Upload folder using huggingface_hub

.gitattributes

@@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+20260303-12:22:50-iql_antmaze-large-play-v0/run_d4rl_offline.log filter=lfs diff=lfs merge=lfs -text
+20260303-12:23:15-iql_antmaze-medium-diverse-v0/run_d4rl_offline.log filter=lfs diff=lfs merge=lfs -text

20260302-12:38:39-iql_halfcheetah-medium-v2/42.txt

@@ -0,0 +1,200 @@
+5000 38.7
+10000 29.5
+15000 30.5
+20000 24.8
+25000 28.4
+30000 40.8
+35000 37.6
+40000 39.5
+45000 41.4
+50000 43.2
+55000 40.5
+60000 42.8
+65000 43.5
+70000 44.0
+75000 45.1
+80000 44.4
+85000 44.5
+90000 44.3
+95000 45.0
+100000 42.9
+105000 45.2
+110000 45.5
+115000 44.5
+120000 45.6
+125000 44.7
+130000 45.0
+135000 45.6
+140000 45.3
+145000 45.8
+150000 45.5
+155000 46.1
+160000 45.5
+165000 45.8
+170000 46.1
+175000 45.8
+180000 47.1
+185000 46.3
+190000 46.5
+195000 46.3
+200000 46.3
+205000 46.6
+210000 46.0
+215000 46.5
+220000 46.8
+225000 46.8
+230000 47.0
+235000 47.1
+240000 46.9
+245000 46.7
+250000 46.7
+255000 47.6
+260000 46.3
+265000 47.3
+270000 47.3
+275000 42.2
+280000 46.4
+285000 47.4
+290000 47.3
+295000 47.5
+300000 47.3
+305000 46.9
+310000 46.9
+315000 47.4
+320000 47.0
+325000 46.8
+330000 47.3
+335000 46.7
+340000 47.3
+345000 47.1
+350000 46.7
+355000 46.5
+360000 46.7
+365000 46.8
+370000 47.1
+375000 47.0
+380000 47.6
+385000 46.8
+390000 46.4
+395000 46.6
+400000 47.1
+405000 46.9
+410000 47.9
+415000 47.4
+420000 47.1
+425000 47.2
+430000 47.2
+435000 47.5
+440000 46.6
+445000 47.6
+450000 47.8
+455000 47.0
+460000 46.5
+465000 46.7
+470000 46.9
+475000 47.1
+480000 47.8
+485000 47.0
+490000 47.7
+495000 46.7
+500000 47.4
+505000 47.7
+510000 47.1
+515000 47.3
+520000 47.3
+525000 47.6
+530000 47.0
+535000 47.4
+540000 47.7
+545000 47.6
+550000 47.6
+555000 47.0
+560000 48.0
+565000 47.7
+570000 47.4
+575000 47.8
+580000 47.6
+585000 47.6
+590000 47.8
+595000 47.5
+600000 47.6
+605000 47.8
+610000 47.4
+615000 47.1
+620000 47.6
+625000 47.4
+630000 46.8
+635000 47.1
+640000 47.4
+645000 47.7
+650000 47.5
+655000 47.8
+660000 47.7
+665000 47.5
+670000 47.5
+675000 47.5
+680000 47.8
+685000 47.4
+690000 47.5
+695000 47.8
+700000 47.3
+705000 47.2
+710000 47.4
+715000 47.7
+720000 47.5
+725000 47.6
+730000 48.3
+735000 47.6
+740000 47.9
+745000 47.7
+750000 47.2
+755000 47.5
+760000 47.9
+765000 48.0
+770000 47.7
+775000 47.7
+780000 47.8
+785000 47.2
+790000 47.7
+795000 47.9
+800000 47.8
+805000 48.1
+810000 47.5
+815000 47.9
+820000 48.3
+825000 47.6
+830000 47.3
+835000 47.6
+840000 47.9
+845000 47.7
+850000 47.5
+855000 47.3
+860000 47.3
+865000 47.9
+870000 47.6
+875000 47.9
+880000 47.5
+885000 47.8
+890000 47.9
+895000 47.8
+900000 47.4
+905000 47.4
+910000 47.2
+915000 48.0
+920000 47.8
+925000 47.5
+930000 47.9
+935000 47.4
+940000 48.1
+945000 47.9
+950000 47.6
+955000 47.6
+960000 48.0
+965000 47.3
+970000 48.1
+975000 47.6
+980000 47.2
+985000 47.8
+990000 48.0
+995000 47.6
+1000000 47.6

20260302-12:38:39-iql_halfcheetah-medium-v2/run_iql.log

@@ -0,0 +1,145 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:39-iql_halfcheetah-medium-v2 runner.logger.experiment_name=iql_halfcheetah-medium-v2 env_name=halfcheetah-medium-v2
+2026-03-02 12:38:40.892033: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 12:38:40.933348: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 12:38:40.933640: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 12:38:40.934774: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 12:38:40.940961: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 12:38:42.765864: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 12:38:48,786	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8265 
+[INFO 12:38:50 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: 66b30de6a0de14d2f5da98cb53abfb8bb96d76936cd911fe86ce72f8
+node_ip: 10.193.248.212
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 12:38:50 FlexiblePlacementStrategy] 
+[INFO 12:38:50 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 12:38:50 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=1637692) 2026-03-02 12:38:52.431580: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=1637692) 2026-03-02 12:38:52.454823: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=1637692) 2026-03-02 12:38:52.454851: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=1637692) 2026-03-02 12:38:52.455486: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=1637692) 2026-03-02 12:38:52.459441: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=1637692) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=1637692) 2026-03-02 12:38:54.422716: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=1637692) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=1637692) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=1637692) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=1637692) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=1637692)   from distutils.dep_util import newer, newer_group
+(pid=1637692) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=1637692) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=1637692) No module named 'flow'
+(pid=1637692) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=1637692) No module named 'carla'
+(pid=1637692) pybullet build time: Jan 29 2025 23:17:20
+(pid=1637692) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=1637692)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=1637692) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=1637692)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=1637692) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=1637692)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+(EmbodiedIQLFSDPPolicy pid=1637692) 
  0%|          | 0/999000 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
  7%|▋         | 68773/999000 [00:00<00:01, 687692.95it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 14%|█▍        | 138215/999000 [00:00<00:01, 691629.00it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 21%|██        | 207378/999000 [00:00<00:01, 679613.06it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 28%|██▊       | 275364/999000 [00:00<00:01, 674034.05it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 34%|███▍      | 343585/999000 [00:00<00:00, 676941.60it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 41%|████      | 411293/999000 [00:00<00:00, 671520.25it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 48%|████▊     | 478713/999000 [00:00<00:00, 672379.75it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 55%|█████▍    | 546179/999000 [00:00<00:00, 673093.37it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 61%|██████▏   | 613496/999000 [00:00<00:00, 668284.93it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 68%|██████▊   | 681865/999000 [00:01<00:00, 672995.56it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 75%|███████▍  | 749176/999000 [00:01<00:00, 671941.24it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 82%|████████▏ | 816378/999000 [00:01<00:00, 614387.74it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
 89%|████████▊ | 885586/999000 [00:01<00:00, 636392.86it/s]
 95%|█████████▌| 950092/999000 [00:01<00:00, 638884.34it/s]
+(EmbodiedIQLFSDPPolicy pid=1637692) 
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:39-iql_halfcheetah-medium-v2",
+      "experiment_name": "iql_halfcheetah-medium-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:39-iql_halfcheetah-medium-v2",
+  "env_name": "halfcheetah-medium-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-12:38:39-iql_halfcheetah-medium-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:39-iql_halfcheetah-medium-v2
+    experiment_name: iql_halfcheetah-medium-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:39-iql_halfcheetah-medium-v2
+env_name: halfcheetah-medium-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-12:38:39-iql_halfcheetah-medium-v2/tensorboard/events.out.tfevents.1772426330.dps799-Legion-Y9000P-IRX9.1633756.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efaeb71c8810db1c1b6a8e9e7f106322d7d6e75b84217fefd446d1f2a843c87d
+size 467877892

20260302-12:38:40-iql_antmaze-umaze-v0/42.txt

@@ -0,0 +1,10 @@
+100000 60.0
+200000 86.0
+300000 80.0
+400000 83.0
+500000 88.0
+600000 85.0
+700000 93.0
+800000 87.0
+900000 87.0
+1000000 91.0

20260302-12:38:40-iql_antmaze-umaze-v0/run_iql.log

@@ -0,0 +1,136 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_antmaze_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:40-iql_antmaze-umaze-v0 runner.logger.experiment_name=iql_antmaze-umaze-v0 env_name=antmaze-umaze-v0
+2026-03-02 12:38:42.232862: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 12:38:42.255981: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 12:38:42.256005: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 12:38:42.256611: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 12:38:42.260626: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 12:38:44.001242: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 12:38:53,443	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8266 
+[INFO 12:38:55 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: eb37689ebdb7105b29ef611ca0cc26b90cbf1a91caaa4897c64ccd5c
+node_ip: 10.193.248.212
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 12:38:55 FlexiblePlacementStrategy] 
+[INFO 12:38:55 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 12:38:55 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=1639712) 2026-03-02 12:38:57.310556: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=1639712) 2026-03-02 12:38:57.335686: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=1639712) 2026-03-02 12:38:57.335718: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=1639712) 2026-03-02 12:38:57.336394: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=1639712) 2026-03-02 12:38:57.340496: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=1639712) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=1639712) 2026-03-02 12:38:59.189780: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=1639712) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=1639712) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=1639712) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=1639712)   from distutils.dep_util import newer, newer_group
+(pid=1639712) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=1639712) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=1639712) No module named 'flow'
+(pid=1639712) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=1639712) No module named 'carla'
+(pid=1639712) pybullet build time: Jan 29 2025 23:17:20
+(pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=1639712)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/envs/registration.py:505: UserWarning: WARN: The environment antmaze-umaze-v0 is out of date. You should consider upgrading to version `v2` with the environment ID `antmaze-umaze-v2`.
+(EmbodiedIQLFSDPPolicy pid=1639712)   logger.warn(
+(EmbodiedIQLFSDPPolicy pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/d4rl/locomotion/ant.py:180: UserWarning: This environment is deprecated. Please use the most recent version of this environment.
+(EmbodiedIQLFSDPPolicy pid=1639712)   offline_env.OfflineEnv.__init__(self, **kwargs)
+(EmbodiedIQLFSDPPolicy pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=1639712)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=1639712)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=1639712) 
+(EmbodiedIQLFSDPPolicy pid=1639712) 
+(EmbodiedIQLFSDPPolicy pid=1639712) 
+(EmbodiedIQLFSDPPolicy pid=1639712) 
+(EmbodiedIQLFSDPPolicy pid=1639712) 
+(EmbodiedIQLFSDPPolicy pid=1639712) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/utils/seeding.py:38: DeprecationWarning: WARN: Function `rng.randn(*size)` is marked as deprecated and will be removed in the future. Please use `rng.standard_normal(size)` instead.
+(EmbodiedIQLFSDPPolicy pid=1639712)   deprecation(
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:40-iql_antmaze-umaze-v0",
+      "experiment_name": "iql_antmaze-umaze-v0"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:40-iql_antmaze-umaze-v0",
+  "env_name": "antmaze-umaze-v0",
+  "seed": 42,
+  "eval_episodes": 100,
+  "log_interval": 1000,
+  "eval_interval": 100000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.9,
+    "temperature": 10.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}
+(EmbodiedIQLFSDPPolicy pid=1639712) Target Goal:  (0.9098989388597016, 8.458488603882019)

