Add sanitized world-model and HGD research checkpoints

.gitattributes

@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+planet/cartpole_swingup/tensorflow/model.ckpt-4504500.data-00000-of-00001 filter=lfs diff=lfs merge=lfs -text
+planet/cartpole_swingup/tensorflow/model.ckpt-4504500.meta filter=lfs diff=lfs merge=lfs -text
+planet/cheetah_run/tensorflow/model.ckpt-4504500.data-00000-of-00001 filter=lfs diff=lfs merge=lfs -text
+planet/cheetah_run/tensorflow/model.ckpt-4504500.meta filter=lfs diff=lfs merge=lfs -text
+planet/reacher_easy/tensorflow/model.ckpt-4504500.data-00000-of-00001 filter=lfs diff=lfs merge=lfs -text
+planet/reacher_easy/tensorflow/model.ckpt-4504500.meta filter=lfs diff=lfs merge=lfs -text

LICENSES.md

@@ -0,0 +1,17 @@
+# Attribution and license notices
+
+The `licenses/` directory contains the license notices shipped with the five
+source code repositories used to produce or load these checkpoints:
+
+| Project | Source notice |
+|---|---|
+| LeWM | [MIT notice](licenses/le-wm.txt) |
+| DINO-WM | [MIT notice](licenses/dino_wm.txt) |
+| PlaNet | [Apache 2.0 notice](licenses/planet.txt) |
+| TD-MPC2 | [MIT notice](licenses/tdmpc2.txt) |
+| DreamerV3 | [Source license](licenses/dreamerv3.txt) |
+
+These source notices are preserved for attribution. They do not create a new
+blanket license for this collection or replace any terms accompanying upstream
+weights and datasets. Consult the linked upstream artifact pages for those terms.
+No new collection-wide license is assigned in this preparation.

README.md

@@ -0,0 +1,87 @@
+---
+language:
+- en
+tags:
+- reinforcement-learning
+- world-model
+- robustness
+- adversarial-evaluation
+- checkpoint
+---
+
+# World-model attack research checkpoints
+
+Frozen checkpoints and supporting configurations for the research project
+**Misranking Imagined Futures: Test-Time Attacks on Frozen World-Model Planners**.
+This collection preserves the available checkpoints used by the local research
+implementations, including upstream pretrained models and locally trained models.
+
+| Directory | Included artifacts |
+|---|---|
+| `le-wm/` | Cube, Push-T, Reacher, and Two-Rooms model weights and architecture configurations; residual HGD and paper HGD defense checkpoints |
+| `dino_wm/` | Released Push-T and Wall/Two-Rooms model checkpoints and training configurations |
+| `planet/` | Cartpole Swingup, Reacher Easy, and Cheetah Run native TensorFlow checkpoints, frozen PyTorch inference exports, and TensorFlow parity fixtures |
+| `tdmpc2/` | The four RGB checkpoints selected by the research reference manifests: Finger Spin at 1,000,000 steps; Cup Catch and Cheetah Run at 1,500,000; Cartpole Balance at 250,000 |
+| `dreamerv3/` | Cheetah Run and Reacher Easy training configurations and immutable checkpoint directories |
+
+These checkpoints have different observation preprocessing, architectures, task
+interfaces, and Python dependencies. Use the corresponding research model loader;
+they are not interchangeable Transformers models. The PlaNet PyTorch exports
+contain 32 inference tensors and require the research implementation of PlaNet.
+The DINO-WM checkpoints also require their DINOv2 encoder dependency and native
+task data. LeWM and defense evaluations require their datasets and simulators.
+
+## Download and integrity
+
+Download the repository using the Hugging Face CLI:
+
+```bash
+hf download mohendy/world-model-attack-checkpoints --local-dir ./checkpoints
+python ./checkpoints/verify_download.py
+```
+
+If the repository is private, authenticate with `hf auth login` first. For a
+durable experiment record, pass `--revision` with the repository commit SHA
+returned by the upload, instead of relying on the moving `main` branch. A partial
+download can be checked with `python verify_download.py --model planet` from the
+downloaded directory.
+
+`manifest.json` records the SHA-256 and size of every checkpoint/configuration,
+plus its original relative location in the research workspace. Its
+`source_relative_path` field provides the destination mapping for the code
+release's artifact layout. Binary model checkpoints preserve their source bytes.
+Configuration output paths are sanitized and relative. HGD checkpoints are
+re-exported with sanitized metadata; their `state_dict` tensors were verified to
+be exactly equal to the originals. Both original and released hashes are recorded.
+
+Copy or link artifacts into the code release using that mapping, and set explicit
+checkpoint/data paths for your machine. Run directories for Dreamer must include
+their `config.yaml` and chosen `ckpt/<checkpoint-id>/` directory. Select the
+immutable directory explicitly instead of a training run's `latest` pointer.
+
+Model-specific `runtime.json` files record the installed Python and package
+versions from the research environment. They are provenance records, not a
+guarantee that all packages resolve unchanged on another platform. The code
+release maintains environment installation instructions and validation results.
+
+## Scope and provenance
+
+Datasets, experiment logs, videos, and per-episode adversarial perturbations are
+not included. Checkpoint availability alone does not reproduce the paper tables:
+the task datasets, episode identities, seed schedules, preprocessing, planner
+configuration, and matched clean reference results must also agree.
+
+Upstream projects and weight sources:
+
+- [LeWM code](https://github.com/lucas-maes/le-wm) and
+  [checkpoint collection](https://huggingface.co/collections/quentinll/lewm).
+- [DINO-WM code](https://github.com/gaoyuezhou/dino_wm) and its
+  [released artifacts](https://osf.io/bmw48/).
+- [PlaNet code](https://github.com/google-research/planet); the local inference
+  exports record source artifact revision
+  `b26a7fe027c018f62b3178c8e126371fbdee33e5` in `attack_export.json`.
+- [TD-MPC2 code](https://github.com/nicklashansen/tdmpc2).
+- [DreamerV3 code](https://github.com/danijar/dreamerv3).
+
+See [LICENSES.md](LICENSES.md) for the accompanying upstream notices. Original
+upstream work and local research additions should be credited separately.

dino_wm/pusht/hydra.yaml

@@ -0,0 +1,96 @@
+ckpt_base_path: ./checkpoints
+training:
+  seed: 0
+  epochs: 100
+  batch_size: 32
+  save_every_x_epoch: 1
+  reconstruct_every_x_batch: 500
+  num_reconstruct_samples: 6
+  encoder_lr: 1.0e-06
+  decoder_lr: 0.0003
+  predictor_lr: 0.0005
+  action_encoder_lr: 0.0005
+img_size: 224
+frameskip: 5
+concat_dim: 1
+normalize_action: true
+action_emb_dim: 10
+num_action_repeat: 1
+proprio_emb_dim: 10
+num_proprio_repeat: 1
+num_hist: 3
+num_pred: 1
+has_predictor: true
+has_decoder: true
+state_based: false
+model:
+  _target_: models.visual_world_model.VWorldModel
+  image_size: 224
+  num_hist: 3
+  num_pred: 1
+  train_encoder: false
+  train_predictor: true
+  train_decoder: true
+debug: false
+plan_settings:
+  plan_cfg_path: conf/plan.yaml
+  planner:
+  - cem
+  goal_source:
+  - dset
+  goal_H:
+  - 5
+  alpha:
+  - 0.1
+  - 1
+env:
+  name: pusht
+  args: []
+  kwargs:
+    with_velocity: true
+    with_target: true
+  dataset:
+    _target_: datasets.pusht_dset.load_pusht_slice_train_val
+    with_velocity: true
+    n_rollout: null
+    normalize_action: true
+    data_path: ${oc.env:DATASET_DIR}/pusht_noise
+    split_ratio: 0.9
+    transform:
+      _target_: datasets.img_transforms.default_transform
+      img_size: 224
+  decoder_path: null
+  num_workers: 16
+encoder:
+  _target_: models.dino.DinoV2Encoder
+  name: dinov2_vits14
+  feature_key: x_norm_patchtokens
+action_encoder:
+  _target_: models.proprio.ProprioceptiveEmbedding
+  num_frames: 1
+  tubelet_size: 1
+  use_3d_pos: false
+proprio_encoder:
+  _target_: models.proprio.ProprioceptiveEmbedding
+  num_frames: 1
+  tubelet_size: 1
+  use_3d_pos: false
+decoder:
+  _target_: models.vqvae.VQVAE
+  channel: 384
+  n_embed: 2048
+  n_res_block: 4
+  n_res_channel: 128
+  quantize: false
+predictor:
+  _target_: models.vit.ViTPredictor
+  depth: 6
+  heads: 16
+  mlp_dim: 2048
+  dropout: 0.1
+  emb_dropout: 0
+  pool: mean
+saved_folder: ./checkpoints
+effective_batch_size: 32
+gpu_batch_size: 1
+wandb_run_id: null

dino_wm/pusht/model_latest.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e909f0cec958fc0b49a79f2e85730ae6b5f83dee61a224705f91883eb3bb74c5
+size 377929744

dino_wm/runtime.json

@@ -0,0 +1,103 @@
+{
+  "python": "3.10.14",
+  "platform": "Linux x86_64",
+  "packages": {
+    "hydra-submitit-launcher": "1.2.0",
+    "submitit": "1.5.1",
+    "typing-extensions": "4.15.0",
+    "cloudpickle": "2.1.0",
+    "hydra-core": "1.2.0",
+    "packaging": "24.0",
+    "antlr4-python3-runtime": "4.9.3",
+    "omegaconf": "2.3.0",
+    "pyyaml": "6.0.1",
+    "accelerate": "0.26.1",
+    "safetensors": "0.4.3",
+    "huggingface-hub": "0.36.2",
+    "tqdm": "4.66.4",
+    "requests": "2.32.3",
+    "certifi": "2024.6.2",
+    "urllib3": "2.2.1",
+    "idna": "3.7",
+    "charset-normalizer": "3.3.2",
+    "hf-xet": "1.2.0",
+    "fsspec": "2024.5.0",
+    "filelock": "3.14.0",
+    "torch": "2.7.0+cu128",
+    "triton": "3.3.0",
+    "setuptools": "70.0.0",
+    "nvidia-cufile-cu12": "1.13.0.11",
+    "nvidia-nvjitlink-cu12": "12.8.61",
+    "nvidia-nvtx-cu12": "12.8.55",
+    "nvidia-nccl-cu12": "2.26.2",
+    "nvidia-cusparselt-cu12": "0.6.3",
+    "nvidia-cusparse-cu12": "12.5.7.53",
+    "nvidia-cusolver-cu12": "11.7.2.55",
+    "nvidia-cublas-cu12": "12.8.3.14",
+    "nvidia-curand-cu12": "10.3.9.55",
+    "nvidia-cufft-cu12": "11.3.3.41",
+    "nvidia-cudnn-cu12": "9.7.1.26",
+    "nvidia-cuda-cupti-cu12": "12.8.57",
+    "nvidia-cuda-runtime-cu12": "12.8.57",
+    "nvidia-cuda-nvrtc-cu12": "12.8.61",
+    "jinja2": "3.1.4",
+    "markupsafe": "2.1.5",
+    "networkx": "3.2.1",
+    "sympy": "1.14.0",
+    "mpmath": "1.3.0",
+    "psutil": "7.2.2",
+    "numpy": "1.26.4",
+    "dill": "0.4.1",
+    "scikit-image": "0.19.3",
+    "pywavelets": "1.6.0",
+    "tifffile": "2024.5.22",
+    "imageio": "2.34.1",
+    "pillow": "10.3.0",
+    "scipy": "1.13.1",
+    "numba": "0.59.1",
+    "llvmlite": "0.42.0",
+    "decord": "0.6.0",
+    "h5py": "3.11.0",
+    "pygame": "2.5.2",
+    "pymunk": "6.8.0",
+    "cffi": "1.16.0",
+    "pycparser": "2.22",
+    "mujoco": "3.2.7",
+    "pyopengl": "3.1.7",
+    "glfw": "2.7.0",
+    "etils": "1.5.2",
+    "zipp": "3.19.1",
+    "importlib-resources": "6.4.0",
+    "absl-py": "2.1.0",
+    "dm-control": "1.0.27",
+    "pyparsing": "3.1.2",
+    "protobuf": "6.33.6",
+    "lxml": "5.3.1",
+    "labmaze": "1.0.6",
+    "dm-tree": "0.1.8",
+    "dm-env": "1.6",
+    "gym": "0.23.1",
+    "gym-notices": "0.0.8",
+    "wandb": "0.25.1",
+    "sentry-sdk": "2.3.1",
+    "pydantic": "2.12.5",
+    "typing-inspection": "0.4.2",
+    "pydantic-core": "2.41.5",
+    "annotated-types": "0.7.0",
+    "platformdirs": "4.2.2",
+    "gitpython": "3.1.43",
+    "gitdb": "4.0.11",
+    "smmap": "5.0.1",
+    "click": "8.1.7",
+    "einops": "0.4.1",
+    "matplotlib": "3.5.3",
+    "python-dateutil": "2.9.0.post0",
+    "six": "1.16.0",
+    "kiwisolver": "1.4.5",
+    "fonttools": "4.53.0",
+    "cycler": "0.12.1",
+    "imageio-ffmpeg": "0.4.9",
+    "torchvision": "0.22.0+cu128"
+  },
+  "metadata_conflicts": []
+}

