Lmod has detected the following error: The following module(s) are unknown: "buildenv-gcccuda/12.1.1-gcc12.3.0" Please check the spelling or version number. Also try "module spider ..." It is also possible your cache file is out-of-date; it may help to try: $ module --ignore_cache load "buildenv-gcccuda/12.1.1-gcc12.3.0" Also make sure that all modulefiles written in TCL start with the string #%Module Already on 'bimamba' Your branch is up to date with 'origin/bimamba'. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Created output directory: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] [2026-04-20 11:59:40,645][pytorch_lightning.utilities.rank_zero][INFO] - Using 16bit Automatic Mixed Precision (AMP) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. [2026-04-20 11:59:40,770][pytorch_lightning.utilities.rank_zero][INFO] - GPU available: True (cuda), used: True [2026-04-20 11:59:40,770][pytorch_lightning.utilities.rank_zero][INFO] - TPU available: False, using: 0 TPU cores [2026-04-20 11:59:40,770][pytorch_lightning.utilities.rank_zero][INFO] - IPU available: False, using: 0 IPUs [2026-04-20 11:59:40,770][pytorch_lightning.utilities.rank_zero][INFO] - HPU available: False, using: 0 HPUs [2026-04-20 11:59:40,771][pytorch_lightning.utilities.rank_zero][INFO] - `Trainer(limit_val_batches=1)` was configured so 1 batch will be used. /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py:74: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Will load checkpoint from /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt Executing task: training out of ['training'] Error executing job with overrides: ['+name=train_stage_b_mamba_joint', 'algorithm=df_video_mamba3stage', '+customized_load=true', '+seperate_load=false', 'experiment.num_nodes=1', 'load=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt', 'dataset.save_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/datasets/minecraft', 'dataset.n_frames=200', '+dataset.n_frames_valid=200', '+dataset.angle_range=110', '+dataset.pos_range=2', '+dataset.wo_updown=false', '+dataset.customized_validation=true', '+dataset.add_timestamp_embedding=true', '+dataset.use_explicit_memory_frames=false', 'algorithm.training_stage=stage_b_diffusion_training', 'algorithm.stage_b_joint_training=true', 'algorithm.stage_b_memory_aux_weight=0.1', 'algorithm.use_mamba_memory_pipeline=true', 'algorithm.use_oracle_pose_eval=true', 'algorithm.enable_memory_noise_curriculum=false', '+algorithm.use_memory_attention=false', '+algorithm.relative_embedding=false', '+algorithm.memory_retrieval_topk=32', 'algorithm.diff_window_size=8', 'algorithm.memory_condition_length=0', 'algorithm.context_frames=100', '+algorithm.n_tokens=8', 'experiment.training.lr=2e-5', 'experiment.training.batch_size=8', 'experiment.training.checkpointing.every_n_train_steps=2500', 'experiment.training.max_steps=30000', 'experiment.validation.val_every_n_step=2500', '+output_dir=/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2/'] Traceback (most recent call last): File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 204, in run() # pylint: disable=no-value-for-parameter File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/main.py", line 94, in decorated_main _run_hydra( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra _run_app( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 457, in _run_app run_and_report( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 223, in run_and_report raise ex File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 220, in run_and_report return func() File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/utils.py", line 458, in lambda: hydra.run( File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/hydra.py", line 132, in run _ = ret.return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 260, in return_value raise self._return_value File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/core/utils.py", line 186, in run_job ret.return_value = task_function(task_cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 200, in run run_local(cfg) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/main.py", line 124, in run_local experiment.exec_task(task) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 172, in exec_task getattr(self, task)() File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 341, in training load_custom_checkpoint(algo=self.algo,checkpoint_path=self.ckpt_path) File "/proj/cvl/users/x_fahkh2/WorldMem_Repro/experiments/exp_base.py", line 74, in load_custom_checkpoint ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu')) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 1065, in load with _open_file_like(f, 'rb') as opened_file: File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 468, in _open_file_like return _open_file(name_or_buffer, mode) File "/proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torch/serialization.py", line 449, in __init__ super().__init__(open(name, mode)) FileNotFoundError: [Errno 2] No such file or directory: '/proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_a_40k_epoch3_step40000_oasis500m_ditreset_noxattn.ckpt' srun: error: node104: tasks 2,4-6: Exited with exit code 1 srun: Terminating StepId=14496.0 [2026-04-20T11:59:41.759] error: *** STEP 14496.0 ON node104 CANCELLED AT 2026-04-20T11:59:41 DUE TO TASK FAILURE *** srun: error: node104: tasks 0-1: Exited with exit code 1 srun: error: node104: task 3: Terminated srun: error: node104: task 7: Terminated srun: Force Terminated StepId=14496.0 Already on 'bimamba' Your branch is up to date with 'origin/bimamba'. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'training': Defaults list is missing `_self_`. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information warnings.warn(msg, UserWarning) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/fabric/__init__.py:40: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=AlexNet_Weights.IMAGENET1K_V1`. You can also use `weights=AlexNet_Weights.DEFAULT` to get the most up-to-date weights. warnings.warn(msg) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lpips/lpips.py:107: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:13,774][pytorch_lightning.utilities.rank_zero][INFO] - Using 16bit Automatic Mixed Precision (AMP) /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. [2026-04-20 12:00:13,890][pytorch_lightning.utilities.rank_zero][INFO] - GPU available: True (cuda), used: True [2026-04-20 12:00:13,890][pytorch_lightning.utilities.rank_zero][INFO] - TPU available: False, using: 0 TPU cores [2026-04-20 12:00:13,890][pytorch_lightning.utilities.rank_zero][INFO] - IPU available: False, using: 0 IPUs [2026-04-20 12:00:13,890][pytorch_lightning.utilities.rank_zero][INFO] - HPU available: False, using: 0 HPUs [2026-04-20 12:00:13,891][pytorch_lightning.utilities.rank_zero][INFO] - `Trainer(limit_val_batches=1)` was configured so 1 batch will be used. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. /proj/cvl/users/x_fahkh2/envs/worldmem/lib/python3.10/site-packages/lightning/pytorch/plugins/precision/amp.py:54: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. INFO: Initializing distributed: GLOBAL_RANK: 6, MEMBER: 7/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:20,106][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 6, MEMBER: 7/8 INFO: Initializing distributed: GLOBAL_RANK: 0, MEMBER: 1/8 [2026-04-20 12:00:20,274][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 0, MEMBER: 1/8 INFO: Initializing distributed: GLOBAL_RANK: 5, MEMBER: 6/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:20,549][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 5, MEMBER: 6/8 INFO: Initializing distributed: GLOBAL_RANK: 2, MEMBER: 3/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:20,698][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 2, MEMBER: 3/8 INFO: Initializing distributed: GLOBAL_RANK: 1, MEMBER: 2/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:20,703][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 1, MEMBER: 2/8 INFO: Initializing distributed: GLOBAL_RANK: 4, MEMBER: 5/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:20,834][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 4, MEMBER: 5/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:21,126][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 3, MEMBER: 4/8 INFO: Initializing distributed: GLOBAL_RANK: 3, MEMBER: 4/8 INFO: Initializing distributed: GLOBAL_RANK: 7, MEMBER: 8/8 Outputs will be saved to: /proj/cvl/users/x_fahkh2/WorldMem_Repro/checkpoints/bimamba_stage_b_joint_ckpt_40k_fixed_v2 Executing task: training out of ['training'] [2026-04-20 12:00:21,613][lightning.fabric.utilities.distributed][INFO] - Initializing distributed: GLOBAL_RANK: 7, MEMBER: 8/8 [2026-04-20 12:00:26,404][pytorch_lightning.utilities.rank_zero][INFO] - ---------------------------------------------------------------------------------------------------- distributed_backend=nccl All distributed processes registered. Starting with 8 processes ---------------------------------------------------------------------------------------------------- wandb: WARNING `resume` will be ignored since W&B syncing is set to `offline`. Starting a new run with run id stage_b_joint_offline. wandb: Tracking run with wandb version 0.17.9 wandb: W&B syncing is set to `offline` in this directory. wandb: Run `wandb online` or set WANDB_MODE=online to enable cloud syncing. INFO: LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,085][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 1 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 1 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 2 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 2 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 3 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 3 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 4 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 4 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 5 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 5 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 6 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 6 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] [2026-04-20 12:00:43,086][lightning.pytorch.accelerators.cuda][INFO] - LOCAL_RANK: 7 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: LOCAL_RANK: 7 - CUDA_VISIBLE_DEVICES: [0,1,2,3,4,5,6,7] INFO: | Name | Type | Params --------------------------------------------------------------------------------- 0 | diffusion_model | DiffusionMamba | 609 M 1 | validation_lpips_model | LearnedPerceptualImagePatchSimilarity | 2.5 M 2 | vae | AutoencoderKL | 229 M 3 | mamba_memory | BiMambaMemory | 4.5 M --------------------------------------------------------------------------------- 614 M Trainable params 231 M Non-trainable params 845 M Total params 3,383.355 Total estimated model params size (MB) [2026-04-20 12:00:44,109][lightning.pytorch.callbacks.model_summary][INFO] - | Name | Type | Params --------------------------------------------------------------------------------- 0 | diffusion_model | DiffusionMamba | 609 M 1 | validation_lpips_model | LearnedPerceptualImagePatchSimilarity | 2.5 M 2 | vae | AutoencoderKL | 229 M 3 | mamba_memory | BiMambaMemory | 4.5 M --------------------------------------------------------------------------------- 614 M Trainable params 231 M Non-trainable params 845 M Total params 3,383.355 Total estimated model params size (MB) INFO: SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,969][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,970][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. INFO: SLURM auto-requeueing enabled. Setting signal handlers. [2026-04-20 12:00:45,970][lightning.pytorch.trainer.connectors.signal_connector][INFO] - SLURM auto-requeueing enabled. Setting signal handlers. /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): /proj/cvl/users/x_fahkh2/WorldMem_Repro/algorithms/worldmem/models/mamba_memory.py:173: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. with torch.cuda.amp.autocast(enabled=False): Training: | | 0/? [00:00