20260302-12:38:40-iql_antmaze-umaze-v0/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:40-iql_antmaze-umaze-v0
+    experiment_name: iql_antmaze-umaze-v0
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-12:38:40-iql_antmaze-umaze-v0
+env_name: antmaze-umaze-v0
+seed: 42
+eval_episodes: 100
+log_interval: 1000
+eval_interval: 100000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.9
+  temperature: 10.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-12:38:40-iql_antmaze-umaze-v0/tensorboard/events.out.tfevents.1772426335.dps799-Legion-Y9000P-IRX9.1633803.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcc80cf4bb2ff8a41f609e2b82b95fd502a0312b7eafb55ca367a41e89fb4e08
+size 467852818

20260302-15:55:04-iql_walker2d-medium-v2/42.txt

@@ -0,0 +1,200 @@
+5000 73.8
+10000 59.8
+15000 32.6
+20000 57.6
+25000 61.6
+30000 68.9
+35000 58.4
+40000 53.2
+45000 66.0
+50000 61.3
+55000 65.4
+60000 52.7
+65000 60.9
+70000 55.6
+75000 59.3
+80000 61.7
+85000 65.0
+90000 56.0
+95000 73.2
+100000 57.2
+105000 60.4
+110000 59.7
+115000 65.8
+120000 73.7
+125000 68.4
+130000 71.5
+135000 68.8
+140000 75.2
+145000 69.4
+150000 72.2
+155000 67.7
+160000 75.5
+165000 81.5
+170000 65.3
+175000 74.0
+180000 71.5
+185000 73.9
+190000 78.2
+195000 78.7
+200000 72.5
+205000 67.5
+210000 69.9
+215000 78.4
+220000 80.7
+225000 79.4
+230000 68.2
+235000 73.3
+240000 82.0
+245000 74.9
+250000 67.6
+255000 74.5
+260000 67.5
+265000 82.5
+270000 79.7
+275000 73.4
+280000 78.0
+285000 80.0
+290000 71.5
+295000 78.4
+300000 84.7
+305000 75.2
+310000 78.9
+315000 61.6
+320000 68.4
+325000 71.4
+330000 67.9
+335000 62.8
+340000 72.2
+345000 80.1
+350000 82.4
+355000 76.3
+360000 85.2
+365000 74.2
+370000 70.2
+375000 81.1
+380000 84.2
+385000 77.8
+390000 81.0
+395000 82.1
+400000 80.9
+405000 85.3
+410000 84.9
+415000 79.7
+420000 82.7
+425000 80.1
+430000 81.2
+435000 76.1
+440000 74.7
+445000 81.5
+450000 71.3
+455000 76.1
+460000 80.2
+465000 76.8
+470000 79.3
+475000 72.0
+480000 76.2
+485000 85.1
+490000 75.6
+495000 72.9
+500000 77.1
+505000 78.8
+510000 77.6
+515000 79.3
+520000 85.4
+525000 71.7
+530000 69.0
+535000 81.4
+540000 67.1
+545000 73.9
+550000 82.4
+555000 77.9
+560000 77.3
+565000 84.0
+570000 81.6
+575000 77.3
+580000 80.1
+585000 81.1
+590000 80.1
+595000 73.5
+600000 79.6
+605000 77.7
+610000 78.5
+615000 85.3
+620000 81.9
+625000 77.4
+630000 79.5
+635000 77.2
+640000 82.9
+645000 85.3
+650000 80.2
+655000 82.7
+660000 83.6
+665000 81.2
+670000 84.6
+675000 81.9
+680000 84.5
+685000 82.5
+690000 84.1
+695000 81.3
+700000 84.7
+705000 79.6
+710000 83.1
+715000 82.6
+720000 75.6
+725000 79.2
+730000 74.7
+735000 83.1
+740000 79.4
+745000 84.2
+750000 85.0
+755000 85.4
+760000 74.2
+765000 80.8
+770000 85.5
+775000 82.9
+780000 80.4
+785000 65.1
+790000 79.6
+795000 81.9
+800000 78.4
+805000 80.6
+810000 76.2
+815000 75.3
+820000 80.9
+825000 73.8
+830000 75.6
+835000 86.1
+840000 66.1
+845000 82.0
+850000 83.2
+855000 72.3
+860000 81.6
+865000 80.8
+870000 82.3
+875000 85.0
+880000 81.7
+885000 79.6
+890000 84.1
+895000 78.4
+900000 82.3
+905000 73.3
+910000 85.8
+915000 85.3
+920000 81.8
+925000 79.9
+930000 76.6
+935000 75.6
+940000 80.9
+945000 80.2
+950000 76.2
+955000 73.1
+960000 80.0
+965000 76.6
+970000 85.8
+975000 80.1
+980000 83.6
+985000 84.6
+990000 79.8
+995000 85.4
+1000000 78.5

20260302-15:55:04-iql_walker2d-medium-v2/run_iql.log

@@ -0,0 +1,143 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-15:55:04-iql_walker2d-medium-v2 runner.logger.experiment_name=iql_walker2d-medium-v2 env_name=walker2d-medium-v2
+2026-03-02 15:55:06.004436: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 15:55:06.030273: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 15:55:06.030303: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 15:55:06.031007: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 15:55:06.035442: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 15:55:07.979787: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 15:55:17,976	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8266 
+[INFO 15:55:20 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: e3db06c217f2aec909d97e1e80278d7d139dd8c5353c1b32e1f16a49
+node_ip: 10.193.248.212
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 15:55:20 FlexiblePlacementStrategy] 
+[INFO 15:55:20 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 15:55:20 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=526529) 2026-03-02 15:55:23.289230: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=526529) 2026-03-02 15:55:23.328259: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=526529) 2026-03-02 15:55:23.328313: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=526529) 2026-03-02 15:55:23.329268: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=526529) 2026-03-02 15:55:23.334914: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=526529) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=526529) 2026-03-02 15:55:25.607595: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=526529) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=526529) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=526529) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=526529) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=526529)   from distutils.dep_util import newer, newer_group
+(pid=526529) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=526529) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=526529) No module named 'flow'
+(pid=526529) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=526529) No module named 'carla'
+(pid=526529) pybullet build time: Jan 29 2025 23:17:20
+(pid=526529) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=526529)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=526529) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=526529)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=526529) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=526529)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
+(EmbodiedIQLFSDPPolicy pid=526529) 
  0%|          | 0/999322 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
  8%|▊         | 77603/999322 [00:00<00:01, 775228.58it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 16%|█▌        | 158464/999322 [00:00<00:01, 794837.86it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 24%|██▍       | 239079/999322 [00:00<00:00, 799979.67it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 32%|███▏      | 319852/999322 [00:00<00:00, 803028.63it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 40%|████      | 401081/999322 [00:00<00:00, 806359.93it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 48%|████▊     | 481718/999322 [00:00<00:00, 803546.65it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 56%|█████▌    | 562075/999322 [00:00<00:00, 802810.42it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 64%|██████▍   | 642950/999322 [00:00<00:00, 804688.10it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 72%|███████▏  | 723518/999322 [00:00<00:00, 804987.75it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 80%|████████  | 804263/999322 [00:01<00:00, 805740.19it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 89%|████████▊ | 884869/999322 [00:01<00:00, 805832.94it/s]
+(EmbodiedIQLFSDPPolicy pid=526529) 
 97%|█████████▋| 965453/999322 [00:01<00:00, 804507.40it/s]
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-15:55:04-iql_walker2d-medium-v2",
+      "experiment_name": "iql_walker2d-medium-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-15:55:04-iql_walker2d-medium-v2",
+  "env_name": "walker2d-medium-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-15:55:04-iql_walker2d-medium-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-15:55:04-iql_walker2d-medium-v2
+    experiment_name: iql_walker2d-medium-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-15:55:04-iql_walker2d-medium-v2
+env_name: walker2d-medium-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-15:55:04-iql_walker2d-medium-v2/tensorboard/events.out.tfevents.1772438120.dps799-Legion-Y9000P-IRX9.520610.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:639a6c568cbf776aa7f457d28d79d7875ab4e8cb4f57d2202269b28bdd1c1bd2
+size 467877892

20260302-17:48:01-iql_halfcheetah-medium-replay-v2/42.txt

@@ -0,0 +1,200 @@
+5000 10.6
+10000 9.0
+15000 9.9
+20000 20.1
+25000 20.7
+30000 27.0
+35000 35.8
+40000 40.8
+45000 25.3
+50000 42.5
+55000 38.8
+60000 42.6
+65000 43.0
+70000 43.3
+75000 40.3
+80000 41.8
+85000 39.1
+90000 44.7
+95000 44.4
+100000 44.7
+105000 44.6
+110000 42.7
+115000 41.0
+120000 45.2
+125000 45.6
+130000 44.7
+135000 42.6
+140000 44.2
+145000 43.3
+150000 38.3
+155000 44.6
+160000 43.4
+165000 45.1
+170000 44.7
+175000 44.4
+180000 40.9
+185000 44.7
+190000 42.8
+195000 44.2
+200000 44.1
+205000 44.9
+210000 44.8
+215000 41.3
+220000 44.7
+225000 44.6
+230000 44.4
+235000 43.4
+240000 43.2
+245000 44.7
+250000 43.5
+255000 44.4
+260000 44.5
+265000 44.4
+270000 40.8
+275000 44.7
+280000 45.4
+285000 43.7
+290000 42.2
+295000 41.8
+300000 43.9
+305000 44.1
+310000 44.5
+315000 45.3
+320000 44.8
+325000 44.4
+330000 45.0
+335000 44.7
+340000 44.3
+345000 44.3
+350000 42.2
+355000 44.7
+360000 44.4
+365000 44.0
+370000 44.7
+375000 44.2
+380000 44.0
+385000 44.9
+390000 41.0
+395000 44.6
+400000 45.0
+405000 44.6
+410000 44.4
+415000 45.1
+420000 45.1
+425000 43.7
+430000 41.4
+435000 43.9
+440000 42.6
+445000 44.3
+450000 43.8
+455000 44.5
+460000 44.4
+465000 43.9
+470000 44.0
+475000 44.5
+480000 44.7
+485000 44.4
+490000 42.5
+495000 44.1
+500000 44.2
+505000 44.2
+510000 44.7
+515000 44.3
+520000 45.1
+525000 44.5
+530000 44.9
+535000 44.7
+540000 45.0
+545000 44.5
+550000 43.8
+555000 45.1
+560000 43.8
+565000 39.2
+570000 44.4
+575000 44.5
+580000 44.2
+585000 44.2
+590000 44.3
+595000 42.2
+600000 43.7
+605000 43.8
+610000 44.1
+615000 45.0
+620000 44.6
+625000 43.8
+630000 44.1
+635000 39.4
+640000 42.9
+645000 44.0
+650000 43.3
+655000 44.1
+660000 43.7
+665000 43.5
+670000 43.6
+675000 44.9
+680000 44.5
+685000 39.5
+690000 43.7
+695000 44.1
+700000 43.1
+705000 43.7
+710000 44.4
+715000 43.5
+720000 43.7
+725000 39.3
+730000 40.2
+735000 44.3
+740000 44.8
+745000 44.4
+750000 44.1
+755000 43.5
+760000 39.5
+765000 44.5
+770000 39.2
+775000 43.1
+780000 43.9
+785000 44.5
+790000 40.8
+795000 40.7
+800000 42.3
+805000 44.1
+810000 43.3
+815000 44.4
+820000 44.0
+825000 44.1
+830000 44.1
+835000 38.9
+840000 44.3
+845000 44.1
+850000 43.7
+855000 43.2
+860000 43.6
+865000 42.1
+870000 42.0
+875000 40.7
+880000 42.7
+885000 43.9
+890000 43.2
+895000 44.3
+900000 43.2
+905000 43.2
+910000 44.7
+915000 37.9
+920000 44.3
+925000 43.2
+930000 43.3
+935000 43.9
+940000 44.2
+945000 41.3
+950000 43.1
+955000 39.5
+960000 44.8
+965000 42.5
+970000 43.9
+975000 43.7
+980000 40.7
+985000 43.0
+990000 44.0
+995000 43.7
+1000000 40.3