dino_wm/wall_single/hydra.yaml

@@ -0,0 +1,96 @@
+ckpt_base_path: ./checkpoints
+training:
+  seed: 0
+  epochs: 1000
+  batch_size: 128
+  save_every_x_epoch: 5
+  reconstruct_every_x_batch: 500
+  num_reconstruct_samples: 6
+  encoder_lr: 1.0e-06
+  decoder_lr: 0.0003
+  predictor_lr: 0.0005
+  action_encoder_lr: 0.0005
+img_size: 224
+frameskip: 5
+concat_dim: 1
+normalize_action: true
+action_emb_dim: 10
+num_action_repeat: 1
+proprio_emb_dim: 10
+num_proprio_repeat: 1
+num_hist: 1
+num_pred: 1
+has_predictor: true
+has_decoder: true
+state_based: false
+model:
+  _target_: models.visual_world_model.VWorldModel
+  image_size: 224
+  num_hist: 1
+  num_pred: 1
+  train_encoder: false
+  train_predictor: true
+  train_decoder: true
+debug: true
+plan_settings:
+  plan_cfg_path: conf/plan.yaml
+  planner:
+  - gd
+  - cem
+  goal_source:
+  - dset
+  - random_state
+  goal_H:
+  - 5
+  alpha:
+  - 0.1
+  - 1
+env:
+  name: wall
+  args: []
+  kwargs: {}
+  load_dir: ''
+  dataset:
+    _target_: datasets.wall_dset.load_wall_slice_train_val
+    n_rollout: null
+    normalize_action: true
+    data_path: ${oc.env:DATASET_DIR}/wall_single
+    split_ratio: 0.9
+    split_mode: random
+    transform:
+      _target_: datasets.img_transforms.default_transform
+      img_size: 224
+  decoder_path: null
+  linear_probe_model_path: null
+  num_workers: 0
+encoder:
+  _target_: models.dino.DinoV2Encoder
+  name: dinov2_vits14
+  feature_key: x_norm_patchtokens
+action_encoder:
+  _target_: models.proprio.ProprioceptiveEmbedding
+  num_frames: 1
+  tubelet_size: 1
+  use_3d_pos: false
+proprio_encoder:
+  _target_: models.proprio.ProprioceptiveEmbedding
+  num_frames: 1
+  tubelet_size: 1
+  use_3d_pos: false
+decoder:
+  _target_: models.vqvae.VQVAE
+  n_embed: 2048
+  n_res_block: 4
+  n_res_channel: 128
+  quantize: false
+predictor:
+  _target_: models.vit.ViTPredictor
+  depth: 6
+  heads: 16
+  mlp_dim: 2048
+  dropout: 0.1
+  emb_dropout: 0.1
+  pool: mean
+saved_folder: ./checkpoints
+effective_batch_size: 128
+gpu_batch_size: 4

dino_wm/wall_single/model_latest.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8441971becdae934fe08de5b163398390a32f6fe1fb0a8df113290e2468e142b
+size 368656057

dreamerv3/cheetah_run/ckpt/20260901T093748F019112/agent.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d94c5824d8c77ba8191ccd5207409986a08bbf7e04ae5d55a9d468c7d158af3
+size 2042806149

dreamerv3/cheetah_run/ckpt/20260901T093748F019112/step.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69901686be1d13680e0c0ddbfd0a0d2f2a70bb0ae46a2e636bf78bf5bc45cce2
+size 17

dreamerv3/cheetah_run/config.yaml

@@ -0,0 +1,185 @@
+agent:
+  ac_grads: false
+  advnorm: {impl: none, limit: 1e-08, rate: 0.01}
+  conhead: {act: silu, layers: 1, norm: rms, output: binary, outscale: 1.0, 
+      units: 1024, winit: trunc_normal_in}
+  contdisc: true
+  dec:
+    simple:
+      act: silu
+      bspace: 8
+      depth: 64
+      kernel: 5
+      layers: 3
+      mults: [2, 3, 4, 4]
+      norm: rms
+      outer: false
+      outscale: 1.0
+      strided: false
+      units: 1024
+      winit: trunc_normal_in
+    typ: simple
+  dyn:
+    rssm: {absolute: false, act: silu, blocks: 8, classes: 64, deter: 8192, 
+        dynlayers: 1, free_nats: 1.0, hidden: 1024, imglayers: 2, norm: rms, 
+        obslayers: 1, outscale: 1.0, stoch: 32, unimix: 0.01, winit: 
+        trunc_normal_in}
+    typ: rssm
+  enc:
+    simple:
+      act: silu
+      depth: 64
+      kernel: 5
+      layers: 3
+      mults: [2, 3, 4, 4]
+      norm: rms
+      outer: false
+      strided: false
+      symlog: true
+      units: 1024
+      winit: trunc_normal_in
+    typ: simple
+  horizon: 333
+  imag_last: 0
+  imag_length: 15
+  imag_loss: {actent: 0.0003, lam: 0.95, slowreg: 1.0, slowtar: false}
+  loss_scales: {con: 1.0, dyn: 1.0, policy: 1.0, rec: 1.0, rep: 0.1, repval: 0.3,
+    rew: 1.0, value: 1.0}
+  opt: {agc: 0.3, anneal: 0, beta1: 0.9, beta2: 0.999, eps: 1e-20, lr: 4e-05, 
+      momentum: true, schedule: const, warmup: 1000, wd: 0.0}
+  policy: {act: silu, layers: 3, maxstd: 1.0, minstd: 0.1, norm: rms, outscale: 
+      0.01, unimix: 0.01, units: 1024, winit: trunc_normal_in}
+  policy_dist_cont: bounded_normal
+  policy_dist_disc: categorical
+  repl_loss: {lam: 0.95, slowreg: 1.0, slowtar: false}
+  report: true
+  report_gradnorms: false
+  repval_grad: true
+  repval_loss: true
+  retnorm: {debias: false, impl: perc, limit: 1.0, perchi: 95.0, perclo: 5.0, 
+      rate: 0.01}
+  reward_grad: true
+  rewhead: {act: silu, bins: 255, layers: 1, norm: rms, output: symexp_twohot, 
+      outscale: 0.0, units: 1024, winit: trunc_normal_in}
+  slowvalue: {every: 1, rate: 0.02}
+  valnorm: {impl: none, limit: 1e-08, rate: 0.01}
+  value: {act: silu, bins: 255, layers: 3, norm: rms, output: symexp_twohot, 
+      outscale: 0.0, units: 1024, winit: trunc_normal_in}
+batch_length: 64
+batch_size: 16
+clock_addr: ''
+clock_port: ''
+consec_report: 1
+consec_train: 1
+env:
+  atari:
+    actions: all
+    aggregate: max
+    autostart: false
+    clip_reward: false
+    gray: true
+    lives: unused
+    noops: 30
+    pooling: 2
+    repeat: 4
+    resize: pillow
+    size: [96, 96]
+    sticky: true
+  atari100k:
+    actions: needed
+    autostart: false
+    clip_reward: false
+    gray: false
+    lives: unused
+    noops: 30
+    repeat: 4
+    resize: pillow
+    size: [64, 64]
+    sticky: false
+  crafter:
+    logs: false
+    size: [64, 64]
+  dmc:
+    camera: -1
+    image: true
+    proprio: false
+    repeat: 1
+    size: [64, 64]
+  dmlab:
+    episodic: true
+    repeat: 4
+    size: [64, 64]
+    use_seed: true
+  loconav:
+    camera: -1
+    repeat: 1
+    size: [64, 64]
+  minecraft:
+    break_speed: 100.0
+    length: 36000
+    logs: false
+    size: [64, 64]
+  procgen:
+    resize: pillow
+    size: [96, 96]
+errfile: false
+ipv6: false
+jax:
+  compute_dtype: bfloat16
+  coordinator_address: ''
+  debug: false
+  enable_policy: true
+  expect_devices: 0
+  jit: true
+  mock_devices: 0
+  platform: cuda
+  policy_devices: [0]
+  prealloc: true
+  train_devices: [0]
+logdir: ./logdir/dreamer/cheetah_run/slurm-23188
+logger:
+  filter: score|length|fps|ratio|train/loss/|train/rand/
+  fps: 15
+  outputs: [jsonl, scope]
+  timer: true
+  user: ''
+method: name
+random_agent: false
+replay:
+  chunksize: 1024
+  fracs: {priority: 0.0, recency: 0.0, uniform: 1.0}
+  online: true
+  prio: {exponent: 0.8, initial: inf, maxfrac: 0.5, zero_on_sample: true}
+  priosignal: model
+  recexp: 1.0
+  size: 5000000.0
+replay_context: 1
+replica: 0
+replicas: 1
+report_length: 32
+run:
+  actor_addr: localhost:{auto}
+  actor_batch: -1
+  actor_threads: 1
+  agent_process: false
+  debug: true
+  duration: 0
+  envs: 16
+  episode_timeout: 180
+  eval_envs: 4
+  eval_eps: 1
+  from_checkpoint: ''
+  log_every: 120
+  logger_addr: localhost:{auto}
+  remote_envs: false
+  remote_replay: false
+  replay_addr: localhost:{auto}
+  report_batches: 1
+  report_every: 300
+  save_every: 900
+  steps: 1100000.0
+  train_ratio: 256.0
+  usage: {gc: false, gputil: false, malloc: false, nvsmi: true, psutil: true}
+script: train
+seed: 0
+task: dmc_cheetah_run

dreamerv3/reacher_easy/ckpt/20260830T195435F325934/agent.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1fb60436f02fcfa6ba1fbd0b5b5078142cb6f6c34b13c42e4a5967ef047b394
+size 2042658588

