+[2026-06-06 22:22:12,920][261293] Fps is (10 sec: 2457.6, 60 sec: 4027.7, 300 sec: 3533.7). Total num frames: 21872640. Throughput: 0: 3910.4. Samples: 21880082. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:22:12,922][261293] Avg episode reward: [(0, '1808.451')]
+[2026-06-06 22:22:13,300][262198] Updated weights for policy 0, policy_version 42721 (0.0015)
+[2026-06-06 22:22:16,367][262198] Updated weights for policy 0, policy_version 42731 (0.0015)
+[2026-06-06 22:22:17,921][261293] Fps is (10 sec: 2047.9, 60 sec: 3788.8, 300 sec: 3533.7). Total num frames: 21880832. Throughput: 0: 3761.9. Samples: 21886346. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:22:17,922][261293] Avg episode reward: [(0, '1750.634')]
+[2026-06-06 22:22:18,755][262198] Updated weights for policy 0, policy_version 42742 (0.0012)
+[2026-06-06 22:22:19,831][262198] Updated weights for policy 0, policy_version 42752 (0.0013)
+[2026-06-06 22:22:22,921][261293] Fps is (10 sec: 2047.9, 60 sec: 3686.4, 300 sec: 3540.6). Total num frames: 21893120. Throughput: 0: 3477.1. Samples: 21899016. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:22:22,922][261293] Avg episode reward: [(0, '1811.080')]
+[2026-06-06 22:22:24,296][262198] Updated weights for policy 0, policy_version 42762 (0.0012)
+[2026-06-06 22:22:26,795][262198] Updated weights for policy 0, policy_version 42772 (0.0012)
+[2026-06-06 22:22:27,920][261293] Fps is (10 sec: 2048.1, 60 sec: 3413.4, 300 sec: 3526.7). Total num frames: 21901312. Throughput: 0: 3156.5. Samples: 21910406. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 22:22:27,921][261293] Avg episode reward: [(0, '1719.713')]
+[2026-06-06 22:22:28,276][262198] Updated weights for policy 0, policy_version 42783 (0.0012)
+[2026-06-06 22:22:30,860][262198] Updated weights for policy 0, policy_version 42793 (0.0011)
+[2026-06-06 22:22:32,343][262198] Updated weights for policy 0, policy_version 42805 (0.0008)
+[2026-06-06 22:22:32,921][261293] Fps is (10 sec: 2457.6, 60 sec: 3310.9, 300 sec: 3526.7). Total num frames: 21917696. Throughput: 0: 3040.1. Samples: 21919212. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 22:22:32,923][261293] Avg episode reward: [(0, '1749.288')]
+[2026-06-06 22:22:33,995][262198] Updated weights for policy 0, policy_version 42821 (0.0008)
+[2026-06-06 22:22:35,312][262198] Updated weights for policy 0, policy_version 42831 (0.0009)
+[2026-06-06 22:22:36,846][262198] Updated weights for policy 0, policy_version 42843 (0.0011)
+[2026-06-06 22:22:37,920][261293] Fps is (10 sec: 3686.4, 60 sec: 3345.1, 300 sec: 3540.6). Total num frames: 21938176. Throughput: 0: 2995.0. Samples: 21942914. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 22:22:37,922][261293] Avg episode reward: [(0, '1776.904')]
+[2026-06-06 22:22:39,036][262198] Updated weights for policy 0, policy_version 42853 (0.0014)
+[2026-06-06 22:22:41,821][262198] Updated weights for policy 0, policy_version 42863 (0.0013)
+[2026-06-06 22:22:42,920][261293] Fps is (10 sec: 2867.3, 60 sec: 3072.0, 300 sec: 3505.9). Total num frames: 21946368. Throughput: 0: 2708.9. Samples: 21955908. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 22:22:42,922][261293] Avg episode reward: [(0, '1755.836')]
+[2026-06-06 22:22:44,684][262198] Updated weights for policy 0, policy_version 42874 (0.0013)
+[2026-06-06 22:22:46,421][262198] Updated weights for policy 0, policy_version 42884 (0.0012)
+[2026-06-06 22:22:47,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2901.3, 300 sec: 3485.1). Total num frames: 21958656. Throughput: 0: 2548.4. Samples: 21961892. Policy #0 lag: (min: 3.0, avg: 17.7, max: 35.0)
+[2026-06-06 22:22:47,922][261293] Avg episode reward: [(0, '1785.048')]
+[2026-06-06 22:22:49,709][262198] Updated weights for policy 0, policy_version 42894 (0.0014)
+[2026-06-06 22:22:52,428][262198] Updated weights for policy 0, policy_version 42905 (0.0013)
+[2026-06-06 22:22:52,923][261293] Fps is (10 sec: 2252.1, 60 sec: 2730.5, 300 sec: 3450.3). Total num frames: 21968896. Throughput: 0: 2347.8. Samples: 21974614. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:22:52,926][261293] Avg episode reward: [(0, '1892.001')]
+[2026-06-06 22:22:54,111][262198] Updated weights for policy 0, policy_version 42916 (0.0012)
+[2026-06-06 22:22:56,252][262198] Updated weights for policy 0, policy_version 42927 (0.0008)
+[2026-06-06 22:22:57,920][261293] Fps is (10 sec: 2457.7, 60 sec: 2594.1, 300 sec: 3436.5). Total num frames: 21983232. Throughput: 0: 2518.1. Samples: 21993396. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:22:57,923][261293] Avg episode reward: [(0, '1752.183')]
+[2026-06-06 22:22:58,211][262198] Updated weights for policy 0, policy_version 42942 (0.0013)
+[2026-06-06 22:22:59,775][262198] Updated weights for policy 0, policy_version 42952 (0.0011)
+[2026-06-06 22:23:02,117][262198] Updated weights for policy 0, policy_version 42962 (0.0011)
+[2026-06-06 22:23:02,921][261293] Fps is (10 sec: 3277.5, 60 sec: 2560.0, 300 sec: 3436.5). Total num frames: 22001664. Throughput: 0: 2569.8. Samples: 22001986. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:23:02,923][261293] Avg episode reward: [(0, '1730.433')]
+[2026-06-06 22:23:03,428][262198] Updated weights for policy 0, policy_version 42973 (0.0011)
+[2026-06-06 22:23:04,970][262198] Updated weights for policy 0, policy_version 42983 (0.0012)
+[2026-06-06 22:23:06,879][262198] Updated weights for policy 0, policy_version 42993 (0.0008)
+[2026-06-06 22:23:07,678][262198] Updated weights for policy 0, policy_version 43004 (0.0010)
+[2026-06-06 22:23:07,921][261293] Fps is (10 sec: 3481.6, 60 sec: 2628.3, 300 sec: 3429.5). Total num frames: 22018048. Throughput: 0: 2730.3. Samples: 22021880. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:23:07,923][261293] Avg episode reward: [(0, '1749.578')]
+[2026-06-06 22:23:09,680][262198] Updated weights for policy 0, policy_version 43014 (0.0010)
+[2026-06-06 22:23:10,615][262198] Updated weights for policy 0, policy_version 43024 (0.0008)
+[2026-06-06 22:23:12,037][262198] Updated weights for policy 0, policy_version 43035 (0.0009)
+[2026-06-06 22:23:12,920][261293] Fps is (10 sec: 3481.9, 60 sec: 2730.7, 300 sec: 3443.4). Total num frames: 22036480. Throughput: 0: 2959.0. Samples: 22043560. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:23:12,922][261293] Avg episode reward: [(0, '1687.005')]
+[2026-06-06 22:23:13,865][262198] Updated weights for policy 0, policy_version 43045 (0.0008)
+[2026-06-06 22:23:14,910][262198] Updated weights for policy 0, policy_version 43055 (0.0008)
+[2026-06-06 22:23:16,236][262115] Signal inference workers to stop experience collection... (4200 times)
+[2026-06-06 22:23:16,270][262198] InferenceWorker_p0-w0: stopping experience collection (4200 times)
+[2026-06-06 22:23:16,324][262115] Signal inference workers to resume experience collection... (4200 times)
+[2026-06-06 22:23:16,325][262198] InferenceWorker_p0-w0: resuming experience collection (4200 times)
+[2026-06-06 22:23:16,457][262198] Updated weights for policy 0, policy_version 43066 (0.0009)
+[2026-06-06 22:23:17,427][262198] Updated weights for policy 0, policy_version 43076 (0.0008)
+[2026-06-06 22:23:17,921][261293] Fps is (10 sec: 3686.4, 60 sec: 2901.3, 300 sec: 3450.4). Total num frames: 22054912. Throughput: 0: 3011.3. Samples: 22054718. Policy #0 lag: (min: 3.0, avg: 19.3, max: 35.0)
+[2026-06-06 22:23:17,923][261293] Avg episode reward: [(0, '1657.890')]
+[2026-06-06 22:23:19,133][262198] Updated weights for policy 0, policy_version 43086 (0.0009)
+[2026-06-06 22:23:21,071][262198] Updated weights for policy 0, policy_version 43099 (0.0010)
+[2026-06-06 22:23:22,552][262198] Updated weights for policy 0, policy_version 43109 (0.0008)
+[2026-06-06 22:23:22,921][261293] Fps is (10 sec: 3686.1, 60 sec: 3003.7, 300 sec: 3450.4). Total num frames: 22073344. Throughput: 0: 2990.3. Samples: 22077482. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:23:22,925][261293] Avg episode reward: [(0, '1743.047')]
+[2026-06-06 22:23:23,593][262198] Updated weights for policy 0, policy_version 43119 (0.0009)
+[2026-06-06 22:23:25,568][262198] Updated weights for policy 0, policy_version 43133 (0.0011)
+[2026-06-06 22:23:27,189][262198] Updated weights for policy 0, policy_version 43143 (0.0009)
+[2026-06-06 22:23:27,920][261293] Fps is (10 sec: 3686.5, 60 sec: 3174.4, 300 sec: 3450.4). Total num frames: 22091776. Throughput: 0: 3190.3. Samples: 22099470. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:23:27,922][261293] Avg episode reward: [(0, '1785.538')]
+[2026-06-06 22:23:28,813][262198] Updated weights for policy 0, policy_version 43153 (0.0008)
+[2026-06-06 22:23:30,057][262198] Updated weights for policy 0, policy_version 43166 (0.0012)
+[2026-06-06 22:23:32,921][261293] Fps is (10 sec: 3072.2, 60 sec: 3106.1, 300 sec: 3429.5). Total num frames: 22104064. Throughput: 0: 3279.1. Samples: 22109450. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:23:32,924][261293] Avg episode reward: [(0, '1805.035')]
+[2026-06-06 22:23:33,101][262198] Updated weights for policy 0, policy_version 43176 (0.0014)
+[2026-06-06 22:23:35,980][262198] Updated weights for policy 0, policy_version 43187 (0.0013)
+[2026-06-06 22:23:36,853][262198] Updated weights for policy 0, policy_version 43198 (0.0009)
+[2026-06-06 22:23:37,920][261293] Fps is (10 sec: 2867.2, 60 sec: 3037.9, 300 sec: 3429.5). Total num frames: 22120448. Throughput: 0: 3311.3. Samples: 22123612. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:23:37,922][261293] Avg episode reward: [(0, '1718.466')]
+[2026-06-06 22:23:38,237][262198] Updated weights for policy 0, policy_version 43208 (0.0008)
+[2026-06-06 22:23:39,590][262198] Updated weights for policy 0, policy_version 43220 (0.0008)
+[2026-06-06 22:23:41,096][262198] Updated weights for policy 0, policy_version 43233 (0.0008)
+[2026-06-06 22:23:42,702][262198] Updated weights for policy 0, policy_version 43244 (0.0009)
+[2026-06-06 22:23:42,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3276.8, 300 sec: 3450.4). Total num frames: 22142976. Throughput: 0: 3479.4. Samples: 22149970. Policy #0 lag: (min: 3.0, avg: 18.1, max: 35.0)
+[2026-06-06 22:23:42,921][261293] Avg episode reward: [(0, '1628.579')]
+[2026-06-06 22:23:43,819][262198] Updated weights for policy 0, policy_version 43254 (0.0007)
+[2026-06-06 22:23:44,710][262198] Updated weights for policy 0, policy_version 43265 (0.0008)
+[2026-06-06 22:23:46,283][262198] Updated weights for policy 0, policy_version 43275 (0.0007)
+[2026-06-06 22:23:47,645][262198] Updated weights for policy 0, policy_version 43285 (0.0008)
+[2026-06-06 22:23:47,920][261293] Fps is (10 sec: 4300.8, 60 sec: 3413.4, 300 sec: 3464.2). Total num frames: 22163456. Throughput: 0: 3580.5. Samples: 22163104. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:23:47,922][261293] Avg episode reward: [(0, '1764.867')]
+[2026-06-06 22:23:48,453][262198] Updated weights for policy 0, policy_version 43295 (0.0009)
+[2026-06-06 22:23:50,307][262198] Updated weights for policy 0, policy_version 43305 (0.0008)
+[2026-06-06 22:23:51,265][262198] Updated weights for policy 0, policy_version 43315 (0.0010)
+[2026-06-06 22:23:52,408][262198] Updated weights for policy 0, policy_version 43326 (0.0009)
+[2026-06-06 22:23:52,921][261293] Fps is (10 sec: 4300.6, 60 sec: 3618.3, 300 sec: 3485.1). Total num frames: 22185984. Throughput: 0: 3672.8. Samples: 22187156. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:23:52,923][261293] Avg episode reward: [(0, '1857.158')]
+[2026-06-06 22:23:53,764][262198] Updated weights for policy 0, policy_version 43336 (0.0007)
+[2026-06-06 22:23:55,358][262198] Updated weights for policy 0, policy_version 43346 (0.0007)
+[2026-06-06 22:23:56,196][262198] Updated weights for policy 0, policy_version 43358 (0.0008)
+[2026-06-06 22:23:57,634][262198] Updated weights for policy 0, policy_version 43368 (0.0007)
+[2026-06-06 22:23:57,921][261293] Fps is (10 sec: 4095.8, 60 sec: 3686.4, 300 sec: 3485.1). Total num frames: 22204416. Throughput: 0: 3754.1. Samples: 22212496. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:23:57,922][261293] Avg episode reward: [(0, '1757.223')]
+[2026-06-06 22:23:59,086][262198] Updated weights for policy 0, policy_version 43378 (0.0008)
+[2026-06-06 22:23:59,955][262198] Updated weights for policy 0, policy_version 43388 (0.0008)
+[2026-06-06 22:24:01,505][262198] Updated weights for policy 0, policy_version 43398 (0.0008)
+[2026-06-06 22:24:02,619][262198] Updated weights for policy 0, policy_version 43408 (0.0008)
+[2026-06-06 22:24:02,923][261293] Fps is (10 sec: 3890.2, 60 sec: 3720.4, 300 sec: 3498.9). Total num frames: 22224896. Throughput: 0: 3790.7. Samples: 22225312. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:24:02,926][261293] Avg episode reward: [(0, '1759.891')]
+[2026-06-06 22:24:03,878][262198] Updated weights for policy 0, policy_version 43419 (0.0008)
+[2026-06-06 22:24:05,372][262198] Updated weights for policy 0, policy_version 43430 (0.0007)
+[2026-06-06 22:24:06,644][262198] Updated weights for policy 0, policy_version 43440 (0.0007)
+[2026-06-06 22:24:07,835][262198] Updated weights for policy 0, policy_version 43451 (0.0010)
+[2026-06-06 22:24:07,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3822.9, 300 sec: 3512.8). Total num frames: 22247424. Throughput: 0: 3835.7. Samples: 22250086. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:24:07,922][261293] Avg episode reward: [(0, '1853.570')]
+[2026-06-06 22:24:09,402][262198] Updated weights for policy 0, policy_version 43461 (0.0010)
+[2026-06-06 22:24:10,773][262198] Updated weights for policy 0, policy_version 43471 (0.0010)
+[2026-06-06 22:24:11,887][262198] Updated weights for policy 0, policy_version 43482 (0.0009)
+[2026-06-06 22:24:12,921][261293] Fps is (10 sec: 4097.0, 60 sec: 3822.9, 300 sec: 3519.8). Total num frames: 22265856. Throughput: 0: 3865.9. Samples: 22273436. Policy #0 lag: (min: 4.0, avg: 19.6, max: 38.0)
+[2026-06-06 22:24:12,922][261293] Avg episode reward: [(0, '1750.713')]
+[2026-06-06 22:24:12,998][262198] Updated weights for policy 0, policy_version 43492 (0.0007)
+[2026-06-06 22:24:14,943][262198] Updated weights for policy 0, policy_version 43502 (0.0008)
+[2026-06-06 22:24:15,903][262198] Updated weights for policy 0, policy_version 43514 (0.0008)
+[2026-06-06 22:24:17,541][262198] Updated weights for policy 0, policy_version 43524 (0.0009)
+[2026-06-06 22:24:17,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3857.0, 300 sec: 3526.7). Total num frames: 22286336. Throughput: 0: 3920.7. Samples: 22285882. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:17,923][261293] Avg episode reward: [(0, '1729.833')]
+[2026-06-06 22:24:19,046][262198] Updated weights for policy 0, policy_version 43534 (0.0010)
+[2026-06-06 22:24:19,838][262198] Updated weights for policy 0, policy_version 43544 (0.0009)
+[2026-06-06 22:24:21,638][262115] Signal inference workers to stop experience collection... (4250 times)
+[2026-06-06 22:24:21,659][262198] InferenceWorker_p0-w0: stopping experience collection (4250 times)
+[2026-06-06 22:24:21,792][262115] Signal inference workers to resume experience collection... (4250 times)