20260302-17:48:01-iql_halfcheetah-medium-replay-v2/run_iql.log

@@ -0,0 +1,130 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:01-iql_halfcheetah-medium-replay-v2 runner.logger.experiment_name=iql_halfcheetah-medium-replay-v2 env_name=halfcheetah-medium-replay-v2
+2026-03-02 17:48:02.939420: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 17:48:02.959815: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 17:48:02.959836: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 17:48:02.960402: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 17:48:02.963932: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 17:48:04.416865: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 17:48:09,603	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8265 
+[INFO 17:48:11 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: eead09c4d75002e1e4d721fb00c6387bda49861e55bd29c96777d06d
+node_ip: 10.193.248.212
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 17:48:11 FlexiblePlacementStrategy] 
+[INFO 17:48:11 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 17:48:11 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=158125) 2026-03-02 17:48:13.128323: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=158125) 2026-03-02 17:48:13.150597: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=158125) 2026-03-02 17:48:13.150626: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=158125) 2026-03-02 17:48:13.151219: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=158125) 2026-03-02 17:48:13.154905: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=158125) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=158125) 2026-03-02 17:48:14.726313: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=158125) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=158125) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=158125) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=158125) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=158125)   from distutils.dep_util import newer, newer_group
+(pid=158125) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=158125) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=158125) No module named 'flow'
+(pid=158125) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=158125) No module named 'carla'
+(pid=158125) pybullet build time: Jan 29 2025 23:17:20
+(pid=158125) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=158125)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=158125) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=158125)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=158125) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=158125)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=158125) 
+(EmbodiedIQLFSDPPolicy pid=158125) 
+(EmbodiedIQLFSDPPolicy pid=158125) 
+(EmbodiedIQLFSDPPolicy pid=158125) 
  0%|          | 0/201798 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=158125) 
 41%|████      | 81980/201798 [00:00<00:00, 819765.60it/s]
+(EmbodiedIQLFSDPPolicy pid=158125) 
 82%|████████▏ | 166010/201798 [00:00<00:00, 831826.25it/s]
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:01-iql_halfcheetah-medium-replay-v2",
+      "experiment_name": "iql_halfcheetah-medium-replay-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:01-iql_halfcheetah-medium-replay-v2",
+  "env_name": "halfcheetah-medium-replay-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-17:48:01-iql_halfcheetah-medium-replay-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:01-iql_halfcheetah-medium-replay-v2
+    experiment_name: iql_halfcheetah-medium-replay-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:01-iql_halfcheetah-medium-replay-v2
+env_name: halfcheetah-medium-replay-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-17:48:01-iql_halfcheetah-medium-replay-v2/tensorboard/events.out.tfevents.1772444891.dps799-Legion-Y9000P-IRX9.155087.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b642f7b2d60305712a77a340f42c680dcf0a799a22958c57d0e93f614103555
+size 467877892

20260302-17:48:05-iql_hopper-medium-replay-v2/42.txt

@@ -0,0 +1,200 @@
+5000 10.9
+10000 13.9
+15000 17.1
+20000 29.3
+25000 20.5
+30000 38.1
+35000 54.6
+40000 32.3
+45000 47.1
+50000 55.0
+55000 30.2
+60000 93.1
+65000 84.6
+70000 91.9
+75000 85.7
+80000 49.4
+85000 58.8
+90000 85.3
+95000 63.9
+100000 90.3
+105000 24.3
+110000 29.0
+115000 78.3
+120000 60.2
+125000 47.8
+130000 98.2
+135000 47.3
+140000 88.4
+145000 84.9
+150000 79.8
+155000 75.3
+160000 96.8
+165000 52.8
+170000 60.2
+175000 90.3
+180000 57.5
+185000 46.4
+190000 95.7
+195000 59.3
+200000 60.4
+205000 91.8
+210000 59.4
+215000 61.9
+220000 65.8
+225000 50.2
+230000 46.2
+235000 49.4
+240000 45.9
+245000 69.7
+250000 53.9
+255000 52.9
+260000 55.0
+265000 46.8
+270000 74.5
+275000 57.8
+280000 61.0
+285000 54.2
+290000 57.2
+295000 51.4
+300000 60.5
+305000 69.8
+310000 78.9
+315000 48.3
+320000 56.9
+325000 69.7
+330000 73.2
+335000 46.7
+340000 98.5
+345000 65.1
+350000 58.4
+355000 47.7
+360000 45.5
+365000 70.2
+370000 52.9
+375000 59.4
+380000 54.3
+385000 52.8
+390000 50.7
+395000 70.9
+400000 86.4
+405000 64.8
+410000 46.5
+415000 65.5
+420000 73.1
+425000 62.4
+430000 40.7
+435000 63.8
+440000 60.2
+445000 58.2
+450000 48.7
+455000 90.0
+460000 47.2
+465000 56.4
+470000 54.8
+475000 99.7
+480000 46.2
+485000 82.8
+490000 60.9
+495000 77.5
+500000 74.3
+505000 54.0
+510000 46.8
+515000 49.1
+520000 73.9
+525000 47.7
+530000 56.6
+535000 60.0
+540000 67.7
+545000 82.6
+550000 68.2
+555000 63.6
+560000 61.2
+565000 41.7
+570000 52.1
+575000 59.1
+580000 63.6
+585000 73.6
+590000 61.7
+595000 66.5
+600000 55.8
+605000 53.0
+610000 52.0
+615000 77.3
+620000 61.6
+625000 59.4
+630000 66.8
+635000 55.8
+640000 90.1
+645000 57.9
+650000 67.1
+655000 64.4
+660000 57.6
+665000 56.4
+670000 64.7
+675000 70.5
+680000 76.3
+685000 74.1
+690000 68.5
+695000 63.3
+700000 55.3
+705000 63.7
+710000 55.1
+715000 68.7
+720000 64.0
+725000 62.6
+730000 67.4
+735000 54.2
+740000 55.6
+745000 66.5
+750000 61.3
+755000 64.8
+760000 63.8
+765000 67.4
+770000 55.5
+775000 62.6
+780000 62.6
+785000 58.7
+790000 74.0
+795000 78.5
+800000 61.1
+805000 63.4
+810000 61.9
+815000 70.8
+820000 73.5
+825000 78.6
+830000 64.9
+835000 68.9
+840000 69.1
+845000 77.5
+850000 68.9
+855000 77.1
+860000 62.5
+865000 68.6
+870000 69.5
+875000 63.0
+880000 80.6
+885000 74.8
+890000 70.0
+895000 70.5
+900000 69.7
+905000 61.8
+910000 64.8
+915000 68.4
+920000 70.2
+925000 71.5
+930000 61.3
+935000 75.4
+940000 60.8
+945000 68.5
+950000 55.5
+955000 72.9
+960000 71.4
+965000 65.6
+970000 68.7
+975000 78.5
+980000 76.4
+985000 62.0
+990000 66.6
+995000 65.3
+1000000 72.1

20260302-17:48:05-iql_hopper-medium-replay-v2/run_iql.log

@@ -0,0 +1,138 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:05-iql_hopper-medium-replay-v2 runner.logger.experiment_name=iql_hopper-medium-replay-v2 env_name=hopper-medium-replay-v2
+2026-03-02 17:48:07.431914: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 17:48:07.453188: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 17:48:07.453210: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 17:48:07.453823: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 17:48:07.457754: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 17:48:09.001081: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 17:48:11,942	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 10.193.248.212:38499...
+2026-03-02 17:48:11,968	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 17:48:11 RLinf] Ray namespace conflict detected. Retrying to initialize Cluster with a new namespace (attempt 1).
+2026-03-02 17:48:12,023	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 10.193.248.212:38499...
+2026-03-02 17:48:12,045	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 17:48:12 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: eead09c4d75002e1e4d721fb00c6387bda49861e55bd29c96777d06d
+node_ip: 10.193.248.212
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 17:48:12 FlexiblePlacementStrategy] 
+[INFO 17:48:12 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 17:48:12 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=158413) 2026-03-02 17:48:14.765819: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=158413) 2026-03-02 17:48:14.786086: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=158413) 2026-03-02 17:48:14.786111: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=158413) 2026-03-02 17:48:14.786679: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=158413) 2026-03-02 17:48:14.790165: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=158413) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=158413) 2026-03-02 17:48:16.262883: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=158413) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=158413) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=158413) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=158413) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=158413)   from distutils.dep_util import newer, newer_group
+(pid=158413) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=158413) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=158413) No module named 'flow'
+(pid=158413) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=158413) No module named 'carla'
+(pid=158413) pybullet build time: Jan 29 2025 23:17:20
+(pid=158413) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=158413)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=158413) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=158413)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=158413) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=158413)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=158413) 
+(EmbodiedIQLFSDPPolicy pid=158413) 
+(EmbodiedIQLFSDPPolicy pid=158413) 
+(EmbodiedIQLFSDPPolicy pid=158413) 
+(EmbodiedIQLFSDPPolicy pid=158413) 
  0%|          | 0/401598 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=158413) 
 19%|█▊        | 74764/401598 [00:00<00:00, 747610.41it/s]
+(EmbodiedIQLFSDPPolicy pid=158413) 
 38%|███▊      | 153169/401598 [00:00<00:00, 769028.29it/s]
+(EmbodiedIQLFSDPPolicy pid=158413) 
 58%|█████▊    | 231238/401598 [00:00<00:00, 774345.07it/s]
+(EmbodiedIQLFSDPPolicy pid=158413) 
 77%|███████▋  | 309246/401598 [00:00<00:00, 776600.63it/s]
+(EmbodiedIQLFSDPPolicy pid=158413) 
 96%|█████████▋| 387446/401598 [00:00<00:00, 778539.88it/s]
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:05-iql_hopper-medium-replay-v2",
+      "experiment_name": "iql_hopper-medium-replay-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:05-iql_hopper-medium-replay-v2",
+  "env_name": "hopper-medium-replay-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-17:48:05-iql_hopper-medium-replay-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:05-iql_hopper-medium-replay-v2
+    experiment_name: iql_hopper-medium-replay-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-17:48:05-iql_hopper-medium-replay-v2
+env_name: hopper-medium-replay-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-17:48:05-iql_hopper-medium-replay-v2/tensorboard/events.out.tfevents.1772444893.dps799-Legion-Y9000P-IRX9.155216.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b9d76674f7964fa892a10f71dc6a2b7483abb8491b013205537c7a9c05c900d
+size 467877892