dreamerv3/reacher_easy/ckpt/20260830T195435F325934/step.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e5c5685c2d39badee13ea957052bbe8a3f71e577e2369e55151ad875842cb2f
+size 17

dreamerv3/reacher_easy/config.yaml

@@ -0,0 +1,185 @@
+agent:
+  ac_grads: false
+  advnorm: {impl: none, limit: 1e-08, rate: 0.01}
+  conhead: {act: silu, layers: 1, norm: rms, output: binary, outscale: 1.0, 
+      units: 1024, winit: trunc_normal_in}
+  contdisc: true
+  dec:
+    simple:
+      act: silu
+      bspace: 8
+      depth: 64
+      kernel: 5
+      layers: 3
+      mults: [2, 3, 4, 4]
+      norm: rms
+      outer: false
+      outscale: 1.0
+      strided: false
+      units: 1024
+      winit: trunc_normal_in
+    typ: simple
+  dyn:
+    rssm: {absolute: false, act: silu, blocks: 8, classes: 64, deter: 8192, 
+        dynlayers: 1, free_nats: 1.0, hidden: 1024, imglayers: 2, norm: rms, 
+        obslayers: 1, outscale: 1.0, stoch: 32, unimix: 0.01, winit: 
+        trunc_normal_in}
+    typ: rssm
+  enc:
+    simple:
+      act: silu
+      depth: 64
+      kernel: 5
+      layers: 3
+      mults: [2, 3, 4, 4]
+      norm: rms
+      outer: false
+      strided: false
+      symlog: true
+      units: 1024
+      winit: trunc_normal_in
+    typ: simple
+  horizon: 333
+  imag_last: 0
+  imag_length: 15
+  imag_loss: {actent: 0.0003, lam: 0.95, slowreg: 1.0, slowtar: false}
+  loss_scales: {con: 1.0, dyn: 1.0, policy: 1.0, rec: 1.0, rep: 0.1, repval: 0.3,
+    rew: 1.0, value: 1.0}
+  opt: {agc: 0.3, anneal: 0, beta1: 0.9, beta2: 0.999, eps: 1e-20, lr: 4e-05, 
+      momentum: true, schedule: const, warmup: 1000, wd: 0.0}
+  policy: {act: silu, layers: 3, maxstd: 1.0, minstd: 0.1, norm: rms, outscale: 
+      0.01, unimix: 0.01, units: 1024, winit: trunc_normal_in}
+  policy_dist_cont: bounded_normal
+  policy_dist_disc: categorical
+  repl_loss: {lam: 0.95, slowreg: 1.0, slowtar: false}
+  report: true
+  report_gradnorms: false
+  repval_grad: true
+  repval_loss: true
+  retnorm: {debias: false, impl: perc, limit: 1.0, perchi: 95.0, perclo: 5.0, 
+      rate: 0.01}
+  reward_grad: true
+  rewhead: {act: silu, bins: 255, layers: 1, norm: rms, output: symexp_twohot, 
+      outscale: 0.0, units: 1024, winit: trunc_normal_in}
+  slowvalue: {every: 1, rate: 0.02}
+  valnorm: {impl: none, limit: 1e-08, rate: 0.01}
+  value: {act: silu, bins: 255, layers: 3, norm: rms, output: symexp_twohot, 
+      outscale: 0.0, units: 1024, winit: trunc_normal_in}
+batch_length: 64
+batch_size: 16
+clock_addr: ''
+clock_port: ''
+consec_report: 1
+consec_train: 1
+env:
+  atari:
+    actions: all
+    aggregate: max
+    autostart: false
+    clip_reward: false
+    gray: true
+    lives: unused
+    noops: 30
+    pooling: 2
+    repeat: 4
+    resize: pillow
+    size: [96, 96]
+    sticky: true
+  atari100k:
+    actions: needed
+    autostart: false
+    clip_reward: false
+    gray: false
+    lives: unused
+    noops: 30
+    repeat: 4
+    resize: pillow
+    size: [64, 64]
+    sticky: false
+  crafter:
+    logs: false
+    size: [64, 64]
+  dmc:
+    camera: -1
+    image: true
+    proprio: false
+    repeat: 1
+    size: [64, 64]
+  dmlab:
+    episodic: true
+    repeat: 4
+    size: [64, 64]
+    use_seed: true
+  loconav:
+    camera: -1
+    repeat: 1
+    size: [64, 64]
+  minecraft:
+    break_speed: 100.0
+    length: 36000
+    logs: false
+    size: [64, 64]
+  procgen:
+    resize: pillow
+    size: [96, 96]
+errfile: false
+ipv6: false
+jax:
+  compute_dtype: bfloat16
+  coordinator_address: ''
+  debug: false
+  enable_policy: true
+  expect_devices: 0
+  jit: true
+  mock_devices: 0
+  platform: cuda
+  policy_devices: [0]
+  prealloc: true
+  train_devices: [0]
+logdir: ./logdir/dreamer/reacher_easy/20260830T125224
+logger:
+  filter: score|length|fps|ratio|train/loss/|train/rand/
+  fps: 15
+  outputs: [jsonl, scope]
+  timer: true
+  user: ''
+method: name
+random_agent: false
+replay:
+  chunksize: 1024
+  fracs: {priority: 0.0, recency: 0.0, uniform: 1.0}
+  online: true
+  prio: {exponent: 0.8, initial: inf, maxfrac: 0.5, zero_on_sample: true}
+  priosignal: model
+  recexp: 1.0
+  size: 5000000.0
+replay_context: 1
+replica: 0
+replicas: 1
+report_length: 32
+run:
+  actor_addr: localhost:{auto}
+  actor_batch: -1
+  actor_threads: 1
+  agent_process: false
+  debug: true
+  duration: 0
+  envs: 16
+  episode_timeout: 180
+  eval_envs: 4
+  eval_eps: 1
+  from_checkpoint: ''
+  log_every: 120
+  logger_addr: localhost:{auto}
+  remote_envs: false
+  remote_replay: false
+  replay_addr: localhost:{auto}
+  report_batches: 1
+  report_every: 300
+  save_every: 900
+  steps: 1100000.0
+  train_ratio: 256.0
+  usage: {gc: false, gputil: false, malloc: false, nvsmi: true, psutil: true}
+script: train
+seed: 0
+task: dmc_reacher_easy

dreamerv3/runtime.json

@@ -0,0 +1,126 @@
+{
+  "python": "3.12.14",
+  "platform": "Linux x86_64",
+  "packages": {
+    "nvidia-cuda-nvcc-cu12": "12.9.86",
+    "google-resumable-media": "2.10.2",
+    "google-crc32c": "1.8.0",
+    "scope": "0.7.1",
+    "uvicorn": "0.52.4",
+    "websockets": "17.1",
+    "watchfiles": "1.2.0",
+    "anyio": "4.14.2",
+    "typing-extensions": "4.16.0",
+    "idna": "3.19",
+    "uvloop": "0.22.1",
+    "pyyaml": "6.0.3",
+    "python-dotenv": "1.2.3",
+    "httptools": "0.8.0",
+    "h11": "0.16.0",
+    "click": "8.5.0",
+    "pillow": "12.3.0",
+    "orjson": "3.12.0",
+    "numpy": "1.26.4",
+    "mediapy": "1.2.7",
+    "matplotlib": "3.11.1",
+    "python-dateutil": "2.9.0.post0",
+    "six": "1.17.0",
+    "pyparsing": "3.3.2",
+    "packaging": "26.3",
+    "kiwisolver": "1.5.1",
+    "fonttools": "4.63.0",
+    "cycler": "0.12.1",
+    "contourpy": "1.3.3",
+    "ipython": "9.17.0",
+    "traitlets": "5.16.1",
+    "stack-data": "0.6.3",
+    "pure-eval": "0.2.3",
+    "asttokens": "3.0.2",
+    "executing": "2.2.1",
+    "pygments": "2.21.0",
+    "psutil": "7.2.2",
+    "prompt-toolkit": "3.0.53",
+    "wcwidth": "0.8.3",
+    "pexpect": "4.9.0",
+    "ptyprocess": "0.7.0",
+    "matplotlib-inline": "0.2.2",
+    "jedi": "0.20.0",
+    "parso": "0.8.7",
+    "ipython-pygments-lexers": "1.1.1",
+    "filelock": "3.32.4",
+    "fastapi": "0.141.1",
+    "annotated-doc": "0.0.5",
+    "typing-inspection": "0.4.4",
+    "pydantic": "2.13.5",
+    "pydantic-core": "2.46.5",
+    "annotated-types": "0.8.0",
+    "starlette": "1.6.0",
+    "elements": "3.22.0",
+    "ruamel-yaml": "0.19.1",
+    "google-cloud-storage": "3.13.1",
+    "requests": "2.34.2",
+    "certifi": "2026.7.22",
+    "urllib3": "2.7.0",
+    "charset-normalizer": "3.5.1",
+    "google-cloud-core": "2.7.0",
+    "google-auth": "2.57.0",
+    "cryptography": "50.0.1",
+    "cffi": "2.1.1",
+    "pycparser": "3.0",
+    "pyasn1-modules": "0.4.2",
+    "pyasn1": "0.6.4",
+    "google-api-core": "2.34.0",
+    "proto-plus": "1.28.4",
+    "protobuf": "7.36.0",
+    "googleapis-common-protos": "1.75.2",
+    "av": "18.1.0",
+    "jaxtyping": "0.3.11",
+    "wadler-lindig": "0.1.7",
+    "colored-traceback": "0.4.2",
+    "cloudpickle": "3.1.2",
+    "mujoco": "3.12.0",
+    "pyopengl": "3.1.10",
+    "glfw": "2.10.2",
+    "etils": "1.14.0",
+    "zipp": "4.1.0",
+    "fsspec": "2026.7.0",
+    "absl-py": "2.5.0",
+    "dm-control": "1.0.45",
+    "tqdm": "4.70.0",
+    "scipy": "1.17.1",
+    "setuptools": "83.0.0",
+    "lxml": "6.1.2",
+    "labmaze": "1.0.6",
+    "dm-tree": "0.1.10",
+    "wrapt": "2.3.0",
+    "attrs": "26.1.0",
+    "dm-env": "1.6",
+    "einops": "0.8.2",
+    "chex": "0.1.90",
+    "toolz": "1.1.0",
+    "jaxlib": "0.7.1",
+    "ml-dtypes": "0.5.4",
+    "jax": "0.7.1",
+    "opt-einsum": "3.4.0",
+    "optax": "0.2.5",
+    "granular": "0.23.3",
+    "msgpack": "1.2.2",
+    "portal": "3.8.1",
+    "ninjax": "3.6.3",
+    "jax-cuda12-plugin": "0.7.1",
+    "nvidia-nvshmem-cu12": "3.7.2",
+    "nvidia-cuda-cccl-cu12": "12.9.27",
+    "nvidia-cuda-nvrtc-cu12": "12.9.86",
+    "nvidia-nvjitlink-cu12": "12.9.86",
+    "nvidia-nccl-cu12": "2.31.2",
+    "nvidia-cusparse-cu12": "12.5.10.65",
+    "nvidia-cusolver-cu12": "11.7.5.82",
+    "nvidia-cublas-cu12": "12.9.2.10",
+    "nvidia-cufft-cu12": "11.4.1.4",
+    "nvidia-cudnn-cu12": "9.25.1.1",
+    "nvidia-cuda-runtime-cu12": "12.9.79",
+    "nvidia-cuda-cupti-cu12": "12.9.79",
+    "jax-cuda12-pjrt": "0.7.1"
+  },
+  "metadata_conflicts": []
+}

