Initial release of ResWAM

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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

README.md

@@ -0,0 +1,71 @@
+---
+license: mit
+library_name: pytorch
+tags:
+  - robotics
+  - vision-language-action
+  - world-action-model
+  - flow-matching
+  - robodojo
+  - bimanual-manipulation
+---
+
+# ResWAM-RoboDojo (Official RoboDojo 42-Task Checkpoint)
+
+Official **ResWAM** multi-task bimanual policy checkpoint (`OPTML-Group/ResWAM-RoboDojo`) trained on the official 35-task **RoboDojo** simulation training set without external robot pretraining, achieving **9.40%** average Success Rate (SR) and **13.41** stage-level Progress Score across all 42 tasks (54 evaluation configurations, 2,100 episodes) in Table 3 of the paper.
+
+- **Paper**: *ResWAM: Learning What Changes, Not What Stays, for World-Action Models*
+- **Authors**: Chongyu Fan, Gaowen Liu, Ramana Rao Kompella, Jason Corso, Sijia Liu
+- **Affiliations**: Michigan State University, Cisco, University of Michigan, IBM Research
+- **Project Page**: [https://chongyu-fan.netlify.app/posts/reswam/](https://chongyu-fan.netlify.app/posts/reswam/)
+- **Code**: [https://github.com/OPTML-Group/ResWAM](https://github.com/OPTML-Group/ResWAM)
+- **HF Collection**: [https://huggingface.co/collections/OPTML-Group/reswam-6af8906dc9b6f21539b30934](https://huggingface.co/collections/OPTML-Group/reswam-6af8906dc9b6f21539b30934)
+
+## Abstract
+
+World-action models (WAMs) augment vision-language-action policies (VLAs) with future prediction so that the policy can anticipate how its actions will affect the physical world. Existing WAMs supervise complete future observations. However, observations typically change little over an action chunk, so the future loss is dominated by copying static background content rather than learning task-relevant robot and object motion, limiting policy learning. Moreover, existing WAMs rely on large video generation models or separate world models, incurring heavy training and inference compute that hinders closed-loop deployment. To improve policy performance, we propose *visual residual learning*, supervising only the feature change from the current frame to the future frame in a frozen DINOv3 space. Compared with complete future features, this residual learning target yields a higher feature rank with richer and more diverse representations of scene changes, and spatially concentrates on the robot arm and manipulated objects rather than the background. To further improve computational efficiency, we present **ResWAM**, a residual world-action model with 2.3B parameters in total, including a lightweight 0.5B dual-branch DiT that jointly trains an action branch and a residual branch via conditional flow matching. Without robot pretraining (*i.e.*, pretraining on external robot action trajectories beyond the benchmark demonstrations), **ResWAM** achieves top average success rates on LIBERO-Plus (81.1%), RoboTwin 2.0 (93.8%), and RoboDojo (9.40% success, 13.41 score) at the lowest training (2.7–4.0 TFLOPs) and inference compute (1.0–1.4 TFLOPs), and leads real-robot Open, Put, and Stack (80.0% average, +10.0% over π<sub>0.5</sub>) at 92.7 ms latency.
+
+## Checkpoint Summary
+
+| Field | Value |
+| :- | :- |
+| **Repository** | [`OPTML-Group/ResWAM-RoboDojo`](https://huggingface.co/OPTML-Group/ResWAM-RoboDojo) |
+| **Checkpoint file** | `model.pt` |
+| **Parameters** | `2.3B` (`RynnBrain-1.1-2B` 16-layer text backbone + `0.5B` dual-branch DiT + frozen `DINOv3-B/16`) |
+| **Action horizon ($H$)** | `32` (14-D ARX X5 bimanual joint position + grippers; executes 10 steps before replanning) |
+| **Compute & Latency** | `3.9 TFLOPs` train / `1.4 TFLOPs` infer / `99.7 ms` per action chunk (10 Euler steps, RTX A6000) |
+
+## RoboDojo Results (42 Tasks, 54 Configurations, 2,100 Episodes)
+
+| Generalization (SR / Score) | Precision (SR / Score) | Long-Horizon (SR / Score) | Memory (SR / Score) | Open (SR / Score) | **Avg. (SR / Score)** |
+| :-: | :-: | :-: | :-: | :-: | :-: |
+| 6.50% / 9.12 | **14.75%** / **21.31** | **15.75%** / **25.52** | **9.00%** / **9.97** | 1.00% / 1.14 | **9.40%** / **13.41** |
+
+## Files in this Repository
+
+- `model.pt`: Full model weights and embedded codec metadata.
+- `codec_statistics.json`: 14-D bimanual action and state z-score normalization statistics (`ZScoreCodec.from_file("codec_statistics.json")`).
+- `resolved_config.json`: Complete resolved model, data, and evaluation configuration.
+
+## Download and Evaluation Usage
+
+```bash
+hf download OPTML-Group/ResWAM-RoboDojo --local-dir ./checkpoints/ResWAM-RoboDojo
+
+python scripts/eval/eval_robodojo.py run \
+    --checkpoint ./checkpoints/ResWAM-RoboDojo/model.pt \
+    --robodojo-root ./third_party/RoboDojo \
+    --output-dir ./outputs/robodojo/eval_p32e10 \
+    --gpus 0,1,2,3
+```
+
+## Citation
+
+```bibtex
+@article{fan2026reswam,
+  title={ResWAM: Learning What Changes, Not What Stays, for World-Action Models},
+  author={Fan, Chongyu and Liu, Gaowen and Kompella, Ramana Rao and Corso, Jason and Liu, Sijia},
+  journal={arXiv preprint},
+  year={2026}
+}
+```