20260302-20:44:57-iql_halfcheetah-medium-expert-v2/42.txt

@@ -0,0 +1,200 @@
+5000 32.4
+10000 20.6
+15000 19.9
+20000 14.3
+25000 26.0
+30000 14.0
+35000 22.5
+40000 23.8
+45000 28.2
+50000 16.2
+55000 24.2
+60000 36.8
+65000 26.7
+70000 34.2
+75000 35.3
+80000 36.3
+85000 45.7
+90000 36.6
+95000 59.8
+100000 39.2
+105000 40.4
+110000 41.7
+115000 36.6
+120000 52.7
+125000 38.1
+130000 45.4
+135000 35.0
+140000 44.6
+145000 63.8
+150000 57.0
+155000 56.0
+160000 41.6
+165000 48.7
+170000 82.7
+175000 60.1
+180000 68.5
+185000 49.9
+190000 50.4
+195000 69.8
+200000 66.3
+205000 64.4
+210000 76.4
+215000 73.5
+220000 76.0
+225000 67.7
+230000 65.5
+235000 70.0
+240000 84.0
+245000 79.6
+250000 85.9
+255000 90.5
+260000 79.1
+265000 80.4
+270000 88.0
+275000 75.8
+280000 83.8
+285000 77.6
+290000 87.3
+295000 58.9
+300000 77.4
+305000 85.5
+310000 75.1
+315000 87.3
+320000 83.9
+325000 90.7
+330000 81.8
+335000 89.9
+340000 87.0
+345000 90.4
+350000 87.6
+355000 76.2
+360000 76.6
+365000 88.8
+370000 91.5
+375000 89.2
+380000 90.8
+385000 85.4
+390000 89.9
+395000 87.1
+400000 91.9
+405000 90.3
+410000 91.3
+415000 90.2
+420000 86.2
+425000 89.3
+430000 89.5
+435000 91.7
+440000 90.8
+445000 88.1
+450000 82.5
+455000 89.1
+460000 89.1
+465000 93.5
+470000 90.2
+475000 91.8
+480000 86.7
+485000 92.5
+490000 91.7
+495000 87.4
+500000 91.4
+505000 92.8
+510000 91.9
+515000 92.7
+520000 86.8
+525000 93.1
+530000 92.9
+535000 92.9
+540000 94.0
+545000 90.6
+550000 92.3
+555000 91.7
+560000 92.2
+565000 80.0
+570000 89.5
+575000 93.0
+580000 92.7
+585000 93.3
+590000 89.7
+595000 92.8
+600000 93.4
+605000 93.5
+610000 92.5
+615000 91.3
+620000 92.5
+625000 92.2
+630000 92.2
+635000 93.5
+640000 93.2
+645000 93.4
+650000 91.0
+655000 92.5
+660000 92.7
+665000 93.7
+670000 93.4
+675000 94.2
+680000 92.9
+685000 94.7
+690000 92.5
+695000 92.9
+700000 93.5
+705000 93.4
+710000 93.5
+715000 92.7
+720000 93.1
+725000 93.7
+730000 86.9
+735000 92.7
+740000 92.9
+745000 92.3
+750000 94.4
+755000 94.2
+760000 93.9
+765000 93.6
+770000 95.0
+775000 93.1
+780000 94.1
+785000 94.2
+790000 88.6
+795000 92.6
+800000 94.0
+805000 93.8
+810000 95.2
+815000 93.3
+820000 93.3
+825000 94.3
+830000 93.4
+835000 94.3
+840000 94.1
+845000 93.6
+850000 94.2
+855000 93.6
+860000 93.3
+865000 94.8
+870000 93.9
+875000 93.4
+880000 93.9
+885000 94.3
+890000 93.6
+895000 93.9
+900000 93.9
+905000 93.4
+910000 94.4
+915000 94.1
+920000 94.0
+925000 94.3
+930000 93.6
+935000 94.0
+940000 94.0
+945000 94.0
+950000 94.3
+955000 94.0
+960000 94.0
+965000 94.0
+970000 94.0
+975000 93.6
+980000 93.7
+985000 94.2
+990000 94.3
+995000 93.9
+1000000 94.6

20260302-20:44:57-iql_halfcheetah-medium-expert-v2/run_iql.log

@@ -0,0 +1,155 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-20:44:57-iql_halfcheetah-medium-expert-v2 runner.logger.experiment_name=iql_halfcheetah-medium-expert-v2 env_name=halfcheetah-medium-expert-v2
+2026-03-02 20:44:58.690183: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 20:44:58.709123: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 20:44:58.709142: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 20:44:58.709671: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 20:44:58.712968: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 20:45:00.001999: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+[2026-03-02 20:45:02,411 W 2371324 2371324] network_util.cc:296: Failed to determine local IP via external connectivity to: 8.8.8.8:53, [2001:4860:4860::8888]:53, falling back to hostname resolution
+2026-03-02 20:45:04,806	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:06 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: e1bbec51f33da5fa4759e1e1942cf8a693863b7fbfb78d1adb590ba0
+node_ip: 127.0.1.1
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 20:45:06 FlexiblePlacementStrategy] 
+[INFO 20:45:06 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 20:45:06 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=2374613) 2026-03-02 20:45:07.644910: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=2374613) 2026-03-02 20:45:07.663693: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=2374613) 2026-03-02 20:45:07.663715: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=2374613) 2026-03-02 20:45:07.664209: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=2374613) 2026-03-02 20:45:07.667342: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=2374613) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=2374613) 2026-03-02 20:45:09.005828: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=2374613) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=2374613) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=2374613) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=2374613) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=2374613)   from distutils.dep_util import newer, newer_group
+(pid=2374613) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=2374613) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2374613) No module named 'flow'
+(pid=2374613) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2374613) No module named 'carla'
+(pid=2374613) pybullet build time: Jan 29 2025 23:17:20
+(pid=2374613) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=2374613)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=2374613) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=2374613)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=2374613) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=2374613)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+(EmbodiedIQLFSDPPolicy pid=2374613) 
  0%|          | 0/1998000 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
  4%|▍         | 89407/1998000 [00:00<00:02, 894030.35it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
  9%|▉         | 181459/1998000 [00:00<00:01, 909593.59it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 14%|█▎        | 273480/1998000 [00:00<00:01, 914422.65it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 18%|█▊        | 365403/1998000 [00:00<00:01, 916314.80it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 23%|██▎       | 459039/1998000 [00:00<00:01, 923534.68it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 28%|██▊       | 552311/1998000 [00:00<00:01, 926650.43it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 32%|███▏      | 644977/1998000 [00:00<00:01, 926561.40it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 37%|███▋      | 737634/1998000 [00:00<00:01, 923821.58it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 42%|████▏     | 830710/1998000 [00:00<00:01, 925979.49it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 46%|████▌     | 923802/1998000 [00:01<00:01, 927498.69it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 51%|█████     | 1016554/1998000 [00:01<00:01, 925953.05it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 56%|█████▌    | 1109151/1998000 [00:01<00:00, 922012.98it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 60%|██████    | 1201357/1998000 [00:01<00:00, 919790.29it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 65%|██████▍   | 1293554/1998000 [00:01<00:00, 920437.94it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 69%|██████▉   | 1385921/1998000 [00:01<00:00, 921402.65it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 74%|███████▍  | 1478955/1998000 [00:01<00:00, 924078.27it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 79%|███████▊  | 1571764/1998000 [00:01<00:00, 925276.95it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 83%|████████▎ | 1664364/1998000 [00:01<00:00, 925485.63it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 88%|████████▊ | 1756914/1998000 [00:01<00:00, 919637.30it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 93%|█████████▎| 1848887/1998000 [00:02<00:00, 917671.73it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
 97%|█████████▋| 1940661/1998000 [00:02<00:00, 915549.48it/s]
+(EmbodiedIQLFSDPPolicy pid=2374613) 
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:44:57-iql_halfcheetah-medium-expert-v2",
+      "experiment_name": "iql_halfcheetah-medium-expert-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:44:57-iql_halfcheetah-medium-expert-v2",
+  "env_name": "halfcheetah-medium-expert-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-20:44:57-iql_halfcheetah-medium-expert-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:44:57-iql_halfcheetah-medium-expert-v2
+    experiment_name: iql_halfcheetah-medium-expert-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:44:57-iql_halfcheetah-medium-expert-v2
+env_name: halfcheetah-medium-expert-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-20:44:57-iql_halfcheetah-medium-expert-v2/tensorboard/events.out.tfevents.1772455506.dps799-Legion-Y9000P-IRX9.2371324.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0db9ea85958589fb8d6820ce26620fe9231581815f041ff43abebeecd0a171e
+size 467877892

20260302-20:45:24-iql_hopper-medium-expert-v2/42.txt

@@ -0,0 +1,200 @@
+5000 17.8
+10000 7.8
+15000 7.2
+20000 23.5
+25000 17.4
+30000 28.9
+35000 42.6
+40000 32.5
+45000 21.5
+50000 15.2
+55000 97.6
+60000 35.2
+65000 14.8
+70000 106.0
+75000 17.8
+80000 89.0
+85000 24.9
+90000 15.8
+95000 81.5
+100000 39.4
+105000 63.3
+110000 12.7
+115000 33.6
+120000 19.6
+125000 2.0
+130000 33.4
+135000 17.6
+140000 45.5
+145000 57.4
+150000 49.4
+155000 30.8
+160000 12.5
+165000 110.4
+170000 103.3
+175000 33.1
+180000 66.5
+185000 110.7
+190000 92.0
+195000 111.2
+200000 12.4
+205000 69.3
+210000 2.1
+215000 85.2
+220000 38.8
+225000 108.4
+230000 106.3
+235000 103.8
+240000 88.6
+245000 75.2
+250000 67.7
+255000 24.0
+260000 34.5
+265000 107.2
+270000 102.1
+275000 86.6
+280000 25.3
+285000 43.4
+290000 9.9
+295000 79.9
+300000 106.6
+305000 104.7
+310000 102.3
+315000 89.3
+320000 90.2
+325000 29.8
+330000 54.0
+335000 68.8
+340000 57.4
+345000 99.5
+350000 96.0
+355000 44.8
+360000 56.6
+365000 56.9
+370000 103.2
+375000 99.6
+380000 90.6
+385000 111.3
+390000 5.6
+395000 85.7
+400000 58.9
+405000 63.6
+410000 107.6
+415000 112.3
+420000 71.0
+425000 111.0
+430000 34.7
+435000 55.5
+440000 99.1
+445000 81.2
+450000 100.2
+455000 104.6
+460000 23.7
+465000 35.5
+470000 111.2
+475000 81.7
+480000 110.8
+485000 85.8
+490000 91.2
+495000 105.4
+500000 93.6
+505000 109.3
+510000 12.5
+515000 64.0
+520000 46.8
+525000 62.7
+530000 34.0
+535000 46.9
+540000 71.2
+545000 39.6
+550000 111.7
+555000 54.6
+560000 39.0
+565000 83.4
+570000 112.4
+575000 110.8
+580000 102.3
+585000 53.5
+590000 111.1
+595000 62.7
+600000 37.1
+605000 73.5
+610000 111.6
+615000 49.8
+620000 111.8
+625000 66.0
+630000 48.7
+635000 29.7
+640000 100.6
+645000 108.4
+650000 112.4
+655000 108.3
+660000 58.3
+665000 65.2
+670000 112.4
+675000 110.3
+680000 97.7
+685000 78.4
+690000 101.2
+695000 51.1
+700000 99.0
+705000 76.0
+710000 111.8
+715000 92.7
+720000 98.8
+725000 109.7
+730000 87.1
+735000 99.3
+740000 85.6
+745000 81.9
+750000 100.1
+755000 105.9
+760000 96.8
+765000 87.9
+770000 100.4
+775000 76.6
+780000 77.2
+785000 90.9
+790000 79.9
+795000 74.4
+800000 72.1
+805000 92.5
+810000 86.5
+815000 76.4
+820000 106.0
+825000 51.6
+830000 47.3
+835000 101.2
+840000 94.9
+845000 106.7
+850000 101.4
+855000 105.7
+860000 79.1
+865000 88.4
+870000 88.0
+875000 91.6
+880000 92.3
+885000 70.8
+890000 74.6
+895000 88.3
+900000 85.2
+905000 86.6
+910000 84.7
+915000 92.1
+920000 81.4
+925000 94.3
+930000 84.8
+935000 67.2
+940000 84.3
+945000 85.7
+950000 99.5
+955000 84.7
+960000 106.1
+965000 112.3
+970000 100.5
+975000 94.6
+980000 93.3
+985000 87.2
+990000 99.1
+995000 112.6
+1000000 99.0