le-wm/cube/config.json

@@ -0,0 +1,49 @@
+{
+    "_target_": "stable_worldmodel.wm.lewm.LeWM",
+    "encoder": {
+        "_target_": "stable_pretraining.backbone.utils.vit_hf",
+        "size": "tiny",
+        "patch_size": 14,
+        "image_size": 224,
+        "pretrained": false,
+        "use_mask_token": false
+    },
+    "predictor": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Predictor",
+        "num_frames": 3,
+        "input_dim": 192,
+        "hidden_dim": 192,
+        "output_dim": 192,
+        "depth": 6,
+        "heads": 16,
+        "mlp_dim": 2048,
+        "dim_head": 64,
+        "dropout": 0.1,
+        "emb_dropout": 0.0
+    },
+    "action_encoder": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Embedder",
+        "input_dim": 25,
+        "emb_dim": 192
+    },
+    "projector": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    },
+    "pred_proj": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    }
+}

le-wm/cube/weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2839a907362f403f9136383016e91774373a295d958ae75121791f22a9fddf89
+size 72291425

le-wm/defenses/cube/paper_hgd/hgd_paper.best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04ed4a9b4d9edea42f4e4ec7b489ae79378890a3c2d505077a97d69d19fd1ec9
+size 132576538

le-wm/defenses/cube/paper_hgd/hgd_paper.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ebae97be8b61cefde85958dadea28d5ae34c959b3cd183f4932f6bfe1debdf2
+size 132574859

le-wm/defenses/cube/residual_hgd/hgd_eps16.best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68f58d28e6b39954331143c2f66e736f402882f15fce35ace8afc8d464c17f07
+size 5917691

le-wm/defenses/cube/residual_hgd/hgd_eps16.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc3af72c1ae2ed328c8907f87e5709b719e00c8dd61e7c32ccb9dc722951772a
+size 5916747

le-wm/pusht/config.json

@@ -0,0 +1,49 @@
+{
+    "_target_": "stable_worldmodel.wm.lewm.LeWM",
+    "encoder": {
+        "_target_": "stable_pretraining.backbone.utils.vit_hf",
+        "size": "tiny",
+        "patch_size": 14,
+        "image_size": 224,
+        "pretrained": false,
+        "use_mask_token": false
+    },
+    "predictor": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Predictor",
+        "num_frames": 3,
+        "input_dim": 192,
+        "hidden_dim": 192,
+        "output_dim": 192,
+        "depth": 6,
+        "heads": 16,
+        "mlp_dim": 2048,
+        "dim_head": 64,
+        "dropout": 0.1,
+        "emb_dropout": 0.0
+    },
+    "action_encoder": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Embedder",
+        "input_dim": 10,
+        "emb_dim": 192
+    },
+    "projector": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    },
+    "pred_proj": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    }
+}

le-wm/pusht/weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48938400ae3464c9680731287f583a9cb516f55a8ec64ea13a91be47fb15b607
+size 72290721

le-wm/reacher/config.json

@@ -0,0 +1,49 @@
+{
+    "_target_": "stable_worldmodel.wm.lewm.LeWM",
+    "encoder": {
+        "_target_": "stable_pretraining.backbone.utils.vit_hf",
+        "size": "tiny",
+        "patch_size": 14,
+        "image_size": 224,
+        "pretrained": false,
+        "use_mask_token": false
+    },
+    "predictor": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Predictor",
+        "num_frames": 3,
+        "input_dim": 192,
+        "hidden_dim": 192,
+        "output_dim": 192,
+        "depth": 6,
+        "heads": 16,
+        "mlp_dim": 2048,
+        "dim_head": 64,
+        "dropout": 0.1,
+        "emb_dropout": 0.0
+    },
+    "action_encoder": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Embedder",
+        "input_dim": 10,
+        "emb_dim": 192
+    },
+    "projector": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    },
+    "pred_proj": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    }
+}

le-wm/reacher/weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb70b1fd5409f8f81875d62f5ee5a20dd220a3128a477de66b5760f475f0f469
+size 72290849

le-wm/runtime.json

@@ -0,0 +1,190 @@
+{
+  "python": "3.10.20",
+  "platform": "Linux x86_64",
+  "packages": {
+    "einops": "0.8.2",
+    "huggingface-hub": "0.36.2",
+    "typing-extensions": "4.16.0",
+    "tqdm": "4.70.0",
+    "requests": "2.34.2",
+    "certifi": "2026.7.22",
+    "urllib3": "2.7.0",
+    "idna": "3.18",
+    "charset-normalizer": "3.4.9",
+    "pyyaml": "6.0.3",
+    "packaging": "26.2",
+    "hf-xet": "1.5.2",
+    "fsspec": "2025.10.0",
+    "filelock": "3.32.0",
+    "hydra-core": "1.3.4",
+    "antlr4-python3-runtime": "4.9.3",
+    "omegaconf": "2.3.1",
+    "h5py": "3.16.0",
+    "numpy": "2.2.6",
+    "matplotlib": "3.10.9",
+    "python-dateutil": "2.9.0.post0",
+    "six": "1.17.0",
+    "pyparsing": "3.3.2",
+    "pillow": "11.3.0",
+    "kiwisolver": "1.5.0",
+    "fonttools": "4.63.0",
+    "cycler": "0.12.1",
+    "contourpy": "1.3.2",
+    "imageio-ffmpeg": "0.6.0",
+    "imageio": "2.37.4",
+    "scipy": "1.15.3",
+    "torchvision": "0.28.0",
+    "torch": "2.13.0",
+    "triton": "3.7.1",
+    "nvidia-nvshmem-cu13": "3.4.5",
+    "nvidia-nccl-cu13": "2.29.7",
+    "nvidia-cusparselt-cu13": "0.8.1",
+    "nvidia-cudnn-cu13": "9.20.0.48",
+    "nvidia-cublas": "13.1.1.3",
+    "nvidia-cuda-nvrtc": "13.0.88",
+    "cuda-bindings": "13.3.1",
+    "cuda-pathfinder": "1.6.0",
+    "cuda-toolkit": "13.0.3.0",
+    "nvidia-nvtx": "13.0.85",
+    "nvidia-nvjitlink": "13.3.33",
+    "nvidia-cusparse": "12.6.3.3",
+    "nvidia-cusolver": "12.0.4.66",
+    "nvidia-curand": "10.4.0.35",
+    "nvidia-cuda-cupti": "13.0.85",
+    "nvidia-cufile": "1.15.1.6",
+    "nvidia-cufft": "12.0.0.61",
+    "nvidia-cuda-runtime": "13.0.96",
+    "jinja2": "3.1.6",
+    "markupsafe": "3.0.3",
+    "networkx": "3.4.2",
+    "sympy": "1.14.0",
+    "mpmath": "1.3.0",
+    "setuptools": "83.0.0",
+    "stable-worldmodel": "0.1.1",
+    "wandb": "0.28.1",
+    "sentry-sdk": "2.66.1",
+    "pydantic": "2.13.4",
+    "typing-inspection": "0.4.2",
+    "pydantic-core": "2.46.4",
+    "annotated-types": "0.8.0",
+    "protobuf": "7.35.1",
+    "platformdirs": "4.11.0",
+    "click": "8.4.2",
+    "hydra-submitit-launcher": "1.2.0",
+    "submitit": "1.5.4",
+    "cloudpickle": "3.1.2",
+    "stable-pretraining": "0.1.7",
+    "opencv-python-headless": "5.0.0.93",
+    "pylance": "9.0.0",
+    "lance-namespace": "0.8.6",
+    "lance-namespace-urllib3-client": "0.8.6",
+    "pyarrow": "20.0.0",
+    "opt-einsum": "3.4.0",
+    "richuru": "0.1.1",
+    "rich": "15.0.0",
+    "pygments": "2.20.0",
+    "markdown-it-py": "4.2.0",
+    "mdurl": "0.1.2",
+    "loguru": "0.7.3",
+    "scikit-learn": "1.7.2",
+    "threadpoolctl": "3.6.0",
+    "joblib": "1.5.3",
+    "minari": "0.5.3",
+    "gymnasium": "1.3.0",
+    "farama-notifications": "0.0.6",
+    "typer": "0.27.0",
+    "annotated-doc": "0.0.5",
+    "shellingham": "1.5.4",
+    "datasets": "4.4.0",
+    "aiohttp": "3.14.3",
+    "yarl": "1.24.5",
+    "propcache": "0.5.2",
+    "multidict": "6.7.1",
+    "frozenlist": "1.8.0",
+    "attrs": "26.1.0",
+    "async-timeout": "5.0.1",
+    "aiosignal": "1.4.0",
+    "aiohappyeyeballs": "2.7.1",
+    "multiprocess": "0.70.18",
+    "dill": "0.4.0",
+    "xxhash": "3.8.1",
+    "httpx": "0.28.1",
+    "httpcore": "1.0.9",
+    "h11": "0.16.0",
+    "anyio": "4.14.2",
+    "exceptiongroup": "1.3.1",
+    "pandas": "2.3.3",
+    "tzdata": "2026.3",
+    "pytz": "2026.3.post1",
+    "transformers": "4.57.3",
+    "safetensors": "0.8.0",
+    "tokenizers": "0.22.2",
+    "regex": "2026.7.19",
+    "zstandard": "0.25.0",
+    "timm": "1.0.28",
+    "prettytable": "3.18.0",
+    "wcwidth": "0.8.2",
+    "requests-cache": "1.3.3",
+    "url-normalize": "3.0.0",
+    "cattrs": "26.1.0",
+    "tabulate": "0.10.0",
+    "lightning": "2.6.5",
+    "pytorch-lightning": "2.6.5",
+    "lightning-utilities": "0.15.3",
+    "torchmetrics": "1.9.0",
+    "ale-py": "0.12.0",
+    "craftax": "1.6.1",
+    "pygame": "2.6.1",
+    "flax": "0.10.7",
+    "treescope": "0.1.10",
+    "tensorstore": "0.1.78",
+    "ml-dtypes": "0.5.4",
+    "orbax-checkpoint": "0.11.39",
+    "uvloop": "0.22.1",
+    "psutil": "7.2.2",
+    "simplejson": "4.1.1",
+    "humanize": "4.16.0",
+    "aiofiles": "25.1.0",
+    "jax": "0.6.2",
+    "jaxlib": "0.6.2",
+    "msgpack": "1.2.1",
+    "etils": "1.13.0",
+    "zipp": "4.1.0",
+    "importlib-resources": "7.1.0",
+    "absl-py": "2.5.0",
+    "optax": "0.2.8",
+    "stable-baselines3": "2.9.0",
+    "gymnasium-robotics": "1.4.2",
+    "pettingzoo": "1.26.1",
+    "mujoco": "3.11.0",
+    "pyopengl": "3.1.10",
+    "glfw": "2.10.2",
+    "moviepy": "2.2.1",
+    "python-dotenv": "1.2.2",
+    "proglog": "0.1.12",
+    "decorator": "5.3.1",
+    "opencv-python": "5.0.0.93",
+    "array-api-compat": "1.15.0",
+    "pygame-ce": "2.5.7",
+    "swig": "4.4.1",
+    "box2d-py": "2.3.5",
+    "minigrid": "3.1.0",
+    "ogbench": "1.2.1",
+    "dm-control": "1.0.44",
+    "lxml": "6.1.1",
+    "labmaze": "1.0.6",
+    "dm-tree": "0.1.10",
+    "wrapt": "2.3.0",
+    "dm-env": "1.6",
+    "shapely": "2.1.2",
+    "pymunk": "7.3.0",
+    "cffi": "2.1.0",
+    "pycparser": "3.0",
+    "lancedb": "0.34.0",
+    "overrides": "7.7.0",
+    "deprecation": "2.1.0"
+  },
+  "metadata_conflicts": [
+    "datasets==4.4.0 requires pyarrow>=21.0.0; observed pyarrow==20.0.0"
+  ]
+}