+[2026-06-06 22:24:21,793][262198] InferenceWorker_p0-w0: resuming experience collection (4250 times)
+[2026-06-06 22:24:21,795][262198] Updated weights for policy 0, policy_version 43556 (0.0008)
+[2026-06-06 22:24:22,921][261293] Fps is (10 sec: 3891.4, 60 sec: 3857.1, 300 sec: 3533.7). Total num frames: 22304768. Throughput: 0: 4128.9. Samples: 22309414. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:22,924][261293] Avg episode reward: [(0, '1738.012')]
+[2026-06-06 22:24:23,335][262198] Updated weights for policy 0, policy_version 43566 (0.0007)
+[2026-06-06 22:24:24,278][262198] Updated weights for policy 0, policy_version 43578 (0.0008)
+[2026-06-06 22:24:26,164][262198] Updated weights for policy 0, policy_version 43592 (0.0011)
+[2026-06-06 22:24:27,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3822.9, 300 sec: 3533.7). Total num frames: 22321152. Throughput: 0: 4039.3. Samples: 22331740. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:27,924][261293] Avg episode reward: [(0, '1666.146')]
+[2026-06-06 22:24:28,448][262198] Updated weights for policy 0, policy_version 43605 (0.0009)
+[2026-06-06 22:24:29,432][262198] Updated weights for policy 0, policy_version 43616 (0.0007)
+[2026-06-06 22:24:31,122][262198] Updated weights for policy 0, policy_version 43627 (0.0008)
+[2026-06-06 22:24:32,789][262198] Updated weights for policy 0, policy_version 43638 (0.0009)
+[2026-06-06 22:24:32,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3993.6, 300 sec: 3547.6). Total num frames: 22343680. Throughput: 0: 4012.7. Samples: 22343676. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:32,923][261293] Avg episode reward: [(0, '1705.269')]
+[2026-06-06 22:24:34,562][262198] Updated weights for policy 0, policy_version 43649 (0.0009)
+[2026-06-06 22:24:36,676][262198] Updated weights for policy 0, policy_version 43661 (0.0011)
+[2026-06-06 22:24:37,662][262198] Updated weights for policy 0, policy_version 43672 (0.0010)
+[2026-06-06 22:24:37,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4027.7, 300 sec: 3547.5). Total num frames: 22362112. Throughput: 0: 3943.7. Samples: 22364622. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:37,925][261293] Avg episode reward: [(0, '1622.681')]
+[2026-06-06 22:24:39,900][262198] Updated weights for policy 0, policy_version 43682 (0.0009)
+[2026-06-06 22:24:41,184][262198] Updated weights for policy 0, policy_version 43692 (0.0010)
+[2026-06-06 22:24:42,921][261293] Fps is (10 sec: 2867.1, 60 sec: 3822.9, 300 sec: 3519.8). Total num frames: 22372352. Throughput: 0: 3778.2. Samples: 22382516. Policy #0 lag: (min: 5.0, avg: 17.0, max: 33.0)
+[2026-06-06 22:24:42,922][261293] Avg episode reward: [(0, '1758.011')]
+[2026-06-06 22:24:43,843][262198] Updated weights for policy 0, policy_version 43707 (0.0011)
+[2026-06-06 22:24:45,541][262198] Updated weights for policy 0, policy_version 43717 (0.0009)
+[2026-06-06 22:24:47,721][262198] Updated weights for policy 0, policy_version 43729 (0.0009)
+[2026-06-06 22:24:47,920][261293] Fps is (10 sec: 2867.5, 60 sec: 3788.8, 300 sec: 3519.8). Total num frames: 22390784. Throughput: 0: 3703.9. Samples: 22391978. Policy #0 lag: (min: 7.0, avg: 18.4, max: 39.0)
+[2026-06-06 22:24:47,922][261293] Avg episode reward: [(0, '1723.933')]
+[2026-06-06 22:24:49,001][262198] Updated weights for policy 0, policy_version 43742 (0.0009)
+[2026-06-06 22:24:50,512][262198] Updated weights for policy 0, policy_version 43752 (0.0009)
+[2026-06-06 22:24:52,479][262198] Updated weights for policy 0, policy_version 43765 (0.0009)
+[2026-06-06 22:24:52,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3720.5, 300 sec: 3519.8). Total num frames: 22409216. Throughput: 0: 3627.4. Samples: 22413320. Policy #0 lag: (min: 7.0, avg: 18.4, max: 39.0)
+[2026-06-06 22:24:52,923][261293] Avg episode reward: [(0, '1856.251')]
+[2026-06-06 22:24:53,642][262198] Updated weights for policy 0, policy_version 43775 (0.0007)
+[2026-06-06 22:24:55,319][262198] Updated weights for policy 0, policy_version 43785 (0.0008)
+[2026-06-06 22:24:57,038][262198] Updated weights for policy 0, policy_version 43797 (0.0008)
+[2026-06-06 22:24:57,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3754.7, 300 sec: 3519.8). Total num frames: 22429696. Throughput: 0: 3625.3. Samples: 22436574. Policy #0 lag: (min: 7.0, avg: 18.4, max: 39.0)
+[2026-06-06 22:24:57,922][261293] Avg episode reward: [(0, '1779.305')]
+[2026-06-06 22:24:58,166][262198] Updated weights for policy 0, policy_version 43809 (0.0009)
+[2026-06-06 22:24:59,416][262198] Updated weights for policy 0, policy_version 43819 (0.0008)
+[2026-06-06 22:25:00,821][262198] Updated weights for policy 0, policy_version 43832 (0.0009)
+[2026-06-06 22:25:02,156][262198] Updated weights for policy 0, policy_version 43842 (0.0008)
+[2026-06-06 22:25:02,920][261293] Fps is (10 sec: 4096.3, 60 sec: 3754.9, 300 sec: 3526.7). Total num frames: 22450176. Throughput: 0: 3634.6. Samples: 22449438. Policy #0 lag: (min: 7.0, avg: 18.4, max: 39.0)
+[2026-06-06 22:25:02,922][261293] Avg episode reward: [(0, '1593.936')]
+[2026-06-06 22:25:03,451][262198] Updated weights for policy 0, policy_version 43852 (0.0008)
+[2026-06-06 22:25:05,106][262198] Updated weights for policy 0, policy_version 43866 (0.0008)
+[2026-06-06 22:25:06,266][262198] Updated weights for policy 0, policy_version 43876 (0.0007)
+[2026-06-06 22:25:07,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3686.4, 300 sec: 3526.7). Total num frames: 22468608. Throughput: 0: 3639.7. Samples: 22473202. Policy #0 lag: (min: 7.0, avg: 18.4, max: 39.0)
+[2026-06-06 22:25:07,923][261293] Avg episode reward: [(0, '1736.082')]
+[2026-06-06 22:25:08,403][262198] Updated weights for policy 0, policy_version 43886 (0.0008)
+[2026-06-06 22:25:09,687][262198] Updated weights for policy 0, policy_version 43899 (0.0009)
+[2026-06-06 22:25:10,909][262198] Updated weights for policy 0, policy_version 43909 (0.0008)
+[2026-06-06 22:25:12,446][262198] Updated weights for policy 0, policy_version 43920 (0.0009)
+[2026-06-06 22:25:12,921][261293] Fps is (10 sec: 3891.1, 60 sec: 3720.6, 300 sec: 3526.7). Total num frames: 22489088. Throughput: 0: 3685.5. Samples: 22497586. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:12,923][261293] Avg episode reward: [(0, '1901.211')]
+[2026-06-06 22:25:14,152][262198] Updated weights for policy 0, policy_version 43933 (0.0009)
+[2026-06-06 22:25:15,472][262198] Updated weights for policy 0, policy_version 43943 (0.0009)
+[2026-06-06 22:25:17,274][262198] Updated weights for policy 0, policy_version 43954 (0.0009)
+[2026-06-06 22:25:17,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3686.4, 300 sec: 3533.7). Total num frames: 22507520. Throughput: 0: 3653.7. Samples: 22508094. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:17,922][261293] Avg episode reward: [(0, '1896.592')]
+[2026-06-06 22:25:18,585][262198] Updated weights for policy 0, policy_version 43965 (0.0012)
+[2026-06-06 22:25:19,678][262198] Updated weights for policy 0, policy_version 43975 (0.0007)
+[2026-06-06 22:25:21,723][262198] Updated weights for policy 0, policy_version 43988 (0.0009)
+[2026-06-06 22:25:22,921][261293] Fps is (10 sec: 3686.3, 60 sec: 3686.4, 300 sec: 3540.6). Total num frames: 22525952. Throughput: 0: 3688.3. Samples: 22530592. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:22,923][261293] Avg episode reward: [(0, '1853.994')]
+[2026-06-06 22:25:23,247][262198] Updated weights for policy 0, policy_version 43999 (0.0009)
+[2026-06-06 22:25:25,547][262198] Updated weights for policy 0, policy_version 44009 (0.0011)
+[2026-06-06 22:25:26,810][262198] Updated weights for policy 0, policy_version 44019 (0.0012)
+[2026-06-06 22:25:26,966][262115] Signal inference workers to stop experience collection... (4300 times)
+[2026-06-06 22:25:26,967][262115] Signal inference workers to resume experience collection... (4300 times)
+[2026-06-06 22:25:27,010][262198] InferenceWorker_p0-w0: stopping experience collection (4300 times)
+[2026-06-06 22:25:27,011][262198] InferenceWorker_p0-w0: resuming experience collection (4300 times)
+[2026-06-06 22:25:27,920][261293] Fps is (10 sec: 3481.6, 60 sec: 3686.4, 300 sec: 3533.7). Total num frames: 22542336. Throughput: 0: 3695.5. Samples: 22548812. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:27,922][261293] Avg episode reward: [(0, '1759.693')]
+[2026-06-06 22:25:28,443][262198] Updated weights for policy 0, policy_version 44029 (0.0011)
+[2026-06-06 22:25:29,874][262198] Updated weights for policy 0, policy_version 44039 (0.0011)
+[2026-06-06 22:25:32,088][262198] Updated weights for policy 0, policy_version 44050 (0.0012)
+[2026-06-06 22:25:32,923][261293] Fps is (10 sec: 3071.3, 60 sec: 3549.7, 300 sec: 3519.8). Total num frames: 22556672. Throughput: 0: 3695.0. Samples: 22558262. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:32,928][261293] Avg episode reward: [(0, '1798.351')]
+[2026-06-06 22:25:32,963][262198] Updated weights for policy 0, policy_version 44060 (0.0012)
+[2026-06-06 22:25:35,207][262198] Updated weights for policy 0, policy_version 44070 (0.0010)
+[2026-06-06 22:25:37,344][262198] Updated weights for policy 0, policy_version 44080 (0.0010)
+[2026-06-06 22:25:37,921][261293] Fps is (10 sec: 2867.2, 60 sec: 3481.6, 300 sec: 3512.8). Total num frames: 22571008. Throughput: 0: 3626.0. Samples: 22576488. Policy #0 lag: (min: 3.0, avg: 23.0, max: 39.0)
+[2026-06-06 22:25:37,923][261293] Avg episode reward: [(0, '1685.397')]
+[2026-06-06 22:25:38,179][262198] Updated weights for policy 0, policy_version 44091 (0.0012)
+[2026-06-06 22:25:40,539][262198] Updated weights for policy 0, policy_version 44101 (0.0009)
+[2026-06-06 22:25:42,780][262198] Updated weights for policy 0, policy_version 44113 (0.0013)
+[2026-06-06 22:25:42,921][261293] Fps is (10 sec: 2867.8, 60 sec: 3549.9, 300 sec: 3505.9). Total num frames: 22585344. Throughput: 0: 3517.6. Samples: 22594866. Policy #0 lag: (min: 0.0, avg: 20.7, max: 34.0)
+[2026-06-06 22:25:42,924][261293] Avg episode reward: [(0, '1478.726')]
+[2026-06-06 22:25:44,442][262198] Updated weights for policy 0, policy_version 44125 (0.0009)
+[2026-06-06 22:25:45,517][262198] Updated weights for policy 0, policy_version 44135 (0.0008)
+[2026-06-06 22:25:47,319][262198] Updated weights for policy 0, policy_version 44145 (0.0008)
+[2026-06-06 22:25:47,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3584.0, 300 sec: 3519.8). Total num frames: 22605824. Throughput: 0: 3468.1. Samples: 22605506. Policy #0 lag: (min: 0.0, avg: 20.7, max: 34.0)
+[2026-06-06 22:25:47,923][261293] Avg episode reward: [(0, '1698.887')]
+[2026-06-06 22:25:48,096][262198] Updated weights for policy 0, policy_version 44155 (0.0008)
+[2026-06-06 22:25:49,893][262198] Updated weights for policy 0, policy_version 44165 (0.0010)
+[2026-06-06 22:25:51,672][262198] Updated weights for policy 0, policy_version 44176 (0.0009)
+[2026-06-06 22:25:52,648][262198] Updated weights for policy 0, policy_version 44187 (0.0009)
+[2026-06-06 22:25:52,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3584.0, 300 sec: 3547.6). Total num frames: 22624256. Throughput: 0: 3431.6. Samples: 22627622. Policy #0 lag: (min: 0.0, avg: 20.7, max: 34.0)
+[2026-06-06 22:25:52,924][261293] Avg episode reward: [(0, '1751.562')]
+[2026-06-06 22:25:54,178][262198] Updated weights for policy 0, policy_version 44197 (0.0009)
+[2026-06-06 22:25:55,703][262198] Updated weights for policy 0, policy_version 44208 (0.0008)
+[2026-06-06 22:25:56,618][262198] Updated weights for policy 0, policy_version 44219 (0.0008)
+[2026-06-06 22:25:57,921][261293] Fps is (10 sec: 3891.2, 60 sec: 3584.0, 300 sec: 3568.4). Total num frames: 22644736. Throughput: 0: 3430.0. Samples: 22651938. Policy #0 lag: (min: 0.0, avg: 20.7, max: 34.0)
+[2026-06-06 22:25:57,923][261293] Avg episode reward: [(0, '1715.082')]
+[2026-06-06 22:25:58,289][262198] Updated weights for policy 0, policy_version 44230 (0.0008)
+[2026-06-06 22:25:59,683][262198] Updated weights for policy 0, policy_version 44242 (0.0010)
+[2026-06-06 22:26:00,860][262198] Updated weights for policy 0, policy_version 44252 (0.0008)
+[2026-06-06 22:26:02,528][262198] Updated weights for policy 0, policy_version 44263 (0.0008)
+[2026-06-06 22:26:02,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3549.9, 300 sec: 3568.4). Total num frames: 22663168. Throughput: 0: 3456.7. Samples: 22663644. Policy #0 lag: (min: 0.0, avg: 20.7, max: 34.0)
+[2026-06-06 22:26:02,921][261293] Avg episode reward: [(0, '1503.547')]
+[2026-06-06 22:26:03,896][262198] Updated weights for policy 0, policy_version 44274 (0.0009)
+[2026-06-06 22:26:04,787][262198] Updated weights for policy 0, policy_version 44284 (0.0009)
+[2026-06-06 22:26:06,207][262198] Updated weights for policy 0, policy_version 44295 (0.0007)
+[2026-06-06 22:26:07,706][262198] Updated weights for policy 0, policy_version 44309 (0.0008)
+[2026-06-06 22:26:07,921][261293] Fps is (10 sec: 4300.7, 60 sec: 3652.3, 300 sec: 3582.3). Total num frames: 22687744. Throughput: 0: 3536.2. Samples: 22689722. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:07,924][261293] Avg episode reward: [(0, '1806.536')]
+[2026-06-06 22:26:08,985][262198] Updated weights for policy 0, policy_version 44320 (0.0007)
+[2026-06-06 22:26:10,344][262198] Updated weights for policy 0, policy_version 44330 (0.0007)
+[2026-06-06 22:26:11,489][262198] Updated weights for policy 0, policy_version 44344 (0.0008)
+[2026-06-06 22:26:12,861][262198] Updated weights for policy 0, policy_version 44354 (0.0009)
+[2026-06-06 22:26:12,921][261293] Fps is (10 sec: 4505.3, 60 sec: 3652.2, 300 sec: 3575.3). Total num frames: 22708224. Throughput: 0: 3713.5. Samples: 22715922. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:12,924][261293] Avg episode reward: [(0, '1832.237')]
+[2026-06-06 22:26:14,209][262198] Updated weights for policy 0, policy_version 44364 (0.0007)
+[2026-06-06 22:26:15,506][262198] Updated weights for policy 0, policy_version 44376 (0.0011)
+[2026-06-06 22:26:17,196][262198] Updated weights for policy 0, policy_version 44386 (0.0007)
+[2026-06-06 22:26:17,921][261293] Fps is (10 sec: 4096.3, 60 sec: 3686.4, 300 sec: 3582.3). Total num frames: 22728704. Throughput: 0: 3764.8. Samples: 22727668. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:17,925][261293] Avg episode reward: [(0, '1845.404')]
+[2026-06-06 22:26:18,460][262198] Updated weights for policy 0, policy_version 44396 (0.0008)
+[2026-06-06 22:26:19,552][262198] Updated weights for policy 0, policy_version 44408 (0.0008)
+[2026-06-06 22:26:20,856][262198] Updated weights for policy 0, policy_version 44418 (0.0007)
+[2026-06-06 22:26:22,218][262198] Updated weights for policy 0, policy_version 44428 (0.0007)
+[2026-06-06 22:26:22,921][261293] Fps is (10 sec: 4096.2, 60 sec: 3720.5, 300 sec: 3568.4). Total num frames: 22749184. Throughput: 0: 3933.0. Samples: 22753472. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:22,923][261293] Avg episode reward: [(0, '1781.283')]
+[2026-06-06 22:26:23,850][262198] Updated weights for policy 0, policy_version 44441 (0.0007)
+[2026-06-06 22:26:24,838][262198] Updated weights for policy 0, policy_version 44451 (0.0008)