20260302-20:45:24-iql_hopper-medium-expert-v2/run_iql.log

@@ -0,0 +1,159 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:24-iql_hopper-medium-expert-v2 runner.logger.experiment_name=iql_hopper-medium-expert-v2 env_name=hopper-medium-expert-v2
+2026-03-02 20:45:25.772403: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 20:45:25.791398: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 20:45:25.791417: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 20:45:25.791950: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 20:45:25.795140: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 20:45:27.196244: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 20:45:29,685	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 127.0.1.1:34719...
+2026-03-02 20:45:29,705	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:29 RLinf] Ray namespace conflict detected. Retrying to initialize Cluster with a new namespace (attempt 1).
+2026-03-02 20:45:29,995	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 127.0.1.1:34719...
+2026-03-02 20:45:30,009	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:30 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: e1bbec51f33da5fa4759e1e1942cf8a693863b7fbfb78d1adb590ba0
+node_ip: 127.0.1.1
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 20:45:30 FlexiblePlacementStrategy] 
+[INFO 20:45:30 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 20:45:30 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=2379103) 2026-03-02 20:45:32.444275: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=2379103) 2026-03-02 20:45:32.462748: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=2379103) 2026-03-02 20:45:32.462773: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=2379103) 2026-03-02 20:45:32.463281: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=2379103) 2026-03-02 20:45:32.466456: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=2379103) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=2379103) 2026-03-02 20:45:33.905964: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=2379103) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=2379103) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=2379103) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=2379103) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=2379103)   from distutils.dep_util import newer, newer_group
+(pid=2379103) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=2379103) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2379103) No module named 'flow'
+(pid=2379103) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2379103) No module named 'carla'
+(pid=2379103) pybullet build time: Jan 29 2025 23:17:20
+(pid=2379103) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=2379103)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=2379103) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=2379103)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=2379103) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=2379103)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
+(EmbodiedIQLFSDPPolicy pid=2379103) 
  0%|          | 0/1998966 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
  4%|▍         | 85522/1998966 [00:00<00:02, 855188.19it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
  9%|▊         | 172354/1998966 [00:00<00:02, 862892.58it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 13%|█▎        | 258644/1998966 [00:00<00:02, 765757.35it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 17%|█▋        | 336368/1998966 [00:00<00:02, 723623.88it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 21%|██        | 419657/1998966 [00:00<00:02, 759662.01it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 25%|██▍       | 496470/1998966 [00:00<00:02, 729426.84it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 29%|██▊       | 570055/1998966 [00:00<00:02, 710746.42it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 32%|███▏      | 641546/1998966 [00:00<00:02, 664602.97it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 36%|███▌      | 718897/1998966 [00:00<00:01, 695501.10it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 40%|████      | 802689/1998966 [00:01<00:01, 736578.35it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 44%|████▍     | 887760/1998966 [00:01<00:01, 769873.89it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 49%|████▊     | 970519/1998966 [00:01<00:01, 786851.76it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 53%|█████▎    | 1051855/1998966 [00:01<00:01, 794691.04it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 57%|█████▋    | 1136363/1998966 [00:01<00:01, 809658.11it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 61%|██████    | 1221086/1998966 [00:01<00:00, 820848.59it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 65%|██████▌   | 1304446/1998966 [00:01<00:00, 824648.04it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 69%|██████▉   | 1388310/1998966 [00:01<00:00, 828824.91it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 74%|███████▎  | 1472368/1998966 [00:01<00:00, 832336.96it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 78%|███████▊  | 1558499/1998966 [00:01<00:00, 841009.53it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 82%|████████▏ | 1642661/1998966 [00:02<00:00, 839944.52it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 86%|████████▋ | 1726698/1998966 [00:02<00:00, 837032.83it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 91%|█████████ | 1810433/1998966 [00:02<00:00, 831543.61it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 95%|█████████▍| 1893616/1998966 [00:02<00:00, 830210.92it/s]
+(EmbodiedIQLFSDPPolicy pid=2379103) 
 99%|█████████▉| 1976657/1998966 [00:02<00:00, 830176.69it/s]
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:24-iql_hopper-medium-expert-v2",
+      "experiment_name": "iql_hopper-medium-expert-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:24-iql_hopper-medium-expert-v2",
+  "env_name": "hopper-medium-expert-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-20:45:24-iql_hopper-medium-expert-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:24-iql_hopper-medium-expert-v2
+    experiment_name: iql_hopper-medium-expert-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:24-iql_hopper-medium-expert-v2
+env_name: hopper-medium-expert-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-20:45:24-iql_hopper-medium-expert-v2/tensorboard/events.out.tfevents.1772455530.dps799-Legion-Y9000P-IRX9.2375845.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1262d59a476211508c6fe9ee86bb16d53a6af657b88e2f4fcde52b343a399231
+size 467877892

20260302-20:45:33-iql_walker2d-medium-expert-v2/42.txt

@@ -0,0 +1,197 @@
+5000 70.3
+10000 86.8
+15000 78.7
+20000 61.7
+25000 85.5
+30000 79.6
+35000 94.3
+40000 97.3
+45000 51.6
+50000 106.3
+55000 85.7
+60000 71.2
+65000 55.7
+70000 69.9
+75000 95.3
+80000 88.3
+85000 108.9
+90000 94.9
+95000 102.7
+100000 108.2
+105000 104.4
+110000 103.0
+115000 108.4
+120000 109.7
+125000 104.8
+130000 110.4
+135000 103.7
+140000 109.2
+145000 108.3
+150000 109.4
+155000 108.9
+160000 109.6
+165000 108.4
+170000 105.8
+175000 105.8
+180000 112.3
+185000 103.4
+190000 108.7
+195000 107.5
+200000 105.1
+205000 104.3
+210000 104.0
+215000 105.6
+220000 105.6
+225000 110.0
+230000 106.0
+235000 109.3
+240000 108.3
+245000 108.7
+250000 107.6
+255000 107.8
+260000 110.9
+265000 108.5
+270000 109.8
+275000 108.6
+280000 103.2
+285000 110.2
+290000 104.1
+295000 110.9
+300000 111.0
+305000 102.0
+310000 104.6
+315000 97.1
+320000 102.6
+325000 104.5
+330000 109.2
+335000 96.4
+340000 107.6
+345000 110.4
+350000 111.2
+355000 104.8
+360000 106.1
+365000 99.2
+370000 111.5
+375000 110.7
+380000 94.1
+385000 110.4
+390000 106.7
+395000 110.6
+400000 109.7
+405000 110.4
+410000 111.7
+415000 109.4
+420000 109.7
+425000 109.6
+430000 109.4
+435000 111.1
+440000 110.2
+445000 108.9
+450000 110.5
+455000 99.4
+460000 110.9
+465000 107.2
+470000 111.1
+475000 110.8
+480000 111.0
+485000 111.9
+490000 108.6
+495000 111.0
+500000 110.0
+505000 110.1
+510000 110.2
+515000 104.1
+520000 100.5
+525000 107.8
+530000 109.5
+535000 109.9
+540000 105.2
+545000 109.9
+550000 110.2
+555000 111.0
+560000 106.7
+565000 110.8
+570000 111.4
+575000 108.9
+580000 110.8
+585000 109.6
+590000 110.5
+595000 104.3
+600000 110.0
+605000 110.2
+610000 110.5
+615000 110.8
+620000 110.7
+625000 99.9
+630000 110.2
+635000 109.2
+640000 110.4
+645000 110.3
+650000 110.6
+655000 110.5
+660000 110.1
+665000 110.1
+670000 110.3
+675000 110.5
+680000 110.0
+685000 110.8
+690000 110.4
+695000 110.2
+700000 110.9
+705000 109.8
+710000 110.4
+715000 110.5
+720000 110.4
+725000 109.8
+730000 110.2
+735000 109.6
+740000 110.0
+745000 110.1
+750000 110.1
+755000 110.8
+760000 110.6
+765000 110.5
+770000 110.2
+775000 110.3
+780000 110.3
+785000 110.3
+790000 109.7
+795000 110.2
+800000 109.9
+805000 109.6
+810000 110.0
+815000 110.1
+820000 109.5
+825000 109.6
+830000 109.9
+835000 110.3
+840000 110.4
+845000 110.4
+850000 109.6
+855000 110.4
+860000 110.0
+865000 110.5
+870000 110.2
+875000 110.1
+880000 109.9
+885000 110.5
+890000 110.0
+895000 110.3
+900000 109.8
+905000 110.2
+910000 110.0
+915000 110.0
+920000 109.8
+925000 109.8
+930000 110.4
+935000 109.7
+940000 110.5
+945000 110.3
+950000 110.3
+955000 110.4
+960000 110.1
+965000 110.6
+970000 110.2
+975000 110.2
+980000 110.2
+985000 110.7