le-wm/tworoom/config.json

@@ -0,0 +1,49 @@
+{
+    "_target_": "stable_worldmodel.wm.lewm.LeWM",
+    "encoder": {
+        "_target_": "stable_pretraining.backbone.utils.vit_hf",
+        "size": "tiny",
+        "patch_size": 14,
+        "image_size": 224,
+        "pretrained": false,
+        "use_mask_token": false
+    },
+    "predictor": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Predictor",
+        "num_frames": 3,
+        "input_dim": 192,
+        "hidden_dim": 192,
+        "output_dim": 192,
+        "depth": 6,
+        "heads": 16,
+        "mlp_dim": 2048,
+        "dim_head": 64,
+        "dropout": 0.1,
+        "emb_dropout": 0.0
+    },
+    "action_encoder": {
+        "_target_": "stable_worldmodel.wm.lewm.module.Embedder",
+        "input_dim": 10,
+        "emb_dim": 192
+    },
+    "projector": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    },
+    "pred_proj": {
+        "_target_": "stable_worldmodel.wm.lewm.module.MLP",
+        "input_dim": 192,
+        "output_dim": 192,
+        "hidden_dim": 2048,
+        "norm_fn": {
+            "_target_": "torch.nn.BatchNorm1d",
+            "_partial_": true
+        }
+    }
+}

le-wm/tworoom/weights.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:566f223624ea4bfb39dbfe6ae731198dd6ea73b7b8919fed6b1ecafca810f7dd
+size 72290849

licenses/dino_wm.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 gaoyuezhou
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

licenses/dreamerv3.txt

@@ -0,0 +1,19 @@
+Copyright (c) 2024 Danijar Hafner
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

licenses/le-wm.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Lucas Maes
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

licenses/planet.txt

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

licenses/tdmpc2.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) Nicklas Hansen (2023).
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