+[2026-06-06 22:26:26,634][262198] Updated weights for policy 0, policy_version 44461 (0.0007)
+[2026-06-06 22:26:27,531][262198] Updated weights for policy 0, policy_version 44472 (0.0009)
+[2026-06-06 22:26:27,921][261293] Fps is (10 sec: 4300.8, 60 sec: 3822.9, 300 sec: 3568.4). Total num frames: 22771712. Throughput: 0: 4084.5. Samples: 22778668. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:27,924][261293] Avg episode reward: [(0, '1592.786')]
+[2026-06-06 22:26:28,729][262198] Updated weights for policy 0, policy_version 44482 (0.0007)
+[2026-06-06 22:26:30,002][262198] Updated weights for policy 0, policy_version 44492 (0.0007)
+[2026-06-06 22:26:31,943][262198] Updated weights for policy 0, policy_version 44507 (0.0009)
+[2026-06-06 22:26:32,920][261293] Fps is (10 sec: 4300.9, 60 sec: 3925.5, 300 sec: 3575.3). Total num frames: 22792192. Throughput: 0: 4128.0. Samples: 22791266. Policy #0 lag: (min: 3.0, avg: 18.9, max: 36.0)
+[2026-06-06 22:26:32,922][261293] Avg episode reward: [(0, '1824.021')]
+[2026-06-06 22:26:33,256][262198] Updated weights for policy 0, policy_version 44517 (0.0007)
+[2026-06-06 22:26:33,976][262115] Signal inference workers to stop experience collection... (4350 times)
+[2026-06-06 22:26:33,977][262115] Signal inference workers to resume experience collection... (4350 times)
+[2026-06-06 22:26:33,998][262198] InferenceWorker_p0-w0: stopping experience collection (4350 times)
+[2026-06-06 22:26:33,999][262198] InferenceWorker_p0-w0: resuming experience collection (4350 times)
+[2026-06-06 22:26:34,308][262198] Updated weights for policy 0, policy_version 44528 (0.0007)
+[2026-06-06 22:26:35,546][262198] Updated weights for policy 0, policy_version 44538 (0.0007)
+[2026-06-06 22:26:36,563][262198] Updated weights for policy 0, policy_version 44548 (0.0008)
+[2026-06-06 22:26:37,920][261293] Fps is (10 sec: 4096.0, 60 sec: 4027.7, 300 sec: 3561.4). Total num frames: 22812672. Throughput: 0: 4203.9. Samples: 22816798. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:26:37,922][261293] Avg episode reward: [(0, '1696.721')]
+[2026-06-06 22:26:38,063][262198] Updated weights for policy 0, policy_version 44560 (0.0009)
+[2026-06-06 22:26:39,254][262198] Updated weights for policy 0, policy_version 44571 (0.0009)
+[2026-06-06 22:26:40,864][262198] Updated weights for policy 0, policy_version 44581 (0.0008)
+[2026-06-06 22:26:42,017][262198] Updated weights for policy 0, policy_version 44592 (0.0007)
+[2026-06-06 22:26:42,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4164.3, 300 sec: 3561.4). Total num frames: 22835200. Throughput: 0: 4240.3. Samples: 22842754. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:26:42,923][261293] Avg episode reward: [(0, '1721.715')]
+[2026-06-06 22:26:43,093][262198] Updated weights for policy 0, policy_version 44603 (0.0008)
+[2026-06-06 22:26:44,773][262198] Updated weights for policy 0, policy_version 44613 (0.0008)
+[2026-06-06 22:26:45,915][262198] Updated weights for policy 0, policy_version 44624 (0.0008)
+[2026-06-06 22:26:47,508][262198] Updated weights for policy 0, policy_version 44636 (0.0009)
+[2026-06-06 22:26:47,920][261293] Fps is (10 sec: 4300.8, 60 sec: 4164.3, 300 sec: 3561.4). Total num frames: 22855680. Throughput: 0: 4260.9. Samples: 22855386. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:26:47,923][261293] Avg episode reward: [(0, '1654.562')]
+[2026-06-06 22:26:48,958][262198] Updated weights for policy 0, policy_version 44646 (0.0008)
+[2026-06-06 22:26:50,127][262198] Updated weights for policy 0, policy_version 44657 (0.0009)
+[2026-06-06 22:26:51,268][262198] Updated weights for policy 0, policy_version 44667 (0.0009)
+[2026-06-06 22:26:52,921][261293] Fps is (10 sec: 3891.3, 60 sec: 4164.3, 300 sec: 3547.5). Total num frames: 22874112. Throughput: 0: 4196.6. Samples: 22878566. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:26:52,923][261293] Avg episode reward: [(0, '1670.928')]
+[2026-06-06 22:26:53,381][262198] Updated weights for policy 0, policy_version 44677 (0.0009)
+[2026-06-06 22:26:54,494][262198] Updated weights for policy 0, policy_version 44688 (0.0009)
+[2026-06-06 22:26:55,748][262198] Updated weights for policy 0, policy_version 44698 (0.0009)
+[2026-06-06 22:26:56,825][262198] Updated weights for policy 0, policy_version 44708 (0.0011)
+[2026-06-06 22:26:57,921][261293] Fps is (10 sec: 3686.3, 60 sec: 4130.1, 300 sec: 3540.6). Total num frames: 22892544. Throughput: 0: 4121.6. Samples: 22901394. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:26:57,922][261293] Avg episode reward: [(0, '1696.077')]
+[2026-06-06 22:26:58,599][262198] Updated weights for policy 0, policy_version 44720 (0.0008)
+[2026-06-06 22:27:00,110][262198] Updated weights for policy 0, policy_version 44732 (0.0008)
+[2026-06-06 22:27:01,536][262198] Updated weights for policy 0, policy_version 44742 (0.0008)
+[2026-06-06 22:27:02,809][262198] Updated weights for policy 0, policy_version 44752 (0.0010)
+[2026-06-06 22:27:02,920][261293] Fps is (10 sec: 3891.4, 60 sec: 4164.3, 300 sec: 3568.4). Total num frames: 22913024. Throughput: 0: 4126.6. Samples: 22913366. Policy #0 lag: (min: 3.0, avg: 23.0, max: 43.0)
+[2026-06-06 22:27:02,921][261293] Avg episode reward: [(0, '1883.668')]
+[2026-06-06 22:27:04,568][262198] Updated weights for policy 0, policy_version 44763 (0.0010)
+[2026-06-06 22:27:06,431][262198] Updated weights for policy 0, policy_version 44773 (0.0012)
+[2026-06-06 22:27:07,823][262198] Updated weights for policy 0, policy_version 44783 (0.0011)
+[2026-06-06 22:27:07,921][261293] Fps is (10 sec: 3686.4, 60 sec: 4027.8, 300 sec: 3582.3). Total num frames: 22929408. Throughput: 0: 4025.2. Samples: 22934604. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:07,923][261293] Avg episode reward: [(0, '1816.826')]
+[2026-06-06 22:27:09,526][262198] Updated weights for policy 0, policy_version 44793 (0.0012)
+[2026-06-06 22:27:10,995][262198] Updated weights for policy 0, policy_version 44803 (0.0012)
+[2026-06-06 22:27:12,922][261293] Fps is (10 sec: 3071.5, 60 sec: 3925.3, 300 sec: 3603.1). Total num frames: 22943744. Throughput: 0: 3864.4. Samples: 22952572. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:12,927][261293] Avg episode reward: [(0, '1821.010')]
+[2026-06-06 22:27:12,977][262198] Updated weights for policy 0, policy_version 44813 (0.0012)
+[2026-06-06 22:27:13,923][262198] Updated weights for policy 0, policy_version 44824 (0.0012)
+[2026-06-06 22:27:16,441][262198] Updated weights for policy 0, policy_version 44834 (0.0010)
+[2026-06-06 22:27:17,920][261293] Fps is (10 sec: 2867.3, 60 sec: 3822.9, 300 sec: 3610.0). Total num frames: 22958080. Throughput: 0: 3780.4. Samples: 22961382. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:17,923][261293] Avg episode reward: [(0, '1645.662')]
+[2026-06-06 22:27:18,266][262198] Updated weights for policy 0, policy_version 44844 (0.0011)
+[2026-06-06 22:27:19,481][262198] Updated weights for policy 0, policy_version 44855 (0.0011)
+[2026-06-06 22:27:21,987][262198] Updated weights for policy 0, policy_version 44865 (0.0011)
+[2026-06-06 22:27:22,921][261293] Fps is (10 sec: 2867.6, 60 sec: 3720.5, 300 sec: 3630.9). Total num frames: 22972416. Throughput: 0: 3604.0. Samples: 22978976. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:22,923][261293] Avg episode reward: [(0, '1649.526')]
+[2026-06-06 22:27:23,568][262198] Updated weights for policy 0, policy_version 44875 (0.0010)
+[2026-06-06 22:27:24,550][262198] Updated weights for policy 0, policy_version 44886 (0.0012)
+[2026-06-06 22:27:26,343][262198] Updated weights for policy 0, policy_version 44896 (0.0012)
+[2026-06-06 22:27:27,923][261293] Fps is (10 sec: 3071.2, 60 sec: 3618.0, 300 sec: 3630.8). Total num frames: 22988800. Throughput: 0: 3457.8. Samples: 22998360. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:27,927][261293] Avg episode reward: [(0, '1767.915')]
+[2026-06-06 22:27:28,732][262198] Updated weights for policy 0, policy_version 44906 (0.0011)
+[2026-06-06 22:27:29,645][262198] Updated weights for policy 0, policy_version 44916 (0.0011)
+[2026-06-06 22:27:31,672][262198] Updated weights for policy 0, policy_version 44926 (0.0010)
+[2026-06-06 22:27:32,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3549.8, 300 sec: 3617.0). Total num frames: 23005184. Throughput: 0: 3370.5. Samples: 23007060. Policy #0 lag: (min: 3.0, avg: 22.8, max: 43.0)
+[2026-06-06 22:27:32,925][261293] Avg episode reward: [(0, '1672.944')]
+[2026-06-06 22:27:33,178][262198] Updated weights for policy 0, policy_version 44936 (0.0009)
+[2026-06-06 22:27:34,536][262198] Updated weights for policy 0, policy_version 44948 (0.0008)
+[2026-06-06 22:27:35,881][262198] Updated weights for policy 0, policy_version 44958 (0.0008)
+[2026-06-06 22:27:37,415][262198] Updated weights for policy 0, policy_version 44968 (0.0009)
+[2026-06-06 22:27:37,920][261293] Fps is (10 sec: 3687.4, 60 sec: 3549.9, 300 sec: 3658.6). Total num frames: 23025664. Throughput: 0: 3347.2. Samples: 23029188. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:27:37,921][261293] Avg episode reward: [(0, '1875.577')]
+[2026-06-06 22:27:38,673][262198] Updated weights for policy 0, policy_version 44980 (0.0010)
+[2026-06-06 22:27:39,945][262198] Updated weights for policy 0, policy_version 44990 (0.0008)
+[2026-06-06 22:27:41,403][262198] Updated weights for policy 0, policy_version 45000 (0.0007)
+[2026-06-06 22:27:42,920][261293] Fps is (10 sec: 3891.3, 60 sec: 3481.6, 300 sec: 3679.5). Total num frames: 23044096. Throughput: 0: 3391.5. Samples: 23054010. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:27:42,922][261293] Avg episode reward: [(0, '1775.875')]
+[2026-06-06 22:27:43,319][262198] Updated weights for policy 0, policy_version 45016 (0.0010)
+[2026-06-06 22:27:44,896][262198] Updated weights for policy 0, policy_version 45026 (0.0008)
+[2026-06-06 22:27:46,333][262198] Updated weights for policy 0, policy_version 45036 (0.0008)
+[2026-06-06 22:27:46,885][262115] Signal inference workers to stop experience collection... (4400 times)
+[2026-06-06 22:27:46,910][262198] InferenceWorker_p0-w0: stopping experience collection (4400 times)
+[2026-06-06 22:27:47,003][262115] Signal inference workers to resume experience collection... (4400 times)
+[2026-06-06 22:27:47,004][262198] InferenceWorker_p0-w0: resuming experience collection (4400 times)
+[2026-06-06 22:27:47,241][262198] Updated weights for policy 0, policy_version 45048 (0.0007)
+[2026-06-06 22:27:47,920][261293] Fps is (10 sec: 4096.0, 60 sec: 3515.7, 300 sec: 3721.2). Total num frames: 23066624. Throughput: 0: 3374.0. Samples: 23065198. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:27:47,921][261293] Avg episode reward: [(0, '1854.535')]
+[2026-06-06 22:27:48,947][262198] Updated weights for policy 0, policy_version 45058 (0.0008)
+[2026-06-06 22:27:50,753][262198] Updated weights for policy 0, policy_version 45072 (0.0008)
+[2026-06-06 22:27:52,055][262198] Updated weights for policy 0, policy_version 45084 (0.0007)
+[2026-06-06 22:27:52,921][261293] Fps is (10 sec: 4095.9, 60 sec: 3515.7, 300 sec: 3735.0). Total num frames: 23085056. Throughput: 0: 3424.2. Samples: 23088692. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:27:52,922][261293] Avg episode reward: [(0, '1693.611')]
+[2026-06-06 22:27:53,568][262198] Updated weights for policy 0, policy_version 45094 (0.0008)
+[2026-06-06 22:27:55,338][262198] Updated weights for policy 0, policy_version 45105 (0.0009)
+[2026-06-06 22:27:56,494][262198] Updated weights for policy 0, policy_version 45117 (0.0011)
+[2026-06-06 22:27:57,735][262198] Updated weights for policy 0, policy_version 45127 (0.0010)
+[2026-06-06 22:27:57,921][261293] Fps is (10 sec: 3891.0, 60 sec: 3549.9, 300 sec: 3741.9). Total num frames: 23105536. Throughput: 0: 3567.2. Samples: 23113092. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:27:57,924][261293] Avg episode reward: [(0, '1882.761')]
+[2026-06-06 22:27:59,281][262198] Updated weights for policy 0, policy_version 45138 (0.0008)
+[2026-06-06 22:28:00,476][262198] Updated weights for policy 0, policy_version 45149 (0.0009)
+[2026-06-06 22:28:02,046][262198] Updated weights for policy 0, policy_version 45159 (0.0008)
+[2026-06-06 22:28:02,920][261293] Fps is (10 sec: 3891.2, 60 sec: 3515.7, 300 sec: 3748.9). Total num frames: 23123968. Throughput: 0: 3628.6. Samples: 23124670. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:02,922][261293] Avg episode reward: [(0, '1908.477')]
+[2026-06-06 22:28:03,459][262198] Updated weights for policy 0, policy_version 45170 (0.0008)
+[2026-06-06 22:28:04,607][262198] Updated weights for policy 0, policy_version 45181 (0.0008)
+[2026-06-06 22:28:05,621][262198] Updated weights for policy 0, policy_version 45191 (0.0010)
+[2026-06-06 22:28:06,976][262198] Updated weights for policy 0, policy_version 45206 (0.0008)
+[2026-06-06 22:28:07,921][261293] Fps is (10 sec: 4300.9, 60 sec: 3652.3, 300 sec: 3769.7). Total num frames: 23148544. Throughput: 0: 3812.9. Samples: 23150556. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:07,922][261293] Avg episode reward: [(0, '1702.766')]
+[2026-06-06 22:28:08,049][262198] Updated weights for policy 0, policy_version 45216 (0.0007)
+[2026-06-06 22:28:09,650][262198] Updated weights for policy 0, policy_version 45226 (0.0008)
+[2026-06-06 22:28:11,162][262198] Updated weights for policy 0, policy_version 45243 (0.0010)
+[2026-06-06 22:28:12,595][262198] Updated weights for policy 0, policy_version 45253 (0.0009)
+[2026-06-06 22:28:12,920][261293] Fps is (10 sec: 4710.6, 60 sec: 3788.9, 300 sec: 3783.6). Total num frames: 23171072. Throughput: 0: 4000.5. Samples: 23178374. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:12,922][261293] Avg episode reward: [(0, '1718.104')]
+[2026-06-06 22:28:13,870][262198] Updated weights for policy 0, policy_version 45265 (0.0010)
+[2026-06-06 22:28:14,735][262198] Updated weights for policy 0, policy_version 45275 (0.0010)
+[2026-06-06 22:28:16,107][262198] Updated weights for policy 0, policy_version 45285 (0.0010)
+[2026-06-06 22:28:17,450][262198] Updated weights for policy 0, policy_version 45300 (0.0009)
+[2026-06-06 22:28:17,921][261293] Fps is (10 sec: 4710.2, 60 sec: 3959.4, 300 sec: 3804.4). Total num frames: 23195648. Throughput: 0: 4115.8. Samples: 23192270. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:17,923][261293] Avg episode reward: [(0, '1694.395')]
+[2026-06-06 22:28:18,332][262198] Updated weights for policy 0, policy_version 45310 (0.0008)
+[2026-06-06 22:28:19,557][262198] Updated weights for policy 0, policy_version 45320 (0.0007)
+[2026-06-06 22:28:22,176][262198] Updated weights for policy 0, policy_version 45345 (0.0035)
+[2026-06-06 22:28:22,923][261293] Fps is (10 sec: 4913.9, 60 sec: 4130.0, 300 sec: 3825.2). Total num frames: 23220224. Throughput: 0: 4310.4. Samples: 23223166. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:22,925][261293] Avg episode reward: [(0, '1733.846')]
+[2026-06-06 22:28:23,249][262198] Updated weights for policy 0, policy_version 45355 (0.0007)
+[2026-06-06 22:28:24,064][262198] Updated weights for policy 0, policy_version 45365 (0.0007)