20260302-20:45:33-iql_walker2d-medium-expert-v2/run_iql.log

@@ -0,0 +1,171 @@
+/home/dps799/Desktop/RLinf/.venv/bin/python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_iql.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:33-iql_walker2d-medium-expert-v2 runner.logger.experiment_name=iql_walker2d-medium-expert-v2 env_name=walker2d-medium-expert-v2
+2026-03-02 20:45:34.481957: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 20:45:34.501270: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 20:45:34.501290: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 20:45:34.501810: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 20:45:34.505206: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 20:45:35.925748: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 20:45:38,577	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 127.0.1.1:34719...
+2026-03-02 20:45:38,615	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:38 RLinf] Ray namespace conflict detected. Retrying to initialize Cluster with a new namespace (attempt 1).
+2026-03-02 20:45:38,911	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 127.0.1.1:34719...
+2026-03-02 20:45:38,928	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:38 RLinf] Ray namespace conflict detected. Retrying to initialize Cluster with a new namespace (attempt 2).
+2026-03-02 20:45:39,210	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 127.0.1.1:34719...
+2026-03-02 20:45:39,225	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 20:45:40 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: e1bbec51f33da5fa4759e1e1942cf8a693863b7fbfb78d1adb590ba0
+node_ip: 127.0.1.1
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 20:45:40 FlexiblePlacementStrategy] 
+[INFO 20:45:40 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 20:45:40 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=2383384) 2026-03-02 20:45:41.958301: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=2383384) 2026-03-02 20:45:41.978448: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=2383384) 2026-03-02 20:45:41.978476: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=2383384) 2026-03-02 20:45:41.979009: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=2383384) 2026-03-02 20:45:41.982450: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=2383384) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=2383384) 2026-03-02 20:45:43.503661: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=2383384) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=2383384) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=2383384) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=2383384) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=2383384)   from distutils.dep_util import newer, newer_group
+(pid=2383384) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=2383384) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2383384) No module named 'flow'
+(pid=2383384) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=2383384) No module named 'carla'
+(pid=2383384) pybullet build time: Jan 29 2025 23:17:20
+(pid=2383384) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=2383384)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=2383384) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=2383384)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=2383384) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=2383384)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
+(EmbodiedIQLFSDPPolicy pid=2383384) 
  0%|          | 0/1998318 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
  4%|▍         | 84613/1998318 [00:00<00:02, 846100.55it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
  9%|▊         | 170557/1998318 [00:00<00:02, 853929.69it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 13%|█▎        | 256254/1998318 [00:00<00:02, 855299.68it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 17%|█▋        | 342660/1998318 [00:00<00:01, 858746.27it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 21%|██▏       | 428535/1998318 [00:00<00:01, 856669.49it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 26%|██▌       | 515104/1998318 [00:00<00:01, 859725.53it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 30%|███       | 601078/1998318 [00:00<00:01, 857659.10it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 34%|███▍      | 686846/1998318 [00:00<00:01, 857488.74it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 39%|███▊      | 772596/1998318 [00:00<00:01, 844892.18it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 43%|████▎     | 858383/1998318 [00:01<00:01, 848846.97it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 47%|████▋     | 943645/1998318 [00:01<00:01, 849986.61it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 52%|█████▏    | 1031015/1998318 [00:01<00:01, 857154.43it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 56%|█████▌    | 1117865/1998318 [00:01<00:01, 860572.71it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 60%|██████    | 1204970/1998318 [00:01<00:00, 863723.42it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 65%|██████▍   | 1291354/1998318 [00:01<00:00, 859340.98it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 69%|██████▉   | 1378347/1998318 [00:01<00:00, 862507.94it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 73%|███████▎  | 1464691/1998318 [00:01<00:00, 862781.95it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 78%|███████▊  | 1551242/1998318 [00:01<00:00, 863595.03it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 82%|████████▏ | 1638225/1998318 [00:01<00:00, 865454.21it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 86%|████████▋ | 1725007/1998318 [00:02<00:00, 866158.34it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 91%|█████████ | 1811626/1998318 [00:02<00:00, 865997.79it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 95%|█████████▍| 1898228/1998318 [00:02<00:00, 863218.59it/s]
+(EmbodiedIQLFSDPPolicy pid=2383384) 
 99%|█████████▉| 1984554/1998318 [00:02<00:00, 827339.42it/s]
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:33-iql_walker2d-medium-expert-v2",
+      "experiment_name": "iql_walker2d-medium-expert-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:33-iql_walker2d-medium-expert-v2",
+  "env_name": "walker2d-medium-expert-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}
+Exception occurred while running EmbodiedIQLFSDPPolicy's function set_global_step: exception is The actor died unexpectedly before finishing this task.
+	class_name: EmbodiedIQLFSDPPolicy
+	actor_id: ecc18de2714c2d45c9dfef9406000000
+	pid: 2383384
+	name: ActorGroup:0
+	namespace: RLinf_2
+	ip: 127.0.1.1
+The actor died because its node has died. Node Id: e1bbec51f33da5fa4759e1e1942cf8a693863b7fbfb78d1adb590ba0
+	the actor's node was terminated expectedly: received SIGTERM
+Exiting main process due to a failure upon worker execution.

20260302-20:45:33-iql_walker2d-medium-expert-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:33-iql_walker2d-medium-expert-v2
+    experiment_name: iql_walker2d-medium-expert-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-20:45:33-iql_walker2d-medium-expert-v2
+env_name: walker2d-medium-expert-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-20:45:33-iql_walker2d-medium-expert-v2/tensorboard/events.out.tfevents.1772455540.dps799-Legion-Y9000P-IRX9.2380089.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0e3d75463cde2f0d0bbc8fcb7ed0e4f7cc8336b279d7cfbee61cffbd8e72e0d
+size 463167076

20260302-23:17:39-iql_walker2d-medium-replay-v2/42.txt

@@ -0,0 +1,200 @@
+5000 11.7
+10000 9.4
+15000 8.2
+20000 7.1
+25000 2.7
+30000 0.9
+35000 2.4
+40000 20.7
+45000 0.9
+50000 5.2
+55000 34.4
+60000 28.4
+65000 6.6
+70000 53.6
+75000 32.6
+80000 59.7
+85000 7.2
+90000 59.6
+95000 76.9
+100000 68.2
+105000 66.1
+110000 73.6
+115000 68.3
+120000 71.4
+125000 49.6
+130000 57.5
+135000 45.6
+140000 53.7
+145000 67.6
+150000 57.5
+155000 76.7
+160000 51.0
+165000 62.0
+170000 71.8
+175000 58.5
+180000 58.1
+185000 43.7
+190000 38.6
+195000 77.0
+200000 55.1
+205000 64.3
+210000 71.5
+215000 56.0
+220000 39.5
+225000 61.6
+230000 61.8
+235000 60.5
+240000 30.9
+245000 73.0
+250000 64.4
+255000 39.8
+260000 46.0
+265000 46.7
+270000 69.4
+275000 63.4
+280000 58.5
+285000 69.6
+290000 75.2
+295000 53.5
+300000 49.5
+305000 73.5
+310000 53.9
+315000 66.1
+320000 59.8
+325000 71.8
+330000 66.6
+335000 60.3
+340000 70.6
+345000 67.8
+350000 73.6
+355000 68.2
+360000 69.5
+365000 66.1
+370000 37.3
+375000 64.3
+380000 71.7
+385000 50.6
+390000 67.8
+395000 73.9
+400000 46.3
+405000 83.9
+410000 35.4
+415000 62.1
+420000 39.5
+425000 39.9
+430000 56.2
+435000 65.9
+440000 55.0
+445000 54.6
+450000 45.5
+455000 69.0
+460000 60.0
+465000 62.6
+470000 71.4
+475000 41.5
+480000 74.7
+485000 69.2
+490000 82.7
+495000 50.7
+500000 58.9
+505000 56.5
+510000 61.2
+515000 62.2
+520000 61.1
+525000 61.1
+530000 50.1
+535000 63.0
+540000 69.2
+545000 74.8
+550000 45.2
+555000 52.1
+560000 62.9
+565000 57.3
+570000 62.4
+575000 78.5
+580000 70.0
+585000 54.5
+590000 61.2
+595000 79.8
+600000 61.1
+605000 63.2
+610000 72.3
+615000 64.4
+620000 66.4
+625000 80.9
+630000 65.3
+635000 61.2
+640000 61.9
+645000 65.2
+650000 46.9
+655000 79.6
+660000 63.8
+665000 65.4
+670000 68.2
+675000 59.7
+680000 64.1
+685000 80.2
+690000 67.3
+695000 54.0
+700000 72.7
+705000 73.2
+710000 54.7
+715000 60.8
+720000 56.5
+725000 54.1
+730000 74.5
+735000 62.1
+740000 69.8
+745000 68.1
+750000 65.6
+755000 76.2
+760000 64.1
+765000 71.8
+770000 69.5
+775000 56.4
+780000 73.0
+785000 69.4
+790000 67.2
+795000 61.3
+800000 67.9
+805000 44.9
+810000 66.1
+815000 53.8
+820000 67.7
+825000 66.1
+830000 70.3
+835000 66.2
+840000 62.2
+845000 52.6
+850000 70.6
+855000 54.9
+860000 60.7
+865000 46.8
+870000 62.3
+875000 73.2
+880000 62.4
+885000 62.3
+890000 63.0
+895000 51.3
+900000 51.5
+905000 67.2
+910000 70.4
+915000 58.6
+920000 57.1
+925000 53.4
+930000 67.3
+935000 62.5
+940000 68.7
+945000 65.9
+950000 64.8
+955000 53.5
+960000 70.3
+965000 70.1
+970000 58.3
+975000 55.5
+980000 61.5
+985000 64.4
+990000 67.4
+995000 61.4
+1000000 57.8

20260302-23:17:39-iql_walker2d-medium-replay-v2/run_d4rl_offline.log

@@ -0,0 +1,133 @@
+python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_d4rl_offline.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-23:17:39-iql_walker2d-medium-replay-v2 runner.logger.experiment_name=iql_walker2d-medium-replay-v2 env_name=walker2d-medium-replay-v2
+2026-03-02 23:17:40.542954: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 23:17:40.561989: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 23:17:40.562008: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 23:17:40.562608: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 23:17:40.565733: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 23:17:41.897618: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 23:17:46,794	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8265 
+[INFO 23:17:48 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: 64918e0a8f811d74a75fd42dbf168b72ef4f85b58a79aafbc9c1fba8
+node_ip: 10.95.166.236
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 23:17:48 FlexiblePlacementStrategy] 
+[INFO 23:17:48 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 23:17:48 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=23957) 2026-03-02 23:17:49.652421: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=23957) 2026-03-02 23:17:49.670788: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=23957) 2026-03-02 23:17:49.670821: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=23957) 2026-03-02 23:17:49.671426: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=23957) 2026-03-02 23:17:49.674534: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=23957) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=23957) 2026-03-02 23:17:51.006930: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=23957) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=23957) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=23957) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=23957) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=23957)   from distutils.dep_util import newer, newer_group
+(pid=23957) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=23957) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=23957) No module named 'flow'
+(pid=23957) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=23957) No module named 'carla'
+(pid=23957) pybullet build time: Jan 29 2025 23:17:20
+(pid=23957) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=23957)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=23957) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=23957)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=23957) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=23957)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=23957) 
+(EmbodiedIQLFSDPPolicy pid=23957) 
+(EmbodiedIQLFSDPPolicy pid=23957) 
+(EmbodiedIQLFSDPPolicy pid=23957) 
+(EmbodiedIQLFSDPPolicy pid=23957) 
  0%|          | 0/301698 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=23957) 
 29%|██▉       | 87492/301698 [00:00<00:00, 874877.03it/s]
+(EmbodiedIQLFSDPPolicy pid=23957) 
 58%|█████▊    | 176304/301698 [00:00<00:00, 882642.59it/s]
+(EmbodiedIQLFSDPPolicy pid=23957) 
 88%|████████▊ | 265899/301698 [00:00<00:00, 888701.85it/s]
+(EmbodiedIQLFSDPPolicy pid=23957) 
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-23:17:39-iql_walker2d-medium-replay-v2",
+      "experiment_name": "iql_walker2d-medium-replay-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-23:17:39-iql_walker2d-medium-replay-v2",
+  "env_name": "walker2d-medium-replay-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-23:17:39-iql_walker2d-medium-replay-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-23:17:39-iql_walker2d-medium-replay-v2
+    experiment_name: iql_walker2d-medium-replay-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-23:17:39-iql_walker2d-medium-replay-v2
+env_name: walker2d-medium-replay-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-23:17:39-iql_walker2d-medium-replay-v2/tensorboard/events.out.tfevents.1772464668.dps799-Legion-Y9000P-IRX9.21071.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc0cd4f85407d6a3927a86d5e4379672b3e0d7263f839bdf96b8be1c5fa3b4b5
+size 467877892