manifest.json

@@ -0,0 +1,481 @@
+{
+  "format_version": 1,
+  "repo_id": "mohendy/world-model-attack-checkpoints",
+  "files": [
+    {
+      "path": "tdmpc2/finger-spin/step-1000000.pt",
+      "model": "tdmpc2",
+      "bytes": 42677713,
+      "sha256": "bbc1c9b519784a661839bfe97707a5c5395ca03655fbc17f24fe13cf23857be8",
+      "source_sha256": "bbc1c9b519784a661839bfe97707a5c5395ca03655fbc17f24fe13cf23857be8",
+      "source_relative_path": "tdmpc2/tdmpc2/logs/finger-spin/1/rgb_seed1_2m/models/step-1000000.pt",
+      "sanitized": false
+    },
+    {
+      "path": "tdmpc2/cup-catch/step-1500000.pt",
+      "model": "tdmpc2",
+      "bytes": 42677713,
+      "sha256": "df67f9c68655c3443c8c5e36799091ee9debfebc4ca11b1a59abf370b4083e47",
+      "source_sha256": "df67f9c68655c3443c8c5e36799091ee9debfebc4ca11b1a59abf370b4083e47",
+      "source_relative_path": "tdmpc2/tdmpc2/logs/cup-catch/1/rgb_seed1_2m/models/step-1500000.pt",
+      "sanitized": false
+    },
+    {
+      "path": "tdmpc2/cheetah-run/step-1500000.pt",
+      "model": "tdmpc2",
+      "bytes": 42833361,
+      "sha256": "5654ea3c9c69a2640848b9d0efc043dc7d55896eeb9de62c7cb5c958f8652cd8",
+      "source_sha256": "5654ea3c9c69a2640848b9d0efc043dc7d55896eeb9de62c7cb5c958f8652cd8",
+      "source_relative_path": "tdmpc2/tdmpc2/logs/cheetah-run/1/rgb_seed1_2m/models/step-1500000.pt",
+      "sanitized": false
+    },
+    {
+      "path": "tdmpc2/cartpole-balance/step-250000.pt",
+      "model": "tdmpc2",
+      "bytes": 42638723,
+      "sha256": "74cf0a67dc4f2487c270381f32f1027a07a3299fa8fbe282cb7b7a632ce84b6e",
+      "source_sha256": "74cf0a67dc4f2487c270381f32f1027a07a3299fa8fbe282cb7b7a632ce84b6e",
+      "source_relative_path": "tdmpc2/tdmpc2/logs/cartpole-balance/1/rgb_seed1_2m/models/step-250000.pt",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/config.yaml",
+      "model": "planet",
+      "bytes": 12461,
+      "sha256": "da3a9fa154e89037fc27ac9b2f88db4d7ae821660a6428ce1cb9240d2dcae324",
+      "source_sha256": "1a91c863aea1d70dd07c625ef7e7736c774b18f4cb6b0a1ddf8347d643dce2d3",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/config.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "planet/reacher_easy/attack_export.json",
+      "model": "planet",
+      "bytes": 464,
+      "sha256": "512ce35d92afb78d89106cfa0279c4cd22110d0cf6b1bd65549c88501052a487",
+      "source_sha256": "512ce35d92afb78d89106cfa0279c4cd22110d0cf6b1bd65549c88501052a487",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/attack_export.json",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/attack_weights.npz",
+      "model": "planet",
+      "bytes": 5342758,
+      "sha256": "a1e9829fbd16c2d1b18336534c561006938a4b59a5f55b0de526f7cebfac5e38",
+      "source_sha256": "a1e9829fbd16c2d1b18336534c561006938a4b59a5f55b0de526f7cebfac5e38",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/attack_weights.npz",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/tf_reference.npz",
+      "model": "planet",
+      "bytes": 112084,
+      "sha256": "22d50e7b59c1c661bc1ffb889c4196bffe9fe6d28f4cf2b62d7700ff99e8d165",
+      "source_sha256": "22d50e7b59c1c661bc1ffb889c4196bffe9fe6d28f4cf2b62d7700ff99e8d165",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/tf_reference.npz",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/config.yaml",
+      "model": "planet",
+      "bytes": 12497,
+      "sha256": "6e645402217101ca5151be97f37448913d6b89787161422a019b80b9f547ab36",
+      "source_sha256": "4e9de7dfe7bf9d706caf6720a164e640ed68221acfbe099386dfed75c4a1ca30",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/config.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "planet/cartpole_swingup/attack_export.json",
+      "model": "planet",
+      "bytes": 464,
+      "sha256": "dc7e1134643a0bce63521befb2eaf708aa83d24fdb8e361d0b8280b1cbecd80e",
+      "source_sha256": "dc7e1134643a0bce63521befb2eaf708aa83d24fdb8e361d0b8280b1cbecd80e",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/attack_export.json",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/attack_weights.npz",
+      "model": "planet",
+      "bytes": 5341958,
+      "sha256": "f75ced2942b489b578c750ed99ce5daa9e14a8727ce911b469a7785ce6231580",
+      "source_sha256": "f75ced2942b489b578c750ed99ce5daa9e14a8727ce911b469a7785ce6231580",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/attack_weights.npz",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/tf_reference.npz",
+      "model": "planet",
+      "bytes": 111936,
+      "sha256": "62b9163ef512d1a397af51cd1320b5721f59ea8cb7283b8be1f320d6dd7035a0",
+      "source_sha256": "62b9163ef512d1a397af51cd1320b5721f59ea8cb7283b8be1f320d6dd7035a0",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/tf_reference.npz",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/config.yaml",
+      "model": "planet",
+      "bytes": 12422,
+      "sha256": "7da0ca0f3486fe410848afd46fd1c000c8ef8241dbe47ec5e5d98ef169f51fa0",
+      "source_sha256": "0e234a9f150d9e0e301e3e8d770cb787c56ec018406ac8b9432b4c9a31551e06",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/config.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "planet/cheetah_run/attack_export.json",
+      "model": "planet",
+      "bytes": 464,
+      "sha256": "0d7d9b3940ead701528d14eca0fa05ed6fca471cb009ded68448b8251cea6740",
+      "source_sha256": "0d7d9b3940ead701528d14eca0fa05ed6fca471cb009ded68448b8251cea6740",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/attack_export.json",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/attack_weights.npz",
+      "model": "planet",
+      "bytes": 5345958,
+      "sha256": "fdf26210c8304305b8b5903d730dbdfe8b56c6a358524dba3c35d15048f4c968",
+      "source_sha256": "fdf26210c8304305b8b5903d730dbdfe8b56c6a358524dba3c35d15048f4c968",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/attack_weights.npz",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/tf_reference.npz",
+      "model": "planet",
+      "bytes": 112676,
+      "sha256": "7bb4dfe7beea345abd585c5db704f40c668e1356c0b7644c4a5135d44ec7d271",
+      "source_sha256": "7bb4dfe7beea345abd585c5db704f40c668e1356c0b7644c4a5135d44ec7d271",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/tf_reference.npz",
+      "sanitized": false
+    },
+    {
+      "path": "dino_wm/pusht/hydra.yaml",
+      "model": "dino_wm",
+      "bytes": 1927,
+      "sha256": "33f40be9f62d4dae9553ea99011692b5b9c20aae9110595d1807d876b13f585a",
+      "source_sha256": "e4d7f21b85c778ab8b83c0bb83bc7625fdfb193a6a0619fef25f4a2afd7f8049",
+      "source_relative_path": "dino_wm/checkpoints/outputs/pusht/hydra.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "dino_wm/wall_single/hydra.yaml",
+      "model": "dino_wm",
+      "bytes": 1920,
+      "sha256": "f0b13250e49408c4890190eda837641e030ff6cda073cbc1238e4d11068c4370",
+      "source_sha256": "4ed3109219ed6820ce51fd19e4c97199552b3443ed58c502fddd6441784967e1",
+      "source_relative_path": "dino_wm/checkpoints/outputs/wall_single/hydra.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "dino_wm/wall_single/model_latest.pth",
+      "model": "dino_wm",
+      "bytes": 368656057,
+      "sha256": "8441971becdae934fe08de5b163398390a32f6fe1fb0a8df113290e2468e142b",
+      "source_sha256": "8441971becdae934fe08de5b163398390a32f6fe1fb0a8df113290e2468e142b",
+      "source_relative_path": "dino_wm/checkpoints/outputs/wall_single/checkpoints/model_latest.pth",
+      "sanitized": false
+    },
+    {
+      "path": "dino_wm/pusht/model_latest.pth",
+      "model": "dino_wm",
+      "bytes": 377929744,
+      "sha256": "e909f0cec958fc0b49a79f2e85730ae6b5f83dee61a224705f91883eb3bb74c5",
+      "source_sha256": "e909f0cec958fc0b49a79f2e85730ae6b5f83dee61a224705f91883eb3bb74c5",
+      "source_relative_path": "dino_wm/checkpoints/outputs/pusht/checkpoints/model_latest.pth",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/cube/config.json",
+      "model": "le-wm",
+      "bytes": 1313,
+      "sha256": "4d446944fe28922cc2c5763f43d4ef9132a457bd89e9a0ce5dbceac183994999",
+      "source_sha256": "4d446944fe28922cc2c5763f43d4ef9132a457bd89e9a0ce5dbceac183994999",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/cube/lewm/config.json",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/cube/weights.pt",
+      "model": "le-wm",
+      "bytes": 72291425,
+      "sha256": "2839a907362f403f9136383016e91774373a295d958ae75121791f22a9fddf89",
+      "source_sha256": "2839a907362f403f9136383016e91774373a295d958ae75121791f22a9fddf89",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/cube/lewm/weights.pt",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/pusht/config.json",
+      "model": "le-wm",
+      "bytes": 1313,
+      "sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/pusht/lewm/config.json",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/pusht/weights.pt",
+      "model": "le-wm",
+      "bytes": 72290721,
+      "sha256": "48938400ae3464c9680731287f583a9cb516f55a8ec64ea13a91be47fb15b607",
+      "source_sha256": "48938400ae3464c9680731287f583a9cb516f55a8ec64ea13a91be47fb15b607",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/pusht/lewm/weights.pt",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/reacher/config.json",
+      "model": "le-wm",
+      "bytes": 1313,
+      "sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/reacher/lewm/config.json",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/reacher/weights.pt",
+      "model": "le-wm",
+      "bytes": 72290849,
+      "sha256": "eb70b1fd5409f8f81875d62f5ee5a20dd220a3128a477de66b5760f475f0f469",
+      "source_sha256": "eb70b1fd5409f8f81875d62f5ee5a20dd220a3128a477de66b5760f475f0f469",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/reacher/lewm/weights.pt",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/tworoom/config.json",
+      "model": "le-wm",
+      "bytes": 1313,
+      "sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_sha256": "2564086e961e7b5c7c04dffc451091115b389a590645ff19653c64fd0bc16e09",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/tworoom/lewm/config.json",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/tworoom/weights.pt",
+      "model": "le-wm",
+      "bytes": 72290849,
+      "sha256": "566f223624ea4bfb39dbfe6ae731198dd6ea73b7b8919fed6b1ecafca810f7dd",
+      "source_sha256": "566f223624ea4bfb39dbfe6ae731198dd6ea73b7b8919fed6b1ecafca810f7dd",
+      "source_relative_path": "le-wm/.stable-wm/checkpoints/tworoom/lewm/weights.pt",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/cheetah_run/config.yaml",
+      "model": "dreamerv3",
+      "bytes": 4403,
+      "sha256": "b28b6a00ac52e29df5d31accca25595942f1ea8e6986e02ad1c27d94daba0ad5",
+      "source_sha256": "f73845c27ef1eff92eff2acf5fd196c9a5bd25aaf43aed4a2a8a16ef78457f74",
+      "source_relative_path": "logdir/dreamer/cheetah_run/slurm-23188/config.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "dreamerv3/cheetah_run/ckpt/20260901T093748F019112/agent.pkl",
+      "model": "dreamerv3",
+      "bytes": 2042806149,
+      "sha256": "1d94c5824d8c77ba8191ccd5207409986a08bbf7e04ae5d55a9d468c7d158af3",
+      "source_sha256": "1d94c5824d8c77ba8191ccd5207409986a08bbf7e04ae5d55a9d468c7d158af3",
+      "source_relative_path": "logdir/dreamer/cheetah_run/slurm-23188/ckpt/20260901T093748F019112/agent.pkl",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/cheetah_run/ckpt/20260901T093748F019112/step.pkl",
+      "model": "dreamerv3",
+      "bytes": 17,
+      "sha256": "69901686be1d13680e0c0ddbfd0a0d2f2a70bb0ae46a2e636bf78bf5bc45cce2",
+      "source_sha256": "69901686be1d13680e0c0ddbfd0a0d2f2a70bb0ae46a2e636bf78bf5bc45cce2",
+      "source_relative_path": "logdir/dreamer/cheetah_run/slurm-23188/ckpt/20260901T093748F019112/step.pkl",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/cheetah_run/ckpt/20260901T093748F019112/done",
+      "model": "dreamerv3",
+      "bytes": 0,
+      "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "source_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "source_relative_path": "logdir/dreamer/cheetah_run/slurm-23188/ckpt/20260901T093748F019112/done",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/reacher_easy/config.yaml",
+      "model": "dreamerv3",
+      "bytes": 4409,
+      "sha256": "5a264f9cfc744b358798b053447f8df1d936a91592b04e96850fe3dda5ca32e4",
+      "source_sha256": "431174ded7bf75a544ba3717917cff2487c3b997c6a4c5c22ddbb9260e88e897",
+      "source_relative_path": "logdir/dreamer/reacher_easy/20260830T125224/config.yaml",
+      "sanitized": true
+    },
+    {
+      "path": "dreamerv3/reacher_easy/ckpt/20260830T195435F325934/agent.pkl",
+      "model": "dreamerv3",
+      "bytes": 2042658588,
+      "sha256": "a1fb60436f02fcfa6ba1fbd0b5b5078142cb6f6c34b13c42e4a5967ef047b394",
+      "source_sha256": "a1fb60436f02fcfa6ba1fbd0b5b5078142cb6f6c34b13c42e4a5967ef047b394",
+      "source_relative_path": "logdir/dreamer/reacher_easy/20260830T125224/ckpt/20260830T195435F325934/agent.pkl",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/reacher_easy/ckpt/20260830T195435F325934/step.pkl",
+      "model": "dreamerv3",
+      "bytes": 17,
+      "sha256": "7e5c5685c2d39badee13ea957052bbe8a3f71e577e2369e55151ad875842cb2f",
+      "source_sha256": "7e5c5685c2d39badee13ea957052bbe8a3f71e577e2369e55151ad875842cb2f",
+      "source_relative_path": "logdir/dreamer/reacher_easy/20260830T125224/ckpt/20260830T195435F325934/step.pkl",
+      "sanitized": false
+    },
+    {
+      "path": "dreamerv3/reacher_easy/ckpt/20260830T195435F325934/done",
+      "model": "dreamerv3",
+      "bytes": 0,
+      "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "source_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "source_relative_path": "logdir/dreamer/reacher_easy/20260830T125224/ckpt/20260830T195435F325934/done",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/tensorflow/model.ckpt-4504500.data-00000-of-00001",
+      "model": "planet",
+      "bytes": 61543552,
+      "sha256": "22b7d2959cab4e3b38fd55a579bb718d59516053f6cb92530c7cabf785cd5ce5",
+      "source_sha256": "22b7d2959cab4e3b38fd55a579bb718d59516053f6cb92530c7cabf785cd5ce5",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/model.ckpt-4504500.data-00000-of-00001",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/tensorflow/model.ckpt-4504500.index",
+      "model": "planet",
+      "bytes": 5797,
+      "sha256": "b34dd58607ccf858446024be305a73757eb9b65a83ccecde57768974ad13a509",
+      "source_sha256": "b34dd58607ccf858446024be305a73757eb9b65a83ccecde57768974ad13a509",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/model.ckpt-4504500.index",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/tensorflow/model.ckpt-4504500.meta",
+      "model": "planet",
+      "bytes": 7264371,
+      "sha256": "cdd30a7fd3c3bff6ca5e0bcce9c87bfd5056a02462730ac3d05d91e9cce4b1df",
+      "source_sha256": "cdd30a7fd3c3bff6ca5e0bcce9c87bfd5056a02462730ac3d05d91e9cce4b1df",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/model.ckpt-4504500.meta",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cartpole_swingup/tensorflow/checkpoint",
+      "model": "planet",
+      "bytes": 93,
+      "sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cartpole_swingup/checkpoint",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/tensorflow/model.ckpt-4504500.data-00000-of-00001",
+      "model": "planet",
+      "bytes": 61545952,
+      "sha256": "4d1cbb7fe86a05756a2433fef62f8b024b8b846c642c7e2354b40a37e0a9f4c0",
+      "source_sha256": "4d1cbb7fe86a05756a2433fef62f8b024b8b846c642c7e2354b40a37e0a9f4c0",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/model.ckpt-4504500.data-00000-of-00001",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/tensorflow/model.ckpt-4504500.index",
+      "model": "planet",
+      "bytes": 5789,
+      "sha256": "5844b59c53652eadad60eca39dbd8e1cd0273510ca78a56b221a7fe904a41df8",
+      "source_sha256": "5844b59c53652eadad60eca39dbd8e1cd0273510ca78a56b221a7fe904a41df8",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/model.ckpt-4504500.index",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/tensorflow/model.ckpt-4504500.meta",
+      "model": "planet",
+      "bytes": 7112337,
+      "sha256": "ca8e8aef416be0b50d17dd164c53f352f1a001e065e2cd3d477c4666a796418e",
+      "source_sha256": "ca8e8aef416be0b50d17dd164c53f352f1a001e065e2cd3d477c4666a796418e",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/model.ckpt-4504500.meta",
+      "sanitized": false
+    },
+    {
+      "path": "planet/reacher_easy/tensorflow/checkpoint",
+      "model": "planet",
+      "bytes": 93,
+      "sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/reacher_easy/checkpoint",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/tensorflow/model.ckpt-4504500.data-00000-of-00001",
+      "model": "planet",
+      "bytes": 61555552,
+      "sha256": "fe62149d086a11b89dcb259d8a6af18b41583747c5b34baecc96ae46d47a0431",
+      "source_sha256": "fe62149d086a11b89dcb259d8a6af18b41583747c5b34baecc96ae46d47a0431",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/model.ckpt-4504500.data-00000-of-00001",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/tensorflow/model.ckpt-4504500.index",
+      "model": "planet",
+      "bytes": 5787,
+      "sha256": "f3c20abc873aa178695036d8d913acc41f1c61d1753f9e87a6fa8fccb8df80f2",
+      "source_sha256": "f3c20abc873aa178695036d8d913acc41f1c61d1753f9e87a6fa8fccb8df80f2",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/model.ckpt-4504500.index",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/tensorflow/model.ckpt-4504500.meta",
+      "model": "planet",
+      "bytes": 7068294,
+      "sha256": "bd92bc2d5226976ffb7cca03252f2ee29753cb40025187cdf29961beb6cdc7c2",
+      "source_sha256": "bd92bc2d5226976ffb7cca03252f2ee29753cb40025187cdf29961beb6cdc7c2",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/model.ckpt-4504500.meta",
+      "sanitized": false
+    },
+    {
+      "path": "planet/cheetah_run/tensorflow/checkpoint",
+      "model": "planet",
+      "bytes": 93,
+      "sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_sha256": "c4b1b9ef54b5185a72880247eb4ab7c71e54cc1b3e15a6ef82cbaf9510ac35d6",
+      "source_relative_path": "planet/checkpoints/planet-dmc-rgb-seed0/cheetah_run/checkpoint",
+      "sanitized": false
+    },
+    {
+      "path": "le-wm/defenses/cube/residual_hgd/hgd_eps16.best.pt",
+      "model": "le-wm",
+      "bytes": 5917691,
+      "sha256": "68f58d28e6b39954331143c2f66e736f402882f15fce35ace8afc8d464c17f07",
+      "source_sha256": "6532554c98c2bfa8982942d06562cf10ed81f96579667855cdf0bac328232480",
+      "source_relative_path": "le-wm/.stable-wm/defenses/cube/eps16_encrel_lam0p1/checkpoints/hgd_eps16.best.pt",
+      "sanitized": true
+    },
+    {
+      "path": "le-wm/defenses/cube/residual_hgd/hgd_eps16.pt",
+      "model": "le-wm",
+      "bytes": 5916747,
+      "sha256": "bc3af72c1ae2ed328c8907f87e5709b719e00c8dd61e7c32ccb9dc722951772a",
+      "source_sha256": "373c0ec72c661465ea7cccadbad0b472443c7dd48ebfd2af4f78052e6457afce",
+      "source_relative_path": "le-wm/.stable-wm/defenses/cube/eps16_encrel_lam0p1/checkpoints/hgd_eps16.pt",
+      "sanitized": true
+    },
+    {
+      "path": "le-wm/defenses/cube/paper_hgd/hgd_paper.best.pt",
+      "model": "le-wm",
+      "bytes": 132576538,
+      "sha256": "04ed4a9b4d9edea42f4e4ec7b489ae79378890a3c2d505077a97d69d19fd1ec9",
+      "source_sha256": "2354e376b01155e7253677e0e16273e31b32416c5d5db3f182423e3bc99ab2cf",
+      "source_relative_path": "le-wm/.stable-wm/defenses/cube/hgd_paper_train1000_val100/checkpoints/hgd_paper.best.pt",
+      "sanitized": true
+    },
+    {
+      "path": "le-wm/defenses/cube/paper_hgd/hgd_paper.pt",
+      "model": "le-wm",
+      "bytes": 132574859,
+      "sha256": "9ebae97be8b61cefde85958dadea28d5ae34c959b3cd183f4932f6bfe1debdf2",
+      "source_sha256": "ca01417994b99b2ef0f9938e4a28b4d6e2a819a52865e2d8246aa5cddca6e087",
+      "source_relative_path": "le-wm/.stable-wm/defenses/cube/hgd_paper_train1000_val100/checkpoints/hgd_paper.pt",
+      "sanitized": true
+    }
+  ],
+  "notes": [
+    "Binary research model checkpoints are byte-identical to local sources, except HGD metadata sanitization with exact state_dict tensor equality verified.",
+    "Configuration output paths are relative; set them for the destination machine.",
+    "Native TensorFlow PlaNet checkpoints, PyTorch inference exports, and TensorFlow parity fixtures are included.",
+    "Datasets, videos, episode perturbations, logs, and credentials are excluded.",
+    "The source_relative_path field maps this repository back to the research code artifact layout."
+  ]
+}