+[2026-06-06 22:28:24,754][262198] Updated weights for policy 0, policy_version 45375 (0.0006)
+[2026-06-06 22:28:26,199][262198] Updated weights for policy 0, policy_version 45385 (0.0007)
+[2026-06-06 22:28:26,711][262198] Updated weights for policy 0, policy_version 45395 (0.0007)
+[2026-06-06 22:28:27,808][262198] Updated weights for policy 0, policy_version 45405 (0.0009)
+[2026-06-06 22:28:27,921][261293] Fps is (10 sec: 5324.6, 60 sec: 4335.1, 300 sec: 3880.8). Total num frames: 23248896. Throughput: 0: 4438.9. Samples: 23253762. Policy #0 lag: (min: 2.0, avg: 20.4, max: 38.0)
+[2026-06-06 22:28:27,925][261293] Avg episode reward: [(0, '1782.301')]
+[2026-06-06 22:28:28,724][262198] Updated weights for policy 0, policy_version 45415 (0.0007)
+[2026-06-06 22:28:29,958][262198] Updated weights for policy 0, policy_version 45425 (0.0009)
+[2026-06-06 22:28:30,740][262198] Updated weights for policy 0, policy_version 45435 (0.0009)
+[2026-06-06 22:28:32,742][262198] Updated weights for policy 0, policy_version 45445 (0.0010)
+[2026-06-06 22:28:32,920][261293] Fps is (10 sec: 4916.4, 60 sec: 4403.2, 300 sec: 3894.7). Total num frames: 23269376. Throughput: 0: 4517.0. Samples: 23268462. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:32,922][261293] Avg episode reward: [(0, '1700.969')]
+[2026-06-06 22:28:34,086][262198] Updated weights for policy 0, policy_version 45455 (0.0011)
+[2026-06-06 22:28:35,063][262198] Updated weights for policy 0, policy_version 45465 (0.0011)
+[2026-06-06 22:28:36,538][262198] Updated weights for policy 0, policy_version 45475 (0.0009)
+[2026-06-06 22:28:37,921][261293] Fps is (10 sec: 3891.5, 60 sec: 4369.0, 300 sec: 3880.8). Total num frames: 23287808. Throughput: 0: 4498.4. Samples: 23291118. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:37,922][261293] Avg episode reward: [(0, '1892.507')]
+[2026-06-06 22:28:38,268][262198] Updated weights for policy 0, policy_version 45485 (0.0010)
+[2026-06-06 22:28:39,362][262198] Updated weights for policy 0, policy_version 45496 (0.0011)
+[2026-06-06 22:28:41,602][262198] Updated weights for policy 0, policy_version 45507 (0.0009)
+[2026-06-06 22:28:42,921][261293] Fps is (10 sec: 3276.8, 60 sec: 4300.8, 300 sec: 3860.0). Total num frames: 23302144. Throughput: 0: 4398.9. Samples: 23311040. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:42,923][261293] Avg episode reward: [(0, '1649.107')]
+[2026-06-06 22:28:43,419][262198] Updated weights for policy 0, policy_version 45517 (0.0013)
+[2026-06-06 22:28:44,241][262198] Updated weights for policy 0, policy_version 45527 (0.0010)
+[2026-06-06 22:28:46,096][262198] Updated weights for policy 0, policy_version 45537 (0.0010)
+[2026-06-06 22:28:47,681][262198] Updated weights for policy 0, policy_version 45547 (0.0010)
+[2026-06-06 22:28:47,921][261293] Fps is (10 sec: 3276.7, 60 sec: 4232.5, 300 sec: 3846.1). Total num frames: 23320576. Throughput: 0: 4386.9. Samples: 23322080. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:47,924][261293] Avg episode reward: [(0, '1894.164')]
+[2026-06-06 22:28:48,754][262198] Updated weights for policy 0, policy_version 45557 (0.0011)
+[2026-06-06 22:28:49,998][262198] Updated weights for policy 0, policy_version 45567 (0.0010)
+[2026-06-06 22:28:52,563][262198] Updated weights for policy 0, policy_version 45577 (0.0010)
+[2026-06-06 22:28:52,921][261293] Fps is (10 sec: 3481.6, 60 sec: 4198.4, 300 sec: 3839.1). Total num frames: 23336960. Throughput: 0: 4262.7. Samples: 23342376. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:52,923][261293] Avg episode reward: [(0, '1782.211')]
+[2026-06-06 22:28:53,292][262198] Updated weights for policy 0, policy_version 45587 (0.0013)
+[2026-06-06 22:28:55,145][262198] Updated weights for policy 0, policy_version 45598 (0.0014)
+[2026-06-06 22:28:57,920][261293] Fps is (10 sec: 3072.1, 60 sec: 4096.0, 300 sec: 3818.3). Total num frames: 23351296. Throughput: 0: 4057.9. Samples: 23360982. Policy #0 lag: (min: 3.0, avg: 22.0, max: 39.0)
+[2026-06-06 22:28:57,922][261293] Avg episode reward: [(0, '1866.716')]
+[2026-06-06 22:28:58,054][262198] Updated weights for policy 0, policy_version 45609 (0.0009)
+[2026-06-06 22:28:59,001][262198] Updated weights for policy 0, policy_version 45619 (0.0010)
+[2026-06-06 22:29:00,568][262198] Updated weights for policy 0, policy_version 45629 (0.0010)
+[2026-06-06 22:29:02,272][262198] Updated weights for policy 0, policy_version 45639 (0.0011)
+[2026-06-06 22:29:02,920][261293] Fps is (10 sec: 3072.0, 60 sec: 4061.9, 300 sec: 3797.5). Total num frames: 23367680. Throughput: 0: 3948.9. Samples: 23369970. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:02,922][261293] Avg episode reward: [(0, '1673.192')]
+[2026-06-06 22:29:04,356][262198] Updated weights for policy 0, policy_version 45649 (0.0010)
+[2026-06-06 22:29:05,085][262198] Updated weights for policy 0, policy_version 45659 (0.0008)
+[2026-06-06 22:29:07,447][262198] Updated weights for policy 0, policy_version 45670 (0.0010)
+[2026-06-06 22:29:07,921][261293] Fps is (10 sec: 3276.8, 60 sec: 3925.3, 300 sec: 3790.5). Total num frames: 23384064. Throughput: 0: 3676.3. Samples: 23388590. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:07,922][261293] Avg episode reward: [(0, '1762.101')]
+[2026-06-06 22:29:09,454][262198] Updated weights for policy 0, policy_version 45684 (0.0012)
+[2026-06-06 22:29:11,146][262198] Updated weights for policy 0, policy_version 45694 (0.0011)
+[2026-06-06 22:29:12,650][262198] Updated weights for policy 0, policy_version 45704 (0.0009)
+[2026-06-06 22:29:12,920][261293] Fps is (10 sec: 3276.9, 60 sec: 3822.9, 300 sec: 3776.7). Total num frames: 23400448. Throughput: 0: 3419.4. Samples: 23407632. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:12,921][261293] Avg episode reward: [(0, '1749.575')]
+[2026-06-06 22:29:14,661][262198] Updated weights for policy 0, policy_version 45714 (0.0009)
+[2026-06-06 22:29:15,517][262198] Updated weights for policy 0, policy_version 45724 (0.0011)
+[2026-06-06 22:29:17,920][261293] Fps is (10 sec: 3072.0, 60 sec: 3652.3, 300 sec: 3762.8). Total num frames: 23414784. Throughput: 0: 3302.0. Samples: 23417054. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:17,922][261293] Avg episode reward: [(0, '1636.130')]
+[2026-06-06 22:29:17,995][262198] Updated weights for policy 0, policy_version 45734 (0.0009)
+[2026-06-06 22:29:18,027][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000045736_23416832.pth...
+[2026-06-06 22:29:18,096][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000025464_13037568.pth
+[2026-06-06 22:29:19,562][262198] Updated weights for policy 0, policy_version 45745 (0.0010)
+[2026-06-06 22:29:20,217][262198] Updated weights for policy 0, policy_version 45755 (0.0009)
+[2026-06-06 22:29:22,145][262198] Updated weights for policy 0, policy_version 45767 (0.0010)
+[2026-06-06 22:29:22,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3584.2, 300 sec: 3776.7). Total num frames: 23435264. Throughput: 0: 3278.5. Samples: 23438650. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:22,923][261293] Avg episode reward: [(0, '1725.016')]
+[2026-06-06 22:29:23,271][262198] Updated weights for policy 0, policy_version 45777 (0.0008)
+[2026-06-06 22:29:24,540][262198] Updated weights for policy 0, policy_version 45790 (0.0010)
+[2026-06-06 22:29:25,643][262198] Updated weights for policy 0, policy_version 45800 (0.0008)
+[2026-06-06 22:29:26,976][262198] Updated weights for policy 0, policy_version 45810 (0.0009)
+[2026-06-06 22:29:27,590][262198] Updated weights for policy 0, policy_version 45820 (0.0009)
+[2026-06-06 22:29:27,921][261293] Fps is (10 sec: 4505.5, 60 sec: 3515.8, 300 sec: 3783.6). Total num frames: 23459840. Throughput: 0: 3434.1. Samples: 23465574. Policy #0 lag: (min: 6.0, avg: 20.5, max: 38.0)
+[2026-06-06 22:29:27,924][261293] Avg episode reward: [(0, '1683.084')]
+[2026-06-06 22:29:29,406][262198] Updated weights for policy 0, policy_version 45832 (0.0009)
+[2026-06-06 22:29:30,852][262198] Updated weights for policy 0, policy_version 45842 (0.0010)
+[2026-06-06 22:29:32,252][262198] Updated weights for policy 0, policy_version 45855 (0.0012)
+[2026-06-06 22:29:32,920][261293] Fps is (10 sec: 4300.8, 60 sec: 3481.6, 300 sec: 3783.6). Total num frames: 23478272. Throughput: 0: 3483.6. Samples: 23478842. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 22:29:32,921][261293] Avg episode reward: [(0, '1725.338')]
+[2026-06-06 22:29:34,761][262198] Updated weights for policy 0, policy_version 45865 (0.0009)
+[2026-06-06 22:29:35,420][262198] Updated weights for policy 0, policy_version 45875 (0.0010)
+[2026-06-06 22:29:36,817][262198] Updated weights for policy 0, policy_version 45885 (0.0009)
+[2026-06-06 22:29:37,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3447.5, 300 sec: 3804.4). Total num frames: 23494656. Throughput: 0: 3492.2. Samples: 23499528. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 22:29:37,924][261293] Avg episode reward: [(0, '1799.830')]
+[2026-06-06 22:29:38,436][262198] Updated weights for policy 0, policy_version 45895 (0.0010)
+[2026-06-06 22:29:40,147][262198] Updated weights for policy 0, policy_version 45905 (0.0010)
+[2026-06-06 22:29:41,035][262198] Updated weights for policy 0, policy_version 45916 (0.0009)
+[2026-06-06 22:29:42,921][261293] Fps is (10 sec: 3481.5, 60 sec: 3515.7, 300 sec: 3804.4). Total num frames: 23513088. Throughput: 0: 3533.6. Samples: 23519996. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 22:29:42,925][261293] Avg episode reward: [(0, '1761.575')]
+[2026-06-06 22:29:43,225][262198] Updated weights for policy 0, policy_version 45926 (0.0009)
+[2026-06-06 22:29:44,988][262198] Updated weights for policy 0, policy_version 45937 (0.0009)
+[2026-06-06 22:29:45,799][262198] Updated weights for policy 0, policy_version 45948 (0.0009)
+[2026-06-06 22:29:47,921][261293] Fps is (10 sec: 3481.6, 60 sec: 3481.6, 300 sec: 3797.5). Total num frames: 23529472. Throughput: 0: 3568.3. Samples: 23530542. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 22:29:47,923][261293] Avg episode reward: [(0, '1685.914')]
+[2026-06-06 22:29:47,950][262198] Updated weights for policy 0, policy_version 45958 (0.0009)
+[2026-06-06 22:29:49,144][262198] Updated weights for policy 0, policy_version 45968 (0.0007)
+[2026-06-06 22:29:49,849][262198] Updated weights for policy 0, policy_version 45978 (0.0009)
+[2026-06-06 22:29:51,163][262198] Updated weights for policy 0, policy_version 45988 (0.0010)
+[2026-06-06 22:29:52,423][262198] Updated weights for policy 0, policy_version 45999 (0.0008)
+[2026-06-06 22:29:52,921][261293] Fps is (10 sec: 4096.1, 60 sec: 3618.1, 300 sec: 3811.4). Total num frames: 23554048. Throughput: 0: 3704.6. Samples: 23555298. Policy #0 lag: (min: 3.0, avg: 21.4, max: 35.0)
+[2026-06-06 22:29:52,923][261293] Avg episode reward: [(0, '1642.591')]
+[2026-06-06 22:29:53,318][262198] Updated weights for policy 0, policy_version 46009 (0.0007)
+[2026-06-06 22:29:54,437][262198] Updated weights for policy 0, policy_version 46019 (0.0009)
+[2026-06-06 22:29:55,773][262198] Updated weights for policy 0, policy_version 46031 (0.0008)
+[2026-06-06 22:29:56,550][262115] Signal inference workers to stop experience collection... (4450 times)
+[2026-06-06 22:29:56,576][262198] InferenceWorker_p0-w0: stopping experience collection (4450 times)
+[2026-06-06 22:29:56,661][262115] Signal inference workers to resume experience collection... (4450 times)
+[2026-06-06 22:29:56,662][262198] InferenceWorker_p0-w0: resuming experience collection (4450 times)
+[2026-06-06 22:29:56,766][262198] Updated weights for policy 0, policy_version 46043 (0.0008)
+[2026-06-06 22:29:57,920][261293] Fps is (10 sec: 4915.3, 60 sec: 3788.8, 300 sec: 3825.2). Total num frames: 23578624. Throughput: 0: 3910.8. Samples: 23583620. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:29:57,922][261293] Avg episode reward: [(0, '1655.452')]
+[2026-06-06 22:29:58,365][262198] Updated weights for policy 0, policy_version 46053 (0.0008)
+[2026-06-06 22:29:59,404][262198] Updated weights for policy 0, policy_version 46064 (0.0008)
+[2026-06-06 22:30:00,568][262198] Updated weights for policy 0, policy_version 46075 (0.0008)
+[2026-06-06 22:30:02,920][261293] Fps is (10 sec: 4096.0, 60 sec: 3788.8, 300 sec: 3818.3). Total num frames: 23595008. Throughput: 0: 4018.7. Samples: 23597896. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:30:02,922][261293] Avg episode reward: [(0, '1668.723')]
+[2026-06-06 22:30:03,335][262198] Updated weights for policy 0, policy_version 46087 (0.0010)
+[2026-06-06 22:30:06,704][262198] Updated weights for policy 0, policy_version 46099 (0.0018)
+[2026-06-06 22:30:07,921][261293] Fps is (10 sec: 2662.3, 60 sec: 3686.4, 300 sec: 3783.6). Total num frames: 23605248. Throughput: 0: 3842.0. Samples: 23611540. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:30:07,923][261293] Avg episode reward: [(0, '1710.542')]
+[2026-06-06 22:30:08,945][262198] Updated weights for policy 0, policy_version 46109 (0.0017)
+[2026-06-06 22:30:12,920][261293] Fps is (10 sec: 1843.2, 60 sec: 3549.9, 300 sec: 3748.9). Total num frames: 23613440. Throughput: 0: 3503.7. Samples: 23623238. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:30:12,921][261293] Avg episode reward: [(0, '1871.987')]
+[2026-06-06 22:30:12,976][262198] Updated weights for policy 0, policy_version 46121 (0.0015)
+[2026-06-06 22:30:15,326][262198] Updated weights for policy 0, policy_version 46136 (0.0012)
+[2026-06-06 22:30:17,921][261293] Fps is (10 sec: 2048.0, 60 sec: 3515.7, 300 sec: 3728.1). Total num frames: 23625728. Throughput: 0: 3339.6. Samples: 23629126. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:30:17,924][261293] Avg episode reward: [(0, '1636.341')]
+[2026-06-06 22:30:19,792][262198] Updated weights for policy 0, policy_version 46149 (0.0014)
+[2026-06-06 22:30:22,786][262198] Updated weights for policy 0, policy_version 46162 (0.0014)
+[2026-06-06 22:30:22,920][261293] Fps is (10 sec: 2252.8, 60 sec: 3345.1, 300 sec: 3707.2). Total num frames: 23635968. Throughput: 0: 3159.8. Samples: 23641720. Policy #0 lag: (min: 3.0, avg: 22.2, max: 35.0)
+[2026-06-06 22:30:22,922][261293] Avg episode reward: [(0, '1812.990')]
+[2026-06-06 22:30:24,777][262198] Updated weights for policy 0, policy_version 46176 (0.0011)
+[2026-06-06 22:30:27,920][261293] Fps is (10 sec: 2048.1, 60 sec: 3106.1, 300 sec: 3693.4). Total num frames: 23646208. Throughput: 0: 2994.3. Samples: 23654740. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:27,921][261293] Avg episode reward: [(0, '1451.845')]
+[2026-06-06 22:30:29,299][262198] Updated weights for policy 0, policy_version 46191 (0.0014)
+[2026-06-06 22:30:31,837][262198] Updated weights for policy 0, policy_version 46204 (0.0014)
+[2026-06-06 22:30:32,922][261293] Fps is (10 sec: 2252.5, 60 sec: 3003.7, 300 sec: 3686.4). Total num frames: 23658496. Throughput: 0: 2896.2. Samples: 23660874. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:32,925][261293] Avg episode reward: [(0, '1786.236')]
+[2026-06-06 22:30:36,236][262198] Updated weights for policy 0, policy_version 46218 (0.0011)