codec_statistics.json

@@ -0,0 +1,108 @@
+{
+  "action_mean": [
+    -0.20078273417564957,
+    0.9277771156992791,
+    0.6894869569948502,
+    -0.34068942096479327,
+    0.06591800516017124,
+    0.0040108836233663,
+    0.7710157369596965,
+    0.17332457336160728,
+    0.824045140543182,
+    0.6208197694009031,
+    -0.3610974871809101,
+    -0.06278889445312434,
+    0.003760568328959391,
+    0.7856019478654002
+  ],
+  "action_std": [
+    0.34326785074077926,
+    0.8879649160338935,
+    0.7303173442632056,
+    0.6492112921281379,
+    0.29917761429299594,
+    0.5848514585306708,
+    0.3433463967463059,
+    0.3201112242117163,
+    0.873740194767336,
+    0.7136994454719052,
+    0.6030868903736086,
+    0.28390183508311156,
+    0.5373599841791166,
+    0.34085015307207406
+  ],
+  "decode_clip": false,
+  "encode_clamp_sigma": 12.0,
+  "gripper_indices": [
+    6,
+    13
+  ],
+  "gripper_normalization": "2x-1_clip_[-1,1]_on_encode",
+  "gripper_semantics": "larger_is_more_open",
+  "joint_indices": [
+    0,
+    1,
+    2,
+    3,
+    4,
+    5,
+    7,
+    8,
+    9,
+    10,
+    11,
+    12
+  ],
+  "joint_normalization": "zscore_mean_std_all_clean_frames",
+  "joint_order": [
+    "left_waist",
+    "left_shoulder",
+    "left_elbow",
+    "left_forearm_roll",
+    "left_wrist_angle",
+    "left_wrist_rotate",
+    "left_gripper",
+    "right_waist",
+    "right_shoulder",
+    "right_elbow",
+    "right_forearm_roll",
+    "right_wrist_angle",
+    "right_wrist_rotate",
+    "right_gripper"
+  ],
+  "padding": "end-of-episode windows repeat the last frame; supervised",
+  "representation": "14-D absolute joint/qpos",
+  "state_mean": [
+    -0.2007454369032377,
+    0.9275954219370752,
+    0.689380351801755,
+    -0.3406735232699966,
+    0.06590374551565717,
+    0.004045320244545627,
+    0.7710789097449021,
+    0.17333555777244486,
+    0.82371858187106,
+    0.6205467270371468,
+    -0.36089229757941016,
+    -0.06272941580737187,
+    0.003887326806600799,
+    0.7856890844310401
+  ],
+  "state_std": [
+    0.34326509850088577,
+    0.8879730441085364,
+    0.7303366967941258,
+    0.6491286947162288,
+    0.29916365964187547,
+    0.5848012863963614,
+    0.3433176166011588,
+    0.32006485002946655,
+    0.8736847808883317,
+    0.7135999425614398,
+    0.6029080227405151,
+    0.28383354337037464,
+    0.5370253190766491,
+    0.3407991216853018
+  ],
+  "version": "reswam-robodojo-absolute-qpos-zscore-joints-affine-grippers-14d-v1"
+}

model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65491e6950499b7552a44b6516a7b0f4576d79e6e0e7dd4af551aa2df0a4657b
+size 9339832400