20260302-23:18:37-iql_hopper-medium-v2/42.txt

@@ -0,0 +1,200 @@
+5000 46.6
+10000 41.7
+15000 54.2
+20000 30.0
+25000 48.4
+30000 44.0
+35000 42.6
+40000 51.1
+45000 39.0
+50000 50.7
+55000 39.5
+60000 43.0
+65000 34.9
+70000 41.6
+75000 40.2
+80000 40.6
+85000 35.6
+90000 36.7
+95000 42.3
+100000 40.6
+105000 38.0
+110000 37.2
+115000 47.7
+120000 57.2
+125000 60.5
+130000 46.0
+135000 48.1
+140000 58.0
+145000 44.3
+150000 47.2
+155000 53.4
+160000 45.6
+165000 54.9
+170000 69.3
+175000 57.4
+180000 64.1
+185000 58.3
+190000 61.4
+195000 62.6
+200000 63.2
+205000 62.2
+210000 74.7
+215000 49.1
+220000 58.9
+225000 62.0
+230000 64.2
+235000 52.4
+240000 53.9
+245000 49.5
+250000 63.6
+255000 51.2
+260000 45.0
+265000 46.8
+270000 66.6
+275000 57.4
+280000 58.2
+285000 49.0
+290000 51.1
+295000 56.1
+300000 53.7
+305000 49.7
+310000 64.1
+315000 56.6
+320000 52.6
+325000 54.2
+330000 56.5
+335000 57.7
+340000 47.1
+345000 63.2
+350000 47.8
+355000 50.8
+360000 54.2
+365000 59.7
+370000 61.6
+375000 69.1
+380000 50.9
+385000 50.1
+390000 54.8
+395000 53.3
+400000 48.2
+405000 46.9
+410000 57.9
+415000 46.5
+420000 61.6
+425000 55.4
+430000 61.2
+435000 60.4
+440000 52.2
+445000 61.9
+450000 59.9
+455000 56.7
+460000 58.3
+465000 63.9
+470000 60.4
+475000 59.4
+480000 50.8
+485000 53.9
+490000 57.0
+495000 61.5
+500000 57.2
+505000 61.6
+510000 54.5
+515000 57.2
+520000 53.9
+525000 60.3
+530000 53.0
+535000 49.1
+540000 53.5
+545000 50.9
+550000 52.5
+555000 59.7
+560000 58.8
+565000 53.0
+570000 62.7
+575000 57.3
+580000 52.8
+585000 56.0
+590000 47.3
+595000 67.7
+600000 48.4
+605000 56.2
+610000 56.1
+615000 65.9
+620000 50.4
+625000 71.6
+630000 62.8
+635000 59.1
+640000 59.9
+645000 55.3
+650000 60.9
+655000 57.9
+660000 71.7
+665000 60.8
+670000 54.2
+675000 53.5
+680000 60.0
+685000 51.5
+690000 59.8
+695000 54.8
+700000 60.6
+705000 56.0
+710000 65.6
+715000 53.6
+720000 57.9
+725000 61.5
+730000 51.6
+735000 61.2
+740000 60.7
+745000 62.0
+750000 61.3
+755000 57.8
+760000 54.6
+765000 67.1
+770000 62.8
+775000 63.4
+780000 56.8
+785000 52.9
+790000 55.1
+795000 58.3
+800000 69.8
+805000 64.5
+810000 60.1
+815000 52.6
+820000 61.3
+825000 65.6
+830000 63.9
+835000 57.2
+840000 66.8
+845000 67.3
+850000 63.0
+855000 52.3
+860000 66.0
+865000 59.1
+870000 58.5
+875000 65.3
+880000 68.1
+885000 58.6
+890000 72.9
+895000 54.4
+900000 54.5
+905000 63.0
+910000 63.4
+915000 67.0
+920000 64.8
+925000 57.9
+930000 71.7
+935000 57.7
+940000 58.1
+945000 67.7
+950000 61.2
+955000 54.9
+960000 63.0
+965000 57.4
+970000 63.3
+975000 52.1
+980000 62.3
+985000 58.6
+990000 68.4
+995000 62.3
+1000000 60.4

20260302-23:18:37-iql_hopper-medium-v2/run_d4rl_offline.log

@@ -0,0 +1,145 @@
+python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_d4rl_offline.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_mujoco_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260302-23:18:37-iql_hopper-medium-v2 runner.logger.experiment_name=iql_hopper-medium-v2 env_name=hopper-medium-v2
+2026-03-02 23:18:39.269717: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-02 23:18:39.288595: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-02 23:18:39.288618: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-02 23:18:39.289198: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-02 23:18:39.292387: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-02 23:18:40.667073: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+2026-03-02 23:18:43,165	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 10.95.166.236:40873...
+2026-03-02 23:18:43,184	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 23:18:43 RLinf] Ray namespace conflict detected. Retrying to initialize Cluster with a new namespace (attempt 1).
+2026-03-02 23:18:43,224	INFO worker.py:1810 -- Connecting to existing Ray cluster at address: 10.95.166.236:40873...
+2026-03-02 23:18:43,238	INFO worker.py:2004 -- Connected to Ray cluster. View the dashboard at http://127.0.0.1:8265 
+[INFO 23:18:44 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: 64918e0a8f811d74a75fd42dbf168b72ef4f85b58a79aafbc9c1fba8
+node_ip: 10.95.166.236
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 23:18:44 FlexiblePlacementStrategy] 
+[INFO 23:18:44 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 23:18:44 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=42431) 2026-03-02 23:18:45.756084: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=42431) 2026-03-02 23:18:45.774978: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=42431) 2026-03-02 23:18:45.775003: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=42431) 2026-03-02 23:18:45.775523: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=42431) 2026-03-02 23:18:45.778704: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=42431) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=42431) 2026-03-02 23:18:47.170493: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=42431) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=42431) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=42431) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=42431) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=42431)   from distutils.dep_util import newer, newer_group
+(pid=42431) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=42431) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=42431) No module named 'flow'
+(pid=42431) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=42431) No module named 'carla'
+(pid=42431) pybullet build time: Jan 29 2025 23:17:20
+(pid=42431) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=42431)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=42431) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=42431)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=42431) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=42431)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=42431) 
+(EmbodiedIQLFSDPPolicy pid=42431) 
+(EmbodiedIQLFSDPPolicy pid=42431) 
+(EmbodiedIQLFSDPPolicy pid=42431) 
+(EmbodiedIQLFSDPPolicy pid=42431) 
+(EmbodiedIQLFSDPPolicy pid=42431) 
  0%|          | 0/999998 [00:00<?, ?it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
  8%|▊         | 84938/999998 [00:00<00:01, 849348.41it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 17%|█▋        | 169873/999998 [00:00<00:01, 806662.59it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 25%|██▌       | 253286/999998 [00:00<00:00, 818841.47it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 34%|███▎      | 335284/999998 [00:00<00:00, 819071.38it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 42%|████▏     | 417261/999998 [00:00<00:00, 812643.57it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 50%|█████     | 500159/999998 [00:00<00:00, 818122.01it/s]
 58%|█████▊    | 583396/999998 [00:00<00:00, 822734.93it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 67%|██████▋   | 667971/999998 [00:00<00:00, 830009.38it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 75%|███████▌  | 754078/999998 [00:00<00:00, 839673.82it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 84%|████████▍ | 839370/999998 [00:01<00:00, 843746.30it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
 92%|█████████▏| 923761/999998 [00:01<00:00, 843248.77it/s]
+(EmbodiedIQLFSDPPolicy pid=42431) 
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260302-23:18:37-iql_hopper-medium-v2",
+      "experiment_name": "iql_hopper-medium-v2"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260302-23:18:37-iql_hopper-medium-v2",
+  "env_name": "hopper-medium-v2",
+  "seed": 42,
+  "eval_episodes": 10,
+  "log_interval": 1000,
+  "eval_interval": 5000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.7,
+    "temperature": 3.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}

20260302-23:18:37-iql_hopper-medium-v2/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260302-23:18:37-iql_hopper-medium-v2
+    experiment_name: iql_hopper-medium-v2
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260302-23:18:37-iql_hopper-medium-v2
+env_name: hopper-medium-v2
+seed: 42
+eval_episodes: 10
+log_interval: 1000
+eval_interval: 5000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.7
+  temperature: 3.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260302-23:18:37-iql_hopper-medium-v2/tensorboard/events.out.tfevents.1772464724.dps799-Legion-Y9000P-IRX9.39583.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a291b2f54dbab165877747c7aca102d6b112343bca5d74a148fb14757f54d0a
+size 467877892

20260303-01:43:05-iql_antmaze-large-diverse-v0/42.txt

@@ -0,0 +1,10 @@
+100000 17.0
+200000 21.0
+300000 45.0
+400000 38.0
+500000 43.0
+600000 43.0
+700000 55.0
+800000 47.0
+900000 50.0
+1000000 47.0

20260303-01:43:05-iql_antmaze-large-diverse-v0/run_d4rl_offline.log

@@ -0,0 +1,138 @@
+python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_d4rl_offline.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_antmaze_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-large-diverse-v0 runner.logger.experiment_name=iql_antmaze-large-diverse-v0 env_name=antmaze-large-diverse-v0
+2026-03-03 01:43:07.073936: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-03 01:43:07.093675: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-03 01:43:07.093702: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-03 01:43:07.094264: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-03 01:43:07.097648: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-03 01:43:08.477278: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+[2026-03-03 01:43:11,227 W 433385 433385] network_util.cc:296: Failed to determine local IP via external connectivity to: 8.8.8.8:53, [2001:4860:4860::8888]:53, falling back to hostname resolution
+2026-03-03 01:43:16,786	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8266 
+[INFO 01:43:18 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: cd63a75e50218d16ebf421239c22fcd7617e0bd0a62af454cee85d8c
+node_ip: 127.0.1.1
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 01:43:18 FlexiblePlacementStrategy] 
+[INFO 01:43:18 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 01:43:18 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=439252) 2026-03-03 01:43:19.971903: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=439252) 2026-03-03 01:43:19.991548: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=439252) 2026-03-03 01:43:19.991571: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=439252) 2026-03-03 01:43:19.992117: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=439252) 2026-03-03 01:43:19.995411: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=439252) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=439252) 2026-03-03 01:43:21.514198: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=439252) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=439252) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=439252) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=439252)   from distutils.dep_util import newer, newer_group
+(pid=439252) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=439252) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=439252) No module named 'flow'
+(pid=439252) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=439252) No module named 'carla'
+(pid=439252) pybullet build time: Jan 29 2025 23:17:20
+(pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=439252)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/envs/registration.py:505: UserWarning: WARN: The environment antmaze-large-diverse-v0 is out of date. You should consider upgrading to version `v2` with the environment ID `antmaze-large-diverse-v2`.
+(EmbodiedIQLFSDPPolicy pid=439252)   logger.warn(
+(EmbodiedIQLFSDPPolicy pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/d4rl/locomotion/ant.py:180: UserWarning: This environment is deprecated. Please use the most recent version of this environment.
+(EmbodiedIQLFSDPPolicy pid=439252)   offline_env.OfflineEnv.__init__(self, **kwargs)
+(EmbodiedIQLFSDPPolicy pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=439252)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=439252)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) 
+(EmbodiedIQLFSDPPolicy pid=439252) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/utils/seeding.py:38: DeprecationWarning: WARN: Function `rng.randn(*size)` is marked as deprecated and will be removed in the future. Please use `rng.standard_normal(size)` instead.
+(EmbodiedIQLFSDPPolicy pid=439252)   deprecation(
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-large-diverse-v0",
+      "experiment_name": "iql_antmaze-large-diverse-v0"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-large-diverse-v0",
+  "env_name": "antmaze-large-diverse-v0",
+  "seed": 42,
+  "eval_episodes": 100,
+  "log_interval": 1000,
+  "eval_interval": 100000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.9,
+    "temperature": 10.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}
+(EmbodiedIQLFSDPPolicy pid=439252) Target Goal:  (33.126732195354386, 24.95509883353427)