+[2026-06-06 22:30:37,920][261293] Fps is (10 sec: 2048.0, 60 sec: 2867.2, 300 sec: 3665.6). Total num frames: 23666688. Throughput: 0: 2636.2. Samples: 23673926. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:37,922][261293] Avg episode reward: [(0, '1814.921')]
+[2026-06-06 22:30:38,475][262198] Updated weights for policy 0, policy_version 46232 (0.0012)
+[2026-06-06 22:30:42,649][262198] Updated weights for policy 0, policy_version 46246 (0.0012)
+[2026-06-06 22:30:42,920][261293] Fps is (10 sec: 2048.3, 60 sec: 2764.8, 300 sec: 3637.8). Total num frames: 23678976. Throughput: 0: 2299.5. Samples: 23687098. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:42,922][261293] Avg episode reward: [(0, '1515.859')]
+[2026-06-06 22:30:45,667][262198] Updated weights for policy 0, policy_version 46261 (0.0012)
+[2026-06-06 22:30:47,921][261293] Fps is (10 sec: 2457.5, 60 sec: 2696.5, 300 sec: 3617.0). Total num frames: 23691264. Throughput: 0: 2131.5. Samples: 23693814. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:47,924][261293] Avg episode reward: [(0, '1638.660')]
+[2026-06-06 22:30:48,655][262198] Updated weights for policy 0, policy_version 46274 (0.0009)
+[2026-06-06 22:30:51,948][262198] Updated weights for policy 0, policy_version 46287 (0.0012)
+[2026-06-06 22:30:52,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2423.5, 300 sec: 3575.3). Total num frames: 23699456. Throughput: 0: 2116.6. Samples: 23706788. Policy #0 lag: (min: 3.0, avg: 18.0, max: 35.0)
+[2026-06-06 22:30:52,924][261293] Avg episode reward: [(0, '1724.928')]
+[2026-06-06 22:30:54,255][262198] Updated weights for policy 0, policy_version 46300 (0.0010)
+[2026-06-06 22:30:57,365][262198] Updated weights for policy 0, policy_version 46312 (0.0013)
+[2026-06-06 22:30:57,920][261293] Fps is (10 sec: 2048.1, 60 sec: 2218.7, 300 sec: 3554.5). Total num frames: 23711744. Throughput: 0: 2160.7. Samples: 23720468. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:30:57,922][261293] Avg episode reward: [(0, '1760.403')]
+[2026-06-06 22:31:00,039][262198] Updated weights for policy 0, policy_version 46323 (0.0014)
+[2026-06-06 22:31:02,084][262198] Updated weights for policy 0, policy_version 46336 (0.0010)
+[2026-06-06 22:31:02,920][261293] Fps is (10 sec: 2457.7, 60 sec: 2150.4, 300 sec: 3512.8). Total num frames: 23724032. Throughput: 0: 2184.4. Samples: 23727422. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:31:02,922][261293] Avg episode reward: [(0, '1782.842')]
+[2026-06-06 22:31:06,364][262198] Updated weights for policy 0, policy_version 46350 (0.0009)
+[2026-06-06 22:31:07,920][261293] Fps is (10 sec: 2457.6, 60 sec: 2184.5, 300 sec: 3485.1). Total num frames: 23736320. Throughput: 0: 2208.5. Samples: 23741102. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:31:07,922][261293] Avg episode reward: [(0, '1779.059')]
+[2026-06-06 22:31:09,273][262198] Updated weights for policy 0, policy_version 46365 (0.0012)
+[2026-06-06 22:31:10,593][262198] Updated weights for policy 0, policy_version 46375 (0.0009)
+[2026-06-06 22:31:11,826][262198] Updated weights for policy 0, policy_version 46385 (0.0008)
+[2026-06-06 22:31:12,920][261293] Fps is (10 sec: 3072.0, 60 sec: 2355.2, 300 sec: 3478.1). Total num frames: 23754752. Throughput: 0: 2373.8. Samples: 23761562. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:31:12,921][261293] Avg episode reward: [(0, '1784.825')]
+[2026-06-06 22:31:13,370][262198] Updated weights for policy 0, policy_version 46397 (0.0012)
+[2026-06-06 22:31:16,901][262198] Updated weights for policy 0, policy_version 46412 (0.0012)
+[2026-06-06 22:31:17,921][261293] Fps is (10 sec: 2662.1, 60 sec: 2286.9, 300 sec: 3436.5). Total num frames: 23762944. Throughput: 0: 2385.1. Samples: 23768204. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:31:17,924][261293] Avg episode reward: [(0, '1805.903')]
+[2026-06-06 22:31:19,969][262198] Updated weights for policy 0, policy_version 46425 (0.0013)
+[2026-06-06 22:31:22,920][261293] Fps is (10 sec: 2048.0, 60 sec: 2321.1, 300 sec: 3401.8). Total num frames: 23775232. Throughput: 0: 2384.5. Samples: 23781230. Policy #0 lag: (min: 2.0, avg: 20.3, max: 34.0)
+[2026-06-06 22:31:22,921][261293] Avg episode reward: [(0, '1797.934')]
+[2026-06-06 22:31:23,196][262198] Updated weights for policy 0, policy_version 46437 (0.0012)
+[2026-06-06 22:31:25,993][262198] Updated weights for policy 0, policy_version 46449 (0.0013)
+[2026-06-06 22:31:27,920][261293] Fps is (10 sec: 2457.8, 60 sec: 2355.2, 300 sec: 3374.0). Total num frames: 23787520. Throughput: 0: 2401.6. Samples: 23795168. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:27,922][261293] Avg episode reward: [(0, '1642.301')]
+[2026-06-06 22:31:28,116][262198] Updated weights for policy 0, policy_version 46462 (0.0012)
+[2026-06-06 22:31:31,871][262198] Updated weights for policy 0, policy_version 46475 (0.0013)
+[2026-06-06 22:31:32,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2321.1, 300 sec: 3339.3). Total num frames: 23797760. Throughput: 0: 2388.8. Samples: 23801310. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:32,924][261293] Avg episode reward: [(0, '1643.859')]
+[2026-06-06 22:31:34,478][262198] Updated weights for policy 0, policy_version 46488 (0.0016)
+[2026-06-06 22:31:37,920][261293] Fps is (10 sec: 2048.0, 60 sec: 2355.2, 300 sec: 3297.6). Total num frames: 23808000. Throughput: 0: 2367.6. Samples: 23813330. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:37,922][261293] Avg episode reward: [(0, '1696.336')]
+[2026-06-06 22:31:39,927][262198] Updated weights for policy 0, policy_version 46505 (0.0014)
+[2026-06-06 22:31:42,577][262198] Updated weights for policy 0, policy_version 46519 (0.0013)
+[2026-06-06 22:31:42,921][261293] Fps is (10 sec: 2048.0, 60 sec: 2321.1, 300 sec: 3262.9). Total num frames: 23818240. Throughput: 0: 2335.0. Samples: 23825542. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:42,923][261293] Avg episode reward: [(0, '1737.551')]
+[2026-06-06 22:31:46,990][262198] Updated weights for policy 0, policy_version 46533 (0.0012)
+[2026-06-06 22:31:47,921][261293] Fps is (10 sec: 1843.1, 60 sec: 2252.8, 300 sec: 3228.2). Total num frames: 23826432. Throughput: 0: 2309.8. Samples: 23831364. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:47,924][261293] Avg episode reward: [(0, '1830.899')]
+[2026-06-06 22:31:50,296][262198] Updated weights for policy 0, policy_version 46548 (0.0012)
+[2026-06-06 22:31:52,920][261293] Fps is (10 sec: 2048.0, 60 sec: 2321.1, 300 sec: 3207.4). Total num frames: 23838720. Throughput: 0: 2275.7. Samples: 23843510. Policy #0 lag: (min: 3.0, avg: 19.1, max: 38.0)
+[2026-06-06 22:31:52,922][261293] Avg episode reward: [(0, '1878.876')]
+[2026-06-06 22:31:53,993][262198] Updated weights for policy 0, policy_version 46562 (0.0013)
+[2026-06-06 22:31:57,103][262198] Updated weights for policy 0, policy_version 46576 (0.0011)
+[2026-06-06 22:31:57,929][261293] Fps is (10 sec: 2046.5, 60 sec: 2252.5, 300 sec: 3165.6). Total num frames: 23846912. Throughput: 0: 2100.4. Samples: 23856098. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:31:57,933][261293] Avg episode reward: [(0, '1927.813')]
+[2026-06-06 22:32:00,322][262198] Updated weights for policy 0, policy_version 46590 (0.0009)
+[2026-06-06 22:32:02,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2252.8, 300 sec: 3151.8). Total num frames: 23859200. Throughput: 0: 2108.2. Samples: 23863072. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:32:02,925][261293] Avg episode reward: [(0, '1781.935')]
+[2026-06-06 22:32:03,374][262198] Updated weights for policy 0, policy_version 46603 (0.0011)
+[2026-06-06 22:32:05,525][262198] Updated weights for policy 0, policy_version 46614 (0.0013)
+[2026-06-06 22:32:07,921][261293] Fps is (10 sec: 2459.5, 60 sec: 2252.8, 300 sec: 3144.9). Total num frames: 23871488. Throughput: 0: 2124.3. Samples: 23876826. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:32:07,923][261293] Avg episode reward: [(0, '1741.684')]
+[2026-06-06 22:32:08,542][262198] Updated weights for policy 0, policy_version 46628 (0.0008)
+[2026-06-06 22:32:12,377][262198] Updated weights for policy 0, policy_version 46642 (0.0011)
+[2026-06-06 22:32:12,920][261293] Fps is (10 sec: 2457.8, 60 sec: 2150.4, 300 sec: 3138.0). Total num frames: 23883776. Throughput: 0: 2119.4. Samples: 23890540. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:32:12,921][261293] Avg episode reward: [(0, '1841.797')]
+[2026-06-06 22:32:14,871][262198] Updated weights for policy 0, policy_version 46655 (0.0012)
+[2026-06-06 22:32:17,921][261293] Fps is (10 sec: 2048.0, 60 sec: 2150.4, 300 sec: 3117.1). Total num frames: 23891968. Throughput: 0: 2127.9. Samples: 23897064. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:32:17,926][261293] Avg episode reward: [(0, '1605.324')]
+[2026-06-06 22:32:17,980][262198] Updated weights for policy 0, policy_version 46668 (0.0011)
+[2026-06-06 22:32:20,261][262198] Updated weights for policy 0, policy_version 46680 (0.0015)
+[2026-06-06 22:32:22,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2150.4, 300 sec: 3103.3). Total num frames: 23904256. Throughput: 0: 2157.7. Samples: 23910426. Policy #0 lag: (min: 0.0, avg: 16.0, max: 32.0)
+[2026-06-06 22:32:22,922][261293] Avg episode reward: [(0, '1763.882')]
+[2026-06-06 22:32:24,075][262198] Updated weights for policy 0, policy_version 46695 (0.0013)
+[2026-06-06 22:32:27,462][262198] Updated weights for policy 0, policy_version 46709 (0.0008)
+[2026-06-06 22:32:27,924][261293] Fps is (10 sec: 2456.8, 60 sec: 2150.3, 300 sec: 3089.3). Total num frames: 23916544. Throughput: 0: 2186.0. Samples: 23923920. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:27,926][261293] Avg episode reward: [(0, '1814.290')]
+[2026-06-06 22:32:30,211][262198] Updated weights for policy 0, policy_version 46722 (0.0011)
+[2026-06-06 22:32:32,920][261293] Fps is (10 sec: 2252.9, 60 sec: 2150.4, 300 sec: 3054.6). Total num frames: 23926784. Throughput: 0: 2217.4. Samples: 23931146. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:32,922][261293] Avg episode reward: [(0, '1843.702')]
+[2026-06-06 22:32:32,988][262198] Updated weights for policy 0, policy_version 46736 (0.0012)
+[2026-06-06 22:32:36,238][262198] Updated weights for policy 0, policy_version 46749 (0.0009)
+[2026-06-06 22:32:37,921][261293] Fps is (10 sec: 2253.4, 60 sec: 2184.5, 300 sec: 3033.8). Total num frames: 23939072. Throughput: 0: 2262.9. Samples: 23945340. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:37,924][261293] Avg episode reward: [(0, '1804.534')]
+[2026-06-06 22:32:39,095][262198] Updated weights for policy 0, policy_version 46762 (0.0012)
+[2026-06-06 22:32:41,387][262198] Updated weights for policy 0, policy_version 46775 (0.0012)
+[2026-06-06 22:32:42,921][261293] Fps is (10 sec: 2457.5, 60 sec: 2218.7, 300 sec: 2999.1). Total num frames: 23951360. Throughput: 0: 2292.0. Samples: 23959220. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:42,924][261293] Avg episode reward: [(0, '1772.054')]
+[2026-06-06 22:32:45,496][262198] Updated weights for policy 0, policy_version 46790 (0.0012)
+[2026-06-06 22:32:47,921][261293] Fps is (10 sec: 2252.9, 60 sec: 2252.8, 300 sec: 2971.3). Total num frames: 23961600. Throughput: 0: 2281.7. Samples: 23965746. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:47,922][261293] Avg episode reward: [(0, '1725.436')]
+[2026-06-06 22:32:48,267][262198] Updated weights for policy 0, policy_version 46803 (0.0011)
+[2026-06-06 22:32:50,597][262198] Updated weights for policy 0, policy_version 46816 (0.0010)
+[2026-06-06 22:32:52,920][261293] Fps is (10 sec: 2252.9, 60 sec: 2252.8, 300 sec: 2943.6). Total num frames: 23973888. Throughput: 0: 2289.9. Samples: 23979872. Policy #0 lag: (min: 1.0, avg: 16.6, max: 37.0)
+[2026-06-06 22:32:52,922][261293] Avg episode reward: [(0, '1802.030')]
+[2026-06-06 22:32:54,195][262198] Updated weights for policy 0, policy_version 46829 (0.0014)
+[2026-06-06 22:32:56,207][262198] Updated weights for policy 0, policy_version 46841 (0.0012)
+[2026-06-06 22:32:57,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2321.4, 300 sec: 2922.7). Total num frames: 23986176. Throughput: 0: 2287.9. Samples: 23993496. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:32:57,922][261293] Avg episode reward: [(0, '1721.438')]
+[2026-06-06 22:32:59,847][262198] Updated weights for policy 0, policy_version 46855 (0.0014)
+[2026-06-06 22:33:02,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2252.8, 300 sec: 2867.2). Total num frames: 23994368. Throughput: 0: 2283.4. Samples: 23999818. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:33:02,923][261293] Avg episode reward: [(0, '1716.326')]
+[2026-06-06 22:33:02,953][262198] Updated weights for policy 0, policy_version 46868 (0.0016)
+[2026-06-06 22:33:06,488][262198] Updated weights for policy 0, policy_version 46881 (0.0011)
+[2026-06-06 22:33:07,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2252.8, 300 sec: 2832.5). Total num frames: 24006656. Throughput: 0: 2262.1. Samples: 24012220. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:33:07,924][261293] Avg episode reward: [(0, '1750.397')]
+[2026-06-06 22:33:09,873][262198] Updated weights for policy 0, policy_version 46896 (0.0012)
+[2026-06-06 22:33:12,921][261293] Fps is (10 sec: 2252.9, 60 sec: 2218.7, 300 sec: 2783.9). Total num frames: 24016896. Throughput: 0: 2245.0. Samples: 24024940. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:33:12,923][261293] Avg episode reward: [(0, '1665.730')]
+[2026-06-06 22:33:13,328][262198] Updated weights for policy 0, policy_version 46909 (0.0010)
+[2026-06-06 22:33:15,463][262198] Updated weights for policy 0, policy_version 46920 (0.0013)
+[2026-06-06 22:33:17,190][262198] Updated weights for policy 0, policy_version 46931 (0.0010)
+[2026-06-06 22:33:17,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2321.0, 300 sec: 2749.2). Total num frames: 24031232. Throughput: 0: 2235.0. Samples: 24031722. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:33:17,927][261293] Avg episode reward: [(0, '1768.568')]
+[2026-06-06 22:33:19,233][262198] Updated weights for policy 0, policy_version 46941 (0.0010)
+[2026-06-06 22:33:21,347][262198] Updated weights for policy 0, policy_version 46952 (0.0014)
+[2026-06-06 22:33:22,921][261293] Fps is (10 sec: 2457.5, 60 sec: 2286.9, 300 sec: 2686.7). Total num frames: 24041472. Throughput: 0: 2281.3. Samples: 24047998. Policy #0 lag: (min: 4.0, avg: 16.9, max: 35.0)
+[2026-06-06 22:33:22,924][261293] Avg episode reward: [(0, '1726.826')]
+[2026-06-06 22:33:24,289][262198] Updated weights for policy 0, policy_version 46964 (0.0012)
+[2026-06-06 22:33:27,724][262198] Updated weights for policy 0, policy_version 46977 (0.0017)
+[2026-06-06 22:33:27,920][261293] Fps is (10 sec: 2253.0, 60 sec: 2287.1, 300 sec: 2658.9). Total num frames: 24053760. Throughput: 0: 2259.6. Samples: 24060902. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:27,923][261293] Avg episode reward: [(0, '1926.286')]
+[2026-06-06 22:33:31,054][262198] Updated weights for policy 0, policy_version 46991 (0.0022)