planet/cartpole_swingup/attack_export.json

@@ -0,0 +1,10 @@
+{
+  "source_revision": "b26a7fe027c018f62b3178c8e126371fbdee33e5",
+  "tensorflow": "2.15.0",
+  "inference_tensor_count": 32,
+  "source_sha256": {
+    "model.ckpt-4504500.data-00000-of-00001": "22b7d2959cab4e3b38fd55a579bb718d59516053f6cb92530c7cabf785cd5ce5",
+    "model.ckpt-4504500.index": "b34dd58607ccf858446024be305a73757eb9b65a83ccecde57768974ad13a509",
+    "model.ckpt-4504500.meta": "cdd30a7fd3c3bff6ca5e0bcce9c87bfd5056a02462730ac3d05d91e9cce4b1df"
+  }
+}

planet/cartpole_swingup/attack_weights.npz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f75ced2942b489b578c750ed99ce5daa9e14a8727ce911b469a7785ce6231580
+size 5341958

planet/cartpole_swingup/config.yaml

@@ -0,0 +1,388 @@
+!!python/object/new:planet.tools.attr_dict.AttrDict
+dictitems:
+  activation: &id001 !!python/name:tensorflow.python.ops.gen_nn_ops.relu ''
+  batch_shape: &id004
+  - 50
+  - 50
+  bound_action: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.bound_action.bound_action ''
+    _kwargs:
+      strategy: clip
+  cell: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple
+    - 30
+    - 200
+    - 200
+    - true
+    - false
+    - 0.1
+    - *id001
+    - 1
+    _fn: !!python/name:planet.models.rssm.RSSM ''
+    _kwargs: {}
+  checkpoint_to_load: null
+  data_loader: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.numpy_episodes.recent_loader ''
+    _kwargs:
+      every: 1000
+  data_reader: !!python/name:planet.tools.numpy_episodes.episode_reader ''
+  debug: false
+  decoder: &id002 !!python/name:planet.networks.conv_ha.decoder ''
+  encoder: !!python/name:planet.networks.conv_ha.encoder ''
+  free_nats: 3.0
+  gradient_heads:
+  - image
+  - reward
+  head_network: &id003 !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.networks.basic.feed_forward ''
+    _kwargs:
+      activation: *id001
+      num_layers: 2
+      units: 200
+  heads: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      image: *id002
+      reward: !!python/object:planet.tools.bind.bind
+        _args: !!python/tuple []
+        _fn: *id003
+        _kwargs:
+          stop_gradient: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  isolate_envs: thread
+  logdir: ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001
+  loss_scales: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      divergence: 1.0
+      global_divergence: 0.0
+      image: 1.0
+      overshooting: 0.0
+      reward: 10.0
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  max_steps: 5000000
+  num_chunks: 1
+  num_layers: 2
+  num_units: 200
+  open_loop_context: 5
+  optimizers: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      main: !!python/object:planet.tools.bind.bind
+        _args: !!python/tuple []
+        _fn: !!python/name:planet.tools.custom_optimizer.CustomOptimizer ''
+        _kwargs:
+          clipping: 1000.0
+          learning_rate: 0.001
+          optimizer_cls: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: !!python/name:tensorflow.python.training.adam.AdamOptimizer ''
+            _kwargs:
+              epsilon: 0.0001
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  os_stop_posterior_grad: true
+  overshooting_distance: 0
+  postprocess_fn: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.preprocess.postprocess ''
+    _kwargs:
+      bits: 5
+  preprocess_fn: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.preprocess.preprocess ''
+    _kwargs:
+      bits: 5
+  print_metrics_every: 5000
+  random_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      test-cartpole_swingup: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          num_episodes: 5
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/test_episodes
+          task: &id005 !!python/object/new:planet.scripts.tasks.Task
+          - cartpole_swingup
+          - !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple
+            - 8
+            - 125
+            - cartpole
+            - swingup
+            - &id006 !!python/object/new:planet.tools.attr_dict.AttrDict
+              dictitems:
+                activation: relu
+                batch_shape: *id004
+                collect_every: 5000
+                divergence_scale: 1.0
+                free_nats: 3.0
+                global_div_scale: 0.0
+                image_bits: 5
+                isolate_envs: thread
+                logdir: ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001
+                main_gradient_clipping: 1000.0
+                main_learning_rate: 0.001
+                max_steps: 5000000
+                model: rssm
+                model_size: 200
+                num_layers: 2
+                num_seed_episodes: 5
+                num_units: 200
+                overshooting_scale: 0.0
+                planner_amount: 1000
+                planner_horizon: 12
+                planner_iterations: 10
+                planner_topk: 100
+                seed: 0
+                state_size: 30
+                tasks:
+                - cartpole_swingup
+                test_collects:
+                - batch_size: 10
+                test_steps: 50
+                train_action_noise: 0.3
+                train_steps: 50000
+              state:
+                _defaults: {}
+                _touched: !!set
+                  __getstate__: null
+                  action_noise_min: null
+                  action_noise_ramp: null
+                  action_repeat: null
+                  activation: null
+                  batch_shape: null
+                  bound_action: null
+                  camera_id: null
+                  checkpoint_every: null
+                  collect_every: null
+                  collect_objective: null
+                  divergence_scale: null
+                  free_nats: null
+                  future_rnn: null
+                  global_div_scale: null
+                  gradient_heads: null
+                  image_bits: null
+                  image_loss_scale: null
+                  isolate_envs: null
+                  loader: null
+                  loader_every: null
+                  logdir: null
+                  main_gradient_clipping: null
+                  main_learning_rate: null
+                  max_steps: null
+                  mean_only: null
+                  min_stddev: null
+                  model: null
+                  model_layers: null
+                  model_size: null
+                  network: null
+                  num_chunks: null
+                  num_layers: null
+                  num_seed_episodes: null
+                  num_units: null
+                  os_stop_posterior_grad: null
+                  overshooting_distance: null
+                  overshooting_scale: null
+                  planner: null
+                  planner_amount: null
+                  planner_horizon: null
+                  planner_iterations: null
+                  planner_topk: null
+                  reward_loss_scale: null
+                  seed: null
+                  state_diagnostics: null
+                  state_size: null
+                  tasks: null
+                  test_collects: null
+                  test_steps: null
+                  train_action_noise: null
+                  train_collects: null
+                  train_steps: null
+                _unlocked: false
+            _fn: !!python/name:planet.scripts.tasks._dm_control_env ''
+            _kwargs: {}
+          - 125
+          - - reward
+            - position
+            - velocity
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+            num_episodes: null
+            save_episode_dir: null
+            task: null
+          _unlocked: false
+      train-cartpole_swingup: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          num_episodes: 5
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/train_episodes
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+            num_episodes: null
+            save_episode_dir: null
+            task: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  savers:
+  - !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      exclude: !!python/tuple
+      - .*_temporary.*
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  tasks:
+  - *id005
+  test_checkpoint_every: 500
+  test_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      test_main_cartpole_swingup: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          exploration: !!python/object/new:planet.tools.attr_dict.AttrDict
+            dictitems:
+              scale: 0.0
+              schedule: !!python/object:planet.tools.bind.bind
+                _args: !!python/tuple []
+                _fn: &id007 !!python/name:planet.tools.schedule.linear ''
+                _kwargs:
+                  min: 0.0
+                  ramp: 0
+            state:
+              _defaults: {}
+              _touched: !!set
+                __getstate__: null
+              _unlocked: false
+          num_agents: 10
+          objective: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: &id008 !!python/name:planet.scripts.objectives.reward ''
+            _kwargs:
+              params: *id006
+          planner: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: &id009 !!python/name:planet.control.planning.cross_entropy_method ''
+            _kwargs:
+              amount: 1000
+              horizon: 12
+              iterations: 10
+              topk: 100
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/test_episodes
+          steps_after: 5000
+          steps_every: 5000
+          steps_until: -1
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  test_dir: 
+    ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/test_episodes
+  test_steps: 50
+  train_checkpoint_every: null
+  train_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      train_main_cartpole_swingup: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          exploration: !!python/object/new:planet.tools.attr_dict.AttrDict
+            dictitems:
+              scale: 0.3
+              schedule: !!python/object:planet.tools.bind.bind
+                _args: !!python/tuple []
+                _fn: *id007
+                _kwargs:
+                  min: 0.0
+                  ramp: 0
+            state:
+              _defaults: {}
+              _touched: !!set
+                __getstate__: null
+              _unlocked: false
+          num_agents: 1
+          objective: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: *id008
+            _kwargs:
+              params: *id006
+          planner: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: *id009
+            _kwargs:
+              amount: 1000
+              horizon: 12
+              iterations: 10
+              topk: 100
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/train_episodes
+          steps_after: 5000
+          steps_every: 5000
+          steps_until: -1
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  train_dir: 
+    ./planet/results/planet-single-seed/cartpole_swingup/seed_00/00001/train_episodes
+  train_log_every: 50000
+  train_steps: 50000
+state:
+  _defaults: {}
+  _touched: !!set
+    __getstate__: null
+    activation: null
+    batch_shape: null
+    data_loader: null
+    data_reader: null
+    decoder: null
+    gradient_heads: null
+    head_network: null
+    heads: null
+    logdir: null
+    loss_scales: null
+    num_chunks: null
+    num_layers: null
+    num_units: null
+    optimizers: null
+    preprocess_fn: null
+    random_collects: null
+    tasks: null
+    test_dir: null
+    test_steps: null
+    train_dir: null
+    train_steps: null
+  _unlocked: false