20260303-01:43:05-iql_antmaze-large-diverse-v0/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-large-diverse-v0
+    experiment_name: iql_antmaze-large-diverse-v0
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-large-diverse-v0
+env_name: antmaze-large-diverse-v0
+seed: 42
+eval_episodes: 100
+log_interval: 1000
+eval_interval: 100000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.9
+  temperature: 10.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260303-01:43:05-iql_antmaze-large-diverse-v0/tensorboard/events.out.tfevents.1772473398.dps799-Legion-Y9000P-IRX9.433385.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:814406f2c1e85f27a6c9eab86a114e5a7eecd9ee84a5ccc21fa1acf8f1652d81
+size 467852818

20260303-01:43:05-iql_antmaze-medium-play-v0/42.txt

@@ -0,0 +1,10 @@
+100000 49.0
+200000 73.0
+300000 75.0
+400000 78.0
+500000 74.0
+600000 75.0
+700000 71.0
+800000 67.0
+900000 70.0
+1000000 64.0

20260303-01:43:05-iql_antmaze-medium-play-v0/run_d4rl_offline.log

@@ -0,0 +1,137 @@
+python /home/dps799/Desktop/RLinf_merge/examples/embodiment/train_d4rl_offline.py --config-path /home/dps799/Desktop/RLinf_merge/examples/embodiment/config/ --config-name iql_antmaze_offline runner.logger.log_path=/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-medium-play-v0 runner.logger.experiment_name=iql_antmaze-medium-play-v0 env_name=antmaze-medium-play-v0
+2026-03-03 01:43:06.681602: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-03-03 01:43:06.701169: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-03-03 01:43:06.701196: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-03-03 01:43:06.701748: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-03-03 01:43:06.704920: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2026-03-03 01:43:08.101950: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+  from distutils.dep_util import newer, newer_group
+<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'flow'
+Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+No module named 'carla'
+pybullet build time: Jan 29 2025 23:17:20
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+  from pkg_resources import parse_version
+/home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
+See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
+  ret = run_job(
+[2026-03-03 01:43:10,792 W 433342 433342] network_util.cc:296: Failed to determine local IP via external connectivity to: 8.8.8.8:53, [2001:4860:4860::8888]:53, falling back to hostname resolution
+2026-03-03 01:43:13,416	INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at http://127.0.0.1:8265 
+[INFO 01:43:14 RLinf] RLinf is running on a cluster with 1 node and 1 accelerator. The nodes' details are: 
+node_labels: []
+node_rank: 0
+ray_id: 9612b837da7da5bd3ebca0c90a105c8c02a42d183207e72c5575dc64
+node_ip: 127.0.1.1
+num_cpus: 32
+python_interpreter_path: /home/dps799/Desktop/RLinf/.venv/bin/python
+hardware_resources:
+- type: Accelerator
+  infos:
+  - type: Accelerator
+    model: NV_GPU:RTX 4060 Laptop GPU
+
+Node groups' details are: 
+label: cluster
+nodes: '0'
+hardware_type: Accelerator
+ignore_hardware: false
+env_configs: null
+
+label: node
+nodes: '0'
+hardware_type: null
+ignore_hardware: true
+env_configs: null
+
+[INFO 01:43:14 FlexiblePlacementStrategy] 
+[INFO 01:43:14 FlexiblePlacementStrategy] Using flexible placement with hardware ranks: [[0]], node groups: ['cluster'].
+[INFO 01:43:14 FlexiblePlacementStrategy] Generated 1 placements: [Placement(rank=0, cluster_node_rank=0, placement_node_rank=0, local_accelerator_rank=0, accelerator_type='NV_GPU', local_rank=0, local_world_size=1, visible_accelerators=['0'], isolate_accelerator=True, local_hardware_ranks=[0], node_group_label='cluster')].
+(pid=437267) 2026-03-03 01:43:16.382259: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+(pid=437267) 2026-03-03 01:43:16.401461: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+(pid=437267) 2026-03-03 01:43:16.401487: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+(pid=437267) 2026-03-03 01:43:16.402017: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+(pid=437267) 2026-03-03 01:43:16.405237: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+(pid=437267) To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
+(pid=437267) 2026-03-03 01:43:18.040915: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+(pid=437267) Gym has been unmaintained since 2022 and does not support NumPy 2.0 amongst other critical functionality.
+(pid=437267) Please upgrade to Gymnasium, the maintained drop-in replacement of Gym, or contact the authors of your software and request that they upgrade.
+(pid=437267) See the migration guide at https://gymnasium.farama.org/introduction/migration_guide/ for additional information.
+(pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/Cython/Distutils/old_build_ext.py:14: DeprecationWarning: dep_util is Deprecated. Use functions from setuptools instead.
+(pid=437267)   from distutils.dep_util import newer, newer_group
+(pid=437267) <frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead
+(pid=437267) Warning: Flow failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=437267) No module named 'flow'
+(pid=437267) Warning: CARLA failed to import. Set the environment variable D4RL_SUPPRESS_IMPORT_ERROR=1 to suppress this message.
+(pid=437267) No module named 'carla'
+(pid=437267) pybullet build time: Jan 29 2025 23:17:20
+(pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/pybullet_envs/env_bases.py:8: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
+(pid=437267)   from pkg_resources import parse_version
+(EmbodiedIQLFSDPPolicy pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/envs/registration.py:505: UserWarning: WARN: The environment antmaze-medium-play-v0 is out of date. You should consider upgrading to version `v2` with the environment ID `antmaze-medium-play-v2`.
+(EmbodiedIQLFSDPPolicy pid=437267)   logger.warn(
+(EmbodiedIQLFSDPPolicy pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/d4rl/locomotion/ant.py:180: UserWarning: This environment is deprecated. Please use the most recent version of this environment.
+(EmbodiedIQLFSDPPolicy pid=437267)   offline_env.OfflineEnv.__init__(self, **kwargs)
+(EmbodiedIQLFSDPPolicy pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/spaces/box.py:84: UserWarning: WARN: Box bound precision lowered by casting to float32
+(EmbodiedIQLFSDPPolicy pid=437267)   logger.warn(f"Box bound precision lowered by casting to {self.dtype}")
+(EmbodiedIQLFSDPPolicy pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/core.py:172: DeprecationWarning: WARN: Function `env.seed(seed)` is marked as deprecated and will be removed in the future. Please use `env.reset(seed=seed) instead.
+(EmbodiedIQLFSDPPolicy pid=437267)   deprecation(
+(EmbodiedIQLFSDPPolicy pid=437267) 
+(EmbodiedIQLFSDPPolicy pid=437267) 
+(EmbodiedIQLFSDPPolicy pid=437267) 
+(EmbodiedIQLFSDPPolicy pid=437267) 
+(EmbodiedIQLFSDPPolicy pid=437267) 
+(EmbodiedIQLFSDPPolicy pid=437267) /home/dps799/Desktop/RLinf/.venv/lib/python3.11/site-packages/gym/utils/seeding.py:38: DeprecationWarning: WARN: Function `rng.randn(*size)` is marked as deprecated and will be removed in the future. Please use `rng.standard_normal(size)` instead.
+(EmbodiedIQLFSDPPolicy pid=437267)   deprecation(
+{
+  "cluster": {
+    "num_nodes": 1,
+    "component_placement": {
+      "actor": "0-0"
+    }
+  },
+  "actor": {
+    "group_name": "ActorGroup",
+    "enable_strategy_checkpoint": false
+  },
+  "runner": {
+    "max_epochs": 1000000,
+    "logger": {
+      "log_path": "/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-medium-play-v0",
+      "experiment_name": "iql_antmaze-medium-play-v0"
+    }
+  },
+  "save_dir": "/home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-medium-play-v0",
+  "env_name": "antmaze-medium-play-v0",
+  "seed": 42,
+  "eval_episodes": 100,
+  "log_interval": 1000,
+  "eval_interval": 100000,
+  "batch_size": 256,
+  "max_steps": 1000000,
+  "tqdm": true,
+  "task": {
+    "mode": "offline"
+  },
+  "algorithm": {
+    "actor_lr": 0.0003,
+    "value_lr": 0.0003,
+    "critic_lr": 0.0003,
+    "hidden_dims": [
+      256,
+      256
+    ],
+    "discount": 0.99,
+    "tau": 0.005,
+    "expectile": 0.9,
+    "temperature": 10.0,
+    "dropout_rate": null,
+    "opt_decay_schedule": "cosine"
+  }
+}
+(EmbodiedIQLFSDPPolicy pid=437267) Target Goal:  (20.69101390655202, 20.534930128523673)

20260303-01:43:05-iql_antmaze-medium-play-v0/tensorboard/config.yaml

@@ -0,0 +1,36 @@
+cluster:
+  num_nodes: 1
+  component_placement:
+    actor: 0-0
+actor:
+  group_name: ActorGroup
+  enable_strategy_checkpoint: false
+runner:
+  max_epochs: 1000000
+  logger:
+    log_path: /home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-medium-play-v0
+    experiment_name: iql_antmaze-medium-play-v0
+save_dir: /home/dps799/Desktop/RLinf_merge/logs/20260303-01:43:05-iql_antmaze-medium-play-v0
+env_name: antmaze-medium-play-v0
+seed: 42
+eval_episodes: 100
+log_interval: 1000
+eval_interval: 100000
+batch_size: 256
+max_steps: 1000000
+tqdm: true
+task:
+  mode: offline
+algorithm:
+  actor_lr: 0.0003
+  value_lr: 0.0003
+  critic_lr: 0.0003
+  hidden_dims:
+  - 256
+  - 256
+  discount: 0.99
+  tau: 0.005
+  expectile: 0.9
+  temperature: 10.0
+  dropout_rate: null
+  opt_decay_schedule: cosine

20260303-01:43:05-iql_antmaze-medium-play-v0/tensorboard/events.out.tfevents.1772473394.dps799-Legion-Y9000P-IRX9.433342.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b485c93683744ea98302f56fbb6395fea85b641c0c7e4c3ed64aafd20a11d1a
+size 467852818

20260303-10:04:45-iql_antmaze-umaze-diverse-v0/42.txt

@@ -0,0 +1,10 @@
+100000 69.0
+200000 70.0
+300000 69.0
+400000 66.0
+500000 58.0
+600000 63.0
+700000 59.0
+800000 65.0
+900000 61.0
+1000000 67.0