+[2026-06-06 22:33:32,920][261293] Fps is (10 sec: 2252.9, 60 sec: 2286.9, 300 sec: 2631.2). Total num frames: 24064000. Throughput: 0: 2246.8. Samples: 24066852. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:32,922][261293] Avg episode reward: [(0, '1603.021')]
+[2026-06-06 22:33:33,061][262198] Updated weights for policy 0, policy_version 47004 (0.0015)
+[2026-06-06 22:33:37,377][262198] Updated weights for policy 0, policy_version 47020 (0.0012)
+[2026-06-06 22:33:37,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2252.8, 300 sec: 2617.3). Total num frames: 24074240. Throughput: 0: 2237.1. Samples: 24080542. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:37,923][261293] Avg episode reward: [(0, '1611.423')]
+[2026-06-06 22:33:40,121][262198] Updated weights for policy 0, policy_version 47033 (0.0015)
+[2026-06-06 22:33:42,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2252.8, 300 sec: 2596.5). Total num frames: 24086528. Throughput: 0: 2239.7. Samples: 24094280. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:42,922][261293] Avg episode reward: [(0, '1668.858')]
+[2026-06-06 22:33:43,467][262198] Updated weights for policy 0, policy_version 47045 (0.0012)
+[2026-06-06 22:33:46,315][262198] Updated weights for policy 0, policy_version 47059 (0.0012)
+[2026-06-06 22:33:47,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2286.9, 300 sec: 2582.6). Total num frames: 24098816. Throughput: 0: 2253.5. Samples: 24101226. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:47,924][261293] Avg episode reward: [(0, '1706.460')]
+[2026-06-06 22:33:48,915][262198] Updated weights for policy 0, policy_version 47072 (0.0011)
+[2026-06-06 22:33:51,482][262198] Updated weights for policy 0, policy_version 47084 (0.0013)
+[2026-06-06 22:33:52,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2252.8, 300 sec: 2568.7). Total num frames: 24109056. Throughput: 0: 2288.7. Samples: 24115210. Policy #0 lag: (min: 3.0, avg: 20.9, max: 37.0)
+[2026-06-06 22:33:52,921][261293] Avg episode reward: [(0, '1784.096')]
+[2026-06-06 22:33:54,101][262198] Updated weights for policy 0, policy_version 47097 (0.0010)
+[2026-06-06 22:33:56,658][262198] Updated weights for policy 0, policy_version 47108 (0.0011)
+[2026-06-06 22:33:57,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2252.8, 300 sec: 2554.8). Total num frames: 24121344. Throughput: 0: 2328.4. Samples: 24129720. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:33:57,923][261293] Avg episode reward: [(0, '1702.308')]
+[2026-06-06 22:34:00,001][262198] Updated weights for policy 0, policy_version 47121 (0.0010)
+[2026-06-06 22:34:02,245][262198] Updated weights for policy 0, policy_version 47134 (0.0010)
+[2026-06-06 22:34:02,920][261293] Fps is (10 sec: 2457.7, 60 sec: 2321.1, 300 sec: 2540.9). Total num frames: 24133632. Throughput: 0: 2342.4. Samples: 24137128. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:34:02,922][261293] Avg episode reward: [(0, '1669.383')]
+[2026-06-06 22:34:04,958][262198] Updated weights for policy 0, policy_version 47146 (0.0011)
+[2026-06-06 22:34:06,994][262198] Updated weights for policy 0, policy_version 47157 (0.0010)
+[2026-06-06 22:34:07,920][261293] Fps is (10 sec: 2662.5, 60 sec: 2355.2, 300 sec: 2534.0). Total num frames: 24147968. Throughput: 0: 2305.5. Samples: 24151746. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:34:07,922][261293] Avg episode reward: [(0, '1689.386')]
+[2026-06-06 22:34:10,002][262198] Updated weights for policy 0, policy_version 47169 (0.0011)
+[2026-06-06 22:34:12,652][262198] Updated weights for policy 0, policy_version 47184 (0.0010)
+[2026-06-06 22:34:12,921][261293] Fps is (10 sec: 2457.4, 60 sec: 2355.2, 300 sec: 2520.1). Total num frames: 24158208. Throughput: 0: 2334.3. Samples: 24165948. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:34:12,923][261293] Avg episode reward: [(0, '1732.107')]
+[2026-06-06 22:34:16,120][262198] Updated weights for policy 0, policy_version 47197 (0.0010)
+[2026-06-06 22:34:17,921][261293] Fps is (10 sec: 2047.9, 60 sec: 2286.9, 300 sec: 2485.4). Total num frames: 24168448. Throughput: 0: 2352.7. Samples: 24172726. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:34:17,925][261293] Avg episode reward: [(0, '1777.535')]
+[2026-06-06 22:34:18,960][262198] Updated weights for policy 0, policy_version 47209 (0.0012)
+[2026-06-06 22:34:20,884][262198] Updated weights for policy 0, policy_version 47220 (0.0010)
+[2026-06-06 22:34:22,921][261293] Fps is (10 sec: 2252.9, 60 sec: 2321.1, 300 sec: 2443.7). Total num frames: 24180736. Throughput: 0: 2351.0. Samples: 24186334. Policy #0 lag: (min: 2.0, avg: 21.9, max: 38.0)
+[2026-06-06 22:34:22,924][261293] Avg episode reward: [(0, '1807.572')]
+[2026-06-06 22:34:24,223][262198] Updated weights for policy 0, policy_version 47233 (0.0009)
+[2026-06-06 22:34:26,829][262198] Updated weights for policy 0, policy_version 47245 (0.0011)
+[2026-06-06 22:34:27,920][261293] Fps is (10 sec: 2457.7, 60 sec: 2321.1, 300 sec: 2422.9). Total num frames: 24193024. Throughput: 0: 2359.9. Samples: 24200478. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:27,923][261293] Avg episode reward: [(0, '1647.995')]
+[2026-06-06 22:34:28,870][262198] Updated weights for policy 0, policy_version 47258 (0.0009)
+[2026-06-06 22:34:30,189][262198] Updated weights for policy 0, policy_version 47268 (0.0007)
+[2026-06-06 22:34:31,612][262198] Updated weights for policy 0, policy_version 47278 (0.0008)
+[2026-06-06 22:34:32,420][262198] Updated weights for policy 0, policy_version 47288 (0.0011)
+[2026-06-06 22:34:32,921][261293] Fps is (10 sec: 3276.8, 60 sec: 2491.7, 300 sec: 2436.8). Total num frames: 24213504. Throughput: 0: 2429.5. Samples: 24210554. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:32,923][261293] Avg episode reward: [(0, '1951.365')]
+[2026-06-06 22:34:32,937][262115] Saving new best policy, reward=1951.365!
+[2026-06-06 22:34:33,752][262198] Updated weights for policy 0, policy_version 47298 (0.0009)
+[2026-06-06 22:34:35,024][262198] Updated weights for policy 0, policy_version 47308 (0.0009)
+[2026-06-06 22:34:36,505][262198] Updated weights for policy 0, policy_version 47322 (0.0009)
+[2026-06-06 22:34:37,693][262198] Updated weights for policy 0, policy_version 47333 (0.0008)
+[2026-06-06 22:34:37,921][261293] Fps is (10 sec: 4300.5, 60 sec: 2696.5, 300 sec: 2450.7). Total num frames: 24236032. Throughput: 0: 2723.5. Samples: 24237768. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:37,925][261293] Avg episode reward: [(0, '1723.178')]
+[2026-06-06 22:34:38,571][262198] Updated weights for policy 0, policy_version 47343 (0.0008)
+[2026-06-06 22:34:40,450][262198] Updated weights for policy 0, policy_version 47358 (0.0009)
+[2026-06-06 22:34:41,271][262198] Updated weights for policy 0, policy_version 47368 (0.0008)
+[2026-06-06 22:34:42,920][261293] Fps is (10 sec: 4300.7, 60 sec: 2833.1, 300 sec: 2464.5). Total num frames: 24256512. Throughput: 0: 3014.4. Samples: 24265366. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:42,922][261293] Avg episode reward: [(0, '1781.747')]
+[2026-06-06 22:34:42,992][262198] Updated weights for policy 0, policy_version 47380 (0.0009)
+[2026-06-06 22:34:44,126][262198] Updated weights for policy 0, policy_version 47392 (0.0008)
+[2026-06-06 22:34:45,411][262198] Updated weights for policy 0, policy_version 47402 (0.0007)
+[2026-06-06 22:34:47,117][262198] Updated weights for policy 0, policy_version 47416 (0.0010)
+[2026-06-06 22:34:47,920][261293] Fps is (10 sec: 4506.0, 60 sec: 3037.9, 300 sec: 2464.5). Total num frames: 24281088. Throughput: 0: 3148.3. Samples: 24278804. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:47,922][261293] Avg episode reward: [(0, '1913.056')]
+[2026-06-06 22:34:48,369][262198] Updated weights for policy 0, policy_version 47427 (0.0008)
+[2026-06-06 22:34:49,371][262198] Updated weights for policy 0, policy_version 47437 (0.0010)
+[2026-06-06 22:34:50,499][262198] Updated weights for policy 0, policy_version 47452 (0.0010)
+[2026-06-06 22:34:51,980][262198] Updated weights for policy 0, policy_version 47462 (0.0008)
+[2026-06-06 22:34:52,920][261293] Fps is (10 sec: 4710.5, 60 sec: 3242.7, 300 sec: 2457.6). Total num frames: 24303616. Throughput: 0: 3448.0. Samples: 24306908. Policy #0 lag: (min: 6.0, avg: 20.0, max: 38.0)
+[2026-06-06 22:34:52,922][261293] Avg episode reward: [(0, '1832.506')]
+[2026-06-06 22:34:52,961][262198] Updated weights for policy 0, policy_version 47472 (0.0008)
+[2026-06-06 22:34:53,921][262198] Updated weights for policy 0, policy_version 47482 (0.0009)
+[2026-06-06 22:34:55,345][262198] Updated weights for policy 0, policy_version 47492 (0.0009)
+[2026-06-06 22:34:56,523][262198] Updated weights for policy 0, policy_version 47502 (0.0009)
+[2026-06-06 22:34:57,647][262198] Updated weights for policy 0, policy_version 47513 (0.0010)
+[2026-06-06 22:34:57,920][261293] Fps is (10 sec: 4710.3, 60 sec: 3447.5, 300 sec: 2485.4). Total num frames: 24328192. Throughput: 0: 3725.5. Samples: 24333596. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:34:57,922][261293] Avg episode reward: [(0, '1836.418')]
+[2026-06-06 22:34:58,867][262198] Updated weights for policy 0, policy_version 47523 (0.0008)
+[2026-06-06 22:35:00,009][262198] Updated weights for policy 0, policy_version 47533 (0.0008)
+[2026-06-06 22:35:01,294][262198] Updated weights for policy 0, policy_version 47547 (0.0008)
+[2026-06-06 22:35:02,601][262198] Updated weights for policy 0, policy_version 47557 (0.0008)
+[2026-06-06 22:35:02,927][261293] Fps is (10 sec: 4707.3, 60 sec: 3617.7, 300 sec: 2527.0). Total num frames: 24350720. Throughput: 0: 3893.4. Samples: 24347950. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:35:02,929][261293] Avg episode reward: [(0, '1830.812')]
+[2026-06-06 22:35:03,442][262198] Updated weights for policy 0, policy_version 47567 (0.0008)
+[2026-06-06 22:35:04,756][262198] Updated weights for policy 0, policy_version 47579 (0.0008)
+[2026-06-06 22:35:06,298][262198] Updated weights for policy 0, policy_version 47589 (0.0009)
+[2026-06-06 22:35:06,955][262198] Updated weights for policy 0, policy_version 47599 (0.0007)
+[2026-06-06 22:35:07,920][261293] Fps is (10 sec: 4710.6, 60 sec: 3788.8, 300 sec: 2582.6). Total num frames: 24375296. Throughput: 0: 4213.9. Samples: 24375960. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:35:07,924][261293] Avg episode reward: [(0, '1680.204')]
+[2026-06-06 22:35:08,190][262198] Updated weights for policy 0, policy_version 47610 (0.0009)
+[2026-06-06 22:35:09,121][262198] Updated weights for policy 0, policy_version 47620 (0.0008)
+[2026-06-06 22:35:10,162][262198] Updated weights for policy 0, policy_version 47630 (0.0007)
+[2026-06-06 22:35:10,921][262198] Updated weights for policy 0, policy_version 47640 (0.0006)
+[2026-06-06 22:35:11,980][262198] Updated weights for policy 0, policy_version 47650 (0.0007)
+[2026-06-06 22:35:12,921][261293] Fps is (10 sec: 4918.3, 60 sec: 4027.7, 300 sec: 2624.2). Total num frames: 24399872. Throughput: 0: 4608.6. Samples: 24407868. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:35:12,923][261293] Avg episode reward: [(0, '1905.968')]
+[2026-06-06 22:35:13,353][262198] Updated weights for policy 0, policy_version 47663 (0.0008)
+[2026-06-06 22:35:14,337][262198] Updated weights for policy 0, policy_version 47674 (0.0008)
+[2026-06-06 22:35:15,347][262198] Updated weights for policy 0, policy_version 47684 (0.0008)
+[2026-06-06 22:35:16,728][262198] Updated weights for policy 0, policy_version 47694 (0.0008)
+[2026-06-06 22:35:17,921][261293] Fps is (10 sec: 4914.6, 60 sec: 4266.6, 300 sec: 2672.8). Total num frames: 24424448. Throughput: 0: 4717.7. Samples: 24422854. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:35:17,924][261293] Avg episode reward: [(0, '1760.235')]
+[2026-06-06 22:35:18,182][262198] Updated weights for policy 0, policy_version 47708 (0.0009)
+[2026-06-06 22:35:19,324][262198] Updated weights for policy 0, policy_version 47718 (0.0007)
+[2026-06-06 22:35:20,068][262198] Updated weights for policy 0, policy_version 47728 (0.0009)
+[2026-06-06 22:35:21,051][262198] Updated weights for policy 0, policy_version 47738 (0.0009)
+[2026-06-06 22:35:22,011][262198] Updated weights for policy 0, policy_version 47748 (0.0008)
+[2026-06-06 22:35:22,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4471.4, 300 sec: 2721.4). Total num frames: 24449024. Throughput: 0: 4755.8. Samples: 24451776. Policy #0 lag: (min: 3.0, avg: 18.5, max: 35.0)
+[2026-06-06 22:35:22,922][261293] Avg episode reward: [(0, '1731.036')]
+[2026-06-06 22:35:23,452][262198] Updated weights for policy 0, policy_version 47758 (0.0010)
+[2026-06-06 22:35:23,986][262198] Updated weights for policy 0, policy_version 47768 (0.0008)
+[2026-06-06 22:35:25,455][262198] Updated weights for policy 0, policy_version 47778 (0.0009)
+[2026-06-06 22:35:26,531][262198] Updated weights for policy 0, policy_version 47788 (0.0008)
+[2026-06-06 22:35:27,921][261293] Fps is (10 sec: 4915.6, 60 sec: 4676.3, 300 sec: 2763.1). Total num frames: 24473600. Throughput: 0: 4784.5. Samples: 24480668. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:27,923][261293] Avg episode reward: [(0, '1637.442')]
+[2026-06-06 22:35:27,933][262198] Updated weights for policy 0, policy_version 47803 (0.0010)
+[2026-06-06 22:35:29,650][262198] Updated weights for policy 0, policy_version 47813 (0.0010)
+[2026-06-06 22:35:30,698][262198] Updated weights for policy 0, policy_version 47823 (0.0009)
+[2026-06-06 22:35:31,756][262198] Updated weights for policy 0, policy_version 47834 (0.0010)
+[2026-06-06 22:35:32,922][261293] Fps is (10 sec: 4505.5, 60 sec: 4676.2, 300 sec: 2804.7). Total num frames: 24494080. Throughput: 0: 4770.1. Samples: 24493460. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:32,925][261293] Avg episode reward: [(0, '1793.832')]
+[2026-06-06 22:35:33,029][262198] Updated weights for policy 0, policy_version 47844 (0.0010)
+[2026-06-06 22:35:34,631][262198] Updated weights for policy 0, policy_version 47854 (0.0009)
+[2026-06-06 22:35:35,527][262198] Updated weights for policy 0, policy_version 47865 (0.0009)
+[2026-06-06 22:35:36,829][262198] Updated weights for policy 0, policy_version 47875 (0.0010)
+[2026-06-06 22:35:37,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4642.2, 300 sec: 2832.5). Total num frames: 24514560. Throughput: 0: 4705.0. Samples: 24518636. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:37,923][261293] Avg episode reward: [(0, '1727.526')]
+[2026-06-06 22:35:38,579][262198] Updated weights for policy 0, policy_version 47885 (0.0008)
+[2026-06-06 22:35:39,627][262198] Updated weights for policy 0, policy_version 47898 (0.0009)
+[2026-06-06 22:35:41,053][262198] Updated weights for policy 0, policy_version 47908 (0.0008)
+[2026-06-06 22:35:42,461][262198] Updated weights for policy 0, policy_version 47918 (0.0008)
+[2026-06-06 22:35:42,920][261293] Fps is (10 sec: 4301.1, 60 sec: 4676.3, 300 sec: 2867.2). Total num frames: 24537088. Throughput: 0: 4668.5. Samples: 24543678. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:42,923][261293] Avg episode reward: [(0, '1782.544')]
+[2026-06-06 22:35:43,658][262198] Updated weights for policy 0, policy_version 47931 (0.0009)