resolved_config.json

@@ -0,0 +1,157 @@
+{
+  "action": {
+    "dimension": 14,
+    "execute_horizon": 10,
+    "horizon": 32,
+    "padding": "end_of_episode_clamp_to_last_frame_supervised",
+    "representation": "14-D absolute joint/qpos"
+  },
+  "benchmark": "robodojo",
+  "data": {
+    "camera_keys": {
+      "head": "observation.images.cam_high",
+      "left_wrist": "observation.images.cam_left_wrist",
+      "right_wrist": "observation.images.cam_right_wrist"
+    },
+    "camera_order": [
+      "head",
+      "left_wrist",
+      "right_wrist"
+    ],
+    "codec": {
+      "decode_clip": false,
+      "encode_clamp_sigma": 12.0,
+      "grippers": "2x-1",
+      "joints": "zscore_mean_std_all_clean_frames",
+      "statistics": "codec_statistics.json",
+      "version": "reswam-robodojo-absolute-qpos-zscore-joints-affine-grippers-14d-v1"
+    },
+    "metadata_frequency_hz": 25,
+    "preprocess_to_256": "direct_bicubic_antialias_no_crop",
+    "vlm_camera_order": [
+      "head",
+      "left_wrist",
+      "right_wrist"
+    ]
+  },
+  "dataset": {
+    "action_dimension": 14,
+    "action_representation": "14-D absolute joint/qpos",
+    "camera_order": [
+      "head",
+      "left_wrist",
+      "right_wrist"
+    ],
+    "cleaning_variant": "all-frames",
+    "control_hz": 25,
+    "episodes": 3500,
+    "frames": 1856102,
+    "gripper_semantics": "larger_is_more_open",
+    "hf_repo": "RoboDojo-Benchmark/RoboDojo",
+    "joint_order": [
+      "left_joint_0",
+      "left_joint_1",
+      "left_joint_2",
+      "left_joint_3",
+      "left_joint_4",
+      "left_joint_5",
+      "left_joint_6",
+      "right_joint_0",
+      "right_joint_1",
+      "right_joint_2",
+      "right_joint_3",
+      "right_joint_4",
+      "right_joint_5",
+      "right_joint_6"
+    ],
+    "sampler": "uniform_concat_every_clean_frame_end_clamped_supervised",
+    "state_dimension": 14,
+    "state_representation": "14-D current qpos",
+    "tasks": 35
+  },
+  "evaluation": {
+    "action_type": "joint",
+    "benchmark": "RoboDojo sim (Isaac Sim 5.1 / Isaac Lab 2.3), official XPolicyLab websocket client",
+    "control_hz": 25,
+    "env_cfg": "arx_x5",
+    "execute_horizon": 10,
+    "num_configs": 54,
+    "num_tasks": 42,
+    "prediction_horizon": 32,
+    "seeds": [
+      0
+    ],
+    "total_episodes": 2100
+  },
+  "flow": {
+    "action_noise": "independent_gaussian",
+    "action_scheduler_shift": 1,
+    "action_tau": "independent_uniform_shifted",
+    "inference_rounds": 10,
+    "synchronous_commit": true,
+    "world_noise": "independent_gaussian",
+    "world_scheduler_shift": 5,
+    "world_tau": "independent_uniform_shifted"
+  },
+  "model": {
+    "dino": {
+      "architecture": "DINOv3-B/16",
+      "combine": "arithmetic_mean",
+      "frozen": true,
+      "model_id": "facebook/dinov3-vitb16-pretrain-lvd1689m",
+      "remove_prefix_tokens": true,
+      "selected_blocks": [
+        8,
+        10
+      ],
+      "spatial_pooling": "2x2",
+      "target_token_ln": true,
+      "tokens": 64
+    },
+    "expert": {
+      "depth": 12,
+      "expert_parameter_count": 493608462,
+      "ffn_dim": 4096,
+      "heads": 16,
+      "hidden_dim": 1024,
+      "state_dropout_prob": 0.0,
+      "use_state_null_tokens": false
+    },
+    "rynnbrain": {
+      "backbone_parameter_count": 1755487104,
+      "repo": "Alibaba-DAMO-Academy/RynnBrain1.1-2B",
+      "text_hidden_size": 2048,
+      "text_layers": 16,
+      "vision_layers": 24
+    }
+  },
+  "optimizer": {
+    "betas": [
+      0.9,
+      0.95
+    ],
+    "clip_grad_norm": 1.0,
+    "expert_min_lr": 1e-05,
+    "expert_peak_lr": 0.0001,
+    "name": "AdamW",
+    "precision": "bfloat16",
+    "vlm_min_lr": 1e-06,
+    "vlm_peak_lr": 1e-05,
+    "weight_decay": 1e-08
+  },
+  "schema_version": "reswam-robodojo-config-v1",
+  "training": {
+    "fsdp": "FULL_SHARD",
+    "global_batch": 256,
+    "gradient_accumulation": 2,
+    "microbatch": 32,
+    "seed": 20260907,
+    "world_size": 4
+  },
+  "world": {
+    "camera": "head",
+    "loss_weight": 0.1,
+    "objective": "endpoint_only_extra_LN_delta",
+    "target_offset": 32
+  }
+}