planet/cartpole_swingup/tensorflow/checkpoint

@@ -0,0 +1,2 @@
+model_checkpoint_path: "model.ckpt-4504500"
+all_model_checkpoint_paths: "model.ckpt-4504500"

planet/cartpole_swingup/tensorflow/model.ckpt-4504500.data-00000-of-00001

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22b7d2959cab4e3b38fd55a579bb718d59516053f6cb92530c7cabf785cd5ce5
+size 61543552

planet/cartpole_swingup/tensorflow/model.ckpt-4504500.meta

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdd30a7fd3c3bff6ca5e0bcce9c87bfd5056a02462730ac3d05d91e9cce4b1df
+size 7264371

planet/cartpole_swingup/tf_reference.npz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62b9163ef512d1a397af51cd1320b5721f59ea8cb7283b8be1f320d6dd7035a0
+size 111936

planet/cheetah_run/attack_export.json

@@ -0,0 +1,10 @@
+{
+  "source_revision": "b26a7fe027c018f62b3178c8e126371fbdee33e5",
+  "tensorflow": "2.15.0",
+  "inference_tensor_count": 32,
+  "source_sha256": {
+    "model.ckpt-4504500.data-00000-of-00001": "fe62149d086a11b89dcb259d8a6af18b41583747c5b34baecc96ae46d47a0431",
+    "model.ckpt-4504500.index": "f3c20abc873aa178695036d8d913acc41f1c61d1753f9e87a6fa8fccb8df80f2",
+    "model.ckpt-4504500.meta": "bd92bc2d5226976ffb7cca03252f2ee29753cb40025187cdf29961beb6cdc7c2"
+  }
+}

planet/cheetah_run/attack_weights.npz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdf26210c8304305b8b5903d730dbdfe8b56c6a358524dba3c35d15048f4c968
+size 5345958

planet/cheetah_run/config.yaml

@@ -0,0 +1,388 @@
+!!python/object/new:planet.tools.attr_dict.AttrDict
+dictitems:
+  activation: &id001 !!python/name:tensorflow.python.ops.gen_nn_ops.relu ''
+  batch_shape: &id004
+  - 50
+  - 50
+  bound_action: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.bound_action.bound_action ''
+    _kwargs:
+      strategy: clip
+  cell: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple
+    - 30
+    - 200
+    - 200
+    - true
+    - false
+    - 0.1
+    - *id001
+    - 1
+    _fn: !!python/name:planet.models.rssm.RSSM ''
+    _kwargs: {}
+  checkpoint_to_load: null
+  data_loader: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.numpy_episodes.recent_loader ''
+    _kwargs:
+      every: 1000
+  data_reader: !!python/name:planet.tools.numpy_episodes.episode_reader ''
+  debug: false
+  decoder: &id002 !!python/name:planet.networks.conv_ha.decoder ''
+  encoder: !!python/name:planet.networks.conv_ha.encoder ''
+  free_nats: 3.0
+  gradient_heads:
+  - image
+  - reward
+  head_network: &id003 !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.networks.basic.feed_forward ''
+    _kwargs:
+      activation: *id001
+      num_layers: 2
+      units: 200
+  heads: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      image: *id002
+      reward: !!python/object:planet.tools.bind.bind
+        _args: !!python/tuple []
+        _fn: *id003
+        _kwargs:
+          stop_gradient: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  isolate_envs: thread
+  logdir: ./planet/results/planet-single-seed/cheetah_run/seed_00/00001
+  loss_scales: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      divergence: 1.0
+      global_divergence: 0.0
+      image: 1.0
+      overshooting: 0.0
+      reward: 10.0
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  max_steps: 5000000
+  num_chunks: 1
+  num_layers: 2
+  num_units: 200
+  open_loop_context: 5
+  optimizers: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      main: !!python/object:planet.tools.bind.bind
+        _args: !!python/tuple []
+        _fn: !!python/name:planet.tools.custom_optimizer.CustomOptimizer ''
+        _kwargs:
+          clipping: 1000.0
+          learning_rate: 0.001
+          optimizer_cls: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: !!python/name:tensorflow.python.training.adam.AdamOptimizer ''
+            _kwargs:
+              epsilon: 0.0001
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  os_stop_posterior_grad: true
+  overshooting_distance: 0
+  postprocess_fn: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.preprocess.postprocess ''
+    _kwargs:
+      bits: 5
+  preprocess_fn: !!python/object:planet.tools.bind.bind
+    _args: !!python/tuple []
+    _fn: !!python/name:planet.tools.preprocess.preprocess ''
+    _kwargs:
+      bits: 5
+  print_metrics_every: 5000
+  random_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      test-cheetah_run: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          num_episodes: 5
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/test_episodes
+          task: &id005 !!python/object/new:planet.scripts.tasks.Task
+          - cheetah_run
+          - !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple
+            - 4
+            - 250
+            - cheetah
+            - run
+            - &id006 !!python/object/new:planet.tools.attr_dict.AttrDict
+              dictitems:
+                activation: relu
+                batch_shape: *id004
+                collect_every: 5000
+                divergence_scale: 1.0
+                free_nats: 3.0
+                global_div_scale: 0.0
+                image_bits: 5
+                isolate_envs: thread
+                logdir: ./planet/results/planet-single-seed/cheetah_run/seed_00/00001
+                main_gradient_clipping: 1000.0
+                main_learning_rate: 0.001
+                max_steps: 5000000
+                model: rssm
+                model_size: 200
+                num_layers: 2
+                num_seed_episodes: 5
+                num_units: 200
+                overshooting_scale: 0.0
+                planner_amount: 1000
+                planner_horizon: 12
+                planner_iterations: 10
+                planner_topk: 100
+                seed: 0
+                state_size: 30
+                tasks:
+                - cheetah_run
+                test_collects:
+                - batch_size: 10
+                test_steps: 50
+                train_action_noise: 0.3
+                train_steps: 50000
+              state:
+                _defaults: {}
+                _touched: !!set
+                  __getstate__: null
+                  action_noise_min: null
+                  action_noise_ramp: null
+                  action_repeat: null
+                  activation: null
+                  batch_shape: null
+                  bound_action: null
+                  camera_id: null
+                  checkpoint_every: null
+                  collect_every: null
+                  collect_objective: null
+                  divergence_scale: null
+                  free_nats: null
+                  future_rnn: null
+                  global_div_scale: null
+                  gradient_heads: null
+                  image_bits: null
+                  image_loss_scale: null
+                  isolate_envs: null
+                  loader: null
+                  loader_every: null
+                  logdir: null
+                  main_gradient_clipping: null
+                  main_learning_rate: null
+                  max_steps: null
+                  mean_only: null
+                  min_stddev: null
+                  model: null
+                  model_layers: null
+                  model_size: null
+                  network: null
+                  num_chunks: null
+                  num_layers: null
+                  num_seed_episodes: null
+                  num_units: null
+                  os_stop_posterior_grad: null
+                  overshooting_distance: null
+                  overshooting_scale: null
+                  planner: null
+                  planner_amount: null
+                  planner_horizon: null
+                  planner_iterations: null
+                  planner_topk: null
+                  reward_loss_scale: null
+                  seed: null
+                  state_diagnostics: null
+                  state_size: null
+                  tasks: null
+                  test_collects: null
+                  test_steps: null
+                  train_action_noise: null
+                  train_collects: null
+                  train_steps: null
+                _unlocked: false
+            _fn: !!python/name:planet.scripts.tasks._dm_control_env ''
+            _kwargs: {}
+          - 250
+          - - reward
+            - position
+            - velocity
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+            num_episodes: null
+            save_episode_dir: null
+            task: null
+          _unlocked: false
+      train-cheetah_run: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          num_episodes: 5
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/train_episodes
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+            num_episodes: null
+            save_episode_dir: null
+            task: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  savers:
+  - !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      exclude: !!python/tuple
+      - .*_temporary.*
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  tasks:
+  - *id005
+  test_checkpoint_every: 500
+  test_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      test_main_cheetah_run: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          exploration: !!python/object/new:planet.tools.attr_dict.AttrDict
+            dictitems:
+              scale: 0.0
+              schedule: !!python/object:planet.tools.bind.bind
+                _args: !!python/tuple []
+                _fn: &id007 !!python/name:planet.tools.schedule.linear ''
+                _kwargs:
+                  min: 0.0
+                  ramp: 0
+            state:
+              _defaults: {}
+              _touched: !!set
+                __getstate__: null
+              _unlocked: false
+          num_agents: 10
+          objective: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: &id008 !!python/name:planet.scripts.objectives.reward ''
+            _kwargs:
+              params: *id006
+          planner: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: &id009 !!python/name:planet.control.planning.cross_entropy_method ''
+            _kwargs:
+              amount: 1000
+              horizon: 12
+              iterations: 10
+              topk: 100
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/test_episodes
+          steps_after: 5000
+          steps_every: 5000
+          steps_until: -1
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  test_dir: 
+    ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/test_episodes
+  test_steps: 50
+  train_checkpoint_every: null
+  train_collects: !!python/object/new:planet.tools.attr_dict.AttrDict
+    dictitems:
+      train_main_cheetah_run: !!python/object/new:planet.tools.attr_dict.AttrDict
+        dictitems:
+          exploration: !!python/object/new:planet.tools.attr_dict.AttrDict
+            dictitems:
+              scale: 0.3
+              schedule: !!python/object:planet.tools.bind.bind
+                _args: !!python/tuple []
+                _fn: *id007
+                _kwargs:
+                  min: 0.0
+                  ramp: 0
+            state:
+              _defaults: {}
+              _touched: !!set
+                __getstate__: null
+              _unlocked: false
+          num_agents: 1
+          objective: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: *id008
+            _kwargs:
+              params: *id006
+          planner: !!python/object:planet.tools.bind.bind
+            _args: !!python/tuple []
+            _fn: *id009
+            _kwargs:
+              amount: 1000
+              horizon: 12
+              iterations: 10
+              topk: 100
+          save_episode_dir: 
+            ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/train_episodes
+          steps_after: 5000
+          steps_every: 5000
+          steps_until: -1
+          task: *id005
+        state:
+          _defaults: {}
+          _touched: !!set
+            __getstate__: null
+          _unlocked: false
+    state:
+      _defaults: {}
+      _touched: !!set
+        __getstate__: null
+      _unlocked: false
+  train_dir: 
+    ./planet/results/planet-single-seed/cheetah_run/seed_00/00001/train_episodes
+  train_log_every: 50000
+  train_steps: 50000
+state:
+  _defaults: {}
+  _touched: !!set
+    __getstate__: null
+    activation: null
+    batch_shape: null
+    data_loader: null
+    data_reader: null
+    decoder: null
+    gradient_heads: null
+    head_network: null
+    heads: null
+    logdir: null
+    loss_scales: null
+    num_chunks: null
+    num_layers: null
+    num_units: null
+    optimizers: null
+    preprocess_fn: null
+    random_collects: null
+    tasks: null
+    test_dir: null
+    test_steps: null
+    train_dir: null
+    train_steps: null
+  _unlocked: false

planet/cheetah_run/tensorflow/checkpoint

@@ -0,0 +1,2 @@
+model_checkpoint_path: "model.ckpt-4504500"
+all_model_checkpoint_paths: "model.ckpt-4504500"

planet/cheetah_run/tensorflow/model.ckpt-4504500.data-00000-of-00001

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe62149d086a11b89dcb259d8a6af18b41583747c5b34baecc96ae46d47a0431
+size 61555552