+[2026-06-06 22:35:45,382][262198] Updated weights for policy 0, policy_version 47941 (0.0009)
+[2026-06-06 22:35:46,473][262198] Updated weights for policy 0, policy_version 47951 (0.0009)
+[2026-06-06 22:35:47,551][262198] Updated weights for policy 0, policy_version 47961 (0.0009)
+[2026-06-06 22:35:47,921][261293] Fps is (10 sec: 4300.8, 60 sec: 4608.0, 300 sec: 2908.9). Total num frames: 24557568. Throughput: 0: 4618.1. Samples: 24555738. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:47,924][261293] Avg episode reward: [(0, '1794.453')]
+[2026-06-06 22:35:48,507][262198] Updated weights for policy 0, policy_version 47971 (0.0009)
+[2026-06-06 22:35:50,361][262198] Updated weights for policy 0, policy_version 47982 (0.0008)
+[2026-06-06 22:35:51,513][262198] Updated weights for policy 0, policy_version 47993 (0.0010)
+[2026-06-06 22:35:52,406][262198] Updated weights for policy 0, policy_version 48003 (0.0010)
+[2026-06-06 22:35:52,921][261293] Fps is (10 sec: 4095.9, 60 sec: 4573.8, 300 sec: 2936.6). Total num frames: 24578048. Throughput: 0: 4551.1. Samples: 24580760. Policy #0 lag: (min: 3.0, avg: 19.5, max: 35.0)
+[2026-06-06 22:35:52,923][261293] Avg episode reward: [(0, '1711.814')]
+[2026-06-06 22:35:54,016][262198] Updated weights for policy 0, policy_version 48013 (0.0008)
+[2026-06-06 22:35:54,615][262198] Updated weights for policy 0, policy_version 48023 (0.0007)
+[2026-06-06 22:35:55,894][262198] Updated weights for policy 0, policy_version 48033 (0.0008)
+[2026-06-06 22:35:56,839][262198] Updated weights for policy 0, policy_version 48043 (0.0007)
+[2026-06-06 22:35:57,921][261293] Fps is (10 sec: 4300.6, 60 sec: 4539.7, 300 sec: 2971.3). Total num frames: 24600576. Throughput: 0: 4473.3. Samples: 24609170. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:35:57,924][261293] Avg episode reward: [(0, '1801.277')]
+[2026-06-06 22:35:58,301][262198] Updated weights for policy 0, policy_version 48053 (0.0009)
+[2026-06-06 22:35:59,863][262198] Updated weights for policy 0, policy_version 48065 (0.0007)
+[2026-06-06 22:36:00,654][262198] Updated weights for policy 0, policy_version 48075 (0.0008)
+[2026-06-06 22:36:01,947][262198] Updated weights for policy 0, policy_version 48085 (0.0007)
+[2026-06-06 22:36:02,770][262198] Updated weights for policy 0, policy_version 48095 (0.0007)
+[2026-06-06 22:36:02,926][261293] Fps is (10 sec: 4708.1, 60 sec: 4574.0, 300 sec: 3012.9). Total num frames: 24625152. Throughput: 0: 4419.5. Samples: 24621752. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:36:02,930][261293] Avg episode reward: [(0, '1648.155')]
+[2026-06-06 22:36:04,213][262198] Updated weights for policy 0, policy_version 48105 (0.0007)
+[2026-06-06 22:36:05,016][262198] Updated weights for policy 0, policy_version 48115 (0.0010)
+[2026-06-06 22:36:06,233][262198] Updated weights for policy 0, policy_version 48125 (0.0010)
+[2026-06-06 22:36:07,183][262198] Updated weights for policy 0, policy_version 48135 (0.0007)
+[2026-06-06 22:36:07,920][261293] Fps is (10 sec: 4710.8, 60 sec: 4539.7, 300 sec: 3026.9). Total num frames: 24647680. Throughput: 0: 4401.9. Samples: 24649862. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:36:07,922][261293] Avg episode reward: [(0, '1838.010')]
+[2026-06-06 22:36:08,496][262198] Updated weights for policy 0, policy_version 48151 (0.0007)
+[2026-06-06 22:36:10,055][262198] Updated weights for policy 0, policy_version 48162 (0.0007)
+[2026-06-06 22:36:10,893][262198] Updated weights for policy 0, policy_version 48172 (0.0008)
+[2026-06-06 22:36:11,595][262198] Updated weights for policy 0, policy_version 48182 (0.0009)
+[2026-06-06 22:36:12,644][262198] Updated weights for policy 0, policy_version 48192 (0.0009)
+[2026-06-06 22:36:12,921][261293] Fps is (10 sec: 4917.5, 60 sec: 4573.9, 300 sec: 3089.4). Total num frames: 24674304. Throughput: 0: 4435.2. Samples: 24680254. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:36:12,923][261293] Avg episode reward: [(0, '1675.575')]
+[2026-06-06 22:36:14,125][262198] Updated weights for policy 0, policy_version 48202 (0.0007)
+[2026-06-06 22:36:14,654][262198] Updated weights for policy 0, policy_version 48212 (0.0007)
+[2026-06-06 22:36:15,494][262198] Updated weights for policy 0, policy_version 48222 (0.0008)
+[2026-06-06 22:36:17,221][262198] Updated weights for policy 0, policy_version 48232 (0.0008)
+[2026-06-06 22:36:17,884][262198] Updated weights for policy 0, policy_version 48242 (0.0007)
+[2026-06-06 22:36:17,921][261293] Fps is (10 sec: 5324.3, 60 sec: 4608.0, 300 sec: 3137.9). Total num frames: 24700928. Throughput: 0: 4507.8. Samples: 24696310. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:36:17,926][261293] Avg episode reward: [(0, '1720.268')]
+[2026-06-06 22:36:18,775][262198] Updated weights for policy 0, policy_version 48252 (0.0007)
+[2026-06-06 22:36:20,410][262198] Updated weights for policy 0, policy_version 48262 (0.0009)
+[2026-06-06 22:36:21,199][262198] Updated weights for policy 0, policy_version 48273 (0.0007)
+[2026-06-06 22:36:22,108][262198] Updated weights for policy 0, policy_version 48283 (0.0007)
+[2026-06-06 22:36:22,921][261293] Fps is (10 sec: 5120.0, 60 sec: 4608.0, 300 sec: 3179.6). Total num frames: 24725504. Throughput: 0: 4601.6. Samples: 24725708. Policy #0 lag: (min: 0.0, avg: 20.5, max: 44.0)
+[2026-06-06 22:36:22,922][261293] Avg episode reward: [(0, '1701.080')]
+[2026-06-06 22:36:23,586][262198] Updated weights for policy 0, policy_version 48293 (0.0007)
+[2026-06-06 22:36:24,129][262198] Updated weights for policy 0, policy_version 48304 (0.0009)
+[2026-06-06 22:36:25,088][262198] Updated weights for policy 0, policy_version 48314 (0.0007)
+[2026-06-06 22:36:26,055][262198] Updated weights for policy 0, policy_version 48324 (0.0008)
+[2026-06-06 22:36:27,619][262198] Updated weights for policy 0, policy_version 48338 (0.0007)
+[2026-06-06 22:36:27,921][261293] Fps is (10 sec: 5120.4, 60 sec: 4642.1, 300 sec: 3235.1). Total num frames: 24752128. Throughput: 0: 4743.2. Samples: 24757124. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:27,924][261293] Avg episode reward: [(0, '1751.872')]
+[2026-06-06 22:36:28,840][262198] Updated weights for policy 0, policy_version 48349 (0.0008)
+[2026-06-06 22:36:30,231][262198] Updated weights for policy 0, policy_version 48359 (0.0009)
+[2026-06-06 22:36:30,366][262115] Signal inference workers to stop experience collection... (4500 times)
+[2026-06-06 22:36:30,382][262198] InferenceWorker_p0-w0: stopping experience collection (4500 times)
+[2026-06-06 22:36:30,389][262115] Signal inference workers to resume experience collection... (4500 times)
+[2026-06-06 22:36:30,390][262198] InferenceWorker_p0-w0: resuming experience collection (4500 times)
+[2026-06-06 22:36:31,524][262198] Updated weights for policy 0, policy_version 48374 (0.0009)
+[2026-06-06 22:36:32,772][262198] Updated weights for policy 0, policy_version 48385 (0.0008)
+[2026-06-06 22:36:32,921][261293] Fps is (10 sec: 4915.1, 60 sec: 4676.3, 300 sec: 3276.8). Total num frames: 24774656. Throughput: 0: 4778.5. Samples: 24770770. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:32,925][261293] Avg episode reward: [(0, '1834.600')]
+[2026-06-06 22:36:34,069][262198] Updated weights for policy 0, policy_version 48397 (0.0009)
+[2026-06-06 22:36:35,677][262198] Updated weights for policy 0, policy_version 48416 (0.0008)
+[2026-06-06 22:36:37,196][262198] Updated weights for policy 0, policy_version 48429 (0.0008)
+[2026-06-06 22:36:37,920][261293] Fps is (10 sec: 4710.5, 60 sec: 4744.6, 300 sec: 3325.4). Total num frames: 24799232. Throughput: 0: 4891.1. Samples: 24800860. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:37,922][261293] Avg episode reward: [(0, '1792.856')]
+[2026-06-06 22:36:38,278][262198] Updated weights for policy 0, policy_version 48444 (0.0009)
+[2026-06-06 22:36:39,874][262198] Updated weights for policy 0, policy_version 48454 (0.0008)
+[2026-06-06 22:36:40,573][262198] Updated weights for policy 0, policy_version 48464 (0.0007)
+[2026-06-06 22:36:41,951][262198] Updated weights for policy 0, policy_version 48479 (0.0008)
+[2026-06-06 22:36:42,921][261293] Fps is (10 sec: 5120.2, 60 sec: 4812.8, 300 sec: 3387.9). Total num frames: 24825856. Throughput: 0: 4938.8. Samples: 24831412. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:42,927][261293] Avg episode reward: [(0, '1721.757')]
+[2026-06-06 22:36:43,408][262198] Updated weights for policy 0, policy_version 48489 (0.0008)
+[2026-06-06 22:36:44,402][262198] Updated weights for policy 0, policy_version 48502 (0.0007)
+[2026-06-06 22:36:46,621][262198] Updated weights for policy 0, policy_version 48515 (0.0014)
+[2026-06-06 22:36:47,921][261293] Fps is (10 sec: 4300.7, 60 sec: 4744.6, 300 sec: 3401.8). Total num frames: 24842240. Throughput: 0: 4987.4. Samples: 24846160. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:47,925][261293] Avg episode reward: [(0, '1946.902')]
+[2026-06-06 22:36:49,814][262198] Updated weights for policy 0, policy_version 48527 (0.0019)
+[2026-06-06 22:36:52,659][262198] Updated weights for policy 0, policy_version 48538 (0.0014)
+[2026-06-06 22:36:52,920][261293] Fps is (10 sec: 2662.5, 60 sec: 4573.9, 300 sec: 3408.8). Total num frames: 24852480. Throughput: 0: 4620.8. Samples: 24857796. Policy #0 lag: (min: 3.0, avg: 20.4, max: 36.0)
+[2026-06-06 22:36:52,922][261293] Avg episode reward: [(0, '1795.539')]
+[2026-06-06 22:36:56,072][262198] Updated weights for policy 0, policy_version 48552 (0.0014)
+[2026-06-06 22:36:57,920][261293] Fps is (10 sec: 1843.2, 60 sec: 4335.0, 300 sec: 3394.8). Total num frames: 24860672. Throughput: 0: 4199.7. Samples: 24869238. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:36:57,922][261293] Avg episode reward: [(0, '1667.930')]
+[2026-06-06 22:36:59,760][262198] Updated weights for policy 0, policy_version 48565 (0.0016)
+[2026-06-06 22:37:02,920][261293] Fps is (10 sec: 1843.2, 60 sec: 4096.3, 300 sec: 3387.9). Total num frames: 24870912. Throughput: 0: 3974.9. Samples: 24875176. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:02,922][261293] Avg episode reward: [(0, '1756.199')]
+[2026-06-06 22:37:03,440][262198] Updated weights for policy 0, policy_version 48578 (0.0012)
+[2026-06-06 22:37:06,333][262198] Updated weights for policy 0, policy_version 48591 (0.0010)
+[2026-06-06 22:37:07,920][261293] Fps is (10 sec: 2252.8, 60 sec: 3925.3, 300 sec: 3387.9). Total num frames: 24883200. Throughput: 0: 3595.7. Samples: 24887516. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:07,922][261293] Avg episode reward: [(0, '1776.523')]
+[2026-06-06 22:37:09,119][262198] Updated weights for policy 0, policy_version 48603 (0.0010)
+[2026-06-06 22:37:11,676][262198] Updated weights for policy 0, policy_version 48616 (0.0011)
+[2026-06-06 22:37:12,921][261293] Fps is (10 sec: 2047.9, 60 sec: 3618.1, 300 sec: 3387.9). Total num frames: 24891392. Throughput: 0: 3198.8. Samples: 24901070. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:12,922][261293] Avg episode reward: [(0, '1761.390')]
+[2026-06-06 22:37:14,726][262198] Updated weights for policy 0, policy_version 48629 (0.0011)
+[2026-06-06 22:37:16,352][262198] Updated weights for policy 0, policy_version 48640 (0.0011)
+[2026-06-06 22:37:17,920][261293] Fps is (10 sec: 2048.0, 60 sec: 3379.2, 300 sec: 3387.9). Total num frames: 24903680. Throughput: 0: 3053.3. Samples: 24908166. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:17,923][261293] Avg episode reward: [(0, '1705.768')]
+[2026-06-06 22:37:20,232][262198] Updated weights for policy 0, policy_version 48652 (0.0010)
+[2026-06-06 22:37:22,165][262198] Updated weights for policy 0, policy_version 48664 (0.0012)
+[2026-06-06 22:37:22,920][261293] Fps is (10 sec: 2457.7, 60 sec: 3174.4, 300 sec: 3387.9). Total num frames: 24915968. Throughput: 0: 2691.4. Samples: 24921974. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:22,921][261293] Avg episode reward: [(0, '1783.464')]
+[2026-06-06 22:37:25,289][262198] Updated weights for policy 0, policy_version 48676 (0.0012)
+[2026-06-06 22:37:27,920][261293] Fps is (10 sec: 2252.8, 60 sec: 2901.3, 300 sec: 3387.9). Total num frames: 24926208. Throughput: 0: 2308.1. Samples: 24935276. Policy #0 lag: (min: 0.0, avg: 16.2, max: 32.0)
+[2026-06-06 22:37:27,922][261293] Avg episode reward: [(0, '1714.586')]
+[2026-06-06 22:37:27,957][262198] Updated weights for policy 0, policy_version 48687 (0.0012)
+[2026-06-06 22:37:30,554][262198] Updated weights for policy 0, policy_version 48700 (0.0010)
+[2026-06-06 22:37:32,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2730.7, 300 sec: 3387.9). Total num frames: 24938496. Throughput: 0: 2132.2. Samples: 24942108. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:32,924][261293] Avg episode reward: [(0, '1691.070')]
+[2026-06-06 22:37:34,863][262198] Updated weights for policy 0, policy_version 48713 (0.0009)
+[2026-06-06 22:37:36,509][262198] Updated weights for policy 0, policy_version 48727 (0.0012)
+[2026-06-06 22:37:37,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2525.9, 300 sec: 3387.9). Total num frames: 24950784. Throughput: 0: 2174.7. Samples: 24955656. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:37,922][261293] Avg episode reward: [(0, '1778.740')]
+[2026-06-06 22:37:39,546][262198] Updated weights for policy 0, policy_version 48739 (0.0013)
+[2026-06-06 22:37:41,980][262198] Updated weights for policy 0, policy_version 48751 (0.0009)
+[2026-06-06 22:37:42,921][261293] Fps is (10 sec: 2457.6, 60 sec: 2286.9, 300 sec: 3394.8). Total num frames: 24963072. Throughput: 0: 2250.2. Samples: 24970498. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:42,923][261293] Avg episode reward: [(0, '1641.664')]
+[2026-06-06 22:37:44,293][262198] Updated weights for policy 0, policy_version 48763 (0.0011)
+[2026-06-06 22:37:47,118][262198] Updated weights for policy 0, policy_version 48776 (0.0012)
+[2026-06-06 22:37:47,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2184.5, 300 sec: 3387.9). Total num frames: 24973312. Throughput: 0: 2281.2. Samples: 24977832. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:47,924][261293] Avg episode reward: [(0, '1849.382')]
+[2026-06-06 22:37:49,665][262198] Updated weights for policy 0, policy_version 48788 (0.0012)
+[2026-06-06 22:37:52,921][261293] Fps is (10 sec: 2252.8, 60 sec: 2218.7, 300 sec: 3387.9). Total num frames: 24985600. Throughput: 0: 2326.7. Samples: 24992216. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:52,922][261293] Avg episode reward: [(0, '1849.450')]
+[2026-06-06 22:37:52,996][262198] Updated weights for policy 0, policy_version 48801 (0.0012)
+[2026-06-06 22:37:55,462][262198] Updated weights for policy 0, policy_version 48813 (0.0009)
+[2026-06-06 22:37:57,549][262198] Updated weights for policy 0, policy_version 48825 (0.0009)
+[2026-06-06 22:37:57,920][261293] Fps is (10 sec: 2662.4, 60 sec: 2321.1, 300 sec: 3408.7). Total num frames: 24999936. Throughput: 0: 2353.1. Samples: 25006958. Policy #0 lag: (min: 2.0, avg: 18.8, max: 34.0)
+[2026-06-06 22:37:57,923][261293] Avg episode reward: [(0, '1523.485')]
+[2026-06-06 22:37:59,111][262115] Stopping Batcher_0...
+[2026-06-06 22:37:59,111][261293] Component Batcher_0 stopped!
+[2026-06-06 22:37:59,112][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000048836_25004032.pth...
+[2026-06-06 22:37:59,113][262115] Loop batcher_evt_loop terminating...
+[2026-06-06 22:37:59,124][262198] Updated weights for policy 0, policy_version 48836 (0.0009)
+[2026-06-06 22:37:59,186][262115] Removing results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000030012_15366144.pth
+[2026-06-06 22:37:59,193][262202] Stopping RolloutWorker_w2...
+[2026-06-06 22:37:59,193][262202] Loop rollout_proc2_evt_loop terminating...
+[2026-06-06 22:37:59,193][261293] Component RolloutWorker_w2 stopped!
+[2026-06-06 22:37:59,196][262115] Saving results/checkpoints_factor_sweeps/deadly_corridor/context_window/deadly_corridor_frame_stack_uniform_u2_6_fs12_seed10/checkpoint_p0/checkpoint_000048836_25004032.pth...
+[2026-06-06 22:37:59,199][261293] Component RolloutWorker_w21 stopped!
+[2026-06-06 22:37:59,199][263229] Stopping RolloutWorker_w18...
+[2026-06-06 22:37:59,199][263232] Stopping RolloutWorker_w21...
+[2026-06-06 22:37:59,199][262206] Stopping RolloutWorker_w4...
+[2026-06-06 22:37:59,200][261293] Component RolloutWorker_w18 stopped!
+[2026-06-06 22:37:59,200][263229] Loop rollout_proc18_evt_loop terminating...
+[2026-06-06 22:37:59,201][263232] Loop rollout_proc21_evt_loop terminating...
+[2026-06-06 22:37:59,201][262206] Loop rollout_proc4_evt_loop terminating...
+[2026-06-06 22:37:59,201][261293] Component RolloutWorker_w4 stopped!
+[2026-06-06 22:37:59,201][263240] Stopping RolloutWorker_w28...
+[2026-06-06 22:37:59,201][262211] Stopping RolloutWorker_w12...
+[2026-06-06 22:37:59,202][263240] Loop rollout_proc28_evt_loop terminating...
+[2026-06-06 22:37:59,202][262211] Loop rollout_proc12_evt_loop terminating...
+[2026-06-06 22:37:59,202][261293] Component RolloutWorker_w12 stopped!
+[2026-06-06 22:37:59,203][261293] Component RolloutWorker_w28 stopped!
+[2026-06-06 22:37:59,203][263230] Stopping RolloutWorker_w19...
+[2026-06-06 22:37:59,204][263230] Loop rollout_proc19_evt_loop terminating...
+[2026-06-06 22:37:59,204][261293] Component RolloutWorker_w5 stopped!
+[2026-06-06 22:37:59,203][262205] Stopping RolloutWorker_w5...
+[2026-06-06 22:37:59,204][263235] Stopping RolloutWorker_w23...
+[2026-06-06 22:37:59,205][263235] Loop rollout_proc23_evt_loop terminating...
+[2026-06-06 22:37:59,205][261293] Component RolloutWorker_w19 stopped!
+[2026-06-06 22:37:59,205][262205] Loop rollout_proc5_evt_loop terminating...
+[2026-06-06 22:37:59,206][261293] Component RolloutWorker_w23 stopped!
+[2026-06-06 22:37:59,206][261293] Component RolloutWorker_w20 stopped!
+[2026-06-06 22:37:59,206][263231] Stopping RolloutWorker_w20...
+[2026-06-06 22:37:59,207][261293] Component RolloutWorker_w1 stopped!
+[2026-06-06 22:37:59,207][263231] Loop rollout_proc20_evt_loop terminating...
+[2026-06-06 22:37:59,208][261293] Component RolloutWorker_w15 stopped!
+[2026-06-06 22:37:59,207][263226] Stopping RolloutWorker_w15...
+[2026-06-06 22:37:59,207][262200] Stopping RolloutWorker_w1...
+[2026-06-06 22:37:59,208][263226] Loop rollout_proc15_evt_loop terminating...
+[2026-06-06 22:37:59,208][261293] Component RolloutWorker_w8 stopped!
+[2026-06-06 22:37:59,209][262200] Loop rollout_proc1_evt_loop terminating...
+[2026-06-06 22:37:59,208][262201] Stopping RolloutWorker_w8...
+[2026-06-06 22:37:59,209][261293] Component RolloutWorker_w10 stopped!
+[2026-06-06 22:37:59,209][262201] Loop rollout_proc8_evt_loop terminating...
+[2026-06-06 22:37:59,208][262208] Stopping RolloutWorker_w10...
+[2026-06-06 22:37:59,210][262208] Loop rollout_proc10_evt_loop terminating...
+[2026-06-06 22:37:59,209][263228] Stopping RolloutWorker_w17...
+[2026-06-06 22:37:59,210][261293] Component RolloutWorker_w17 stopped!
+[2026-06-06 22:37:59,210][263228] Loop rollout_proc17_evt_loop terminating...
+[2026-06-06 22:37:59,211][261293] Component RolloutWorker_w24 stopped!
+[2026-06-06 22:37:59,210][263236] Stopping RolloutWorker_w24...
+[2026-06-06 22:37:59,211][263236] Loop rollout_proc24_evt_loop terminating...
+[2026-06-06 22:37:59,216][261293] Component RolloutWorker_w14 stopped!
+[2026-06-06 22:37:59,216][261293] Component RolloutWorker_w31 stopped!
+[2026-06-06 22:37:59,216][262214] Stopping RolloutWorker_w14...
+[2026-06-06 22:37:59,216][264253] Stopping RolloutWorker_w31...
+[2026-06-06 22:37:59,217][261293] Component RolloutWorker_w16 stopped!
+[2026-06-06 22:37:59,217][262214] Loop rollout_proc14_evt_loop terminating...
+[2026-06-06 22:37:59,217][264253] Loop rollout_proc31_evt_loop terminating...
+[2026-06-06 22:37:59,218][261293] Component RolloutWorker_w7 stopped!
+[2026-06-06 22:37:59,217][262209] Stopping RolloutWorker_w7...
+[2026-06-06 22:37:59,217][263227] Stopping RolloutWorker_w16...
+[2026-06-06 22:37:59,218][262209] Loop rollout_proc7_evt_loop terminating...
+[2026-06-06 22:37:59,218][261293] Component RolloutWorker_w3 stopped!
+[2026-06-06 22:37:59,219][263227] Loop rollout_proc16_evt_loop terminating...
+[2026-06-06 22:37:59,218][262203] Stopping RolloutWorker_w3...
+[2026-06-06 22:37:59,220][262203] Loop rollout_proc3_evt_loop terminating...
+[2026-06-06 22:37:59,224][261293] Component RolloutWorker_w30 stopped!
+[2026-06-06 22:37:59,224][263242] Stopping RolloutWorker_w30...
+[2026-06-06 22:37:59,225][262212] Stopping RolloutWorker_w11...
+[2026-06-06 22:37:59,226][262212] Loop rollout_proc11_evt_loop terminating...
+[2026-06-06 22:37:59,225][261293] Component RolloutWorker_w11 stopped!
+[2026-06-06 22:37:59,226][263242] Loop rollout_proc30_evt_loop terminating...
+[2026-06-06 22:37:59,225][262199] Stopping RolloutWorker_w0...
+[2026-06-06 22:37:59,226][262199] Loop rollout_proc0_evt_loop terminating...
+[2026-06-06 22:37:59,226][261293] Component RolloutWorker_w0 stopped!
+[2026-06-06 22:37:59,227][261293] Component RolloutWorker_w25 stopped!
+[2026-06-06 22:37:59,228][261293] Component RolloutWorker_w9 stopped!
+[2026-06-06 22:37:59,227][263237] Stopping RolloutWorker_w25...
+[2026-06-06 22:37:59,229][263237] Loop rollout_proc25_evt_loop terminating...
+[2026-06-06 22:37:59,228][262204] Stopping RolloutWorker_w9...
+[2026-06-06 22:37:59,230][262204] Loop rollout_proc9_evt_loop terminating...
+[2026-06-06 22:37:59,232][263238] Stopping RolloutWorker_w26...
+[2026-06-06 22:37:59,232][263233] Stopping RolloutWorker_w22...
+[2026-06-06 22:37:59,233][263238] Loop rollout_proc26_evt_loop terminating...
+[2026-06-06 22:37:59,232][261293] Component RolloutWorker_w26 stopped!
+[2026-06-06 22:37:59,233][263233] Loop rollout_proc22_evt_loop terminating...
+[2026-06-06 22:37:59,232][263239] Stopping RolloutWorker_w27...
+[2026-06-06 22:37:59,234][263239] Loop rollout_proc27_evt_loop terminating...
+[2026-06-06 22:37:59,233][263241] Stopping RolloutWorker_w29...
+[2026-06-06 22:37:59,234][261293] Component RolloutWorker_w22 stopped!
+[2026-06-06 22:37:59,235][263241] Loop rollout_proc29_evt_loop terminating...
+[2026-06-06 22:37:59,235][261293] Component RolloutWorker_w27 stopped!
+[2026-06-06 22:37:59,235][261293] Component RolloutWorker_w29 stopped!
+[2026-06-06 22:37:59,241][261293] Component RolloutWorker_w6 stopped!
+[2026-06-06 22:37:59,241][262210] Stopping RolloutWorker_w6...
+[2026-06-06 22:37:59,242][262210] Loop rollout_proc6_evt_loop terminating...
+[2026-06-06 22:37:59,246][261293] Component RolloutWorker_w13 stopped!
+[2026-06-06 22:37:59,246][262213] Stopping RolloutWorker_w13...
+[2026-06-06 22:37:59,247][262213] Loop rollout_proc13_evt_loop terminating...
+[2026-06-06 22:37:59,277][262115] Stopping LearnerWorker_p0...
+[2026-06-06 22:37:59,278][262115] Loop learner_proc0_evt_loop terminating...
+[2026-06-06 22:37:59,278][261293] Component LearnerWorker_p0 stopped!
+[2026-06-06 22:37:59,332][262198] Weights refcount: 2 0
+[2026-06-06 22:37:59,334][262198] Stopping InferenceWorker_p0-w0...
+[2026-06-06 22:37:59,335][262198] Loop inference_proc0-0_evt_loop terminating...
+[2026-06-06 22:37:59,335][261293] Component InferenceWorker_p0-w0 stopped!
+[2026-06-06 22:37:59,336][261293] Waiting for process learner_proc0 to stop...
+[2026-06-06 22:38:07,606][261293] Waiting for process inference_proc0-0 to join...
+[2026-06-06 22:38:07,607][261293] Waiting for process rollout_proc0 to join...
+[2026-06-06 22:38:07,608][261293] Waiting for process rollout_proc1 to join...
+[2026-06-06 22:38:07,608][261293] Waiting for process rollout_proc2 to join...
+[2026-06-06 22:38:07,608][261293] Waiting for process rollout_proc3 to join...
+[2026-06-06 22:38:07,609][261293] Waiting for process rollout_proc4 to join...
+[2026-06-06 22:38:07,609][261293] Waiting for process rollout_proc5 to join...
+[2026-06-06 22:38:07,610][261293] Waiting for process rollout_proc6 to join...
+[2026-06-06 22:38:07,610][261293] Waiting for process rollout_proc7 to join...
+[2026-06-06 22:38:07,611][261293] Waiting for process rollout_proc8 to join...
+[2026-06-06 22:38:07,612][261293] Waiting for process rollout_proc9 to join...
+[2026-06-06 22:38:07,614][261293] Waiting for process rollout_proc10 to join...
+[2026-06-06 22:38:07,616][261293] Waiting for process rollout_proc11 to join...
+[2026-06-06 22:38:07,617][261293] Waiting for process rollout_proc12 to join...
+[2026-06-06 22:38:07,617][261293] Waiting for process rollout_proc13 to join...
+[2026-06-06 22:38:07,618][261293] Waiting for process rollout_proc14 to join...
+[2026-06-06 22:38:07,619][261293] Waiting for process rollout_proc15 to join...
+[2026-06-06 22:38:07,619][261293] Waiting for process rollout_proc16 to join...
+[2026-06-06 22:38:07,620][261293] Waiting for process rollout_proc17 to join...
+[2026-06-06 22:38:07,621][261293] Waiting for process rollout_proc18 to join...
+[2026-06-06 22:38:07,621][261293] Waiting for process rollout_proc19 to join...
+[2026-06-06 22:38:07,622][261293] Waiting for process rollout_proc20 to join...
+[2026-06-06 22:38:07,623][261293] Waiting for process rollout_proc21 to join...
+[2026-06-06 22:38:07,623][261293] Waiting for process rollout_proc22 to join...
+[2026-06-06 22:38:07,624][261293] Waiting for process rollout_proc23 to join...
+[2026-06-06 22:38:07,625][261293] Waiting for process rollout_proc24 to join...
+[2026-06-06 22:38:07,625][261293] Waiting for process rollout_proc25 to join...
+[2026-06-06 22:38:07,626][261293] Waiting for process rollout_proc26 to join...
+[2026-06-06 22:38:07,627][261293] Waiting for process rollout_proc27 to join...
+[2026-06-06 22:38:07,627][261293] Waiting for process rollout_proc28 to join...
+[2026-06-06 22:38:07,628][261293] Waiting for process rollout_proc29 to join...
+[2026-06-06 22:38:07,629][261293] Waiting for process rollout_proc30 to join...
+[2026-06-06 22:38:07,630][261293] Waiting for process rollout_proc31 to join...
+[2026-06-06 22:38:07,630][261293] Batcher 0 profile tree view:
+batching: 1981.7066, releasing_batches: 203.5042
+[2026-06-06 22:38:07,631][261293] InferenceWorker_p0-w0 profile tree view:
+wait_policy: 0.0001
+  wait_policy_total: 65.0990
+update_model: 64.0135
+  weight_update: 0.0009
+one_step: 0.0268
+  handle_policy_step: 6092.5528
+    deserialize: 2414.9108, stack: 14.8420, obs_to_device_normalize: 1822.4902, forward: 1040.7867, send_messages: 244.3134
+    prepare_outputs: 478.7376
+      to_cpu: 352.8365
+[2026-06-06 22:38:07,631][261293] Learner 0 profile tree view:
+misc: 0.1356, prepare_batch: 612.4141
+train: 2304.4804
+  epoch_init: 0.4360, minibatch_init: 7.9598, losses_postprocess: 116.2419, kl_divergence: 139.9722, after_optimizer: 181.9521
+  calculate_losses: 1209.1651
+    losses_init: 0.2526, forward_head: 75.6356, bptt_initial: 755.8484, tail: 70.5440, advantages_returns: 12.9654, losses: 127.7074
+    bptt: 155.1954
+      bptt_forward_core: 151.7918
+  update: 634.4732
+    clip: 53.5167
+[2026-06-06 22:38:07,631][261293] RolloutWorker_w0 profile tree view:
+wait_for_trajectories: 0.3005, enqueue_policy_requests: 122.4550, env_step: 3019.6744, overhead: 53.6382, complete_rollouts: 0.8382
+save_policy_outputs: 79.4437
+  split_output_tensors: 26.9117
+[2026-06-06 22:38:07,631][261293] RolloutWorker_w31 profile tree view:
+wait_for_trajectories: 0.2812, enqueue_policy_requests: 99.3231, env_step: 2994.5721, overhead: 48.5690, complete_rollouts: 0.8313
+save_policy_outputs: 74.5881
+  split_output_tensors: 25.0954
+[2026-06-06 22:38:07,633][261293] Loop Runner_EvtLoop terminating...
+[2026-06-06 22:38:07,634][261293] Runner profile tree view:
+main_loop: 6520.4336
+[2026-06-06 22:38:07,634][261293] Collected {0: 25004032}, FPS: 3834.7